UIL U-EN G-95-2558

INTAKE TUNING, ETHANOL CONVERSION, AND EMISSIONS ANALYSIS OF

A 620 cc FOUR STROKE V-TWIM ENGINE

PHILIP ROBERT GUZIEC

Power Affiliates Program

Department of Electrical and Computer Engineering
University of Dlinois at Urbana-Champaign
Urbana, Dlinois 61801

PAP-TR-95-9
August 1995

ii

FOREWARD

This technical report is a reprint of the thesis written by Philip Robert Guziec as partial
fulfillment of the requirements for the degree of Master of Science.in Mechanical Engineering
at the University of illinois.
Robert White
Thesis Advisor
August 1995

Acknowledgements

To Professor White, for your guidance and assistance throughout my

undergraduate and graduate education, thank you.

To Leland McWhorter,

I appreciate your patience and hard work in fabricating the intake

manifold.

To Peter Nelson, your considerable help with my fuel injection

computer design and programming was invaluable.

To Tim Frazier, I

appreciate your assistance in using the emissions analysis equipment

under your care.
thank you.

To my parents, for your support and encouragement,

Finally, I would like to thank both my parents and my advisor

for their tolerance of the duration of my research and writing.

iii

Table of Contents
Page

1.0

Introduction...................................................................................................... ~ ...... 1

2.0

Theory and Design ................................................................................................ 5

2.1 Intake Manifold ............................................................................................ S
2.2 Ethanol Conversion ...................................................................................... 9

2. 3 Operating Condition ..................................................................................... 9

3.0

Experimental Setup ........................................................................................... 1 1

4.0

Intake Manifold Tuning Results ................................................................... 1 4

4 .1 Helmholtz Tuning Results ...................................................................... 1 4
4. 2 Throttle Plate Effects on Resonance .................................................. 2 1

5.0

Ethanol Conversion Results and Discussion ............................................. 3 5

5.1

Spark Timing Comparison Between Gasoline and Ethanol. ............... 3 5

5.2

Thermal Efficiency Comparison of Gasoline and Ethanol .................. 3 8

6.0

Emissions Results ................................................................................................ 4 2

6.1 Primary Emissions .................................................................................... 4 3
6. 2 HC and NOx Emissions .............................................................................. 4 5

7.0

Conclusions ............................................................................................................ S 2

References ......................................................................................................................... 5 5

Appendix 1......................: 5 6
Appendix 2........................................................................................................................ 5 7
AI'J)endix 3....................................................................................................................... 5 8
Appendix 4........................................................................................................................ 5 9
Appendix 5................................................................................................................ 6 0

iv

1.0

Introduction .

Acceleration and hill climbing power requirements of a typical

United States passenger vehicle normally dictate the required peak
power for the selection of the vehicle's engine.

The equation for the

power required by a vehicle is as follows:

Power(hp)
Where:

= {CrMvg+Mv[a+g*sin(theta)]+.5(rho)CdAvSv2} Sv/550
Cr = coefficient of rolling resistance [lbf/lbf]
M v = mass of vehicle [slugs]

(1.1)

g = acceleration of gravity [ftts2]

a
acceleration of vehicle [ftts2)
theta = angle of incline of road
rho = density of air [slug/ft3)
Cd = drag coefficient of vehicle
Av
frontal area of vehicle [ft2]
Sv
speed of vehicle [ft/sec]

=
=

It is interesting to note that the additional power required due to

the component of gravitational acceleration of an inclined road acting .
directly against the vehicle's motion is indistinguishable from the
additional power required due to accelerations of the vehicle.

Due to low

maximum speed limits in the U.S., the conditions above lead to engines
that are typically sized to produce far more power than what is required
by the vehicle at normal cruising speeds.

For example, a typical 3500 lb

vehicle with a Cr of .012, a 21 ft2 frontal area with a Cd of .34 traveling

at 60 mph on a level road requires only 12.7 kW (16.6 hp) while an
acceleration of .1 g requires an additional 41.7 kW (55.9 hp) and a 5%
grade requires a further 20.9 KW (28.0 hp).

These transient power

requirements result in an engine that is larger and heavier than needed

1

to supply the average power required by the vehicle, and consequently

one which is run primarily under heavily throttled conditions.

The

engine is also mechanically connected to the drive wheels by a

transmission causing any change in the power or speed requirements of
the vehicle to require a corresponding change in the engine power or
speed.
Series type Hybrid f.lectric Vehicles (HEVs) have an electric motor
driving the wheels with the electric power supplied by a combination of
batteries and a power source known as an Auxiliary f.ower ll.nit (APU),
which converts chemical energy to electrical energy.
consist of an Internal

~ombustion

The APU can

f.ngine (ICE) or. gas turbine coupled to

a generator, a fuel cell, or any other means of converting chemical

energy to electrical energy.

The University of Illinois selected an ICE for

the HEV because it is an inexpensive well known technology which is

readily available.

Since there is no mechanical connection between the

APU and the drive wheels, the speed and power of the APU need not
directly correspond to the speed and immediate power requirements of
the vehicle.

This decoupling provides unique opportunities for the

isolation of engine operating conditions from the transients of the road

speed and power of the vehicle.
In a series type HEV, the battery pack can be chosen with a
capacity to provide or store the difference between the power
requirements of the vehicle and the power output of the APU.

This

storage capacity allows the selection of an APU which produces the

overall average power required by the vehicle.

In the case of an ICE

powered APU, a constant power level can be selected and optimized for
minimum Jirake S.pecific Euel C.onsumption (BSFC) and tailpipe emissions.
2

Optimization for minimum BSFC and minimum emissions are related in

that minimum specific fuel consumption reduces overall emissions by
reducing the quantity of fuel burned and percentage emissions are
related to BSFC, as will be discussed in chapter 6.

Proper selection of

the generator control strategy can relieve the engine controller of the
task of following rapid transients, allowing for more accurate fuel and
spark control and more uniform loading of the catalyst.
S_eries type HEV s can also be provided with a battery pack capacity
sufficient for a .limited range without the APU in operation, allowing for
daily short range commuting from an at home recharging station.
Hybrid operation can be made completely transparent to the driver,
with operation of the APU triggered automatically by an on board micro
controller at a selected state of charge of the battery pack.

The APU

should be started near the limit of the electric range and, by estimating
the rate of discharge of the battery pack, the time until APU start can be
estimated.

With the abundant electrical energy available from the

battery pack, the APU can be prepared for starting by preheating the
catalytic converter and Exhaust Q.as Qxygen (EGO) sensor to operating
temperature.

This can provide a considerable reduction in cold start

emissions[!].

The overall effect on emissions reduction can be

significant; since cold start emissions account for 60% to 80% of the
overall emissions of a modern automobile on the federal emissions test
procedure[2].

Preheating of the catalyst and EGO sensor and starting of

the APU can be controlled entirely by the on board micro controller,

eliminating the delay experienced with preheated catalytic converter
installations in traditional automobiles.

The engine selected for the Univers"ity of Illinois hybrid electric

vehicle is a Kawasaki FP 6200 small industrial engine used in John Deere
riding mowers.

The FD 6200 is a modern small industrial and off road

vehicle V-twin engine designed to meet 1997 off road vehicle emissions
requirements. The engine was selected for its rated power output, light
weight, and emissions conscious design.

The engine has an aluminum

block and heads, with two pushrod activated valves per cylinder, a
nominal compression ratio of I 0:1, and weighs 42 kilograms.

The engine

as manufactured by Kawasaki is equipped with a single throttle body

fuel injection system flowing into a simple two leg manifold controlled
by an open loop speed-density control computer.

Baseline testing

showed that the system resulted in an equivalence ratio maldistribution

of up to 20% between the cylinders due to the throttle plate location, the
throttle plate angle, the injector inlet location, and the timing of the fuel
injection pulse relative to the unevenly timed intake strokes of the two
cylinders.

Consequently, an intake manifold and fuel injection system

was designed to replace the one supplied by the manufacturer in order

to increase maximum power, reduce fuel maldistribution between the
cylinders, and reduce overall emissions.
Ethanol was selected as the primary fuel for the HEY due to the
potential for emissions reduction and because of its status as an
alternate fuel and a renewable resource.

2.0 Theory and Design

2.1

Intake

Manifold

The replacement intake manifold was designed with tuned runner

length and a closed loop port fuel injection system in order to increase
peak power and efficiency and reduce emissions.

The system was

optimized about the manufacturer's rated speed of 3600 RPM due to

noise and durability considerations.

Wide Q.pen Throttle (WOT)

operation was chosen as optimum for maximum power output and

minimum BSFC based on testing of the engine as supplied by the
manufacturer.
The intake manifold was designed to minimize intake airflow
pressure losses and was tuned for peak airflow at the design RPM using
Helmholtz [4] resonance theory.

Individual port fuel injection was

selected to minimize fuel maldistribution and allow the separation of the

intake runners.

Due to the uneven, 0 - 270 - 720, firing order of the V

twin engine, the intake runners for the two cylinders were kept
completely separated up to the air filter.

Two throttle plates were used,

one in each intake runner, and a small balance tube between the
runners provided equalization of intake runner

pressur~s

between the

two ports at idle arid a common measurement point for the manifold
pressure sensor.

Thus, for the purpose of this analysis, the cylinders

were treated as two single cylinder engines.

The simple Helmholtz

~odel

was selected for intake tuning

because it is the dominant. effect in single cylinder intake tuning and can
quickly and accurately predict the RPM at which the tuning peak occurs.
According to Thompson and Engleman[3], "the total breathing curve is
5

the result of the increase in breathing due to Helmholtz type resonance

with organ pipe oscillation ripples superimposed upon the Helmholtz
tuning."

Organ pipe contributions to the tuning effect are ignored in this

analysis.

The equation for Helmholtz tuning RPM for a single inlet pipe is:

RPM= 162/K * Cs * Sqrt[A/(L*VD)] * Sqrt [(R-1)/(R+l)] (2.1)

Where:

K is a constant (2.0 to 2.5) depending on valve timing

C s is the speed of sound in the inlet (ft/sec)
A is intake runner area, (in2)
L is the length of the inlet runner from the intake valve (in)
Vo is volum~tric displacement of the cylinder (in3)
R is compression ratio of the engine

The effective engine cylinder volume at mid-stroke, Vo/2*[(R+l)/(R-l)],

is used as the volume for the Helmholtz resonator and is included in the
equation and constant 1348.

