EPA-460/3-77-015
August 1977
CHARACTERIZATION
AND RESEARCH INVESTIGATION
OF METHANOL AND METHYL
FUELS
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Mobile Source Air Pollution Control
Emission Control Technology Division
Ann Arbor, Michigan 48105
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35) , Research Triangle Park, North Carolina
27711; or, for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency
by the University of Santa Clara, Department of Mechanical Engineering,
Santa Clara, California 95053, in fulfillment of Grant No. R803548-01.
The contents of this report are reproduced herein as received from the
University of Santa Clara, Department of Mechanical Engineering.
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-460/3-77-015
11
-------
EPA-460/3-77-015
CHARACTERIZATION
AND RESEARCH INVESTIGATION
OF METHANOL AND METHYL FUELS
bv
R.K. Pefley, L.H. Browning, W.E. Likos,
M.C. McCormack, and B. Pullman
Department of Mechanical Engineering
University of Santa Clara
Santa Clara, California 95053
EPA Grant No. R803548-01
EPA Project Officer: R.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 1977
-------
CHARACTERIZATION AND RESEARCH INVESTIGATION
OF METHANOL AND METHYL FUELS
FINAL REPORT
EPA Grant No. R803548-01
Project Officer - R. J. Garbe
R. K. Pefley
L. H. Browning
M. L. Hornberger
W. E. Likos
M. C. McCormack
B. Pullman
-------
INDEX
Page
List of Figures ii.
List of Tables iv.
Abstract v.
I. Preface 1.
II. Recapitulation of Work Objectives 2.
III. Summary of Work and Projected Planning 6.
IV. Discussion of Program Results 23.
IV.1 Steady State Engine Performance and Exhaust
Emissions Characterization - Methanol vs
Indolene 23.
IV.2 Alternate Fuel Induction Systems 54.
IV.3 Methanol versus Indolene - A Comparison Based
on Simulation of the Federal Emission Test
Procedure and the Federal Highway Fuel Economy
Test Procedure 67.
IV.4 T hermochemical Engine Process Modeling 77.
IV.5 Cold Starting and Lean Burning 91.
IV.6 Engine Wear and Crank Case Blow by 99.
V. Conclusions 105.
VI. References 107.
-------
11.
LIST OF FIGURES
Page
III.l: NOX Emissions - Methanol Vs. Jet Fuel in Gas Turbines 19
III.2: CO Emissions - Methanol Vs. Jet Fuel in Gas Turbines 20
IV.1: Test Engine - Dynamometer Configuration 25
IV.2: Comparative WOT Indicated Power - Indolene (OEM) Vs.
Methanol (OEM and Alternate Systems) 28
IV.3: Comparative WOT Brake Power - Indolene (OEM) Vs. Methanol
(OEM and Alternate Systems) 29
IV.4: WOT Comparative Thermal Efficiency for OEM and Alternate
Fuel-Air Induction Systems 30
IV.5: 14" Hg. Manifold Vacuum Comparative Thermal Efficiency for
OEM and Alternate Fuel-Air Induction Systems 33
IV.6: Exhaust Emissions Sampling System 35
IV.7: C02, CO, and 02 Vs. Cylinder Equivalence Ratio - Indolene
Vs. Methanol 37
IV.8: Comparative WOT Steady State NOX Emissions - Indolene (OEM)
Vs. Methanol (OEM and Alternate Systems) 39
IV.9: Comparative 1/3 Brake Load Steady State NOX Emissions -
Indolene (OEM) Vs. Methanol (OEM and Alternate Systems) 40
IV.10: Comparative WOT Steady State UBF Emissions - Indolene
(OEM) Vs. Methanol (OEM and Alternate Ssytems) 42
IV.11: Comparative 1/3 Brake Load Steady State UBF Emissions
Indolene (OEM) Vs. Methanol (OEM and Alternate Systems) 43
IV.12: Comparative Exhaust Aldehydes 45
IV.13: Cylinder to Cylinder Variations in Equivalence Ratio -
_ — Methanol Vs.--Lndo1ene — - — - — 47-
IV.14: Effects of Maldistribution on Spark Timing, Power, Air
Flow Rate, and Fuel Flow Rate 49
IV.15: Effects of Maldistribution on Power - Calculated Vs. Actual 50
IV.16: Effects of Maldistribution on NOX - Experimental Data 53
IV.17: Calculated Maldistribution Effect on N0y 53
-------
111.
Page
IV.18: Equivalence Ratio Vs. Percent of Wide Open Throttle -
WHB Induction System 56
IV.19: Air Flow Rate Vs. Percent of Wide Open Throttle - WHB
Induction System 53
IV.20: Electronic Fuel Injection System 59
IV.21: Effects of Injection Timing on Power, Thermal Efficiency
and Emissions 61
IV.22a: Conceptualized Diagram of the Dresserator Induction System 64
IV.22b: Manifold Adapter Bracket 65
IV.23: Comparative Thermal Efficiency Across the Speed Range at
Wide Open Throttle - Indolene Vs. Methanol 66
IV.24: Fuel Economy Vs. Equivalence Ratio for Alternate Fuel-Air
Induction Systems on Methanol With the Indolene and
Methanol OEM Results - Urban Driving 69
IV.25: Fuel Economy Vs. Equivalence Ratio for Alternate Fuel-Air
Induction Systems on Methanol With the Indolene and
Methanol OEM Results - Highway Driving 72
IV.26: Exhaust Emissions Vs. Equivalence Ratio for OEM and Alternate
Fuel-Air Induction Systems for the Hot 1972 FTP
Simulation 74
IV.27: Comparison of AC Coupled Data and Integrated Data 79
IV.28: Match of Computer Predicted Results to Actual Pressure Traces 80
IV.29: Comparison of Acutal Vs. Predicted Performance Data 83
IV.30: Comparisons Between Experimental and Predicted Dry Exhaust
Emissions 84
IV.31: Computer Predicted Performance for 2000 RPM and 4000 RPM 86
IV.32: Computer Predicted Dry Emissions for 2000 RPM and 4000 RPM 87
IV.33: Computer Predicted MBT Compression Ratio Effect 88
IV.34: Computer Predicted Spark Retard Effects 89
IV.35: Prototype Methanol Cold Start System 93
IV.36: Effect of Fuel Type on Engine Wear 103
-------
TV.
LIST OF TABLES
Page
III.I Energy Based Fuel Economy as a Percentage Improvement
Over OEM Performance Using Indolene from Computer
Simulation of FTP (Hot 1972 Procedure) 10
III.2 Vehicle Emissions Comparison from Computer Simulation of FTP
(Hot 1972 Procedure) 10
IV.1 Base Line Test Matrix 26
IV.2 Exhaust Emission Measurement Technique 36
IV.3 Simulated Cases 68
IV.4 Simulated FTP Results 70
IV.5 Simulated HFETP Results 71
IV.6 Performance and Emissions Comparison at Match
Point (WOT and 2000 rpm) 82
IV.7 History of Engine Events 102
-------
V.
ABSTRACT
The work reported herein deals with several aspects of using pure methanol as
an alternate fuel.
A stock (OEM) Pinto engine mounted on a dynamometer hasbeen used to compare
methanol with Indolene* in terms of power, efficiency, and emissions for a
variety of speeds and loads.
Although the engine was designed for use with gasoline, it was found that
methanol was generally superior in power, thermal efficiency and reduced
emissions with the exception of aldehydes.
Study of engine wear showed no serious consequences from the use of methanol.
This is consistent with the evidence from our two road vehicles which have
now been operating for more than 5 and 6 years respectively on pure methanol.
Measured maldistribution of the air-fuel mixtures among the cylinders of the
test engine and variation of the mixture with speed and load reveal that
these variations must be reduced for improved mileage and emissions from
either gasoline or methanol. The problem is caused by the venturi carburetor
and intake manifold.
Evaluation of alternates to the venturi carburetor and intake manifold.
Three different fuel metering systems were tested for a variety of speeds and
loads using the dynamometer mounted engine. They were all found to provide
superior steady state performance on methanol when compared with the OEM car-
buretor system with enlarged fuel jets for methanol.
One prototype (WHB)** fuel system solves the maldistribution problem among
the cylinders and allows a 20-30% reduction in the idle speed. Forthcoming
modifications are expected to tighten control of the air-fuel ratio as
*An unleaded gasoline of closely specified composition.
**The initials of the inventor, William H. Beekhuis.
-------
VI .
functions of speed and load to a satisfactory degree at which time this
system will be tested on a vehicle.
Mileage and emissions from a Pinto vehicle equipped with the various fuel
metering systems have been computer predicted for the Federal emissions test
procedure using laboratory engine measurements.
Again the alternate fuel preparation systems showed significantly superior
performance on methanol in comparison with OEM-Indolene performance. The
evidence indicates that lean-burning, methanol-fueled vehicles may meet
statutory limits of emissions if tight control is maintained on the air-fuel
ratio and spark advance. A Pinto vehicle has been acquired to provide actual
evidence for comparisons with some of the computer predicted results.
A computer has been used to simulate the test engine's thermokinetic com-
bustion events.
The computer model predicts power, fuel economy and emissions with air-fuel
ratio, compression ratio, spark advance and speed as parameters. After
matching the computer's prediction for a specific set of conditions, to the
engine's performance, it was used to predict trends in power, fuel economy
and NOX emissions due to changes in air-fuel ratio, speed, and compression
ratio. The trends were very similar to actual engine trends where they
could be compared. The evidence clearly suggests use of higher compression
rat-ios with_me-thano-l While-the trends -are satisfaetor-y-,-the model does
not as yet satisfactorily deal with "squish" type combustion chamber effects
nor quench zone effects. It will be modified to improve these modeling
features.
Gas turbine converted to operate on methanol.
A small (60 hp) gas turbine has been converted to run on methanol. The con-
version was easily accomplished but atomization of the fuel was found to be
important in obtaining a reduction in CO and NOX for methanol in comparison
with jet engine fuel. The reduction in these pollutants was found to
correlate with other experimental evidence and computer modeling studies
upon using combustor inlet temperature as a basis for comparison.
-------
VII .
Environmental factors of marine and aquatic methanol spills and photo-
chemical smog are under study.
Preliminary experimentation relative to marine spills indicates that methanol
is naturally present in that environment. It appears at this early stage of
investigation that damage to the ecosystem from a major coastal spill may be
localized and of short duration.
A photochemical smog chamber is being equipped with instrumentation to per-
form a comparative photochemical smog study between methanol and gasoline
automotive exhaust. Results from this study will appear in subsequent
reports.
-------
I. PREFACE
This report is a continuation in a series of reports^»2>3 evaluating
methanol as an alternate liquid fuel with primary emphasis on automobile
use. The work was started in 1968 and has progressed continuously since
then. Over that time period the work has been partially funded by govern-
ment agencies (EPA, ERDA, NASA and NAPCA), partially by the City of
Santa Clara and the University. It has also been supported in part by
donations from the Ford Motor Company, General Motors Corporation,
Chrysler Corporation and Dresser Industries.
Although this report is being written primarily to fulfill contractual
obligations with EPA and ERDA, it is also intended to comprehensively
report on all aspects of our methanol program whether they have external
sponsorship or not. The reasons for this are twofold. First, it serves
as a total record of our activities. Second, it accentuates our belief
that engineering developments associated with alternate energy planning
must pay heed to a broad range of issues such as safety and environmental
hazards while at the same time addressing specific issues such as engine
performance and emissions.
This report relates the past year's efforts to the objectives established
at the beginning of the year, to our prior work, and to our plans for
follow-on programs.
We wish to acknowledge the contributions of our professional associates,
Dr. M. A. Sweeney, Dr. P. D'Eliscu and our student associates: J. Nebolon,
K. Overby, D. Rourk and J. Villemarie.
-------
II. RECAPITULATION OF WORK OBJECTIVES
Is methanol a superior fuel to gasoline in terms of performance, emissions,
economics, safety, and the environment if one or the other is to be made
from coal and bio-mass energy resources in the coming years? This is the
overall issue to which our following specific program objectives are related
and which serves as the general basis for the subsequent discussion and
analysis of our work.
I I.I Engine Performance and Emissions Characterization—Gasoline Versus
Methanol: The objective here is to first establish a baseline of per-
formance and emissions from a stock Pinto 2300 cc. engine using Indolene
(a reference grade gasoline). Then, using the same test matrix, evaluate
changes in performance and emissions when operating on methanol with
low cost modifications to the stock air-fuel preparation system.
II.2 Alternative Fuel Induction Systems: Previous work has shown the
maldistribution of the air-fuel mixture among the cylinders in a
multicylinder engine to be a serious problem. Since some modifica-
tions to the fuel preparation system are mandatory in changing a
gasoline fueled engine to methanol, the objective here is to explore
alternates to the combination of venturi-carburetor and intake-manifold
which appear attractive in alleviating the maldistribution problem
while improving performance and emissions from methanol. The first
phase of this exploration uses the same engine test set up and test
matrix used for establishing baseline engine performance and
emissions. The alternate fuel preparation systems selected for
evaluation are:
a. Electronically controlled fuel injection system.
b. Dresserator multi-cylinder Shockwave carburetor.
C. WHB individual cylinder Shockwave carburetion system.
II.3 Simulation of Urban and Highway Driving in a Pinto: What is a
meaningful way of efficiently comparing the extensive evidence
accumulated from engine dynamometer tests of the various fuel
-------
preparation systems? This objective question has led us to explore
the use of experimental performance and emissions data in conjunction
with a computer simulation of the Federal driving cycles for a Pinto
vehicle. The composite evidence of efficiency and emissions from
each fuel preparation system could then be easily contrasted, however,
causes for the contrasts would require more detailed analysis of the
data.
II.4 Thermo-chemical Engine Process Modeling: Computer simulation of the
thermo-chemical events in an open cycle heat engine can be a valuable
aid in predicting changes in performance and emissions which will
result from use of an alternate fuel such as methanol. Our objective
is to create such a computer model and tune it by use of specific
engine parameters (those of the 2300 cc. Pinto) and performance test
points. Then, use it to forecast changes in performance and emissions
due to spark timing, compression ratio, equivalence ratio , speed, and
load. Successful modeling, as attested to by comparison with experi-
mental evidence can greatly reduce future test program time and cost
in searching for the best engine modifications in accommodating
methanol as a fuel.
II.5 Cold Starting and Lean Burning: Cold starting and lean burning are
partially related issues for cold starting problems are often caused
by the air fuel-vapor mixture being outside the flammability range.
Methanol in pure form is not as suitable as gasoline for cold starting
because of this. The objective here is to explore ways of improving
the cold start ability of methanol. Improved fuel nebulization to
prevent fuel droplets from separating from the air stream, catalytic
partial dissociation into CO+ 2\\2 to obtain a gaseous fuel with broad
flamability limits, and intake manifold warming combustors are pre-
ferred avenues of exploration as they require no fuel modifications.
Special starting fuels and fuel blending agents are considered as
alternate possibilities.
-------
Lean burning of methanol under automobile cruise operation has shown
promise as a means of reducing NOX emissions and increasing fuel economy.
The generation of dissociated methanol (CO + 2^) by waste heat from
the exhaust and use of this fuel to enhance this possibility has been
of interest since the program's inception. The objective here is to
continue investigation of methods for satisfactory extension of lean
burning as it relates to both cruise operation and cold starting.
11.6 Engine Wear Rate and Crankcase Blow-by: There may be changes in
engine life expectancy due to shifts in friction and wear character-
istics resulting from the use of alternate fuels such as methanol. The
objective here is to track test engine wear rates by inspection for
metals accumulation in the lubricating oil during the time the engines
are being used for emissions and performance evaluations. Any evi-
dence of concern would be cause for more focused investigations.
Analysis of the constituents in crankcase blow-by gases can also
enhance understanding of wear rates. Our emissions test facility is
being used to explore this evidence during the engine characterization
studies.
11.7 Road Vehicles Performance: We have three vehicles operating on the
road with methanol fuel. Two use pure methanol and one uses methanol-
gasoline blends. The objective is to obtain real world experience with
methanol fueled vehicles.
II.8 Supplementary Objectives: As our confidence in the suitability of
methanol as an alternate fuel has grown, some important additional
activities have sprouted which address the following questions:
a. What are the differences in photochemical reactivity between
gasoline and methanol exhaust emissions?
b. What are the comparative biological hazards of methanol and
petroleum marine spills?
-------
c. Can small power gas turbines be easily modified to run on
methanol and what effects does this fuel have on performance
and emissions?
-------
III. SUMMARY OF WORK AND PROJECTED PLANNING
Highlights of evidence resulting from our investigations are pre-
sented in this section of the report. It is also intended to pro-
vide a synoptic interpretation for guidance in follow-on program
planning.
III.1 Engine Performance and Emissions Characterization--Gasoline*
versus Methanol: The experimental phase of the program has focused
on the steady state mapping of performance and exhaust emissions
of a dynamometer mounted, 2300cc, Ford Pinto engine. A signifi-
cant portion of this work was detailed in a previous report-^ and
is augmented by the detailed discussion in Section IV.1. of this
report.
The Indolene* and methanol baseline comparison with OEM** equipment
indicated that gains in thermal efficiency and reduced exhaust
emissions (with the exception of aldehydes) are obtained while
operating on methanol at the same engine speed, load, and
equivalence ratio ($)***. The following summary quantifies the
important engine performance and exhaust emission comparisons.
A. Performance
a) The power from methanol ranges from 5% to 11% higher
than that obtained from gasoline.
b) The indicated thermal efficiency from methanol is on
the average 10% higher than from gasoline for the
observed ranges of speed, load and ($).
B. Exhaust Emissions
a) NOX emissions from methanol average a factor of 2 lower
than gasoline NOX emissions.
* Indolene was used as the reference gasoline in the majority of the work.
** OEM is used to identify original equipment of the manufacturer.
***Equivalence ratio (*) is defined as stoichiometric v actual air-fuel ratio.
Fuel system modifications are required to obtain this comparison.
-------
7
b) The CO emissions in the lean region are about the same for
methanol and gasoline, but lower for methanol in the rich
region.
c) The hydrocarbon emissions, herein referred to as unburned
fuel (UBF), show small but discernable advantage favoring
methanol.
d) The aldehyde emissions are higher for methanol than for
gasoline by a factor ranging from 1.2 to 10 depending on
the equivalence ratio.
All of the observed comparative performance and emissions character-
istics were subject to the inherent cylinder to cylinder variations
in * of this particular 4-cylinder engine. In general, this mal-
distribution as we will call it, was worse when operating on methanol
with variations in (*) approaching ± 40% in some of the worst cases.
A maldistribution index (MI) was defined which allows estimation of
these effects on the power output and thermal efficiency. Based on
the observed maldistribution for both Indolene and methanol it
appears that additional relative gains for methanol ranging from 2 to
5% in thermal efficiency (and power) are possible if the maldistribu-
tion is eliminated in the range 0.81*1 1-0.
