. v N: 79121.1»
graham...
uh... . J

{.hfiflx}. 3.5..." if... _.
. WV.» . ...._._......m.m.m..w kazflAn1..
.mw..§-.¢n. .........,§;.

. r. .3...me
9...... ..w . {arena a
4.5....

{mafia .J..§........:,3m.

can! s .fil, H
. {I v... c
:Wanf.» ......,....~.....1 .
. s

j .
333$}
3 V

.‘ . .. . . ..uvfi...
_ . . . 4.. w it LI}... :5. ,
. . . . a. .sfismcfifi.
2.; :55." ..:. . v . . £953-...
. .05.». u»... Lahnawfi. m.
*4 » Jammwnkr
5......urmu..w .

DIS. .0
Iv Ht.
:1: . v.3 .

”53»
:. 7?‘
§

1!

, "g

E‘
Q
:
t

9

{Vitri-
. lgfink,
-. I .“ . .
:. tn... . . i
a. A... . .I. .-:...
. awawzuui i in 2.... 1.3:... e
.551... u 30:! a." ..:..«r... krona hm! .
u... . «3.9.31.3. . . . -

K.
. .. ‘2
2. 4.3.6

E}!

.. if.
. r
5

., .
U? ..«vv...
1“:
. .
t

l

or-.. turn».
.... 4.. ..S.......:.uh..
so. ...

3%..

5

2|
0

'33.

. . VI»: 5 r! I I I
..:...v... 3.3.14.5. .-..-i..wlnu.. h
. hi] I. .thl. o» anoinflo-c‘. .. II
.> .

bu .‘ . .
..: .. . .

‘I «L‘Ibvurr4~..nx
.. .. . . Ilrbt..- u
.

l. #8
O I.
.11.... an”... 5%
...ll’tfl. “-9.996.
n .. . . . A A
1 . . .. n . . .
..ua ..m...H.H.u-.u. .. .. .,
.3. ..T n- ... . .

u. :11! vlh.» . u?

10;

c
.
.

3

{‘5‘
l 43!. 1311.. - .7. 3..

.
Iii? v"
I p
(I. 1.0.9.3.: : H483.

‘QILL
.

0 'cl‘31'143Io'

. . 7:43.191..- ...... n-
.1 A ..... It! I... .x .HL..H\JEF1.V¥F.M.PJI5 {Iii .
. a A It In

.. . It. r .1“
VIII-(Iv. .2. .I.I.rr.. uuk ..: . .Oo. . . rfilbflgv..)..v\v
Iv Cllllit‘tlv‘I’H-‘n .l :L’flkultl'f l I! II .I i
.rf..v17 u... ..l.l

I 3.71 a...»

. . . v .1! .o..l 3...... . .
. 7 I‘ll... '03- ii. I:
. . 0.. .1l... .l .L 9q'L‘I-‘l
r, A'tbll. . 7-7:..-“ "14.1,.

--.......

I.
23

z“

....r........ 41....-.nl....1-;- .! 2 .. ;
. I 2.... i. 3-.-- 1I4l...f.....
-.:.-. .21--.. (.-:..ilxziv
..vt..1....-.-p..vh«e!!..) .13:

.1

Vt

W.

lio‘.

in?
t

.1. A .9
. . . 3.th-
.L..~ Munrlfllum‘ohvli
J. .1. ..L. an... «L... -
v.1! ..I . . .

'f‘ .
r1" y .
[51‘

‘

\-
if

. I
.1: ’

.1 1
'F? ft-f

“I

:'
§
l

$2.310. .4. .I:
r ...:- Till-til 1
.l. .03.): .-. 3-3.3.13-..
a .1. i- ..l. .. 3:14.- -.Dirllx..-
A...” 3.1!”? v.-- 3.2.2.1.... x -
. sl, 1.-.: .

 

 

vnnnlal‘- ..ti.

;.. , L \ . .,
-.thivtuflde .n Lwl...L.|oHo.MHl§l.!t‘o

. W- .o

 

{1.7. --r u... v.-ovq.vl . 11.. V. nu VIOI.‘.UIHJ>I‘00
.Ocviltl 3.x I'VIKIXV..- I. WNW“..- 1A.?“

u Il’lAVAlbvl.o .. it. .[ tr.‘ 1. . I. \nl I:

’ ..:... 0‘:

 

- ..:.Uuhsmouhdlviullx ugfiun
..niwwflmmmgr u, , . .

vi.| 1GC v.

 

SIIRARTYLB

eeese IlllllllllllllllllllllIlllllllllllllllll|

I 31293 01716 738

llalllllzllll

 

 

 

This is to certify that the

thesis entitled

EFFECTS OF AMBIENT CONDITIONS ON THE
EMISSIONS OF A SMALL CARBURETED OVERHEAD
VALVE FOUR STROKE UTILITY ENGINE

presented by
Erik C. Bertrand

has been accepted towards fulfillment
of the requirements for

M. S . degree in Mechanical Engineering

   

Major professor

Date August 14, 1997

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

‘—

LIBRARY I
‘ Michigan State
University

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE ‘ DATE DUE

 

091514

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1/98 chlRC-Datoouam.“

 

EFFECTS OF AMBIENT CONDITIONS ON THE EMISSIONS OF A SMALL
CARBURETED OVERHEAD VALVE FOUR STROKE UTILITY ENGINE

By

Erik C. Bertrand

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Mechanical Engineering

1997

\ I'll ‘ ALI.

[I'D {I‘D} ,

I I.

u?!’

ABSTRACT

EFFECTS OF AMBIENT CONDITIONS ON THE EMISSIONS OF A SMALL
CARBURETED OVERHEAD VALVE FOUR STROKE UTILITY ENGINE

By

Erik C. Bertrand

Small hand-held utility engines are increasingly becoming a factor in emissions
production and regulation. This paper outlines a laboratory study of the effects of
ambient conditions on the steady state emissions and performance of a typical air-cooled,
overhead valve, carbureted, single cylinder four-stroke engine used for lawn and garden
applications. In this study, the ambient temperature and relative humidity (and thus
specific humidity) were thermostatically controlled, with one varied while the other was
held nearly constant. The engine testing was performed at WOT under steady-state

conditions at a fixed engine speed using a hydraulic dynamometer.

Experimental results show that in carbureted engines, the main effect of changes
in ambient conditions is to change the intake air properties and therefore the metered
air/fuel ratio (AFR). Moderate changes in ambient conditions lead to small changes in the
AFR but large changes in emissions output. Definite trends existed in the air-fuel ratio,
engine torque, C0, C02 , NOX emissions, as the ambient conditions were varied. For
example, a 0.009 decrease (0.015 to 0.006) in the specific humidity resulted in a 50%

reduction in NOX emissions.

ACKNOWLEDGMENTS

I have a few important people I would like to thank:

Dr. Giles Brereton, my advisor, for his guidance, intuition and ambition. Without
his help graduate school would have been much more difficult, if not unlikely. Dr.
Harold Schock, for giving me the chance to become associated with the MSU Engine
Research lab. My wife, Nancy, for sticking by me throughout the latter part of my
undergraduate education and graduate school, despite the lack of attention. My father,
Marc, for his financial and educational input. Without his advice, it would not have been
possible to attend undergraduate or graduate school. My mother, Kathy, whose good
spirit and positive thinking made an overwhelming task seem almost effortless. Tom
Stuecken for the countless brackets, fixtures and components that were fabricated along
with the relentless pursuit of perfection. My Grandfather, although no longer on this
earth, would have been proud. Lastly, I would like to thank the Environmental Protection

Agency, Ann Arbor, Michigan for supporting this work.

iii

TABLE OF CONTENTS

ACKNOWLEDGMENTS .................................................................................................. iii
TABLE OF CONTENTS ................................................................................................... iv
LIST OF TABLES ............................................................................................................. vi
LIST OF FIGURES ........................................................................................................... vii
NOMENCLATURE ........................................................................................................... ix
1 . INTRODUCTION ........................................................................................................... 1
1.1 Overview ................................................................................................................... 1
1.2 Four Stroke Operation ............................................................................................... 3
1.3 Emission Characteristics of Small O.H.V. Four Stroke Engines .............................. 5
1.3.1 Emissions from rich air / fuel ratios ................................................................... 5
1.3.2 Emissions from high temperature combustion ................................................... 6

1.4 Literature Review ...................................................................................................... 6
1.5 Objectives ................................................................................................................ 10

2. EXPERIMENTAL METHODS .................................................................................... 12
2.1 The Test Engine ...................................................................................................... 12
2.2 Climate Control ....................................................................................................... 12
2.3 Experimental Apparatus .......................................................................................... 13
2.4 Experimental Procedure .......................................................................................... 14
2.4.1 Measured Quantities ......................................................................................... 15
2.4.2 Calibration ........................................................................................................ 15
2.4.3 Temperature Variation ..................................................................................... 17
2.4.4 Humidity Variation ........................................................................................... l7

3. RESULTS ...................................................................................................................... 18

iv

3.1 Theoretical Equations .............................................................................................. 18

3.1.1 Carburation ....................................................................................................... 18
3.1.2 Compression ..................................................................................................... 25
3.1.3 Combustion ...................................................................................................... 28

3.2 Theoretical Results .................................................................................................. 29
3.3 Experimental Results .............................................................................................. 35
3.3.1 Humidity Variation ........................................................................................... 35
3.3.2 Temperature Variation ..................................................................................... 43

4. CONCLUSIONS ........................................................................................................... 51
Appendix A ....................................................................................................................... 53
A.1 Fuel Flow Measurement ........................................................................................ 53
A.2 Dynamometer and Controls ................................................................................... 55
A3 Environmental Controls ........................................................................................ 57
AA Dilution Tunnel ..................................................................................................... 58
A5 Emission Measurement ......................................................................................... 59
A6 Data Collection ...................................................................................................... 62
Appendix B ....................................................................................................................... 64
3.] Heat Transfer ......................................................................................................... 65
B.2 Work ...................................................................................................................... 66
B.3 Internal Energy ....................................................................................................... 68
B.4 Change in Gibbs Free Energy ................................................................................ 68
Appendix C ....................................................................................................................... 70
Appendix D ....................................................................................................................... 72
Appendix E ........................................................................................................................ 75
LIST OF REFERENCES .................................................................................................. 76

LIST OF TABLES

TABLE I: CARE EXHAUST EMISSION STANDARDS ................................................ 3
TABLE 2: BASIC FOUR STROKE OPERATION ............................................................ 3
TABLE 3: CALIFORNIA CURRENT EMISSION STANDARDS ................................. 10
TABLE 4: THE TEST ENGINE ....................................................................................... 12
TABLE 5: EQUIPMENT USED ....................................................................................... 13
TABLE 6: EMISSION MEASUREMENT TECHNIQUES ............................................. 62
TABLE 7: TABLE OF WOSHNI COEFFICIENTS ......................................................... 65

vi

LIST OF FIGURES

FIGURE 1: COMPOSITION OF EXHAUST AND FUEL AS A FUNCTION OF

CARBON NUMBER [6] ............................................................................................ 8
FIGURE 2: EXHAUST EMISSION SOURCES [4] .......................................................... 8
FIGURE 3: OVERVIEW OF TEST CELL ....................................................................... 14
FIGURE 4: CROSS SECTION OF CARBURETOR THROAT ...................................... 19
FIGURE 5: PRESSURE - VOLUME DIAGRAM ........................................................... 27
FIGURE 6: ACTUAL AFR VS. DESIGN AFR ................................................................ 30
FIGURE 7: COMBUSTION TEMPERATURE VS. ACTUAL AFR .............................. 30
FIGURE 8: COMBUSTION TEMPERATURE VS. DESIGN AFR ................................ 31
FIGURE 9: ACTUAL AFR VS. DESIGN AFR ................................................................ 32
FIGURE 10: PEAK COMBUSTION TEMPERATURE VS. ACTUAL AFR ................. 33
FIGURE 11: PEAK COMBUSTION TEMPERATURE VS. DESIGN AFR ................... 33
FIGURE 12: CO/C02 RATIO VS. DESIGN AFR (l) ..................................................... 34
FIGURE 13: CO/C02 RATIO VS. DESIGN AFR (2) ..................................................... 34
FIGURE 14: ROOM TEMPERATURE VS. SPECIFIC HUMIDITY ............................. 36
FIGURE 15: ENGINE TORQUE VS. SPECIFIC HUMIDITY ....................................... 37
FIGURE 16: BSFC VS. SPECIFIC HUMIDITY .............................................................. 38
FIGURE 17: CO VS. SPECIFIC HUMIDITY .................................................................. 39
FIGURE 18: C02 VS. SPECIFIC HUMIDITY ................................................................ 39
FIGURE 19: NOX VS. SPECIFIC HUMIDITY ............................................................... 40
FIGURE 20: HC VS. SPECIFIC HUMIDITY .................................................................. 41

vii

FIGURE 21

FIGURE 22:
FIGURE 23:
FIGURE 24:
FIGURE 25:
FIGURE 26:
FIGURE 27:
FIGURE 28:
FIGURE 29:
FIGURE 30:
FIGURE 31:
FIGURE 32:
FIGURE 33:
FIGURE 34:
FIGURE 35:
FIGURE 36:
FIGURE 37:
FIGURE 38:
FIGURE 39:
FIGURE 40:
FIGURE 41:
FIGURE 42:

AFR VS. SPECIFIC HUMIDITY ................................................................ 42

CO/C02 RATIO VS. SPECIFIC HUMIDITY ............................................ 42
SPECIFIC HUMIDITY VS. ROOM TEMPERATURE ............................. 44
ENGINE TORQUE VS. ROOM TEMPERATURE ................................... 45
BSFC VS. ROOM TEMPERATURE ......................................................... 45
CO VS. ROOM TEMPERATURE .............................................................. 46
C02 VS. ROOM TEMPERATURE ............................................................ 47
NOX VS. ROOM TEMPERATURE ........................................................... 47
HC VS. ROOM TEMPERATURE .............................................................. 48
AFR VS. ROOM TEMPERATURE ............................................................ 49
CO/C02 RATIO VS. ROOM TEMPERATURE ........................................ 50
VIBRATING FLOW TUBE (SINGLE FLOW TUBE SHOWN) ............... 54
FLUID FORCES REACT ING IN VIBRATION OF FLOW TUBE ........... 54
END VIEW OF FLOW TUBE SHOWING TWIST ................................... 54
FUEL FLOW CART .................................................................................... 55
DYNAMOMETER HYDRAULIC CIRCUIT ............................................. 56
TEMPERATURE AND HUMIDITY CONTROLS .................................... 57
DILLUTION TUNNEL ............................................................................... 59
ANALYZER CROSS SECTIONS .............................................................. 61
DATA COLLECTING COMPUTER (486 PC) .......................................... 63
DATA COLLECTING ALGORITHM ........................................................ 63
WORK INPUTS AND OUTPUTS .............................................................. 66

viii

NOMENCLATURE

English Symbols

 

N_amg Meaning Units
a Moles of Air kmoles
aa Moles of Air (actual) kmoles
a d Moles of Air (design) kmoles
b Moles of Water Vapor kmoles
c Speed of Sound m/ sec
7;! Mass Flow Rate kg / sec
mm, Mass of Air kg

m M, Mass of Fuel kg

n . Moles of Air kmoles

air

ix

a,

P REF

Moles of Fuel

Specific Volume

Moles of C02

Moles of C0

Moles of HC

Moles of NOX

Specific heat at constant volume

Specific heat at constant pressure

Equilibrium Constant

Molal Mass

Mach Number

Reference Pressure

Pressure

Pressure at start of compression

kmoles

meters3 / kg

kmoles

kmoles

kmoles

kmoles

kJ/kg—K

kJ/kg-K

kg / kmole

m/ sec

atm

atm

atm

Greek Symbols

Symbol

Reference Pressure

Universal gas constant

Specific gas constant

Reference Temperature

Temperature

Temperature at start of compression

Volume

Volume Flow Rate

Meaning

Specific Humidity

Density

xi

atm

kJ/kmol—K

kJ/kg—K

meters3

meters3 / sec

 

kgH2 O/kgAIR

kg / meter 3

AFR

AFR

AFRd

BSF C

C0

C0,

HC

N0,

0H V

WOT

Meaning

Air Fuel Ratio

Actual Air Fuel Ratio

Design Air Fuel Ratio

Brake Specific Fuel Consumption

Carbon Monoxide

Carbon Dioxide

Hydrocarbons

Oxides of Nitrogen ( N0 and N02)

Over Head Valve

Wide Open Throttle

xii

 

kgAIR/ngUEL

kgAIR/ngUEL

kgAIR/ngUEL

gr/kW -hr

1. INTRODUCTION

1.1 Overview

Internal combustion engines come in a variety of shapes, sizes and configurations.
From tiny 1.6cc (.001638 liter) model aircraft engines to 434,257cc (434 liters) large ship
powering diesels, they all have one thing in common: they use some type of fuel to do
“useful” work for mankind. In addition there are many designs of internal combustion
engines using mainly two types of fuel: gasoline and Diesel. Four stroke engines and
their cousins the two stroke models are similar in many ways. For example, they both use
a piston cylinder arrangement to compress the air / fuel mixture to some fraction of its
original volume. Likewise, they can run on either gasoline or Diesel fuel, and can be
liquid or air cooled. In a four stroke engine, the cylinder filling and emptying is

controlled largely by the camshaft timing and the moving piston.

In a two stroke engine, the cylinder filling and emptying are controlled by the inlet
and exhaust ports cast into the cylinder walls. The two stroke engine has no clear
defining line between the four combustion events that happen in the four stroke engine. It
is this lack of cylinder event definition that may lead to the decline of the two stroke
engine, in the small engine / utility arena. Two stroke hydrocarbon emissions in exhaust
gases are typically found to be at least ten times greater for similar engine displacements
and power outputs. For many years, the automobile has been considered as one of the
most serious polluters, powered by internal combustion engines. However this may no

longer be true.

California Air Resources Board (CARB) studies found that two—cycle engines in
California produce 48,000 kg per day of total exhaust hydrocarbons, and 150,000 kg per
day of carbon monoxide [21]. Using these estimates, hydrocarbon and carbon monoxide
emissions from utility engines in the state are equal to the emissions produced by 3.5

million new cars driven 26,000 km ( 16,000 miles).

CARB has since proposed standards to regulate emissions from several types of
utility engines that are smaller than 225 cc and less than 19 kW (25 hp). These categories
cover lawn and garden equipment and general utility equipment such as pumps,
generators and compressors. The 1995 standards for lawn and garden equipment includes
a category for hand-held engines with a displacement of 20 cc to 50 cc. These standards
are 241 g/kW-hr Hydrocarbons (HC), 804 g/kW-hr Carbon Monoxide (C0), and 5.4
g/kW-hr Oxides of Nitrogen (NOX ). The 1999 standards are 67 g/kW-hr HC, 174 g/kW-
hr C0, and 5.4 g/kW-hr NOX. As a result of these new regulations, small engine

manufacturers are required to reduce emission output from their engines, see Table 1.

It is important to understand the role ambient conditions play in evaluating
emission output from these sources. As emission regulations for small engines are
phased in, it will be important to specify the testing conditions under which regulations
must be met. The test parameters include the ambient conditions (atmospheric pressure,
air temperature and humidity) under which the test was carried out. It has been known for
a long time that temperature and humidity affect engine performance and emission output
[2]. In addition, the valid testing ranges of ambient conditions acceptable for measuring
emission output, have not yet been set by CARB or the Environmental Protection Agency
(EPA). It is for these reasons that this study has been undertaken; to better understand
how ambient conditions affect emission output levels and to provide some insight as to

appropriate ambient testing parameters.

Table 1: CARB Exhaust Emission Standards
(gr/brake kW-hr)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

YEAR gig“: HC + NOx HC NOX co
1995 I 16.1 - - 402
II 13.4 - - 402
Ill - 295 5.4 804
IV - 241 5.4 804
V - 161 5.4 402
1999 I, II 4.3 - - 134
III, IV, V - 67 5.4 174
CLASS DESCRIPTION
I Non hand held, less than 225cc displacement
ll Non hand held, greater or equal to 225cc displacement
III Hand held, less than 200C displacement
IV Hand held, 200C to less than 50cc displacement
V Hand held, greater or egal to 50cc displacement

 

1.2 Four Stroke Operation

Four stroke engines utilize four distinct processes for each combustion event and

are summariZed in table 2.

Table 2: Basic Four Stroke Operation

 

 

EVENT PISTON ACTION VALVE OPEN
DIRECTION INTAKE EXHAUST
INTAKE DOWN FRESH CHARGE IS X

DRAWN INTO THE CYLINDER
THE MIXTURE IS COMPRESSED
TO A FRACTION OF ITS ORIGINAL VOLUME
THE MIXTURE IS IGNITED
PRODUCTS ARE PRODUCED
THE PRODUCTS ARE
EXPELLED FROM THE CYLINDER

 

COMPRESSION UP

 

EXPANSION DOWN

 

 

 

 

EXHAUST UP

 

 

 

 

The first step in the process is the intake stroke. During this process, only the

intake valve is open and the piston is moving down within the cylinder. This creates a

vacuum and draws air in past the carburetor, through the intake ports, around the intake
valve and into the cylinder. Next, the intake valve closes and the piston starts to move
upwards. This compresses the fuel / air mixture within the cylinder prior to ignition. The
final pressure and temperature within the cylinder is a function of many things but most
important are compression ratio, port design and mixture properties. When the piston is
close to Top Dead Center (TDC), a spark is delivered to the spark plug initiating
combustion within the cylinder. This combustion drastically increases the mixture
temperature and therefore the pressure within the cylinder and thus forces the piston
downward. When the piston is near Bottom Dead Center (BDC), the exhaust valve is
opened allowing blowdown of the cylinder. Blowdown happens when the pressure inside
the combustion cylinder is higher than the pressure within the exhaust manifold, thus
allowing the newly created exhaust products to escape. The process is further helped
along when the piston starts moving upward decreasing the volume within the cylinder
and further forcing exhaust products out of the cylinder. Unlike a two stroke engine, the
short circuiting process does not happen. Short circuiting happens when fresh intake
charge is expelled through the exhaust port, which results in a significant source of
unburned hydrocarbons. In a four stroke engine, the pressure inside the intake manifold
is less than atmospheric (under WOT conditions, the pressure is approximately
atmospheric). Since the pressure inside the engine cylinder is greater than atmospheric,
when the intake valve is opened, a small amount of exhaust gases flow back into the
intake port. This flow will reverse when the piston starts moving downward again,

creating lower pressures within the cylinder.

This clear definition of cylinder events leads to higher combustion temperatures
than two stroke engines, (due to the lack of intake dilution) and correspondingly higher
NOX emissions. In addition, the fuel consumption is approximately one third that of a

two stroke engine and HC output is a factor of ten less. These significant differences lead

to an engine that is more fuel efficient and one that produces less HC emissions.
Conversely, the four stroke engine is one that is more expensive to produce due to the

increased parts list involved (camshaft and related valve gear).

1.3 Emission Characteristics of Small O.H.V. Four Stroke Engines

There are two main factors that influence emission output from small four stroke
engines, relative to a stoichiometric mixture: 1) the operating air / fuel ratio and 2) the

high relative combustion temperatures.