According to Thompson and Engleman[3], a

length of pi/2 times the diameter of the runner should be added to the
length to account for the portion of the free air in the mouth of the tube
which moves at significant velocity.

In this design, an intake bell mouth

is used to prevent a vena contracta from reducing the effective area of

the inlet of the intake runner.

According to Thompson and Engleman[3],

an intake bell mouth also serves to dampen out the organ pipe
oscillations and improve the magnitude of the Helmholtz supercharge.
According to Engleman[4], if an intake bell mouth is used, the end of the
effective intake runner is in the bell mouth.

For the calculations in this

study, the end of the intake bell mouth was used as the effective length
of the runner.
Round aluminum tubing was chosen for the inlet runners for
simplicity of fabrication, and an inside diameter of 1.15 inches was
selected to most closely match the dimensions of the inlet port to the
cylinder heads.

This tubing diameter resulted in a mean intake gas

velocity based on 100% volumetric efficiency of 43 ft/sec which is

significantly lower than Engleman's reconunended 200 ft/sec [4].
Thompson and Engleman's [3] paper on single cylinder intake tuning,
however, contradictorily states that larger inlet runner areas result in
greater maximum supercharge.

For the purposes of this design, 43

ft/sec was accepted as a compromise required to properly match the

intake runner to the intake port without an abrupt expansion of runner
area and corresponding pressure drop.
For the optimum length of intake runner to fit within the space
constraints of the application, it was necessary that the intake runners
bend a number of times.

The centerline length of the tubing bends are

taken as the design length for the intake runners.

The bends may result

in some loss of supercharge, but should not change the RPM for
maximum tuning [4]. In order to minimize pressure losses in the bends,
the manifold is constructed of mandrel bent tubing of approximately
constant cross sectional area in the bend.
Figure 2.1 is a plot of calculated tuning RPM vs. intake runner
length using the recommended K factor of 2.1.[4]

The sensitivity of

tuning RPM to runner length at the design engine speed of 3600 RPM is
about 70 RPM per inch change in length.

This insensitivity, combined

with the broad range of effect of Helmholtz resonance tuning [3] and [4],
7

indicates that there should- be significant Helmholtz effect at the tuning

RPM despite any error in calculation due to inaccuracies in the runner
length due to bends, varying area of the inlet port, or choice of the valve
timing factor K.

Figure 2.1 shows that the theoretical tuning length is

approximately 26 inches from the intake valve to the end of the bell
mouth.

Tuning RPM vs. Intake Runner Leg nth

5000

: : : ::-r ~-: :r: ~- ::;-: ::r :- : : r : :

1trtt1! rtt~ trrtii~-riti~-r~

~~:-~~ ~-!- -~--t--~~-~~-t-~-~---~---~--~---~-~- -~---:-~

4000

: : =: !:;~:~!=~~cr~~~-:J:::~::::: :_:.~ r1-;:j~ ~:~ r=~=~:~ :~-~t:_=~ :1:~: :i

3000

. :r:::r::::r:::::r:::r:::r::r:::r~::r::::r:::r:::r:::: : : :r::::r:::F:=F=Tr:::r:::r-r:::r:::r:::i

a.

1I r1rr~r;

a:

..rr1 ..r rr r r r T r r ~r :

1;~-1rr-!r1rr;

11r.,..1!-1rr1r1;

tt-~-rti!ttt1t rtii~-tt t~r ~- -~

-~~rt+i;rtti~ ttt~tr++ ttr ~

2000

Tun 1ng RpM

.. .. ....r..... t--1..-i......r.....i .....

t.....i......i...... ......r......
~

1. tt~j;tttj;ri
-+-+~t-+-!--+++~7-f~

,_________... it-tjrittjrii

1000

. :r::1~:~:=~~~~1~~~~:=r1~~~ ;It~I=-~~~~t::t:~I-:=: J~:.;t:.F:l

0
0

10

20
30
Runner Leg nth (inches)

Figure

2.1

40

50

2.2

Ethanol

Conversion

Using ethanol as a fuel requires an increase in fuel flow rate by a

factor of 1.57 to accommodate the change in air fuel ratio from 14.6 for
gasoline to 9.3 for a blend of 95% ethanol and 5% gasoline [5].

This can

be accomplished by resizing the fuel injectors, increasing the fuel

pressure, or increasing the pulse duration of the fuel injectors if they are
of adequate size.

In order to convert the throttle body intake system to

ethanol operation for baseline data collection, the fuel pressure was
increased to increase the fuel flow rate.

When port fuel injectors were

used for the new intake system, the fuel injectors were sized for the
greater fuel flow rate required while still operating at the rated fuel
pressure for the injectors.

MBT spark timing may change due to

differences in ignition delay and flame front propagation of ethanol

relative to gasoline.

The higher octane ratjng of ethanol may also allow

increased spark advance if gasoline spark advance is retarded. from MBT

due to knock limits.

Ethanol also has a much leaner flammability limit

[5], allowing greater flexibility in choosing an equivalence ratio for

minimum BSFC and emissions.

2.3

Operating

Conditions

Modern automobile engines spend the majority of their life

operating at low power levels and stoichiometric air fuel ratios.

At high

power levels where maximum power is a concern and emissions

requirements are not applicable, the engines are run at equivalence
ratios near 1.17 for both maximum power and internal engine cooling.
9

The test engine as supplied from the manufacturer was designed for use
in a riding mower application where the primary operation would be at
high load for long periods.

Consequently, the engine operated at

equivalence ratios of approximately 1.1 to 1.2 at all times.

Operating at

maximum power with stoichiometric or lean air fuel ratios is not a

normal condition in either automotive engines or industrial applications
of small engines, and such operation results in higher thermal loading
and less internal cooling than originally accounted for in the engine
design.

Such high thermal loads can result in a number of wear and

durability problems as well as temperature problems.

These problems

can include but are not limited to rapid valve seat wear, cylinder wall
wear, cooling problems, and cylinder hot spots which can cause
preignition.

Further, ethanol does not have the lubricating effect that

gasoline does, potentially exacerbating wear problems.

No data was

collected on the extent of these effects in the test engine.

One problem encountered in the test engine was preignition.

The

preignition was so severe that the engine continued to run and produce
half power with the spark plugs wires disconnected.

The assumed cause

of the preignition was overheated spark plugs as the spark plug

electrodes were completely melted away.

In this instance, the problem

was remedied by a change to spark plugs with a much colder heat range.
Spark plug electrode problems with alcohol fuels are well known, and
different materials are commonly used.

10

3.0

Experimental

Setup

The dynamometer used for the testing was a Mid-West

Dynamometer and Engineering Company eddy current type capable
of absorbing a maximum of 250 horsepower.

The dynamometer was

controlled by a Dyne Systems Dyn-Loc digital controller using

feedback from an inductive speed pick-up mounted on the
dynamometer shaft and a BLH Electronics model V35B load cell.
tests were conducted at constant RPM.

All

Constant RPM dynamometer

operation closely approximates the type of continuous operation

expected in a series type hybrid which buffers the engine from
transient fluctuations of load and speed.
The engine was cooled by routing the coolant through a water
to water heat exchanger with heat rejected to a continuously running
supply of cold tap water.

The flow rate was manually adjusted to

maintain temperature of the coolant intake into the block between

65 C and 75 C, corresponding to the manufacturer's thermostat
opening temperature of 75 C.

This temperature is lower than

typical modern automotive thermostats which typically operate

above 95 C and was probably selected by the

manufa~turer

to help

reduce component temperatures during continuous high load

operation.
An exhaust manifold was fabricated with two 1 inch diameter
runners 18 inches in length flowing into a round chamber of 2 inch
diameter and 24 inches long.
laboratory exhaust system.

This manifold was connected to the

The exhaust backpressure was not

measured during testing but was near atmospheric.

11

The emissions

measurement samples of exhaust gases were collected with .25 inch

diameter stainless tubing passing through compression fittings into
the exha.ust runners 2 inches from the exhaust port and bent at a
small radius to a 1 inch long straight section pointing upstream at the
exhaust ports.

The flow was routed through heated lines to

emissions measurement apparatus capable of measuring percentages

of carbon monoxide, carbon dioxide, and oxygen, and PPM of
unburned hydrocarbons and oxides of nitrogen.

The equivalence

ratio was calculated from the exhaust constituents.[4]

The fuel flow rate was measured using a calibrated burette
mounted atop a sealed tank containing the high pressure fuel pump
and return line from the pressure regulator.

The fuel pressure

regulator in conjunction with a mechanical fuel pressure gauge were

used to set the fuel pressure at the 270 kPa rating of the injectors.
The burette was filled prior to each measurement with a boost pump
and, with the engine running at constant conditions, a stopwatch was
used to measure the time required to empty the burette.

This fuel

measurement technique produced results repeatable within .I %.

Mean intake manifold pressure was measured with a mercury
manometer attached to the common pressure measurement point of the
intake manifold.

A..007 inch metering orifice from a carburetor was

used to dampen out cycle to cycle manifold pressure variations to the

manometer and fuel injection system pressure sensor.

Individual intake

port pressure traces were measured at a tap in the intake runner wall 3
inches from the intake valve at the flange which mounted the intake
manifold to the cylinder head.

The pressure was measured using a

Motorola MPX2200AP piezoresistive pressure sensor.

12

The pressure

range of the sensor is 0 to 2 bar absolute with linearity within . 0.5

percent over laboratory temperature ranges.

The output voltage was

amplified with an Analog Devices AMP-02 instrumentation amplifier to

a final output voltage of 0 to 4 volts.

The voltage was measured using

an HP 54501 100 MHz digital storage oscilloscope and printed on an HP

thinkjet printer.

The pressure sensor was connected to the intake

manifold with twelve inches of 1/8 inch ID vacuum tubing.

The

frequency response of the vacuum tubing and pressure sensor setup was
not measured, however pressure drops at intake valve opening shown
on figures 4.13 to 4.21 are vertical within the resolution of the plotter,
indicating a high enough frequency response for this application.
All raw data which were collected are included in appendices 1
through 5.

13

4.0

4.1

Helmholtz

Intake

Tuning

Manifold

Tuning

Results

Results

From the Helmholtz theory calculations, see Figure 2.1, the tuning
length for Helmholtz resonance at 3600 RPM is approximately 26 inches
from the intake valve to the end of the intake trumpet.

To examine off

design performance and to allow for the inaccuracies inherent in the K

factor of the model, see section 2.1, WOT torque was measured over the
effective RPM range of the engine with intake runner lengths of 18.5,
21.5, 24, and 27 inches.