The maldistribution of * creates widely scattered engine emissions
evidence when the composites of all four cylinders are examined. This
is particularly true of NOX. As previously reported^, the composite
NOX vs composite * almost has the appearance of having been produced
by a random function generator.
Fortunately, the test program has also included individual exhaust
cylinder measurements. From these measurements, the maldistribution
effects on exhaust emissions of the NOx. UBF and CO have been resolved.
111.2 Alternative Fuel-Air Induction Systems for Methanol: In addition to
the baseline comparisons of Indolene and methanol in the OEM and
slightly modified OEM equipment, three alternative fuel-air induction
systems for methanol have been under investigation. They have been
described previously^ and identified as the WHB shock wave system,
the Dresserator Inductor (Dresser Industries), and the electronic
fuel injection (EFI) system. All three systems have been installed
-------
and tested on the engine used in the baseline comparison. Their
collective data indicates improvements in thermal efficiency, power,
and exhaust emissions relative to both methanol and Indolene base-
lines. In addition, one system (WHB) has completely eliminated the
maldistribution phenomena.
The WHB Shock Wave Carburetion System provides for individual cylinder
metering of the fuel for each intake stroke by using the strength of
the rarefaction wave generated by the intake process and attenuated by
throttle setting. It clearly solves the maldistribution of fuel-air
mixture among the cylinders. Typical variations are t 2-3%. It also
improves on full throttle OEM torque at higher engine speeds, because
it provides less intake flow resistance. At lower speeds due to this
lower intake flow resistance and excessive intake valve duration,
some reversed flow of fuel and air occurs causing a reduction in
torque as full throttle is approached. This system needs fuel orifice
placement modifications to provide a more nearly constant air-fuel
ratio over the speed and load range. Its performance in terms of
simulated CVS testing is summarized in Table III.l.
The Dresserator Carburetor was found to provide very good thermal
efficiency, particularly at part load conditions. This is attributed
to improved air fuel mixing associated with the shock wave which is
generated by the mixing nozzle. The carburetor was mounted on the
intake manifold only after carefully testing for the best position
from a maldistribution viewpoint. The maldistribution was still
comparable with the OEM maldistribution and is obviously due to the
intake manifold configuration. This carburetor also displayed serious
fuel standoff problems at high-torque, low-speed conditions again con-
firming the evidence that the camshaft provides late intake valve
closing for the lower engine speed range. This system needs improved
air-fuel distribution to provide fully satisfactory performance. Its
present performance which is incorporated in the CVS simulation is
also shown in Table III.l.
-------
Electronic Fuel Injection (EFI) was selected to provide ease of
control of maldistribution and equivalence ratio. The system
allows adjustment of the Injection timing and adjustment of the
amount of fuel injected for each cylinder. This flexibility is
appealing. However, balancing air-fuel ratios among the cylinders
for each speed and load required an tterative procedure between
the EFI controller and the cylinder exhaust sample measurements.
This proved too time consuming to be practical. Extensive per-
formance emissions mappings with this system await resolution of
this problem so CVS simulation of the data from this system is not
as yet available. However, some steady state test results are
presented in Section IV.
III.3 Simulation of Urban and Highway Driving in a Pinto The ultimate
objective of comparisons among the alternate fuel systems is to
identify the system or systems which will give best vehicle economy
and emissions while maintaining drivablHty. With this in mind,
the steady-state performance and emissions data generated for the
various fuel-air induction systems operating on methanol have been
used to predict fuel economy and exhaust emissions utilizing the
Federal Emission Test Procedure (FTP) and Federal Highway Fuel
Economy Test Procedure (HFETP). To perform this task, a computer
program developed by the Jet Propulsion Laboratory was modified
to include the HFETP and some additionally important factors.
These program modifications are detailed in Section IV.3.
The indolene baseline data was also utilized in the predictions
and provided the basis for relative comparison with the alternate
systems. Table III.l summarizes the fuel economy results on an
energy basis (miles traveled per million Btu of fuel used (mi/106 Btu)),
It illustrates the advantages of methanol and the improvements
derived from the alternate fuel-air induction systems for stoichi-
metric air to fuel ratio (*=1.0) and lean * where thermal efficiency
is high.
-------
TABLE III.l
Energy Based Fuel Economy as a Percentage Improvement Over OEM Performance Using Indolene* from
Computer Simulation of FTP (Hot 1972 Procedure)
Equivalence
Methanol -
Methanol -
Methanol -
Ratio $
OEM system**
Dresserator**
WHB***
1.0
8
18
19
0.9
16
28
0.8
20
29
0.7
22
26
*OEM result at $ = 1.0 as predicted by the computer simulation, approximates the values for 1975 Pintos.
**A further improvement of 2 to 5% is to be expected upon elimination of the maldistribution of air-fuel
mixture in the range 0.8 <_® ^ 1,0.
***This system varied in * with a time averaged value of 0.98,
TABLE III.2
Vehicle Emissions Comparison from Computer Simulation of FTP (Hot 1972 Procedure)
Emissions (gms/mi)
Equivalence Ratio $
CO
Indolene-OEM system 17.0
Methanol -OEM system 13.10
Methanol -Dresserator 11.9
Methanol -WHB 24.02
1.0
UBF
2.18
1.13
1.53
2.16
NOX
4.8
1.91
CO
2.30
2.03
1.79 1.83
2.59 ;
0.9
UBF
1.61
1.14
1.72
NOX
6.0
3.18
3.19
CO
2.05
1.78
1.64
0.8
UBF
1.95
1.92
3.14
NOX
CO
3.27
1.49
1.61
2.0
1.90
0.7
UBF
3.34
5.93
NOX
.45
.35
-------
11
Exhaust emission comparisons are presented in Table 111,2 for
carbon monoxide (CO), unburned fuel (UBF), and oxides of nitrogen
(NOX) on a grams permile{gms/mi) basis. These results represent
the case of no maldistribution (MI = 0.0) since individual cylinder
data was used as input for the simulation. An interesting result
from the simulation predictions is near attainment of the Federal
statutory NOX standard of .40 gms/mi for the OEM system and the
surpassing of the standard for the Dresserator Inductor operating
on methanol at $ * 0.7. The methanol CO and UBF results are
similar to those with Tndolene with the CO emission standard of
3.4 gms/mi being easily met in the lean operating region for all
systems including the Indolene baseline. At stoichometric * the
OEM, Dresserator, and WHB systems on methanol all show 3/4 lower
levels of UBF than Indolene shows on the OEM equipment. Note that
the UBF results are mixed to some extent . Methanol typically
shows rising values as * = 0.7 is approached. In general, UBF and
CO results need to be interpreted cautiously due to transient
emissions behavior which is not taken into account by the steady
state data used as input for the simulation.
In summary, it appears that two of the alternate systems show promise
in increasing fuel economy and reducing exhaust emissions by operat-
ing at very lean values of * on methanol. Methanol's lean burning
capabilities extend the operable range of $ to 0.7 and possibly
lower, thus making it an excellent candidate fuel for meeting the
statutory NOX standard.
Although these two alternate systems provide improvements in energy
based fuel economy and emissions while operating on methanol,
significant further improvement appears possible so that one more
iteration on the design of each will be tested before they are
placed on an automobile for road evaluation.
III.4 Thermochemical Engine Process Modeling: Thermochemical modeling of
engine processes offers a low cost way of studying influences of
compression ratio, spark advance and equivalence ratio on engine per-
formance and emissions. Over the past two years, a computer model
has been developed which uses thermodynamic and chemical kinetic
equations along with engine parameters to predict mean effective
pressure, spark advance for mean best torque, thermal efficiency,
and engine exhaust emissions.
-------
12
After finely tuning the model to the engine performance for a
specific speed and load, the computer model was used to predict
performance at other speeds. Subsequently, it was used to pre-
dict compression ratio effects on performance and emissions.
The results are very encouraging both from a modeling and engine
performance point of view. The model predicted reasonable trends
due to speed changes with the exception of spark advance for mean
best torque. This discrepancy is believed to be due to the squish*
type combustion chamber design of the 2300 cc engine which was not
adequately simulated in the initial computer program.
More importantly, the model predicted a thermal efficiency and
power increase of 14% when the compression ratio was raised from
8.44:1 to 14:1. The NOX emissions increased 18% on a volumetric
basis. However, when the spark was retarded 3° from MBT, the
volumetric NOX at 14:1 compression ratio fell to the level for
8.44:1 with MBT spark timing. This condition still provides a
13.7% increase in thermal efficiency and power and an attractive
12.4% reduction in ISNOX** compared to the 8.44:1 CR.
Although the computer model predicts cylinder heat transfer, it
does not currently model the wall quench zone relative to chemical
reactions. Hence, it needs further development before it will
correctly predict aldehydes and hydrocarbons. As previously
mentioned, the squish chamber of the Pinto engine also needs to be
more accurately modeled. With these improvements, there are
opportunities for simulation studies relative to the control of
hydrocarbon and aldehyde emissions and cold start phenomena.
The computer can also be used to study performance effects due
to the presence of water in the fuel whether it has been
intentionally or unintentionally added as well as the effects of
other blending agents or impurities.
*The cylinder head has a valve recessment zone and a flat faced
zone in close proximity to the piston at top dead center.
**ISNOX is NOX on the basis of grams per indicated horsepower hour.
-------
13
III.5.Cold Starting and Lean Burning: Pure methanol has a vapor pressure
temperature relationship combined with a high heat of vaporization
which causes cold starting difficulties when the ambient temperature
drops into the range of 32-41° F (0-5° C).
Equilibrium thermodynamic studies indicate the potential for
generating a cold start gaseous fuel by decomposing methanol. This
gaseous fuel which is primarily 2H2+CO has wide flammability limits
and offers the possibility of serving as a pilot fuel for cold
starting and lean burning if it is concentrated in the vicinity of
the spark plug. For cold starting, it is estimated that a flow
rate range of 0.1 to 0.6 gram/sec would be desirable. Higher
flow rates are desired, after the engine is running,for continued
lean burn pilot fuel. However, the generation of this gaseous
fuel under cold start conditions is seen as the more difficult
problem and is receiving our primary attention.
Bench type experiments using battery power to heat a small bore,
stainless steel decomposition tube have resulted in decomposition
percentages as high as 50-60% at flow rates of approximately 0.5
grams/sec. This is in the desirable range. The hardware required
to achieve this result appears to be reasonably adaptable to
automotive use. However, the power and energy requirements from
the battery appear excessive—being in the range of a kilowatt for
1-5 seconds. These bench experiments are discussed in detail in
Section IV.5.
An engine experiment has been devised for determining the least
amount of pilot gaseous fuel necessary for cold starting. A cold
chamber is available which allows engine cold soaking and start up.
Hollow electrode spark plugs have been made available.*
Dissociated methanol (2H2+CO)will be supplied through the hollow
electrode to create a rich gaseous mixture near the spark plug
during the engine cranking process.
*Donated by Mr. E. Leshner of Fuel Injection Development Corporation
-------
14
Future plans include conducting the engine cold start experiments
and making an assessment as to the feasability of dissociating
methanol using battery energy for cold starting. If the results
are favorable, the hollow electrode sparkplugs will also be
studied as a means of introducing dissociated methanol as a
pilot fuel for lean burn control during normal engine operation.
111.6. Engine Wear Rate and Crankcase Blow by: Conflicting evidence is
found in the literature as to engine wear rates using methanol
as a fuel in comparison with gasoline.
Oil samples from our test engine have been monitored for accumula-
tion of metals since the test program's inception two years ago.
Atomic absorbtion spectrophotometry has been used for the metals
detection. The engine has experienced repeated speed and load
cycling on both methanol and gasoline fuels.
The evidence to date is summarized as follows:
(a) The influence of the variations of speed and
load on the wear rates of iron, chromium, and
lead were strong enough to mask any variations
in wear rates due to use of different fuels.
(b) In the case of copper, it appears that methanol
causes somewhat higher wear rates. However,
because the differences are small (see Section
IV.6), it will require specific wear rate tests
to fully establish the degree of difference, if
any.
It is intended that the wear rates will continue to be monitored
throughout the ensuing test program.
The composition of the blowby gas which escapes from the com-
bustion chamber into the crankcase of an engine is of importance
in engine wear. These gas constituents come into intimate con-
tact with the lubrication oil and the strong oxidizing
-------
15
constituents such as nitrogen dioxide can be particularly
harmful. As an element of our comparative study of methanol and
gasoline, an effort is being made to analyze the blowby gases from
our test engines.
To date, data has only been taken on the methanol fueled engine.
However, there is data in the literature for gasoline fueled
engines which allows a rough comparison. This qualitative evidence
indicates that the N02 concentrations in the blowby gases from our
methanol fueled engine are about the same as those from gasoline.
This is an interesting piece of evidence because the oxides of
nitrogen in the exhaust of a methanol fueled engine are well below
those from gasoline. Further study is necessary to fully confirm
this evidence and to provide an adequate explanation if it is found
to be correct.
III.7 Vehicle Performance: Three street vehicles, which have been des-
cribed in some detail elsewhere4, have been operated on methanol for
extended periods of time. A 1970 American Motors Gremlin is now in
its 7th year of operation on pure methanol. It has operated through-
out that time period without any failures in major engine or fuel
system components. One car in a pool of cars operated by meter
readers in the City of Santa Clara has been using pure methanol as a
fuel since its purchase over five years ago. It is a 1972 Valiant.
It too has operated with no failures of major fuel system or engine
components throughout its operational life. A third car operated
by the City of Santa Clara is in its third year of operation on methanol
gasoline blends. The blends have varied from 0 to 20% by volume.
Typically it operates on a 10% blend. It too has experienced no major
fuel system or engine component failures. This car is also 1972
Plymouth Valiant.
The vehicles operating on pure methanol have received modifications
to the carburetor jetting to correct for the shift in air-fuel ratio
requirements due to the fuel change. More heat has also been supplied
-------
16
to the intake manifold because of the high heat of vaporization
requirement for methanol. The vehicle operating on blends has not
been altered.
During the past year, an interesting problem has appeared. A
supply of surplus methanol from one of the ERDA laboratories was
donated to our project. The methanol has been stored for approximately
seven years in 55 gallon steel drums. Upon inspection, the methanol
appeared to have a slight discoloration and was known to contain some
water. To avoid phase separation, this methanol was not used in the
vehicle operating on blends, but it was used in the two vehicles
operating on neat methanol. One of these vehicles experienced stoppage
on the road. Inspection revealed a gelatanous mass collected ahead of
the 3 micron porosity fuel filter.
Analysis leads us to the conclusion that a compound, hydrousferric
oxide, is the cause of the trouble. Characteristically, this compound
first formed as very small particles which apparently passed through
the filter in the supply line used for fueling the vehicle. Heat and
agitation associated with vehicle road operation accelerated coagula-
tion such that the precipitated material would not pass through the
fuel line filter and the vehicle starved for fuel. By filtering this
contaminated methanol through activated charcoal, it is being restored
as a fuel suitable for our vehicle use.
The vehicles will be continued in operation and will be the subjects
of mechanical inspection and further CVS testing during the ensuing
12 months.
111.8.a. Photochemical Reactivity Studies: The photochemical reactivity of the
emissions from methanol fueled combustion processes has not been
adequately investigated. Toward that end, an investigation is being
initiated using a smog chamber made available by the ERDA,
Bartlesville Energy Research Center. This chamber is being equipped
with appropriate instrumentation and is expected to be operational
within this calendar year.
-------
17
The objective of the tests now being planned is to compare the
photochemical reactivity of exhaust gases from three fuels:
methanol, Indolene, and Chevron unleaded gasoline with and without
catalysts for each fuel. The reactivity index will be a composite
based on smog chamber measurements of hydrocarbon consumption rate,
NO consumption rate, N02 peaking time, and ozone formation. Exhaust
samples will be taken from a single vehicle run on the Federal
driving cycle.
Ill.S.b. Biological Hazards of Methanol Spills: During the past year
Dr. P. D'Eliscu of the Biology Department with University funds
has investigated four aspects of methanol toxicity relating to
marine or estuarine species:
(a) Acute toxicity to substrate-forming inverte^
brates, algae, free-living vertebrates and
surface water plankters.
(b) Chronic exposure to methanol of selected
invertebrates and subsequent reproductive
changes.
(c) Chronic exposure of various crustacean and
subsequent molting ratio changes.
(d) Comparison of gasoline versus methanol
toxicity under varying conditions.
As a sample of early results, a concentration of about one percent
methanol in sea water proved to be toxic to many common components
of rocky intertidal, mud flat, and estuarine ecosystems if heavy
metals were eliminated from methylation. Lower levels of methanol
proved toxic if metal contamination is considered.
Preliminary assessment of methanol toxicity to small marine and
estuarine organisms is encouraging to environmentalists. The
effects of immediate spills or leaks would probably be minimal
except in a very proximal area where concentrations reach one
percent. Since methanol is quite miscible, volatile and degradable,
-------
18
gross environmental impact from moderate spills is unlikely.
Individual rare species may be more significantly affected, however.
Continuing studies include a comparison of gasoline and methanol
as coastal marine pollutants, the effects of tidal period, and
aeration recovery of organisms from short term exposure.
IH.8.C. Conversion of a Small (60 hp) Gas Turbine to Methanol: A computer
modeling study of a gas turbine combustor contrasting emissions of
NOX and CO from methanol fuel with simulated jet fuel has previously
been reported.5 This model predicts an 80% reduction in NOX and a
45% reduction in CO for homogeneous and stoichiometric conditions
in the primary zone of the combustor which served as a model. For
that study, air entered the combustor at a pressure of 15 atmospheres
and at a temperature of 800° F.
During the past year a 60 horsepower Solar T-45M-13 gas turbine engine
which powers a water pump has been modified to operate on methanol
fuel. The conversion was accomplished by doubling the orifice area
of the sleeve metering valve within the fuel regulator. Exhaust
concentrations of NOX, NO, CO, hydrocarbons, C02 and oxygen were
measured during turbine operation on Jet A fuel and on methanol.
Fuel injection pressure (and subsequent droplet size) was varied by
changing simplex fuel nozzles during the experiments. An air-assist
atomizing nozzle was also tested in an effort to obtain the smallest
possible droplets.
The following two bar graphs (Figures III.l and .2) summarize combustor/
turbine emissions from this and five other experiments and contrast
them with the computer predictions. In each graph is plotted the
relative amount of pollutant emissions for methanol as compared to
distillate fuel. The relative emissions are plotted against com-
bustor inlet temperature (CIT). Two effects can be seen in the graph
of relative NOX emissions. The smallest droplets, produced by air-
aided atomizing nozzles, reduced relative NOX levels by 60%-80% for
methanol for all combustor inlet temperatures shown. The larger
-------
241%
ISO X •
_,
UJ
u.
uj 125 % •
1-
_J
52 100 % •
O
u)
- 0 •« -
Z UJ 75 %
ui or
ox|
Z 50 % -
UJ ~*
H <
< x
uj uj 25% •
tr z
0
SIMPLEX 223 PSI Ep. NQ UM|T .
