The Economic Effectiveness of
Mandatory Engine Maintenance for
Reducing Vehicle Exhaust Emissions
Volume III
Inspection/Maintenance Procedures Development
August 20, 1971
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, I NC.
THIRTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVIRONMENTAL PROTECTION AGENCY
AIR POLLUTION CONTROL OFFICE
5600 FISHERS LANE
ROCKVILLE, MARYLAND 20852
SCOTT RESEARCH LABORATORIES, INC
f. o. max JMI«
•AN •CMNAKDIHO. CALIFORNIA M4M
ONE SPACC fARK • KtDOHOO KACH CALIFORNIA •
-------
The Economic Effectiveness of
Mandatory Engine Maintenance for
Reducing Vehicle Exhaust Em issions
Volume III
Inspection / Maintenance Procedures Development
August 20, 1971
I N SUPPORT OF:
APRAC PROJECT NUMBER CAPE-13-68
FOR
COORDI NATI NG RESEARCH COUNCIL, INC.
THI RTY ROCKEFELLER PLAZA
NEW YORK, NEW YORK 10020
AND
ENVI RONMENTAL PROTECTION AGENCY
AIR POLLUTION CONTROL OFFI CE
5600 FI SHERS LANE
ROCKVI LLE, MARYLAND 20852
," ,"
APPROVED BY %?~aNI14/@~
RI CHARD R. KOPPANGtJ
PROJECT ENGINEER
~~~
NEAL A. RICHARDSON
PROJECT MANAGER
TRW
)I~I SCOTT RESEARCH LABORATORIES, INC.
~ P. O. .0. 8011.
SAN .XRNARDINO. CALI..ORNIA .-
SYSrE- _"
ONE SPACE PARK' REDONOO BEACH. CALIFORNIA 90278
-------
PREFACE
This report consists of three volumes entitled: liThe Economic
Effectiveness of Mandatory Engine Maintenance for Reducing Vehicle
Exhaust Emissionsll. The following are the titles given for each
volume:
. Executive Summary, Volume I
. Modeling of Inspection/Maintenance Systems, Volume II
.
Inspection and Maintenance Procedures Development, Volume III
The first volume summarizes the general objectives, approach and
results of the study. The second volume presents the analytical modeling
of a mandatory inspection/maintenance system and simulation results
obtained using that system model. The experimental programs conducted to
develop input data for the model are described in Volume III.
The work presented herein is the product of a joint effort by TRW
Systems Group and its subcontractor, Scott Research Laboratories. TRW,
as the prime contractor, was responsible for overall program management,
experimental design, data management and analysis, and the economic-
effectiveness study. Scott conducted the emission instrument evaluation
and acquired and tested all of the study vehicles. Scott also provided
technical assistance in selecting emission test procedures and in evalu-
ating the test results.
i
-------
TABLE OF CONTENTS
1.
INTRODUCTION AND SUMMARY....................................
ENGINE PARAt~ETER SURVEY................. . . . . . . . . . . .. . . .. . . . .
2. 1 Parameter Survey Summary Resul ts .......................
2.2 Parameter Selection Criteria .......... .................
2.3 Vehicle Sample Identification and Acquisition ..........
2.4 Vehicle Inspection Procedure ........ .... ...............
2.4.1 Test Station Arrangement ........................
2.4.2 Measurement Procedure and
2.
Test Equipment..................................
3.
PARAMETER SCREENING EXPERIMENTS.............................
3.1 Test Results and Conclusions ..... ...... ................
3.1.1 Screening Experiment ............................
3.1.2 Choke Parameter Experiments.....................
3.2 Parameter and Diagnostic Mode
Selection Criteria ............. ........................
3.3 Experimental Design ........ ...................... ......
3.3.1 Power Train Selection ...........................
3.3.2 Experimental Approach ...........................
3.4 Test Procedure.........................................
3.4.1 Vehicle Preparation ............. ................
3.4.2 Measurements and Special Test Equipment .........
3.4.3 Orthogonal Test Procedures......................
3.4.4 Cold Engine Parameter Tests .....................
3.4.5 Test and Measurement Quality Control............
3.5 Data Reduction and Analysis Procedure. .................
4.
DE FIN IT I V E EX PER I ME NT .......................................
4.1 Test Results and Conclusions ...........................
4.1.1 HC Emissions Response ............ ...............
4.1.2 CO Emissions Response ...........................
4.1.3 NO Emissions Response ... ........ ...... ..........
4.1.4 Emission Di agnosti c Si gnatures ..................
4.2 Emissions Stability Tests and
Experimental Error.....................................
4.3 Air Cleaner Tests ........... .... II II ...................
4.4 Experimental Design .......... ...... ....................
4.5 Data Reduction Procedures .......... ....................
4.6 Test Procedures........................................
iii
Page
1-1
2-1
2-1
2-12
2-17
2-22
2-22
2..22
3~1
3-1
3-1
3-22
3-27
'3-30
3-30
3-30
3-34
3-34
3-35
3-38
3..42
3-42
3..44
4-1
4-1
4-1
4-6
4-6
4-7
4-14
4-17
4-23
4-29
4-34
-------
TABLE OF CONTENTS (continued)
4.6.1 Vehicle Preparation ...... ...... .....,........... .
4.6.2 Measurement and Special Test
Equipment...................................... .
4.6.3 Orthogonal Test Procedures ........ ...... ........
4.6.4 Air Cleaner Restriction Experiment ...... .... ....
4.6.5 Test and Measurement Quality Control. .... .......
REFERENCES
...... ... ............ .... ... .... ....... ........ ... ......
APPENDIX A SCREENING EXPERIMENT DATA SUMMARY........... ..........
APPENDIX B DEFINITIVE EXPERIMENT DATA SUMMARY....................
lV
Page
4-34
4-35
4-35
4-38
4-39
R-1
A-1
B-1
-------
1.
INTRODUCTION AND SUMMARY
The experiments performed to develop, characterize, and rank vehicle
inspection and maintenance procedures are described in this volume. The
data from these experiments are the fundamental building blocks for the
economic-effectiveness model of Volume II.
The experimental data were developed from the following sequence of
investigations as indicated in the flow diagram, Figure 1-1.
.
An engine parameter
general maintenance
systems
A screening experiment to select the most frequent and
extensive malfunctioned engine parameters
field survey to characterize the
state of vehicle emission control
.
.
A definitive experiment to select and evaluate the
most effective parameters to maintain in an inspec-
tion as determined by combining the data from the
previous two experiments
These experiments were designed to yield a consistent set of data to
support the emissions models. Inspection procedures using both direct
engine parameter and emission diagnosis were developed with these data.
The approach is represented mathematically by the following equation relating
a change in emissions (composite or diagnostic) to the frequency, extent, and
sensitivity of the parameter malfunction.
~e.
J
f~ (P 1 )
C P1
ae.
~ dP i +
ae.
ap J dP 2 + . . +
2
Jp' (P n)
C P
n
ae.
ap J dP n
n
fP' (P 2)
C P2
1- 1
-------
Literature Survey
Parameter Selection Criteria
26 Engine parameters, Pi'
identified
GM Idle Adjustment
Da ta Set
228-Vehicle test set
Mode emissions, ern
Engine parameters, Pi .
P.
,
Engine Parameter
Field Survey
Malfunction Frequency and Extent
244-Vehicle test set
26 Enqine parameters/vehicle
P(Pi}
Screening Experiments
Orthogonal Tests
Emission Response to Malfunctions
11 Vehicles
13 Malfunctioned parameters/vehicle
33 Diagnostic modes, em
32 Emission tests/vehicle
~e.
-..J.
,JP.
Definitive Experiments
Orthogonal Tests
Emission Response to Malfunctions
11 Vehicles
5 Malfunctioned parameters/vehicle
24 Emission tests/vehicle
Figure 1-1.
Experimental Program Flow Diagram
1-2
P ( Pi), em
Economic-effectiveness
'computer model
ae.
em' .=...:..J..
op.
,
-------
where:
.th . .
= J emlSSlon, either composite or diagnostic
. th
= 1 parameter
= cut point rejection criterion on parameter distribution
= distribution of parameter P. in the general vehicle
population 1
The distribution, P(P.), is developed from the parameter survey;
1
the emission response, ae.jaP., from the statistically designed experiments.
J 1 .
e.
J
p.
1
CPo
1
P( P.)
1
The following conclusions were drawn from the experimental program.
'Parameter Survey
The most frequently occurring out-of-specification parameters which are
known to have or suspected of having significant effects on emissions are:
Parameter
Related Subsystem (See Table 2-8)
Electrical
Electrical
Perfonnance
. Ignition Timing
. Vacuum advance
. Medium load induction
% CO at 30 and 45 mph
. Idle speed
. Idle fue1-to-air ratio (% CO)
. Air cleaner restriction
. Float level
Carburetor
Carburetor
Carburetor
Carburetor
Based on the literature survey data, these additional parameters have
significant effects on emission, although malfunction frequencies are low:
. Misfire
. Air injection
. Positive crankcase
vent valve
Electrical
Emission control
Emission control
Statistically Designed Experiments
The above engine parameters were studied in statistically designed
experiments where in the parameters were systematically malfunctioned
and exhaust emissions measured. The results are:
o The most effective engine parameters to maintain in the general
population are: secondary ignition system (spark plugs, wiring
when causing misfire); idle timing, rpm and fue1-to-air ratio; air
1-3
-------
.
cleaner, positive crankcase vent valve; and air injection system.
Air cleaner restriction has a highly variable impact on CO emissions.
.
Idle fue1-to-air ratio {idle CO} affects CO emissions significantly,
having slight to strong nonlinear influences on emissions.
.
Basic timing significantly affects NO and HC emissions under most
engine loading conditions.
Misfiring is the predominant malfunction influencing HC emissions.
.
.
Fast idle modes are not substantially different from closed throttle
in diagnosing malfunctions.
Emissions measured under differing engine loads will diagnose the
malfunctions listed above; however, no measurement uniquely points
to a single malfunction.
.
Choke Experiments
Composite emissions are not very sensitive to choke blade setting
with heat riser operating nominally.
. Composite emissions are very sensitive to blade settings with the
heat riser valve frozen open.
.
Air Cleaner Experiments
Eight of the eleven power trains evaluated had linear emission
responses to air cleaner restrictions.
Experimental Technique
.
The results of the air cleaner, screening, and definitive experiments
were generally in good agreement with each other.
.
Emission responses to engine parameters for power trains
representative of the general population were generally in
close agreement with each other for all parameters except
cleaner restriction and vacuum advance.
.
Properly designed orthogonal experiments will provide accurate
and precise data on the sensitivity of emissions to engine
parameters with experimental error approxinlately within vehicle
emissions variability.
1-4
-------
2.
ENGINE PARAMETER SURVEY
This section describes the field survey conducted to characterize the
general maintenance state of vehicle emission control systems. The survey was
conducted for TRW by Scott Research Laboratories in the following sequence:
. A literature search was performed to identify those engine
parameters which affect vehicle emissions.
. A sample of 244 vehicles was acquired and inspected for these
engine parameters from which 227 valid tests were obtained.
. The results were summarized in a statistical format com-
patible with the economic-effectiveness model.
The survey results are discussed first.
2. 1
PARAMETER SURVEY SUMMARY RESULTS
The basic data obtained during the parameter survey have not been
included; however, the main conclusions of this program are summarized in
Tables 2-1, 2-2, and 2-3. The raw data are available but not reported since
the primary use is in determining frequency of malfunction, the mean and
standard deviation of the parameters.
The most frequent out-of-specification parameters from the parameter
survey, shown in Table 2-1, were ignition timing (basic and vacuum advance)
point dwell, medium load induction system (as reflected by CO), idle speed,
idle CO, air cleaner restriction, and float level. In addition, based on
literature survey information, misfire, air injection system and positive
crankcase vent valve are significatnt malfunctions in terms of their effect
on emissions, even though nominally their failures were found in less than
12% of the population.
The ignition electrical subsystem variables (coil and required voltage)
have been transformed into a single parameter, the ratio of coil available
to plug required voltage, to obtain a parameter more relevant to misfire.
That is, for vehicles with low ratios, misfire in the secondary might occur.
The loaded mode CO and misfire essentially reflect subsystem performance.
For example, misfire under load may be the result of low ratios of available-
to-required voltages (values of 150% were found to result in misfire at load),
2-1
-------
Table 2-1.
Parameter Survey Summary Results
%' % Cars Rejection
Criteria
Parameter
Satisfactory Unsati sfactory Reported Deviation
from Spec.
