FINAL SUMMARY REPORT
on
CHEMICAL AND PHYSICAL CHARACTERIZATION OF
AUTOMOTIVE EXHAUST PARTICULATE MATTER
IN THE ATMOSPHERE
to
COORDINATING RESEARCH COUNCIL
(CRC-APRAC PROJECT NO. CAPE-19-70)
and
ENVIRONMENTAL PROTECTION AGENCY
(CONTRACT NO. 68-02-0205)
June 14, 1973
Period Covered: June 25, 1971, to June 30, 1972
by
C. W. Melton, R. I. Mitchell, D. A. Trayser
J. F. Foster, Project Director
BATTELLE
Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Bactelle is not engaged in research for advertising, sales promotion,
or publicity purposes, and this report may not be reproduced in full or in part
for such purposes.
-------
llBatteile
June 14, 1973
Mr. A. E. Zengel
Project Manager
Coordinating Research Council
30 Rockefeller Plaza
New York, New York 10020
Dear Mr. Zengel:
Enclosed are 300 copies of the Summary Report on "Chemical and Physical
Characterization of Automotive Particulate Matter in the Atmosphere" for
the contract year July 1, 1971, to June 30, 1972. The report has been
approved in this revision by the APRAC/CAPE-19-70 Project Group.
Yours very truly,
-------
TABLE OF CONTENTS
Page
SUMMARY o 1
INTRODUCTION 4
OBJECTIVE 5
EXPERIMENTAL APPARATUS AND PROCEDURES 5
Project Vehicles 5
Conditioning Procedures 6
Summary of Conditioning Operations <, 6
Details of Conditioning Operation of the Project Cars ... 8
Test Cycle 8
Dilution Ratios 9
Exhaust Gas Monitoring » 10
Particulate Sampling From the Dilution Tunnel 10
Fuels 11
Comparison of Vehicles for Similarity 15
Standardized Operating Procedures . . . . 15
Particulate Sampling and Comparison Procedures 15
Procedures After the Transition to Different Fuels . . o . . . . 16
Initial Tune-Up and Servicing 16
New Baseline Data 17
Collection of Particulates 19
Completion of Conditioning With Unleaded and Leaded Fuels. ... 19
Dilation Tunnel 19
Construction and Assembly 20
Velocity and Gas Mixing Profiles 22
Aerosol Mixing Profiles 22
Modification of the Dilution Tunnel. . 30
Positive Pressure Operation o . . 30
Change in Dilution Ratio 30
Aerosol Mixing Profile. . . . . . 30
Residence Chamber <> 31
Apparatus for Aerosol Monitoring 37
Filters and Impactors . . . o o 37
Single Particle Counter 38
-------
TABLE OF CONTENTS (Continued')
Page
EXPERIMENTAL RESULTS AND DISCUSSION 42
Comparative Measurements of Vehicles 42
Gaseous Emissions From Steady-State Operation 42
Gaseous Emissions From LA-4 Cycles 46
Inorganic Composition of Exhaust Particulates 49
Morphology of Exhaust Particles 49
Organic Analysis of Exhaust Emissions by HPLC and GC. . 49
Exhaust Particle Sizes 55
Total Particulate Mass Loadings 55
Factors Affecting Light Scattering in the Tunnel . 61
Preliminary Residence Chamber Measurements . . . 69
Test Conditions . . . . 69
Aerosol Concentrations 71
Particle Count . . 71
Particle Mass , . . 73
Particle Morphology ..<.. 75
MAJOR ACCOMPLISHMENTS . . 77
Standardization of Test Autos 77
Construction and Operation of Test Facility 78
Conditioning of Cars 79
Preliminary Particulate Measurements ,, . . « 79
FUTURE WORK 82
LIST OF TABLES
Table 1. Characteristics of Fuels 12
Table 2. Determination of Aerosol Concentration Profiles in
the Dilution Tunnel at 560 FPM 26
Table 3. Distribution of Dye Aerosol at Replicate
Sampling Sites 32
Table 4. Exhaust Gas Composition from the Exhaust Pipe at
50-MPH Steady Cruise During Conditioning of
White Car (Unleaded Fuel) 44
-------
LIST OF TABLES (Continued)
Table 5. Exhaust Gas Composition From the Exhaust Pipe at
50-MPH Steady Cruise During Break-in of Blue Car
(Unleaded Fuel) 45
Table 6. Compositions of Integrated Exhaust Samples From Single
Modified LA-4 Cycles 48
Table 7. Comparison of Inorganic Compositions of Particles
Collected From the Dilution Tunnel 50
Table 8. Summary of Major Peaks for Five HPLC Chromatograms
of Extracts From Particulates 54
Table 9. Combined Weights of Auto Exhaust Particulates From
Unleaded Fuels From Four Modified Cold-Start LA-4
Cycles on "Metricel-DM" Filters and on the Cascade
Impactor and Backup Filter ..... 58
Table 10. Weight of Auto Exhaust Particulates Collected on
Metricel-DM Filters of Different Sizes and on
Cascade Impactor 60
LIST OF FIGURES
Figure 1. Conditioning of Project Cars in Preparation for
Testing 7
Figure 2. Air-Fuel Ratio Versus Air Flow 18
Figure 3. Original Layout of Dilution Tunnel, With Induced
Draft Blower 21
Figure 4. Velocity and Mixing Profiles at Sampling Station 6. . . 23
Figure 5. Sampling Probe Assembly Used for Aerosol Mixing
Studies in the Dilution Tunnel 25
Figure 6. Dilution Tunnel Mixing Profiles for Aerosols 0.87
Micron in Diameter 27
Figure 7. Dilution Tunnel Mixing Profiles for Aerosols 2.0
Microns in Diameter 28
-------
LIST OF FIGURES (Continued')
Page
Figure 8. Dilution Tunnel Mixing Profiles for
Aerosols 5.9 Microns in Diameter 29
Figure 9. Number Code for Sampling Port Locations,
Looking Toward Automobile 33
Figure lOa. Modified Test Apparatus with Positive Pressure
Sampling ........................ 34
Figure lOb. Modified Layout of Test Apparatus with
Positive Pressure Sampling . . 35
Figure lOc. Layout of Residence Chamber and Purge-Circulation
System 36
Figure 11. Photograph of Single-Particle Counter 40
Figure 12 Calibration of Single Particle Counter
with Signal Generator 41
Figure 13. Calibration of Single Particle Counter
with Polystyrene Latices 43
Figure 14. HC Concentrations Sampled From the Exhaust at
50 MPH Steady Cruise During Break-in of Project Cars . . 47
Figure 15. Typical Carbon-Black-Type Exhaust Particles From
Unleaded Fuel Collected From the Tunnel on 0.25 ^m
Stage of Impactoro . 51
Figure 16. Typical Tar-Droplet-Type Exhaust Particles From
Unleaded Fuel Collected From the Tunnel on 0.5 ^m
Stage of Impactor 51
Figure 17. Particle Size Distribution in Auto Exhaust Generated
by the White Car with Unleaded Gasoline RE-141B 56
Figure 18. Particle Size Distribution in Auto Exhaust Generated
by the Blue Car with Unleaded Gasoline RE-141B 57
Figure 19. Particle Size Distribution in Automobile Exhaust
Generated by the White Car x^ith Unleaded Gasoline
at 50 MPH Steady State « 62
Figure 20. Particle Size Distribution in Automobile Exhaust
Generated by the Blue Car with Leaded Gasoline at
50 MPH Steady State 63
-------
LIST OF FIGURES (Continued)
Figure 21. Typical Light-Scattering Patterns for Different
Initial Temperature Conditions, Blue Car 65
Figure 22. Correlation Between Light-Scattering and Initial
Choke-Box Temperature, Blue Car 67
Figure 23. Light-Scattering Patterns for Consecutive Partial
Modified LA-4 Cycles Using Controlled-Choke
Schedule 1 - Blue Car 68
Figure 24. Light-Scattering Patterns for Consecutive Partial
Modified LA-4 Cycles Using Controlled-Choke
Schedule 2 - Blue Car 70
Figure 25, Particles in Residence Chamber, Run 4 72
Figure 260 Particle Size Distribution of Diluted Automobile
Exhaust in Chamber (Run 4) 74
Figure 27. Particles Collected on the 1/4 ^m Impactor Stage
After Six Hours in the Residence Chamber 76
Figure 28. A Particle Typical of Those Collected on the 1.0- m
Impactor Stage After Six Hours Residence Time. ..... 76
-------
CHEMICAL AND PHYSICAL CHARACTERIZATION OF
AUTOMOTIVE EXHAUST PARTICULATE MATTER
IN THE ATMOSPHERE
By
C. W. Melton, R. I. Mitchell, D. A. Trayser
J. F. Foster, Project Director
SUMMARY
Two matched autos have been made operative to generate exhaust
particulate matter. These 1970 Fords are equipped with measured and matched
1971 351 CID engines, automatic transmissions, and 2-barrel carburetors.
These cars were chosen to represent a typical and large volume sales car
model. To achieve reproducible operation of the cars for generating the
exhaust particles, a chassis dynamometer with an automatic driver programmed
for the 1972 FTP driving schedule is used,, Forced-air cooling of the
exhaust system is used during the tests to maintain operating temperatures
similar to measured values during operation on the highway.
A typical unleaded fuel was purchased in sufficient quantity to
supply the projected needs of the project. A portion of this fuel was then
leaded with TEL motor mix to 2.49 g/gallon lead. Thus both fuels have
identical compositions except for the added TEL motor mix.
Initially, each car was driven for 4000 miles with unleaded fuel
on a (modified) Durability Driving Schedule. Periodic tests during this
stabilization period established their matched condition. Stabilization
was completed after operation for another 4000 miles on the Durability Driving
Schedule with unleaded fuel in one car and leaded fuel in the other. An
exhaust gas dilution tunnel has been built and calibrated. The diluted
exhaust passes through a measuring and mixing orifice and along the length
of the 36-foot stainless steel tunnel which has a 23-inch diameter. Typically,
-------
the exhaust is diluted an average of 20:1 or 30:1 by appropriate control of
the tunnel air flow.
Exhaust gas composition is monitored continuously for CO, HC,
CO , and NO in the tunnel and a CVS type bag sample is also taken in order
to appraise reproducibility of vehicle operation.
A 2100 cu. ft. residence chamber has been constructed for
collection of a proportional sample of the tunnel flow. Final dilution is
about 300:1 in order to minimize wall effects and be more representative of
atmospheric dilution. Walls of the chamber are flexible opaque film
mounted on a collapsible frame to minimize photochemical interactions, and
to maintain negligible pressure difference between sample and surroundings
while the chamber is in use to collect, store, and then withdraw samples.
Temperature, humiditys gaseous composition, and particle content in the
chamber are monitored over a range of residence times. Particles are
collected simultaneously for detailed analysis.
Measurements of the mass of particulate emissions collected on
Metricel-DM membrane filters showed variations in the collected amounts
with the face velocity of the exhaust as well as with the operating cycle
and fuel used.
Particle-size distributions were measured in the Battelle impactor,
which collects fractions on six impact stages with cut-offs at sizes ranging
from 8 to 0.25 micrometers mass mean equivalent diameter. An absolute filter
is used to back up the last stage of the impactor.
The properties of the aerosol particles in diluted exhaust were
examined in exploratory measurements while the aerosol was aged in the
residence chamber. The overall results in these preliminary studies suggest
-------
that the smallest particles below the light-scattering range agglomerate
or grow during the first two hours to a size sufficient to scatter light.