1.3.1 Emissions from rich air/ fuel ratios

Small engines require air / fuel ratios that are slightly richer than stoichiometric,
typically around 12:1 to 13:1. The abundance of fuel present within the mixture has three
effects: 1) it increases HC output; 2) it increases CO output; and 3) it lowers the relative
cylinder combustion temperature. In ideal combustion, all the fuel and oxygen is

converted into CO2 and H20 , according to equation (i):

C811,8 +125.(02 +3.76N2) —> 8-CO, +9-H20+47-N2 (i)

In fuel rich combustion, there is relatively less oxygen, so fuel ends up being

burned as CO2 , CO and unburned HC , according to equation (ii):

CgHl8 +a(02 + 3.76* N,) —+ wCO2 +xC0+ yHC+zH20+a3.76- N2 (ii)

This equation by no means is a complete model of incomplete combustion. It is intended

to show some of the possible combustion products when fuel rich combustion occurs.

The HC ’5 present in the equation are illustrative of all HC ’3 that result from excess fuel.
Any excess products, other than CO2 , H20 and N 2 will result in a decrease in
combustion temperature. The rich air / fuel mixtures inducted into these engines causes

an increase in CO , HC and a decrease in combustion temperature.
1.3.2 Emissions from high temperature combustion

The inherent nature of four stroke engines is to have higher combustion
temperatures than their two stroke counterparts. This is primarily due to the lack of short
circuiting during the intake process. Since there is less exhaust gas present within the
cylinder, the result is higher combustion temperatures and a corresponding increase in the
formation of NOx . Primarily this (high NOx ) is the main drawback of the four stroke
engine, besides the increased initial cost. NOX formation can be controlled with
conventional emission control methods, such as Exhaust Gas Recirculation (EGR), and is

found on automotive production four stroke engines today.
1.4 Literature Review

Many papers have been written on the subject of small engine emissions.
However most four stroke engines are larger in displacement and of older design than the
one tested in this study. In the early seventies, several papers were written on the subject
of small engine emissions. During the early years of emission regulation (1972), small
engines were considered only to contribute a small part of the total C0, HC and

NOX production.

Eccleston, and Hum, [6] first analyzed exhaust emissions from small utility

internal combustion engines, noting how the exhaust hydrocarbon composition differed

between two stroke and four stroke engines. More specifically the hydrocarbon
composition within the exhaust of the two stroke engine closely mirrored that of the fuel.
This was not the case for the four stroke, where the fuel was broken down into more basic
carbon elements (see figure 1). This is a direct result of the short-circuiting event that
happens during the scavenging process, where part of the intake charge escapes through

the exhaust port.

Donohue, Hardwick, Newhall, Sanvordenker and Woelffer, [4] discuss the
relevance of the emissions from these engines, estimating hydrocarbon emission output
from these sources to be less than 1% of the total annual hydrocarbon output from all
sources (see figure 2). Note: The scaling of the x axis, the output from the small engines,
is an order of magnitude less. In addition they estimate of the total US. carbon monoxide
output, totaling 100,000,000 tons/year, small engines contribute approximately 1,160,000
tons/year or approximately 1%. It is important to keep in mind that this data was
formulated in 1968; current automotive production of emission products has been greatly

reduced.

Hare, Springer, Oliver and Houtman, [9] estimated that emission production from
small utility engines constituted only a small portion of the total pollution from all
sources. They further estimated (as percentages of 1970 national totals from all sources)
small engines account for 0.4% of HC’s, 0.8% of C0 and 0.006% of NOX. It was
noted they rank second to highway vehicles in terms of the current number of engines in
use. With very vigorous sales numbers each year, future weighting of emission output

from these engines was likely to become more pronounced.

 

Fraction of Unbumed HC's Vs. Carbon Number of Fuel

 

 
 
 

 

 

  

 

 

.w 35 l
‘1’ so - IFUEL
E 25 (L uz-cvcua
g 20, let-CYCLE
2 15-
O
5 10 -
|
IL 0
1 2 3 4 5 6 7 e 9 1o 11
than

 

 

 

Figure 1: Composition of Exhaust and Fuel as a function of carbon number [6]

 

 

 

 

 

 

 

 

Exhaust Output Vs. Source .00
(:1 HC
I NOX
ta
0
>.
E
C
O
*—
TOTAL AUTOMOTIVE SMALL ENGINE
(x 1.0e6) (x 1.0e6) (x 1.0e5)
Source

 

 

 

Figure 2: Exhaust Emission Sources [4]

In addition to previously being considered only a minor contributor in the global
emission production, these engines produce very high amounts of HC, NOx and CO
for the power generated. Several papers have been written on the subject of reducing the
output from these useful sources. White, Carroll, Hare and Lourenco [10] found that
several traditional methods are effective in reducing emissions output from these engines.

For example, leaner carburetor settings will reduce the CO output from these engines.

Conversely, leaner intake charges typically increase combustion temperatures and
therefore the NOX output, which is also a regulated emission species. In addition the
leaner mixture increased cylinder head temperature and therefore may decrease engine

life.

These results were confirmed by Sun, Assanis, and Brereton [12], who stated that
traditional emission control approaches already in use in automobiles, work quite
effectively. Some examples are catalytic converters, leaner mixture ratios and
oxygenated fuels. Perhaps the biggest factor is the rich mixture ratios required by these
small engines. The rich mixture ratios enhance engine life, promote engine cooling and
extend engine power of small air-cooled engines. The disadvantage of this is high HC
and C0 emissions, increased cylinder temperatures, a reduction in engine life and the

thermal destruction of any catalytic converter so attached.

Mooney, Shinn Hwang, Daby and Winberg [5] also reconfirmed the above
conclusions. Specifically the reduction in mixture requirements were given the utmost
attention. They speculate that meeting CARB Tier 1 and EPA Phase 1 emission
regulatory levels might be accomplished with engine modifications only. CARB Tier 2
will probably require the application of catalytic converter technology to these engines.
In order for the successful application of converter technology several areas of “lean
burn” utility engines need to be addressed. For example: making these engines run
leaner; catalyst durability and thermal stability; improving utility engine oil consumption;

cooling air management / heat shielding and high temperature mounting systems.

Current California exhaust emission standards are quite stringent, compared to
previous allowable pollution levels. Table 3 shows some of the allowed emission levels

and compares them to current automotive standards. It should be noted that a typical

10

chain saw (tuned) emits 600 times the allowable CO and 250 times the allowable HC +
NOX of a typical automobile on a brake specific basis. It is apparent that the practice of
operating these engines at richer mixtures must be addressed before the reduction of

exhaust components can be achieved.

Table 3: California Current Emission standards

 

1.5 Objectives

In general, there has been little work done in the area of ambient conditions and
the effects on small engine emissions. Only a small amount of empirical work has been
done in the area of automobile four stroke engines, but it is of little relevance to small
engines. It would be of great benefit to understand the relationship between ambient
conditions and engine emissions, in order to propose a standardized set of emission

testing procedures. The general objectives of this thesis are:

1) To measure emission output as functions of ambient (pressure, temperature

and humidity) conditions.

ll

2) To explore the degree to which theoretical approaches can be used to reliably
predict four stroke small engine emissions’ dependence on ambient

conditions.

3) To explain why changes in ambient conditions produce changes in measured

emissions and engine power output.

2. EXPERIMENTAL METHODS

2.1 The Test Engine

The engine used throughout the testing procedure was the Ryobi Pro4Mor Over
Head Valve (OHV) engine. This is the first portable hand-held gasoline engine capable
of meeting all proposed U.S. exhaust emission standards for lawn and garden equipment,

including 1999 CARB standards. A table summarizing the engine characteristics is given

 

 

 

 

 

 

 

below (see table 4).
Table 4: The Test Engine
BORE 3.2 cm
STROKE 3.26 cm
DISPLACEMENT 26.2 cc
RATED kW .75 @ 7000 RPM
CONFIGURATION OHV - PUSH ROD
ROCKER ARM TYPE STAMPED STEEL, BALL TYPE
MASS 3.6 kg

 

 

 

 

2.2 Climate Control

The Ryobi Pro4Mor OHV engine was run over a range of ambient conditions.
For example the temperature was varied from 9.53 ‘C to 29.03 ‘C (holding the humidity
constant) and the humidity was varied from 0.01487 kg HZO/kg AIR to 0.00578 kg
HZO/kg AIR (holding the temperature constant). Two types of tests were performed:

one at constant ambient temperature while varying humidity; and the other at constant

12

l3

ambient humidity while varying temperature. As a result, it was possible to measure the

emission output as a function of room temperature, room relative humidity or both.

2.3 Experimental Apparatus

In order to measure the emission output from small engines, several pieces of

equipment are required. They are summarized in table 5 below:

Table 5: Equipment Used

 

 

 

 

 

 

 

 

 

 

 

 

 

FUNCTION APPARATUS
Engine Controls Micro-Dyn Dynamometer Model 15
Exhaust-gas collection Dilution tunnel with 100 scfm critical-flow nozzle
Fuel flow measurement Micro Motion Fuel Meter 0800068100
CO and C02 measurement Horiba Infrared Analyzer AlA-23
NOx measurement Horiba Chemiluminescent Analyzer CLA-22A
HC measurement Horiba Flame Ionization Analyzer FIA-23A
Temperature Controls Honeywell Industrial Temperature Controls T-775-A
Humicfly Controls Honeywell Industrial Humidig Controls T-775-E
Data Collection IBM 486 PC With 3 Digital to Analog Boards

 

 

A brief description of each piece of equipment follows. The engine was run on a
hydraulic dynamometer which allowed it to run at a programmed constant speed (RPM),
while measuring the torque output produced by the engine. Fuel was supplied to the
engine by a custom-built fuel cart, which measured the mass flow rate of fuel consumed
by the engine. Fuel was supplied to the engine at constant pressure. The intake of the
dillution tunnel was positioned so that all the exhaust products and large quantities of
room air were drawn into the tunnel, without placing additional suction on the muffler
exit. This diluted mixture of exhaust gas was drawn through a critical flow nozzle by a

ring compressor, which regulates the total flow rate of ambient air and exhaust gases. A

14

sample of this diluted gas was fed to the emission analyzers while the remainder was
expelled from the test room. The analyzers measured the diluted gas for concentrations

of HC, NOX , CO2 and CO. Figure 3 gives a general overview of the test cell. For

more complete details on equipment used, refer to Appendix A.

   
 
 

Emissions Monitoring Equipment

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

HC NOx C0 C02

I r ............ Dynamometer
: : . f
I 1

I 1

I 1

I I

I 1

I 1

l : r ............................... ..

I | I

I | I

i I i

a -------

 

 

 

 

 

 

 

Data Processing PC

-------- Voltages

 

Exhaust Ducting

 

Temperature and Humidity
Controlled Room

Figure 3: Overview of Test Cell

2.4 Experimental Procedure

For each test conducted (either humidity or temperature ramp) the room was set to
the desired value (temperature or humidity), and allowed to stabilize overnight. After the

desired value had been attained, the necessary equipment was switched on and calibrated.

15

No two tests were performed on the same day and the engine was not operated for more

than 1.5 hours continuously.

For each test, the engine was operated at Wide Open Throttle (WOT) or
approximately 7000 rpm. During this time the engine and emission analyzers were
allowed to stabilize before any measurements were taken. Once the equipment had

reached steady state, the temperature or humidity could be altered.

2.4.1 Measured Quantities

Many quantities had to be measured to calculate the emission output from the

engine. A brief rundown of the important items are listed below.

The concentrations of all emission species involved were measured. They
include: HC, NOx , CO2 and CO. Additional quantities that are measured include: the
speed of the engine, torque output, the atmospheric and differential pressure. The
differential pressure is the stagnation pressure ahead of the critical flow nozzle. The
temperatures include: the temperatures of the nozzle (temperature of air just ahead of the
critical flow nozzle), cylinder head, exhaust, fuel, room air (at engine air inlet) and engine
oil. Relative humidity and fuel flow rates were also quantified. See Appendix E for a

sample listing.