Basic Timing 23.8 76.2 227 ~:i:2 deg
Total Advance 65.2 34.8 227 ~:i:2 deg
Mech. Advance 70.5 29.5 227 ~:i:2 deg:
Vacuum Advance 67.0 33.0 227
Plug Required Volt. 93.8 6.2 226 >2 kv
Coil Avail. Volt. 86.7 13.3 226 ~2 kv
Point Dwell 58.6 41. 4 227 ~1 deg
30 mph Cruise, % CO 82.0 18.0 127 >2.5% CO
Plug Required Volt. 92.1 7.9 227 >2 kv
Misfire Rate 97.3 2.7 225 >010
45 mph Cruise, %CO 81. 0 19.0 127 >2. 0% CO
.
Plug Required Volt. 88.9 11. 1 226 >2 kv
Misfire Rate 95.6 4.4 225 >0%
60 mph Cruise, WOT
Plug Required Volt. 86.3 13.7 227 >2 kv
Misfire Rate 90.3 9.7 227 >0%
Idle Speed 29.5 70.5 227 ~:i:30 rpm
Idle CO 44.1 55.9 227 >3.510
Manifold Vac., in. Hg 80.9 19.1 225 ~2 in. Hg
Air Cleaner Restriction 77.1 22.9 227 In red zone
of AC
t
Float Level 65.6 34.4 221 ~:i:0. 06 in.
Heatriser Valve 78.2 21. 8 225 Frozen
PCV Performance 88.1 11. 9 227 <0 in H20
Air Pump Performance 90.4 10.6 50 . +
~O. 2% CO
Vacuum Leaks 97.7 2.3 224 Disconnected
+CO . -CO
dIsconnected connected
2-2
-------
Variable
Nwnber
1
2
3
4
5
6
7
8
9
10
Statistical Summary
All Years, Engine Modification Systems
Estimate of
Standard
Deviation
S
5. 8
6. 7
Table 2-2.
Parameter
(units)
6. Timing, deg
Vacuwn Advance, deg
% Available Voltage
(Coil Avail/Plug Req'd), %
*
Misfire Rate at 30 mph, %
at 45 mph, %
at 60 mph, %
ICO %
6. IRPM, rpm
A/ C Restriction, deg
PC V Pressure. in. H20
Air Pump Pressure. psi
Odometer Reading. mi
Estimate
of Mean
X
0.7
19. 8
308
12
15. 3
17.6
3.69
6
27
- 0.4
N.A.
2 867 1
155
o
5.4
8.4
1. 94
95
45
0.6
N.A.
18514
Sample
Size
136
136
133
6
11
22
136
136
136
136
N.A.
136
::
-------
severe mistiming, or short circuits in the secondary (plugs or wires)
or combinations there of.
Tables 2-2 and 2-3 summarize the mean values and standard deviations
of all vehicles and give a breakdown by air injection and engine modification
vehicle engine parameters. The difference between the mean value of the
actual measurement and the manufacturer's specification or the nominal (0)
state of the engine is an indication of the severity of the malfunction.
Parameters of most interest in an enforced inspection/maintenance pro-
gram are those which are frequently out of specification and result in high
emissions. High emissions will result if a parameter with small effect on
emissions is grossly out of specification with regard to extent and frequency,
or if a less frequent malfunction has a great effect on emissions. In the
former, either one- or two-sided inspections would be effective (using rejec-
tion criteria on both sides of a specification tolerance band). In addition,
a one-sided inspection of parameters with large variability on the side which
reduces emissions may be economically effective. The parameters presented in
Tables 2-2 and 2-3 fulfill this requirement. Those which could be effectively
inspected in one- or two-sided inspection are idle CO and timing. The remain-
ing parameters could most effectively be inspected using a one-sided test.
Of the parameters evaluated, some are constrained by physical design
or definition to values greater than some constant value, usually zero.
Because of this, their distribution functions are not normally distributed,
particularly when the mean values are close to zero. All of the emission
measurements, air cleaner restriction, air pump restriction, and positive
crankcase ventilation (crankcase pressure) have skewed distributions
resulting from constraints. A frequency diagram for air cleaner restric-
tion, as shown in Figure 2-1, approaches a binomial distribution. The
preponderance of air cleaners (50%) are at the nominal state. An extremely
long "tail" over the full range of the instrument measurement capability
starts well to the left of the mean value. One can assume that a certain
percent of the population has neglected air cleaner maintenance.
2-4
-------
N
I
CJ'1
:Q 80
<
U
&&..
o
o
Z 60
l
>-
u
Z
w
:J
o
w
~ 40
120
227 P~RAMETER SJRVEY VEHIC~ES
AIRCLEANER RESTRICTION MEASURED
WITH MODIFIED AC TESTER
X
I I
I
I .
100
20
o
o
40
20
60 80 100 120
AIR CLEANER RESTRICTION, DEGREES
140
160
180
Fi gure 2-1.
Frequency Diagram for Air Cleaner Restriction
-------
2.1.1
Primary and Secondary Ignition Subsystem Evaluation
Some of the parameters measured (loaded CO and misfire, for example)
reflect the overall condition of a subsystem rather than a specific compo-
nent. A more detailed breakdown of the component performance for those
vehicles having misfire appears in Table 2-4. The misfire at load is com-
pared against the ratio of avai1ab1e-to-required voltage, deviation from
manufacturer's specification of coil available voltage and the dynamic com-
pression test. ,The following conclusions were drawn:
.
The cylinder balance (dynamic compression) test reflected
that 2.2% of the total population of vehicles were misfiring
at 1500 rpm, no load (5 failures out of 226 vehicle samples).
'.
A shorted secondary ignition component (wire or plug) was
the predominant mode of failure.
.
Misfire is generally poorly correlated to coil available
voltage at low engine load.
.
On 25% of the vehicles misfiring at wide open throttle,
low ratios of avai1ab1e-to-required voltage were also
evident.
In summary, there appears to be poor correlation between the electrical
subsystem variables measured and misfire. Approximately 50% of the vehicles
misfiring under loads equivalent to the Federal cycle were also misfiring
under the 1500 rpm, no-load condition. This suggests that an idle HC diag-
nostic inspection might be an effective procedure for identifying this
malfunction.
2.1.2 Induction Subsystem Analysis
A similar analysis was performed for the induction system components
known to affect loaded mode CO. What was sought was an estimation of the
number of induction system failures likely to be in the main and power
circuits of the carburetor rather than in related components, such as PCV,
air cleaner and float level. Refer to Volume II, Table 3-2 for table of
failures. Half of the failures shown in the table are related to external
2-6
-------
Table 2-4. Typical Electrical System Malfunctions and Their Correlation to Misfire
Parameter Survey Study
Misfire, % Deviation of Coil Available
Car No. 30 mph 45 m ph WOT Available, kv to Required Volt. Dynamic Compression *
23 12 12 24 -10 2.44 9 7 11 9 11 @) 9 9
65 0 0 36 4 3. 00 8 8 10 10 9 9 10 10
66 12 12 12 4 2. 57 8 7 3 7 9 9 @ 7
114 12 12 12 4 4.50 8 8 10 8 8 (0 8 8
127 0 0 24 0 2. 28 10 10 9 10 11 9 10 9
N
I
......
160 0 20 20 4 1. 63 10 8 7 9 7 9 10 7
177 0 0 6 4 2. 25 10 9 10 9 11 10 11 9
192 0 0 12 -8 1. 53 Transistor ignition
211 12 24 24 -4 2. 00 5 5 0 6 5 5 5 3
212 0 12 12 -3 4. 18 9 5 6 10 5 7 8 8
229 12 12 12 0 3. 54 8 7 7 8 <£) 9 6 8
243 12 25 25 0 2. 93 6 7 8 5 5 6 6 7
247 0 0 12 0 2. 60 7 8 7 6 8 6 8 7
* rpm drop when cylinder spark plug shorted
-------
component failure rather than metering malfunctions. This analysis indi-
cated that between 9 and 10% of the parameter survey vehicles had these
carburetion malfunctions, assuming the rejection points of 2.5 and 2.0% CO
at 30 and 45 mph road loads, respectively. The unknown is how many of these
vehicles had rich fuel adjustments when new or rebuilt. It
is difficult (with the exception of the accelerator pump/power jet actuator)
to hypothesize failure mechanisms for carburetor main meter circuits which
would increase fuel flow. Data similar to that of Volume II, Table 3-2
indicate that only 50% of the vehicles were common to both sets of data
(vehicle sets obtained by rejecting the 30 and 45 mph loaded CO mode emis-
sions data). It is a demonstrated fact that idle CO adjustments of suffi-
cient severity will affect CO emission in the lower loaded modes, depending
on specific power train configuration. The data indicated that 88% of the
vehicles which failed the 30 mph CO criteria had idle CO emissions as high
or in excess of 5%. These rich adjustments, on an average, would be anti-
cipated to increase CO at 30 mph from 0.5 to 1.0%. Carry-over of rich,
idle CO in the heavier loaded 45 mph mode is substantially less than in
the 30 mph mode.
2.1.3 Air Pump Malfunction Evaluation
Two diagnostic approaches to characterizing air pump malfunctions
were evaluated:
.
Application of the manufacturers' field test specifications
.
Inference from CO emission measurements under load
In the former, a pressure measurement at the air pump discharge (either dead
headed or free flowing) is made. Low pressures indicate a leaking pump or
bypass valve. The emission measurement technique requires CO to be mea-
sured under load with the pump operating and disconnected. This is a more
reliable indication of malfunction, little or no changes being indicative
of inoperability. The loaded mode was selected to minimize the influence
of the idle adjustment on air injection effectiveness. These adjustments,
particularly idle CO (Section 4), have a significant effect on pump per-
formance and confuse a diagnosis in the idle mode. Comparisons of vehicles
2-8
-------
rejected by both criteria indicate extremely poor correlation between pro-
cedures. Of the six known air injection system failures, only two were
identified with the pump discharge measurement. It is not known why the
pump pressure diagnosis was so poorly correlated. However, the procedure
only diagnoses those components upstream of the pump discharge. Restric-
tions or leaks in the manifolding and injection nozzles are not diagnosed.
2.1.4 Comparison with GM Data
The APRAC CAPE-13 Committee supplied the GM idle adjustment program
basic data set. This data set on 228 emission-controlled California
vehicles consisted of as-received and post-idle adjustment mode and com-
posite emissions, scope diagnosis, and idle rpm and timing measurements.
These data were compared with idle adjustments from the engine param-
eter survey for air injection and engine modification vehicles, respec-
tively (see Table 2-5). The mean biases from specification or nominal
state and their standard deviations agree fairly well between data sets.
The most significant deviations are between engine modification vehicles.
for idle rpm.
Estimates of failure rates for the air injection system and misfire
at idle were made. Air pump malfunctions for the GM set were estimated
from mode dilution correction values. The hypothesis applied to fail
this subsystem is that the dilution factor should exceed unity
for a nominal system since excess pump air dilutes the exhaust gas.
Application to the idle mode must be done with some care as idle adjust-
ments can produce dilution corrections of 1.1 or less, even when the air
injection system is functioning properly. This is most dramatically shown
by using the failure criterion, dilution correction equal to or less than
1.1 on the as received vehicles (see Table 2-6).
2-9
-------
N
I
.....
o
Table 2-5. Comparison of Parameter Survey with General Motors Idle Adjustment Data
Parameter -
GM Data TRW Parameter Surve\ Data
Mean I Standard Error I Sample Size Mean I Standard Error Sample Size
Timing, Degrees
Air Injection 0.55 3.45 124 1.3 5.1 91
Engine Modification 0.89 4.54 105 0.7 5.8 135
RPM
Air Injection -54 85 124 -32 88 91
Engine Modification -28 77 105 6 95 135
Idle CO,%
Air Injection (pump 3.42t 2.21t 124 3.77 1.85 91
disconnected)
Engine Modification 4.3 3.27 105 3.69 1.94 135
Air Pump,% 5.8* 10**
Not Operating
Misfire,% at Idle 2.2 2.2
-
*
Inferred from dilution correction at 30 mph mode.
**Inferred from no change in CO when pump hose disconnected, 50-vehicle sample.
tCorrected for disconnected air pump.
-------
Table 2-6,
Air Injection System Diagnosis Using Several Inspection Policies Based on Dilution Correction
GM Idle Adjustment Program
Inspection Policy, Maintenance Vehicles Actual No.* "**
Errors of
Dilution Correctiont S ta te Rejected Fa i1 ures COl1llliss ; on
1.1 at idle As received 27 7 20
1.0 at idle As received 7 6 1
N 1.1 at idle Idle parameters 7 7 0
I
..... adjusted
.....