While varying intake runner length, the test

conditions were stoichiometric equivalence ratio with ethanol fuel (E95)

and the manufacturers nominal spark advance of 21 degrees.

The

manufacturers fuel injection computer was used to simultaneously

trigger injector drivers for the two fuel injectors.

Equivalence ratio was

controlled by varying fuel pressure to achieve approximately .5 volts on

the EGO sensor, indicating near stoichiometric operation.

The

imprecision inherent in this control method resulted in some scatter in

the data and lower torque values than observed during later testing.
The torque data are plotted in Figure 4.1 with the vertical axis
truncated to better show the effect of the different lengths.
Increasing the intake runner length increases the low RPM torque
and decreases the high RPM torque.

The shape of the torque curve

appears to be limited by the conservative camshaft timing.

The

intake valves do not open until Iop ILead Center (TDC) and remain
open for 236 degrees.

Further increases in torque may be possible at

the design speed with changes in valve timing.

14

Torque vs. RPM

50

48

- 46
E

- 44
z

42

2000

2400

2800

3200

3600

4000

RPM

Figure

4.1

In order to show the small overall change in torque due to change in

length at the design speed, torque is plotted vs. length at 3600 RPM in
figure 4.2.

There is a maximum at 24 inches in length, however it is

barely noticeable.
An overall intake runner length of 24 inches from intake valve to
bellmouth was selected based on figure 4.1 as the optimum for the
design speed of 3600 RPM and all further testing was done at this
length.
15

Torque vs. Length at 3600 RPM

50.00

-~-~ ~--~ ~-~

40.00

-- ... . .... - ..................... - .

~ ~-~- ~- ~-~r~ ~-~ -~ -~- ~--1~-~--~-~ ~ ~~ -~-~ ~ ~-~- ~

. ...1

. ..... .. . ,.... - . . r .. : . . . .j

- 30.00
E

- 20.00
z

10.00
0. 000 -+------~---~-------------~
18

20

24

22

26

28

Inches

Figure

4.2

In order to compare the overall torque curve of the tuned intake

manifold to that of. the manufacturer's manifold, torque as a function of
RPM for the tuned manifold was estimated from the data of figure 4.1.
The correction was necessary to account for the increase in torque
observed during later testing due to optimized spark advance and fuel
control.
figure 4.1

This correction was made by multiplying the torque values .in

at 24 inches length by the ratio of torque value observed

during later testing at 3600 RPM to the value in figure 4.1 at 3600 RPM.
16

This estimation technique may result in inaccuracies in the estimated

torque curve.

As can be seen in figure 4.3, the change in the estimated

peak torque RPM is small, from 2400 with the manufacturer's manifold
to 2800 with the tuned manifold, but the change in shape and
magnitude of the torque curve due to the tuned manifold is significant.
A significant increase in maximum torque with the tuned manifold over
the manufacturer's manifold was estimated, with an experimentally
measured increase of 28% from 40.5 to 51.9 N*m at the design speed of
3600 RPM.

This increase in torque appears to be relatively independent

of tuning length, over the lengths tested, and may be partly attributable

Torque vs. RPM

60

......... . .......... , . ... . . ...... l ........ . . .

. ........... , , , , ...... .. . . . , . . . .. . . ....... .. .. .... .. , . . .. . .

50

20

o 2 Leg
<> New Intake

10

0
1 000

1400 1800 2200 2600

RPM

Figure
17

3000 3400 3800

4.3

. . , . . . . . . . . . . .....

to reduced pressure losses in the intake system and more uniform

fuel distribution as well as to the tuning effect.

The volumetric

efficiency of was calculated from BSFC, power, speed, and air fuel ratio to
be 90% at 3600 RPM.

Intake manifold pressure data versus time were gathered at WOT

at varying RPM to verify the existence and proper application of the
Helmholtz resonance effect.

The manifold pressure was measured at the

intake runner wall three inches from the intake valve at the interface
between the intake manifold and intake port in the cylinder head.

The

results are shown in figs 4.4 through 4.8 with vertical markers at IVO
and IVC and each large vertical division representing .25 bar.

It can be

seen in figure 4.6 that at 3600 RPM, the calculated and experimental
tuning maximum, the measured pressure peaks at the same time as the
intake valve closes.

At lower engine speeds, the pressure peaks before

intake valve closing, indicating that the resonance frequency is too high
for the engine RPM and the pressure peak arrives too early.

This trend

continues to the point where at 2400 RPM an entire resonance period

can be .seen during the time that the cylinder is open.

Due to mechanical

design limit characteristics of the engine, the maximum engine speed is

limited to 4000 RPM.
intake valve closes.

At 4000 RPM the pressure also peaks before the

The reason for this is unknown.

These results and

intake port pressure plots are consistent with the data and pressure
plots of Thompson and Engleman [3] and clearly verify the presence and
proper application of the Helmholtz tuning effect.

The resonance

frequency when the intake valve is closed can be seen in figure 4.4
through 4.8 to remain independent of RPM as it is solely a function of
intake runner geometry.

18

Intake Port Pressure at 4000 RPM and Wide

Open Throttle
hp

stoppea

VIIOrker:H 1
Val'lrlol!r I C 1

2.4S3'rsv
1.

~~7~0

S tOD IOrKer:
~tl'lrt

~rkAr:

de!~~

t~

19'". 3'0001S

50000~

~.eoooo:c

ihhdi.:, t

Figure

4.4

Intake Port Pressure at 3600 RPM and Wide

Open Throttle

Ven~er:c

1 i

Vort.:er 1 < t l
del t~ V< 1 )

2. !9063' v
t .48438 .\l
906.15'>.;

stop aarl<er:
~laC'l

aarur:
\&

\. ~

~'>-.~s

llaett~

t:

9t. i'4)'t ' Pr:

~~1

Figure
I 9

r.:r. 4000iu
3 ..S00001lS

4.5

Intake Port Pressure at 3200 RPM and Wide

Open Throttle

;-

Vman:er2< 1 )
Vmerker 1 < 1 1
delta V( 1 ~

2.31250 v
1...53 \ 25.. '1
'~. ~St)t&\~

stop .er~er i
s.t~c

. aar~er.:

"&Ha t-~
i tete l t" t-:

Figure

Zi:91?Q0111S

tS . 5tjQQII s
\.1-. 400Ca.
a(>. o4S? ~

4.6

Intake Port Pressure at 2800 RPM and Wide

Open Throttle

lr At
jl

8rkers --~

.mill
-~~j ---------------off

stert aerker

- -n

1:0 ~ ~0~~0._. ._~._:._. ._.~

.._!.._
..._
, ..._
, ..._
, :

Vorr.er1< 1
IJJDorker I ( 1
jet tc V( 1

.'(\C':>O

-I

.111s

- f"----------

2s_..:_.a._~~o.'. , ~ ~ . . . ; . . . . : S~~ ~~~.; ~ ~; r.___s_t-op_ll_e_r_k_e_r-~~~ -

5. 01) as/dh
2.20313 v
stop a.arter:
1.53125 V
st~rt erker:
671.880mV
delte t:
I /tte-~ ~~ t:

Figure
-20

4. 7

27.2000iis
13 ,1)000tnS
1.4 ;201~011 s

70. 4z..?5 Itt

Intake Port Pressure at 2400 RPM Wide Open

Throttle
,_

r..~- .-.. -, .- . . ~:;-.. -... -

~'I

-~""'!'"'-:-'~~~--:--:--::~~.. i . . . . . . ... 1
.. .
.

,..

erkers -

off

lla:.

-: - - - - - -1

.. . :-:
. .

..:~- , 'o/1ar1<er 2 - . . .

W!A@e"SW!J
---------,::.-.r---,.-,--- :I 1 - - l
-iC V1arker l

'.
I
1":..
-- ,----...,..... ----:_::.a.----f
~,
" \
I ...
.:...::-----
.,
-- ....
'- j
(
t1- _____
'~,./
,__
'-J
,.
. . . . . . . . . . .!... / _ _- _ _ _ _ ,~-...._ "!:..
- ________
I

l~

I
1

~l

! _ _ _ _ __

,~

___

-~l

~C

]c

...........L..;-..-;.J...;...:...;_;._L..........
.

._.,_;_._._~ ..-L.~_..._._,_. _. - ~!
25 . 0C(H) rn
~.01)

Ml.jl\1

}r
Si~ .0C4)C ~$ I
. , . , , . , ,

va.,rk:er2( 1 )

2. IS6ZS V

$(OP

1ar~er:

40~ ()t)Ol)MS

Vmerker 1 c 1 )

1 .6rl91a '!

s.l~c . l

a.uk.er. :

2! ..4000115

delt~V<1)

546.870mV

~elta

t:

i/"et te t :

Figure

WMHiijinM9

Dt

arkers

ill

-~~..l . __
off_ _
-_
_.

:::.

;~ ,

---~

s tort oorker

-~

stop 11er1<er -

I 6. 6000ru
60.~410

Hz

4.8

The maximum intake pressure reached can be seen in figure 4.9 to

increase with increasing RPM, indicating the increased potential for
tuning at higher engine speeds.

This increase in peak pressure can be

attributed to the increase in intake air velocity resulting from the

increase in piston speed.

4.2

Throttle

Plate Effe.cts on Resonance

During mapping of the fuel injection system, it was observed that

the torque produced and fuel required increased significantly with
increasing throttle angle near WOT conditions.

The sensitivity of torque

to throttle changes is expected to decrease as throttle opening increases

for a single throttle plate nearing wide open conditions [5].
21

In order to

further investigate the cause of this effect, torque and intake manifold
pressure were measured as a function of throttle

~ogle

at a constant

3600 RPM.

Peak Runner Pressure vs. RPM

1.25

: I~--~-~ :~J~~-~--~--~ -~-J~~-==~=~-

1.00

0.75

- . ..

.. ...........................i ... .................. --~--- : .. ........... .......... .. .. -~- -- . - ... ....... .... ... .

as

CD

0.50

0.25
0. 00 -+----+------t----+------+----1
2000

2500

3000

RPM

3500

4000

4500

Figure 4.9
Throttle open area perpendicular to the intake runner tubing, calculated
from throttle angle[4], was used as an estimate of throttling effect.

The

percentage of maximum manifold pressure and percentage of maximum

torque are plotted against percentage of maximum throttle open
figure 4.1 0.

are~

in

Although intake manifold pressure stops increasing when

the throttle opening reaches 40% of the maximum area, torque continues
to increase significantly up to WOT.