Thereafter light-scattering measurements indicate little change in the
aerosol. Other measurements, by a single-particle counter, also indicate
growth and agglomeration. Some particle growth appears to continue as long
as 24 hours, accompanied by precipitation of the largest particles.
Samples were withdrawn from the chamber after six hours residence
and passed through 142-mm and 47-mm filters, and an impactor with back-up
76-mm filter. The weight gains recorded after filtration of identical
volumes of gas in concurrent samples (60 minutes at 1 cfm) varied widely
and correlated qualitatively with the diameters or areas of filters used.
No conclusions were possible concerning the absolute weight concentration
of the filterable aerosol particles.
The studies on characterization of auto exhaust particles are
being continued. The changes attributable to leaded vs. unleaded fuel,
residence time, relative humidity, and the concentrations of two common
atmospheric pollutants are to be studied in a series of tests designed to
show statistically significant differences in particle properties. The
variability of weights of particles collected on filters resulting from
variations in filter properties and in collection conditions represents a
serious handicap in the search for quantitative interpretations of the
characteristics of exhaust particulate matter in the atmosphere. Studies
will be continued with comparison of results on a relative scale, and efforts
to resolve the problem of measuring the absolute mass concentration of auto
exhaust particulates will be continued.
-------
INTRODUCTION
(1 2)
This report describes a third year ' of studies directed toward
determining the fate of auto exhaust particles in the atmosphere. Experi-
mental apparatus and procedures have been developed, and analytical data
have been accumulated on particle characteristics and morphology. In the
year covered by this report, preparation, instrumentation and deposit
conditioning of two automobiles has been completed in preparation for
systematic examination of particulate emissions during the coming year.
Environmental variables are to be studied for influences on the properties
and fate of aerosols emitted by the automobiles and diluted by mixing with
the ambient atmosphere.
This report is presented in two major sections describing first
the experimental apparatus and procedures, and then the experimental results
with discussion of their significance and interpretation. Some of the
information on apparatus, equipment and operations in the preceding annual
report is repeated in order that this summary report shall be comprehensible
without major dependence on preceding annual reports.
(1) C. W. Melton, et al., "Physical-Chemical Characteristics of Particles
Associated with Polynuclear Aromatic Hydrocarbons Present in Automobile
Exhaust", Final Summary Technical Report for the period January 24, 1969,
to March 31, 1970, to Coordinating Research Council (APRAC-CAPE-12-68-
Neg. 59), January 29, 1970.
(2) J. F. Foster, et al., "Chemical and Physical Characterization of Automotive
Exhaust Particulate Matter in the Atmosphere", Final Summary Report for
the period July 1, 1970, to June 24, 1971, to Coordinating Research Coun-
cil (CAPE-12-68-Neg. 59 and CAPE19-70) October 6, 1972.
-------
OBJECTIVE
The objective of the CAPE-19 project is to characterize exhaust
particulates generated under "real-life" conditions from both leaded and
unleaded gasolines, with the long-range goal being the definition of the
fate of automobile-generated particulates in the atmosphere. Along this
line, attempts will be made analytically to characterize organic-inorganic
associations which exist in automobile exhaust particles generated under
various conditions and after having interacted with prevalent extraneous
nuclei such as silica (SiCL).
EXPERIMENTAL APPARATUS AND PROCEDURES
Project Vehicles
Two 1970 Fairlane Fords with matched 1971 engines were prepared
for exhaust-gas generation; one was run with unleaded fuel and the other
with leaded fuel. Each vehicle was equipped with a 351 CID V-8 engine,
2-barrel carburetor, and automatic transmission. In order to make the two
automobiles as nearly comparable as possible, specially matched and measured
1971 engines were made available by the Ford Motor Company and were installed
in the two vehicles by project personnel at Battelle0
To identify each car in this report, one of them will be referred
Lo as the white car; only unleaded fuel was used in its operation. The
other will be referred to as the blue car; only leaded fuel was used in its
operation after completion of deposit conditioning runs and collection of
bace-line data on unleaded fuel.
-------
Both vehicles were checked after delivery for driveability and
mechanical condition, and were found to be satisfactory. The checkout
included both visual examination and city-street and highway driving.
Preliminary studies of exhaust particulate collection and development of
procedures were conducted with the white (unleaded car) to gain experience
with operating the vehicle in a test mode, before installation of the 1971
measured engine. Both vehicles were provided with instrumentation for
monitoring performance, as discussed in the preceding Final Summary Report
for 1970-1971. Vehicle history and further modification during the current
year are described on the following pages.
Conditioning Procedures
Summary of Conditioning Operations
Figure 1 summarizes the conditioning operations. Both the blue
and white cars were conditioned for deposits using a modified Durability
Driving Schedule for a total of 4000 miles, and using unleaded Fuel No.
RE141A. During the conditioning period, comparisons were made of automobile
operating parameters, and exhaust emissions in samples of diluted exhaust
taken from a dilution tunnel, to appraise the comparability of the two cars.
Later, new fuels, one leaded and the other unleaded, were obtained and the
conditioning was repeated. First, both cars were run 200 miles on the new
unleaded fuel, RE-141B} to determine how the vehicles operating with the new
fuel compared Lo one another. Then the blue leaded car was switched to the
new leaded fuel, RE-141C, and 4000 additional miles were accumulated while
running according to the Durability Driving Schedule. The white unleaded
-------
_to
CD
T3
CD
TJ
O
CD
en
e
=>
\-
o
LU
-3
O
cr
Q.
fee
tr
<
o_
LU
QC
Q.
C
a
CD
C
o
o «
* _
cu E
o
o
O
m
ro
CD
£
0)
E
o
Q.
Q
C
T
o
CD
i_
CD
o
0)
V)
00 H-
o
CO
O
O
O
CD
>
i_
Q
CD
C
CO
Q
w
H
03
fn
O
M
H
W
CO
CJ
H
O
W
O
Pn
O
Q
2
O
o
P
O
h
-------
8
car was run again with the new unleaded gasoline RE-141B to accumulate the
same mileage as the blue leaded car. At this point, the cars were considered
adequately conditioned and ready for generation of exhaust particulate for
residence chamber experiments.
Details of Conditioning Operation of
the Project Cars
The initial operation of each project car for 4000 miles used
part of the proposed Durability Driving Schedule, given in the July 15,
1970, Federal Register. The 70-mph lap was not used. The initial condition-
ing runs for deposit stabilization used unleaded fuel in both cars. The
4000 miles were logged in two months with the white car and in two weeks
with the blue car.
The ignition timing, points dwell, points gap, and idle speed were
checked on each car at the beginning and at the end of the conditioning
period. Also, the condition of the spark plugs and ignition points were
checked at the end. No significant wear or deterioration was observed.
After 4000 miles, the ignition timing of the white (unleaded) car was found
to be retarded about 5 degrees from manufacturer's specification. Timing
was reset and a test run made to determine if the change had any effect on
exhaust emissions. No change in HC or CO in the exhaust was observed.
Test Cycle
One test cycle was selected and used in all studies described here
to generate particles from each of the two cars. The selected cycle was
based on the 1972 Urban Driving Schedule with modified vehicle preconditioning
-------
and starting procedures. Instead of the overnight cold soak called for in
the Federal Test Procedure (FTP), a rapid cooldown procedure was used. The
car was operated through a full cycle to establish uniform operating tem-
peratures, then cooled rapidly by using a heat exchanger for the engine
coolant and blowers for the exhaust system, radiator, and choke box. Cool-
down was continued until all temperatures except the oil sump were 80 F or
lower. A thermocouple was installed in the choke box to permit monitoring
air temperature at the thermostatic spring during each cold-start modified
LA-4 cycle. Approximately 20 minutes were required for the rapid cooldown
to lower the temperatures to 80 F.
After rapid cooldown, the car was started and run through the
test cycle. The starting procedure was modified from the standard FTP by
adding 20 seconds to the initial idle period right after engine start to
minimize the chance of engine stumble or stall.
Dilution Ratios
Tests with the project cars are described throughout this report
in which the progress of conditioning or the characteristics of gaseous and
particulate emissions were measured. When a dilution tunnel was used, the
dilution ratios given as part of the data were derived from the following
measured parameters. These values are valid except when otherwise stated
for specific tests.
In general, the dynamometer load was 6 HP during conditioning
procedures. Thereafter, the test fuel was changed in January, 1972, from
RE-141A (unleaded) to either RE-141B (unleaded) or RE-141C (leaded), and the
dynamometer load was increased to 12 HP.
-------
Car Operation
Dynamometer Load, hp
(Set at 50 mph)
Exhaust Gas Flow,
scfm (average)
Tunnel Gas Flow,
scfm
Dilution Ratio
10
Modified Modified
IA-4 LA-4
Cvcle Cycle
12
25
30
50 MPH
Steady State
46
50 MPH
Steady State
12
54
600
24
600
20
1,500
33(32.6)
1,500
28(27.8)
Exhaust Gas Monitoring
At intervals during the conditioning of the project cars, the
hydrocarbon (HC) and carbon monoxide (CO) concentrations in the exhaust gases
were measured while the vehicle operated at 50-mph steady-state cruise. Exhaust-
gas data (shown in Tables 4 and 5, pages 44 and 45) are for a fully warmed up
engine. HC was measured with a Beckman 109A Nondispersive Infrared Analyzer,
and CO was measured with an Olson-Horiba MEXA-200 Nondispersive Infrared
Analyzer. Exhaust-gas sampling was done with a 1/8-inch stainless steel
probe located in an exhaust-gas diverter valve« Before entering the analyzers,
the sample gas passed through an ice-bath cold trap and a glass fiber filter
to remove water vapor and particulates,,
Particulate Sampling From the Dilution Tunnel
Particulate sampling runs with the dilution tunnel were made at the
beginning of the conditioning schedule for each car and at approximately
1000-mile intervals thereafter. These runs were made with the car operating
at 50-mph steady cruise, and with a dilution tunnel air flow of 1500 scfm
-------
11
(for a dilution ratio of about 33 to 1). All runs were made with a fully
warmed up engine and exhaust system, and all sampling was done at the
35-foot sampling station (Station 6, Figure 3, page 21) in the tunnel.
Total particulate samples were collected for one hour on 47-mm
diameter silver membrane and cellulose ester membrane filters (0.45 M-m and
0.80 \m nominal pore diameter, respectively), using a sampling flow rate of
1 scfm. Particle size fractions were collected for 1/2 hour using the
Battelle cascade impactor at a sampling flow rate of 1/2 scfm.
Fuels
Two different fuels, RE-141A and RE-143A, were purchased from
Mobil Research and Development Corporation in July, 1971. These fuels were
unleaded and leaded gasolines respectively and their organic compositions
had been adjusted in order to create two fuels having very similar octane
numbers. Initial conditioning of the two vehicles was accomplished with
unleaded fuel RE-141A. The CAPE-19 Project Group later decided that compari-
sons should be made between unleaded and leaded fuels of the same organic
composition, and that the octane number need not be the same. Consequently,
another pair of fuels, RE-141B and 141C, were purchased from Mobil Research
and Development Corporation in January, 1972. These gasolines were reported
to have the same organic constituents, and differed only by the addition of
TEL Motor Mix (2.49 g Pb/gal) in fuel RE-141C. The composition of RE-141A,
-B, and -C are given in Table 1. Leaded fuel RE-143A was never used.