2.4.2 Calibration

The dynamometer, emission bench, and fuel cart were all calibrated initially and
checked periodically. The torque calibration on the dynamometer was performed by

hanging a 1 kg mass (2.2 lb.) a distance of 45.72 cm (18 inches) from the centerline. This

16

known torque generated a voltage and was used along with the zero point, to generate a

linear relationship between torque and voltage.

Calibration of the fuel flow meter was performed by Micro Motion. It was
checked by flowing fuel through the meter for a known time. This volume of fuel
collected (in units of cm3 / sec) was converted to gr/hr and compared to the meter
reading. This verified the Micro-Motion calibration. Other outputs from the fuel flow
cart include the ambient temperature and humidity levels. They were calibrated in a
similar manner to the torque meter. For the temperature output, the room was set to a
reference temperature, verified with four thermometers and compared to the voltage
output. Likewise for the humidity, with the exception that only two meters were used to
double check the room humidity and then compared to the voltage output. This allowed
algebraic linear relationships between temperature, humidity and voltage, to be found that

could be used for subsequent data reduction.

Calibration of the emission analyzers was performed in a similar manner. The
meters were calibrated daily or before each run using two gases of known concentrations.
A “zero” gas (pure N 2 ) was supplied to the analyzer to set the zero point and the “span”
gas supplied to the analyzer set the full scale reading. The meters were all tested for
linearity, using a gas divider and the appropriate span gas. A gas divider controls (or
decreases) the proportions of span and zero gas fed to the meter. This allows one to
record analyzer voltage outputs, for a range of known concentrations and plot the results
to check for linearity. Three of the four meters varied linearly with the span gas (C02 ,
HC and NOX ) while the fourth (CO) was found to be well approximated by a second
order polynomial. The algebraic equations inferred from these calibrations were used to

convert analyzer output voltages to concentrations.

17

2.4.3 Temperature Variation

In order to perform the temperature transient, the room was set to a specified cold
temperature (0 ‘C) and allowed to stabilize overnight. With all the equipment running the
coolers were turned off, the heater was switched on, and the room was allowed to warm
up. As the room warmed up, the emission characteristics of the engine were measured.

The ambient humidity level increased slightly as the room temperature increased.

2.4.4 Humidity Variation

Likewise for the humidity variation. In order to perform the humidity ramp, the
room was set to a specified humidity level (0.015 kgHzO/kgAIR) and allowed to
stabilize overnight. With all the equipment running, the humidifier was turned off, which
allowed the ambient moisture levels to drop. As the moisture level dropped, the emission
characteristics of the engine were measured. As with the temperature transient, the

ambient temperature level increased slightly as the room humidity decreased.

The room humidity was controlled by a steam generator and a Honeywell model
T775-E remote humidity sensor. It was easier to de-humidify the room than to humidify,
so for each test the humidity was set to a high initial value and allowed to stabilize
overnight. Then, with all the equipment running, the steam generator was turned off and
the humidity level was allowed to drop. This gave the desired rate of decrease in ambient

humidity and allowed engine emissions to be measured as a function of specific humidity.

3. RESULTS

The results chapter is divided into three parts: a theoretical section concerned with
thermodynamic analyses of gasoline combustion; a theoretical results section and an

experimental section in which measurements are reported and discussed.
3.1 Theoretical Equations

In order to do combustion analysis, one must first specify the properties of the
reactants. For example, during the combustion of moist air + fuel, the Air Fuel Ratio
(AFR) and specific humidity (to) determine the number of moles of air (a) and water
vapor (b) in the reactants. According to equation (1), fuel reacts with a moles of air and b
moles of water vapor to form combustion products. The following sections (3.1.1, 3.1.2
and 3.1.3) describe the thermodynamic modeling used for carburation, compression and

combustion, respectively.
C8H,8 + a(02 + 376- N2 ) +bH20 —> PRODUCTS (1)
3.1.1 Carburation

Assuming, for convenience, that 1 kmol of fuel is present, the moles of air and
water vapor are completely determined by the specific humidity and the AFR according to

relations (2) and (3) below.

muzo _ "Hzoano _ 18b orb _ 290(1)
mAlR "AIR M AIR 29“ 18

 

 

 

0):

(2)

18

19

mm = a(32 + 3.76- 28)
mm, 1.114

 

 

AFR = = 1.2042 - a (3)

Thus, with the specific humidity and AFR set, the LHS of the combustion
equation is fully determined. In the following discussion, it is shown how, when the

carburetor throat is the limiting area, carburetors regulate the quantity (I’ll/J6 ) or (VA/E ).

Figure 4 gives the reference points used in the following equations.

 

 

 

 

Figure 4: Cross section of Carburetor throat

Treating the flow from intake (position 1) to throat (position 2) as steady, friction
less and incompressible (due to the compressibility effects being negligible, see appendix

C for notes on compressibility), the Bernoulli equation states that,

a Lax;

p+2 p 2 (4)

 

20

Since Vlz/Vz2 zO and P1 —P2 is almost constant for a carburetor at a given

throttle setting. Then JEVZ = CONSTANT , or for a fixed throat area (F = VA),

{LJE = CONSTANT. (5)
Starting with, the ideal gas equation,

PV=nRT,

under steady flow conditions differentiating both sides with respect to time leads to,

o O _ 0 ° _ T
PV'MIX =nMIX RMXTOI' VMIX =nMIx Kiwi?)

where 1‘ is the volume flow rate and n is the molal flow rate of air and water vapor.
Substituting the above relation into (5) and setting nmx =(nAm+nH,o) yields the

following relation (with R absorbed in the CONSTANT ),

. . T
(n AIR 'I" n H20 )(F) p MIX = CONSTANT , (6)

with p mx as the density of air and water vapor. Furthermore, since P was essentially

constant throughout each experiment run in this study, equation (6) reduces to,

(5A,R+r°z,,,o )T pm, = CONSTANT (7)

Next, letting n AIR + n H20 equal a+ b (as in equation (1)),

21

(gubJT pm, =CONSTANT. (8)

Substitution for the density of the mixture in terms of convenient parameters,
requires additional derivation. Starting with the density of the mixture, me , which

equals:

mAIR +mH20 _ pAlRVAlR +pH20VH20 PMRnAm + pHZOnHZO

P MIX = -
V V "TOTAL "TOTAL

 

we can write,

 

__ Pmnxm + pnzonnzo

 

p _
MIX "TOTAL "TOTAL

or dividing by the density of air,

pMIX = "AIR + pflzo ""20 . (9)

pAlR "TOTAL pAlR "TOTAL
Next, from the ideal gas law,

Pv—fl-‘R'T=>—PM—fl=>—PM— (i)
‘ M RT _ V RT ' p

which leads to,

 

 

(ii)

 

 

 

 

 

 

 

mAIR 31%.
Bill. _ M AIR _1_§ M ”20
— 9
pAlR [mm + mHZO J 29[m,1m + mH20 ]
. MAIR Muzog \ MAIR MII20 J
"TOTAL "TOTAL

dividing the numerator and denominator by the mass of air, yields

"1
mAIR #—

P MIX _ mAIR M AIR 18 m4”? M ”20

_ + —
p AIR mAIR mH~0 29 mAIR mHoO
+ ‘ + ‘
M AIRmAIR M I-lzomAIR M AIRmAIR M HZOmAIR

 

 

 

 

 

or,

pm _ 1 18 (1) 1+0)

— +— = . IO
pAIR M I + (.1) 29M I + (.0 1+§m ( )
AIR M AIR M H20 H20 I 8

29

 

 

 

 

 

From the definition of specific humidity [23],

muzo 18 PH30

mAIR — 29 PAIR

 

 

23

 

29cc
PMIX : PAIR + PH20 : PAIR(1+—1-8_)
P _ PAIR _ PMIX
4’" " R T ' 2%)
"R (”—18 )RMRT

and for experiments conducted at constant PM“ ,

1

pAIROC °
29w

1 -—-T
(+18)

Substituting for p Mm and p Am , as indicated in (8) and above into (10),

 

 

o o l
(a+b)TI (+0”) , =CONSTANT.

11an

Next, substituting for b in terms of a from (2),

 

 

29 1
a(l+—m)TI (+23)) , =CONSTANT
#TG+ECO)
OI,

_ CONSTANT

(/(1+(n)T ' (11)

 

24

We define the molal flowrate of air at T = TREF , u) = 0 as the design air flow rate, ad . In

which case ad TREF = CONSTANT or,

_ TREF
" ‘adii(1+m)T (12)

Since the fuel flow rate is determined by the vacuum at the carburetor throat, which is

essentially fixed at a given throttle opening, then from (3),

 

 

 

 

 

 

 

 

AFRd AFR“
= 1.2042 =
ad ad
or,
a = AFR" . (13)
d 1.2042
which leads to,
AFRd
a = L40“ and b = 29“”
1+0)
TREF
and,
_ AFRd
AFR“ — (l4)

 

 

 

(«In

25

Equation 14 states that for any given AFRd (i.e. for a given fixed-jet carburetor in
which the throat area is the limiting area to air flow), the actual AFR (AFRO) is a function
of temperature and specific humidity. The AFR“ is the actual AFR in a constant pressure

test.

If there is an increase in specific humidity, the molal flowrate of air (a) will
decrease therefore the AFR“ will also decrease (become richer). Likewise, if the
temperature becomes greater than the reference temperature ( TREF ), the AFR“ will also

decrease (become richer).

Since the number of moles of a (air), b (H20) are determined using specific
humidity ((1)), Tm , T and AFRd , the molal flow rate of a and b is a function of AFRd ,
(o and temperature. This allows the final temperature and pressure, at the end of

compression to be calculated, for various operating ((0 , Tm ,T, AFRd) conditions.

3.1.2 Compression

Modeling the compression process can be accomplished in many ways: one can
choose to treat the process as ideal (isentropic, constant C. ); one can have isentropic
compression with temperature dependent properties, or one can have the actual case, with
specified heat transfer and temperature dependent properties. It was decided to model
compression as isentropic with temperature dependent properties, due to the difficulty in
either determining the entropy produced in such a process, or equivalently, prescribing

the heat transfer.

From the first law,

26

dQ—dW=dE

with

dE=dU.

If the process is ideal (i.e. reversible),

dW = PdV
dQ = TdS

Substituting the above relations into (15) yields, the Gibbs relation:

V
TdS—PdedUordS-fldf-=-d—;J-

(15)

(16)

(17)

Using the ideal gas law (P/ T = R/v ), constant entropy (d5 = 0) and the definition for

C, (C, = dU/dT) one obtains:

5d -CvdT
v v— T

 

Upon integrating both sides and rearranging, the following will result:

:2. _ 2.6;»
—R1n[vl]_.[l TdT

(18)

(19)

where CV is a flT) and is mass weighted for each species in the reactants, according to:

27

n
v.mix _ 2n
i=1 mi

 

(20)

Once the temperature dependence of the various specific heats are known, they
can be integrated from a reference temperature to any final temperature, T2 , and thus the
Right Hand Side (RHS) of the above equation (19) is a function of T2. Since the volume
ratio is the compression ratio, which is constant, the LHS is also known. This sets up an
iteration process, where the final temperature is guessed and compared to the value on the
LHS. The process is repeated until the LHS matches the RHS, at that point the guessed
temperature is the temperature which is sought and the process repeats for different initial
temperatures, humidity ratios and AFR’s. The compressed charge temperature T2 and
pressure P2 are the conditions at which combustion calculations begin, according to

figure 5 below.