36 PPM tPA NUX LIMI ' •
75 PPM AT 15% 02
140 %
MEDIUM SIZED DROPLETS
SIMPLEX 340 PSI
24 PPM
96 TO __._
- ^^H ««tf\&JM«l l^r»» MMPMI««VI««ftl
SMALLEST DROPLETS • DISTILLATE REFERENCE
SIMPLEX 400 PSI • _.,_,
19 PPM • FUEL
1
SOLAR •
T 45 M B
1
*o qt H i 1
"OE n • QM
32 % M370OIB 1 1 cnnn 1
— i nviMenna H [ FORD | jQ
-------
UJ
U.
W
3OO
25O X -
200 % -
2 iso % -
O
°
UJ o
> Z 100 % -
UJ UJ
or Z
50% -
400 '
117%
LARGEST DROPLETS
929 PPM-SIMPLEX
87% SMALLEST DROPLETS
383 PPM-SIMPLEX
77%
AIR ATOMIZING
NOZZLE
300 PPM
36 PPM
EPA co LIMIT:
90 PPM AT 15 % 02
220%
63 PPM
100%
COMPUTER
PREDICTION
DISTILLATE
REFERENCE FUEL
COMPUTER
PREDICTION
METHANOL
200
FIGURE Ht.2 :
300
400
500 600 700 800 900
COMBUSTOR INLET TEMPERATURE (°F)
CO EMISSIONS - METHANOL VS JET FUEL IN GAS TURBINES
1000
1100
1200
ro
o
-------
21
droplets, produced by simplex-fuel pressure atomizing nozzles, pro-
duced NOX emissions from methanol in the range from 96% to 241% of
distillate fuel NOX emissions. Therefore, at low CIT the initial
droplet size is very important for actually achieving the low NOX
potential of methanol combustion that has been predicted by computer
study and produced by turbines at higher CIT values.
A detailed study of the droplet burning contrasts between methanol
and jet fuel would be necessary to accurately explain the contrasts
seen in this evidence. However, it is believed to be strongly related to
the relatively high heat of vaporization of methanol (506 Btu/lbm vs
110 Btu/lbm for distillate fuel) and the doubling of the fuel flow
rate for methanol to achieve the same energy release rate. The com-
bination of these two factors indicates that for equal droplet sizes
the methanol drops will take longer to evaporate compared to distillate
fuel droplets. The proportionately larger amount of burning in the
form of droplet combustion, which occurs near stoichiometric, high
temperature conditions, results in the excess NOX production for
methanol with large initial droplet sizes. It is thought that the
Delavan Swirl-Air nozzle produced very small droplets (<20yM) which
overcame the droplet evaporation limitation of the simplex nozzles
and reduced the percentage of droplet burning and hence NOX production.
The relative carbon monoxide emissions were also decreased with smaller
droplets in the present experiment. However, it is evident from the
experiments shown in Figure III.2, that the relative CO emissions are
strongly dependent on CIT. This trend is clearly evident for all
experiments using the more efficient air-aided atomizing nozzles.
The high absolute levels of CO at low CIT values are caused by the flame
quenching effects of the cool secondary air mixing with the combustion
products. This quenching effect is greatly aggravated in the T-45M-13
by asymmetric airflow into an elbow type combustor. As the CIT values
rise, among the more modern and symmetrical combustors reported in the
literature, the quenching effects are strongly reduced and relative CO
values correspondingly decrease. Methanol's predicted lower CO
-------
22
emissions were achieved at very low and at very high CIT, as the
graph indicates. The absolute values of NOX and CO are well below
the EPA standards of 75 PPM NOV and 90 PPM CO for all of the engines
/\
with air-aided atomizing nozzles.
Engine power output was achieved at consistently lower (40° F) exhaust
temperatures with methanol. This resulted from the increased cooling
of incoming combustor air caused by methanol's high latent heat of
vaporization and by increased mass flow through the turbine with
methanol. The practical consequence is that power output may be
increased in a turbine which is limited in allowable sustained turbine
inlet temperature. Alternatively, at the same power output the
maintenance interval may be significantly increased due to lower
temperatures at the turbine inlet.
In conclusion, the NOX emissions of a small gas turbine with low CIT
were reduced by as much as 68% with methanol fuel and careful fuel
atomization. The carbon monoxide emissions were also reduced 23%.
There is general agreement in trend between this evidence and the
experimental and computer simulation evidence found for turbines
operating at higher CIT values. Either engine power or maintenance
intervals may be increased due to reduced turbine inlet temperatures
with methanol. Droplet burning studies should be conducted to fully
understand the best burning technique for liquid fed methanol flames
so that its excellent burning characteristics can be fully exploited.
-------
23
IV.DISCUSSION OF PROGRAM RESULTS
The purpose of this section of the report is to provide a more detailed
explanation of the component elements of our program, the methods of pro-
cedure in the investigations, and interpretations of the evidence derived
from the various phases of the program to date.
IV.I. Steady State Engine Performance and Exhaust Emissions Character-
ization - Methanol vs Indolene
An experimental parametric evaluation of a Pinto 4-cylinder 2300 cc
engine operating on Indolene and methanol is presented in this
section. In section IV.1.1 comparative engine performance and
exhaust emissions are contrasted for the "as supplied" or OEM
equipment operating on Indolene, and the modified (enlarged car-
buretor jets and increased manifold heating) OEM equipment operat-
ing on methanol. Additional comparative data for three alternate
fuel-air induction systems operating on methanol is also presented.
The maldistribution of the fuel-air mixture among cylinders in con-
ventionally manifolded multicylinder engines is one problem that
was eliminated with two of the three alternate systems. To
properly treat the experimental data of the system with maldistribution,
it became necessary to define a maldistribution index. This topic is
covered in section IV.1.3. Operational and design aspects of each
of the three alternative systems are discussed in section IV.2.
IV.1.1. Comparative Steady State Performance and Exhaust Emissions: The
objective of the comparative mappings of performance and emissions
is to draw contrasts between gasoline and methanol. These contrasts
are drawn in terms of engine power, thermal efficiency, lean burn
capability, and the pollutant exhaust emissions of carbon monoxide
griC--;C£Pj)-> -.M.lb.ur.n.ed fuel (UBF), oxides of nitrogen (NOX) and aldehydes
or ^(HAMQl-o rThji?;2(?feie§-li^xis •^P1?-?l-fpom.•£w?r.4ire-Gtt&rvs. ^wJiich. can
be summarized in rq4estto^fot?m;-.u%)(-Wi;th-'minima;l- modifications,:.can
a conventional gasoline engine yield equivalent or better operation
while operating on methanol? b) Can additional improvements on
-------
24
methanol be attained through the use of alternate practical fuel-
air induction systems?
IV.1.1.1. Experimental Procedure: A 1975 California Ford Pinto 2300 cc
engine was used to drive a laboratory electric dynamometer and
instrumented for the measurement of engine control and output
variables and exhaust emissions. A schematic which shows this
equipment and the points of measurement and control for the
steady state testing of the OEM equipment is shown in Fig. I V.I.
The intake manifold and heater were removed for tests with two of
three alternative systems, and exhaust "headers" replaced the
OEM exhaust for two of the three alternate systems.
Five different configurations of air-fuel preparation systems were
investigated. These are summarized below:
1) The stock OEM equipment operating on a reference gasoline
(indolene), however, without E6R.
2) The stock OEM equipment with enlarged carburetor jets to
permit double the fuel flow rate for operation with methanol
as a fuel, and externally controlled intake manifold heating.
3) The stock intake manifold with the Dresserator Inductor
instead of the stock carburetor with methanol as a fuel and
externally controlled manifold heating.
4) An electronic fuel injection (EFI) system with its own in-
take manifold operating on methanol.
5) The WHB individual cylinder induction system (no intake
manifold) operating on methanol.
The Dresserator and WHB systems utilized a header type exhaust
system during portions of the testing. The first four configurations
were mapped for engine performance and exhaust emissions according to
the steady state test matrix shown in Table IV. 1.
-------
MANIFOLD
HEATER
SPARK
TIMING
CONTROL
HEAT
EXCHANGER
BLOWER
VALVE
EXHAUST SAMPLE VALVES
(I PER CYLINDER, I COMPOSITE)
» TO
FUEL PRESSURE AV__
EMISSIONS
BENCH
1
INTAKE MANIFOLD |
(BOTTOM VIEW)
INLET
PRESSURE
DRAFT
GAUGE
COOLING
WATER
P.C.V.
EAR.-P.C.V. SLOT^
LAMINAR
FLOW
ELEMENT
HEATED LINE
FUEL
FUEL FLOW
TRANSDUCER
EXHAUST
SYSTEM
CHART
RECORDER
THROTTLE
CONTROL
MANIFOLD
VACUUM
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
I >
THERMOCOUPLE LOCATIONS
f. I
FIGURE Iff-1: TEST ENGINE-DYNAMOMETER CONFIGURATION
I. CYL I
2. CYL I
3. CYL 2
4. CYL 2
5. CYL 3
6. CYL 3
7. CYL 4
8. CYL 4
-HEAD 9. WATER TEMP -PLENUM
-PLENUM 10. WATER TEMP -HEAD
-HEAD II. OIL TEMP
-PLENUM 12. L.F. E. AIR TEMP
-HEAD 13. CYL I EXHAUST
-PLENUM 14. CYL 2 EXHAUST
-HEAD IS.CYL 3 EXHAUST
-PLENUM 16. CYL 4 EXHAUST
17, EXHAUST EMISSIONS SAMPLE
ro
tn
-------
26
TABLE IV. 1.
BASE LINE TEST MATRIX
Engine
Speed
(RPM)
Idle
1000
1500
2000
2500
3000
3500(4)
4000(4)
*(D
0.6(5)
0.7(5)
0.3
0.9
1.0
1.1
1.2
Load
(Fraction) (2)
1/4
1/3
2/3
Full
Spark
Setting
(°BTDC)
MBT<3>
Other(4)
Ambient
Temperature
(°F)
70
Other(4)
Notes: (1) Equivalence ratio, i.e., stoichiometric -r actual A/F.
(2) Fraction load is designated by intake manifold vacuum:
Full = 0" hg vac; 2/3 = 7"; 1/3 = 14", 1/4 = 18"
(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
The prinicpal independent variable in this testing was the fuel-air
equivalence ratio (*). With the exception of the WHB system, tech-
niques were utilized which permitted a wide range of * values for
each configuration. While operating on methanol, the different con-
figurations were tested in the range of * = 0.6 and * = 0.7 to
explore methanol's lean burning capabilities.
The range of $ for the WHB system was fixed by the geometry of
the fuel metering hardware, and thus * could not be the primary
independent variable for tests with this system. By changing the
fuel metering hardware, nominally rich and lean test points (relative
to a standard sets of jets) were obtained.
-------
27
The test matrix also emphasized part load (high manifold vacuum)
and low speeds since the combination of these two is the normal
engine operating mode for everyday urban driving. The steady
state maps generated from this testing were then used as input
to a computer program which predicted fuel economy and emissions
for the Federal Test Procedure (FTP) and the Highway Fuel Economy
Test Procedure (HFETP). The results are detailed in Section IV.3.
IV.I.1.2. Power, Thermal Efficiency and the Lean Misfire Limit: Figures IV.2.
and IV.S.display, for wide open throttle (WOT), indicated and brake
horsepower. Figure IV.4.presents indicated thermal efficiency.
These are shown as functions of $ for the five fuel-air induction
systems previously cited. Here * is the composite or average
value of the four cylinders. A new variable termed the maldistribu-
tion index (MI) appears with each system curve. This index is
defined out of necessity to quantify effects on power, thermal
efficiency, and emissions of the inherent maldistribution of the
fuel-air mixture among the cylinders.
The value of MI accompanying each curve is defined as the mean
deviation (across the range of * values) for the two most deviant
* cylinder values. A more complete definition of maldistribution
can be found in Section IV.1.3. For now it is sufficient to note
that increasing values of MI mean greater maldistribution, and the
effect on power and thermal efficiency can be estimated fairly
accurately.
Power and the Lean Misfire Limit: Figure IV.2,indicates that at
WOT methanol yields an increase in indicated power ranging from 6%
(OEM-methanol @ = 0.8) to 11% (WHB @ * = 1.0) over the OEM-Indolene
results. All the alternate systems yield higher power than either
OEM-Indolene or OEM-methanol systems.
-------
or
IU
UJ
CO
or
o
0
UJ
o
z
60.0
56.0
52.0
48.0
44.0
40.0
36.0
32.0
28.0
24.0
20.0
1
»
-I1
r°
li
i
JEFI
\MEOH)
1
/
/'
/ ,
*•
/ *
/
RESSER*
MEOH)
J
T
/
/
/ /
/ /
* /
)••
*•
f
*
TOR
(OEM
l(MEO
kM^
.INDOLENI
/^
/^
X
/
LEAN
H)
-ii
^
c x-^
X
X
RICH^
•u^"* '
—
— —
^--^
:+">+ +
^
+ +
Q DENOTES REAL LEAN LIMIT
C DENOTES EFFECTIVE LEAN LIMIT
PINTO 23OO CC ENGINE
MBT SPARK TIMING
2000 RPM WOT
c
4-» -f •+ C
E
_ii v
(MALDISTRIBUTION
INDEX)
)EM INDOLENE (0.06)
JEM METHANOL (0.17)
IRESSERATOR (0. IS )
Fl (0.01)
VHB (0. 01)
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
FIGURE ET.2 COMPARATIVE WOT INDICATED POWER - INDOLENE (OEM) VS METHANOL
(OEM AND ALTERNATE SYSTEMS)
ro
Co
-------
gj
I
§
60.0
56.0
52.0
48.0
44.0
40.0
36.0
32.0
28.0
24.0
20.0
/
/
/ ;
/
i
/
/ \
//.
/ +
f i-
/ *
+
+
y
X
•
X
.' X
/ x--'
,-v//
^ /
^ / >
/
LEAN
^
X.' S*
V /
/
/
/"
RICHr
*-¥*J!
^^"
- ...^
-e-f ••• -
• -i-
PINTO 2300 CC ENGINE
MBT SPARK TIMING
2000 RPM WOT
C
+ 44 -»• C
E
— . .— ..* \
( MALDISTRIBUTION
INDEX)
)EM INOOLENE (0.06)
)EM METHANOL (0.17)
IRESSERATOR (0.15)
:FI (o.oi)
VHB (0.01)
ro
10
0.50 0.60 0.70 0.80 0.90 1.00 1. 10 1.20 1.30 1.40 1.50
FIGURE Iff. 3 : COMPARATIVE WOT BRAKE HORSEPOWER - INDOLENE (OEM) VS
METHANOL (OEM AND ALTERNATE SYSTEMS)
-------
42
40
38
o
UJ
2 36
u.
u.
UJ
tr
UJ
UJ
<
o
o
34
32
30
28
26
24
x
/ 4
/ I
, j
/ +
•*•
*
I
M * * «
<*" "^
**,
PINTO 23OO CC ENGINE
MBT SPARK TIMING
2000 RPM WOT
(Ml )
OEM INDOLENE (0.06)
OEM METHANOL(O.IT)
+ •»** DRESSERATOR (0.15)
EFI (0.01)
WHB (0.01)
*
^\ K
\"
^"" -^«
^
f
\
\
J.EAN
•
\
N *.
\ V
"*>
\^
\^
\
RICH^
'\
•
\
•
\
\
\
i*
\
.
\
\
\
•\
\\
•\
*.\
4
\
\
\
\
\
\
\
". \
m \
" \
\
\
\
050 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40
I
FIGURE DT-4: WOT COMPARATIVE THERMAL EFFICIENCY FOR OEM AND ALTERNATE
FUEL- AIR INDUCTION SYSTEMS
CO
o
-------
31
The graphical evidence is typical I.C. engine behavior. As *
becomes lean, there is a fall-off in power. For the systems with
significant maldistribution, the apparent or effective lean mis-
fire limit occurs at higher equivalence ratios as is indicated in
the top left hand portion of the curve. The value of * for the
cylinder approaching lean misfire is also indicated. The EFI
system was capable of operating at the leanest * since each injec-
tor could be set to eliminate maldistribution. The WHB system
lean limit does not appear since jet geometry did not permit *
values much leaner than * = 0.9. Note that methanol's true lean
limit was found to be close to * = 0.58 for both the EFI and
Dresserator data, while the OEM-gasoline data indicates its
characteristic lean limit near * = 0.77. The MI values of 0.15
for the Dresserator and 0.17 for the OEM-methanol indicate that
in the lean region gains in power of 6 to 10% are possible if
maldistribution is eliminated. This would also extend the lean
operating limit of both systems and simultaneously raise the
thermal efficiency. The implications of this effect are important
since it is in the lean region where peak thermal efficiency occurs.
The brake power results at WOT in Figure IV.3 show trends similar
to the indicated power. The WHB system shows slightly higher values
because the pumping losses at WOT are somewhat less than those of
the other systems.
Thermal efficiency (Fuel economy on an energy basis): The indicated
thermal efficiencies* in Figure IV.4-correspond to the power curves
of Figure IV.I. All of the alternate systems and the OEM-methanol
configuration show significant improvements in thermal efficiency
over the OEM-Indolene results (15% for WHB at = 1.0, 10% for
Dresserator at $ = 0.9 and * = 0.8, and 9% for the OEM-methanol at
$ = 1.0). Location of the peak efficiency depends on the system
used and its MI value, but all occur in the lean region as expected.
For this engine speed of 2000 RPM the EFI shows the leanest peak of
*Based on a lower heating value of 19,032 btu/lbm for Indolene and
8570 btu/lbm for methanol.
-------
32
all the systems at * * 0.75. The Dresserator yields the highest thermal
efficiency (38.5%), but because of the maldistribution, the fall off at
leaner * values is very abrupt. The maldistribution index of 0.17 for
the Dresserator indicates that if the maldistribution were eliminated,
an additional gain in thermal efficiency in the range of 5 - 10% is
possible. The WHB system shows the highest thermal efficiency at
stoichiometric * of all the systems and thus makes it a very attractive
system. The exhaust headers were responsible for a 6% increase in
thermal efficiency at this speed and throttle setting. The Dresserator
on the other hand also utilized the headers but showed no benefit in
thermal efficiency over the stock exhaust system. If maldistribution
was eliminated from the OEM-methanol and the Dresserator-methanol con-
figurations, it is possible that these two would match the WHB the
thermal efficiency at $ = 1.0.