-------
12
TABLE 1. CHARACTERISTICS OF FUELS
(Data supplied by Mobil Research and
Development Corporation on Inspection
Tests and Elemental Analyses of Gasolines
Shipped to Battelle Memorial Institute in
July, 1971, and January, 1972)
Blend Designation
Name
Research Octane No. (RON)
Motor Octane No. (MON)
Vapor Pressure, Reid, Micro (D-2551)
TEL as Lead, ppm (M-1059)
TEL as grains Pb/gal. (M-951)
Sulfur, °L wt.
Chlorine, ppm (M-600)
Phosphorus, ppm (M-798)
Nitrogen, ppm (M-1042, Col
API Gravity (D-287)
ASTM Distillation (D-86)
Initial Boiling Pt., °F
57o Distilled
1070 "
207» "
307* "
407, "
507, "
607o "
707o "
807, "
907, "
957o "
End Point
.)
-16*
63*
77*
142*
188*
207*
231*
241*
275*
292*
335*
369*
7*
RE-141B
Battelle '72
Nonleaded
93.6
85.4
9.0
0.7
(0.002)
0.036
1.1
<1.
21
60.8
93
118
132
156
182
207
228
244
260
286
331
366
400
RE-141C
Battelle '72
Leaded
100.0
91.7
10.6
-
2.49
-
Present
1.
21
60.3
92
116
130
154
178
202
223
240
256
284
330
374
403
Nonleaded Break- In Fuel
RE-141A
Battelle '71
CRC White
94.0
65.3
9.6
1.2
(0.003)
0.020
1.4
1.
Not Sought
Not Sought
-
125
-
.
«
218
w
.
-
324
-
"
* Simulated distillation by gas chromatography of RE-141B (SIMDIST ASTM Method D-2887).
-------
13
TABLE 1. (Continued)
(Gas Chromatographic Analysis C5 and Lighter;
Mass Spec HC Type Analysis Cg and Heavier;
Chromatographic-UV Analysis BaP and BaA.
Gasolines Shipped to Battelle Memorial Institute
in July, 1971, and January, 1972)
(All results in liquid volume percent based
on total sample)
Blend Designation
RE-141B
Battelle '72
RE-141C
Battelle '72
Nonleaded Break-In Fuel
RE-141A
Battelle '71
Name
Components ;
Propylene
Propane
Isobutane
Isobutylene + Butene-1
n- Butane
trans- Butene-2
cis-Butene-2
3-Methylbutene-l
Isopentane
Pentene-1
2-Methylbutene-l
n- Pentane
trans- Pentene-2
cis-Pentene-2
Total C5 & Lighter
Nonleaded
0.01
0.08
0.70
0.07
5.00
0.07
0.09
0.06
10.88
0.16
0.34
2.56
0.41
0.22
(20.7)
Leaded
0.01
0.07
0.69
0.06
5.02
0.07
0.09
0.06
10.99
0.16
0.34
2.59
0.42
0.23
(20.8)
CRC White
0.01
0.07
1.04
0.18
4.31
0.35
0.35
0.08
9.13
0.32
0.52
2.71
0.62
0.39
(20.1)
C<5 & Heavier, Mass Spec PONA
Paraffins
Monoolefins
Cycloolefins & Diolefins
Monocycloparaffins
Dicycloparaf fins
Alkylbenzenes
Alkylindanes & -tetralins
Alkylnaphthalenes
Total C6 & Heavier
Approximate Distribution of
C6
C7
C8
C9
CIG
Cll
C12
Total Alkylbenzenes
43.8
3.9
0.6
1.6
0.2
27.5
0.8
0.8
(79.3)
Alkvlbenzenes by Mass Spec
1.8
6.9
9.0
6.8
2.2
0.8
0.1
(27.6)
43.6
4.0
0.6
1.8
0.2
27.2
0.9
0.9
(79.2)
1.7
6.5
9.1
6.8
2.2
0.8
0.1
(27.2)
42.6
4.5
0.8
2.6
0.2
27.8
0.8
0.6
(79.9)
1.9
6.6
9.6
7.0
2.1
0.6
0.
(27.8)
-------
14
TABLE 1. (Continued)
Blend Designation -
Name
RE-141B
Battelle '72
Nonleaded
RE-141C
Battelle '72
Leaded
Nonleaded Break-In Fuel
RE-141A
Battelle '71
CRC White
Total Sample GC + MS PONA Summary
Paraffins
Olefins
Naphthenes
Aromatics
Grand Total CG & MS
63.1
5.9
1.8
29.2
100.0
63.0
6.0
2.0
29.0
100.0
59.9
8.1
2.8
29.2
100.0
Distil la tion-Chromatographic-UV Analysis
Benz (a ) anthracene
Benso(a)pyrene
3.3 ppm
1.0 ppm
Not
Sought
Not
Sought
Footnotes;
RE-141B and RE-141C were prepared from a single 6000 gallon lot which was composed
of five blending components. These were similar to, but not exactly the same, as
those in RE-141A. Those components which tend to be unstable in long term storage
had antioxidant and metal deactivator additives (grades approved for use in military
gasoline) added as they were produced at the refinery.
Half of the 6000 gallon lot was put into clean drums and labeled RE-141B. The other
half had Tetraethyllead Motor Mix added and, after mixing, -was drummed, then labeled
RE-141C. TEL Motor Mix contains ethylene dichloride (1.0 theory) and ethylene dibromide
(0.5 theory) according to the usual specifications. No other additives were added during
blending of either RE-141B or -C. BaA and BaP in Re-141C should be the same as that
determined by analysis of RE-141B. These polynuclear aromatics were not determined in
RE-141A.
-------
15
Comparison of Vehicles for Similarity
Standardized Operating Procedures
The project cars after break-in (see Figure 1) were run under two
conditions to generate exhaust for comparison by analytical methods. One
condition was the modified LA-4 cycle and the other was a 50-mph steady
cruise mode.
A cold-start for the modified IA-4 cycle was achieved by use of the
rapid-cooldown technique developed for the break-in operation and already
described. This procedure was developed to insure that engine and exhaust
system temperatures at the beginning of each modified LA-4 cycle run were
consistent.
Particulate Sampling and Comparison Procedures
Particulate sampling runs were also made after break-in was
completed, using the dilution tunnel and operating the car through the
modified LA-4 cycle. A dilution tunnel air flow of 600 scfm was used to
give an approximate 24 to 1 dilution ratio, and the driving cycle was driven
from a cold (ambient temperature) start. Filters for total samples, the
cascade impactor for size classification of samples, and sampling flow rates
through filters and the cascade impactor were the same as for the 50-mph
steady cruise dilution tunnel runs (see pages 10 and 11). One cold-start
driving cycle provided sufficient samples for characterization.
Light scattering measurements were made on the diluted exhaust-
gas with an integrating nephelometer during selected tunnel runs. No
measurable change from filtered-air alone was observed during the 50-mph
steady cruise runs. Only slight momentary increases in light scattering were
observed at the beginning of the cold-start driving cycle operation.
-------
16
Mainly gross analytical data were employed in organic analytical
comparisons. High-pressure liquid chromatograms and gas chromatograms were
used as empirical "fingerprints" of the exhaust composition, with no attempt
to identify and compare specific organic compounds.
In summary, the basis for comparison of the two project cars for
similarity were measurements of HC and CO in the raw exhaust, measurement
of light scattering in the tunnel, particle morphology, inorganic and
organic composition of collected particulate samples, and measurements of
particulate mass emissions. The cars were judged to be acceptably similar.
Procedures After the Transition to Different Fuels
Initial Tune-Up and Servicing
When another set of two comparable fuels, RE-141B and -141C, was
specified for use in the experimental studies, both project cars were
serviced and then each was operated with the new unleaded fuel (RE-141B) for
approximately 200 miles on the modified Durability Driving Schedule used
in the 4000-mile conditioning operation (refer to page 8). The servicing
consisted of changing the oil and oil filter, inspecting the spark plugs
and points, and tuning to manufacturer's specifications.
In the servicing, it was found that two of the spark plugs and
one spark plug cable of the blue car were deteriorated to the point of
causing erratic operation. It is possible that this condition existed to
some degree during previous particulate sampling runs, and resulted in
abnormal black soot deposits noted occasionally on collection filters. No
other problems were noted with either car.
-------
17
The 200-mile durability operation was conducted to condition the
engines and exhaust systems to the new fuel before making additional runs.
New Baseline Data
Before switching the blue car to leaded fuel, and then accumulating
additional mileage on both cars, a number of different tests were made on
each car to establish and compare baseline operating characteristics with
the RE-141B fuel. These were (1) continuous HC and CO measurements during
modified LA-4 cycles, (2) fuel consumption, (3) HC, CO and NO concentrations
from modified LA-4 cycle composite bag samples, and (4) air-fuel ratios at
various operating conditions.
The HC and CO concentrations obtained continuously during modified
LA-4 cycles from the two cars were acceptably similar although the values
for the blue car were somewhat higher than for the white car. However,
variations among runs with the same car tended to be nearly as great as the
differences between the two cars.
Fuel consumption was compared on the blue car between two modified
LA-4 cycles, one with a hot start and the other with a cold start. The hot-
start fuel consumption was 3.65 and the cold-start fuel consumption was
4.25 Ib.
Air and fuel consumption were measured on both cars at warmed up
steady-state conditions over the speed range using a Meriam Laminar Flow
Element to determine air flow and a Kent-Moore volume type Gas-per-Mile
gauge to measure gas flow. From these measurements, the relative air-fuel
ratios were estimated. Figure 2 shows a plot of computed air-fuel ratio as
a function of air flow, in which the cars appear to be similar with regard
to carburetion.
-------
20
18
18
o
DC
u_
16
14
12,
o White car, unleaded fuel
Blue car, leaded fuel
Dynamometer setting - 6 hp at 50 mph
i i i
10
20 30
Air Flow, scfm
40
50
60
FIGURE 2. AIR-FUEL RATIO " ERSUS AIR FLOW
-------
19
Collection of Particulates
Particulates were collected from the new RE-141B fuel for charac-
terization according to morphology, total mass, size distribution, and
organic analysis, while the cars were operated on modified cold-start LA-4
cycles. A consistent operating procedure was followed for each cycle. The
results of these runs are described on pages 47-58.
Completion of Conditioning With
Unleaded and Leaded Fuels
The white (unleaded) car and the blue (leaded) car completed an
additional 4000 miles conditioning on the new fuels using the modified
Durability Driving Schedule. HC and CO concentrations at 50-mph steady
cruise were measured periodically. At approximately 1000-mile intervals,
particulate samples were collected to determine size distribution and total
mass. For the particulate sampling runs, the cars were run at 50-mph steady
cruise for 4 hours. The tunnel air flow was set at 400 cfm to increase the
particulate concentration by decreasing the dilution ratio to about 7.4:1.
Dilution Tunnel
Exhaust gases and particulates issuing from an automobile tailpipe
during over-the-road operation are rapidly quenched and diluted by the
ambient atmosphere in the highly turbulent airstream near the rear of the
vehicle. To simulate the real environment under reproducible experimental
conditions and to permit accurate sampling of automobile exhaust, the
dilution tunnel was constructed to use the technique developed by Habibi at
the Du Pont Petroleum Laboratory .
(1) Habibi, Kamran, "Characterization of Particulate Lead in Vehicle Exhaust-
Experimental Techniques", Environmental Science and Technology, Vol. 4,
NOo 3, March 1970, pp 239-253.