1,2 at 30 mph As received 5 5 0
*
Total number of vehicles in set with air injection system malfunctions was
**
"Vehicles which were rejected but not actually failed.
tO'l t. t' - 15
1 U 10n correc 10n - 6HC + 0.5CO + C02
seven.
-------
2.2 PARAMETER SELECTION CRITERIA
The primary task in developing reliable inspection and maintenance pro-
cedures is to select engine parameters most likely to have adverse effects
on emissions. Selection of the parameters used in the development program
was based on results of a comprehensive survey of the literature (References
1 - 10). Forty-four components and conditions (see Table 2-7) were found to
have potential effects on exhaust emissions. Besides effects on exhaust
emissions, parameter selection criteria included consideration of inspection
and maintenance costs and ~heir ability to malfunction or slip out of adjustment.
Final selection was made of 26 inspection parameters capable of being
di agnosed with a hi gh probabi 1 ity. These items appea"r on the engi ne
parameter inspection sheet (see Table 2-8) used in this program.
An inspection procedure was developed and implemented to identify
the source area of indicated failure and, wherever possible, the specific
component or part causing the trouble. This procedure purposely required
the use of conventional diagnostic equipment, except for a chassis dyna-
mometer.
Some malfunctions could occur without detection under the inspection
procedure used. These include primary ignition circuit condenser;
distributor point cam and drive; carburetor choke system and pump circuit;
air cleaner hot air door thermostat; air injection pump gulp or dump valve,
check valves and hoses; combustion chamber deposits; cooling system temp-
erature; and exhaust system back pressure. Also, the procedure could
detect fuel system malfunctions but fail to clearly identify their specific
sources, such as carburetor deposits, bowl vent, hot idle compensator and
fuel supply. For these reasons, the criteria for engine parameter selection
stressed high probability of detection and identification of malfunctions
or maladjustments.
2-12
-------
Table 2-7.
Engine Parameters Which Affect Exhaust Emissions
1.0 Ignition System
1.1 Electrical, Secondary
, Spark plugs
, Wires
, Distributor cap
, Rotor
, Coil (secondary)
1.2 Electrical, Primary
, Points
, Condenser
, Coil (primary)
, Primary resistance (ba 11 as t )
, Alternator circuit
1.3 Distributor, Mechanical
, Timing
, Poi nt dwe 11
, Vacuum advance
, Mechanical advance
, Cam
, Distributor drive (excessive end play)
2.0
Carburetor
2.1 Idle Circuit (speed & mixture)
2.2 Low Speed (off idle)
2.3 Choke (including fast idle)
2.4 Float Circuit
2.5 High Speed (part load)
2.6 High Speed (full load)
2.7 Depos its
2.8 Pump Circuit
2.9 Air Cleaner Element
2.10 Air Cleaner Thermostat
2.11 Fuel Supply (filter & pressure)
2.12 Bowl Vents
2.13 Hot Idle Compensator
2-13
-------
Table 2-7.
Engine P~rameters Which Affect Exhaust Emissions (Cont.)
3.0 Exhaust Emission Devices
3. 1 Air Injection Reactor Systems;
. Pump (including relief valve)
. Gulp valve (or bypass)
. Check valves
. Hoses
3.2 Engine Modifications
. Vacuum advance control valve
. Vacuum. retard (idle only)
. Mechanical retard (idle only)
. Fuel shutoff (decel. only)
4.0 Engine
4.1 Combustion Chamber Deposits
4.2 Valve Condition (compression)
4.3 Intake Manifold Vacuum (leaks)
4.4 Blowby (PCV systems & rings)
4.5 Heat Riser Control
4.6 Cooling System (temperature)
4.7 Exhaust Back Pressure (muffler)
2-14
-------
Table 2-8. Engine Parameter Inspection
Performed by
Scott Research Laboratories, Inc.
Car No. 114
1.
Vehicle Identification No.
1. 1 Year, Make 68 Dodge
1. 2 Model Monaco 500
1. 3 License No. 6H3429 New York
1. 4 ado., Miles 22803
1. 5 Name N. E. Knox
1. 6 Address 593 Courtland Dr.
1. 7 Engine Size, CID 383
1. 8 Carb, Make (fbbls) Carter 4V
1. 9 Transmission Auto
2.
Electrical Inspection
1. 10 Exh. Emission System Eng. Mod
1. 11 Crankcase Ems System Closed
San Bernardino, California
Satisfactory
Measured Spec. Departure Yes No
+ 11 +5 X
45 41-50 X
26 23-26 X
19 18-24 X
16 20 X
36 28-32 X
29 28-33 X
L 8 2. 8 3. 10 4. 8
5. 8 6. 0 7. 8 8. 8
2. 1 Basic Timing, deg
2. 2 Total Advance, deg
2. 3 Mech. Advance, deg
2. 4 Vacuum Advance, deg
2.5 Plug Req'd Volt, kv
2. 6 Coil Avail. Volt, kv
2. 7 Point Dwell, de g
2. 8 Dynamic Compression, %
Change in Speed
3. Performance Diagnosis
3. 1 30 mph Cruis e at 16 in. hg CO, % 1.5 X
. Plug Req'd Volt, kv 8 X
. Misfire Rate, % 12 X
3. 2 45 mph, at 8 in. hg CO, % 1. 8 X
. Plug Req'd Volt, kv 9 X
. Misfire Rate, % 12 X
3. 3 60 mph, WOT
. Plug Req'd Volt, kv 12 X
. Misfire Rate J % 12 X
2-15
-------
Table 2 -8.
Engine Parameter Inspection (Cont. )
4.
Carburetor Inspection
4. 1 Idle Speed, rpm (N or D)
4. 2 Idle CO, %~'
4. 3 Manifold Vacuum, in. hg
4. 4 Air Cleaner, angle deg
4. 5 Float Level, in.
4. 6 Heat Riser Valve
(None, Free, Froze)
5.
Emission Control
5. 1 PCV Perf. at Idle, in. H O':'~'
2
5. 2 Air Pump Performance, psi
5. 3 Vac Leaks (Yes or No)
5. 4 Idle Speed (rpm in N or D)
Remar ks:
Car No. 114
Measured Spec.
560N 650N
5. 0
17
5
7/32
Satisfactory
Departure Yes No
X
X
7/32
x
X
X
X
Free
-0. 2 X
None
No X
560-N 650-N X
.'.
"'Without Air on Cars with Air Pumps.
..I~ "..
"""Vacuum is minus (-), and Pres sure is plus (+).
2-16
-------
2.3 VEHICLE SAMPLE IDENTIFICATION AND ACQUISITION
Various sample sources from which to select vehicles for characterizing
their general maintenance states were considered. Requirements for the source
selection were:
.
A large population of 1966-69 California vehicles
.
Minimum acquisition costs
Typical maintenance states
.
Vehicle sources considered were: 1) random selections from within the
San Bernardino area; 2) constrained selections from large industrial
parking lots; and 3) constrained selections from commercial sources such
as rental agencies or used car lots. The second approach was selected.
Although a completely random fleet of vehicles would statistically
be preferable. the cost and logistics of acquiring these vehicles was
judged to be excessive. For example. an identification. pickup. delivery.
and possibly an incentive system would be required since vehicles would be
held as long as 4 hours. Incentive has the potential danger of biasing the
sample toward the more poorly maintained vehicles. The attendant scheduling
problems would also increase the cost of the experiment. Therefore. this
approach was rejected. The use of commercial vehicles (i.e.. rental vehicles
or used car lots) was rejected on the basis of both cost and poor matchinq of
maintenance states within the general population. The identification and
scheduling of these vehicles is significantly less costly than a random
selection.
The best compromise was to borrow vehicles from industrial parking
lots during working hours. Enough vehicles were picked up in the morn-
ing to establish a test reserve to assure maximum use of manpower and
facilities. To the extent that the industrial employees represent a
micro-scale of the general population. their vehicles are a valid sta-
tistical sample.
2-17
-------
To support this decision a questionnaire was circulated to TRW
Systems and Aerospace Corporation employees to determine their auto main-
tenance habits, vehicle attributes, and willingness to loan their vehicles at
no cost to the program (see Figure 2-2). These results indicated sufficient
vehicles were available to form a test fleet of 250 cars reflecting national
population characteristics in terms of make, model and year. The average
indicated tune-up maintenance interval for the fleet was slightly in excess
of 12 months.
Vehicles were acquired from two large industrial sites -TRW Systems
and Aerospace Corporation parking lots ~~fi the vicinity of the Scott
Research laboratory (SRl) San Bernardino test facility using the following
procedure:
.
Vehicles were selected randomly from the inventory of
available vehicles as reflected by the questionnaire.
.
A reserved area within the parking lot was allocated
for those vehicles to be tested within a given day.
.
The vehicle owner was notified three days before by
mail and with a follow-up call the day before his vehicle
was tested to deliver his vehicle to the designated area.
.
SRl drivers retrieved and returned vehicles from the
reserved parking lot.
Two hundred and forty-four vehicles were acquired from which 227 valid
tests were obtained. The distribution of the acquired vehicles relative
to the US population is shown in Figure 2-3 and a detailed matrix in
Table 2-9. The population matching is good across model and model years
with the exception of American Motors Corp. (AMC) vehicles. Since AMC
vehicles are a small percent of the sample, this deviation will not
significantly affect the statistical results. The 1966-67 and 1968-69
vehicle groups are well matched to the popu1ation_j(i.~., 50% in each
group).
2-18
-------
Name I
I I I I I I I I I I I I I I I
LAST NAME. LEAVE ONE SPACE, THEN FIRST INITIAL.
Employee number I
I
18
Do not u-rite in boxes containing decimal points
Automobile make. I
2'
year (' 66 or later, please) L-.L.....I
.8
model. I
87
Engine displacement (if knou:n): cubic incbes I I
4.
I . I
borsepou-er I I
8.
I . I
number of cylinders: (cbeck one) eigbt L.....J
87
six L.....J
88
four L.....J
8.
Transmission: automatic L.....J
80
manual L.....J
81
Air conditioning: yes L.....J
82
no L.....J
88
When purcbased: year L-.L.....I
84
montb L-.L.....I
88
wbat state' I
68
How many years longer do you expect to keep tbis automobile? I
I . I
Present speedometer mileage I
I . I
Average mileage driven per year I I
17
I . I
Number of miles per day spent commuting I
I . I
28
Average number of miles per day car is driven on tbe freeway I
82
I . I
Principal driver: male L.....J
38
female L.....J
3.
married L.....J
40
sl'ngle L.....J
41
age: 16 - 25 L.....J
43
26 - 40 L.....J
44
41 - 65 L.....J
..
Qver 65 L.....J
48
Wbat brand name of gasoline do you normally purcbase? I I
. 48
regular L.....J
81
premium L-J
62
Wbo normally tunes up your car (spark plugs, distributor, carburetor)?
auto dealer L.....J
64
service station L.....J
68
diagnostic station L.....J
6.
yourself L-J
87
How often? less tban 6000 miles L-J
68
6000 - 12000 miles L-J
89
more tban 12000 miles L.....J
70
Are you willing to participate in tbis type of program witb tbe car described above?
yes L.....J
72
no L-J
7'
After you bave finished. please fold the questionnaire so that the return address shows and deposit it in the company mail.
Figure 2-2.
Sample Questionnaire
2-19
-------
Test Vehicle Population - 244
Distribution, Percent 0 10 20 30 40 50
American Motors
U. S. Population -
Test Sample .
Chrysler
U. S. Population
Test Sample
Ford
U. S. Population
Test Sample
General Motors
U. S. Population
Test Sample
Foreign
U. S. Population
Test Sample -
Figure 2 -3.
Test Vehicle Population Compared with
U. S. Vehicl-e Population
2-20
-------
Table 2-9. Engine Parameter Maladjustment Survey Test
Vehicle Population
Population
Number of Vehicles Distribution%
Make 66 67 68 69 70 Total Sample u. S.