Indicated thermal efficiency evidence, for 1/3 load, from all the systems
at 2000 RPM is seen in Figure IV.5. The trends are similar to the WOT
results with a few exceptions. In general, all the alternate systems and
the OEM-methanol configuration show significant gains over the OEM-
Indolene baseline results. Dresserator and WHB systems yield a 15% gain
in thermal efficiency at * = 1.0. The MI values at the 1/3 load setting
are lower for the systems with inherent maldistribution (OEM-Indolene,
OEM-methanol, and Dresserator). As expected, the effective lean misfire
limit occurs at higher values of * than the WOT conditions for both
methanol and Indolene with the exception of the OEM-methanol results.
This exception is a result of maldistribution at the WOT condition which
masks the true lean limit. The Dresserator and WHB systems show the
highest thermal efficiencies (38% @ = 0.75 and 37% @ * = 0.87, respectively)
but the fall off at leaner values of $ is more dramatic than the WOT cases.
Unlike the WOT results for the EFI, the part load results show a thermal
efficiency lower than the OEM-methanol which approaches the OEM-Indolene
results in the very rich region. Also, unlike the WOT results, the WHB
system shows results similar to the Dresserator. Test data with the OEM
exhaust system indicated that neither the Dresserator nor the WHB system
gained thermal efficiency at part load when the exhaust headers were
installed.
-------
50
40
38
36
34
rr 32
ec.
UJ
O
UJ
30
28
o
o
* 26
24
22
1
ft [ORE
V LJME
* i
v
4
•f
-f
+
4
4-
+
4
t
SSERATOR
OH)
OEM
(MEOH)
» * - .
y
O DENOTES REAL LEAN LMIT
DENOTES EFFECTIVE
LEAN LIMIT
PINTO 2300 CC ENGINE .
MBT SPARK TIMING
2000 RPM
(Ml)
OEM INDOLENE (.022)
OEM METHANOL(.09)
+ + + •> DRESSERATOR (0.04)
EFI (0.02)
WHB (0.02)
®~~KINDOL
L
u
/^;,
~^^^
/*
/
/
ENE)
LEAN
;\
\ ^
\
*
— ^
Rjgt
'v
sS
^
\
S,^^ v
^\
\\
1
\
\
\
\
\
\
\
<
\
\
\
V
0.50 0.60 0.70
0.80
1.10
1.20 1.30
0.90 1.0.0
I
FIGURE ELS 14" HG MANIFOLD VACUUM COMPARATIVE THERMAL EFFICIENCY FOR
OEM AND ALTERNATE FUEL - AIR INDUCTION SYSTEMS
1.40
CO:
CO
-------
34
The Dresserator again shows a high thermal efficiency at very lean
$ and a characteristic abrupt drop off in thermal efficiency past
that peak. Dresserator data at all speeds and loads shows this
severe drop off in thermal efficiency. The OEM-methanol and EFI
results do not display this type of behavior. It is claimed that
the Dresserator produces a highly atomized and homogeneous air-
fuel mixture. Whether or not this is indirectly confirmed by the
thermal efficiency results pends further analysis of the data since
differences in maldistribution and MET spark setting cloud this
issue. The drop off is due to a substantial power decrease, yet,
no misfire occurred. Further analysis is under way to explain why
the Dresserator unit shows this characteristic hump in thermal
efficiency at all speeds and loads while the OEM-methanol and EFI
systems do not.
In summary, these power and thermal efficiency results are typical
of the results seen at other speeds. In general, the alternative
systems yield improvements in fuel economy (energy basis) of from
5 to 18 percent over the OEM-Indolene baseline at like values of
$ across the speed and load range. The WOT power from methanol
ranges from 2 to 10% higher than from Indolene. The lean flarnmability
characteristics of methanol permit very lean steady state operation
that is well beyond the limit of gasoline.
IV.1.2. Comparative Exhaust Emissions: Like the power and thermal efficiency
of the four-cylinder engine test results, exhaust emissions were
highly influenced by the variations in among the cylinders for the
steady state test points. This maldistribution can effectively
randomize the composite (average of the four cylinders) emissions
measured during steady state testing when plotted against the
averaged value of $. To preclude this kind of result, emission
measurements were taken for each of the individual cylinders. The
results are presented on a grams per indicated horsepower-hour
(gms/ihp-hr) basis as functions of each cylinder's value of *••;
The measured emissions include the principal combustion species
C02» 02, and h^O (deduced) and the pollutant species CO, NO, NOX, DBF,
and aldehydes (HCHO). Figure IV.6.is a schematic of the exhaust
sampling system and Table IV.2,indicates the method of detection
and equipment used in these measurements.
-------
ZERO
GAS
FLOW
METERS
EXHAUST
DUMP
BACKFLUSH
AIR SUPPLY
02
ANALYZER
co2
ANALYZER
HIGH CO
ANALYZER
LOW CO
ANALYZER
MID RANGE
CO
ANALYZER
CONTROL
VALVES ,
®... fv\
\SJ
CAL/T
GAS L.
CAL
GAS
®/rv
\L)
CAL
GAS
e
|
(3
r®"
T
O
r
e
PUMPS
,
] INDICATES HEATED
I PORTION OF SYSTEM
EXHAUST
DUMP
I !
rA
o
?r
Vl7
IP
)P
'ERS
?(b
CO
^
i
BR
S
3
NDE
••M*
»
>
]
>
»
]
1NE
OL
3°F
>
>
3
:NS
-C
-®
r?»
/o\
A4^
30'HEATED LINE 400°F
CONDENSATE
DUMP
HEATED
400° F
INSULATED : .
LINE
CYLINDER 4321
COMPOSITE
EXHAUST 00—^
SAMPLE
N0-N0x_ ANALYZER J j
IMPINGER
CONTAINING
I MBTH
SOLUTION
FOR
ALDEHYDE
DETECTION
co
en
FIGURE Iff.6 : EXHAUST EMISSIONS SAMPLING SYSTEM
-------
36
TABLE IV.2
Exhaust Emission Measurement Technique
Species
C02
CO
02
UBF
NO/NOX
aldehydes
Method
NDIR*
NDIR
polarographic
heated FID**
chemi 1 umi nescent
MBTH***
(wet chemistry)
Continuous
Monitoring
yes
yes
yes
yes
yes
no
Manufacturer
and Model
Beckman-315B
Beckman 31 5B
Beckman-OM-11
Beckman -402
Beckman-951H
B & L Spectronic 20
*NDIR-nondispersive infrared
**FID-flame ionization detector (measured response factor of
0.85 for CH3OH)
***MBTH-3-methyl-2 benzothiazolone hydrazone (total aldehydes as HCHO)
Comparative C02» CO, and Og: The exhaust emission species of
C02, CO, and 02 are three of the combustion species which follow
fairly predictable behavior. Figure IV.7. indicates that their "dry"
mole percentages are usually functions of * only. The results are
curve fits of individual cylinder measurements for all the data from
all the fuel-air induction systems investigated. The methanol results
show slight differences from the Indolene. The different systems on
methanol produced virtually invariant results. Characteristic of
all the systems is the scatter of data around stoichiometric $. The
Dresserator C02 points which are circled illustrate the trend. It
is believed that this scatter is indicative of the nonhomogeneity
of the inducted charge. If this is true, then both lean and rich
burning occurs within any cylinder. Thus, the C02 results take on
values lower than the peak value of C02 that would be attained with
a homogeneous charge at * = 1.0. Further study is needed to determine
whether or not the experimental evidence can be used to yield a
quantitative statement about the. degree of stra,tificattonA
-------
37
8
g
g
g
g
g
g
g
g
g
8
i
/
/
i
/
/
LEAH
^*"
X
'A
m
— -.
i
t
^1
I:
"V
N
\i
DOLENE-
*
O-
V
Y
\
«»«.«.>
— icnum.
. X
\ 1
PINTO Z3OO CC LNulNL.
METHANOL CURVE INCLUDES
DATA FROM OEM, DRESSERATOR,
WHB, AND EFI SYSTEMS.
MALDISTRIBUTION INDEX « 0.0
0.90 0.60 0. TO 0.80 0-90 1.00 l.»
1.30 I.4O 1.90
ff
K
% oz vs
EQUIVALENCE RATIO
O.9O 0.60 0.70 O.6O O.90 1.00 I.K) 1.20 I. SO 1.40 I.SO
% CO VS
EOUVALENCC RATIO
).9O 0.60 0.7O 0.80 0.90 1.00 1.10 1.20 1.30 1.40 I.5O
$
FIGURE 1ST. 7: C02, CO, AND 02 VERSUS CYLINDER EQUIVALENCE RATIO.
INDOLENE VERSUS METHANOL.
-------
38
Comparative NOV: Through sampling of the individual cylinders of
the Pinto engine and curve fitting the MBT data points, comparative
MI = 0.0 (no maldistribution) curves of NOX emissions for the
various systems were plotted as a function of the cylinder $ value.
In Figure IV.8, the WOT results at 2000 RPM show that methanol
yields lower values of NOX than the OEM-Indolene data. The WHB
system shows the lowest peak values which are about one-half the
magnitude of Indolene results. The EFI (and perhaps the WHB system)
show the lowest NOX at stoichiometric *. Note that the maldistribu-
tion index has not been indicated since cylinder data was used to
construct the curves. The close agreement in results for the
methanol systems indicates that mixture preparation may be com-
parable for OEM-methanol, EFI and Dresserator systems or that mixture
preparation doesn't dramatically effect in-cylinder NOX formation.
The WHB results near $ = 0.9 may indicate a difference in fuel-air
preparation in comparison to the others, since spark advance was
nearly the same (32.5 to 34°BTDC) for all the systems at this
-------
20.0
18.0
16.0
14.0
12.0
10.0
X
o
8.0
6.0
4.0
2.0
0-0
LEAN
RICN
PINTO 2300 CC ENGINE
MBT SPARK TIMING
2000 RPM - WOT
(MALDISTRIBUTION INDEX-0.0)
OEM INOOLENE
OEM METHANOL
+ f4 + DRESSERATOR
EFI
O WHB
0-50 0.60 0.70 0.80 0.90 I.OO 1.10 1.20 1.30 1.40 ISO
FIGURE H-8: COMPARATIVE WOT STEADY STATE NOX EMISSIONS - INDOLENE (OEM)
VS METHANOL (OEM AND ALTERNATE SYSTEMS)
GO
-------
II.0
10.0
I
Q.
I
3.0
2.0
1.0
0.0
PINTO 2300 CC ENGINE
MBT SPARK TIMING
2000 RPM-1/3 BRAKE LOAD
(MALDISTRIBUTION INDEX t 0.0)
OEM INDOLENE
OEM METHANOL
+ +*-+ DRESSERATOR
EFI
O
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
* CYLINDER
IGURE ET.9 COMPARATIVE 1/3 BRAKE LOAD STEADY STATE NOX EMISSIONS -
INDOLENE (OEM) VS METHANOL (OEM AND ALTERNATE SYSTEMS)
-------
41
all the methanol systems show lower peak values of NOX at the
part load settings. In the very lean region ($ = 0.7) the NOX
is under 0.5 gms/ihp-hr. This low value along with the corres-
ponding WOT value of approximately 1.0 gm/ihp-hr provides the
incentive for operating the engine at this equivalence ratio to
meet the Federal NOX statutory standard of .40 gms/mi. The FTP
simulation results (Section IV.3) indicate that Dresserator and
the OEM-methanol configurations can better and nearly meet
respectively the standard at this equivalence ratio.
Comparative DBF: Unburned fuel comparisons at WOT and 1/3
brake load for 2000 rpm appear in Figures IV. 10. and IV.11.
Indolene is reported as CH]^ and methanol as CH30H on a grams
per indicated horsepower-hour basis. Again, individual cylinder
data has been curve fit and displayed as a function of the cylinder
$ value. The systems operating on methanol show lower values of
UBF when compared with the OEM-Indolene. The benefit is small but
discernible. The WHB results are not shown at WOT because exhaust
headers were also used with this fuel system. The headers, in con-
junction with the cam which supplies significant valve overlap,
aspirated fuel from the intake to the exhaust. However, in moving
to the part load condition (Figure IV.11) the results from the WHB
system were comparable with the other alternative systems on
methanol. Again, all the other methanol systems showed a slight
benefit over the OEM-Indolene and were comparable among themselves.
Characteristic of lean operation on liquid fuels is the increasing
nature of UBF emissions as the lean mixture limit is approached.
At * = 0.7 the engine is not seeing any misfire on methanol, yet
UBF is quite high.
In interpreting these results, it should be mentioned again that
the exhaust gases were sampled near the exhaust valve on the
downstream centerline of the exhaust port (1" from the valve). This
implies that although the results show correct comparative
-------
20.0
18.0
16.0
14.0
~. 12.0
I
i
0
CD
10.0
8.0
6.0
4.0
2.0
0-0
PINTO 2900 CC ENGINE
MBT SPARK TIMING
2000 RPM
( MALDISTRIBUTIO
OFU II
OEM ft
•M- + + DRESS!
EFI
\
\
\
14 INDEX >
4DOLENE
0.0)
(ETHANOL
iRATOR
\
T \
\ \
\ X
X \
\^
\
V
- -^
LEAN
^
^~^T.I
RICHf
— "^ ***
*
*>f
/ t
' £
-/
t
ro
0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40
$ CYLINDER
FIGURE IT. 10: COMPARATIVE WOT STEADY STATE UBF EMISSIONS INDOLENE (OEM)
VS METHANOL (OEM AND ALTERNATE SYSTEMS)
-------
CC
I
I
Q.
X
9.0
8.0
7.0
6.0
OT 5.0
u.
m
4.0
3.0
2.0
1.0
0.0
\
•
\
\
•
\
\
•
^,
A
\\\
\^
yN.
y^^
i
LEAN
«^*J
PINTO 2300 CC ENGINE
MBT SPARK TIMING
2000 RPM - 1/3 BRAKE LOAD
( MALDISTRIBU1
OFM
OEM
+ -*•+•+• + ORES
EFI
WHB
RICH
X
*~S
**K~
/
^S'
1
*
<"/
" ''/
* //
x /• >
^
/ ^
/
ION INDEX * 0.0)
INDOLENE
METHANOL
SERATOR
'/
i
/
>
0.50 0.60 0.70 0-80 0.90 1.00 1. 10 1.20 1.30 1.40 1.50
CO
FIGURE Iff. II: COMPARATIVE 1/3 BRAKE LOAD STEADY STATE UBF EMISSIONS
INDOLENE (OEM) VS METHANOL (OEM AND ALTERNATE SYSTEMS)
-------
44
evidence, they include a systematic bias that is introduced by
two effects. First, there is stratification of UBF in the
exhaust gases due to the manner in which exhaust gases are
expelled from the cylinder on the exhaust stroke. Secondly,
oxidation of UBF takes place in the exhaust system in the
presence of excess oxygen, with the effect being more pro-
nounced at the higher loads. Measurements of the composite
exhaust sample downstream indicate that these reported results are
at least two times higher than results that would be measured at
the tailpipe of an automobile which didn't have an oxidation
catalyst.
Comparative Exhaust Aldehydes: Figure IV.12. is a plot of
exhaust aldehydes (as formaldehyde) as a function of the com-
posite or average «> for the four cylinders across the speed
and load range. The bar with each system curve represents the
range of the data. Few individual cylinder measurements were
made due to the time involved in performing the wet chemical
technique (MBTH*) necessary to determine total exhaust alde-
hydes. As a result, the maldistribution of fuel and air plays
a role in obscuring the comparative evidence for the systems
with significant MI. In general, the aldehyde results for
methanol fueled systems show values of aldehydes ranging from
two to ten times greater than the aldehydes associated with
the OEM system operating on Indolene. The WHB system shows the
highest aldehydes which is in agreement with UBF findings. In
contrast to Indolene results, all the systems operating on
methanol show increasing aldehydes at rich values of <*>. Though
not shown, the EFI system produced aldehyde results as high as
4 gms/ihp-hr at * = 0.9 when the injection timing was close to
top dead center. Injection timing could also be manipulated in
the WHB system by selecting the proper main jet geometry. For
the range of jets observed, the aldehydes did not vary
significantly.
*2 Methyl-3-Benzothiazolone-Hydrazone
-------
45
a:
a.
3.20
2.80
2.40
2.00
O
I
O
eu
o
CO
1.60
1.20
CO
LJ
O
UJ
_i
<
0.80
0.40
0.00
DRESSER
\
&TOR
t
I
^K
^
7
*
N
#•
*
\
V
\
X*.
LEAN
.^-DATA
t
X
^^.
^
PINTO 2300CC ENGINE
MBT SPARK TIMING
1500-4000 RPM
1/3 1
—V W
+ -f* 4
— .. — .
RICH^
RANGE
^.^^J-^
— — . .
-OAD TO WOT
OEM INDOLENE
OEM METHANOL
DRESSER ATOR
WHB
•
">^
«
/
s
1
-OEM
0.6 0.7 0.8 0.9 1.0 M
^COMPOSITE
1.2
1.3
FIGURE 1Z.I2: COMPARATIVE EXHAUST ALDEHYDES VS
EQUIVALENCE RATIO
-------
Though aldehyde emissions are higher with methanol, their
magnitudes are small in comparison to other exhaust species.
Nonetheless photochemical smog consequences with increased
aldehydes is not known and thus methods for reducing this one
pollutant from methanol fueled engines is being studied.
IV.I.3. Definition of a Maldistribution Index (MI): Maldistribution
or variations in equivalence ratio (*) among the individual
cylinders of multicylinder engines is characteristic of
practically all production SI engines which have an intake
manifold with a venturi type carburetor. Poor nebulization
or atomization of the fuel droplets combined with inadequate
mixing with the air stream leads to this phenomena. The
4-cylinder Pinto engine used in generating the steady state
maps of performance and emissions is no exception to this rule.
Figures IV.13. shows the variations as a function of manifold
vacuum at two engine speeds for the OEM baseline tests on Indolene
and the OEM baseline tests on methanol, respectively. The average
$ of the four cylinders is used as a normalizing parameter
($ cyl/$ average). Characteristically, both Indolene and methanol
fuels show increasing maldistribution as the full load condition
(0.0 in Hg) is reached. This maldistribution can be bounded in
terms of the leanest and richest cylinder which gives rise to a
definition. The maldistribution index (MI) is defined as:
IYIJ E $ (richest cylinder) -
-------
AVO
0 -
© -
A -
X -
© -
0.70
0.80
0.90
1.00
1.10
1.20
1.30
SPEED
2000
RPM
3000
RPM
MAN VAC
(IN HG)
0'
7"
14"
0"
7"
14"
MALDISTRIBUTI
INDOLENE
0.08
0.05
0.025
0.06
0.065
0.03
ON INDEX
METHANOL
0. 17
0.05
0.09
0.11
0.09
0.05
u a
§ O
a
10
I.I
0.9
0.8
>
^
<^
i
k 1
^^
!