-------
20
The design, construction, and initial negative pressure operation
of the auto exhaust dilution tunnel were described in detail in the 1971
final report to CRC, but are summarized here to obviate reference to that
report. During this report period, the tunnel was modified for positive
pressure operation as also described in this section.
Construction and Assembly
The exhaust-gas dilution tunnel is 36 feet long by 23 inches
diameter, similar to the Du Pont tunnel. Each 6-foot section was formed
from a 6* x 61 sheet of 16-gage stainless steel by rolling it into a cylinder
and welding the longitudinal seam. Mild steel flanges 1-1/4 inches wide
were welded to each end of the tunnel sections for bolting them together.
Figure 3 is a schematic plan view of the dilution tunnel as origi-
nally assembled with blower at the downstream end, which gave negative
pressures at the sampling stations. This view gives sampling station locations
and the tunnel-section numbering system, which were unchanged when the tunnel
was later modified for positive pressures. At each sampling station, a
1/4-inch pipe coupling was attached to the tunnel surface at the top and
another on the side 90 degrees from the top. Probes were inserted through
these fittings for velocity and gas concentration measurements during check-
out of the tunnel. These fittings are also available for gas sampling during
vehicle operation. To accommodate particle-sampling probes, 6-inch-diameter
bosses with 4-inch openings into the tunnel were attached to the tunnel
surfaces at sampling stations 2, 4, and 6. The bosses are located approxi-
mately 30 degrees away from the bottom of the tunnel. Cover plates close
the openings when not in use.
-------
21
W
0
= 0)
.
iAs 2
/ - ' V - -
o o
2 5
o o
Q
W
O
P
O
25
H
M
w
0
M
0
H
!3
o
a
^
o
CO
g
e
g
> O
I "°
u- g
o
u
-------
22
Velocity and Gas Mixing Profiles
Velocity profiles in the dilution tunnel were measured at the
major sampling stations at two tunnel flow rates and two exhaust-gas flow
rates, using shop air to simulate the exhaust gas flow. The velocity
measurements were made with a Thermal Systems, Inc., anemometer using a
hot-film sensor projecting into the air stream. Mixing profiles were
determined at 580 feet per minute using CO as a tracer gas added to the
simulated exhaust gases and a Beckman NDIR CO Analyzer to measure the CO
concentrations in the tunnel.
Both velocity and mixing profiles were measured by traverses of
probes in vertical and horizontal planes. In each traverse, 14 positions
were measured for the velocity profiles and 15 positions for the mixing
profiles. Tests were run separately for horizontal and vertical velocity
traverses, without attempting to adjust flow to exactly the same value in
each test.
Figure 4 shows the velocity and mixing profiles measured at
Station 6. These curves show that mixing of exhaust gases and dilution air
was good and that velocity profiles were satisfactorily flat in the cross
section up to eight inches from the center line of the tube.
The velocity and mixing profiles were not affected by changes of
"exhaust-gas" flow rate.
Aerosol Mixing Profiles
Mixing analyses were repeated with an aerosol at the high flow
velocity used in the gas mixing studies above, in order to determine the
variation in concentration of the aerosol particles across the vertical and
-------
23
800
700
600
500
a
c
400
u
_o
a>
300
-X 1( X
Velocity (high flow)
Left-hand scale)
i-V v Y V
V A. A ^A. A
V "r1 ** nr *-"
y V v
200
100
CO concentration at 580 fpm
(Right-hand scale) *>
Velocity (low flow)
(Left-hand scale)
Vertical traverse
Horizontal traverse
200
160
E
ex
a
120 §
2
£
80
u
o
O
o
40
8 12 16
Sampling Point, inches
20
22Z
8
FIGURE A. VELOCITY AND MIXING PROFILES AT SAMPLING STATION 6.
-------
24
horizontal tunnel profiles. The sampling probe assembly (Figure 5) was a
ring bolted between any two tunnel sections to hold 13 fixed sampling
tubes, 3/8 inch in diameter, at the center and on vertical and horizontal
axes at 5-, 8-, and 10-1/2-inch radii. The exit ends of the tubes were
connected to absolute filter holders. Each filter holder contained a
critical flow orifice which was designed for 0.5 + 0.01 cfm, and was connected
to a vacuum line.
During these profile studies, the sampling probe assembly was
placed between the tunnel sections and the air velocity in the duct was
adjusted to 580 ft/min, a velocity intermediate between the two values for
high and low flow. The test aerosol was introduced in place of the auto-
mobile exhaust. A conventional aerosol can was used to discharge fluorescent
dye dissolved in Freon-12 and toluene. The particle size of the aerosol was
determined with the Battelle cascade impactor for each run and was varied
from run to run between 0.87 to 5.9 |om mass-mean diameter by varying the
dye concentration. The amounts of dye collected on each impaction stage
and on the absolute filter were determined with a fluorophotometer.
Table 2 is a summary of the measurements at 5-1/2, 17-1/2, and
29-1/2 feet from the point of aerosol generation. The data show that the
greatest deviation was obtained in the larger particles at the first sampling
point. The coefficient of variation of the mean (the normalized standard
error) was less than 4 percent for any test and averaged less than 2 percent.
Figures 6, 7, and 8 show plots of the mass variation of dye
concentration for the 0.87, 2.0, and 5.9 micro" dye aerosols for three
sampling positions. The variation for the 5.9 micron particles at the first
sampling station was greater than all others.
-------
25
FIGURE 5. SAMPLING PROBE ASSEMBLY USED FOR AEROSOL MIXING
STUDIES IN THE DILUTION TUNNEL
-------
26
TABLE 2. DETERMINATION OF AEROSOL CONCENTRATION PROFILES
IN THE DILUTION TUNNEL AT 560 FPM
Size Concentration Distribution
Travel Distance
feet
5 1/2
5 1/2
5 1/2
17 1/2
17 1/2
17 1/2
17 1/2
17 1/2
17 1/2
29 1/2
29 1/2
29 1/2
29 1/2
A MMD(a),
Mm
0.87
2.0
5.9
0.85
1.3
2.0
2.8
3.6
5.9
0.87
2.0
5.9
5.9
AMC(b),
X
109.8
434.5
206.5
239.2
237.8
446.1
617 . 5
333.3
325.3
266.8
358.2
438.5
463.1
SD(C),
a
3.77
30.4
26.5
12.5
14.5
19.8
35.8
22.1
15o5
18.3
14.7
34.2
26.6
SEM(d) ,
a/v/rf
1.04
8.44
7.35
3.61
4.17
5.5
9.9
6.6
4.31
5.07
4.08
9.5
7.4
Coefficient of
Variation of
the Mean,
percent
0.05
1.95
3.56
1.51
1.76
1.23
1.61
1.98
1.33
1.90
1.14
2.17
1.59
(a) Mass Mean Diameter.
(b) Arbitrary fluorescent units (proportional to microgram/m ).
(c) Standard Deviation.
(d) Standard Error Mean.
-------
27
Horizontal Axis
Vertical Axis
29:
15
10
5
0
5
10
15
fe>
£
e
o
o>
fe 10
o.
c
o
s i?4-
0>
c
I
u
o
I 5
7= 10
a
I '5
15
10
10
15
10 15 20 25 05 10 15 20 25
Sampling Point Along Duct Diameter, inches
FIGURE 6. DILUTION TUNNEL MIXING PROFILES FOR AEROSOLS 0.87 MICRON
IN DIAMETER
-------
28
Horizontal Axis
Verlical Axis
15
10
5
0
5
10
15
0
£
c
o
c
-------
29
V)
o
c
o
o
29*
172
I5
10
5
0
5
10
Horizontal Axis
Vertical Axis
V
0 5 10 15 20 25 05 10 15 20 25
Sampling Point Along Duct Diameter, inches
FIGURE 8. DILUTION TUNNEL MIXING PROFILES FOR AEROSOLS
5.9 MICRONS IN
-------
30
Modification of the Dilution Tunnel
Positive Pressure Operation
The original dilution tunnel arrangement, shown in Figure 3,
had the blower located at the tunnel exit to induce air flow, which gave
a negative pressure in the tunnel. With this configuration, it would be
difficult to pull a sample from the tunnel to be transferred to a residence
chamber without affecting the particulates, so the tunnel was modified
according to suggestions received from Dr. John B. Moran . The blower
was replaced with a smaller 1600 cfm blower installed at the inlet to
produce a positive pressure in the tunnel which would serve to force a
portion of the diluted exhaust gases into the residence chamber. The new
blower has adequate capacity to supply up to 1200 cfm nominal dilution flow
at 1-inch 1^0 positive pressure for modified 1A-4 cycle operation.
Change in Dilution Ratio
The modified tunnel is operated with a flow of 900 scfm to give
a 30:1 dilution ratio for the modified LA-4 cycle rather than at the previous
20:1 ratio. It is then possible to achieve more easily a final dilution of
300:1 in the residence chamber described below.
Aerosol Mixing Profile
The aerosol profile in the modified dilution tunnel under positive
pressure was measured at the 29-1/2-foot position, as described above. In
this case, the automobile was running on the chassis dynamometer at a steady
35-rnph and 3-hp as the dye was introduced.
(1) Consultation with Dr. Moran, EPA-NERC, Division of Chemistry and Physics,
Research Triangle Park, North Carolina 27711.
-------
31
The uniformity of distribution of the aerosol on the filters is
shown in Table 3. The variations were minor and the exhaust was considered
to be mixed essentially completely throughout the cross-section of the
tunnel. Nonuniform dispersion of particles within the tunnel is not,
therefore, a major source of sampling error.
Residence Chamber
Figure lOa shows a perspective view, Figure lOb shows the layout
of the test apparatus with a residence chamber placed at the end of the
positive-pressure dilution tunnel, and Figure lOc shows details of residence-
chamber probes and purge system. The residence chamber was constructed of
6-mil black polyethylene film with heat-sealed seams. Its dimensions are
about 9 ft. x 12 ft. x 20 ft. for a filled volume of about 2100 cu ft.
The flexible chamber is suspended within a lightweight external framework,
and the bottom is supported by an independent frame which can be raised to
partially collapse the chamber for purging. While diluted exhaust gases are
fed into the chamber the bottom frame is lowered as required to maintain
a constant back pressure on the proportional sampling system.
In the circulation and sampling system, the purge blower (360 cfm
at 1-1/4 inches static pressure) is used to purge the residence chamber
after a run and to pass the new charge of chamber air through a particulate
filter (99.97 percent efficiency at 500 cfm and 0.9 inches static pressure)
a drier or humidifier, and an activated charcoal absorber, to control initial
contamination and humidity in the chamber.
A 2-inch-inside-diameter PVC pipe was used to carry the sample
from the tunnel to the chamber. For convenience in preliminary residence
chamber studies, the line was installed temporarily in the tunnel between
-------
32
TABLE 3. DISTRIBUTION OF DYE AEROSOL AT REPLICATE
SAMPLING SITES
, , Photometer Reading
Filter Location Arbitrary Units
1
2
3
4
5
6
7
8
9
10
11
12
13
80.5
81.2
80.5
78.3
83.0
79.4
85.6
78.0
76.4
85.5
74.0
75.5
82.3
a = 3.59(b) , ,
(c)
Percent coefficient of variation - 4.48
(a) See Figure 9 for location diagram.
V/7/R - R")2.
(b) a = v_g(R " R) » where R is the arbitrary
n-1
photometer reading.