American Motors
Corporation
Rambler 1 1 1 0 0 3 1.2 3.6
3 D 3. 6
Chrysler
Corporation
Chrysler 1 1 1 5 0 8 3.3 2.4
Dodge 8 4 6 7 0 25 10. 2 6. 5
PI ym outh 6 1 4 1 0 12 4. 9 7. 7
15 "6 11 13 0 45 18. 4 16. 6
Ford Motor
Company
Ford 21 10 9 15 3 58 23.8 19. 7
Line oln 1 1 0 0 0 2 0.8 O. 0
Mercury 6 1 0 2 0 9 3. 7 3. 6
28 IT "9 17 3 69 28. 3 23. 3
General Motors
Corporation
Buick 5 3 4 3 0 15 6. 2 6. 5
Cadillac 1 2 1 0 0 4 1.6 2. 0
Chevrolet 15 9 13 11 0 48 19. 7 23. 0
Oldsmobile 6 4 7 1 0 18 7.4 6. 1
Pontiac 9 5 4 4 1 23 9. 4 9. 3
36 TI 29 19 1" 108 44. 3 46.9
Foreign
Volkswagen 2 0 4 3 0 9 3. 7 5. 6
Others 1 0 4' 5 0 10 4. 1 4. 0
"3 0 8' 8' 0 19 7:8 9:6
Grand Total 83 42 58 57 4 244 100. 0 100. 0
2-21
-------
2.4 VEHICLE INSPECTION PROCEDURE
2.4.1
Test Station Arrangement
The parameter inspection was conducted at two test stations. Idle
% CO and float level were measured at a second test site for accessibility
to exhaust analysis equipment. The inspection procedure (see Figure 2-4)
provided maximum test efficiency for a two-man crew with quick access to
all test equipment. Engine warmup was assured as all test vehicles were
driven approximately five miles from the industrial parking lot to the
Scott Research Laboratories' test facility.
2.4.2 Measurement Procedure and Test Equipment
The procedures and equipment used for the parameter inspection were
based on accepted shop procedures. Commonly available equipment was used
wherever possible as it was impractical to attempt the survey on a more
sophisticated level. Specifications used were obtained primarily from
manufacturers' tuneup guides and shop manuals (see Figure 2-5). Certain
information was not always available, however, which made it necessary
to establish some specifications or criteria for purposes of quantifying
inspection results. These specifications were used when neither manu-
facturers' or general shop manuals provided complete information. The
actual procedures used are described in Table 2-10 and refer to the
individual items on the inspection sheet of Table 2-8. In addition to
the procedures detailed in Table 2-10, observations were made of diagnoses
based on ignition scope patterns.
Complete specifications were not available for required spark plug
voltage, available coil voltage, performance diagnosis, idle % CO, air
cleaner restriction, PCV system performance at idle, and air pump perfor-
mance at idle. For these inspection items, specifications were derived
from whatever literature was available. Specifications for spark plug
required voltage and coil available voltage were derived from the ignition
scope operating manual. The PCV system and air pump performance specifica-
tions were derived from Automobile Smog Control Manual, by Harold T. L. Glenn.
The AC air cleaner tester indicated that a reading of 90 to 135 degrees was
the upper range for an acceptable air cleaner element. The upper limit for
2-22
-------
N
I
N
W
Pick up . Install vehicle on Connect ignition
vehicle &- ... chassis dyno-inspect ... scope-perform
deliver .. PCV system, air cleaneI .. electrical inspection,
to Scott t'abs element, heat riser inspect idle RPM &
valve & air pump vacuum leaks .
"
Make performance ... Deliver vehicle to Measure
... ...
"... inspection "... 2nd test station- ... float
measure idle % CO level
;
Figure 2-4.
Parameter Inspection Procedure Sequence
-------
m7
196B DODGE-PlYMOUTH-C~IRVSLER
ENGINE 383 C... In.
..-- AT~e ~
10
563
7~ ,4 TOeD
2, I 8
.~8T~
1-8-4 3-6-!>-7.2
Z..o L""
.060"..210'"
H
T urquo...
0383
LB383HP
10.~1
1»165
25
Firing O,der
Hyd. Lihen
O,y Lah
5th DiClit Code
E ngln. Cokw
Serin
Type
Compo A.tio
Comp. PSI
lma.. v...)
BATTERY 1211 NIg. Grd.
S..nd,1,d 24MB59
Spec. Equip. 27MB70A
59AH
70AH
Cranking Vol,.ga Imin)
S,.".r 0.- (cranking'
9.5v
180-2OGA 1.....1
N
I
N
~
CHARGING SYSTEM
Output (min)
Op.:ro,ing VOl lOge GI 700F
lme_red" blttoryl
S-
26~A
13. 7-15.211
WI Ait CoNS.
36A
13.7-15.211
..Je!L
APM
M~nu.' TranI. 6SO Ai' Cond. oH
Aw.O. Tron.. (loll 650 Air Cond. off
Unsnap ~II joint connac:tiOft ., KC'."'" bell c'8ftk
bIIlo.. letting-Adj..11 '0 1IIIt. _II".
~ 10....,
A' Idle 28.33 .
Varl..110n 2 Idilt. Vie. lin. dilCOnMCtld1
IGNITION TIMING 100\1<"')
0111. V... L,n. OiICOn_'ed
Menu.' Tr.n.. OC TOC . Idl. RPM
A..,o. Tran..' 5 8TOC . Itli. RPM
IGNITION ADVANCE 1~.2IiCIORPM)
Cont. . V... C8IIt. Only
44-54 26-30
41-51n1. 23-2IS~
Menu.1 TronL
A"to. TC'8ftI,.
IDLE MIXTURE
Ai, F lIel Ratio
FUEL PUMP
PrOlL
Vol.
V...
3
\/8
HP
383 CU. IN. ENGINE
W/4 BBL. CARB.
14.~1Imin'
3\\.5 PSI ... 500 RPM
1 pt. 30 _. II 500 RP"
10".500 APM 1..81 line di.._ted at....
Fil'.. Idi--. .I_t tVll8lln line - ...81......
end c..b. R""". every - yun 01 24.000 mila
Fill., 11...1 hlnkl PlMtic lilter on - 01 ~ "'....
IGNITION COIL I"". 7oo.8OOf'1
ChrYII...Prn'oIit. 2444242
Pro. n... 1.6~1.79
S". R... 94()() 11700
Tn' Sf! Line 8
Ch,ysler.E..e. 2444241
Pro. R... 1.41.1.55
Sec. Rn 92()() 10100
Tn' So, Lin. 8
BaU.t Rni"or
cn,ysl.. 2095501 0.S.0..
IGNITION CURRENT
ERJino SloPPed 3.OA
Idling 1.8"
HI TENSION WIRE RESISTANCE loIWNI
Coil cobl. in'''1 Coi'her primary I_inelef C818-
con,..t 01 dIll. c." - 25.000 _.
Coil c.ble 'emoved-IS.ooo mea.
SPork Plug cobl.. -30.000 -..
SECONDARY RESISTANCE UMia.
CRANKING CIRCUIT RESISTANCE
'''''''.ted Circuit
Bet. POS. POll '0 Sol. Be,. T......-.Io
IPos. Cobl. only- .211'
1501. SW. only- . hI
IE..h connec,iot>- .0.1
GrOUnd Circuit
S'..'or Housing 1o 8.'. Nog. POII-.2v
(Grd. coblo only- .2111
IEng. grd. circ"il- .101
IE.." c-,ion- .0.\
STARTER FREE RUNNING CURRENT DR.W
lOA mea. ., 1 10 RPM 192!to 2600 ......
SOLENOID CURRENT DRAW
Pull in Coil 14..16.OA . lor
HoIcI in ColI 11..,2.III.Ior
FORM NO. 6~017
SUN EUCTlIC CORPORAnON-cHICAGO, IlliNOIS 60631
0... I1odrio c---. 'III
SPARK PLUGS
Chlmpion
M-
Gap
T orQU.
CARBURETOR
Man".1 T,ana.
Auto. T,.nL
DISTRIBUTOR
Chryslor
ROtation
SQring Tension 10>'
Gap Cinchl
0-11 10........
V..;,,'ion (Degreesl
Cond- Copocily
J11Y or J1O'f
P.3-4P
.035"
30 FtlLbL
OIrysler.2863841
ChrysI...2883801
Con...A VS44265
c:.t...AV~lI
M.nu.. T,.nl.
2857356
CC
17.20
.01..019
28-33
2
,
A"to. Tr.nL .
2857358
CC
17-20
.014-.019
2W3
2
.2!t-.285MF0
MECHANICAL ADVANCE
2857358 2857358
OiSl. RPM Oitt. Dig. Oist. RPM DilL Oeg.
375-625 0 37!to625 0
525 0.6% 625 N"
800 11~.13" 760 '0.12
2500 17.1& 2500 ,....,."
VACUUM AOVANCE
0iIt. Del. In. V...
o 507"
...9 12
.12 I."
In.Vae.
.7"
12
I."
Figure 2-5.
Oilt. Del.
o
1-8
.12
Typical Tuneup Guide
-------
Table 2-10.
Inspection Procedures
*
Item Parameter
2.1 Basic Timing
Procedure
The vacuum advance and retard lines
were disconnected and the timing
was read on the engine pointer with
a timing 1 ight.
2.2 Total Advance
Vacuum lines were reconnected and
the engine speed held at 2500 rpm.
The timing light was adjusted to
synchronize with TDC, the ignition
scope console indicated the total
advance, and the basic timing was
subtracted from the meter reading.
The vacuum lines to the distributor
were disconnected and the procedure
for 2.2 was repeated.
2.3 Mechanical Advance
2.4 Vacuum Advance
The observed mechanical advance was
subtracted from the observed total
advance.
2.5 Coil Available Voltage
The average firing voltage at 1500 rpm,
in neutral, as observed on the ignition
scope, was recorded.
2.6 Plug Required Voltage
One spark plug wire was disconnected
and the firing voltage for that
cylinder at 1500 rpm was recorded.
The average dwell was read on the
console dwell meter.
2.7 Point Dwell
2.8 Dynamic Compression
3.1
30 mph Performance
Diagnosis
By means of the ignition scope console,
each spark plug was individually shorted
at 1500 rpm and the console indicated
the percent reduction in rpm.
The chassis dynamometer was set so that
the vehicle speed was 30 mph at 16 in.Hg
manifold vacuum. The exhaust %CO
concentration was read, the average spark
plug firing voltage was observed, and
the number of spark plugs misfiring and
frequency of misfire were determined as
a percentage of total required firings.
*
Keyed to Table 2-8.
2-25
-------
Table 2-10.
Inspection Procedures (Cont.)
3.2 45 mph Diagnosis
The vehicle was operated at 45 mph at
8 in. Hg mani fo 1 d vacuum and the
observations from 3.1 were repeated.
The vehicle was operated at 60 mph,
wide open throttle. The observations
from 3.1, except % CO, were repeated.
3.3 60 mph WOT Performance
Diagnosis
4.1
Idle Speed
The idle rpm was observed on the
ignition scope with the transmission
either in neutral or drive, as
specified by the manufacturer.
The idle % CO was recorded.
4.2 Idle % CO
4.3 Manifold Vacuum
The manifold vacuum at idle was read
with a vacuum gauge.
The air cleaner filter element was
installed on an AC air cleaner tester
and the relative pressure drop was
observed on a 0 to 180 degree scale. .
4.4 Air Cleaner, Angle
4.5 Float Level
The float level was measured using
the procedure specified in the shop
manua 1 .
4.6 Heat Riser Valve
If there was a heat riser valve it was
checked for sticking.
5. 1
PCV Performance at Idle
The crankcase vacuum was measured at
the dipstick tube using a vacuum
gauge reading in inches of water.
The dead head pressure of one of the
air pump outlet hoses was read in
inches of mercury.
5.2 Air Pump Performance
5.3 Vacuum Leaks
The individual vacuum hoses were closed
off to check for leaks in the system
(see 5.4).
5.4 Idle Speed
The idle rpm was recorded to determine
the magnitude of the leakage compared
to the idle rpm in 4.1.
2-26
-------
idle % CO readings was 3.5%.
Any amount of misfire was unacceptable.
The limit for acceptable variation of dynamic compression was 2%
between all cylinders. This was the recommended tolerance given in the
ignition scope operating manual.
The equipment required for the parameter survey consisted of con-
ventional diagnostic equipment, an electric chassis dynamometer, and
two nondispersive infrared carbon monoxide analyzers. The diagnostic
equipment included an Autoscan 4000 Ignition Analyzer, a 10-15 in. Hg
pressure gauge, a 0-5 in. H20 vacu~m g~uge, and an AC air cleaner tester.
The AC air cleaner tester was a type used in the service trade and indi-
cates a relative degree of restriction. With this type of tester (see
Figure 2-6), air flow is induced through the filter element and the
pressure drop created is indicated on an arbitrary scale of increasing
restriction. The air flow is produced from regulated shop air through
an aspirator underneath the filter element. The tester is calibrated
by adjusting the shop air flow to produce a calibration setting using
a fixed opening to the aspirator system. For this program, a protractor
was placed over the pointer to facilitate reading the amount of air
cleaner restriction. Figure 2-7 shows actual aspirator system vacuum
data plotted against the indicator angular position.