22

Percent Mean Manifold Pressure,

Percent Torque vs. Percent Intake Area
100

_so
)(

cu

~ 60

~:::~:::c:~~:~r~::::~~:: ~: :~ : I: : ~:=I: ~ ~:~: : : : : : : : :I: : : : : : : : : : : : :.: :r: : : : :.:. :

4o

-
a..

o Percent Torque
<> Percent Manifold Pressure

20

..

'----~11111111!1---~----

0
0

20

40

60

80

100

1 OO*(A/Amax)

Figure

4.10

In order to better understand the increase in torque at near

constant manifold pressure, torque was plotted against manifold
pressure in figuro 4.11.

Torque increase becomes increasingly more

sensitive to mean manifold pressure increase as mean manifold pressure

increases.

For comparison purposes, a simulation of airflow rate vs.

intake manifold pressure for a 4700 cc eight cylinder engine with a two
barrel carburetor is taken _from Heywood [5] and shown in figure 4.12.
For intake _manifold pressures above 25o/o of atmospheric, airflow is
essentially linear with intake manifold pressure.
23

If the assumption is

Percentage Maximum Torque

vs. Manifold Pressure
100
80
)( 60
ctS

rrjr-r-r-r-r-~;!9.:

1-

~ 40

........... L............... L..............:................ L.............. L. ..............J. ............... ,................! ..... .......... i. ...............l
:

: ,

0: . .

............ T.............. T...............:............... T............... 1................ r ;;;o-:................r.... ......... ...............

.

j,

;
:
: ,
.
:
.
:
:
;.,
:
;
:
ii i ~ i,i ; i ii
;
:
:
:
;
,
:
:
.
:
:
:
:
:
:
: ,
:
.
:
:
~
~
:
i
~,
i
.
.
~
i
-
:
:................. .; ................................................~
:
:
;
:..................................
, :
:
:
:
;
:
i
~
! ,
~
i
.
:
,
:

20
0
0

0.2

0.4

0.6

0.8

Manifold Pressure (bar)

Figure

4.11

accepted that torque is approximately proportional to airflow, then

torque for the eight cylinder engine is essentially proportional to
manifold pressure.

This contrasts significantly with the data for the

tuned manifold and. supports the hypothesis that the throttle plates
interfere significantly with the resonance effect.

24

In order to further understand sensitivity of torque to throttle

position, intake runner pressure was measured versus time while
varying throttle angle at a constant 3600 RPM, see Figs 4.13 through
4.20.

The vertical markers indicate intake valve opening and intake

valve closing and each large vertical division is .25 bar.

Air Flow Rate Past a Throttle vs. Intake

Manifold Pressure with Throttle Angle
and Engine Speed

Throttl~

angle t/;

Intake manifold pressure, cmHg

Figure

25

4.12

Intake Runner Pressure at 3600 RPM

'./Mart<er:c 1
VarJ.:er 1 < t
de 1 u v < 1

r.t.4000ils

3 .S.OOQOas

9000$

~()...

9 t. 7~J"t ' Pr:

Figure 4.13
100 Percent of Maximum Torque

Intake Runner .Pressure at 3600 RPM

"-~'

Drln t in'J

. - - . . . . . . . - -. - - .
. -. ,-

.. .......
. .. . . .. .- ,

......
,
.-

_,_,_....__..~.~-

. -
,....,.
,
. -.-......
.

I
1

[ A'/

~-

, "'!""
.

!
. ,- : "

~. r~l

."'"!"
: - , !""!
..

\ l \ f \ (
r-.- - - - - - - - - - - - \~-

1::

__,

~-~

/~ :

r\.
~\

.J -

./'

\..

~-

._./

.; ..___ _ _ _ __
1

-~[ 1

1~~~-C

at oerkers

r
..__._...._
.

.. .

;
. _;
._

0.00000
Va~rker2~

Verker

. . .

t l
del to vc 1 )
1(

!;_

off

:=:

~j

stort.pp!;p

i- .

_. .

~~ ~.;_;.,. :io~
- .............._.;......_..~....._....so.... ~co
. ............~; 'u

.._2.._5.......
.

.i..
'

=
-., . _.....
.
... . .

'*Ml

:1 V1ar1cer

;..

'
' -

.:;

;:_:\---- -;.:3.( Verker

*MHM" }

.._;

--

IDa.

l:.
:
~r"\,----- ~- --- ~~- -- ;.:.:c--------- ::r"~---

ar~ers

oft

"!"-:
.

o. o..;._
o , . i .

....

s. c~ ..s/dl '~"
2 ; S437S' 1i

s too 1arker:

1 . 453n 'J
591). 62C~~t

~.tart sac~.e.r :

~4.!1)0015

oe H.~

t.:

2.3 . 40.01)&.~
t ()., 9JYJOit e

tiOelt~

t :

91.7.:.131 Hz

Figure 4.14
92.8 Percent Maximum Torque

26

top

HMw. s J

Intake Runner Pressure at 3600 RPM

~~-~-~-~-s~-t~-o~-c~c7e~d~~~~-~-~-~;~,~-~-~--~-~~-~-~-~- ~- ~- ~- ------~~~---,-,,~[!' - A~f~erkjiH

...
---- - - - - - - - - - - '1
t1
:
. :1(- vaen<er z f~~-- -?.---------/---r-\----;;:.~ --- ~;.:.:~ ---~;:::.,-- ---- --~r , .....~ I
I

..___,

1:

..

'

/'

....

""-"

...

.., - - - - - - - -

..,-1-

\lllat'~Q.r.

r----:-'-.---r-------------~]1 1

miWfl
1

~~ At erkers

:-

otr

-=i

II]

--

- - ------...J-~

~-r

}l

~:Ooi~o .

' . ; '

2S:~~o0 .~.......Lw. ' ' SQ:~ ~~~stop

s. C'>
V111~n:er~(

Vaorker 1( 1
del

t~

V< 1

:2o563 v

as/oh~

*M'S''.# .

;,m;.+]

uop erker:

19. IOOOms

s.ter.t. a.ark.er:

e . :oo.o.o u

612.500V

deltt t:
1/de 1 ta t :

tO. 9000115
91.7.:131 Hz

1 .:15313

start aarker

Figure 4.15
85.0 Percent Maximum Torque

Intake Runner .Pressure at 3600 RPM

lv:

stocced

~----~~- .,. ----~--. -.. ,, ~

. ~~
, -.T.~: .

' : , -.-.-.~~-.-.~,--~-r~~~

i-

l:

~k-'-l_ :

o. 00_000

45". ~::10015

vorker:.!< f >
Ve.rker 1 c. '\ )
de 1 te v c ~- )

34.d.OOOIL$
lt.\.94)~~9

9T. ?4'3 t Wi

Figure 4.16
68.9 Percent Maximum Torque
27

'"--------

Intake
Runner Pressure at 3600 RPM
hp stoppeo
~-

... ,......... ,..,.. ,..""'~ ...,...... ,. ,... ,..,':!l4~f~~rlri!&iill

lr
vaerker
~ ~ pn--=s
1-: ---- , - ---------~----~--------------------------./,
\
?-r
f ___ :_':.~.)_-' -~- ____________ ~:1r
1-

:2 -

!:

!;

:,--

('-

jr

~~

:l

.lt ftr'~&T"S - -

off

--

~~:...-------
aa.rker t f
MMi.NIDMi J

,........_____

sfort oorker

' ,,,,,,,.e
______

I
. .. , . ... , - .. ~_.!f._s..lo~ er~r --~
L.;...w_.J~..L-:.~------~~~ ..
__,
0.0000()
JS.CJQOO S
so ~004')0. ~ ..
5. (It) lt~.i~l y
Verl<er2< 1 )
I. 79666 \[
stop aer:er :
T7. uOOOMS
Vaaer.J.~er I( 1)
t. . .\6.750 'i
~tart a.~r.k.e.r:
6.. \OOOO&S
delt~'!C1'~
61.\9-. 38(\M.y
ot l \ & t ~
H~. ~~') .. :.

!"

..

llct!lt~t ~

9l.7.Cl it:

Figure 4.17
59.3 Percent Maximum Torque

Intake Runner .Pressure at 3600 RPM

stop pea

hp

~r- ,

,.
f-

., .

a.,

. : I l . i I . I ! .. [
1

; I

-lr

:_ .

i= 1

~~f~elr"~-

!>'----------'

-t..

Vaerker 2 - - . .

~t 1

wwwp#q j

~\
~,
1-~.
1
.~.-- I'----------~'-----------------------!...----------~-"
,,
,..
.,
/~
I *.Jaa.r~ar 1 ~
1 IMI@EFJ
- I
...
I
~

:-

-.

.:' :

,I

',..