(c) Percent coefficient of variation = ~s~ x 100
-------
33
FIGURE 9. NUMBER CODE FOR SAMPLING PORT LOCATIONS,
LOOKING TOWARD AUTOMOBILE
(See Table 3)
-------
34
C/3
H
H
cn
O
CM
H
M
P-i
H
CO
w
H
ta
H
O
II
w
O
t-l
-------
35
H
H
|
3
Q
M
b
M
O
-------
36
CO
>H
CO
O
M
H
U
W
Q
w
o
w
o
CO
W
Pi
fn
O
o
3
O
o
o
-------
37
sections 5 and 6, 30 feet downstream from the mixing orifice, and entered
the chamber at one end 4 feet from the top, projecting 10 feet into the
chamber. The total length of this sample line was about 30 feet with two
90-degree elbows. The sample-flow velocity through the line was enough
to prevent most particles less than 4 or 5 microns from depositing on
surfaces. The final sample line was much shorter and entered the chamber
from the side as shown in Figure lOa.
Sampling from the residence chamber for particulate collection,
for visibility measurements, and for gas analysis is done through two
1/2-inch stainless steel probes extended into the center of the chamber.
The sample line from the dilution tunnel to the residence chamber
is a 2-inch ID plastic (PVC) pipe with the probe's inlet reduced for iso-
kinetic sampling.
Apparatus for Aerosol Monitoring
Filters and Impactors
The basic apparatus used to determine the particulate loading of
the diluted automobile exhaust includes absolute filters and impactors. The
particulate samples were collected on various types of filters contained in
47-mm and 142-mm-diameter filter holders. These filter holders were equipped
with critical-flow orifices which were calibrated for a 1.0 cfm flow rate.
Two types of impactors (the Andersen and a special Battelle
impactor) were used. The Andersen impactor was purchased late in this year's
program and few data were obtained with it because the large mass of the
collection surfaces made accurate weighing difficult. Thin shim stock was
then obtained for use as impaction slides in subsequent tests.
-------
38
A special cascade irapactor was designed and constructed for
particle size analysis of the exhaust particulates. This impactor has a
sampling rate of 1.0 cfm and uses a critical flow orifice to fix the flow
and particle-size cut-off on the last stage. The cut-off size of this
stage is 0.25 micron. Cut-off sizes of the upstream stages increase by
a factor of two for each successive stage from 0.25 to 16 microns. The
material smaller than 0.25 micron is collected on an absolute filter (76-mm
2
diameter, active area 21.2 cm ). The major advantage of this impactor is
that it has an unusually sharp classification, so that greater precision
is possible in particle size classification for the submicron aerosol
particles which comprise the bulk of the automobile exhaust from the rela-
tively new automobile engines.
Single Particle Counter
In order to detect slight changes in the particle size of an
aerosol, it is necessary to compare two measurements of a large particle
population within a fairly short time interval. To achieve this goal, a
single-particle counter was assembled at Battelle, and was used to study
these auto exhaust aerosols.
Basically, the instrument consists of the optics section from a
Bausch and Lomb Dust Counter, a signal processor, and a multi-channel
analyzer. In operation, the aerosol particles are passed through a light-
scattering cell and the scattered light is monitored by a photo multiplier
tube. When only one particle at a time passes through the small view volume,
it is possible to measure the amount of light scattered by it and to classify
it according to the amount of energy received by the photo multiplier tube.
-------
39
The larger the particle, the greater the quantity of light scattered. A
measure of relative particle size is obtained.
Figure 11 is a photograph of the assembled particle counter. The
particle counter is shown in the lower right portion of the photograph.
The filter circuit and the signal processor circuit are in the two boxes
to the left of it. Other apparatus shown consists of the automatic timer,
computer memory, amplifier, printer, and high speed analog-to-digital
converter. During operation, it is possible to view the pulses as they are
being counted on the oscilloscope. An automatic printer permanently records
the data after any selected time period.
As a preliminary check to determine if the apparatus functioned
properly, a pulse simulating a particle was passed through the system and
was recorded by the multichannel analyzer. This pulse was attenuated by
about 3 decibels for each trial check, representing progressively smaller
sizes. Figure 12 is a log-log plot of the pulse voltage versus the channel
in which the pulse was counted which is straight between Channels 10 and 125.
The calibrating procedure to relate particle size to pulse voltage
used aerosols prepared from various Dow Polystyrene Latices of uniform
particle size. The polystyrene latices were diluted with distilled water
and atomized into a large Mylar bag using a medical nebulizer. Many of the
atomized droplets contained no particles and very few contained more than
one. After the water evaporated, the bag contained a mono-dispersed poly-
styrene aerosol of the characteristic particle size. The aerosols were
sampled through the single-particle counter with the output of the photo-
multiplier tube fed into the 256-channel analyzer.
-------
40
FIGURE 11. PHOTOGRAPH OF SINGLE-PARTICLE COUNTER
-------
41
40
30
20
g. '0
15
O 8
o>
~Q. c
E
o
o
.c
Q_
o
0)
*-
5
3
6
CO
10
20 30 40 60 80 100
Channel Number
200
300
FIGURE 12. CALIBRATION OF SINGLE PARTICLE COUNTER
WITH SIGNAL GENERATOR
-------
42
Figure 13 is a plot of the data. In order to cover the entire
particle size range of interest, it was necessary to use two different
scales of the B & L Particle Counter. The photomultiplier output for
particles larger than 1.5 microns is proportional to the 1.64 power of the
particle diameter, whereas for particles smaller than 1.0 a., the output is
proportional to the particle diameter to the 0.55 power.
The particle counter was not calibrated for automobile exhaust,
because it would be difficult to measure the size distribution of exhaust
and to stabilize a sample for use as a standard calibration mixture. However,
the particle counter does measure particle concentration of each channel and
shows shifts in particle size or changes in light-scattering properties
during a long-term residence test.
EXPERIMENTAL RESULTS AND DISCUSSIONS
Comparative Measurements of Vehicles
Gaseous Emissions From Steady-State Operation
At intervals during break-in of each of the two vehicles, hydro=
carbon (HC) and carbon monoxide (CO) concentrations in the exhaust gas were
measured at 50-mph steady state, with nonleaded fuel No. RE-141A. The
results for the white car are presented in Table 4, and for the blue car
in Table 5.
There was no significant progressive change in the exhaust
emissions of the white car. The HC and CO concentration levels at the
beginning point (110 miles) and at the end point (3930 miles) were the same.
The HC concentration at about 920 miles was almost 40 percent above the
-------
43
c
o
o
6
af
CO
D
o.
0.
0.25
20
30 40 50 60 80 100
Chonnel Number
200 300
FIGURE 13. CALIBRATION OF SINGLE PARTICLE COUNTER
WITH POLYSTYRENE LATICES
-------
44
TABLE 4. EXHAUST GAS COMPOSITION FROM THE EXHAUST PIPE
AT 50-MPH STEADY CRUISE DURING CONDITIONING
OF WHITE CAR (UNLEADED FUEL)
Accumulated
Miles
110
577
706
919
1480
1705
1720
1940
2200
2425
2590
2704
3045
3374
3600
3760
3893
3912
3920
3930
HC
Concentration,
ppm C
1100
1095
1065
1395
1050
990
1080
1127
1050
1050
1005
1050
1050
660
795
900
990
1050
1050
1080
CO
Concentration,
percent
0.7
0.7
0.6
1.0
0.8
0.8
0.7
0.9
0.8
1.45
0.9
-
0.8
1.2
1.6
0.7
0.75
0.55
0.5
0.75
Manifold Exhaust
Air Gas
Pressure, Temperature,
in Hg F
17.0 440
458
456
18.7 462
462
18.9 462
438
463
462
467
465
469
458
462
466
19.2 480
18.5 424
424
18.5 404
18.5 446
-------
45
TABLE 5. EXHAUST GAS COMPOSITION FROM THE EXHAUST PIPE
AT 50-MPH STEADY CRUISE DURING BREAK«IN OF
BLUE CAR (UNLEADED FUEL)
Accumulated
Miles
58
602
770
953
1938
2176
2476
2629
2745
2946
3174
3394
3674
HC
Concentration,
ppm C
750
885
810
750
970
1100
980
1042
1027
900
1065
1080
975
CO
Concentration,
percent
0.6
1.07
1.15
0.95
0.8
0.7
0.5
0.6
0.6
0.7
0.75
0.7
0.7
Manifold Exhaust
Air Gas
Pressure, Temperature,
in Hg F
18.8 454
19.1
19.2
19.2
19.4
19.5
19.5
19.5
19.5
19.5
19.4
19.5 430
19.4 490
-------
46
average, and 560 miles later has dropped nearly to the average. The cause
is not known. The low HC concentration value at 3375 miles (about 40
percent below the average) followed a two-week lay-off period. During the
next 500 miles, the HC concentration rose again to the average value.
HC emissions from the blue car increased 40 percent during the
first 2000 miles, at which point the concentration was nearly the same as
the white car. At 4000 miles, HC emissions from the blue car had increased
60 percent.
For the next 4000 miles, the blue car was fueled with leaded fuel
No. RE-141C and the white car was fueled with unleaded fuel No. RE-141B.
During this period, hydrocarbon analyses were performed periodically under
50-mph steady-state conditions. Figure 14 shows a comparison of hydrocarbon
emissions from the two vehicles. The HC emissions for the white car did not
change significantly in 4000 miles of additional operation on unleaded fuel
but the HC emission from the blue car, operated with leaded fuel, increased
an additional 90 percent, so that at 9000 miles, HC emission was three
times higher than at the beginning.
Gaseous Emissions From Modified LA-4 Cycles
Integrated bag samples were collected from the dilution tunnel
during single modified IA-4 cycles on unleaded gasoline RE-141B. Both over-
night soak (about 17 hours) and rapid cooldown cycles were run. A large (about
12 cubic foot) polyethylene bag was used with a vacuum pump drawing a sample
from the tunnel at a rate of about 1/2 cfm. The tunnel flow was 600 cfm.
Table 6 presents the results of these tests. Emissions after
rapid cooldown or overnight soak were the same.
-------
47
O
o »-
c, S
11
c
"2_ c
(U
-o >«-
Is
o
o
m
CM
o
o
o
CM
o
o
m
o
o
o
o
o
o
o
o
o
CO
o
o
o
N-
o
o
o
10
V)
o -2
O ^z
o
o
o
ro
o
o
o
CM
o
o
o
0>
O j?
M
o
w
CO
M
i=>
Pi
u
>-i
Q
<
W
H
co
o
u~l
H
<
H
CO
Z)
X
w
H
CO
a ^
q
-------
48
TABLE 6. COMPOSITIONS OF INTEGRATED EXHAUST SAMPLES
FROM SINGLE MODIFIED IA-4 CYCLES
Test HC(a),
Vehicle No
Blue Car 1
2
3
White Car 1
2
3
4
5
6
(a) Instruments
Conversion
ppmC g/mi
61 1.9
-
45 1.4
31 1.0
36 1.1
38 1.2
24 0.8
72 2.3
44 1.4
were calibrated
factor: g/mi =
1
x 24700
CO
ppm
167
150
100
110
125
75
50
150
175
to read
3
/Tin , .
mm
GMV _(I
g/mi
9.9
8.9
5.9
6.5
7.4
4.4
3.0
8.9
10.0
NO
ppm
5
-
-
4
4
5
4
-
-
concentratioi
23 min
x 7^
4W) g
. X
mi
= 2
4
g/mi Conditions
0.32 Overnight soak
Overnight soak
Rapid cooldown
0.25 Overnight soak
0.25 Rapid cooldown
0.32 Overnight soak
0.25 Rapid cooldown
Overnight soak
Rapid cooldown
a.s in ppm.