The methods and equipment used in the engine parameter survey
provided reliable evaluation of the components and adjustments. While
measurement accuracy was not of laboratory quality, the errors were low
enough to provide a pass/fail evaluation with a high level of confidence.
2-27
-------
Air
Outlet
Indicator
900
I FAI R "
I /
I /
/
I /R EPLACE .
1800
Filter
Air Aspirator System
Figure 2-6.
AC Air Cleaner Tester
.15
.
=
c
I
~
o .10
i
~
i
:)
:)
u
C
>05
a.
11/
~
1/1
)-
1/1
r I t
60 100 . 140 180
INDICATOR ANGLE. DEGREES
o 20
Figure 2-7.
Air Cleaner Tester System
Vacuum Versus Indicator
Angle
2-28
Shop
Air
-------
3.
PARAMETER SCREENING EXPERIMENT
An experiment to screen the most frequent and extensive malfunc-
tioned engine parameters was performed. This screening was based on
the emission response of representative power trains to simulated
engi ne parameter malfunctions. The diagnostic 'power of emi ssions
measured under varying engine loads was also investigated. The
following subsections discuss the results and experimental procedure
applied to develop these data.
3.1
TEST RESULTS AND CONCLUSIONS
The test results and conclusions are presented for two experiments
performed to screen engine parameters;
. Orthogonal test evaluation of eleven parameters
. Choke parameter experiment
3. 1 . 1
Screening Experiment
The screening experiment, as previously noted, had as its objective
the identification of the most effective procedures and sensitive emis-
sion diagnostic modes to be studied in the more definitive experiment.
A complete summary of the statistically reduced data for the three
hot cycle emissions (HC, CO and NO) is presented in Tables 3-1 through
3-6. The vehicle descriptions are keyed through the vehicle number and
presented in Table 3-7. The summary data are comprised of the emissions
change, 6E, that results from a change in parameter level, 6P. The units
of ~E for composite (volumetric) emissions is parts per million (ppm) and
for mass emissions, grams per mile. These data were abstracted from a
more complete data tabulation which may be found in Appendix A. In addi-
tion, an effectiveness index, EI, was developed for each of the parameters
and is used as a figure of merit for selecting those parameters to be carried
into the definitive experiment. This index is essentially the emissions re-
ductions which could be affected when all of the vehicles in the general
population have a specific parameter, (Pi)' restored to its nominal or
specification value. The mathematical definition of the index is:
3-1
-------
w
,
N
Tab 1 e 3- 1.
Summary of Orthogonal Test Data,HC Composite
Emission Sensitivity to Malfunction
(90% Confidence Level and Above)
Screening Experiment
r . +7 Misfire+2.5% 1CO+20% I RPM+50 0 Leak+lcfm
Vehicle lmlng-l0 0 -0.5% -100 Vacuum Adv. spec 0
No.
~E E1 ~E E1 ~E E1 ~E E1 ~E E1 ~E E1
SOlA 137 7.3 486 66.3 -158 -16E -146 6.5 --- --- --- ---
502 246 13.1 472 64.2 -137 -204 --- --- --- --- 177 12.2
503 --- --- 454 61. 7 --- --- --- --- --- --- 143 9.8
504A 221 11.7 335 95.7 --- --- -77 4.6 --- --- --- ---
505A 103 5.5 314 83.5 --- --- --- --- --- --- --- ---
506 116 6. 1 303 86.8 64 95 -140 8.4 107 -1.8 -72 -3.3
507 246 13.0 273 72.7 --- --- -212 12.7 --- --- 132 8.5
508 180 9.6 296 78.7 -97 -145 -71 4.29 107 -1.59 --- ---
509A 280 14.8 264 70.3 --- --- -256 15.4 58 -1.02 --- ---
510 309 16.3 221 58.6 -288 -429 -217 13.0 268 -5.63 247 13.2
511 --- --- 189 53.9 --- --- --- --- --- --- --- ---
X,ppm 206
x = mean value of response, all statistically significant vehicles
A = air injection system
6E = change in emission in going from the low to high value of the parameter, ppm
EI = (Pi-Pspec) 6E/6Pi
-------
Table 3-2.
HC Mass Emission Sensitivity to Malfunction
(90% Confidence Level and Above)
Screening Experiment
Change in Emission Level, 6E, grams per mile
(N .
I
(N
T. . +7 Vac Advspec * Misfire265% I C02. 0 50 F1oat+1/4 A/C180 PCVP 1 ug Ai rp1 ug Leak1gfm
Vehicle lmlng-10 pt Res.. IRP~100
o -0.5 -1/8 0 spec spec
SOlA --- --- 1. 18 4.44 -2.03 --- --- 2.33 --- --- ---
502 1.60 -0.67 --- 4.97 - 1. 30 --- 1.33 --- --- --- 1.62
503 -1. 76 ** --- --- 4.91 --- --- --- --- 1.63 --- ---
504A 2.05 --- --- 3.70 --- -1.33 --- --- --- 1. 24 ---
505A 0.66 --- --- 3.15 --- --- --- --- --- --- ---
506 -1.07** --- --- 3.27 --- -1.16. --- --- --- --- ---
507 1.39 --- --- 3.36 --- -1.62 --- --- --- --- 2.36
508 --- --- --- 2.91 --- --- --- --- --- --- ---
509A 1. 22 --- --- 2.32 --- -1. 92 --- 0.48 0.54 0.87 ---
510 2.69 1.09 --- 2.28 -2.84 -2.96 1.80 --- 2.30 --- ---
511 --- --- 0.27 1. 74 --- -1.03 -0.51 --- --- --- ---
- - -
- g'!1s ~ QJD Q£) QJD
X ml
*
Pt. resistance originally expressed as Q impedance, after Vehicle No. 505 expressed as:
Available Primary Kv x 100
Required Secondary Kv
** -
Not considered in X
see notes, Table 3-1
-------
Table 3-3. ~ummary of Orthogonal Test Data
CO Hot Composite
(90% Confidence level and Above)
Screening Experiment
(,.)
I
~
Vehicle I CO, + 2% , - O. 5 % Float,+1/4,-1/8 A/C, 180, 0 PCV, Plug Spec ~~~l~~C~ leak,l cfm, 0
No. ~E % EI % ~E % EI % ~E % EI % ~E % EI % ~E % EI % ~E % EI %
50l. .575 .616 -.827 -.044 0.368 .081 --- --- .554 .053 -.648 -.015
502 .428 .637 - . 561 -.030 1.145 .252 .473 .056 N/A N/A --- ---
503 .286 .426 -.623 -.033 1.741 .383 1 .325 .158 N/A N/A -.468 -.032
504 .434 .646 -.300 - . 01 6 --- --- 0.240 .029 .290 .028 -.188 -.013
505 .249 .371 -.286 -.015 --- --- 0.209 .025 .291 .028 - . 129 -.027
506 --- --- -.825 -.044 0.604 .143 0.31 .037 N/A N/A --- ---
507 .623 .927 -.522 -.028 0.268 .070 0.383 .046 N/A N/A -. 146 -.009
508 .333 .495 -.824 -.044 0.507 .122 0.315 .037 N/A N/A --- ---
509 .293 .436 -.444 -.024 --- --- 0.706 .084 .508 .049 --- ---
510 .565 .841 - . 321 -.017 --- --- 0.476 .057 N/A N/A --- ---
511 .504 .750 --- --- N/A N/A --- --- N/A N/A --- ---
X,% .429 G -.523 -.028 0.772 G 0.493 G .411 8 -.363 -.017
See notes, Table 3-1.
-------
Table 3-4. Summary of Orthogonal Test Data
CO Mass Emission Sensitivity to Malfunction
(90% Confidence Level and Above)
Screening Experiment
Change in Emission Level, 6E, grams per mile
(.oJ
I
U'1
Vehicle No. Timing Vac Adv pt Res Misfire lCO lRPM Float A/C PCV Air Leak
Pump
SOlA -16.7 --- --- --- 16.9 --- -34.6 18.0 --- 28.9 ---
502 -17.2 --- 3.9 --- 19.0 --- -11.6 25.7 8.8 --- ---
503 -69.4 --- --- --- --- --- -23.0 72.5 49.6 --- -22.9
504A - 21 .8 --- --- 8.6 17. 1 --- --- 8.3 13.6 10.5 ---
SO SA - 20 . 0 --- 4.3 --- --- 11 . 1 - 7.5 --- 9.9 14.4 ---
506 -83.2 - 27 . 1 --- --- --- --- -32.6 37.3 --- --- ---
507 --- --- --- --- 15.0 --- -13.2 --- --- --- ---
508 -29.4 - 6.5 --- --- --- --- -15.6 18.8 --- --- ---
S09A -18.1 --- --- --- 10.4 11.3 --- --- 19.9 16.7 ---
510 --- --- --- --- 15.8 --- --- --- 13.6 --- ---
511 --- --- --- 6.6 --- --- - 7.5 --- --- --- ---
- - - -
X, .9!. -34.5 6) -18.2 8 8 8
m;
See notes, Table 3-2.
-------
Table 3-5. Summary of Orthogonal Test Data
NO Composite Emission Sensitivity to Malfunction
Screening Experiment
UJ
I
0'1
Timi ng Vacuum Advance Float PCV
Vehicle No. 6E EI 6E EI 6E EI 6E EI
--
505 889 47.1 243 -6.5 197 10.5 --- ---
506 255 13.5 236 -3.9 151 8.06 --- ---
507 524 27.8 130 -3.3 104 '5.6 -190 -22.7
508 564 29.9 530 -7.9 178 9.5 -171 -20.4
509A 412 21.8 120 -2. 1 154 8.2 -247 -29.4
510 591 30.2 677 -14.3 128 6.9 -161 -19.2
X, ppm 536 ~ 322 -6.33 152 Q}:) -192 -22.9
See notes, Table 3-1.
-------
Table 3-6. Summary of Orthogonal Test Data
NOX Mass Emission Sensitivity to Malfunction
(90% Confidence level and Above)
Screening Experiment
Change in Emission leve1, 6E, grams per mile
tA)
I
......
Vehicle No. Timi ng Vac Adv pt Res Mi s fi re ICO IRPM Float AIC PCV Air leak
SOlA 2.16 --- --- --- --- --- loll --- -0.75 --- 1.96
502 2.88 --- 0.51 --- --- --- --- -1.14 --- --- ---
503 1. 51 --- --- 0.41 --- --- 0.41 -1.34 -0.72 --- ---
504A 2.63 --- --- --- --- --- --- --- --- -0.39 ---
505A 3.70 --- -0.48 0.80 --- 0.86 --- --- --- -1.08 ---
506 1.85 --- --- --- --- --- 0.76 -0.78 --- --- ---
507 2.98 --- --- --- --- 0.68 --- --- --- --- ---
508 2.01 0.69 --- --- --- --- 0.44 -0.52 -0.61 --- ---
509A 1.31 --- --- --- --- --- 0.71 --- -0.93 --- ---
510 2.43 1.22 --- --- --- --- --- --- -0.76 --- ---
511 1.14 --- -0.07 --- --- 0.18 0.15 --- --- --- ---
- -
X 2.23 0.59 -0.75
See notes, Table 3-2.
-------
W
I
CD
Table 3-7.
Orthogonal Test Program
Power Train Attributes
Cold Start Composite Emissions
As Received Tuned
Vehicle No. Make Model-Yr. CID CARB. Mileage HC CO HC CO
501 Chev 66 Impala 327A Roch 4 67880 331 1.93 258 1.14
502 Ford 68 Custom 302 Ford 2 33361 268 1.90 290 1.53
503 Dodge 68 Charger 318 Ca rt 2 36815 304 0.69 303 1. 21
*
504 Merc 67 Cougar 289A Ford 2 24216 2299 1.83 354 2.26
505 Ford 66 Mustang 200A Ford 1 18008 456 1. 76 318 1.35
506 01ds 68 98 455 Roch 4 45760 314 1. 74 452 2.76
507 Chev 68 Ma1ibu 327 Roch 4 19128 463 3.56 219 0.99
508 Dodge 66 Po1ara 383 Stro 2 62801 333 1.56 345 1.49
509 Chev 69 Nova 230A Roch 1 13452 284 1. 75 201 1.47
510 Ford 68 Ranchero 390 Ford 2 18006 354 1.45 283 1.21
511 VW 68 1500 92 SOL X 1 50340 1373 * 6.94 374 2.18
*
Misfiring vehicle.
-------
EI.. =
lJ
00 de.