^
-.
"
,
^
P
t
1 234 | 234
CYLINDER NUMBER CYLINDER NUMBER 1
2000 RPM 3000 RPM
< Q
3
o o;
x m
1.3
1.2
0.9
0.8
1234
CYLINDER NUMBER
2000 RPM
234
CYLINDER NUMBER
3000 RPM
INDOLENE FUEL
METHANOL FUEL
FIGURE ET.I3 : CYLINDER TO CYLINDER VARIATIONS IN EQUIVALENCE RATIO - METHANOL VERSUS INDOLENE
-------
48
Effect on Power and Thermal Efficiency; In Figure IV.13 the
average MI for the range at each speed and load corresponding to
the graphical evidence appears in tabulated form for both fuels.
The values of MI indicate that at the same speed and load methanol
is showing a higher degree of maldistribution in the OEM intake
manifold/carburetor combination. Figure IV.14, illustrates the
effects that this maldistribution has on engine operating variables
and outputs. Here, ignition timing, indicated power, fuel flow rate,
air flow rate, and the maldistribution appear as functions of $ (average)
The maldistribution, which appears at the bottom of the figure, has
been plotted on a vertical axis instead of a horizontal axis. The
cylinder numbers appear by each point. In this particular set of
seven steady state data points, four of the points show severe
maldistribution (MI = 0.20 to 0.30). With severe maldistribution,
the MBT spark timing appears overadvanced, and the power is de-
creased dramatically relative to the points of little maldistribution.
Power is down by 11% at * = 0.9 and 9% at $ = 1.0. Since the fuel
flow rate is not effected by the maldistribution, the power drop can
be translated directly into a thermal efficiency decrease of 11% and
9% respectively. This decrease in power (and thermal efficiency) at
a given average * can be attributed to the cylinders that are operat-
ing lean combined with non-MBT spark timing that exists for all four
cylinders. The contributions from these two factors are illustrated
in Figure IV.15. In this figure, engine power for a no maldistribu-
tion case is plotted as a function of $ (highest curve). Assuming
that MBT conditions are maintained for each cylinder, maldistribution
effects can be calculated by assuming some form of maldistribution
among the cylinders. Two cases are plotted which represent MI = 0.10
and MI = 0.20. It is assumed that the values of * for each cylinder
deviate linearly from* average. For the assumed symmetrical
maldistribution, the calculated percentage decrease in power
(MBT-each cylinder) is tabulated along with the observed percentage
decrease in power (non MBT-each cylinder). The differences between
calculated and observed values indicates that the additional deficit
due to non-MBT conditions more than doubles the deficit due to
-------
49
45
u
o
CD 35
e
or
8.
V)
25
7.0
CD
£d 6 0
< 6.0
O _
1.4
1.0
0.6
MBT SPARK TIMING
PINTO 2300 CC ENGINE
METHANOL FUEL
0.60 0.70 0.80 0.90
1.00
$
1.10
1.30
I
a.
in
V)
cc
o
u
Q
1.40
1.20
1.00
0.80
03
1.40
FIGURE IZ.I4: EFFECTS OF MALDISTRIBUTION ON SPARK TIMING, POWER, AIR FLOW
RATE, AND FUEL FLOW RATE
-------
PINTO 2300 CC ENGINE
2500 RPM- 7" HG MANVAC
FUEL: METHANOL
EFI
0.70
0.80
0.90
% DECREASE IN POWER AND
THERMAL EFFICIENCY
M 1 *
.10
.10
.10
.10
.20
.20
.20
.20
I
.60
.90
1.00
1.10
.80
.90
1.00
1.10
CALC
AT MBT
0.5
0.8
1.2
0.7
3.0
2.5
ACTUAL
4.5
3-4
II**
6.V*
X- MALDISTRIBUTION INDEX (Ml)
DEFINED FOR A SYMMETRICAL
MALDISTRIBUTION
FOR Ml - .10, $CYL/$AVG VALUES
ARE 1.10, 1.033, .967, AND .900
FOR Ml :.20,$CYL/$AVG VALUES
ARE 1.20, 1.067, .934, AND .600
XX-DATA FROM FIG. 13
*
1°
b
01
*
b
Ul
*
b
or
UJ
I
Q.
UJ
CO
g
O
UJ
<
O
O
30.0 ?
1.00
I.K)
FIGURE JZ-15: EFFECTS OF MALDISTRIBUTION
CALCULATED VS ACTUAL
1.20 1.30
ON POWER -
en
O
-------
51
maldistribution alone. Other maldistribution variations can be
assumed and superimposed on the graph; however, the symmetric case
presented here appears to closely approximate the actual maldistribu-
tion seen with the OEM-Indolene, OEM-methanol, and Dresserator systems.
Thus, the table in Figure IV.I5 is used in this report to estimate
the expected gains in fuel economy if maldistribution is eliminated
in the OEM (Indolene and methanol) and Dresserator systems. For
average * values leaner than 0.8 no estimate of expected improvement
in thermal efficiency is attempted since the leanest cylinder is
usually undergoing intermittent misfire.
Effect on Exhaust Emission: The maldistribution index as defined
appears from a qualitative standpoint to be somewhat unsatisfactory
in explaining trends in average exhaust emission level. The follow-
ing discussion is somewhat academic, since measurements from the
individual cylinders were utilized in displaying emission character-
istics; nevertheless, the discussion is important in trying to in-
terpret the simulation of the FTP (Section IV.3.) relative to actual
exhaust emissions measured in the FTP. The importance lies in the
fact that FTP emissions are based on the mixed or average levels of
emissions which are collected at the exhaust tailpipe (i.e. the effects
of maldistribution are included). CO and UBF trends with maldistribu-
tion can be explained through calculations similar to those utilized
to generate the power curve (Figure IV.I5.). The two figures below
illustrate in a qualitative way the expected effects.
«—LEAN
INCREASING
MALDISTRIBUTION
RICH—
O
o
— LEAN
INCREASING
MALDISTRIBUTION
RICH—
-------
52
The NOX emissions have a highly non-linear character when plotted
against the cylinder * as is indicated in Figure IV. 8 and IV. 9.
Average values of NOX when maldistribution is present produces
results such as seen in Figure IV. 16. Here is an illustration of
actual test data based on the measurement of the average (a composite)
NOX for methanol and indolene at 2000 RPM and WOT. The peak of
Figure IV. 8 for the OEM-Indolene now has two maxima due to the
maldistribution. The OEM-methanol data which had a mean MI value
almost double that of the indolene shows a more divergent character
of the two maxima. The fact that smooth curves could be drawn for
the two cases indicates that the maldistribution symmetry remained
fairly constant for the range of * investigated. Figure IV. 17
predicts the methanol results of Figure IV. 16. based on calculated
mean values of NOX for an assumed constant maldistribution which is
indicated. This maldistribution approximates the WOT methanol mal-
distribution at 2000 RPM (see Figure IV. 13.). Any number of maldistri-
bution combinations could be plotted and thus the random nature of a
four cylinder average NOX plot versus * can be anticipated.
Summary: Maldistribution of the fuel-air mixture among the cylinders
of this four cylinder engine appears to effect power and thermal
efficiency in a predictable manner except at very lean operating condi-
tions. The maldistribution index (MI) provides a means to quantify
the effect for average $ >_ 0.8 while operating on methanol. The
exhaust emissions can't be accurately predicted for the simple defini-
tion of maldistribution used; however, by assuming a constant maldis-
tribution symmetry average exhaust emission curves can be generated
from the single cylinder measurements made.
The variations in exhaust emissions (especially NOX) that can be seen
with varying degrees of maldistribution provides the rationale for using
MI = 0.0 emission results for predicting emissions in the simulated FTP
and HFETP driving cycles (see Section IV.3.). Though few automotive
engines operate at steady state conditions with no maldistribution,
calculations on the basis of MI = 0.0 provide the boundary conditions
which real engines may approach. Through design of fuel-air metering
systems which yield virtually no maldistribution (such as the WHB system)
-------
53
16
15
14
13
12
II
I 10
OL
£ 9
x 4
* 3
/
/
/
/-^
/ '
/
f
/ "%
/
/
k J
^ S
\
^Ml -O.I7
\
l-»v
^\MI -0.08
\
\
\
>
^^ "
\
\_--
'"\
^^
2000 RPM - WOT
4 CYLINDER 2300CC ENGINE
MBT SPARK TIMING
ME
iMn
.
r^^
FHANOL
OLENE
0.70
0-80
0.90
1.10
1.20
1.00
BCOMP
FIGURE IE. 16: EFFECT OF MALDISTRIBUTION ON NOX~ EXPERIMENTAL DATA
1.30
1.40
IS
co
z
o
CO
co
CO
X
o
10
SINGLE CYL.
COMPOSITE DATA
WITH MMJSTRIBUTI'
ILLUSTRATED
2000 RPM
WOT
CYLINDER NUMBER
(
234
I
/
/
/
/
/
/
/
/
/
/
\
\
\
i
\
\
\
i
t
i
0.8
0.9
1X30
$
I.I
1.2
FIGURE ET.I7 CALCULATED MALDISTRIBUTION EFFECT ON NOX
FOR A CONSTANT MALDISTRIBUTION SYMMETRY
-------
these predicted emission levels may be attained eventually.
54
IV.2. Alternative Fuel Induction Systems: From the investigations with
the modified OEM carburetor several areas of improvement are
needed to design the optimum fuel-air induction system for methanol
operation. One of the most significant of these areas is the mal-
distribution of air and fuel among the cylinders. The consequences
are loss of performance 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.
In the investigation of alternative air-fuel induction systems, one
must incorporate gains in thermal efficiency and reduced exhaust
emissions with the realities of the production automobile. These
factors include economics as well as an ability of the system to adapt
to present day emission controls such as PCV and EGR. The best air fuel
induction system, thus, must be economic as well as having precise control
of air-fuel ratio through better metering, nebulization and mixing.
IV.2.1. WHB System: 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, initiat-
ing 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.
-------
55
Subsequent wave reflections are prevented by 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 max-
imize the effectiveness of this pressure gradient in atomizing the fuel
issuing from the orifice. Visual examination indicates mean fuel droplet
diameters of the order of 10 microns were produced by a CFR engine at
900 RPM 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 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 manifold as in a conventional system because no
significant changes in mixture flow direction occur along the path.
WHB System Performance: Due to the relative simplicity of the system and
its ability to essentially remove maldistribution among the cylinders, the
WHB induction system shows promise as a superior fuel-air inductor for
methanol operations. Even though the WHB system requires an inductor for
each cylinder, its simplicity places it economically competitive with the
OEM system.
In order to obtain low emissions and lean operation, close control of
air-fuel ratio is required. In Figure IV.18, the variation of $ with per-
centage throttle opening is shown with engine speed as a parameter.
Cylinder maldistribution at 2000 and 3500 RPM are also shown. It is
clear that there is negligible influence of speed on the equivalance ratio.
Hence this fuel preparation system has acheived the design objective of
holding a balanced air-fuel ratio among the cylinders over the desired
speed range. However, the excurisions in air-fuel ratio from 0.8 to 1.3
with throttle opening are excessive for good performance and emissions
control.
-------
1.30
1.20
1.10
I.OO
0.90
o.eo
0.70
1.05
1.00
0.95
n
4- 4
|\
\
\
7
1 VI
9'
L
X
/'
J
1
X j
_„-
NO IND
IN THIS
-"-"
VIDUAL CY
RANGE 0
. —
UNDER DA
F THROTTL
FA AVAILAE
.E POSITIO
— -^ _^
«
"^ "^
-•^..^
PINTO 2300 CC ENGINE
MBT SPARK TIMING
640 RPM
+ + ++ 1000 RPM
2000 RPM
3000 RPM
3500 RPM
3LE
NS
2000 8 3500 RPM
MALDISTRIBUTION -
%
\
t
EXPANDED SCALE
20 30 40 50 60 70 80
PERCENT AREA OF WIDE OPEN THROTTLE
90
100
C71
CFi
FIGURE 1ST. 18: EQUIVALENCE RATIO VS PERCENT OF WIDE OPEN THROTTLE - WHB INDUCTION SYSTEM
-------
57
Due to the reduction of flow resistance of the WHB system com-
pared to the OEM carburetor manifold system, coupled with the
excessive inlet valve duration (276°), excessive air flow reversal
was encountered. Figure IV. 19 shows the mass flow of air
inducted versus percentage of throttle port opening. Note that the
maximum amount of air inducted occurs at the 70% point for 2000 RPM.
In the region between 70% and WOT (100%), the amount of inducted air
declines. This is due to the excessive open duration of the intake
valve which allows time for reversed flow to occur. This reversed
air flow carries some fuel past the carburetor inlet resulting in
significant sputter. Evidence like Figure IV.19 generated at higher
engine speeds shows that the OEM camshaft was designed for high speed
operation. Consequently, a new camshaft is being designed with less
intake valve open time for better operation at urban driving speeds.
Testing of the WHB system will resume with the fuel circuit modifications
and the new camshaft.
IV.2.2.Electronic Fuel Injection (EFI) System: The electronic fuel injection
system employed for these tests is shown in Figure IV.20. Basically
it consists of Bosch electronic fuel injectors and an analog control
unit designed by personnel at the ERDA Lawrence Livermore Laboratory.
Referring to Figure IV.20, there is one injector per cylinder, mounted
to spray fuel approximately at the backside of the inlet valve. The
injectors are supplied with fuel from a continuous return, pressure
regulated system. The fuel pressure regulator is referenced to mani-
fold pressure so that a constant pressure differential of 28 psi is
maintained across the injectors regardless of the manifold pressure.
The injectors operate either fully open or closed. The opening and
closing times of the injectors have been minimized by the design of the
electronic circuit driving them, The controller determines how long the
injectors are to remain open based on inputs from a manifold pressure.
transducer, rpm signal and inlet air temperature.
The metering circuitry can be adjusted to yield a wide variation of
equivalence ratios. The injection pulse width is typically four to
fifteen milliseconds in going from idle to WOT at 3000 rpm. It is
-------
7.0
6.0
5.0
4.0
CD
3.0
2.0
1.0
THEORETIC
THEORETIC
AL NO LOS
AL NO LOS
/
/,
/
5 VALUE FC
/
£ VALUE F(
/"
R WOT- ZCK
>* \
IR WOT -IOC
10 RPM
^-
0 RPM
PINTO 2900 CC ENGINE
MBT SPARK TIMIN6
+ -H- + 1000 RPM
2000 RPM
20 40 60 80
PERCENT OF WOT
100
FIGURE ET.I9: AIR FLOW RATE VERSUS PERCENT OF WIDE OPEN
THROTTLE - WHB INDUCTION SYSTEM
00
-------
FUEL TANK
FUEL
PRESSURE
REGULATOR
VIBRATION
DAMPER * 1
COLD
START
VALVE
INJECTORS
I EACH CYLINDER
AIR TEMP
SENSOR
THROTTLE
BODY
INTAKE
MANIFOLD
PRESSURE
WATER
TEMP
SENSOR
ENGINE SPEED
FROM TRIGGER
IGNITION SWITCH
en
v£>.
ELECTRONIC CONTROL UNIT
FIG ET. 20 ELECTRONIC FUEL INJECTION SYSTEM
-------
60
possible to manually adjust the starting point of the injection
pulse, and the individual cylinder pulse widths. The manifold for
the EFI system was designed and built in our laboratory. The primary
considerations in design were to achieve uniform air flow among the
cylinders and to locate the injectors close to and pointed at the
backside of the inlet valves in order to minimize wall wetting.
It was felt that four equal length runners connected to a common
plenum would provide uniform air flow to each cylinder. The runners
were approximately nine inches long. The location of the injectors
became problematic in that intake port entry angle for the inboard and
outboard cylinders is different and the injectors are not equally directed
at the valves. The EFI test program encountered several problems.
Ignition noise from both the secondary and primary circuits was a per-
sistent source of intermittent trouble. Effective safeguards against
the noise were only found near the end of the reporting period.
Although measures were taken to provide uniform distribution of air and
fuel in the manifold design, it was found necessary to adjust the indi-
vidual cylinder injection times to balance the equivalance ratio in all
cylinders. Due to the sensitivity of the trim potentiometer, time delay
in the emissions instrumentation and interactions of engine variables,
this balancing operation became a serious time sink requiring up to
three quarters of an hour to set a given engine condition. Due to these
difficulties, the test matrix was limited to mostly 2000 rpm.
The EFI system is unique in that the point of fuel injection can be
varied. This provides differing cases of fuel preparation, i.e., per-
cent of fuel vaporized and/or fuel-air stratification.
Figure IV.21. demonstrates the effect of the starting point of injection
on the levels of unburned fuel, CO, and formaldehyde. Point of injec-
tion is measured as crankshaft degrees before top dead center of the
induction stroke. Note the intake valve effective opening is at 23°BTDC.
Exhaust equivalence ratio is stated on the graph and is seen to vary
slightly. Power and indicated thermal efficiency remain nearly constant.
Clearly, the effect on emissions is due to mixture preparation.
-------
51
52.0
INDICATED
POWER 51-0
a~
-i
>
— a-
THERMAL
EFFICIENCY
PPM
.350
.300
2000
1800
1600
1400
1200
1000
800
600
400
200
-G-
2000 RPM- WOT
FUEL METHANOL
34° SPARK
TINLET = 77 °F
20° 60° 100° 140° 180° 220° 260° 300°
TIMING AT START OF INJECTION-°BTDC OF INDUCTION STROKE
FIGURE BL2I: EFFECTS OF INJECTION TIMING ON POWER, THERMAL EFFICIENCY 8 EMISSIONS
-------
62
Matthes and McGill (Ref, 6) and others have suggested that non-
homogeneity in the cylinder can shift the knee of the CO curve.
The nonuniform charge will shift the CO versus equivalence ratio
curve toward the lean side. The CO emissions may be explained by this
mechanism. Injecting fuel into the manifold before the valve is open
allows time for vaporization and a more homogeneous charge to be
inducted. On the other hand, injection at or near the time of valve
opening leads to a stratified charge of liquid droplets. Impingment
of the drops against the cylinder wall would lead to an excessively
rich quench zone resulting in the high unburned fuel and formaldehyde
emissions which were observed.
Based on these observations, the majority of EFI data was acquired with
the injection point set @ 90° BTDC of the induction stroke. Although
this system is tedious to operate, it has produced some informative
data as discussed in Section IV.1. It is planned that data will continue
to be generated with this system in subsequent phases of the program.
IV.2.3. Dresserator Inductor: The Dresserator Type III prototype is a third
generation experimental carburetor built by Dresser Industries. It
uses a sliding venturi to maintain sonic conditions at the throat of the
carburetor for a wide range of manifold vacuums. The fuel is delivered
from a slotted bar which is mounted directly above the venturi and spans
the entire length of the carburetor's cross-sectional area. The choked
condition created at the throat of the carburetor and the subsequent
shock wave are used to finely atomize the fuel spray into a non-wettable
fog, allowing a more homogeneous mixture of fuel and air to enter the
engine cylinder. This well atomized mixture reduces cycle to cycle
variations in peak cylinder pressures as well as reducing the droplet
fallout as the mixture passes through the intake manifold.