3 3
0.0283 |p x (ppm) ^j-
-3
.11 x 10 (ppm) (MW) .
-------
49
Inorganic Composition of Exhaust Particulates
Exhaust particulates from unleaded fuel were sampled at 1 cfm
from the tunnel during one modified LA-4 cycle after 20:1 dilution by
filtered air. The quantities of inorganic elements collected on the
Milipore filters were measured at very low concentrations (at the p,g level)
by optical emission spectroscopy. The quantitative accuracy of these
measurements is about * 50 percent, but at such low concentrations, this
degree of uncertainty is not significant. Furthermore, the filtered
dilution-tunnel air stream without exhaust particulates contains the same
elements in about the same low concentrations as the background concentra-
tions in filtered air, as shown in Table 7.
Morphology of Exhaust Particles
The particle types and size ranges for both cars were the same
when both cars were fueled with unleaded gasoline. Primarily, there are
three types of particles: carbon black, tar droplets, and pyrolyzed chunks
of carbonaceous material. The pyrolyzed material was the least prevalent
emitted from both cars. Typical particles resembling carbon black and tar
droplets are shown in Figures 15 and 16. Almost all of the exhaust particles
were smaller than 1.0 |j,.
By the criterion of particle morphology, the two engines perform
very similarly.
Organic Analysis of Exhaust Emissions
by HPLC and GC
After break-in of the project cars on unleaded fuel RE-141A, the
cars were run through modified LA-4 cycles and at 50-mph steady-state to
-------
50
TABLE 7. COMPARISON OF INORGANIC COMPOSITIONS OF PARTICLES
COLLECTED FROM THE DILUTION TUNNEL
(a)
(Micrograms per filter after passing 23 cu. ft.
sample at 1 cfm)
Pb Zn Si Fe Ma Al Ca Cu
Filtered dilution tunnel
air only(t^ O.5 <0.5 3.0 0.5 1.0 0.5 3.0 1.0
Exhaust from unleaded fuel
in blue car sampled
from dilution tunnel(c)
-------
51
20.000X
J20168
FIGURE 15. TYPICAL CARBON-BIACK-TYPE EXHAUST PARTICLES
FROM UNLEADED FUEL COLLECTED FROM THE
TUNNEL ON 0.25 \Xa. STAGE OF IMPACTOR
20,OOOX U- in -M J20083
FIGURE 16. TYPICAL TAR-DROPLET-TYPE EXHAUST PARTICLES
FROM UNLEADED FUEL COLLECTED FROM THE
TUNNEL ON 0.5 \)a\ STAGE OF IMPACTOR
-------
52
generate samples for organic analysis by high-pressure liquid chromatography
(HPLC) and gas chromatography (GC). First, fuel RE-141A, then RE-141B, was
used to generate samples for organic analysis. In the set of runs with
fuel RE-141B, a 1967 Chevrolet was also used to generate particulate to
determine if the organic fraction from its exhaust differed from that of the
white and blue cars.
Exhaust particulate matter was collected from the tunnel on glass
fiber filters, and the filters were extracted in Soxhlet apparatus with
methylene chloride. Concentrates were transferred to small aluminum pans
(~5 mg tare), dried, and weighed on an electrobalance, after which solutions
were prepared for analysis by HPLC and GC. Gas chromatograms were developed
using a 10 ft x 1/8-inch (stainless steel) column of OV-17 on "Gas Chrom Q"
solid substrate. The temperature was programmed from 100 C to 250 C at
6/C minute. For HPLC, Waters Associates ALC-100 Analytical Liquid Chroma-
tograph was used with a 9 ft x 1/8-inch column of oxyproprionitrile (OPN)
on Porasil and with a mobile phase of 0.25 percent isobutyl ketone in iso-
octane. Peak detection for HPLC was accomplished by ultraviolet photometric
monitoring at 254 nm.
Extracts from exhaust particulates generated under different
operating conditions and by different cars were examined by both HPLC and
GC and the chromatograms were compared to identify any compositional differ-
ences. Such comparisons would indicate, along with other diagnostics,
operating similarities or differences between the two cars. In order to
evaluate the ability of these techniques to detect compositional differences,
the exhaust particulate from a 1967 Chevrolet automobile was also carried
through the same analytical procedures, and the data were compared with those
for the two matched project cars.
-------
53
Analyses by HPLC of particulate extract from the three cars are
summarized in Table 8. All cars were operated for two consecutive modified
LA-4 cycles using unleaded gasoline RE-141B, and with the choking action
not under positive control after the cold start. Several major components
are common to all samples but the ratios of these components varied widely.
Moreover, the differences between the two matched cars were no greater than
those between repeated tests on either car alone. Furthermore, the same
ratios for the 1967 car differed no more than those of the matched cars.
Therefore, the organic composition of all the particulates as indicated by
HPLC is similar, although the capabilities of the analytical method for
detecting systematic differences is limited by the variability found in
replicates.
Gas chromatography was also performed on exhaust particulate
samples generated by the three cars operated on unleaded gasoline RE-141B.
The gas chromatograms did not reveal significant compositional differences
among organic components from the three vehicles. The particulate extract
was shown to be a highly complex mixture with components eluting as a broad,
largely unresolved envelope.
The chromatograms were made with the total organic fraction of the
particulate, and the complexity of this organic mixture obscured compositional
differences which might relate to differences in fuel composition or to
engine operation. Complexity might be reduced by prefractionation of the
total extract using techniques such as thin-layer chromatography or chemical
separation. Such prefractionation would be advisable, if further samples
are to be analyzed by gas chromatography.
-------
54
S3
CJ
C
ti
3
B
o
14-1
"O
0)
r-H
a
g
cd
co
4J
5-1
cS
4J
en
O
O
CO
a)
o
o
-a
a)
o
B
M
pd o
CM
0) M-l
*T3 O
CD c!
i-l O
C -H
3 r *
n r-1
^-1 T-l
a) -a
fa J->
oo
g
Automobile and Run Number
/^
vO
CO
*_
!-i
«
CJ
0)
4-1
H
00
CO
&
}-i
CO
CJ
0)
4J
H
P
3
r^
vO
ON
r 1
r-~
01
^tT=
5-1
rt
CJ
CO
-------
55
Exhaust Particle Sizes
Mass mean equivalent diameters (MMED) of exhaust particles were
determined with removable film substrates laid in each stage of the
Battelle cascade impactor. The material was collected during four (six
in Run 42) cold start modified LA-4 cycles. Cumulative weight percent
versus equivalent particle diameter is given in Figures 17 and 18 for
each car. Both were fueled with nonleaded gasoline RE-141B. The mean
size with unleaded gasoline varied substantially in each car. Variations
were mostly a function of the amount of undersize collected on the
absolute filter backing up the last stage of the impactor. This filter
collects some adsorbed material from the exhaust and its weight is variable
from test to test.
Total Particulate Mass Loadings
The exhaust particles collected from each car run through four
modified LA-4 cycles on unleaded gasoline were weighed to determine mass
emissions. The effects of different face velocities on the collected mass
of particulates were investigated by making simultaneous collections on
47-mm and 142-mm diameter Metricel-DM filters and on a Battelle cascade
impactor with a 76-mm Metricel backup filter. All sampling rates were 1 cfm.
The 1967 Chevrolet was also vised in this series to compare
collections from four modified LA-4 cycles with leaded and with unleaded
gasolines. The results are presented in Table 9.
Later experiments compared the mass of particulates emitted during
four hours at 50-mph steady-state from the blue car (leaded fuel RE-141C)
and white car (unleaded fuel RE-141B). Again the effects of face velocity
-------
56
99
98
95
90
Q.80
Q
§ 70
cD 60
| 50
" 40
c
y 30
CD
Q.
_- 20
CD
£ 10
>
'S 5
2
I
0.5
0.2
O.I
0.05
Run 42
£
3
u
0.01
0.2 0.3 0.4
FIGURE 17.
0.6 0.8 I 2 34 68
Equivalent Particle Diameter, microns, Dp
10
20
PARTICLE SIZE DISTRIBUTION IN AUTO EXHAUST GENERATED
BY THE WHITE CAR WITH UNLEADED GASOLINE RE-14IB
-------
57
99
98
90
a
o 80
c
o
70
.£ 60
o
in
o
CD
a.
CD
CD
20
10
| 5
E
o
0.5
0.2
O.I
0.05
0.01
0.2 0.3 0.4 0.6 0.8 I 2 34 68 10
Equivalent Particle Diameter, microns, Dp
FIGURE 18. PARTICLE SIZE DISTRIBUTION IN AUTO EXHAUST GENERATED
BY THE BLUE CAR WITH UNLEADED GASOLINE RE-14IB
20
-------
58
TABLE 9. COMBINED WEIGHTS OF AUTO EXHAUST PARTICULATES FROM UNLEADED FUELS
FROM FOUR MODIFIED COLD-START(a) LA-4 CYCLES ON "METRICEL-DM"
FILTERS AND ON THE CASCADE IMPACTOR WITH BACK-UP FILTER
Automobile
Blue Car,
Unleaded fuel
Ditto
ii
ii
it
Mean
White Car,
Unleaded fuel
Ditto
ii
ii
n
" (d)
Mass
47-mm Filter^6'
0.362
1.578
0.520
0.444
0.638
0.446
0.556
0.423
0.327
0.540
0.810
E/mi
0.007
0.032
0.010
0.009
0.013
.014+. Oil
0.009
0.011
0.008
0.007
0.011
0.011
of Particles Collected
142-mm
1.209
2.083
-
0.805
1.100
1.117
0.750
0.740
0.995
1.112
1.603
., . Impactor and
Filter (e) 76-mm Filter(e)
-g/mi mg
0.024
0.042 1.783
0.566
0.016 0.441
0.022 0.805
.026+. 013
0.022
0.015 0.960
0.015 0.663
0.020 -(c)
0.022 -(c)
0.032 0.653
e/mi
_
0.036
0.011
0.009
0.016
.018±.013
.
0.019
0.013
-
-
0.013
Mean
.010±.002
.023±.007 .015+.004
1967 Car,
Unleaded fuel
1.130 0.023
3.255 0.069
1.052 0.022
1967 Car,
Leaded fuel
1.0251 0.021
0.555 0.012
1.051 0.022
(a) Cold-start cycle after rapid cooldown.
(b) Filter holder came apart during test.
(c) Negative weight obtained on high velocity stage.
(d) Six LA-4 cycles were collected on stainless steel shim stock.
(e) Filter constants: Dia., mm - 47 142 76
Active area, cm2 - 9.6 125 21.2
Face Velocity at 1 cfm, cm/s -49.2 3.8 22.3.
(f) Conversion factor:
_jpg total flow 1
1000 sample flow distance, mi
g/mi.
Exhaust flow, 30 cfm; tunnel flow, 600 cfm; LA-4 cycle, 7.5 mi.
1
1000
~T~ X 4x7 "5" = °'020 [except footnote (d)].
Mean collection
for two project cars: 47-mm filtci, 0,58 mg or 0.012 g/mi; 142-mm
filter, 1.10 mg or 0.023 g/mi.
-------
59
were studied by collecting parallel samples on 47-mm and 142-mm-diameter
filters and in the Battelle cascade impactor including the 76-mm back-up
filter. The results are shown in Table 10.