L P (Pi) dP ~ dP i
1
de.
(Pi-Pi spec) dP~
1
::
where:
EI. .
lJ
P(Pi)
= effectiveness index for parameter Pi
= probability of finding parameter, Pi'
at level dPi+Pi (from engine parameter
survey)
de/dPi
= change in emission/change in parameter level
(from orthogonal screening experiment)
Only those values of ~E and EI which were found to be statistically
significant at the 90% confidence level are shown. The average value of
~E and El are shown at the bottom of the respective parameter columns.
HC Emissions Response (Tables 3-1 and 3-2)
The El values indicate that the most effective parameters to maintain
for HC emission control are misfire, idle timing, leak rate and idle rpm.
These values are highlighted by circles, positive values indicating emission
decreases. As stated, the data indicated that it might be effective to
maintain manifold leak rate, however, only five vehicles appeared to be respon-
sive to the malfunction. The response of vehicle 510 to this malfunction is
particularly spurious, the ~E value being 2.0 times larger than the average
value. The HC emission sensitivity to the most, influential engine
parameters ranged between~50% of the mean of all statistically significant
vehicle responses. The most anomalous results are the responses of HC to
idle CO, particularly for vehicle 510. Response values of 20% of those actually
measured would normally be anticipated. (see References 1 and 2). A partial
explanation of this is the fact that, in reality, HC.s usually respond in a very
highly nonlinear fashion to idle CO changes. As this experiment was run at
two levels of the parameters, the HC emission response is overestimated.
Vehicle HC emissions were not found to be very sensitive to increased point
resistance(available to required voltage), air cleaner clogging, carburetor
float level, or positive crankcase ventilation gross restrictions.
3-9
-------
The He mass emissions substantiate these general conclusions and the
general ranking of parameter sensitivity, Table 3-2. Both leak rate and
idle CO effects on He mass emission are substantially reduced relative to
their equivalent composite values, as are the number of statistically sig-
nificant occurrences.
CO Emissions Response (Tables 3-3 and 3-4)
The most effective engine parameters to be maintained to control CO
emissions as reflected by the value of the effectiveness index are idle
CO, air cleaner, pev, and air injection with the latter two significantly
reduced based on their comp~site values. beth manifold leak rate and float;
level have low, negative influences on emission levels. The basic response
to float level is indicated to be negative relative to an increasing fuel
level in the bowl. That is, the additional head caused by fuel level
results in fuel enriched carburetion and, hence, increased (negative changes)
CO. Retarded timing was found to have a moderate to substantial increasing
influence in CO composite and mass emissions, respectively, contrary to its
effect on He and NO. None of the ignition related parameters were found to
affect CO.
NO Emissions Response (Tables 3-5 and 3-6)
The most significant parameters affecting NO are basic timing and float
level. Those malfunctions which tend to enrich carburetion effected moderate
reductions in NO, presumably because of peak combustion temperature reduc-
tions. A disconnected vacuum advance essentially retards the spark under
moderate acceleration and cruise conditions and results in a decrease in NO
emissions. This decrease varied considerably from vehicle to vehicle for
composite values; decrements ranged from !20% of the average vehicle value.
Surprisingly, vacuum advance had a negligible influence on mass weighted
emissions. None of the other ignition or induction related parameters
appeared to have a significant impact on NO emissions.
3-10
-------
In summary, the following conclusions resulted:
. The most effective parameters (as measured by the effectiveness
index) to maintain in an enforced program are the idle adjust-
ments of F/A ratio, timing, and rpm; the secondary ignition
system (spark plugs and secondary wiring), and the induction
system components of air cleaner, PCV, and air injection
system on vehicles so equipped.
Genera1iy, both composite and mass emissions respond equiva-
lently to the engine parameters evaluated.
.
. Float level and vacuum advance have strong effects on CO and
NO emissions, respectively, although their malfunction fre-
quencies are such that they would not have a significant
impact on an inspection/maintenance program.
Emission Diagnostic Signatures
Thirty-three diagnostic modes were concurrently evaluated in the
orthogonal experiment. Emission responses were obtained which are similar
to those for the composite and mass emissions. Modes were sought which
are significantly more selective to one class of malfunction (i.e., engine
subsystem) or to specific components or adjustments.
Typical emission responses to the range of engine malfunctions simu-
lated in the experiment are shown in Tables 3-8 through 3-14. Modes are
grouped by degree of engine loading (closed throttle, low load, medium
load, and high load). It should be noted that certain malfunctions such
as PCV, air pump, and air cleaner restrictions were set at their extremes,
nominal and inoperative, while the field they will be in varying states.
of deterioration. Therefore, emission responses to these parameters would
generally be less than those shown.
HC emission response to parameter levels are presented in Tables 3-8
through 3-10 for typical test vehicles. Again, only those emission re-
sponses indicated to be at the 90% confidence level or above are shown. As
with the composite emissions, misfire and timing have the strongest effects
3-11
-------
on HC in the various modes. For vehicles equipped with an air injection
system, malfunctioning of this system also significantly affects mode
emissions. All parameters but misfire tend to have diminishing effects
on HC emissions with increasing load. This would suggest that a high1y
loaded mode would be least confounded for making a misfire diagnosis. It
was also shown in the parameter survey study, Section 2, that increas-
ing load resulted in more frequent misfiring. It is significant to note
the strong influence of timing through all of the modes.
Typical CO mode emission responses to parameters are shown in Tables
3-11 through 3-14. The lower loaded modes generally reveal idle CO malad-
justments. These may be confounded by other induction system related
malfunctions, particularly air injection system, pcy and severely clogged
air cleaners. Air cleaner influences generally become progressively
stronger with load, thus suggesting that a highly loaded mode CO emission
measurement would be most selective in diagnosing this malfunction. This
is due to the progressively higher pressure drop across the element giving
rise to sub-atmospheric pressures within the carburetor intake. Internally
vented carburetor bowls are suspected of responding to this pressure,
effectively raising the fuel level in a manner similar to a float maladjust-
ment. There is also the possibility that the reduced air flow affects car-
buretor fuel metering but this does not explain why some vehicles fail to
respond to air cleaner blockage.
In order to assess whether parts of an acceleration mode have more
diagnostic content than others, several of these modes were segregated
into three equal parts. A typical result is shown in Table 3-15 for
the 2 and 3 mph/sec, 0-60 mph modes. The two parameters most sensitive
to increasing acceleration load are basic timing and air cleaner re-
striction. There appears to be no substantial difference between modes
using different rates of acceleration.
Fast idle modes, 1500 and 2500 rpm. were also evaluated. These
results showed no significant diagnostic content not already in the
idle mode. There is an indication that CO emissions may be more sen-
sitive to air injection system malfunctions in this mode.
3-12
-------
The above discussion may be summarized as follows:
.
Within the loaded mode group, all modes tend to have the same
response to malfunction.
.
The low and moderately loaded modes for the CO emmisions are
sensitive to induction system type malfunctions, such as float
setting, air leak, air cleaner, and PCV, although basic timing
can sometimes confound the diagnosis.
.
For air injection controlled vehicles. a malfunction of this
device can confound a CO diagnosis for idle adjustment in the
closed throttle modes or induction system malfunctions at load.
The fast idle modes (1500 and 2500 rpm) are not substantially
different from the closed throttle modes in diagnostic power.
.
.
Closed throttle HC emissions will generally diagnose malad-
justed idle rpm and timing, although an air injection system
malfunction on vehicles so equipped will confound the diagnosis.
.
Basic timing dramatically affects NO emissions under any
loaded condition and is the only diagnostic mode selective
to a specific malfunction.
Intermittent misfire greater than 2.5% of total firings will
predominate over most other malfunctions simulated in all
modes for HC emissions.
.
3-13
-------
Table 3-8.
Diagnostic Mode HC Emissions Sensitivity
to Malfunction, Vehicle 506
Screening Experiment
Change in Emission 6E, ppm .
W
I
......
Mode* Misfire Timi ng Leak RPM ICO
Id1e(F) 81 181 --- -165 ---
Id1e(1) --- 238 --- -301 ---
Idle(11) 112 159 --- -159 106
Idle (27) --- 147 99 -188 106
30-15(F) 168 141 --- -249 250
50-20 (F) --- --- --- -319 253
30F . (F) 174 43 --- --- ---
24C (28) 181 42 --- --- 20
30A (19) 188 46 --- --- ---
30S (32) 194 38 --- --- ---
+2 mphjsec (13: 191 53 --- --- ---
40C (29) 187 --- --- -21 ---
455 (31) 196 --- --- --- ---
0-25 (F) 152 89 --- -59 ---
1 5- 30 (F) 176 56 --- --- 22
0- 30A (18) 163 80 --- -75 60
605 (30) 217 69 --- --- ---
GOP {25} 181 --- --- --- ---
+3 mphjsec (7) 187 64 --- --- ---
~
* Numbers in parentheses correspond to mode type and location
shown in Figure 3-7, "F" designates Federal 7 mode.
-------
(.otJ
I
-'
U'1
Table 3-9.
Diagnostic Mode HC Sensitivity to Ma1f~nction, Vehicle 508
Screening Experiment
Change in Emission 6E, ppm
Mode Misfire Timi ng leak RPM ICO Float A/C PCV
Idle F 298 154 --- --- --- --- --- ---
I d1 e 11 392 237 --- -145 168. --- --- ---
Idle 27 324 128 --- -56 79. --- --- ---
30- 15 --- 386 --- --- 389. -409 --- ---
50-20 311 363. --- -780 --- --- 228 294
30F 313 71 --- --- --.., -70 40 ---
24C 308 64 --- --- --- -72 35 36 .
30A 335 100 --- -53. 46. -75 86 ---
+2 mph/sec 379 148 --- -88 52. --- 67 ---
40C 321 46 --- --- --- -47 51 ---
0-25 309 110 --- --- --- -36 --- ---
1 5- 30 301 94 --- --- --- -48 --- 27
0-30A 369 138 --- -91 81. -66 92 ---
60S 359 119 --- --- --- -30 37 ---
60P 328 77 --- --- 33.0214 -44 73 ---
+3 mph/sec 332 95 --- --- --- -27 33 28
-------
W
I
.....
0'1
Table 3-10.
Diagnostic Mode HC Sensitivity to Malfunction, Vehicle 509
Change in Emission 6E, ppm
Mode Misfire A/C ICO Fl oa t RPM Ai r Inj Leak Ti mi ng PCV
Idle F 233.07 --- --- --- -235.80 160.00 --- 214.82 ---
I dl e 1 233.60 --- --- --- -162.96 --- --- 1 77 . 70 ---
Id1 e 11 233.30 --- --- --- -255.47 162.56 --- 236.67 ---
Idle 27 220.99 --- --- --- -188. 15 140.06 --- 191. 39 ---
50-20 --- --- --- --- -1605.44 653.31 --- 935.79 ---
60-20 773.68 --- --- --- -1402.89 712.98 302.57 --- -413.57
2'm/s
30F 247.02 --- --- -49.93 -55.85 115.69 --- 87.29 ---
24C 245.51 --- --- --- --- 91.40 --- 1 22.05 ---
30A 247.53 --- --- -62.08 -73.74 100.05 --- 91 .03 ---
+2 mph/sec 257.08 --- --- --- -53.21 79.96 --- 114.67 ---
40C 257.37 --- --- --- --- 11 7. 34 --- 127. 71 ---
0-25 237.03 --- --- --- -124.11 151 .62 --- 159.09 ---
1 5- 30 243.56 --- --- --- -52.61 113.74 --- 136.24 60.33
0-30A 240.45 --- --- --- -158.15 1 23.91 --- 126.72 ---
60S 230.77 --- --- --- -39.88 82.60 --- 93.48 ---
60P --- --- --- --- --- --- --- --- ---
+3 mph/sec 287. 71 --- --- --- --- --- --- --- ---
-------
Table 3-11.
Diagnostic Mode CO Emission Sensitivity to Malfunction, Vehicle 503
Change in Emission t.E, 1.
W
I
....
'oJ
Mode leak ICO Float AIC PCV
Idle F --- 2.11 --- --- 0.62
Idle 1 --- 1.49 --- 0.71 0.63
I dl e 11 --- 2.21 --- --- 0.98
Idle 27 --- 2.17 --- 0.52 0.84
30-15 --- 2.00 --- 0.83 0.58
50-20 --- 2.12 --- -0.37 0.67
30F --- --- --- 2.65 1.66
24C --- --- --- 1. 70 1.25
30A --- --- --- 2.08 ---
30S --- --- --- 2.17 ---
+2 mph/sec -0.85 --- --- 3.58 1.09
40C --- --- --- 3.01 1. 76
45S --- --- --- 2.97 ---
0-25 -0.61 --- -0.76 1.90 1. 57
15- 30 -0.86 --- 2.31 1. 71 -0.86
0-30A -0.48 0.46 -0.60 1. 76 1. 34
60S --- --- --- 3.06 ---
60P --- --- --- 3.51 2.05
+3 mph/sec --- --- --- 4.08 ---
-------
eN
I
.....