 

Pec

 

 

Psc

 

 

 

thc Vbdc

Figure 5: Pressure - Volume Diagram

28

3.1.3 Combustion

Several computer codes were used in generating the theoretical data. Along with
the several FORTRAN codes utilized, a program called STANJAN [14] was used to
analyze, and predict, the resulting combustion products, properties and composition.
STANJAN is a set of FORTRAN computer programs developed at Stanford University
for the analysis of chemical equilibrium problems arising in combustion and other
applications. The program has the ability to treat mixture of ideal gases, condensed
phases and separated incompressible solids or liquids, using the method of element
potentials and the JANNAF thermo-chemical database. A driver program was written to
predict the end of compression temperatures and pressures, assuming isentropic
compression with temperature dependent properties. This program varied the specific
humidity and starting temperature to determine the final pressure and temperature of the
reactants. It then called STANJ AN to calculate the composition and equilibrium state of
the products, assuming a constant volume, constant energy process. The condition for
equilibrium was that the Gibbs function of the mixed products was minimized. In
can'ying out these calculations, it was necessary to specify the possible species which

comprise the products of hydrocarbon combustion.

The assumed products are taken from Heywood [19], pp. 93, figure 3-10-c. This
figure illustrates possible products at a combustion temperature of 2750 K, 30 atm for
various equivalence ratios. The list of the possible products is: N2 , H20 , CO2 , CO,
H2, OH, NO, N02, 02 and H.

The theoretical C0/ C02 ratio was calculated using chemical equilibrium
equations, taken from Spindt [8] and Heywood [19]. A good correlation between

temperature and data suggests the reaction is frozen at a temperature of 1700 K [8], [19].

29

This suggests a value for K p of 3.388, where K p is the equilibrium constant. See

appendix D for a sample of equations used.

3.2 Theoretical Results

Using the assumptions and equations of the previous section, a series of
carburation, compression and combustion calculations were carried out at different

ambient conditions.

The effect of ambient temperature on AFR (at (1) = 0.005), as a function of design
AFR is shown in figure 6. The design AFR (AFRd) is the AFR that one obtains using
specified carburetor parameters, i.e. for a specified throat diameter and jet size. The
actual AFR (AFRa) is the one delivered to the engine by the carburetor, after the effects
of temperature and humidity have been factored in. One can see that as the room
temperature increases, the AFR“ becomes richer. This is expected, as the air becomes
slightly less dense, there is more fuel available and thus the AFR is altered to a more rich
setting.

Figure 7 shows the peak combustion temperature vs. AFR at a constant a).

This plot indicates that for a fixed AFR, the initial temperature shifts the peak temperature
upwards. The temperature increases to a maximum around an AFR“ of 14:1 and then the
combustion temperature starts to fall off, as expected. It is well known that best power is
obtained at a slightly rich setting, stoichiometric being 14.7:1. The peak temperature at
an AFR of ~14:l would theoretically deliver the maximum temperature and thus the

maximum power. The effect of humidity is to shift the entire curve down as water vapor

is added to the mixture.

30

Actual AFR Vs. Design AFR

 

 

 

 

 

 

 

 

 

 

(w = 0.005)

16

15 re

14 ee

13 i
E
'3 12 fl Tl Warmer
a
2

11 1e

10 ~

9 3:.
8 u I i i i l
9 10 ll 12 l3 14 15
Design AFR
Figure 6: Actual AFR Vs. Design AFR
Peak Combustion Temperature Vs. Actual AFR
(w = 0.005)

2800
Q 2700 -_
c,
2
g 2600 ‘"
8. TI Warmer
E
‘2 2500 e
o
"3
.8
E 2400 'r
0
U 320
E .
O- 2300 280

2200 44— —I e _.- l —1 I l 1

8 9 10 l I 12 I3 14 15 16
Actual AFR

Figure 7: Combustion Temperature Vs. Actual AFR

31

Figure 8 shows the combustion temperature as a function of the AFRd. Again,
the design AFR is the AFR that one obtains using specified carburetor parameters, i.e. for
a specified throat diameter and jet size. It does not include correction factors for ambient
temperature and humidity. This curve is generated using the same code, except the AFR
is not modified for ambient conditions. On this plot, Tl =280 yields the highest
temperature, while Tl =320 yields the lowest temperature. It is interesting to note how the
curves transpose after an AFRd 13:1. This is not the case when the peak temperature is
plotted versus AFR, , indicating that carbureted engines performance cannot be

optimized without taking into account the local ambient conditions.

Peak Combustion Temperature Vs. Design AFR
(w = 0.005)

 

2800

     

I

to N
O) \1
8 8
I I

N
u.
8

I

2400 ‘

 

Peak Combustion Temperature (‘ K)

 

 

I J
I I

9 10 ll 12 l3 14 15
Design AFR

Figure 8: Combustion Temperature Vs. Design AFR

32

Figure 9 shows the effect of ambient humidity on AFR (at T = Tm, ), as a function
of AFRd. Adding moisture to the ambient moisture levels did not effect the AFR“ as
strongly as the temperature change did. The effect of increasing humidity has less of an
effect on the AFRO compared to increasing ambient temperature. The temperature
change resulted in a 6.7% decrease in the theoretical AFRa , a corresponding humidity

change netted only a 0.78% drop in the theoretical AFR“ .

Actual AFR Vs. Design AFR

 

 

 

 

(Tl =Tret)

15

14 «e

13 ~-
E mINCREASING
<
7.1 12
a
O
<

11

10 ‘P 0.0

0.015
9 1r i 1 l 1
9 10 ll 12 13 14 15
Design AFR

Figure 9: Actual AFR Vs. Design AFR

Figures 10 and 11 show the peak combustion temperature vs. AFR“ and AFRd,
respectively (at T = TRE, ). The effect of adding moisture to the surroundings, shifts the
peak combustion temperatures lower. This is due to the fact that the energy of the
products is raised, so a lower temperature is required for chemical equilibrium at the

same mixture energy.

33

Peak Combustion Temperature Vs. Actual AFR

 

  

 

 

 

 

   

(T l = Tref)

2800

2750 “"
M
(.2 2700 "
2
E 2650 4- (”INCREASING
32600 .
O
E; 2550 i
o
-.:1
‘3 2500 ‘
E
8 2450 ‘
E 2400 .
o.

2350 ‘

2300 4 + F i 1

9 10 11 12 13 14 15
Actual AFR
Figure 10: Peak Combustion Temperature Vs. Actual AFR
Peak Combustion Temperature Vs. Design AFR
(T1 = Tref)

2800

2750 if
c, 2700 1*
B 2650 - (”INCREASING

   

 

 

1»-
-(t-

 

9 10 11 12 l3 I4 15
Design AFR

Figure l 1: Peak Combustion Temperature Vs. Design AFR

CO / C02 Ratio

CO / C02 Ratio

 

34

CO / C02 Ratio Vs. Design AFR
(0) = constant)

 

 

 

 

Design AFR

Figure 12: CO/COZ Ratio Vs. Design AFR (1)

CO / C02 Ratio Vs. Design AFR
(Tl = Constant)

 

    

0) increasing

 

 

 

9 10 ll 12 13 I4 15

Design AFR

Figure 13: CO/CO2 Ratio Vs. Design AFR (2)

35

Figure 12 and 13 show the theoretical C0/ C02 ratio as functions of AFRd.
Figure 12 illustrates the effect of increasing ambient temperature at a constant humidity.
As the temperature becomes warmer, the air becomes less dense, thus altering the AFR“
to a more rich setting. As the AFR“ becomes richer, more CO is produced and less CO2
is formed (due to the lack of oxygen). Likewise figure 13 shows the effect of increasing
ambient humidity, while holding ambient temperature levels constant. This figure shows
the effect on AFR“ as humidity is increased. This enrichment causes more CO to be

produced and less CO2 to be made, thus increasing the C0/ C02 ratio.

While the variation of C0/ C02 with changing humidity is slight, the

dependence on ambient temperature is significant, especially at rich AF R ’3 (figure 12).

3.3 Experimental Results

Three engine tests were performed: an idle test varying the surrounding
temperature; a WOT test varying the humidity at constant temperature; and a WOT test

varying the temperature at constant humidity.

The tests carried out at WOT are the most applicable to theoretical analysis and

their results are described in detail below.

3.3.1 Humidity Variation

Figure 14 displays room temperature versus specific humidity. This shows that

room temperature remains essentially constant throughout the entire test. The

temperature varied a total of 1.4%.

36

Engine performance changed little with modest changes in the surrounding

humidity. The humidity was varied from ~ 0.015 (kgHzO/kgAIR) to .006

(kgI-IzO/kgAIR) or -150%. Throughout this range, the torque output remained

essentially constant, varied slightly from .75 N-m. Figure 15 shows torque vs. specific

humidity. At the higher humidities, torque decreases slightly with increasing humidity;

this is consistent with the dry air mass in the cylinder decreasing with increasing 0) as

predicted by (14).

Brake Specific Fuel Consumption (BSFC) is also an important engine operating

parameter. It is given by the following equation:

fuel fl0w(gr/ hr)
p0wer(kW)

 

BSFC =

Room Temperature Vs. Specific Humidity
(T1 = Constant)

A
O

 

5128188
e1fi4
‘0
O
‘I
O
O

’0. O

TemperatureCC)
NRNNMM
N asooow
e4ieii

 

N
O

_ l I
I I

0.003 0.005 0.007 0.009 0.01 I 0.013

w (kg Hzo/kg AIR)

Figure 14: Room Temperature Vs. Specific Humidity

 

0.015

(21)

37

Figure 16 shows BSFC vs. specific humidity. It is shown to increase slightly over
the range of the test. This is expected, as the moisture level is increased, the AFR will
decrease slightly. This slight decrease in AFR (richer mixture) causes the air flow rate to

rise, causing the engine torque to slightly increase.

Torque Vs. Specific Humidity
(Tl = Constant)

 

1.00
0.95 e
0.90 .-
A 0.85 A)
- 0.80 ..
g 0.75 —~ .. ° ° 0 .
g 0.70 -- " .
“ 0.65 -.
0.60 ~—
0.55 “i
0.50
0.003 0.005 0.007 0.009 0.01 1 0.013 0.015

\°°
O

 

 

I I A
I I I

 

~0-
~0—

w (kg HZO/kg AIR)

Figure 15: Engine Torque Vs. Specific Humidity

CO is plotted versus ambient humidity as in figure 17. CO increases slightly
with increasing specific humidity, and this is expected. As the air becomes more humid,
the AFR ratio will decrease, resulting in a richer mixture. As the mixture becomes richer,
more CO is produced, due to the decreasing amount of oxygen, which is needed for

stoichiometric combustion.

38

BSFC Vs. Specific Humidity
(Tl = Constant)

1600
1500 i
1400 -~ °

1300 «P °
1200 5~ 0

”00 “b 0. o o 0 °
1000 5*

 

BSFC (gr/Kw-hr)

0.,

 

 

 

800
0.003 0.005 0.007 0.009 0.01 1 0.013 0.015

~1-
—1I—

1 l 4
I T T

w (kg HZO/kg AIR)

Figure 16: BSFC Vs. Specific Humidity

CO2 is plotted against specific humidity in figure 18. As the specific humidity
level increases the mixture AFR decreases (becomes richer), resulting in less oxygen
available for combustion. With less oxygen, less CO2 is produced. This is seen in the

figure.

The trends for NOX reveal a similar pattern to that observed for CO. Figure 19
illustrates how the NOX level decreases as the humidity increases (AFR decreasing). As
the AFR decreases, the combustion temperature drops. When the combustion
temperature drops, the formation of NOx is lowered significantly, due to the large

temperature dependence.

C02 (gr/hr)

39

CO Vs. Specific Humidity
('1‘ l = Constant)

 

 

l I l I l

 

 

I I I I I

0.005 0.007 0.009 0.01 I 0.013

w (kg HZO/kg AIR)

Figure 17: CO Vs. Specific Humidity

C02 Vs. Specific Humidity
(T I = Constant)

0.015

 

1600
1500 4»
1400 i
1300 ._
1200 ._
1100 4
1000

O

A I 4% A l

 

 

 

800 .
0.003

I I I I I

0.005 0.007 0.009 0.01 1 0.013

w (kg HZO/kg AIR)

Figure 18: C02 Vs. Specific Humidity

0.015

40

Finally the HC production is shown to increase drastically, from 21.73 (gr/hr) to
31.88 (gr/hr). This corresponds to a increase of 46.7%, over a humidity change of
0.01487 kg AIR/H20 to 0.00578 kg AIR/H20 (figure 20) The decreased AFR with (0
is probably a contributing factor. Also the decreasing combustion temperatures are likely

to result in lower exhaust gas temperatures, and possibly reduced post-exhaust oxidation.