'-

-;I

~ - 1 - - _;:_,_ - . - - - - - - - - - - - - - - - ......_.-- ... .:?-------~

-j

~~~ ~~ _er~~S
l or f - - .

;: I

Ei!..

~:-

;-

-:...:

~.~ ~ 25 .

i . . : . ;
0.000(10

.....;.....;'.....;_;...;.....;:

': .

Verker:.:: o: f :.

Vaorker 1( '\. ,
de I t ~

iG
1r

v< 1

.1

.........
. ....
. - ........
. .....
. ......_. .............:.........;....o......;......;.....;'....
' ' ' ' ' '

5.
1.67155
1. 1562-S 1/

5.0.. 000.0

I)Ot)O "S
O.i s-l~t\1

"

~15.S'30aV

stop ter"9r:
~t~t

llar~e.r:

:j91 \ ~ t:
i .'aeJ-t~ t !

t3.

..S. I

zovus

2..30000~
l-e . ~-~

9l.

i'~~l

W:

Figure 4.18
53.9 Percent Maximum Torque

28

stort oerker ~

ste>p

oarkor

ill
I

Intake
Runner Pressure at 3600 _RPM
hp stopped

e. '))
1 1
V111arJ.:~r 1 r 1 )
de I t.: 'i ' 1 ' /

V11en:er~:

:r,-.~

r. SoZSO V
1 . t:2.50l 'i
4~'? . ':itt)'v

ct-!..,
r9. !OKI'II s

1an:er:
ea.r1c.11r :

stop
-.lac l

i3.~ .:lOCIOO~~

de- I t ~ t :
~ /ae

t'~

H~. 9-f)(ll:.\$

?r . 7 ~ r ifz

Figure 4.19
50.3 Percent Maximum Torque

Intake Runner .Pressure at 3600 RPM

stop pea

hp

r '

" ' T , f' ' ; ' , f ' , f , ,, ; , , '

l ,

,- < , , f.' ,

""1~- .:l~(~<rkilinill

~~

-=i'r

!::

JL ,

~:

-:\

~/."::'--

vaarker 2

;;;;;;!

"'G!*'"' I

J~
'~ar.kQr\
:-7)
.. f 1
'5C'f'nerfEF"' I

..,

~:~ - -- -- .J.-- .._,.::: ~./: --- - --- -- --- --~_: .

t- - - - -

;:
~~-

;-

..J -

:'

---..:.. - , - -

::

-~

:..

V~r~~ .;r2( 1
Vll.,rker 1< 1

delt~V(1

.. . , . .

z~)._

5-. ~tC)
f . ;ZQ3'T3 V

921.. 875~a.V.
2Sl . 255-!!P'J

ItS fd

, . .. ,

. .

. ,

. .. 1 ..

stop
t.

~t.~r

14r~er:

zo.2,)1Jvms

art~er :
~1-H~ \~

9. 300)0a.~
,0 . 9(1001U
~ r . 1 .:~ r If:

L'ar: 1 t ~ r: :

Figure 4.20
37 Percent Maximum Torque
29

off

--

..

s t er t e r k e r

-J

iE!:I:fliAf

-------~

_::~r tt'lp erker ;:;;;!

..~ . - - - ---~~oo-;;-l
i ~:

_______I

:;l
-:...._

-jr

,..

0. OOiOO

- ~- .:~t :sr~ers

~
- - - - - - - -

The amplitude of the intake pressure wave is plotted against

percentage throttle open area in figure 4.20.

Even when mean manifold

pressure is relatively constant above 40 percent throttle open area, see

figure 4.10, the amplitude of the pressure wave increases with
increasing throttle opening, presumably due to the throttle plate
interference with the pressure waves.

Port Pressure Amplitudes

Throttle Open Area

Intake
0.8

:r:rr:::::r::::r:::r:::r:::r;rrJ::I: J::r ~ : _! r:,

l ~ l i ! l ! l l ~ l \ ! l i ~ ! i :
+! T J tr j t" t i : !

0.7
0.6

:.

.. -~ ....... ~ ....... -~ ... ! ~ .... ~ ......~ ....... !........-~ ...... ~-- ... ~ ... ~- ......-~ ....~- ....-~ .......~ ..... '.-~ ....... ~- ...... -~ ...... ~

0.5

l -~- .o -~-- -1 @ f.\ ! ......1- ..f.~

T o .., .. , ~t ~ -\ , ~ ' Pmax' - Pmin ,
j:I:r:::r:I:::r::r:I : :~~n- Pmin

vs.

0.4
0.3

0.2
0.1
0.0
0

20

60

40

80

100

A/A max

Figure

4.20

Below approximately 40 percent throttle, the pressure wave amplitude

begins increasing with decreasing throttle open area, opposite from the
trend established at larger throttle openings.

30

This increase in amplitude

can be seen to be due to the reduction in minimum pressure as seen in

the minimum intake stroke pressure also plotted on figure

4~20.

The

reduction in minimum pressure is due to the throttle plate choking

during the inrush of the large volume of air during the intake stroke.
The difference between pressure at IVC and minimum pressure is also
plotted in figure 4.20.

This difference can be seen to converge with the

amplitude data for large throttle openings.

Figures 4.15 through 4.19

show that the reduction in positive pressure at IVC is due to the throttle
plate damping the intake stroke and delaying the positive pressure
wave until after IVC.
One can make the analogies of the volume of air rushing in during
the intake stroke as a mass, the volume in the cylinder as a spring, and
the throttle plate as a damper.

Accepting these analogies, a throttle

open area setting slightly less than 40 percent appears to be the setting
for critical damping, with a corresponding phase shift seen in figures
4.12 through 4.19.

If it is assumed that the resonance wave centers on

the mean manifold pressure, the positive amplitude of the pressure

wave divided by the negative amplitude of the pressure wave will give
an indication of the magnitude of the damping effect of the throttle plate
on the efficiency of the supercharge.

Figure 4.21 shows that at WOT, 90

percent of the negative pressure wave amplitude is reflected in a

positive pressure wave one half a resonance period later and that nearly
100 percent of it is used to raise the intake port pressure at IVC.

That

response falls to 70 percent of the negative pressure wave reflected as a

positive wave at 43 percent throttle open area, but with the maximum
manifold pressure approximately equal to the pressure at IVC.

Below

approximately 40 percent open area, the positive pressure wave falls to

31

approximately 30 percent of the negative pressure wave, while the

percentage of the negative wave reflected as a positive peak peak at IVC
continues to fall and becomes negative.

This is because the phase shift

due to damping is actually so large that the intake valve closes before
the pressure reaches the average value.
The damping of the amplitude of the small pressure waves on the
closed intake valve can be seen on the moderately throttled pressure
traces, figures 4.17 and 4.18, providing further evidence that the
throttle plates interfere significantly with resonance waves in the intake
runner.

Pressure Amplitude vs. Throttle Open ArE I

0
0

-~
:;:,

o (Pmax-Pmean)/(Pmean-Pmin)
<> (Pivc-Pmean)/(Pmean-Pmin)

100

tt:! -;1 1-t[~

80

: 0

- - - r- r-- r r -r -~-- - - - - - -- - -1

-r .. r -------:

c:

tt [l6~j Lt[ .........:

::..~..:::r:::::::..:::::t:: . .:. . . . . . . .:.:.::r . :...: ...:.:.:. : ........::r:..::: ......::r..:..:.:::.::.:::.:::...::......:...:...:....... ,

g> 60
Q)

ctS
0>

40

::::::::r::::::::::r~J::::::a:r::::c::::::r::r::::-r :

20

-+ ~ + -o1 - + ! -- - -- -- - - +- - -- + ~ :
~
i
:
~
~
~
~
~
~
~

Q)

--1-rrT------r-r----rr-r------:

"0
:;:,

0>
ctS

-~ t r -- ; ! 1 ~- f - -- r - :

l
i

0....................l

~--- -;

~ -- - -

~
l
:
:
j
:
............... ~ ................. -- ~ - ......... ,.. <\.................;. ................ ;

~-20

r..fo++j+-11-r---i

>

-1---r-r--rrrr-1..~ :

Q)

=~-40

0.

20

40

60

Percent Intake Area

Figure

32

4.21

80

100

Average intake manifold pressure and cylinder port pressure at

IVC are plotted against percent maximum . torque in figure 4.22 in order
to better verify the effect of manifold resonance and the correlation
between cylinder port pressure at IVC and torque.

At 60 percent

torque, average manifold pressure becomes relatively independent of

torque at a value slightly below atmospheric and also crosses the curve
of intake port pressure at IVC.

Intake port pressure at IVC continues to

rise up to maximum torque with a maximum observed port pressure at

IVC of 1.19 bar, see figure 4.22, indicating significant tuning effect.

For

the test engine intake port pressure at IVC seems to be a better

predictor of torque than average manifold pressure.

This . relationship

Mean Manifold Pressure, Port Pressure at

vs. Percent Maximum Torque

..... i

1.2

r:;~D~;

............................L................................J.................................J............a. ~. ~.......J................................. !

1.0

...co

.i. .

.. .

i
~

1
,
~1~~0
:~- V1
1

0~
/"

o'

l
~

0.8

--~----~~r-~:~or----rl

0.6

. . . . . .... . . /-f.. . . . . . . .. .. . . e;.;. .. . .. . . .... . ...... . . . ,...... .. .... . . . ... . .

.0

'l

!
l

o Pressure at IVC
o Mean Manifold Pressure

0.4
0.2

J.... ..... ................ ...... .

-+----+----+---~----+----1

20

40

60

Percent Maximum Torque

Figure

33

4.22

80

100

validates the use in section -4.1 of manifold pressure at IVC due to

resonance as an indicator of the effectiveness of helmholtz tuning.
Based on the preceding data, it appears that the throttle plates in
the individual intake runners act as a normal throttle to reduce average
intake manifold pressure, but also have a wider range of throttling effect
due to their damping of the intake manifold velocity wave during the
intake stroke.

34

5.0

Ethanol

Conversion

Results and

Discussion

In order to convert the throttle body intake system to ethanol

operation, the fuel pressure was increased in order to increase the fuel
flow rate.

When port fuel injectors were used for the new intake

system, the fuel injectors were sized for the required higher fuel flow
rate at their rated pressure.

5.1

Spark Timing Comparison Between Gasoline and

Ethanol

In order to achieve peak efficiency and maximum power while

operating with a given fuel, spark timing must be optimized.

In order to

determine optimum spark advance for gasoline and ethanol, the test
engine was run at the design speed of 3600 RPM, stoichiometric air/fuel
ratio, and WOT while varying spark advance and measuring torque.
Figures 5.1 and 5.2 show the results of varying spark advance B_efore
Iop Jlead Center (BTDC) on torque output.

The torque axis is truncated

From figure 5.1 and 5.2, M.aximum B..rake

to better show the effect.

Torque (MBT) spark timing at stoichiometric equivalence ratio is 12

degrees BTDC for E95 while gasoline has a knock limit at 16 degrees
BTDC.
advance.

Higher octane gasoline may have a MBT value at a greater spark

Since the laminar flame speed of ethanol is slightly lower than

that of gasoline [6], the effect cannot be attributed to a shorter ignition

delay for ethanol and must be attributed to more a rapid flame speed
during the turbulent phase of combustion resulting in a more . rapid rise
in combustion pressure.

This may have beneficial effects on emissions,

as discussed in section 6.2.

Gasoline spark advance is knock limited in

35

the test engine as MBT is approached, at WOT and stoichiometric air/fuel

ratio using 92 pump octane ((research + motor)/2) unleaded fuel.

This ts

not unexpected, as WOT at a stoichiometric air/fuel ratio is a severe

operating condition and the test gasoline has a pump octane rating of 92
compared to E95's pump octane rating of 98 [5].

Torque vs. Spark Advance

Ethanol Fuel
55
53
z51

Q)

::J

5 49

....

47

8
12
Advance BTDC

Figure

16

20

S.l

Since the engine as supplied by the manufacturer operates at an

equivalence ratio of approximately 1.2 and a spark advance of 21
degrees at WOT, the original knock limit may occur at a greater spark
advance due to the ~icher equivalence ratio and due to lower peak
pressures resulting from the reduced cylinder filling efficiency of the

36

original intake manifold.

The severe operating conditions of the

-

modified test engine are required, however,

for high power levels and

effective catalyst action while operating on gasoline, as discussed in

section 6.1.

Corrected Torque vs. Spark Advance

92 Pump Octane Gasoline
Fuel
.
55
53
~

,:
:
:

:
:
:

~I~

~~~

:
:
:

:
~

:
:

:
:

:
:

::

- i - ! - ' - jt~--~-~- ~-- - ~t~-l-~6~-- ~ 6,~ '_!