Another objective of the Dresserator is to maintain a constant air-fuel
ratio. This is accomplished by maintaining choked conditions in the
throat of the carburetor through the reduction in throat area. Thus
the use of the sliding venturi to maintain sonic conditions at the
throat of the carburetor not only aids in fuel atomization but is also
intended to hold the air and fuel flow conditions constant. Although
the model tested did not totally fulfill this objective, it could be
adjusted for the desired values of *.
-------
63
The Dresserator carburetor is shown in Figure IV.22a. It should be
noted that the prototype model tested is experimental and has been
designed to operate only at steady state. It therefore contains no
acceleration enrichment circuit found on conventional carburetors.
Adaptation to the Test Engine: In order to best adapt this carburetor
system to the Ford Pinto 2300 cc 4 cylinder engine, a special sliding
intake manifold adapter was built. This allowed relative movement of
the carburetor in a direction normal to the centerline of the engine
between the lower portion called the base and the upper portion called
the slider. A descriptive drawing of this adaptor is shown in Fibure IV.22b.
Tests were run at a variety of engine speeds and loads at seven different
slider positions (0-6cm) to determine the best carburetor position for
minimum cylinder to cylinder maldistribution. The effects of two
different carburetor fuel distribution bars, the 8 slot and the 4 slot
methanol fuel bars, were also examined to determine the optimum carburetor
position and fuel bar combination for steady state methanol operation. As
a result of these tests, the 8 slot fuel bar with the carburetor at the
5 cm slider location were chosen due to the superior results indicated on
methanol fuel at the 2000 rpm, 1/3 and full brake load conditions. Sub-
sequent tests of the Dresserator system were conducted with this fuel
bar-slider position combination.
Dresserator Performance: The most outstanding feature of the Dresserator
induction system is the improved thermal efficiency due to reduced fuel
flow rate needed for a given speed and load setting. It shows the best
fuel economy for the simulated CVS driving cycle of all the air-fuel
induction systems tested. (See Section IV.3.). The thermal efficiency
curves shown in Figure IV.23. indicate peak thermal efficiencies in the
range of 0.78 _< * <_ 0.85. After this point, however, the dropoff in
thermal efficiency is quite rapid, a characteristic not seen in other
air-fuel induction systems. Further investigation is needed to determine
if the effect of maldistribution and resulting non-MBT spark setting or
mixture preparation is responsible for this phenomenon.
-------
64
INTAKE AIR SUPPLY
FUEL DELIVERY POINT
ATOMIZED A/F MIXTURE
TO INTAKE MANIFOLD
FIGURE 12.22 A: CONCEPTUALIZED DIAGRAM OF THE
DRESSERATOR INDUCTION SYSTEM
-------
65
\
v-t
SYY.
DRESSERATOR
m~— SLIDER
BASE
MANIFOLD
FIGURE ET.22B: MANIFOLD ADAPTER BRACKET
-------
PINTO 2300 CC ENGINE
WOT MBT SPARK TIMING
OEM- INOOLENE
DRESSERATOR - MEOH
0,50 0.60 0.70 0.80 0.90 1.00 1.10
24
1.20 1.30 1.40
en
FIGURE BT.23: COMPARATIVE THERMAL EFFICIENCY ACROSS THE SPEED RANGE AT WIDE
VS VETH,/WO'_ (Hf^OH)
-------
67
At WOT operation particularly at lower speeds, severe standoff of fuel
droplets was experienced. This was caused by excessive intake valve
open time causing reversed flow in the induction system at manifold
vacuums of less than 3 inches Hg. By advancing the camshaft 5° and
installing exhaust headers, this standoff problem was reduced. Redesign
of the camshaft would tend to minimize this problem and this is under way.
While the Dresserator was not optimally designed for operation on methanol,
it did show thermal efficiency increases over the other systems tested.
With some necessary redesign of the intake manifold, camshaft and fuel
delivery bar, the Dresserator shows promise for good operation in a
methanol fueled automobile.
IV.3. Methanol versus Indolene - A Comparison Based on Simulation of the Federal
Emission Test Procedure and the Federal Highway Fuel Economy Test Procedure
IV.3.1. Introduction: Predictions of fuel economy and exhaust emissions for transient
engine behavior using steady state engine maps of fuel economy and emissions
trends have been made possible in recent years through the use of computer
simulation programs. The Federal Emission Test Procedure (FTP) and Federal
Highway Fuel Economy Test Procedure (HFETP) are prescribed urban and high-
way driving cycles which can be simulated with the computer. The
versatility of these programs is found in their ability to predict fuel
economy, emissions and performance for a wide range of engine operating
variables and vehicle hardware variables.
In the simulation presented in this section, use was made of a program
developed by the Jet Propulsion Laboratory for the prediction of fuel
economy and emissions from the steady state engine maps of performance
and emissions. Results for a Ford Pinto utilizing the OEM system on
Indolene, the modified OEM system on methanol, and the alternate systems
(with the exception of EFI) on methanol are presented.
IV.3.2. Simulation Results: Table IV.3. indicates the simulation cases of the
FTP and HFETP that were run. Each of the fuel air induction systems was
run for a number of fixed equivalence ratios with the exception of the
WHB system. (This latter system had a varying equivalence ratio.) Since
-------
68
the input data for their simulations was based on fully warmed up steady
state engine data, the results represent a hot start for the 1972 FTP.
Table IV.3. - Simulation Cases
Run# Fuel Equivalence Ratio ($) Carburetion-Fuel/Remarks*
1 Indolene 1.0 OEM Baseline (gasoline)
2 " 0.9 ....
3 " 0.8
4 Methanol 1.0 Modified OEM Baseline (Methanol)
5 " 0.9
6 " 0.8
7 " 0.7
8 " 1.0
9 " 0.9
10 " 0.8
11 " 0.7
12 " variable
13
*No E6R throughout
"very lean
operation
Dresserator Inductor (Methanol)
WHB
"very lean
operation
induction system (Methanol)640 rpm
900 rpm
Runs 1 through 7 provide comparative results for the OEM-Indolene and
OEM-methanol configurations. Of the alternate fuel-air induction systems
chosen for investigation, only the Dresserator and WHB systems appear.
There was insufficient steady state data for the EFI to permit a simulation.
Fuel Economy: Figure IV.24. displays the fuel economy on an energy basis
(miles traveled per million BTU's of fuel consumed) for the FTP or urban
driving cycle as a function of $. All 13 simulations are displayed with
their characteristic time averaged maldistribution indices. The WHB
maldistribution was virtually zero. Table IV.4 summarizes the fuel
economy as percentage gains over Indolene on an energy basis. Three
general points, which are characteristic of methanol and SI engines
become clearly evident from the figure and table:
1. All fuel preparation systems operating on methanol show marked
combustion efficiency improvements in comparison to Indolene.
-------
D
CD
<0
O
CO
UJ
_J
270
250
230
210
"- 190
O
O
UJ
U
170
150
130
f n
l.U
t
<
(.0
>
ra u
^—- —
r4) *
(.0
c
(.0
>74) ( (
laf -^
(.0
^
LEAN
65)
t.02)^DJj
57J\\
^ (l
8)^^
>!
(.
RICH
061)
>
>
D56)
)
>ie)
URBAN DRIVING SIMULATION
FORD
(_ )= M
$ = Tto
© - (
© - (
A -
0 -
• -
I FTP)
PINTO 2300 CC
^DISTRIBUTION INDEX
HE AVERAGED $ FOR WHB
DEM INDOLENE
DEM METHANOL
DRESSERATOR
WHB 640 RPM IDLE -$=0-98
UVHB 900 RPM IDLE -1=0.98
0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50
I
FIGURE ET.24: FUEL ECONOMY VS EQUIVALENCE RATIO FOR ALTERNATE
FUEL-AIR INDUCTION SYSTEMS ON METHANOL WITH THE
INDOLENE AND METHANOL OEM RESULTS - URBAN DRIVING
en
-------
70
2. All systems including the OEM-Indolene show improved combustion
efficiency as the fuel-air mixture is leaned.
3. Significant additional gains in thermal efficiency are possible
if maldistribution is eliminated.
Table IV.4. Simulated FTP Results
Equivalence % Increase in Fuel Economy Relative to OEM-Indolene*
Ratio
4>
0.7
0.8
0.9
1.0
OEM-Methanol
22
20 (23)
16 (19)
8 (11)
Dresserator
26
*** 29 (33)
28 (32)
18 (22)
WHB**
-
-
-
19
*OEM-Indolene results at * = 1.0
**time averaged $ for WHB was 0.98 (640 RPM idle case)
***numbers in parenthesis indicate expected improvements if mal-
distribution of * among cylinders was eliminated.
The Dresserator shows the highest gain (29% @ $ = 0.80) when compared
with the OEM-Indolene results at * = 1.0 (the estimated operating * of
an actual 1975 "California" Ford Pinto). Even in comparison to the best
Indolene results (at $ = 0.80) the Dresserator still shows a 20%
advantage in fuel economy. If maldistribution is eliminated, than an
additional gain in fuel economy of from 5 to 7 mi/106 BTU can be expected.
(The basis for these estimates can be found in Section IV. 1.3.). The
estimated percentage improvements for the various systems are the
quantities in parenthesis in Table IV.4. Expected improvements at * = 0.7
are not displayed for want of sufficient data to provide the estimates.
The 33% estimated improvement for the Dresserator emphasizes the influence
of maldistribution.
The WHB system displayed no maldistribution. Hence, no gains are
expected in this sense. However, its ability to support a lower idle
speed is clearly beneficial. In cases 12 versus 13, the reduced idle
speed provided a 40% savings in idle fuel consumption and a corresponding
increase of 5 mi/106 BTU in FTP fuel economy, Figure IV.4. Results for
-------
71
highway driving (HFETP) in Figure IV.25. show similar trends to the
FTP results. A summary of these results is presented in Table IV.5,
Table IV.5. Simulated HFETP Results
Equivalence % Increase in Fuel Economy Relative to OEM-Indolene*
Ratio
* OEM-Methanol Dresserator WHB**
0.7
0.8
0.9
1.0
22
19 (23)***
16 (20)
10 (10)
25
28 (35)
25 (30)
18 (22)
-
-
20
-
*OEM-Indolene results at 4> = 1.0
**time averaged * for WHB was 0.87
***estimate for no maldistribution
Improvements up to 28% are seen over the OEM-Indolene results. At
* = 0..9 the WHB system shows a 20% improvement. The values in parentheses
are estimated values of improvement in fuel economy if maldistribution is
eliminated from the OEM-methanol and Dresserator systems. At * = 0.80
for the Dresserator we could expect to see an additional improvement of
5% or a value of 238 mi/106 BTU. This is a 35% improvement over the
OEM-Indolene results at * = 1.0 or a 25% improvement for the OEM-Indolene
at its effective lean operating limit.
Figures IV.24. and IV.25. indicate, with the exception of the Dresserator
data, that the peak fuel economy appears at the leanest equivalence ratio.
This does not seem consistent with the steady state source data some of
which is shown in Figures IV.4. and IV.5. However, increasingly lean
operation requires lower manifold vacuums for the same speeds and loads
that are dictated by the FTP. With the exception of the Dresserator,
the increased efficiency due to lower manifold vacuums more than offsets
the decrease in efficiency due to leaning the mixture as anticipated
from Figures IV.4 and IV.5.
-------
00
o
o
UJ
jji
UJ
u.
I
250
230
210
2 190
N.
CO
UJ
£ 170
ISO
130
0.6
(.09*
*
<
{,01
r
»
(.1!
i-"~~U
M) *
(.0!
t
(.03
3)
>—^U
r?
6) ^^<
(.oe
w^<
(.02
LEM
>73)
"X,
^x
(.0
^^^
(.c
RICH
166)
»
»7)
1
51)
HIGHWAY DRIVING SIMULATION
FORD
1 )B|
$ • Tl
o -
0 -
A -
0 -
• -
(HFETP)
PINTO 2300 CC
MALDISTRIBUTION INDEX
ME AVERAGED <£ FOR WHB
OEM INDOLENE
OEM METHANOL
DRESSERATOR
WHB 640 RPM IDLE -f =.87
WHB 900 RPM IDLE-1>=.87
0.7
0.8
0.9
1.0
I.I
1.2
1.3
1.4
1.5
FIGURE T5T. 25: FUEL ECONOMY VS EQUIVALENCE RATIO FOR ALTERNATE
FUEL-AIR INDUCTION SYSTEMS ON METHANOL WITH THE INDOLENE
AND METHANOL OEM RESULTS - HIGHWAY DRIVING
ro
-------
73
In summary, both FTP and HFETP simulation results predict fuel economies
for Indolene that are within the range of actual Pinto vehicle performance.
Methanol yields increased fuel economy on an energy basis over Indolene.
The alternate fuel-air induction systems operating on methanol (Dresserator
and WHB) yield additional gains. Lean operation is not only possible but
desirable from a fuel economy viewpoint. Elimination of maldistribution
will simultaneously further the lean operating equivalence ratio for both
methanol and gasoline and raise the fuel economy. Because methanol has a
lean burn advantage and a higher energy conversion efficiency than gasoline,
these effects can significantly help offset the fact that it only has about
half the energy per unit volume as gasoline.
Driveability in the FTP: Driveability or the lack of it is an all important
factor in operating any automobile in the lean region. Actual vehicle
driveability quality is difficult to reproduce from computer simulation of
the Federal Test Procedure. Poor driveability can be caused in part by
erratic air-fuel mixture preparation and auto ignition. It can also be
caused by inadequate power. These conditions are usually encountered at
very lean operation. Our computer simulation assumes a very accurate
control on mixture which in turn allows MBT spark setting for all speeds
and loads. This eliminates auto ignition and misfire as a driveability
constraint. However, the computer simulation does recognize inadequate
power. In all the simulated cases power requirements were met, with the
exception of a short time interval for $ = 0.7.
Exhaust Emission Comparisons: Figure IV.26 shows the comparative emissions
evidence for all the simulation cases in urban driving (the FTP). The
results indicate that by changing fuels, the modified OEM system on
methanol yields improvements in all three Federally regulated exhaust
emission species. The most significant result displayed is that the
statutory NOX standard of .40 gms/mi can be met by operating near
* = 0.7. The alternative systems operating on methanol yield results
similar to the modified OEM equipment with a few exceptions which are
discussed in the following paragraphs.
-------
74
UJ
OT
.
m
3.0
2.0
1. 0
0.0
0.6 0.7 0.8 0.9 1.0
32.0
28.0
j 24.0
2
>x
CO
0 20.0
CO
UJ
CO 12.0
8 8.0
4.0
0.0
I.I
EAN
(3.4)
RICH
FEDERAL
-§V TUTORY
0.6 0.7 0.8 0.9 1.0 I.I
I
1.2
6.0
0.0
URBAN DRIVING SIMULATION (FTP)
FORD PINTO - 2300 CC
.MALDISTRIBUTION INDEX * 0.0
f = TIME AVERAGE $ FOR WHB
$ CONSTANT FOR ALL OTHER SYSTEMS
O - OEM INDOLENE
0 - OEM METHANOL
A - DRESSERATOR METHANOL
Q - WHB METHANOL
0.6 a?
0.8
0.9 1.0 I.I 1.2
FIGURE ET.26: EXHAUST EMISSIONS VS EQUIVALENCE RATIO FOR OEM AND ALTERNATE
FUEL-AIR INDUCTION SYSTEMS FOR THE HOT 1972 FTP SIMULATION
-------
75
NOX results show that at similar equivalence ratios methanol yields
roughly half the gms/mi that Indolene yields. The urban cycle, time
averaged, value of $ = 0.98 for the WHB system represents a range of
values from 0.82 at idle to 1.26 at higher speed and load conditions.
The meeting of the statutory NOX standard near * = 0.7 exploits
methanoTs lean operation ability and offers an alternative to the
state of the art NOX control strategies of EGR which penalizes fuel
economy or, a reducing catalyst which requires operating near
stoichimetric 4> also penalizing fuel economy.
Unburned fuel and carbon monoxide results show comparable behavior
for both Indolene and methanol (with or without an alternative fuel-
air induction system). CO emissions show that the statutory standard
is met by operating at * = 0.9 or leaner equivalence ratios on either
Indolene or methanol. The CO results for WHB at * = .98 are predictable
since much of the time the engine is operating at rich *. All the
systems show similar UBF trends, that is, increasing values as *
approaches the lean limit. The Dresserator shows the highest UBF of
all the systems at the lean limit, and all the systems are showing
emissions greater than the Federal statutory standard of 0.41 gms/mi.
It should be kept in mind, however, that the results presented here are
based on exhaust samples taken near the exhaust valve, and thus oxida-
tion of the UBF in the exhaust manifold has not been taken into account.
On the basis of measurements downstream of the exhaust manifold, the
steady state data would indicate that values of UBF would be lower by
a factor of two at the tailpipe.
Even though aldehydes were not calculated in this simulation, they are
expected to follow the general behavior of the steady state data just
as CO, NOX and UBF did. We should expect to see higher values of
exhaust aldehydes on methanol, and these should increase as leaner *
values are approached. But, again, the levels are relatively small in
comparison to the other pollutant emissions.
-------
76
Summary: Methanol shows superior performance to Indolene in terms of
fuel economy across the $ range investigated in these simulations of
urban and highway driving. There is additional benefit from utilizing
alternate air-fuel induction systems. The Dresserator and WHB system
show the best results, and the Dresserator may perform even better if
the maldistribution associated with it can be eliminated.
The simulation program is a convenient tool for calculating comparative
evidence for the FTP and HFETP. It is known that the match up of
simulated and actual emissions is generally poor, but the approach used
here in the simulation provides the limiting cases in emission behavior.
That is, if * were maintained constant in the engine, and maldistribu-
tion was eliminated, then these results would be attained. Cold start-
ing effects need to be incorporated into future simulations to adequate-
ly simulate CO and UBF behavior; however, the NOX simulation results
are representative of both hot and cold starts. This implies that,
upon elimination of maldistribution, the NOX standard of 0.4 gm/mi will
be met in an FTP cold-start simulation of lean engine operational on
methanol. This interesting possible alternative to EGR or reduction
catalyst control of NOX is appealing because it exploits methanol's
superior combustion properties relative to gasoline.
Further study of this computer generated evidence is contemplated. It
will include EGR influence and lean operation using a Pinto vehicle
powered by methanol and operating on the FTP cycle as well as steady
state operation of dynamometer mounted engines. The intent is to find
the best combination of fuel preparation system and equivalence ratios
for methanol fuel vehicles.