NOx Vs. Specific Humidity
(T1 = Constant)

 

N Ox (gr/hr)

0.5 5*

 

 

 

0.0 i t i + i
0.003 0.005 0.007 0.009 0.01 1 0.013 0.015

w (kg HZO/kg AIR)

Figure 19: NOx Vs. Specific Humidity

Figure 21 shows the AFR versus the specific humidity. The AFR is shown to
decrease as the humidity is increased. This minor change in AFR from 11.92 to 12.41

(3.95%) causes significant changes in the combustion products and process.

41

HO Vs. Specific Humidity
(T1 = Constant)

 

l6 1” °
15 .- u
14 .. .3

HC (gr/hr)
{'3
O

12 .. g.
11 ._
10 ~-

 

 

8 i 4 A I I
I T I T 77

0.003 0.005 0.007 0.009 0.01 I 0.013 0.015

 

w (kg HZO/kg AIR)

Figure 20: HC Vs. Specific Humidity

Lastly, figure 22 shows the ratio of CO/CO2 as measured by the instruments (the
diamond symbol) and the theoretical prediction (the square symbol). Here the AFRd was
set to 14.25 in order for the appropriate magnitudes to become similar. This is thought to
arise from the compression model used. The ideal assumptions employed predicted
combustion temperatures which are believed to be higher than the actual values (due to
neglecting heat transfer). Since C0 formation is extremely sensitive to combustion
temperature [19], a slight reduction in temperature causes drastic decreases in C0
production. In addition C0 has been shown to increase with specific humidity, CO2
must decrease with increasing specific humidity in order for the ratio of CO/CO2 to
increase. This trend is shown in the figure. CO increases with decreasing AFR, while

CO2 decreases with decreasing AFR.

AFR (kg AIRI kg FUEL)

CO / C02 RATIO

42

AFR Vs. Specific Humidity
('1‘ l = Constant)

 

 

 

 

 

 

 

 

 

 

 

15.0
14.5 —~
14.0 «F
13.5 -r
13.0 -- O
’0
D O .
12.5 ~~ ’ °I°o
O O O
O
12.0 I I I I 4.
0.003 0.005 - 0.007 0.009 0.011 0.013 0.015
w (kg HZO/kg AIR)
Figure 21: AFR Vs. Specific Humidity
CO/C02 Vs. Specific Humidity
(T1 = Constant)
0.25
O
O
O
0.2 -- 3" ' ' '
J'- n I '
.0
0.15 ._
0.1 (
0.05 e _
o CO/C02 (Measured)
I CO/C02 (Theoretical, AFRdr.l4.25)
0 1 i + I 1
0.003 0.005 0.007 0.009 0.01 1 0.013 0.015

w (kg HZO/kg AIR)

Figure 22: CO/C02 Ratio Vs. Specific Humidity

43

3.3.2 Temperature Variation

The temperature variation proved to be a more difficult test to perform. The data
has been truncated to include only the data that was maintained at constant humidity. As
the room temperature was increased, the moisture level rose to the outside humidity value
and leveled off. Thus it was hard to obtain strictly temperature dependent properties.
Figure 23 illustrates the specific humidity dependence on temperature. This figure
illustrates how the humidity value increased with increasing room temperature. The

theoretical trend in AFR with T is restated below for convenience.

(14)

 

 

Engine torque is shown to decrease with increasing room temperature, as shown
in figure 24. This is expected as the air is becoming less dense and the AFR is also
decreasing (mixture is becoming richer). The richer mixture and the decreasing density
of the air have a combined effect on engine performance. First the decreasing density of
the air lessens the cylinder filling capabilities and second the decreasing AFR lowers
combustion temperatures thus decreasing engine performance. The AFR decreases by a

total of 1.96%.

Specific Humidity Vs. Room Temperature
(w = Constant)

 

0.0040
0.0038 5L
0.0036 5* 0 °
0.0034 ~-
0.0032 ._
0.0030 ‘—
0.0028 ~~
0.0026 *-
0.0024 ~—
0.0022 re
0.0020 t i i i

20 22 24 26 28 30

. . . 090000000000

Specific Humidity (kg HZO/AIR)

 

 

 

Temperature (‘ C)

Figure 23: Specific Humidity Vs. Room Temperature

BSFC also varied with increasing room temperature as demonstrated with figure
25. BSFC decreased as a direct result of the decrease in engine torque (The AFR
decreases by a total of 1.96%). With the AFR decreasing only slightly and the torque
falling off significantly (torque decreased 17%), the BSFC has to increase. BSFC is a

function of the fuel flow rate and power produced by the engine, as shown in equation 20.

The emissions also reflected the change in the surrounding temperature. C0
production increased 25% with only a 1.96% decrease in the computed AFR. Figure 26
shows the CO as a function of room temperature. As the temperature becomes warmer,

the mixture becomes richer, therefore more CO is produced (due to the lack of oxygen).

BSFC (ngkW-hr)

88

1600
1500
1400

‘8’

1200

—I
~
I
T

8

45

Torque Vs. Room Temperature
(w = Constant)

 

I I

 

0000
000.

 

 

-1>-
«I-
J-
db

22 24 26 28

Temperature (‘C)

Figure 24: Engine Torque Vs. Room Temperature

BSFC Vs. Room Temperature
(w = Constant)

30

 

i-

d)-

l
f

4»

l
f

l
I

 

O
.o0°0°
0....

l
I

-11-
-1I-

 

 

N
O

di—

22 24 26 28

Temperature (‘C)

Figure 25: BSFC Vs. Room Temperature

30

46

CO Vs. Room Temperature

 

 

 

 

(w = Constant)
240
220 _-
200 --
E 180 -_
59
8 160 T o . . o 0 °
0 o .
I40 3“ . o
9 O
120 ~— . 0 °
0
100 1r t ‘fi I
20 22 24 26 28 30
Temperature (‘ C)

Figure 26: CO Vs. Room Temperature

The production of CO2 was also effected by the changing ambient conditions.
The production of CO2 decreased with the increase in ambient conditions, as expected.
With a decreasing AFR, the mixture becomes richer. As the mixture becomes richer,
there is less oxygen available to produce CO2 , so more CO is made. Figure 27

illustrates the decrease in CO2 production.

Likewise, NOx was similarly effected. Figure 28 shows NOX as a function of
the surrounding temperature. NOX decreased 16.1%, but more resolution in the meter
was needed. The span gas chosen was too rich, therefore the measured concentration
shows step decreases as the room temperature is increased. Regardless of resolution, the
NOX decreased as a direct result of the decreasing AFR. The decrease in AFR results in

lower combustion temperatures, thus less NOX is produced.

CO2 (gr/hr)

NOx (gr/hr)

47

C02 Vs. Room Temperature
(w = Constant)

 

 

 

 

 

 

 

 

 

1600
1500 re
0
1400* ...°°°O‘ooo°oooo
1300 r-
1200 rr
1100 1r
1000 re
900 _
800 I I I I
20 22 24 26 28 30
Temperature (‘C)
Figure 27: C02 Vs. Room Temperature
NOx Vs. Room Temperature
(w = Constant)
2.5 . , . O
o o 9 ° 9 9
2.0 ‘i' o o o . o . .
1.5 r
1.0 ..
0.5 re
0.0 rr—— - I --~I— i l —
20 22 24 26 28 30

Temperature (‘C)

Figure 28: NOx Vs. Room Temperature

48

HC production is illustrated in figure 29. The production of HC is not as drastic
as the humidity change, never the less, the small decrease in AFR results in a small

increase in HC production. This increase corresponds to a jump of 4.52%.

The AFR has already been discussed in some detail, but it is shown in figure 30.
The AFR decreases only 1.96% over the testing range, but this rather minor change in
AFR brings about modest changes in the emission levels produced by small engines in

general.

'HC' Vs. Room Temperature
(w = Constant)

 

 

 

 

18
17—
16(-
15+
@144-
3913*
Eizee
II~~
13: O .oo.°o.°...’....
8 I I i l
20 22 24 26 28 30

Temperature (‘C)

Figure 29: HC Vs. Room Temperature

49

AFR Vs. Room Temperature
(w = Constant)

 

15.0

14.5 re

14.0 1*

13.5 "F . . . Q Q

13.0 r“

AFR (kg AIR/FUEL)

12.5 r-

 

 

 

12.0 i i I l
20 22 24 26 28 30

Temperature (‘C)

Figure 30: AFR Vs. Room Temperature

Finally the measured CO/CO2 ratio (diamond symbols) is compared to the
theoretical results (square symbols). This is shown in figure 31. The AFRd had to be
raised to 14.25 in order for the magnitudes to be nearly identical. Again, this is thought
to arise from the compression model used. The ideal assumptions employed predicted
combustion temperatures which are believed to be higher than the actual values (due to
neglecting heat transfer). Since CO formation is extremely sensitive to temperature, a
slight reduction in temperature causes drastic decreases in CO production. The ambient
temperature change is thought to bring about more effects than just changes in ambient
energy levels. It is known to be responsible for fuel evaporation rates, heat transfer from

the engine, air density and others.

CO / C02 (ratio)

50

CO / C02 Vs. Room Temperature
(w = Constant)

 

 

 

 

 

 

 

 

0.25
0.20 re
0.15 ie
010i» ..---:""""
' I I O O
o 9 °
0.05 «L
o CO/C02 (Measured)
I CO/C02 (Theoretical, AFRd=14.5)
0.00 l 4 1 I
20 22 24 26 28

Temperature (‘C)

Figure 31: CO/C02 Ratio Vs. Room Temperature

30

1)

2)

3)

4)

4. CONCLUSIONS

In general, changes in ambient conditions lead to small changes in the actual
AFR in carbureted IC engines. These AFR changes produced modest changes
in measured emission levels. However, the effect of ambient humidity on
flame temperature is particularly large. For example a 1.96% increase in

AFR“ (becoming leaner) resulted in a 50% increase in N0x production.

The performance and emissions measurements: BSFC; C0; C02; NOX;
AFR and C0/C02 ratio all exhibited clear trends with changes in ambient
conditions. Increasing specific humidity or temperature caused an increase in

BSFC, C0 and, CO/CO2 while reducing C02, NOX and AFR.

CO/CO2 formulations are highly temperature and AFR dependent therefore
they are difficult to predict from theory. Since the isentropic compression
model was assumed not to include any heat transfer from the charge, this
estimate of the compressed charge temperature is probably higher than its
actual value. Selecting the lower limit at which equilibrium thermodynamics
applies appears to be 1700 K (Spindt [8], Heywood [19]). This leads to a

reasonable agreement with experimental results.
Changes in ambient temperature not only effect air density, but also fuel

evaporation rates, heat transfer from the engine and compression properties.

In contrast changes in humidity seen only to affect the charge composition and

51

52

its properties. Therefore effects of changing ambient temperature on

emissions are potentially much more complex than changing humidity.

5) More research may be needed in the area, in order to better understand how

changes in ambient humidity and temperature effect engine performance.