51
z

+1: rr1+-;,.,(-~++ri:ir ~
:

.:
~

:
:
j

: l

47

::

... ;,::

; ...

... ~,
:

:
:

...:...
:
1

::
j

.. l i

45
0

:
:
:

:
:
j

::

l : : l : i : l : : : l
:

tj :;;w!' ~ tj t ~

::
:

::
:

::
:

::

::
:

. l
:

..t .... ~ t ~tj~~t j

:
:

8
12
Spark Advance (BTDC)

I
16

KL
NI

OM
CI

KT
Figure

37

5.2

20

I
1

5.2 Thermal Efficiency Comparison of Gasoline

and

Ethanol

Brake specific fuel consumption varies with equivalence ratio at

WOT conditions with fixed spark advance.

Because the equivalence ratio

was calculated from exhaust port emissions which were measured

. independently for the two cylinders, an average equivalence ratio for
the engine is calculated by averaging the equivalence ratio of the two

cylinders while assuming constant airflow volume to the two cylinders.

In fig 5.3, WOT BSFC at 3600 rpm for both ethanol and gasoline is
plotted against the average equivalence ratio of the engine.

Although

WOT operation allows both power output and BSFC to vary with
equivalence ratio, WOT represents the design condition for HEV
operation and is therefore the best test condition for this application.
BSFC of ethanol was measured at spark advance settings of 10 degrees
BTDC which was the MBT spark advance when the tests were
uncorrected for atmospheric conditions, and at the manufacturer's
nominal advance of 21 degrees BTDC.

Increased spark advance can be

seen to increase torque and reduce BSFC when running at lean

equivalence ratios.

Gasoline BSFC was measured at 10 degrees advance

to provide a direct comparison with ethanol, although BSFC was also

measured with gasoline knock limit timing of 17 degrees.
was measured over a narrower

r~nge

Gasoline BSFC

of equivalence ratios because the

engine began to misfire and could not maintain the dynamometer speed
at 3600 RPM when running at equivalence ratios leaner than .85.

The

minimum observed BSFC of ethanol was 372 g/Kw*hr and occurred at an

equivalence ratio of .8 and a spark advance of 21 . degrees.
38

The

minimum observed BSFC of gasoline was 263 g/kw*hr at an equivalence

ratio of .92 and a spark advance of 17 degrees.

500
400
~300

.:.::.

........

~200

100

0.5

0.6

0.7
0.8
0.9
1
1.1
Average Equivalence Ratio

Figure

1.2

1.3

5.3

Thermal efficiency was calculated from BSFC and the heating value
of the fuel and plotted in figure 5.4 in order to get a non-dimensional
comparison between ethanol and gasoline.

At an equivalence ratio of

.91, the calculated thermal efficiency was 33.5% for ethanol based on a
lower heating value of 27.8 MJ/kg for a 95% ethanol/5% gasoline blend
and 31.1% for gasoline based on a lower heating value of 44.0 MJ/kg [5].
The thermal efficiency while operating with ethanol rose to 34.9% at an
39

equivalence ratio of .79 with a spark advance of 21 degrees BTDC.

The

higher thermal efficiency achievable when operating with ethanol

relative to gasoline is the result of the combination of a higher thermal
efficiency during stoichiometric WOT operation and a further increase
due to the ability to run at lean equivalence ratios without severe
misfire where higher thermal efficiency is achievable.

Thermal Efficiency vs.

Equivalence Ratio
. . . . . . . . . . . . . . . . :=r.:=o::::od~--~~-=--~T-~---------T-------------------

0.35

<:5 ;

g0.25

::::J::::::::

~:

... ,. ...

...

Q)

=
w 0.20
0

cu

E
~

" :

0.15

Q)

:::::::-: .
-

0.10

.- '
:

~:: --~t~~n:~,:-~ff:1:

<>

.s:::.

~6....

...... ... }6~ -- -1~ 'T.~

: 01
:
l
:

0.30

Ethanol Eft 21

0
Gasoline Eff 1....

! ~ ~

.......... :. ............,......... .,~-~-G_a_sol-in_e_E.ff-17

0.05

0.5

0.6

0.7

0.8

0.9

1.1

Average Equivalence Ratio

Figure

1.2

1.3

5.4

Operation at lean air fuel ratios results in some loss in maximum

power output.

WOT Torque is plotted versus equivalence ratio in figure

5.5 at different spark advance settings.

40

Operating at an equivalence

ratio of .8 instead of 1.0 results in a decrease in torque output of 8%

from 52 N*m to 48 N*m, however fuel consumption falls from 400
g/kw*hr to 372 g/Kw*hr, a decrease of 9.5%.

Torque vs.
60

Average Equivalence

::: -:-. ::

Ratio

:
: ..... . ........ :: ....:

t j t t : j ~

...................:......... ..............

: ~ t .

.................... ) .......... ............. , ...... ..... . ........ . .. .(......................... ...... . . .......... .. j ... . ....... ... .. . ......... ... .. . .... .. . . ...: . .. . . .. . . . :

::

: : : :n: :~: r: : : : :~,~!~ ~ :~ ~: : :-: :n : J: : ;: : : i

30

20

: : : -: n:J: : : : : ~: : : : : ~ ::~:~:: ~:~:~: ~~

10

..................:.............. ......:...........
. . ... . ....... t....................j...........

..................f................;~+. . . ... . . . . . t...................+. . . . ... . . . . ;. . . . .... . . . . j.................... ~ ..... ..............:

..... ............. ; ..... ..... ....... ... j, ..... ... .. .. .. ..... ~ ... ............ .... ~ ...... ... ........... : ...... ....... ....... : ...... .............. ~ .... ..... .... ..... .:

::::::::::~I:~~::~:~::::::r:::::::

Gasoline Torque 10

b.

GaSoline Torque 17

0
0.5

0.6

0.7

0.8

0.9

1.1

Average Equivalence Ratio

Figure

41

S.S

1.2

1.3

6.0

Emissions

Results

The uncatalyzed emissions of the engine were measured directly at

the exhaust port of each cylinder.

The basic exhaust gas composition of

the engine consists of C02, CO, and 02, with small percentages of HC and

NOx.

Primary Exhaust Constituents

vs. Equivalence Ratio
15

r:r

--~1 0
>-

"C

::::J

>
>.c

$+

'

6..

:~~

:~~

1~

:;i fiJ

........,... ~d ..........;............... .... . . . . .. :l ..................l

ltf.~

....... ......) .......... ~1-~l... ................l ....................L.. . . . . . . . .J. ................... ~

~-- ,- ~ - ~--- :

Q)

a..

l
I

tr

: .- ,

~;

0.8
0.9
1
Equivalence Ratio

1.1

1.2

1.3

lSl

co 1 GAS 17

eJ

C02 1 GAS 17
02 1 GAS 17

0
0.5

0.6

0. 7

CO 1 ETH 10

e CO2 ETH 10

<>

C02 1 ETH 10
02 1 ETH 10
CO 1 ETH 21

+ C02 1 ETH 21
6 02 1 ETH 21

C02 2 ETH 10
02 2 ETH 10

CO2 ETH 21
e C02 2 ETH 21
02 2 ETH 21

Figure
42

6.1

C02GAS17

o C02 2 GAS 17
0

02 2 GAS 17

6.1

Primary

Emissions

The primary exhaust components C02, CO, and 02 are- plotted vs.
equivalence ratio at WOT in figure 6.1 with different indicators for both
gasoline and ethanol, for both cylinders, and all spark advance values.
The emissions of these components appear to be essentially independent
of fuel type, cylinder, and spark advance.

The data follow expected

trends and correlate almost exactly with the data published in Heywood
[5] and shown in figure 6.2.

Primary Exhaust Const.i tuents

vs. Equivalence Ratio
0.16
0.14
... 0.12

--AOO
x+a

DltL
St, Sp
HitS

_,?- ~ ~
,..~6

MIT

::

rS

0.10

8
g 0.08
.I

..c
~

_, r6

#c~

~00

~ ~~0

~ 0.04

~~ 0
0~ ~0> ,.,.

0.02

<

H/C- 2-2.25

o..s

~ ~6. . . .

0.06

0.00

..

I.

0.6

0.7

0.8

---.

vo/

-~

0.9

1.0

1.1

IX~. '...;
.

,'

ooo../c ,.,,

~ ~,d ,.,o

/~

,_, ..,'

,.. , "'

, ..Hl

oO
oo

1.2

1.3

1.4

l .S

Eabault equivaJenc:c ratio

Figure

6.2

Since HEV architecture allows engine operation independent of

vehicle demands, viewing emissions in a brake specific format provides
a better understanding of the compromises involved in minimizing
emissions.

BSFC values and air/fuel ratio were used to convert from

43

percentages to units of g/K_w*hr.

This is a better primary measure of

emissions for series hybrid operation.

From the results plotted in figure

6.3, it can be seen that CO emissions are minimized by operating at lean

equivalence ratios.

While operating with ethanol fuel at the two

equivalence ratios where thermal efficiency is maximized for a given

spark advance, C02 emissions were approximately equal at 58000
glkw*hr. Due to the fixed carbon content of the fuel, if CO and HC
emissions are small and consequently ignored, spe9ific C02 emissions are
solely a function of specific fuel consumption as effected by optimization
Specific C02 emissions while

of spark advance and equivalence ratio.

Primary Specific Emissions

vs. Equivalence Ratio
80000

~ ~~ ~ : -

:.

.~

T
.

60000

umo

-...

.:

.:

-~<>1(:.-r
-

+ lu+u:~~~" : ~uu m'

. . .+ .

oOuooou

.:

.;

+u,

.
:

u Oro;

uu

o C02 eth 10

.
:

o CO eth 10

b.

02 eth 10
CO eth 21

~40000

..............:.........._ ... :.................. j.................. ).................. f"""""f""'"'

+ C02 eth 21

......

. . .. . . . t"! . . . . . \ \ 1\~ . . . .

,::,t

C)

= ~

. . . ... . . ll. . . . . . . . . T~ .................!i. . . .