-------
77
IV.4. Thermochemical Engine Process Modeling: Thermochemical modeling
of SI engine processes is important to help provide an explanation and
hence indicate methods of improvement in engine performance and reduc-
tion in pollutant emissions produced by the experimental Pinto test
engine and found in the literature. This modeling is being accomplished
by the use of thermodynamic and chemical kinetic computer simulation
which has been described in a previous report.3
IV.4.1.Prediction of NOX vs. C.R.: One such- trend that has been the
subject of much controversy over the last two years has been
the compression ratio effect on NOX emissions for methanol
fueled SI engines. This controversy came about over engine
experiments conducted by W. E. Bernhardt^ of Volkswagen. In
these tests, he found that the volumetric NOX emissions de-
creased with higher compression ratios in methanol fueled SI
engines. F. Pischinger^ subsequently reported similar results.
These results differed from the normal logic in which com-
pression ratio increases cause peak cylinder temperature in-
creases and thus increases in the production of NOX. Neither,
however, offered an explanation for their deviation. It was
obviously an interesting question which directly related to
our computer studies of compression ratio effects and was
incorporated in these studies.
In order to match the computer simulation to actual engine data,
cylinder pressure-time traces were obtained from the Pinto test
engine. Cylinder number one of this engine was instrumented with
a model 601B Kistler pressure transducer mounted in a model 640
spark plug adaptor and connected to a model 504E charge amplifier
and a Tektronix model 5011 storage osci Hi scope.
Even with the forced air cooling of the 601B pressure transducer,
temperature instabilities resulted in signal drift and, therefore,
made it impossible to store more than one cycle on the oscilliscope
-------
78
screen. This problem was overcome by using the AC coupling mode on
the oscilliscope; however, this led to distortion of the signal trace
as discussed by Lancaster et al. 9 in order to retrieve the original
data from the distorted trace, an inverse Laplace transform was used.
The AC coupling mode on the oscilliscope can be described as a high
pass filter with a corner frequency (fc) of 1.6 hertz. The Laplace
transform describing a high pass filter is as follows:
P'(t) = L (p(t))
where p(t) is the actual pressure trace
P'(t) is the resulting AC coupled trace
and L(s) = s
s+a
where
a = I/T = 2irfc = 10 sec "]
The inverse transform is thus
p(t) = L-l (P'(t))
where +
J $
thus p(t) = p'(t) + a/* p'(u)du
The AC coupled trace which was the average of 100 engine cycles, was
then numerically integrated to give the original trace. The AC
coupled and the integrated traces are shown in Figure IV.27.
The corrected pressure trace obtained from the Pinto 2300 cc, 4-cylinder
engine was then used to determine coefficients for both flame speed and
heat transfer equations. As shown by comparing pressure-time traces in
Figure IV.28., the combustion rate was well matched during most of the
combustion process. However, the peak pressure values computed by
the model are higher than those shown in tKe actual pressure trace.
-------
79
60.00
52.50
METHANOL
0.90
WOT
2000 RPM
SA- 34*BTDC
8.44:1 CR
O - AC COUPLED
-I- - INTEGRATED
u 22.50
7.50
0.00
-360
360
-240 -120 0 120 240
CRANK ANGLE DEGREES (ATDC)
FIGURE TST.27: COMPARISON OF AC COUPLED DATA AND INTEGRATED
DATA
-------
80
60.00
52.50
45.00
37. 50
ui
OT
CO
tU
30.00
>! 22.50
o
15.00
O - ACTUAL
-|- - COMPUTER
7.50
0.00
-360 -240 -120 0 120 240
CRANK ANGLE DEGREES (ATDC)
FIGURE 17.28: MATCH OF COMPUTER PREDICTED RESULTS TO
ACTUAL PRESSURE TRACES
360
-------
81
This discrepancy results from the fact that the model describes a
flat combustion chamber and the Pinto engine actually has a squish
chamber.
Comparisons between the actual test point and the model's matching
predictions of performance and emissions are shown in Table IV.6.
Since no quench zone modeling is included in the program, the Otto
cycle model shows practically no unburned hydrocarbons in the bulk
charge. An interesting point to note is that the program also shows
practically no formaldehyde (CH20) in the bulk gases, thereby showing
it as a quench phenomenon. This can be verified by noting that
formaldehyde is an intermediate step in the combustion reactions and
therefore would only exist in the quench layer.
This squish-type combustion chamber of the actual engine tends to in-
crease turbulence during the beginning of the combustion process which
has been modeled well by the computer simulation. However, as the
piston passes TDC, decreased turbulence and increased heat transfer
occur, causing drastic changes in flame speed and unburned mixture
temperature which are not accounted for in the computer model. This
effect not only reduces the peak cylinder pressure but also causes
the predicted maximum brake torque (MBT) spark advance curves to be
offset or retarded relative to the actual MBT spark advance curve for
the Pinto engine as shown in Figure IV.29. While this spark advance
distortion causes the computer model to predict higher power and
thermal efficiency at MBT spark timing than actual values, the com-
puter model variations with equivalence ratio are consistant with
actual trends. Dry exhaust emission comparisons (Figure IV.30) show
that while predicted C02 and 02 values match experimental values,
predicted CO is much lower than the experimental CO. This indicates
too fast a reaction rate constant for the CO oxidation reaction.
CO + OH ->• C02 + H
The rate used here is a slower rate than was used in the gas turbine
studies. Yet, the predicted CO values are still too low. This issue
-------
82
TABLE IV.6
PERFORMANCE AND EMISSIONS COMPARISON
AT MATCH POINT (WOT & 2000 RPM)
Engine Data Computer Model
Performance
$ 0.90 0.900
SA 34°BTDC 34°BTDC
POWER(KW) 16.7 17.4380
CHARGE WGT (QMS) 0.684 0.68416
nTH (%) 34.8 36.26
ISFC (g/KWH) 519 498.07
EXH TEMP (K) 916 924.5
Emissions c
UBF (PPM) 429 2.3 x 10"°
CH20 (PPM) 177 4
CO (%) 0.07 0.02
CO? (%) 13.2 13.411
02 (%) 2.26 2.161
He (%) 0.04 0.045
MO (PPM) 2410 2516
NOX (PPM) 2530 2522
NO (g/KWH) 11 11.560
NOX (g/KWH) 18 17.769
-------
83
UJ
O
a.
UJ
X
o
O
20
18
16
14
45
PINTO ENGINE
METHANOL
2000 RPM
WOT
8.44:1 CR
40
UJ
X 35 -
30
0.8 0.9
FUEL-AIR EQUIVALENCE RATIO
FIGURE H.2S: COMPARISON OF ACTUAL VS PREDICTED PERFORMANCE
DATA
-------
84
0.0 L
PINTO ENGINE
METHANOL
2000 RPM
WOT
8.44:I CR
NOX RANGE
(EXPERIMENT)
FUEL AIR EQUIVALENCE RATIO
I.O
FIGURE ET30: COMPARISONS BETWEEN EXPERIMENTAL AND
PREDICTED DRY EXHAUST EMISSIONS
-------
85
requires further study. Predicted MBT NOX emissions, as expected
due to the spark advance offset, are also lower than actual values.
However, trends with equivalence ratio match experimental data
reasonably.
The next test of the model was to predict performance and emission
parameters for the test engine at 4000 rpm. Predicted performance
data for 2000 rpm and 4000 rpm are compared in Figure IV. 31. As
expected IMEP and thermal efficiency increase with engine speed.
Predicted dry emission comparisons (see Figure IV. 32) show that
NOX and CO emissions decrease with increasing engine speed while
02 and C02 emissions show no change relative to engine speed.
Limited experimental data at 4000 rpm support these findings.
With the model showing reasonable trends as functions of equivalence
ratio and engine speed, it was then used to predict the performance
and emission trends with increasing compression ratio. As shown in
Figure IV. 33, the power and thermal efficiency increased 14% when
the compression ratio was raised from 8.44:1 to 14:1, while NOX
emissions increased 18% on a volumetric basis. It is interesting to
note that these factors combine over the range of compression ratios
mentioned, to give a net increase of indicated specific NOX (ISNOX)
of only 2.7%. This information is contrary to that reported from
experimentation which indicated a drop in volumetric NOX in going
from a compression ratio of 9.7:1 to 14:1. A possible explanation
for this experimental trend may be that preignition or knock reac-
tions have caused the experimentalists to retard the spark from the
MBT value. To support this argument, the predicted spark retard
performance and exhaust emissions are shown in Figure IV. 34. With
a 3° spark retard from MBT, the volumetric NOX at 14:1 compression
ratio is equal to that at 8.44:1 with MBT spark timing. This still
shows a 13.7% increase in thermal efficiency and power with a 12.4%
reduction in ISNOX. Consequently, the experimental results
reported by Bernhardt and by Pischinger can be obtained with a
slight spark retard. Since only a 3° retard was necessary to dupli-
cate the NOX trend and with a power reduction of only 0.4%, the
reported experimental conclusions could also have resulted from a
missettingof MBT spark advance.
-------
86
METHANOL
WOT
MBT SPARK
8.44:1 CR
T IN « 282 °K
O - 2000 RPM
X - 4OOO RPM
o
Q
ffi
o
UJ
O
1
V)
FUEL AIR
0.9
EQUIVALENCE RATIO
FIGURE ET.3I : COMPUTER PREDICTED PERFORMANCE FOR
2000 RPM AND 4000 RPM
-------
87
DRY EMISSIONS
METHANOL
WOT
MBT SPARK
8.44= I CR
TlN= 282° K
O - 2000 RPM
X - 4000 RPM
1600
FIGURE 17,32
FUEL-AIR EQUIVALENCE RATIO
COMPUTER PREDICTED DRY EMISSIONS FOR
2000 RPM AND 4000 RPM
-------
2000 -
^ 1800 •
Q.
Q.
x 1600
O
1450 -
o:
UJ
21
S 5 19
o
o
17 -
45 -
<
oc 3
uj 38
H >. 40 H
S g
o £
O U.
z uj 35
METHANOL
$= 0.900
WOT
2000 RPM
MBT SPARK
89 10 II 12
COMMPRESION RATIO
13
12.0
II.0
10.0
20
o
o
18 o
e
CJ
O
CO
X
o
16
14
12
o
i
(T
£
CO
10
14
FIGURE Ttf.33: COMPUTER PREDICTED MBT COMPRESSION RATIO
EFFECT
-------
89
8|
x.
*£.
o
CO
Q *
U «^
5 c
u ui
i ^
= a.
12.0 '
10.0
8.0 -
6.0 •
22 -
20 -
18 -
i*
S 5
b O
ft
O QJ
36
10
MBT
8
* 8.44: I \SNOX
JSNO
8.44: I POWER
8.44:1
6
METHANOL
$• 0.900
WOT
2000 RPM
14:I CR
X - MBT FOR 8.44: I CR
- 1800
- 1600
1400
1200
• 1000
800
0
TDC
SPARK ADVANCE CBTDC)
FIGURE ET.34: COMPUTER PREDICTED SPARK RETARD EFFECTS
-------
90
We have concluded that our thermokinetic computer model of a specific
SI engine does correctly predict trends in performance and emissions
as functions of equivalence ratio and speed when compared with actual
data. However, the absolute values of the predicted parameters are
not in as good agreement. This is believed due to oversimplified
modeling of the combustion chamber.
The computer model predicts an increase in volumetric NOx emissions
with CR at MBT spark timing, which is contrary to reported experimental
results. However, the computer model reveals that the volumetric
NOx emissions at a CR of 14:1 can be reduced to those of a CR of 8.44:1
and MBT spark timing by retarding the spark 3° from MBT spark setting
at 14:1 CR. There remains a 13.7% increase in power and efficiency at
this retarded spark setting. Assuming the TDC surface to volume ratio
is held constant, it is expected that the cylinder hydrocarbon and
aldehyde emissions will not be significantly affected by increasing
the compression ratio. The high CR will also enhance cold starting.
Hence, the very important conclusion is reached that the CR of a
methanol fueled engine should be as high as possible without intro-
ducing autoignition.
IV.4.2 Continuation of Modeling Study: One of the least studied areas of
methanol combustion is the quench zone which is believed to be the
source of most of the hydrocarbon and aldehyde emissions. Therefore,
we plan to use the model to look at the quench phenomenon in methanol
fueled SI engines. Another area which holds much interest in the
use of methanol in automotive engines is the cold start problem.
This is another area we plan to investigate with the computer model.
Other areas which should also be included in our future plans are
better modeling of the squish type combustion chamber, droplet
combustion, use of charge stratification and the spark initiation
phenomenoa.
-------
91
IV.5 COLD STARTING AND LEAN BURNING
IV.5 Cold Starting and Lean Burning: Like gasoline, methanol fueled
engines become progressively more difficult to start in winter temper-
atures. However, the onset of this difficulty occurs at a higher
ambient temperature for methanol (30 to 40° F) than for gasoline
(0 to 20°F). Gasoline will permit startup at these lower temperatures
because of its volatile components. Volatile components such as butane
can also be blended with methanol for cold starting enhancement. However,
this is not considered satisfactory as it weakens the safety of methanol
from explosive ignition associated with spills and tank leakage. In
pure form, methanol is much less hazardous than gasoline in this regard.
To date, cold starting of methanol fueled automotive engines has usually
been accomplished by use of an auxiliary gaseous fuel. For example, the
vehicle operated by the City of Santa Clara, which is now entering its
sixth year of operation, has always used propane for cold weather starts.
These auxiliary fuels present no special problems to the experienced
operator. However, they are seen as a source of consumer complaint if
they are to be incorporated in vehicles operated by the public.
Our early research efforts^ showed that methanol in dissociated form (CO+2H2)
could be mixed with methanol in any percentage and used successfully as an
automotive fuel. Because of the wide flammability limits of H£ and the fact
that dissociated methanol is a gaseous fuel, it has appeal for cold starting
providing it can be generated under cold start conditions from the liquid
methanol stored in the vehicle's fuel tank using the vehicle's battery power.
It is toward this goal that our research efforts over the past year have
been principally directed.
IV.5.1 Battery Power Generation of Dissociated Methanol: Equilibrium studies
from our computer, which do not consider activation energies, show
that methanol can be considered thermodynamically unstable even at
room temperature, and, if all possible products are considered, it
-------
92
decomposes to a variety of compounds at low temperatures, eventually
ending with only carbon monoxide and hydrogen at the higher temperatures.
It is of further interest that once started, decomposition at lower
temperatures can provide an energy output. While the latter is a
potentially intriguing method of sustaining a decomposition reaction,
our initial approach was based on higher temperature decomposition
in which the reaction is endothermic and energy must be supplied.
This simple thermal approach was also considered more practical since
it appeared likely to be accomplished without the need for catalysts.
Following the decision to use a thermal decomposition approach, other
basic questions arose. How much methanol must be decomposed and how
much energy is required? Based on typical engine cranking speeds, it
was estimated that a fuel flow rate of 0.6 grams/sec would occur.
Since this represents the total fuel rate at cranking speed, must
all or only a fraction of this methanol be dissociated? Initially it
was assumed that all should be dissociated if possible.
The present test setup for exploring the practicality of the thermal
dissociation concept is shown schematically in Figure IV.35. The
apparatus has undergone a series of changes since inception but basically
all variations involved a means to pump, meter and initiate methanol
flow. Similar means were required for air or nitrogen which was
often injected simultaneously with the methanol as it was found to
enhance dissociation. Also required was a means to initiate battery
input and measure its electrical current.
The present apparatus is described as follows. Compressed nitrogen
is used to pressurize the methanol storage tank, using a null-type
gas regulator to maintain constant pressure. The liquid is forced
from the bottom of the tank first through a rotometer where flow
is measured and then to an electrically operated solenoid valve,
located slightly downstream of the methanol pressure gauge. The
solenoid valve leads directly to a hypodermic-needle-sized tube that
-------
CH3OH
ROTOMETER
SOLENOID
VALVE
FUEL
PUMP
METHANOL
FUEL
SUPPLY
GAS
REGULATOR
\ DECOMPOSITION TUBE
.250" O.D., 304 STAINLESS
STEEL HEATED TUBING,
12" LENGTH , 10" RADIUS
H"3 INDICATES HEATED
PORTION)
HEAT SOURCE-
12 VOLT TRUCK
BATTERY
WATER
LEVEL
COMPRESSED AIR
SUPPLY
CO + H2+CH4
FLAME ARRESTOR
WATER CONTAINER
GO
FIGURE IZ.35: PROTOTYPE METHANOL COLD START SYSTEM
-------
94
injects the liquid into the decomposition tube. Often air or nitrogen
is injected into the latter tube and it flows from the compressed gas
injector leading directly into the decomposition tube. The tube itself
is heated as a shunt to the truck battery through the contacts shown.
Battery current is initiated via a starter-solenoid and then passes
through an ammeter with a calibrated shunt for measuring current
flow. The run interval and time of initiation of the battery, methanol
and gas inputs are controlled with an interval timer, and the methanol
flow, especially, is timed during the run (and also during calibration)
with an electric stop watch. In most of the tests, the effluent from
the decomposition tube was passed directly into a wash water bath that
dissolves the water soluble components, and then the insoluble gases
exit the container to a tube where they are ignited by an electric
spark generator. In these tests, the degree of decomposition is
estimated by weighing the contaminated wash'water, and some qualitative
information is ascertained by observing the extent of the flame ,and
further traces of insoluble non-gaseous products. In a few later
tests, the effluent from the decomposition tube was passed into the top
of an enclosed container that was completely filled with water. The
insoluble gases displace the water into a separate container, and the
amount carried over is taken as a measure of the extent decomposed.
Samples of gases are also removed from the top of the enclosed container
and analysed on a gas chromatograph.
Tests were performed using various tube configurations. Because of
problems in attaining reproducibility in flow rates of methanol, it
is now realized that the early test results are not satisfactory. The
problem stemmed from the formation of air bubbles in the methanol line
at points where restrictions occurred, and these bubbles affected the
methanol flow rates. It is presently believed that the bubbles were a
consequence of degassing of dissolved gases on passage through a point
of low pressure. Their effect on flow rates has been largely eliminated
by the scheme indicated in Figure IV.35 in which the flow is controlled
by a tiny orifice placed within the decomposition tube. As a consequence,
-------
95
air bubbles which may be evolved do so downstream of the injector where
they have no effect on flow rate.