Appendix A

Appendix A

Apparatus

A.l Fuel Flow Measurement

The fuel flow cart uses a Micro-Motion mass flow meter. This meter operates
using the Coriolis effect. The flow tubes of the Coriolis mass flow sensor are driven at
their natural frequency by a magnet and drive coil attached to the center of the bent u—
tube, see figure 32. The vibrating motion of the u-tube, combined with the momentum of
the fluid flowing within the tube, induces a Coriolis force that causes each flow tube to
twist in proportion to the rate of mass flow through the tube during each vibrational cycle,
see figure 33. Since one leg of the flow tube lags behind the other leg during this twisting
motion, the signals from sensors on the two tube legs can be compared electronically to
determine the amount of twist, see figure 34. The transmitter measures the time delay
between the left and right pickoff signals using precision circuitry and a high frequency
crystal controlled clock. This delta time value is multiplied by the flow calibration factor
to determine the mass flow rate then the resulting flow rate is displayed by the Digital
Rate Totalizer (DRT) and sent to the PC [24]. Other sensors on the flow cart measure the
temperature and relative humidity within the room and this data is also sent to the PC for
“storage”. See Figure 35 for a fuel flow schematic. After the fuel leaves the flow cart, it

is sent to the carburetor.

53

54

 

 

 

 

 

 

 

 

Figure 32: Vibrating flow tube (single flow tube shown)

 

 

 

 

 

Figure 33: Fluid Forces Reacting in Vibration of Flow Tube

 

 

 

 

 

 

 

 

Figure 34: End View of Flow Tube Showing Twist

 

 

 

55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pressure
Storage Fuel Regulator
Tank Cooler
. Flow
Eng” Float Meter to 486 PC
Bowl ‘ __>

 

 

 

 

 

l____> Display

 

 

 

 

 

 

Figure 35: Fuel Flow Cart

A.2 Dynamometer and Controls

The dynamometer measures various engine operating parameters, including
engine torque, engine RPM, cylinder head temperature, exhaust temperature, oil
temperature and fuel temperature. The dynamometer is a Micro-Dyn model 15, which is
capable of dissipating approximately 20.4 N-m (15 ft-lbs) of torque. The maximum
engine operating speed for this dynamometer is 14,000 rpm. It is a low inertia, hydraulic
dynamometer which dissipates the energy produced by the engine through an oil to air
heat exchanger. The dynamometer is computer controlled, with settings for speed,
throttle opening or other test operating conditions (such as transient operation). It uses a
hydraulic pump driven by an electric motor as the working fluid. The fluid flow rate is

controlled by the computer via a stepper motor, which governs the engine RPM (figure 36

56

shows a fluid flow schematic). With the RPM set by the flow rate, one is ready to load
the engine, by setting the throttle opening to the desired setting. In between the engine
and hydraulic pump is the torque sensor, a Lebow model 1804-500. This torque meter
measures the torque output from the engine, and sends the appropriate electrical signal to
the computer. The torque output from the engine can be displayed either as the
instantaneous torque through one engine cycle, or as an integrated “mean” torque using

an RC circuit.

 

 

 

 

Manual
Pump Bypass

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

‘
Fluid
Reservoir
Note: Driven Motor is
Driven by Engine
Heat Exchanger ‘ Driven
Motor

 

 

 

 

 

 

 

Figure 36: Dynamometer Hydraulic Circuit

57

A.3 Environmental Controls

The temperature and humidity within the room were monitored and controlled by
a set of Honeywell Industrial controls. These controls were specifically designed for this
purpose. A Honeywell model T775-E controls the humidity within the room while a
Honeywell model T775-A monitors the temperature. They are connected to a steam

generator, a heating unit and a refrigeration unit. See figure 37 for a schematic.

 

 

 

i Humidifier I

2, *1

NC Unit Heater

 

 

 

 

 

 

 

 

 

 

 

Figure 37: Temperature and Humidity Controls

58

AA Dilution 'Ilunnel

The dilution tunnel consists of a 4” diameter pipe, an air bleed valve (which
allows excess air to be drawn into the compressor, assuming the nozzle has a smaller
flow rate than the compressor), a nozzle and a ring compressor. There is also a secondary
air inlet valve, down stream of the primary air inlet valve, to adjust the suction that is
imposed on the exhaust exit. Some engines are susceptible to decreases in exhaust exit

pressure and hence the performance can be altered if this is the case.

When the exhaust leaves the engine, it is drawn into the dilution tunnel along with
large quantities of room air. The effect of this dillution is to freeze any further reaction
by the exhaust products, and removes the need for a heated sample line. The flow rate in
the tunnel is governed by a critical flow nozzle, placed ahead of the ring compressor.
This nozzle can be changed to tailor the flow rate, depending on the desired dilution ratio.
For example, with the NOX meter set on its lowest range and a four stroke engine idling,
very little NOX will be produced. If greater accuracy is desired, the nozzle can be
changed to lower the amount of dilution air in the tunnel, thus increasing the relative
amount of NOX present. This will increase the value read by the meter, but not change
the actual output by the engine. The sample line is taken from just ahead of this nozzle
where there are static temperature and pressure probes. These are used for determining
the flow rate, using Bemoulli’s equation. After the exhaust gases pass through the ring
compressor, they are sent through an exhaust duct, and expelled to the atmosphere (see

Figure 38).

59

 

Emission Sample Line
‘ (to Horiba Instruments)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Exhaust and Room Air Static Pressure
and Temperature
-—-—> Taps ('1‘ o 486 PC)
Engine
Nozzle
Dynamometer

‘_ Air Bypass Valve

— (if Needed)

0
_—_>

Compressor To Exhaust Ducting

 

 

 

 

 

Figure 38: Dillution Tunnel

A.5 Emission Measurement

With the sample coming from just ahead of the critical flow nozzle, after the
exhaust mixture has traveled through the exhaust tubing for almost 15 feet, one can
assume it is well mixed. The sample line takes a small sample and pipes it over to the
Horiba Bag Bench. Each analyzer measures the species of interest using a different
principle, measuring each constituent as a volume fraction of the total exhaust gas
sample. The C0 and C02 meters, employ a non-destructive measurement technique.
This involves passing specific wavelength of infra-red light (IR) through the sample (see
figure 39). An IR detector on the opposing side, detects how much IR light has passed
through the sample. Since C0 and C02 absorb IR light (over different frequency

ranges), the greater the concentration of C0 or C02 in the sample, the less IR light will

pass through. From the Amagat model of ideal gases, the volume fraction is proportional
to the mole fraction present, and therefore the volume fractions of C0 and C02 can be

found.

The HC meter uses a different approach and is “tuned” to specific carbon
compositions within the species. For example, the composition of two stroke exhaust
mirrors the composition of the two stroke fuel. On the other hand, the composition of
four stroke exhaust, is mostly composed of Cl and C2 molecules [6]. Therefore one
must use a span gas, that closely resembles the exhaust gas in question. If this is not
done, false readings can result. The analyzer detects the amount of HC’s present in the
sample using the flame ionization technique. The flow of a sample with carbon
compounds through a hot flame leads to ionization of the gas and an ionization current.

This current is proportional to the volume fraction of H C ’3 present within the sample.

The NOX meter uses a chemical illumenescent process. Nitrogen compounds in
the exhaust gas are actually a mixture of N0 and N02 , which is usually written as NOX .
In the analyzer, the N02 is first converted to NO. The resulting sample then reacts
with 03, which is generated by an electrical discharge through oxygen, at a low pressure
in a heated vacuum chamber. The light emitted by the reaction is measured by a photo-
multiplier and indicates the NOX concentration within the sample. N0 and N02 are not
stable at room temperature. It is for this reason we refer to them simply as NOX . If one
takes and measures a sample of N0 and N02 and places it in a bag for storage, the
concentrations of each species will change with time and temperature. See table 5 for an

overview of the emission measurement techniques.

61

 

 

 

 

Thermocouple

QI’Lq-onm...

Collector ‘
electrode

Burner . .; ‘

  

I - Sample gee

 

 

 

 

 

 

1 Receiver chember with eompeneetton
epeeee V. end V3. 2 Flow eeneor.

3 Meeeurlng cell. 4 Routing chopper die:
with motor. 5 lettered emitter

 

 

 

 

 

 

 

 

 

 

 

 

Figure 39: Analyzer Cross Sections

From Top to Bottom: Flame Ionization Detector, Non Dispersive Infra-Red Analyzer and
Chemiluimenscence Detector

62

Table 6: Emission Measurement Techniques

 

 

 

 

 

 

 

 

 

TYPICAL
GAS TECHNIQUE RANGE
CO Non dispersive 0-3000
Infra Red (NDIFl) ppm
Non dispersive o
002 Infra Fled (NDIR) (”20 A
NOx Chemiluminescence 11:?
UNBURNED Flame Ionization 1-10,000
HYDROCARBONS (HC's) Detector (FID) ppm

 

A.6 Data Collection

After all the components are measured, the resulting voltages are digitized
through one of three Data Translation digital to analog (D/A) boards. The IBM 486 PC
uses two voltage conversion boards and one thermocouple board for reading all the
pertinent room data to the FORTRAN code for analysis (see figure 40). In addition to
these three D/A boards, the fuel flow cart transmits its data through an RS-232 serial port
into the IBM 486 PC. Next, the program SAMPLE takes ten samples of each value in
question, and uses the average to calculate a single data point. Then SAMPLE is run two
more times to obtain two more data sets, again each one is an average over ten individual
samplings. Convert was written to read in all the data values outputted from our program
SAMPLE. This code read in the raw data (in table form) took the average for three data
points, and output the data (in column form) to a specified file name. The average of all

three results represents one data point in my data. See figure 41 for clarification.

63

 

 

Torque and Emission inputs

 

  

 

RS-232 (From Fuel Cart)

Pressure, RPM
and other Analog
. vagxagg‘yszs V" .ng
Inputs

 

 

Thermocouple
Inputs

 

Screw Terminal Panels

 

Figure 40: Data Collecting Computer (486 PC)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I
Program Single
‘0 SamP'“ ’ SAMPLE ' Output
I
+
fl
Program Single
'0 53mph” ’ SAMPLE ’ Output
1 .

 

 

 

 

Program
10 Samples —-> SAMPLE

 

 

 

’ Single
Output

 

 

 

 

 

 

 

 

l

Average of Three
Single Outputs
(is one data point)

 

 

 

 

Figure 4]: Data Collecting Algorithm

 

 

Appendix B

Appendix B

Scaling Analysis

In order to verify or check the constant U, constant v assumption, a scaling

analysis of the first law will be performed.

From the first law,

dQ—dW=dE (l4)

Dividing by the change in Gibbs free energy,

dQ dW_dE
AG AG‘AG

 

(22)

Each term in equation (22) must be estimated, and the relative magnitudes are then

compared.

65

B.1 Heat Transfer

Starting with the heat transfer, dQ, which is estimated using a Woshni [15], [16]

correlation. Starting with the basic heat transfer equation,

(1... =hA(T.. — T)

a sensible estimation for the heat transfer coefficient (h) is as follows,

8 (VMOT + VCOMB )0'8

 

 

h : 0'0032P0. 30.2 T053
where,
VMOT = CIVPISTON
VdT:
VCOMB = C2 PIVI (1" PMOT)

C, and C2 are constants, and given in the following table,

Table 7: Table of Woshni Coefficients

 

 

 

I Compression Combustion - Expansion I
[C. 2.28 2.28 |
IE 0.00 0.00324 |

 

 

 

The average piston speed is erozv and is defined by, Stroke/ Time. Va, is the

displacement volume, T, , P, and V, are evaluated at a reference condition, say inlet valve

66

closing. P is the cylinder pressure and PM, is the pressure under motoring conditions.
PW, , is assumed to be the pressure at the end of compression. B is the cylinder bore and

T is the cylinder temperature. The total heat transfer is estimated to be 391.7223 J / kg .
B.2 Work

Work in and out of the engine during the compression / expansion process is

assumed to follow the integral:

w = deV (23)

The combustion is assumed to occur over a finite period of time, or a specific
amount of crank angle degrees. Heywood, [19] estimates for rich mixtures, the
combustion rate is higher than for lean mixtures. Small engines are set to run on the rich
side, and therefore the crank angle needed for combustion is assumed to be 50° (pp. 382,
figure 9-8-a). A schematic of a P-v diagram is included below (figure 42). The net work

out, W

net ’

is the total area under the curve, or

 

Vsc thc Vec

Figure 42: Work inputs and outputs

67

Where the Subscripts in the picture are defined as following:

1. P - Pressure
V - Volume
sc — Start of Combustion

tdc - Top Dead Center

9‘95”!"

ec - End of Combustion

V“. is point 1, V“ is point 2. The pressure change is assumed to be polytropic, so

the pressure rise follows the equation:
&
(PV)CV = CONSTANT (24)

Since the peak pressure is known, the pressure at start of combustion can be
backed out using an estimated volume (specified crank angle). Likewise the end of
combustion pressure can also be found. With the pressures and volumes known, the
work input (piston is moving upwards, pressure is rising) and the work output (piston is
moving down, pressure is falling) can be found. The figure illustrates the work output is

much larger than the work input, as it should be.