Calculating from the effective filtration areas of the 47-mm and
142-mm-diameter filters, the ratio of the face velocities at 1 cfm total flow
rate is about 13:1. Because the collection efficiency of both filters is
extremely high, little difference in the total collections would be
anticipated. Nevertheless, large differences were observed between the
total mass of particulate collected by the two filter diameters. The 47-mm
filter almost always collected less material than the 142-mm filter, under
otherwise comparable conditions. This face velocity effect has been observed
by others, but the mechanism is unknown.
In view of the difficulty of obtaining consistent collection of
unleaded auto exhaust particulate at differing face velocities, comparisons
of total mass loadings based on filter collections are arbitrarily made with
samples taken at the lower face velocities by the 142-mm filter. Comparisons
of total mass emissions on the 142-mm filter collections at 1 cfm from data
in Tables 9 and 10 above indicate that the particulate mass loadings from
the project cars are much lower at steady 50-mph cruise than with the modified
LA-4 cycle. The absolute amounts of the particulates collected are much
lower than the amounts reported by others with different cycles and different
cars. The lower amounts probably come from a combination of causes,
including differences in cooldown procedure, carburetor idle, jet adjustment,
and choke action.
(1) See review by K. Habibi, Environmental Science and Technology, Vol. 7,
pp. 223-233 (1973) .
-------
60
TABLE 10. WEIGHT OF AUTO EXHAUST PARTICULATES COLLECTED
ON METRICEL-DM FILTERS OF DIFFERENT SIZES AND
ON CASCADE IMP ACTOR
(4-hour samples at 50 mph steady-state)
Test
No.
Project
Car Gasoline
Mass of Particulates Collected
Cascade
47-mm Filter 142 -mm Filter Impactor
43
47
48
50
White
White
Blue
Blue
a/mi
g/mi
g/mi
Unleaded (RE-141B) 0.761 0.0015 0.965 0.0019 0.703 0.0014
Unleaded (RE-141B) 0.403 0.0008 0.241 0.0005 0.733 0.0015
Mean 0.0012 0.0012 0.0015
Leaded
-------
61
To determine MMED data at 50 mph from the white unleaded car
fueled with unleaded gasoline RE-141B, and from the blue leaded car fueled
with leaded gasoline RE-141C, samples were taken for four consecutive hours.
The results are plotted in Figures 19 and 20. Sixty-six weight percent of
the leaded exhaust particulates were smaller than 0.25 pjn whereas only
approximately 35 weight-percent of the unleaded particulates were smaller
than 0.25 pm.
When the white car was fueled with unleaded gasoline, there was a
substantial difference between Runs 37 and 42 (Figure 17) in the proportionate
amounts of small sized and undersize particles collected. However, there was
good agreement over the whole size range from the blue car in Runs 28, 30,
and 33 (Figure 18). Approximately 56 weight percent of the exhaust
particles from unleaded fuel are below 0.25 pm.
Factors Affecting Light Scattering in the Tunnel
Consistency in running cold-start LA-4 cycles for particulate
sampling is important, but a procedure which reproduces exhaust aerosol
characteristics has been difficult to establish. Repeated measurements of
light scatter in the dilution tunnel, mass emissions measurements, chemical
composition, plus examination of shade and distribution of collected material
on the filters revealed a lack of consistency in quantities and appearance
even with the carefully duplicated rapid cooldown procedure. Therefore,
the influence on exhaust aerosols of engine temperatures, cooldown time, and
cold-start procedures was investigated briefly.
This investigation included variations in cooldown time, selective
cooling, and controlling choke action. In connection with choke action, it
-------
62
99
98
95
90
Oa80
I 70
» 60
1 50
40
30
c
cu
O)
a
20
10
-------
63
99
95
90
^N4
50
O
48
c
o
.C
V
"o
E
(U
Q.
70
40
»- 20
x:
en
0)
a
E
O
10
5
0.5
0.2
O.I
0.05
'.01
0.2 0.3 0.4
0.6 0.8 I 2 34 68 10
Equivalent Particle Diameter, microns, Dp
20
FIGURE 20. PARTICLE SIZE DISTRIBUTION IN AUTOMOBILE EXHAUST GENERATED BY
THE BLUE CAR WITH LEADED GASOLINE AT 50 MPH STEADY STATE
-------
64
had been decided with the CAPE-19 Project Group that controlled choke
operation would be a part of the operating procedure. The blue leaded car
operating on leaded fuel RE-141C was used in this study and the choke was
modified during the course of the experiments.
A total of 24 modified LA-4 cycles were run with various conditions
of cooldown with the unmodified choke. Variations included length of cool-
down time (from three minutes to over-the-weekend), laboratory ambient
temperature (70 F and 40-50 F), and selective cooling (cooling water, exhaust
system, radiator, choke box, and oil pan). The tunnel was operated at 600 cfm
to give an average dilution ratio of 20:1 in the tunnel for modified LA-4
cycles, which were run with the dynamometer preset at 12 HP load at 50 mph.
At this setting, the modified LA-4 cycle generates exhaust gas at an average
rate of 30 cfm. The Sinclair-Phoenix photometer was used to measure light
scattering at the downstream end of the tunnel. The car was usually only
operated through the first ten minutes of the cycle when most light-scattering
effects are observed. The full cycle was run in a few experiments to estab-
lish the light-scattering pattern of the full cycle. Light-scattering effects
were negligible after ten minutes of the cycle.
Figure 21 shows light-scattering curves recorded from three runs
of widely different conditions. The first curve resulted from an overnight
soak, the next curve from a 17-minute cooldown with all cooling on, and the
last curve from a 12~minute cooldown,. The influence of the different cool-
down conditions is significant. These curves of Figure 21 are reasonably
typical of all the data, although there were exceptions that did not follow
quite as clear a trend of light-scattering versus cooldown time as illustrated.
Each of these three curves shows distinct "bursts" of light scattering
following the acceleration parts of the first five modes of the cycle.
-------
65
After overnight soak
c
JZ
o»
c
CD
O
O
C/)
g>
H
O
o
After 17-minute cooldown
After 12-minute cooldown
Modified LA~4 Cycle Time
FIGURE 21. TYPICAL LIGHT-SCATTERING PATTERNS FOR DIFFERENT
INITIAL TEMPERATURE CONDITIONS, BLUE CAR
-------
66
Figure 22 shows the area under the light-scattering curves of all
24 test runs plotted against the choke box temperature at the start of the run.
The lower the initial temperature, the greater the light scatter. At the
choke-box temperatures in the range 40 to 60 F, light scattering was extremely
sensitive to small changes in temperature, whereas, at higher choke-box
temperatures, the effect is less. The light-scattering intensity is recorded
on a log scale by the instrument and areas under the curves plotted in
Figure 22 were measured without converting to a linear scale. Qualitative
correlation between choke-box temperature and light scatter is evident.
After these tests, the choke was modified for controlled operation.
The entire assembly containing the bi-metal coil spring was replaced by a shaft
which could be rotated by hand to move the choke through its 60 degrees of arc
from full closed to full open. A quadrant plate with 5-degree marks was
mounted over the shaft to indicate choke position. A leaf spring connected
the shaft with the slotted choke lever to permit the choke plate to be moved
by the action of air flow into the carburetor.
Another 24 modified LA-4 cycle runs were made with manually con-
trolled choke action. Most of these runs were made using a constant cooldown
period of 12 minutes; in this period, all temperatures but the oil dropped
below 100 F. Several different choke-opening schedules were tried for these
test runs. The car started with difficulty and idled roughly if the choke
was fully closed; hence, all runs were made with the choke initially at 5
degrees open.
Figure 23 shows light-scattering data for three consecutive
12-minute cold-start partial modified LA-4 cycles. The choke-opening
schedule (Schedule 1) used for these test runs was as follows: start at
5 degrees open, move 10 degrees/minute to 55 degrees. Thus, the choke was
-------
67
14 p
40
60 80 100 120
Initial Choke-Box Temperature, F
140
FIGURE 22, CORRELATION BETWEEN LIGT1T-SCATTERING AND INITIAL
CHOKE-BOX TEMPERATURE, BLUE FORD
-------
in
c
0)
en
c
i_
o>
o
o
00
o
o
68
Modified LA-4 Cycle Time
FIGURE 23. LIGHT-SCATTERING PATTERNS FOR CONSECUTIVE PARTIAL MODIFIED
IA-4 CYCLES USING CONTROLLED-CHOKE SCHEDULE 2 - BLUE CAR
-------
69
almost fully open in 5 minutes. The light scattering is moderate and
consistent.
Figure 24 shows light-scattering data for another series of
three more similar cold-start modified LA-4 cycles. The choke-opening
schedule (Schedule 2) used for these test runs was as follows: start
at 5 degrees open, move 5 degrees/minute to 20 degrees, move 20 degrees/
minute to 60 degrees. The fully open position is again reached in 5 minutes,
but more choking takes place in the first 3 minutes than for the runs
illustrated in Figure 24. The light-scattering intensity is greater
because of the increased choking but the cycles are still quite consistent.
This investigation demonstrated that choke control gives accep-
table light-scattering reproducibility in repeated test runs.
Preliminary Residence Chamber Measurements
Test Conditions
After three exploratory tests in which no particles were collected
from the residence chamber, Run 4 gave preliminary data on the particle content
of the chamber. The sample line was installed and the tunnel sample-point
pressure selected before the sample flow into the chamber could be measured.
Sample flow was greater than anticipated, so that the final dilution in the
chamber was about 150:1. Although this was not the target dilution ratio of
300:1, the results from the following experiment were useful in evaluating
systems performance and feasibility.
Instruments used during Run 4 described below were: Sinclair-
Phoenix Photometer, Integrating Nephelorneter, Particle Mass Monitor, Conden-
sation Nuclei Monitor, and Single Particle Counter. In addition, particulate
-------
70
o
CO
£
li
C7>
O
Modified LA-4 Cycle Time
FIGURE 24. LIGHT-SCATTERING PATTERNS FOR CONSECUTIVE PARTIAL MODIFIED
LA-4 CYCLES USING CONTROLLED-CHOKE SCHEDULE 2 - BLUE CAR
-------
71
samples were collected from the chamber using 142-mm and 47-mm filters
and the Battelle cascade impactor.
During the chamber run, data were recorded periodically, beginning
before the start of the cycle and up to 6-1/2 hours of residence time.
Additional data were recorded the next day up to 24 hours of residence time.
Aerosol Concentrations
The white project car was operated on unleaded fuel RE-141B for
one modified LA-4 cycle to generate exhaust particles in the tunnel for
sampling into the residence chamber.
Figure 25 shows graphically the results from the Integrating
Nephelometer, Particle Mass Monitor, and Condensation Nuclei Monitor for
the first 6-1/2 hours of residence time, plotted as a percentage of the
maximum values recorded. Measurements at 21 to 24 hours of condensation
nuclei count and of light scatter by the integrating nephalometer showed no
further change from the readings at 6-1/2 hours. The particle Mass Monitor
did not seem to be working properly at 24 hours residence time.
Light scatter by the Sinclair-Phoenix Photometer (not shown)
were very similar to data from the Integrating Nephelometer, with a peak
value reached at 1.6 hours and no further change at 24 hours.