! 02

m'

............ ... . . .Lx:,; u~- ..." .J

=

20000

......... ....t.................~. ..........~ ..., ........ .: . -~-,~---; ........

:

0.6

0.7

0.8

~:

02 eth 21

o CO gas 17

C02 gas 17

0
0.5

0.9

Average PHI

Figure 6.3
44

1.1

1.2

1.3

1.4

operating on gasoline are slightly higher at 62000 g/kw*hr due to the

lower thermal efficiency of gasoline.

Oxygen emissions are not - a

pollutant and do not need to be minimized.

6.2 HC and NOx Emissions

Modern three way catalytic converters are capable of reducing
both HC and CO emissions via an oxidation reaction and NOx emissions
via a reduction reaction.

High efficiency for both

r~actions

is restricted

to a narrow range of equivalence ratio around stoichiometric with lean

air/fuel ratios resulting in low NOx reduction conversion efficiency and
rich air/fuel ratios result in low HC and CO oxidation efficiency.

For an

aged catalyst, the conversion efficiencies of HC and NOx are about 90% at
an equivalence ratio of 1.001 [5].

Peak efficiency of a three way catalyst

is restricted to equivalence ratios of .999 to 1.007 for a window of 80%

efficiency[5]._ This operational range can be broadened to about .06
equivalence ratio units by cycling the fuel flow at .5 to 1 hertz [5].
Unfortunately the equivalence ratio for optimal catalyst efficiency
does not correspond with the equivalence ratio for minimum BSFC and
also results in high pre-catalyst NOx emissions.

Due to the lean

flammability limit of gasoline, engines cannot operate at sufficiently lean

equivalence ratios to lower pre-catalyst NOx emissions.

Thus, high pre-

catalyst NOx emissions and higher than optimum fuel consumption must
be accepted in order to achieve low post catalyst emissions of HC, CO, and

NOx.
Part per million emissions of hydrocarbons and oxides of nitrogen
were measured at WOT at spark advances of 10 and 21 degrees for
ethanol and 17 degrees for gasoline.
45

Emissions were measured from

each cylinder independently.

Percentage hydrocarbon emissions are

plotted in figure 6.4 against equivalence ratio.

Percentage hydrocarbon

emissions during ethanol operation can be seen to reach a minimum at

an equivalence ratio of approximately .8.

Increasing spark advance

reduces the hydrocarbon emissions at air fuel ratios leaner than .8

may be attributed to increased combustion time and efficiency.

This

In the

test engine, the hydrocarbon emissions of cylinder number I are

approximately 50% greater than the emissions for: cylinder number 2.

Hydrocarbon Emissions
vs. Equivalence Ratio
2500

--,-----....,-- -.,-----,r-- ................. .

l-~l+f+~-1

::::==~:=:::::::::::t::::::::=::::j::::::::==ff.:::~=:=:=i::=:::::::=:t::=::::::::::t:::=::::::

o HC 1 ETH 10
o HC 1 ETH 21

::::::::::::f:::::::::::::t::::::::::::::t::::::::::::::t::::::::::::::i::::::::::::::t::::::::::::::l:::::::::::

. 4 -l. i .............. .;. .............. t............. .;. ..............i .......... .

~-~ ................!+!+!

2000
c
0

~1500

HC 2 ETH 10
HC 2 ETH 21

:~-~~I::;:::!~::::;:;l:=:~~[~~~~t:: *
:~~~~i~]i~l~-~!~]~I~:~:::T~~:~:;;I::;;~;;::J::;:~~~:J
l:l

+ HC 1 GAS 17

Q)

c..

~1000
cu
c..

500

..........t ..~ ..L.............L. ...........l0 ..--~L. ............L............:...........

HC 2 GAS 17

:::::=:=~:::::::::=::r::::::::::=t::::::=::=:L:::::=:::i=::::::::::::r:::::::==t:=::::::::::L::=::::::~::::::::::=::j:::::::::::::j

0
0.5

0.7

0.9

1.1

1.3

1.5

Equivalence Ratio

Figure

6.4

The reason for this is not known but may be due to fuel injector
placement and fuel vaporization, operating temperature of the cylinder,
exhaust valve sealing, or wear condition of the cylinder allowing more

46

.,

,,,

oil to remain in the combustion chamber.

Hydrocarbon emissions when

using ethanol as a fuel were found to be approximately 50% of the

hydrocarbon emissions when using gasoline at the equivalence ratios
tested.

One possible reason for lower HC emissions is the oxygen content

of ethanol promoting more complete combustion.

The reduced spark

advance required by ethanol relative to gasoline implies more rapid

combustion and that the combustion is more complete while the
combustion temperatures and pressures are high. . These conditions
would promote more complete combustion.

More rapid combustion

would also reduce NOx emissions if the shorter duration of high

temperature and pressure due to reduced spark advance outweighed
any increase in peak temperature and pressure.
Percentage emissions of oxides of nitrogen are plotted in figure 6.5.
NOx emissions can be seen to peak at an equivalence ratio of
approximately .9 and to increase with increasing spark advance at . rean
equivalence ratios.

Increasing spark

advan~e

results in increased peak

pressures and temperatures, increasing NOx emissions.

Cylinder 2

emissions of NOx are approximately 20% higher than for cylinder 1.

Since equivalence ratio is calculated independently for each cylinder and
used as the independent variable for the plot, emissions

~ariations

solely

due to cylinder to cylinder variations in equivalence ratio should fall on

the same curve.

Since cylinder 2 can be seen to consistently run at a

leaner equivalence ratio than cylinder 1, it may have a higher

volumetric efficiency due to variations in port design and intake
manifold fabrication.

This higher initial cylinder charge would result in

higher peak pressures and temperatures.

Cylinder 2 also has lower HC

emissions than cylinder 1, possibly indicating better sealing of the

47

cylinder and further promoting higher peak pressures.

NOx emissions

-are below 1000 ppm at equivalence ratios below approximately .65.

en~ugh

Further optimization of spark advance may lower NOx emissions

so that reduction by catalyst is not required, however MBT spark timing

for a given equivalence ratio results in higher pressures and
temperatures and therefore higher NOx emissions.

One disadvantage to

operation at this point is a reduction of approximately 25% in power

output relative to MBT timing at an equivalence

ra~io

of 1.0 as seen fig

5.5.

NOx Emissions vs. Equivalence Ratio

5000

l
:

1
1

1
1

... \ :

jt7:.......~ ................1j

3000

I // j , .'\.\ ! b.
. . . . . . . .). . . . . .. . . . . ti . . . . . . . . .
...i........
+
!
l:t v~_:,,-e~'Q.,\~\ *

c:

Q)

+* :

I:
~

a..

t.. ......-~ ........~ ...~ ..

NOX2 ETH 10
NOX 2 ETH 21
NOX 1 GAS 17
NOX2GAS17

A...!---~-----~\-~t~~---~--~r~-~--- ~--~-T!I!I!.....
I!I!I!I.....!I!I!I!I~....~.....~ _ _...

2000

a..

~;

[ ','

. . . . . . . ./. . . . . .;:. {------/L.L___________L... . . . . . . . ..~--~~~~~J- . . . . . . . .

1000

<f; ~

I/~(/

0
0.5

0.6

0.7

0.8

0.9

1.1

Equivalence Ratio

Figure
48

<> NOX 1 ETH 21

4000
.Q

C'O

,'

.6. ~

o NOX 1 ETH 10

11r--11

6.5

'1'

I~

1.2

I
1.3

In order to give a better measure of emissions for HEV operation,

average brake specific emissions for the engine were calculated from
BSFC, average air/fuel ratio for the cylinders, and average percentage
emissions.

Specific emissions of HC and NOx as a function of equivalence

ratio are plotted in figure 6.6.

Ethanol can be seen to reduce maximum

NOx emissions by 29% from 19 g/Kw*hr to 13.5 g/kw*hr when operating

near stoichiometric.

This reduction may be due to the significantly

higher heat of vaporization of ethanol relative to gasoline lowering the

peak temperature of combustion.

The energy required to vaporize the

ethanol in a stoichiometric mixture of ethanol and air is 3.89 times the

energy required to vaporize the gasoline in a stoichiometric mixture of
gasoline and air[5].

This energy results in a temperature drop due to

adiabatic vaporization of a stoichiometric charge of 21.4 C for gasoline

49

Specific HC and NOx Emissions

vs. Equi.v alence Ratio
20.00

::f:s..::o

I
I

~l

1
i

o HC eth 10
o NOx eth 10
HC eth 21
1:1 NOx eth 21
+ GasHC
x GasNOx

ll

i '.

!:

lx
:

I\

l:

:,t-~-t

!;

!i

l
: I

~~

\ l

:i ~\ :i

:i

-----~--------Y---JL._ _ _ _ _ t------~-1~---~---

.s=

.......

:
:
........................................
l
l

15.00

II

10.00

.~
i

C>

/i

\l v

!: G>:/
+ + - - . 1'. 'P i
i
'
:
:
:\
----+--~:-J-:---1-------------~::i------1

5.000

i
=

0.5

l
~
:

<1

0.6

0.7

""G .. -

o :';jl""'\_
l::.T - - - :i~ -e

::
;

:1;

:i
!

0.8

0;.9

1.1

1.2

1.3

........

Equivalence Ratio

Figure

and 76.5 C for ethanol.

6.6

The reduced NOx. emissions may also be due to

0

the more rapid combustion and reduced spark advance which may
reduce the time spent at high temperatures and pressures.

Specific NOx

emissions have a much higher peak for ethanol when operating at 21

degrees advance, but specific NOx emissions below that of gasoline can
be achieved by operating at equivalence ratios below .7.

Specific HC

emissions of ethanol are approximately 50% of those for gasoline over

the ranges tested, and a minimum of 2.8 g/Kw*hr can be achieved with
'

ethanol vs. a minimum of 7.5 g/kw*hr for gasoline.

50

An alternative operating point when using ethanol in this

application is at an equivalence ratio of approximately .68 to .75 and a
spark advance of between 10 and 21 degrees.

This region of operation

has low BSFC and low uncatalyzed specific NOx and HC emissions.
Increasing spark advance from 10 degrees to 21 degrees reduces BSFC
and increases torque at the penalty of increased NOx emissions as can be
seen in figures 5.3, 5.5, and 6.6.

Reducing equivalence ratio results in

lower NOx emissions but increased HC emissions and reduced torque as

can be seen in figure 5.5.

There is an excess of oxygen at this point,

allowing easy oxidation of HC emissions if NOx emissions are already

significantly lowered.