The displacement was found using the physical engine parameters and a relation
found in Taylor [17]. The swept volume is a function of cylinder bore, stroke and rod
length. The stroke length prior to TDC was estimated as 0.00049635 m and that after
TDC was 0.00373235 m, assuming a total combustion time of 0.00119 seconds (50 crank
degrees @ 7000 RPM).

68

With these assumptions and equations, the work input during the combustion
process was estimated to be 299.144 J / kg , the work out during the combustion /

expansion was estimated to be 510.529 J / kg , so the net work, W is the difference of

net ’

the two, or 211.38 J/kg.

B.3 Internal Energy

The internal energy of the system, E (= U) is assumed to be constant throughout

the combustion process. Therefore the RHS of the first law (equation 22) is zero.

B.4 Change in Gibbs Free Energy

Gibbs free energy is defined to be:

G=H—TS (25)

and H is U-PV. Substituting into (25)

GzU—PV—TS (26)

The change in Gibbs free energy is the difference between two points or,

AG = G, — a, (27)

Since U 2 = U, and V2 = V, , the difference in the free energy is:

AG =(P1V. -T.S. )-(P2V2 -T252)

69

since the volume ratios are the same, the above reduces to,

AG=(P,-P2)V—7,S,+T2S2 (28)

From combustion analysis [14],

1. P, = 1.5189 e6Pa

P2 = 8.6444 e6 Pa
T, = 549.69 K
T, = 2883.87 K

s, = 6.6596 e3 J/kg — K

.O‘P'PP’.“

s2 = 8.7624 e3 J/kg— K

The change in Gibbs free energy, AG , is estimated as 2.0903 e7 J / kg .

From (22), substituting in the appropriate values:

2.9. d_W._
AG AG-
391 211

 

2.09e7 _ 2.09e7 E 0

This shows that our constant U, constant v assumption is valid.

Appendix C

Appendix C

Compressibility Estimation

In order to neglect the effects of compressibility through the carburetor throat,
they must not contribute a significant part to the flow equation. Heywood [19] states:
“For flow rates less than about 60 percent of the choked flow, the effects of

compressibility on the mass flow rate are less than 5 percent (pp. 909).”

Choked flow is defined as flow when the Mach number is equal to one, or

M: :1 (29)

n|<

where 117! is the Mach number, c is the speed of sound and V is the observed

velocity. If the Mach number is equal to 1, V = c. From:
m = pAV (30)
The velocity of air through the orifice can be found. Assuming 100% volumetric

efficiency (worst case), a four stroke engine takes in its displacement in air every other

revolution, or,

70

71

 

 

 

rev lch arg e ch arg e l min 26cm3 cm3
7000 , = 3500 , = 1516
mm 2rev mm 603cc ch arg e sec

 

1516cm3( 1'" )3—00015171'3— (31
100cm " )

S86 SCC

Equation (31) is the volume flow rate of air through the venturi, assuming 100%
volumetric efficiency, and only dry air flowing through the restriction. In order to obtain

the velocity, divide the volume flow rate by the area it is flowing through,

 

‘ pA " pA ’ A
or,
0.0015l7m3/sec _ 53 631_m_
2.8257e —5m2 — ' sec .
In terms of the Mach number,
. V 53.631
M = - = — = . 622
c 343.3 015

or 15.6%. Heywood states the effects of compressibility are only relevant when the Mach
number is about 60%, so the effects of compressibility are neglected in this study. Mach

number value from Munson, Young and Okiishi, table B-4, pp. 809 [20].

Appendix D

Appendix D

Equilibrium Calculations

Consider a general chemical reaction equation [19]:
yaMa +beb+...=y,M, +ymMm+... (33)
This can be re-written as:

2M.- =0 (34)

which leads to,

 

zln L " ——AGO -an
Po ‘ RT_ P

Where K p is the equilibrium constant at constant pressure, P is the operating pressure
and P0 is a reference pressure. The equilibrium constant for a given reaction is then

obtained from the equation:

72

73

[ylllym
[ya 1b?)

H

 

KP:

L—J

For fuel rich combustion, the equilibrium of C0 and C02 during expansion and

the subsequent exhaust processes is often assumed to follow the “water-gas” reaction [19]

[8]:
C0, + H, <—-K"—> C0 + H,0 (35)
with the equilibrium constant defined as,

_ [c0][H,0]
” - [00211112] '

For an AFR of 12.5,
(2,151,, + 1042(0, + 3.76N,) —> aco, +bH,0 + cCO + dH, + 39.18N,
A Carbon balance yields: a + c = 8
A Hydrogen balance yields: 2b + 2d = 1.8
An Oxygen balance yields: 2a + b + c = 20.83

In addition the equilibrium states (bc)/(ad) = 3.388 = K p. These four equations can be

solved to obtain,

74
c2 - 1936‘ + 47.3 = 0 (36)

which gives c = 2.89, a = 5.12, b = 7.72, and d = 1.29. The total number of moles are
found by summing the parts, or a + b + c + d + 39.3 = 56.3, so the mole fractions are:
C02 = 0.0908; H20 = 0.137; C0 = 0.051; H, = 0.023 and N2 = 0.698. This procedure
can be expanded to include moist air. Equation (14), restated below for convenience,
states that the actual AFR (AFR, )is a function of the design AFR (AFR, ) , temperature

and humidity.

 

 

(14)

 

Thus, the theoretical C 0/ C02 ratio can be found for different AFR, , temperatures and

humidities.

Appendix E

Appendix E

Measured Quantities

Below is an output file from the program SAMPLE. These are the measured and

calculated quantities.

Emissions ngrating Conditions Temmratures
Fuel flow (gr/hr) 524.7 Speed (rpm) 6797. T_nozzle(‘C) 21.4
HC(C_1) (ppm) 0.2220 Power (brake kW) 0.4819 T_cyl (‘C) 161.5
NOX (ppm) 0.0587 Torque (N-m) 0.6766 T_exhaust (‘C) 249.1
C02 (%) 0.4009 Rel. Humidity (%) 62.25 T_fuel (‘C) 20.0
C0 (ppm) 0.2604 P_diff (kPa) 0.122 T_room (‘C) 12.9
CH4 (ppm) 0.0000 P_abs (kPa) 98.5 T_oil (‘C) 69.6

CO(gr/kW-hr) NOx(gr/kW-hr) HC(gr/kW-hr) C02(gr/kW-hr)
CO(gr/hp-hr) NOx(gr/hp—hr) HC(gr/hp-hr) CO2(gr/hp-hr)
46.7 2.4 14.85 2242.84
48.3 1.8 11.08 1673.16

CO(gr/hr) NOx(gr/hr) HC(gr/hr) C02(gr/hr)

31.2 1.1 7.16 1081.17
BSFC (gr/kW/hr): 1088.53 Fuel burned (%): 98.64
Dilution Factor 36.20 AFR: 14.00

Dilution Factor (EPA)35.1 1

[CO]/[C02]: 0.0288

75

LIST OF REFERENCES

1)

2)

3)

4)

5)

6)

7)

LIST OF REFERENCES

Blumberg, P and Kummer, J .T., “Prediction of NO Formation in Spark Ignited
Engines - An Analysis of Methods of Control”, Combustion Science and Technology,

Volume 4, pp. 73-95, 1971.

Robinson, J .A., “Humidity Effects on Engine Nitric Oxide Emissions at Steady-State

Conditions”, SAE Paper 700467, 1970.

Sodre, J .R. and Yates, D.A., “Species and Time-Resolved Measurements of Exhaust

Hydrocarbons from a SI Engine”, SAE Paper 971016, 1997.

Donohue, J.A., Hardwick, G.C., Newhall, H.K., Sanvordenker, KS. and Woelffer,
N.C., “Small Engine Exhaust Emissions in the United States”, SAE Paper 720198,
1972.

Mooney, J .J., Shinn Hwang, H., Daby, K.O. and Winberg, J.R., “Exhaust Emission
Control of Small 4-Stroke Air Cooled Utility Engines an Initial R & D Report”, SAE

Paper 941807, 1994.

Eccleston, B.H. and Hum, R.W., “Exhaust Emissions from Small, Utility, Internal

Combustion Engines”, SAE Paper 720197, 1972.

Brereton, G.J., DeAraujo, A. and Bertrand, E., “Effects of Ambient Conditions on the

Emissions of a Small Carbureted Four-Stroke Engine”, SAE Paper 961739, 1996.

76

77

8) Spindt, R.S., “Air-Fuel Ratios from Exhaust Gas Analysis”, SAE Paper 650567,
1965.

9) Hare, C.T., Springer, K.J., Oliver, W.R. and Houtman, W.H., “Small Engine
Emissions and Their Impact”, SAE Paper 730859, 1973.

10) White, 1.1., Carroll, J.N., Hare, C.T., and Lourenco, J.G., “Emission Control
Strategies for Small Utility Engines”, SAE Paper 911807, 1991.

11) Sun, X., Brereton, G., Morrison, K. and Patterson, D., “Emissions Analysis of Small

Utility Engines”, SAE 952080, 1995.

12) Sun, X., Assanis, D. and Brereton, 6., ”Assessment of Alternative Strategies for
Reducing Hydrocarbon and Carbon Monoxide Emissions from Small Two-Stroke

Engines”

13) Brereton, G.J., Morrison, K., Chishty, H.A., Schwartz, G. and Patterson, D.J.,
“Measured Emissions of Small Engines under Steady State and Transient Operation”,

SAE Paper 941806, 1994.

14) Reynolds, R.C., “STANJAN”, a computer program used for analyzing combustion

problems, Dept. of Mechanical Engineering, Stanford University

15) Woshni, G., “A Universally Applicable Equation for the Instantaneous Heat Transfer
Coefficient in the Internal Combustion Engine”, SAE Paper 670931, 1967.

78

16) Ramos, 1.1., “Internal Combustion Engine Modeling”, Hemisphere Publishing

Corporation, 1989.

17) Taylor, CE, “The Internal Combustion Engine in Theory and Practice”, Volumes 1

and 2, M.I.T. Press, Revised Edition 1985.

18) Zeggeren, V., and Storey, S.H., “The Computation of Chemical Equilibria”,
Cambridge University Press, 1970.

19) Heywood, J.B., “Internal Combustion Engine Fundamentals”, McGraw Hill Series in
Mechanical Engineering, 1988.

20) Munson, B.R., Young, DR and Okiishi, T.H., “Fundamentals of Fluid Mechanics”,
John Wiley and Sons Publishing, 1990.

21) California Air Resources Board (CARB) homepage:

http://arbis.arb.ca.gov/homepage.htm.

22) “Test Procedure for Measurement of Gaseous Exhaust Emissions from Small Utility

Engines”, SAE J 1088 Recommended Practice, 1983.

23) Sonntag, R.E., VanWylen, G.J., “Introduction to Thermodynamics”, John Wiley and
Sons, Inc., 1991.

24) Micro Motion Model RFT9739 Instruction Manual, 1992.

"7111111111111111111115