00
Table 3-12.
Diagnostic Mode CO Sensitivity to Malfunction, Vehicle 508
Screening Experiment
Change in Emission fiE, %
Mode Mi s fi re Timi ng Leak RPM ICO Float A/C PCV
Idle F --- --- --- --- 2.29 --- 0.55 ---
Idle 11 --- --- --- --- 2.88 --- --- ---
Idle 27 --- --- --- 0.38 2.75 --- 0.57 ---
30-15 --- --- --- --- 1. 59 --- 0.53 -0.42
50-20 --- --- --- --- 1.77 --- 0.75 ---
30F --- -0.61 --- --- --- --- --- ---
24C --- -0.93 --- --- --- -0.99 --- 0.68
30A --- --- --- --- --- -0.63 --- ---
+2 mph/sec --- -1.37 --- --- --- -0.51 0.65 ---
40C --- -2.78 --- --- --- --- 1.47 ---
0-25 --- -0.81 --- --- --- -0.94 0.58 ---
1 5- 30 --- -0.89 --- --- --- -0.98 --- 0.55
0-30A --- -0.79 --- --- --- --- --- ---
60S --- --- --- --- --- --- 0.73 ---
60P --- --- --- --- 0.72 --- 1. 16 0.98
+3 mph/sec --- -0.97 --- --- --- -0.78 0.84 ---
-------
Table 3-13.
Diagnostic Mode CO Emission Sensitivity to Malfunction, Vehicle 509
Change in Emission.~E, %
W
I
.....
1.0
Mode Misfire AIC ICO Float RPM Air Inj leak Timing PCV
Idle F --- --- 1.49 --- --- 1.40 --- --- ---
Idle 1 --- --- --- --- --- --- --- --- ---
Idle 11 --- --- 1. 34 --- --- 1.20 --- --- ---
Idle 27 --- --- --- --- --- --- --- --- ---
50-20 --- --- 1.60 --- --- 1.54 --- --- ---
60-20-2 m/s --- --- 0.87 --- --- 1.56 --- --- ---
30F --- --- --- --- --- --- --- --- ---
24C --- --- --- -0.53 --- --- --- -0.81 0.88
30A --- --- --- -0.59 --- 0.47 --- --- 0.34
+2 mphjsec --- --- --- -0.98 0.72 0.83 --- -1 .67 0.86
40C --- 1.06 --- --- --- 1.54 --- -1.50 ---
0-25 --- --- --- --- --- --- --- --- ---
1 5- 30 --- --- --- --- --- --- --- --- ---
0-30A --- --- --- -0.46 0.43 0.82 --- -0.51 0.89
60S --- --- --- --- --- --- --- --- ---
60P --- --- --- --- --- 0.18 --- --- ---
+3 mph/sec --- 1.20 --- --- --- 1.59 --- --- ---
-------
(AI
I
N
o
Table 3-14. Diagnos~ic Mode CO Sensitivity to Malfunction,. Vehicle 510
Change in Emission 6E, %
Mode Mi s fi re Ti mi ng Leak RPM ICO Float AIC PCV
Idle F --- --- --- --- 2.28 --- --- ---
Idle 1 -0.367 --- --- --- 2.25 --- --- ---
Idle 11 --- --- --- --- 2.65 --- --- ---
Idle 27 --- --- --- --- --- --- --- ---
30-15 --- --- --- --- 2.26 --- --- ---
50-20 --- --- -0.253 --- 2.13 --- --- -0.29
30F --- --- --- --- --- -0.37 --- 0.39
24C --- --- --- --- --- -0.38 --- 0.51
30A --- --- --- --- --- -0.40 --- 0.45
30S --- --- --- --- --- --- --- ---
+2 mph/sec --- -1.56 --- --- --- --- 1.05 ---
40C --- --- --- --- --- -0.33 --- 0.51
45S --- --- --- --- --- --- --- ---
0-25F --- --- --- --- 0.52 0.37 --- 0.59
15-30F --- --- --- --- 0.30 -0.49 --- 0.67
0-30A --- --- --- --- 0.41 -0.37 --- 0.55
60S --- --- --- --- --- -0.65 2.17 0.70
60P --- 0.29 --- --- --- -0.36 --- 0.55
+3 mph/sec --- --- --- --- --- --- 2.16 0.75
-0.25 --- 0.30
-0.32 --- 0.56
-------
Table 3-15.
Sensitivity of Parts of Acceleration Modes to Malfunctions, Vehicle 503
Change in Emission, ~E
W
I
N
HC Misfire CO Timi ng CO AI C CO PCV
(ppm) (%) (%) (%)
L = 0-20 mph 212 -0.46 0.88 0.77
+3 mph/sec M = 20-40 mph 260 -0.95 1. 74 0.63
H = 40-60 mph 251 -0.83 2.03 ---
L = 0-20 mph 213 -0.40 0.77 0.85
+2 mph/sec M = 20-40 mph 258 -0.92 1. 37 0.74
H = 40-60 mph 272 - 1 .31 1. 79 0.54
-------
3.1.2 Choke Parameter Experiment
Two commonly malfunctioning and maladjusted choke parameters, choke
blade setting and heat riser valve, were investigated as a subexperiment.
The complete Federal emissions test is required for this determination as
the choke is only on below approximately a 1000F block temperature. The
Federal emissions of these tests were statistically evaluated and the results
are presented in Table 3-16 for both Federal composite and mass emissions.
Slopes with "F" ratios indicating a confidence level of 90% or greater are
starred. The results indicate that:
. There is a wide variability in the data as indicated by the
low statistical significance on the majority of the tests.
. Mass emissions are more highly correlated to these malfunctions
than composites.
. General trends are correct, closing the choke results in in-
creasing (positive) HC and CO emissions on 80% of the vehicles;
failed open heat riser valves result in increasing (negative)
HC and CO emissions on 70% of the vehicles.
. The influence of these adjustments varies considerably from
car to car.
.
Heat riser valve position appears to have the predominant
effect on composite emissions while choke blade has the pre-
dominant effect on mass emissions.
Some of the difficulties in the data may be seen in the plots of
emissions versus choke blade setting for the two levels of heat riser
valve position, Figures 3-1 through 3-6. The emissions responses at both
levels of the heat riser valve are seen to be quite different, indica-
ting substantial interactions between parameters may be occurring. The
rich (most closed) choke blade setting is extremely anomalous; slopes
at this point at the two levels of heat riser valve position having
opposite slope signs.
The average effects of making choke adjustments were estimated by
averaging the full data set, contrary to the screening test where only
statistically significant data were used. These results seem to indicate
the most cost-effective choke maintenance would be to free heat riser
valves. The effectiveness index for this treatment yields values of 10 ppm
and 0.05% for HC and CO, respectively.
3-22
-------
Table 3-16.
Emission Response to Choke Parameters
HC
Composite Emissions Change Mass Emissions Change
Heat Riser2 Choke B1ade1 Heat Riser Choke Blade
F Ratio Slope F Ratio Slope F Ratio Slope F Ratio Slope
501 5.94 -47.3 0.46 -08.8 0.24 -0.60 0.72 0.56
502 0.21 6.7 0.95 12.3 0.68 -0.50 2.80 0.85
503 1. 27 -303.0 1.20 236.3 0.97 -2.54 0.88 2.09
* * *
504 11.13 58.0 0.68 12.3 11.43 1.43 15.14 1.31
505 0.01 3.3 1.34 41.0 1.03 0.69 3.46 -0.46
*
506 1.09 -132.7 0.25 14.3 6.13 -1.84 1.55 0.03
507 0.96 26.3 1. 19 24.5. 0.02 0.06 2.03 0.59
*
508 0.54 -190.7 0.79 7.0 6.07 -0.36 6.23 -0.12
* *
509 7.42 -100.0 4.06 41.5 9.93 - 1 . 83 1.30 0.05
*
510 1.45 102.7 2.26 -70.8 6.94 2.09 1.07 -0.39
511 1.98 - 96.3 0.17 -23.0 1.08 -0.54 0.65 0.06
X - 41. 3 > 26.0 -0.35 < 0.42
CO
501 1.09 - 0.82 0.01 0.04 0..47 -18.17 O. 10 2.25
* *
502 3.05 - 0.14 15.70 0.28 2.01 4.50 42.02 16.60
503 1. 59 - 0.43 2.28 0.45 0.09 - 5.50 1. 56 . 14.83
* *
504 1.62 0.23 6.58 0.37 1. 55 10.70 8.61 21.28
*
505 0.78 - O. 21 0.68 -0. 16 0.40 - 0.60 48.69 -2.10
*
506 3.99 - 0.21 4.93 0.05 5.97 - 11. 37 0.16 -0. 1 0
507 .0.04 - 0.03 2.37 0.19 0.50 - 8.80 0.78 8.93
508 1. 91 0.27 0.94 0.10 0.73 1 5. 70 1. 11 -0.03
509 0.12 0.07 0.09 0.05 0.44 - 3.57 0.04 0.63
510 0.81 - 0.15 2.71 0.10 3.69 12.07 3.80 10.45
511 1.86 - 1.09 1. 57 - 0.11 1.03 - 11 . 63 1.41 1.10
X - 0.23 > 0.12 - 1.51 < 6.68
lChoke blade slope based on increasing closure.
2Heat riser slope is difference between nominal and fail
*
Statistically significant at the 90% confidence level.
open condition.
3-23
-------
400
350
~
a..
a..
..
U
:I:
300
250
2.5
2.0
~
o
o
u
1.5
1.0
-0.30
VEHICLE EMISSION RESPONSE TO CHOKE
ADJUSTMENT AND MALFUNCTIONS
VEHICLE 504
COLD START TESTS
HEAT RISER
NOMINAL
- - FAILED OPEN
""".
,-"
,,"
'-,
""......
.....
..",.
Figure 3-1. HC Emission Response
",
",
---
- ......
./'
./'
/
-0 . 20
-0.10 0
CHOKE BLADE SETTING, INCH
. 0.10
-'0.20
Figure 3-2.
CO Emission Response
3-24
-------
350
300
~
Q.
Q.
..
U
:I:
250
200
2.0
1.5
*
6
u
1.0
0.5
-0.3
VEHICLE EMISSION RESPONSE TO CHOKE
ADJUSTMENT AND MALFUNCTIONS
VEHICLE 507
COLD START TESTS
- NOMINAL
- - FAILED OPEN
-
F;gure 3-3. HC Em;ss;on Response
-tl.2
-0.1 0
CHOKE BLADE SETTING, INCH
0.2
0.1
F; gure 3-4.
CO Em1ss;9n Response
3-25
-------
450
400
~
Q.
Q.
.. .350
u
:J:
300
2.0
1.5
rF.
..
o
u
1.0
0.5
-0.3
VEHICLE EMISSION RESPONSE TO CHOKE
ADJUSTMENT AND MALFUNCTIONS
VEHICLE 508
COLD START TESTS
- NOMINAL
- - FAILED OPEN
.
.- _. - - .
---
/
/
./
/
"'"
-
Figure 3-5.
HC Emission Response
.--
----
-0.2
-0.1 0
CHOKE BLADE SETTI NG, INCH
0.2
0.1
Figure 3-6.
I
CO Emission Response
3-26
-------
3.2 PARAMETER AND DIAGNOSTIC MODE SELECTION CRITERIA
Parameters to be IIma1adjusted" during the screening experiment were
selected to provide a broad spectrum of engine system parameters. The
goal was to identify maladjustments which would produce measurable effects
on exhaust emissions. In determining the least number of parameters to
accomplish the overall program objectives, the following procedure was
used: first, engine parameters were selected which significantly effect
exhaust emissions; second, the parameters were grouped according to
similar types of effects on emissions. Table 3-17 lists the parameters in
11 groups of similar effects and indicates the approximate ranking of their
effects on exhaust emissions.
For the 11 main effects to be incorporated as test parameters, a
single maladjustment was identified to simulate a malfunction class which
could be adjusted over the normal range of parameter deviation. Rather
than concentrating on deviating single components or adjustments, param-
. eters were selected which best represented the gross malfunction charac-
teristics of the subsystem it affected. The adjustment limits were based
on parameter survey data, the desire to avoid extreme driveability problems,
sufficient emission sensitivity to assure that the effect is greater than
the experimental noise and ability to set the deviations in conjunction
with similar parameters.