The procedure of heating the decomposition tube by shunting it directly
across the battery appears to have certain simplicities which are
desirable for an automobile application. However, it also results
in restricting tube dimensions to those with acceptible electrical
resistances. We have found, both experimentally and from calculations,
that dimensions that represent a tube of 1/4" O.D., .01" wall and length of
12 to 14" is satisfactory. Our first reliable experiements indicate
that simply injecting the methanol into such a tube, along with an air
q
flow between 1 and 6 ft /hr., with an electrical input from about
220 to 300 amps, and at methanol flows of about 0.5 grams/sec, for
about 3.8 sees., results in about 10% decomposition of the methanol
to water insoluble gases. Despite many diverse inter-relations of the
tube dimensions and flow rates, we did not succeed in improving this
value. It was, moreover, quite apparent that minor changes in dimensions
of the tube, generally resulted in a lower electrical input or decreased
temperature rise of the tube, could easily lower this value. It was
often decreased to negligible amounts by fairly small changes in
geometry. Our thermocouple measurements of tube temperatures during the
more successful runs ranged from 600 to 800°C. The effluent from the wash
water container burned with a clear blue flame. After we adopted stainless
steel tubes in our lines from the decomposition tube to the wash container,
there was rarely more than negligible insoluble non-gaseous residue
produced, and we never found more than a trace of formaldehyde in the
wash water.
In continuing experiments, we sought to improve decomposition by thermally
insulating the tube. Radiation tubes, in which the decomposition tube
was surrounded by a larger one of reflective materials, produced very
small increases. However, thermally insulating by wrapping with glass
cloth effectively increased the yield to a range of 15 to 20%, compared
to 10% for the non-insulated tubes. Experiments were therefore performed
in which the fuel was more assurredly heated. They comprised locating
the 1/4" tube within a 1/2" tube, with methanol injected into the smaller
tube and then passing via holes into the annul us between the tubes before
-------
96
being led to the wash water. The concentric tubes were electrically
heated for lengths of time required to attain selected temperatures,
prior to injection of the methanol. Under these circumstances,
decomposition of about 10% was achieved at only 500°C, and well over
25% at 900°C. While this experimental device does not qualify for
use on an automobile and it requires a much greater electrical input
than normally available, it indicated that great improvements in
decomposition can be achieved by heating methanol to moderate temperatures
for sufficiently long stay-times.
Based on the above, a decomposition device was constructed of similar
design but with characteristics that might be satisfactory for use on an
automobile. For reasons already discussed, these seem to be limited to
dimensions corresponding to 1/4" diameter by .01 wall by 12", and we
chose to simulate the latter with two concentric tubes, 3/16" and 1/16"
O.D. In essence, the methanol is injected into the smaller of the
tubes, passes via orifices at the end of the tube into the annul us
between it and the larger tube, and then exits near the point of injection.
This apparatus resulted in decompositions between 50 and 60%, which
represents very attractive progress.
In addition to tests with the above device in which the effluent was
passed directly into water and the water-solubles determined by
weighing, the effluent was passed into the closed water container indicated
in the bottom right of Figure IV.35. The displaced water indicated
decomposition of over 40%. Analysis of the collected gas showed about
17% CO, 41% H2, and 2-1/2% methane. The CO and H,, concentrations
appear, within experimental error, of the 1:2 ratio required of complete
dissociation of methanol; but the total gas concentration found is less
than 100%, which may indicate the presence of other gases.
This vein of exploration has demonstrated that methanol may be thermally
decomposed in an apparatus of practical dimension for use in an automobile.
-------
97
However, the electrical energy input is too high to meet cold starting
requirements. Studies will continue with emphasis on ways to decrease
this energy requirement.
IV.5.2 Cold Start and Lean Burn Experiments; Initially it was assumed that
all the methanol flowing at cranking speeds must be dissociated to
achieve ignition and burning under cold start conditions. But is
this the case? is u possible that a pilot amount of gaseous fuel
can sustain a flame in the very lean air-methanol vapor mixture
encountered with cold start conditions? Two avenues of exploration
of this issue are available.
Experiments have been performed by others using a hollow electrode
spark plug to create a stratified charge in the cylinder with the
rich zone in the proximity of the spark plug.1cl It is the intent
of our experiments to attempt to inject and hold a gaseous mixture
of dissociated methanol (CO + 2^) in the proximity of the hollow
electrode spark plug for supporting combustion of methanol under
cold-start, lean-burn conditions. Percentage of the fuel injected
through this electrode will be varied to determine the least per-
centage necessary as a function of ambient temperature down to
-10°F or -230C.
Chemical chains and reaction rate data already exists in our com-
puter for the combustion of dissociated methanol. By incorporation
of a stratified charge configuration for the simulated gases in the
combustion chamber, it may be possible to study the cold-start,
lean-burn problem with the computer. If so, limited experimental
testing could be used to verify the computer predictions and the
computer data would,, at relatively low cost, provide the bulk of
the information relative to the percentage of dissociated methanol
required as pilot fuel for cold starting.
-------
98
Lean burning of a homogenous mixture under normal engine
operation will continue to be explored experimentally in our
dynamometer tests and computer modeling studies. If the use of
the hollow electrode spark plug proves satisfactory in the
cold start studies, it will then also be incorporated in our
lean burn studies for normal operation. This is an important
issue for our early studies^ have shown that engine operation
on equivalence ratios as low as 0.4 are possible using
dissociated methanol.
-------
99
IV.6. Engine Wear Rate and Crank Case Blow By: Investigators have
reported a variety of evidence relative to the influence of methanol
fuel on engine wear rates. The summation of this evidence is
inconclusive. ^' ^' ^3
One very favorable piece of evidence, that there is no significant
difference between methanol and gasoline, relates to our methanol
fueled vehicles. The 1970 model American Motors Gremlin is now in
its seventh year of operation on pure methanol with no detectable
engine wear problems. Indeed there have been no failures in either
engine components or major fuel system components. A 1972 Plymouth
Valiant is now in its 6th year of operation on pure methanol. During
its entire life, it has been operated in stop-go driving by the meter-
readers for the City of Santa Clara. This is a very severe wear mode
of operation. Yet, there have been no failures in either engine com-
ponents or major fuel system components.
Other evidence in the literature suggests that methanol may be incom-
patible with some of the materials presently used in the internal com-
bustion engine. It is known that methanol will attack the terne plate
coating on the fuel tank. Other workers^' ^5 have reported wear or
the lack of it while using methanol and methanol-gasoline blends, but
their results are cloudy as engine wear has not been studied in a con-
trolled test until recently.3' 16
Our present engine tests utilizing methanol fuel include an engine wear
rate test designed to analyze the change in engine wear rate due to the
use of methanol fuel in comparison with unleaded gasoline fuel. In this
context engine wear rate is defined as the rate of accumulation of copper,
lead, iron and chromium metals in the lubricating oil of the engine due
to the loss of these metals from the engine.
-------
100
The choice of these metals is based on their significance in the engine
parts. In the early phases of the tests, aluminum and tin were also
measured but showed negligible evidence of wear and so were dropped from
the tests.
In order to monitor metal loss from the engine, a warm oil sample (approxi-
mately 200 grams) was removed from the lubricating system approximately
every ten engine hours and tested for metals level. The oil samples were
diluted in xylene, and analyzed by atomic absorption spectroscopy for lead,
copper, chromium and iron concentration. Atomic absorption spectroscopy
is a common technique for oil analysis used by industry. ' '
During the course of the test, the engine load and speed were varied in a
regular pattern as explained elsewhere in this report. The engine was
started and stopped frequently. Fuels were also used for various lengths
of time. Thus, the effects of engine start-up, load and speed are
approximately the same for the methanol and gasoline runs.
The accumulated wear metals in the oil as functions of engine hours and
fuel type are plotted on a semilogrithmic graph, Figure IV.36. Atomic
absorption data corrected for oil dilution (due to oil addition) and
oil concentration (due to oil losses) is the basis for this evidence.
Events that may have influenced this data are keyed to Figure IV.36. by
event numbers listed in Table IV. 7.
The only wear rate contrast that may be significant from tests is the
possible increase in the wear rate of copper due to the use of methanol
fuel. The slope of the accumulated copper curve has distinct transitions
at 260, 330, 415, 510 and 550 engine hours where the fuel was changed.
Though the wear rate of both iron and chromium varied during the test,
there is no consistent correlation between fuel and wear rate of either
the iron or chromium.
-------
o
o
o
X
o
CO
UJ
<
UJ
s
s
I 5
q 4
1000
500
400
300
2OO
100
50
40
30
20
10
• - COPPER
-|- - IRON
A — LEAD
• - CHROMIUM
300 400 500
ENGINE OPERATING HOURS
600
700
FIGURE EC. 36: EFFECT OF FUEL TYPE ON ENGINE WEAR
-------
102
TABLE IV.7 HISTORY OF ENGINE EVENTS
N0_. ENGINE HOURS EVENTS
0 000 Indolene Fuel
1 030 Oil Leaks
2 171-180 Oil Leaks
3 180 Oil Pressure Drop
4 182 Filter Change
5 185 Oil Change
6 216-225 Oil Leaks
7 225-236 Extensive Oil Leak
8 246-257 Oil Leak
9 263 Oil Change—New Filter
10 265 Methanol Fuel
11 265-278 Oil Leaks
12 278-282 Oil Circuit Failure-
total pressure loss
13 282 Oil Changed
14 329 Chevron Unleaded Fuel
15 362-376 Oil Leak
16 365 Filter Change
17 401 Methanol Fuel
18 409 Chevron Unleaded
19 414 Methanol Fuel
20 418-434 Oil Leak
21 444 Oil Leak
22 466-498 Oil Leak
23 510 Chevron Unleaded
24 519 Methanol Fuel
25 524 Chevron Unleaded
26 558 Methanol Fuel
27 567 Oil Leaks
28 648 Oil Change, Filter Change
29 732 Oil Change
-------
103
The wear rate of lead rapidly increased after the oil pressure failure (event 12)
but eventually becomes negligible. It is concluded that the wear rate
of the lead bearings is unaffected by the fuel.
The calculated results using this evaluation technique are very sensitive
to the engine oil volume. During the course of the study, the amount
of oil in the engine at any given time was only known within +_ 204 grams
and the oil losses which were assumed to occur at a constant rate were
uncertain within +102 grams. These uncertainty errors decreased as the
test went on due to refinement of sampling techniques so that the uncertainty
due to losses was reduced to 50 grams and the uncertainty of the amount
of oil in the engine was reduced to 100 grams. It is estimated by error
analysis that uncertainties in the final accumulated wear concentrations
do not exceed +_ 5%.
In order to strengthen the validity of this particular wear analysis,
oil filters were removed and analyzed for wear metal content. It was
found that the oil filter removes between 25 and 50 percent of the wear
metals from the oil and does so at a constant rate of removal except
when extremely high metal concentrations are present (i.e., during the
oil pressure loss). Thus, the presence of the oil filter in the oil
line affects the magnitude of the actual metals lost but it does not
alter the relative wear rate evidence.
Atomic absorption analysis was also conducted on the large particles
(>.45 microns) for several oil samples (#51, 54, 55). The metals level
of the large particles of chromium, iron and lead represented only 18-
27% of the total metals content in the oil. However, the large copper
particles were up to 80% of the level of copper in the oil for these
samples. These samples were only taken when methanol was used as fuel.
This evidence casts significant uncertainty on the conclusion of increased
copper wear from methanol.
-------
104
Analysis of engine wear rates as functions of fuel yields the following
conclusions:
a. The wear of copper parts may be affected by methanol fuel.
b. The wear rate of iron, chromium and lead parts does not change due
to the use of methanol fuel.
c. Engine wear studies should investigate large (>.45 micron) particles
as well as small particles to evaluate wear.
d. The presence of oil filters in the oil system does not significantly
effect wear rate data.
It is not presently known if the lost copper is from the bearings or oil
pump. This will be investigated when the test engine is disassembled.
Further wear studies using the same technique of oil analysis will continue
to be conducted for the test engines and future test cars in order to
broaden our data base and provide more accurate evidence.
-------
105
V. CONCLUSIONS
The past year's work has provided further evidence of methanol's
advantages over gasoline. Some of the conclusions which follow
from this work reinforce previous observations and some appear for
the first time.
1. Methanol is generally superior to gasoline in power, thermal
efficiency, and reduced emissions with the exception of
aldehydes.
2. Maldistribution of the air-fuel mixtures among the cylinders
of a multicylinder engine and variation of the mixture with
speed and load must be reduced for good performance and
emissions control for both gasoline and methanol fueled
engines.
3. Prototype hardware (WHB system) has been used in our ex-
periments that solves the maldistribution problem and allows
a 20-30% reduction in idle speed. However, it does not as
yet control the air-fuel ratio as function of speed and load
to a satisfactory degree.
4. Three different fuel metering systems (WHB, Dresserator, and
EFI) were found to provide superior steady state performance
on methanol when compared with the OEM carburetor system with
enlarged fuel jets.
5. Computer simulation of the engine's thermokinetic combustion
events shows reasonable predictive trends in power, thermal
efficiency and NOx emissions due to changes in equivalence
ratio, speed, and compression ratio. It does not as yet
satisfactorily deal with "squish" type combustion chamber
effects nor with quench zone effects. However, the evidence
from the study clearly suggests use of higher compression
ratios with methanol.
6. Computer simulation of the Federal emissions test procedure
using steady state data from our experimental engines has
produced evidence that lean-burning, methanol-fueled vehicles
-------
106
may meet statutory limits of emissions if tight control is
maintained on the equivalence ratio.
7. There do not appear to be any major wear contrasts with the
possible exception of copper for engines using methanol or
gasoline.
8. There does not appear to be any significant difficulty in
converting small gas turbines to run on methanol with equal
power at lower turbine inlet temperatures and with marked
reduction in NOX.
9. Preliminary studies relative to marine spills indicates that
methanol is naturally present in that environment. Other than
localized and short term damage it may not cause harm to the
ecosystem in the event of a major coastal spill.
-------
107
IV. REFERENCES
1. R. E. Fitch and J. D. Kilgroe, "Investigation of a Substitute Fuel
to Control Automotive Air Pollution," Final Report, CETEC Corp.
NAPCA Contract No. CPA 22-69-10, Mountain View, Ca., 1970.
2. R. K. Pefley, et. al., "Study of Decomposed Methanol as a Low
Emission Fuel," Final Report, Santa Clara Univ., EPA, Contract
No. EHS 70-08, Santa Clara, Ca., 1971.
3. R. K. Pefley, A. E. Bayce, L. H. Browning, M. C. McCormack, and
M. A. Sweeney, "Characterization and Research Investigation of
Methanol and Methyl Fuels in Automobile Engines," Report ME-76-2,
Santa Clara Univ., EPA Grant No. R. 803548-01, Santa Clara, Ca.,
1976.
4. R. K. Pefley, et. al., "Methanol-Gasoline Blends—University View
Point," Engineering Foundation Conference, Henniker, New Hampshire,
1974.
5. H. G. Adelman, L. H. Browning, and R. K. Pefley, "Predicted Exhaust
Emissions from a Methanol and Jet Fueled Gas Turbine Combustor,"
AIAA Journal, Vol. 14, No. 6, June, 1976, pp. 793-798.
6. W. R. Matthes, R. N. McGill, "Effects of the Degree of Fuel
Atomization on Single-Cylinder Engine Performance," Society of
Automotive Engineers, # 760117, Feb. 23-27, 1976.
7. W. E. Bernhardt, "Engine Performance and Exhaust Emission Character-
istics from a Methanol-Fueled Automobile," 1975, GMR Symposium,
Michigan, 1975.
8. F. Pischinger, "Discussion Contribution," Methanol as an Alternate -
Fuel Proceedings, Stockholm, Sweden, 1976, pp. 113-116.
-------
108
9. D. R. Lancaster, R. B. Krieger, and J. H. Lienesch, "Measurement and
Analysis of Engine Pressure Data," SAE Paper No. 750026, SAE Trans-
actions, Vol. 84, 1975, pp. 155-172.
10. M. D. Leshner, W. H. Baisley, and E. Leshner, "A Fuel Vapor
Injector/Igniter System," Final Report, Fuel Injection Development
Corp., ERDA Contract # E (04-3)-1237, Bellmawr, New Jersey,
February, 1977.
11. J. C. Ingamells and R. H. Lindquist, "Methanol as a Motor Fuel,"
preprint, Chevron Research Laboratory Report J 4802, Richmond, Ca.,
May, 1974.
12. J. D. Rogers, "Ethanol and Methanol as Automotive Fuel," Report
No. P813-3, E. I. Du Pont De Nemours and Co., Inc., 1973.
13. R. M. Tillman, Data Presented at the Bureau of Mines Sponsored
Technical Meeting, Denver, Colorado, Sept., 1974.
14. R. R. Adt, Jr., R. D. Doepker, and L. E. Poteat, "Methanol-
Gasoline Fuels for Automotive Transportation," University of Miami.
15 . "On the Trail of New Fuels-Alternative Fuels for Motor Vehicles,"
Federal Ministry for Research and Technology, Bonn, Germany, 1974.
16. "Engine Lubricants for Use in Methanol Fueled Highway Vehicles,"
Southwest Research Institute, San Antonio, Texas, October, 1976.
17 . E. A. Means and D. Ratcliff, "Determination of Wear Metals in
Lubricating Oils by Atomic Absorbtion Spectroscopy," Atomic
Absorbtion Newsletter, Vol. 4, No. 1, January, 1965.
18 . "Testing Used Engine Oils-the Why and How," Chevron Research
Bulletin, Chevron Research Laboratory, Richmond, Ca., 1973.
19. "Scheduled Oil Sampling as a Maintenance Tool," Paper No, 720372,
presented at Earthmoving Industry Conference, Society of Automotive
Engineers, April 1972.
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverie before completing)
I. REPORT NO.
M.E. 77-1
3. RECIPIENT'S ACCESSION NO.
4. TITLE ANDSUBTITLE
Characterization and Research Investigation of
Methanol and Methyl Fuel: Final Report
5. REPORT DATE
August 1977
6. PERFORMING ORGANIZATION CODE
7-AUTH8.RISL Pefley, L. H. Browning, M. C. McCormack,
M. L. Hornberger, W. E. Likos, B. Pullman
8. PERFORMING ORGANIZATION REPORT NO
9. 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, MI 48105
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
An automotive engine mounted on a dynamometer is used to generate power, efficienc
and emissions maps which compare methanol with gasoline. This data is also used in
conjunction with a computer program to predict vehicle performance and emissions.
Methanol is found to offer advantages over gasoline. Two alternates to the stock fuel
preparation system are also evaluated. They show improvements over the stock system.
Computer modeling of the thermokinetic events in the test engine using methanol
has allowed predictions of power, efficiency and emissions as functions of com-
pression ratio, spark advance, air-fuel ratio and speed. High compression ratios
appear beneficial.
The report also considers engine wear, cold start aspects of methanol. It also
presents some gas turbine evidence which favors methanol over commercial turbine
fuel.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Methanols
Methyl Alcohol
Exhaust Emissions
18. DISTRIBUTION STATEMENT
19. SECURITY CLASS (This Report)
Unclassified
Not Restricted
21. NO. OF PAGES
108
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
------- |