Further testing in this operating region . with

varying spark advance may provide a better compromise than the spark
advance values tested.

51

?.0

Conclusions

The simple two leg manifold and throttle body fuel injection
system supplied by the manufacturer have a number of disadvantages
for the test engine.

These disadvantages include fuel maldistribution, up

to approximately 20%, and a reduction in power relative to a tuned

intake manifold.
Helmholtz theory accurately predicted the length of intake runner
required for maximum tuning at the design RPM.

No further

information about intake manifold characteristics was required and the

results were obtained with simple calculations.

The complete tuned

intake system increased torque by 28% over the simple two leg manifold
at the design speed.

The combination of port fuel injection and the

separation of intake runners reduced cylinder to cylinder fuel

maldistribution from approximately 20% to a range of 1% near
stoichiometric to a maximum of 4% at lean equivalence ratios.
In the test engine, a partially closed throttle plate in the individual
tuned intake runner of a cylinder significantly reduced torque and
intake port pressure at IVC with only a small change in intake manifold
pressure.

There was a strong relationship between intake manifold

pressure at IVC and torque.

This indicates that throttle plate

disturbances significantly interfere with the torque increase due to

resonance even if mean gas velocity is low and that care has been taken
to minimize disturbances due to fabrication welding and intake runner
bends.
Ethanol increased torque in the test engine by 5%

r~lative

to

gasoline from 49.7 to 52 N*m and increased thermal efficiency from

52

31.1% to 33.5 % at WOT, MBT spark timing, and stoichiometric

equivalence ratio.

The wider flammability limit of ethanol allowed

operation at the leaner equivalence ratio of .8 with a resulting increase

in the maximum calculated thermal efficiency to 34.9%.
In the test engine, ethanol reduced maximum NOx emissions by
29% from 19 g/Kw*hr to 13.5 g/kw*hr when operating near
stoichiometric equivalence ratios.

This can be attributed to the greater

heat of vaporization of ethanol resulting in lower. combustion

temperatures.

Specific NOx emissions have a much higher peak for

ethanol when operating at increased spark advance and equivalence

ratios of .9 to .7, but specific .NOx emissions below that of gasoline can be
achieved by operating at equivalence ratios below .7 with increased
spark advance.

Over the equivalence ratios tested for both fuels, specific

HC emissions in the test engine while burning ethanol are reduced to

approximately 50% of those while burning gasoline with a minimum of
2.8 g/Kw*hr achievable with ethanol at lean equivalence ratios vs. a
minimum of 7.5 g/kw*hr achievable for gasoline.
CO emissions were minimized by operating at lean equivalence
ratios. C02 emissions are approximately 63000 g/kW*hr for gasoline
during stoichiometric operation at MBT spark timing.

C02 emissions

were 58000 g/kW*hr for E95 during both stoichiometric, MBT operation
and at the maximum observed thermal efficiency point of 21 degrees
advance and .8 equivalence ratio.

This trend is expected since specific

C02 emissions are dependent only on specific fuel consumption and fuel
carbon content.
Ethanol fuel offers an alternative to the traditional operating
strategy used to reduce emissions while operating on gasoline.
53

When

using ethanol in this application, operation at an equivalence ratio of

approximately .68 to .75 and a spark advance of between 10 and 21
degrees results in low
emissions.

BS~C

and low uncatalyzed specific NOx and HC

Increasing spark advance from 10 degrees to 21 degrees

reduces BSFC and increases torque at the penalty of increased NOx

emissions.

Reducing equivalence ratio results in lower .NOx emissions

but also increased HC emissions and reduced torque.

There is an excess

of oxygen at this point, allowing easy oxidation of HC emissions if NOx

emissions are already significantly lowered.

Further testing in this

operating region with varying spark advance may provide a better

compromise than the spark advance values tested.
The higher octane rating of ethanol relative to gasoline allows E95
operation at MBT while gasoline encounters a knock limit in the test
engine.

Ethanol provides a larger margin of safety from knock during

the severe operating conditions of WOT and stoichiometric air fuel ratio
which are encountered when optimizing operation for a three way
catalyst.

54

References

1. Whittenberger, William A., Kubsh, Joseph E.

"Electrically heated metal

substrate durability," S.A.E. Paper 910613. 1991.

2. Hochmuth, John K.; Bruk, Patrick L.; Tolentino, Cesar; Mignano,

Michael J.

"Hydrocarbon Traps for Controlling Cold Start Emissions,"

S.A.E. Paper 930739, 1993.

3.

Michael P. Thompson, Helmuth W. Engleman. "The Two Types of

Resonance in Intake Tuning," A.S.M.E. Paper 69-DGP-11. 1969.

4. Helmuth W. Engleman. "Design of a Tuned Intake Manifold," AS ME

Paper 73-WAA/PGP 2. 1973.

5. Heywood, John B., Internal Combustion Eniine Fundamentals, McGraw

Hill, New York, 1988.
6. Marks, Lionel S., Marks' Standard Handbook For Mechanical
Enaineers, McGraw Hill, New York, 1978.

55

Ethanol
percent percent

co

0.14
2.38
0.66
0.32
0.29
0.13
2.50
2.15
1-.76
1.60
3.54
3.22
6.37
6.08
0.13
0.09
0.11 .
0.09
0.13
0.10
0.12
0 .09
0.11
0.09
0.10
0.09
0.13
0.12

12.39
13.20
13.71
14.13
13.61
13.81
12.87
13.28
13.23
13.63
12.24
12.66
10.68
10.95
12.24
11.96
11.22
10.99
9.94
9.93
12.69
12.38
11.50
11.26
10.35
10.12
8.88
8.62

Appendix 1
Emissions Data

ppm

ppm

percent

t-C
371
681
981
556
847
519
1135
755
1042
717
1182
858
1369
1080
867
610
843
632
1009
733
696
497
779
577
905
644
1384
1259

N)(

02

45
90
2616
3360
2767
3470
1344
1753
2137
2037
1095
1265
329
386
2048
1762
653
699
151
138
3996
4622
2775
2928
801
982
92
123

CYUNDER POWER (kw)

2.87
0.34
0.97
0.81
1.28
1.22
0.60
0.35
0.70
0.41
0.48
0.26
0.36
0.24
2.82
3.15
4.00
4.28 ..
5.46
5.42
2.20
2.57
3.63
3.90
4.95
5.19
6.62
6.86

1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2

0.0
0.0
17.6
17.6
17.2
17.2
17.5
17.5
0.0
0.0
18.3
18.3
18.3
18.3
14.8
14.8
..
13.2
13.2
10.4
10.4
16.1
1 6.1
15.1
15.1
13.4
13.4
10.2
10.2

ADVNa.
10
10
10
10
10
10
10
10
21
21
10
10
10
10
10
10
10
10
10
10
21
21
21
21
21
21
21
21

Gasoline

CD

en

0.50
0.84
0.95
0.39
2.56
2.76

13.69
13.88
13.45
13.81
12.67
12.80

t-C
2069
1258
2000
1197
2412
1294

N)(

oz

3496
3731
3828
4361
2035
1860

1.50
0.92
1.93
1.41
0.99
0.58

56

CYUNDER POWER (kW) ADVNCE

1
2
1
2
1
2

17.3
17.3
17.0
17.0
17.6
17.6

17
17
17
17
17
17

Appendix 2
Fuel Consumption
Ethanol
Average injection Pulse (microseconds)
injave
spark
02 sensor
power
advance
voltage
(kW)
8919
10
0.10
18.5
9850
10
0.88
19.6
9307
10
0.50
19.3
6974
10
0.03
10.8
8018
10
0.04
16.0
7496
10
0.04
14.1
10106
10
0.92
19.4
21
8592
0.10
18.4
9766
21
0.88
19.0
21
9267
0.50
18.9
21
6908
0.03
13.5
21
8018
0.05
17.0
7496
21
0.04
15.4
10106
21
0.92
19.0
gasoline
5805
5690
5301
5994

10
10
10
17

0.50
0.10
0.88
0.50

18.2
17.9
18.8
18.7

57

Data
Fuel Volume 224 ml

torque
(N*m)
48.9
52.0
51.2
28.6
42.4
37.4 .
51.5
48.7
50.4
50.1
35.8
45.1
. 40.8
50.4

48.3
47.5
49.9
49.6

time
(seconds)
91.5
80.0
85.0
118.0
101.0
99.5
77.0
91.5
82.0
87.0
116.5
100.5
109.5
77.0

coolant
temp
68
64
66
73
70
72
66
65
62
64
73
74
72
62

130.5
136.5
116.0
128.0

72
73
73
68

Variable

RPM

4000
3800
3600
3400
3200
3000
2800
2600
2200
2000
1800
1600

Appendix 3
Length Torque

Data

Manufacturer's
18.5 inch
21.5 inch
24 inch
27 inch
Manifold
torque (N*m) torque (N*m) torque (N*m) torque (N*m) torque (N*m)
42.7
42.7
42.2
42.0
43.8
43.8
43.5
43.6
45.6
45.4
45.8
44.8
40.5
45.8
46.8
. 42.1
46.3
46.4
47.0
47.8
47.8
47.9
43.4
48.0
48.4
49.2
49.1
44.5
49.0
49.1
49.0
50.0
45.3
47.9
47.9
48.1
48.8
45.9
45.8
46.4
46.1
46.3
46.1
45.4
44.5
42.3

58

Appendix 4
Variable Spark Advance Data
Ethanol

T atm ( C)"

25
Advance (BTDC)
0

Patm(mbar)

Torque

10
11
12
15
20

1000
998
992
992
992
998

4 7.63286454
50.88741179
50.90944979
50.76846978
50.62846879
50.18415791

Gasoline

T atm ( C)

Advance (BTDC)
0

Patm(mbar)

Torque

996
996
996
996
996

44.92198372
48.95218357
49.44367135
49.54196891
48.36239822

25

10
15
17

20

59

Appendix 5
Throttle Angle and Maniold Pressure Data for Pressure Traces
T atm ( C)

Patm (mbar)

25

1000.00

Torque

Throttle Angle

44.3
41 . 1
37.7

0
15
35

0.60
0.80
0.90

T atm ( C)

Patm (mbar)

25

989.00

Torque

Throttle Angle

30.5
22.3
23 . 9
16.5
26.3
34.0
42.7
0.0

40
55
50
75
45
25
0
90

Manifold Vacuum (in Hg)

Manifold Vacuum (in .Hg)

1.30
3.05
2.50
6.65
2.05
1.00
0.50
18.00

60