Particle Count
In some experiments, light scatter was also measured with a
Single Particle Counter. The instrument has a lower size limit of about
0.3 nm, and light scatter from particles in the range 0.3 to 4.0 jj,m is
-------
72
,
(2 4J
o
0)
H t-l
'§0
E m
o o
9A|;D|9J
-------
73
measured in 256 size classes. A particle count can be made over any time
period; however, the dilute aerosol in these experiments requires 10
minutes to obtain an adequate size distribution. The data were presented
on an oscilloscope as the counts accumulated and at the completion of the
(10-minute) count the totals were recorded numerically. These data are
described qualitatively below but are not shown in detail because Run
No. 4 was exploratory.
The total particles counted by the Single Particle Counter came
to a constant count at about two hours. However, in the smaller size
range of the classifications of the Particle Counter (not shown), there was
initially a slight decrease and then a continual increase for the next 24
hours of the test. Conversely, there was a noticeable decrease in the
number of particles in the larger range (2-4 p,m) during the whole 24-hour
period. This pattern indicates that the smallest particles below the count-
ing range were growing by agglomeration to a detectable size and the largest
particles were precipitating out of the aerosol.
Particle Mass
Three concurrent 60-minute samples were taken after four hours
residence (R = 4 hr) in the chamber, each at a rate of 1.0 cfm, using the
Battelle cascade impactor and two different sizes of absolute filters. (The
60-minute samples were taken from R-30 minutes to R+30 minutes.) Figure 26
is a plot of the data from the cascade impactor. The curve shows that
73 percent of the mass was less than 0.25 micron. The weight found at each
of the six impactor stages varied between 17 and 25 micrograms, and there
were 363 |ig on the 76-mm backup filter. A total of 497 micrograms was
-------
74
99
98
95
90
80
70
1, 60
'55
$ 50
-o 40
§ 30
i_
£ 20
o>
>
3 I0
I 5
2
1
0.5
0.2
O.I
0.05
n r»i
-O
'
X
) |
^iii'-'
-
=
J
^^
>-^
White Unleade
f~i 10 1 n r- i/i
f uei
Resic
UllUT
Date
- nc 11
ence t'
ion rat
- June
^
d Car
HB (L
me -
io
> 30,
^
^
S
/{
/
\
>
jnleaded)
4 hours
150:1
1972
/
0.2 0.3 0.4
0.6 0.8 I 2 34
Droplet Diameter, microns
8 10
20
FIGURE 26. PARTICLE SIZE DISTRIBUTION OF DILUTED AUTOMOBILE
EXHAUST IN CHAMBER (Run 4)
-------
75
collected by the impactor. The 142-mm absolute filter collected 705
micrograms, whereas the 47-ram-diameter absolute filter collected 96
micrograms. Because of these wide variations of 96, 497, and 705 p,g
in amounts collected, additional runs must be made with duplicate samples
to examine the source of the variations.
Particle Morphology
Particle morphology was determined by transmission electron
microscopy performed on collections made after six hours residence. Almost
all particles collected were under 1.0 jjjn. The most prevalent particle
found at the 0.25 (j,m stage was carbon black (Figure 27). The particle
type most prevalent in the 0.5 and 1.0 u,m size ranges was the droplet
structure containing crystal growths appearing to have been nucleated by
a small particle (Figure 28). Both the carbon black and droplet-type
particle found in the residence chamber have been found previously in
exhaust from unleaded gasoline in the dilution tunnel (Figures 15 and 16).
-------
76
J20956
FIGURE 27. PARTICLES COLLECTED ON THE 1/4-pm
IMPACTOR STAGE AFTER SIX HOURS IN
THE RESIDENCE CHAMBER
J20950
20,OOOX
FIGURE 28. A PARTICLE TYPICAL OF THOSE COLLECTED
ON THE 1.0-iam IMPACTOR STAGE AFTER
SIX HOURS RESIDENCE TIME
-------
77
MAJOR ACCOMPLISHMENTS
1. Standardization of Test Autos
Two matched autos have been made operative to generate exhaust
particulate matter. These 1970 Fords are equipped with measured and matched
1971 351 CID engines Model 351C, automatic transmissions, and 2-barrel
carburetors. The autos are as nearly identical as possible except for color,
which identifies the white car as using only unleaded gasoline, and the
blue car only leaded gasoline following a break-in on unleaded gasoline.
These cars were chosen to represent a typical and large-volume sales U.S.A.
car model. To achieve reproducible operation of the cars for generating
the exhaust particles, they are driven under consistent dynamometer load
conditions in the laboratory, using an automatic driving system, controlled
by a tape, to repeat precisely the selected driving cycle in each test run.
The cars are instrumented with thermocouples at strategic positions to show
that temperatures are normal during conditioning before each test, after
an overnight wait for the car to cool to a reproducible initial state, and
during operation in the test run. Forced air cooling is used during the
tests to maintain operating temperatures at levels closely similar to
measured values during operation on the highway.
A typical unleaded fuel was purchased in sufficient quantity to
supply the projected needs of the project to its completion, and this fuel
was also used for preparing leaded fuel. Thus, both fuels have identical
compositions except for added lead compound and scavenger in the leaded
fuel.
-------
78
2. Construction and Operation of Test Facility
As the exhaust issues from the tailpipe, the particles are
quenched and diluted with atmospheric air to simulate release into the
free atmosphere. A dilution tunnel has been built and calibrated to
receive the exhaust into a controlled flow of filtered air. The exhaust
and air pass through a measuring and mixing orifice and along the length
of the 36-foot stainless steel tunnel past sampling ports at 6-foot
intervals. Typically, the exhaust is diluted 20:1 or 30:1 by appropriate
control of the tunnel air flow.
Instrumentation was assembled and is in use to monitor exhaust
gas composition continuously for CO, HC, CO-, and NO in the tunnel in order
to appraise reproducibility of operation. A composite sample of gas from
the tunnel is withdrawn into a storage bag during cyclic operation of the
car for analysis to determine average composition of exhaust during
variable modes of operation.
A residence chamber has been constructed and used to isolate a
composite sample of the tunnel flow diluted further ten-fold to a final
dilution of about 300:1 for the exhaust. A volume of 2100 cu ft
of the final dilution is collected and can be held in the chamber for
extended periods. The gas is sampled periodically to examine composition,
particle content, and light-scattering properties for evidence of the physical
and chemical characteristics of exhaust particles as a function of their
exposure time in the atmosphere. Walls of the chamber are flexible, opaque
film mounted on a collapsible frame to minimize photochemical interactions,
and to maintain negligible pressure difference between sample and surround-
ings while the chamber is in use to collect, store, and then withdraw samples
-------
79
of diluted exhaust. Instrumentation is extensive for measuring and
recording temperature, humidity, gaseous composition, and particle content
by various procedures.
3. Conditioning of Cars
Each of the two cars has been conditioned for deposit stabiliza-
tion for a total of about 9000 miles. Initially, each car was driven for
4000 miles with unleaded fuel on a (modified) Durability Driving Schedule.
Periodic tests during the stabilization run and after 4000 miles established
their matched condition. Stabilization was completed after operation for
another 4000 miles on the Durability Driving Schedule with unleaded fuel in
one car and leaded fuel in the other. Samples are generated for examina-
tion of exhaust particles by the modified LA-4 cycle from a cold start or
from operation at 50 mph cruise mode.
4. Preliminary Particulate Measurements
Experimental measurements of the mass of particulate emissions
collected on Metricel-DM 450 membrane filters showed variations in the
collected amounts with the face velocity of the diluted exhaust approaching
the filter, as well as with the operating cycle and fuel used. Mean values
of particulate emissions from the matched cars during modified LA-4 cycles
using unleaded fuel were 0.010±0.002 (white) and 0.014t0.011 (blue) g/mi
on the 47-mm filter, and 0,023+0.007 (white) and 0.026±0.013 (blue) g/mi
on the 142-mm filters. Comparable collections on 47-mm filters during
50 mph cruise were 0.0012 g/mile from unleaded fuel in the white car, and
0.007 g/mile from leaded fuel in the blue car.
-------
80
Particle-size distributions of particles were measured in the
Battelle impactor, which collects fractions on six impact stages with
cut-offs at sizes ranging from 8 to 0.25 micrometers mass mean equivalent
diameter. The undersize particles are collected on an absolute filter
backing up the last stage with a mean pore diameter of 0.45 |am. Weights
collected from four consecutive cold-start modified LA-4 cycles in each
of five experimental runs showed a predominance of undersize from unleaded
fuel with a mean of 56 percent of the total weight on the backup filter
The morphology of particles from the dilution tunnel with each
car operating on unleaded fuel was examined and compared to determine the
similarity of the matched cars. Two types of particles predominated in
the samples. Particles on the 0.25 micrometer stage appeared similar to
carbon black, and particles on the 0.5 micrometer stage resembled tar
droplets with a spherical envelope surrounding a crystalline core.
The chemical nature of the particles was examined for both inorganic
and organic constituents to measure metals content of the exhaust particles
and to search for detectable amounts of polynuclear aromatic hydrocarbons
or other significant organic components. The amounts of samples that could
be collected were so small from the diluted aerosol from unleaded gasoline
exhaust that metals detected were not significantly above background levels
in the filtered dilution air. No single organic compounds nor significant
classes of compounds could be identified in the extract from the small
quantity of exhaust particulates available by the chromatographic methods
used.
-------
81
The properties of the aerosol particles in diluted exhaust were
examined in exploratory measurements while the aerosol was aged in the
residence chamber. The number of condensation nuclei decreased steadily
after the chamber was charged and dilution mixing was completed. Conversely,
the light-scattering properties of the aerosol, as measured by the inte-
grating nephelometer, increased markedly in the first 1.6 hours of aging,
decreased between 1.6 and 2.5 hours, and then remained approximately con-
stant. The overall results in these preliminary studies suggest that the
smallest particles below the light-scattering range agglomerate or grow
during the first two hours to a size sufficient to scatter light. There-
after, light-scattering measurements indicate little change in the aerosol.
Other measurements by a single-particle counter, which classifies the counts
into separate size ranges, indicate growth and agglomeration, with increas-
ing numbers in the sizes detectable by this instrument. Some particle
growth appears to continue as long as 24 hours, accompanied by precipitation
of the largest particles, as the count decreases in the larger size classes.
Samples were withdrawn from the chamber after six hours residence
and passed through 142-mm and 47-mm filters, and an impactor with backup
76-mm filter. The weight gains recorded after filtration of identical
volumes of gas in concurrent samples (60 minutes at 1 cfm) varied widely
and correlated positively with the diameters or areas of filters used. No
conclusions were possible concerning the absolute weight concentration of
the filterable aerosol particles.
Morphology of the particles collected in an impactor after six
hours aging in the residence chamber was examined. They were similar to
particles collected in earlier runs from the dilution tunnel.
-------
82
FUTURE WORK
The studies on characterization of auto exhaust particles will
be continued with the use of the residence chamber for holding a diluted
portion of the generated auto exhaust to examine the aerosol for changes
in properties of the particles in the dark. The changes attributable to
leaded vs. unleaded fuel, residence time, relative humidity, and the con-
centrations of two common atmospheric pollutants will be studied in a
series of tests designed to show statistically significant differences in
particle properties.
The variability of weights of particles collected on filters
resulting from variations in filter properties and in collection conditions
represents a serious handicap in the search for quantitative interpreta-
tions of the characteristics of exhaust particulate matter in the atmosphere,
Studies will be continued with comparison of results on a relative scale,
and efforts to resolve the problem of measuring the absolute mass concen-
tration of auto exhaust particulates will be continued.
------- |