Several diagnostic modes have been recommended for determining mal-
functioning components, adjustments, or subsystems known to influence
vehicle emissions (References 3,4,5 and 6). All combinations of driving
modes (i.e., acceleration, cruise, and d~ce1erations) have been integrated
into a cycle driveable on a chassis dynamometer. This driving cycle
contains acceleration rates ranging from 2 to 3 mph/sec, cruise ranges
from 30 to 60 mph at road load, and cruises at above normal road load
conditions (Figure 3-7). The time positions of the actual modes in thi~
driving cycle are also shown.
3-27
-------
Table 3-17.
Test Parameter Selection
Co.)
I
N
(X)
Parameters Selected Rough Rank Parameters Selected Rough Rank
1. Bas i c Timi ng (3) 7. Air Cleaner Restriction (9)
Distributor Drive (wear) (19) 8. Vacuum Advance (17)
Mechanical Advance (17~ Vacuum Advance Control (17)
Mechanical Retard (17 Vacuum Retard (17)
2. Ignition Primary Circuit (2) 9. Air Injection System (29)
Point Dwell (5) Air Pump Pressure (29)
Point Wear (2)
Condenser (25) Air Pump Hoses (29)
Primary Resistor (26) Check Valve (29)
Coi 1 Primary (21) Gulp Valve (7)
Alternator Circuit (27) 10. Intake Manifold Leaks (22)
3. Ignition Secondary Circuit (1) 11. PCV Valve (9)
Spark Plugs (1) 12. Cold Test Parameters
Plug Wires (11) a. Air Cleaner Thermostat (14)
Coi 1 Secondary (21) b. Heat Riser (6)
Distributor Cap & Rotor) (8) c. Choke Relief (10)
4. Idle Circuit Speed ~ (4) 13. Parameters Excluded
Throttle Position (decel) (4) a. Accelerator Pump (29)
5. Idle Circuit Air/Fuel ~ (7) b. Hot Idle Compensator (29)
6. Carburetor (Air/Fuel) (12) c. Carburetor Bowl Vents (30)
d. Fuel Sys tern (18)
Low Speed (off idle) Circuit (13) e. Fuel Cutoff (31)
High Speed Circuit-Part Load (12) f. Engine Deposits (27)
High Speed Circuit-Full Load (12) g. Piston Rings (16)
Enriched Point (14) h. Valves (13)
Float Level (28) i. Exhaust Back Pressure (23~
j. Cooling System Temperature (22
-------
1. NO lOAD WHEElS $TOPPED
3,000 R.P .M.
s. KEYMODE (CLAYTON)
DISCONNECT
INERTIA ,
WHEELS I
@/
I 60 M.P.H. 5. fig
'.6. SYNTHETIC
~
I
@.
~ 2. M.P.H. 10 H.P.
.w M. P. H. -30 H.P.
@
'~
45 M.P.H. \ fJ2\
10. Hg ~
30 M.P.H. \ @
16" He ~
15 M.P.H.
16"
o
50
100
150
TIME (SECONDS) .
63
Figure 3-7.
Diagnostic Mode. Emissions
Screening Experiment
3-29
-------
3.3
EXPERIMENTAL DESIGN
An experimental program was designed to study the 12 variables
(13 for the air injection system, AIR) defined in Table 3-17. Design
criteria imposed were that the experiment should:
. Reflect typical power train types within the general population.
. Develop fundamental, generalized data from which optimum engine
parameters and emission signatures can be identified.
. Maximize the amount and kind of data per dollar spent.
3.3.1 Power Train Selection
The vehicle power trains selected for testing were required to
fit approximately the fOllowing population statistical criteria:
.
.
Ratio of 8 to 6 cylinder vehicles
Ratio of air injection system to engine modification
emission control devices
Percent by manufacturer
.
A power train matrix complying with these criteria requiring
11 test vehicles was selected (Table 3-7). Five of the eight V-8
engines are middle sized in terms of cubic inch displacement (CID):
three are large V-8's in heavier vehicles. The Volkswagen 1300 was
selected to reflect the foreign made vehicles within the U.S. population.
3.3.2 Experimental Approach
As emissions tests are expensive to perform, an experimental
approach which maximizes data utility per dollar spent is essential to
a cost effective program. The analysis of the engine parameter experi-
ment data resulted in the selection of 11 parameters affecting hot
cycles and two variables affecting cold cycles. A complete characteri-
zation of these variables (main effects and first-order interactions)
would require 213 or 8192 tests if the variables are investigated at
two levels. This level of testing is obviously not practical. A
statistically designed experiment was therefore employed with the
following advantages:
3-30
-------
. More information acquired per test point
. A measure of experimental error obtained
. Interaction effects investigated economically
Since the first objective is primarily to screen
tive engine parameters and emissions diagnostic modes,
was divided into two parts:
the most attrac-
the experiment
.
Eleven parameters relatively insensitive to
engine " temperature
. Two choke parameters sensitive to engine
temperature
Scre~ning Experiment Orthogonal Tests
A 1/64 replicate of a 2 11 factorial design was selected for the
first experiment. Although this design has low statistical power, it
does satisfy the basic requirements for the experiment. The experi-
ment requires 32 emissions tests on each of the 11 IIrepresentativeli
automobiles selected for the program. This design provides for the
separate orthogonal estimation of all 11 main effects and up to 15
of the possible 55 two-factor interactions. These interaction estimates,
however, are self-confounded in groups of three or four so that to
establish one as being significant, it will be necessary to discount
the reality of the others in the particular group (Reference 7).
Also, a random testing order was followed during the experiment to
preclude the possibility of timewise biasing of the desired contrasts.
The test matrix is shown in Table 3-18.
For each measured exhaust emission mode, mass and volume
weighted, a complete analysis of variance was conducted to estimate
each of the 11 main effects and the two-factor interaction terms.
In addition to the variance analysis, a least squares response
surface equation of the following form was fitted to each data set:
1"1
Y = B + E B.X. + I B'jX.X. (for selected i,j pairs)
k 0 .111 ..11J
1= 1,J
3-31
-------
Table 3-18.
Orthogonal Test Matrix - Screening Experiment
(.0)
I
W
N
Float
Run Timi "9 Vacuum Pt. Res. Ign - Sec. Id1e CO. Id1e RPM leve1 Air C1eaner PCV Valve Air Pump Man. leak
No. DEGDEV Advance Avail. Kv - 2 5 (I 'Usfire) (I Dev.) Deviation Deviation Restriction Res triction Restriction CID (1CF")
Req'd Kv . AC Tester 3M
1 -10 Nonn Yes 2.5 2.0 Rich -201 -1/S" Rich 0 0 1001 Yes
2 -10 Nonn 110 2.5 -0.5 lean +101 -1/S" Rich 0 1001 0 Yes
3 7 Nonn Yes 0 -0.5 lean -201 -1/S" Rich 180° 1001 1001 Yes
4 -10 Nonn Yes 2.5 2.0 Rich -201 -lIS" Rich 180° 1001 0 No
5 7 0 No 2.5 2.0 Rich -201 -lIS" Rich 180° 1001 0 No
6 -10 Nom Yes 0 -0.5 lean +101 +1/4" lean 180° 1001 0 No
7 7 Ilona No 2.5 -0.5 lean -201 +1/4" lean 0 1001 1001 No
S -10 0 Yes 0 2.0 Rich +101 -1/S" Rich 0 1001 1001 No
9 7 Nonn No 2.5 ~0.5 lean -201 +1/4" lean 180° 0 0 Yes
10 7 Nonn Yes 2.5 2.0 Rich +101 ' +1/4" lean 0 0 0 No
11 7 0 110 0 -0.5 lean +101 +1/4" lean 0 0 1001 Yes
12 -10 Nonn Yes 0 -0.5 lean +101 +1/4" lean 0 0 1001 Yes
13 7 0 Yes 0 2.0 Rich -201 +1/4" lean 0 1001 0 Yes
14 7 Nonn Yes 2.5 2.0 Rich +101 +1/4" lean 180° 1001 1001 Yes
15 7 0 No 0 -0.5 lean +101 +1/4" lean 180° 1001 0 No
16 -10 0 No 2.5 2.0 Rich +101 +1/4" lean 180° 1001 1001 Yes
17 7 0 No 2.5 2.0 Rich -201 -1/8" Rich 0 0 1001 Yes
1S -10 0 No 0 -0.5 lean -201 -1/S" Rich 180° 1001 1001 Yes
19 7 Nonn No 0 2.0 Rich +101 -1/S" Rich 180° 0 0 Yes
20 7 Nonn No 0 2.0 Rich +10i: -1/8" Rich 0 1001 lOOt No
21 7 0 Yes 2.5 -0.5 lean +101 -lIS" Rich 0 1001 0 Yes
22 -10 0 No 0 -0.5 lean -201 -1/8" Rich 0 0 O' No
23 -10 0 Yes 2.5 -0.5 lean -201 + 1/4" lean 180° 0 0 Yes
24 -10 0 Yes 0 2.0 Rich +101 -1/8" Rich 180° 0 0 Yes
25 -10 Nonn No 2.5 -0.5 lean +101 -1/8" Rich 180° 0 1001 No
26 7 0 Yes 0 2.0 Rich -201 +1/4" lean 180- 0 1001 No
27 -10 0 Yes 2.5 -0.5 lean -201 +1/4" lean 0 1001 1001 No
28 7 Nonn Yes 0 -0.5 lean -201 -1/8u Rich 0 0 0 No
29 -10 0 No 2.5 2.0 Rich +101 +1/4" lean 0 0 0 No
30 -10 Nonn No 0 2.0 Rich -201 +1/4" lean 180° 0 1001 No
31 -10 Nonn No 0 2.0 Rich -201 +1/4" Lean 0 IDOl 0 Yes
.
32 7 0 Yes 2.5 -0.5 lean +101 -,/8" Rich 180° 0 1001 No
-------
where:
B. = slope coefficient
1
Yk = mode or weighted emission measurement
Xi = parameter adjustment, deviation from specification
From this equation, it is possible to generate
surface for each emission specie and to assess
of each of the independent variables, X..
1
To minimize test costs, these emissions tests were performed
for the hot cycles of the Federal procedure. An identical engine warm-
up procedure was followed for every test to control the initial tem-
perature of the engine block. This key temperature was shown in the
Vehicle Emissions Surveillance Study to significantly affect emissions.
an emission level response
the relative significance
Hot cycle testing also improves the precision of the test since the
more variable carburetor choke effects are not present. Detailed test
procedures and emissions response surfaces are discussed in the
following subsection.
Choke Tests
The 11 engine parameters selected for study in the orthogonally
designed experiment are not substantially influenced by power train
temperature. A subexperiment on the 11 vehicles was performed to
investigate the carburetor choke subsystem. This experiment evaluated
the effect of choke linkage maladjustments on vehicle emissions as
measured by the full, cold start procedure (i.e., starting with a
power train at ambient conditions). Both heat riser valve and the
choke relief (blade angle setting) were evaluated at two and three
levels of their parameter setting, respectively. This was done using
a full factorial design requiring six emissions tests. All other
adjustments and failure simulations were returned to manufacturers'
recommended states except for the choke"adjustments.
3-33
-------
Base Emissions
To establish limits on the maximum emissions decrements to be
affected through maintenance, the 11 vehicles selected for the
statistically designed experiment underwent two additional tests. These
were a complete diagnosis of the general engine maintenance state and a
Federal emissions test for both the vehicle as received and after a
major engine tuneup. The maintenance effectiveness in reducing
emissions is simply the difference in the two emissions measurements.
3.4
TEST PROCEDURE
3.4.1 Vehicle Preparation
Before a test vehicle was run on the parameter screening experi-
ment, it received an oil and filter change, then was tested for exhaust
emissions from a cold start and with the diagnostic modes. The vehicle e
was then given a thorough diagnostic inspection (based on the parameter
survey procedure), a major tuneup, and repaired as indicated by the
diagnostic inspection. The cold start and diagnostic mode exhaust
emission tests were then repeated. This procedure provided information
on the initial and post tuneup state of the vehicle as to both exhaust
emissions and engine parameters. Also, the data were used to expand
the parameter survey and evaluate the effectiveness of the inspection
procedure.
Inspection for the parameter screening experiment was more
detailed than the parameter survey with respect to carburetor diagnosis.
This was because % CO readings were available from the diagnostic modes
before the tuneup. In addition, time requirements for inspecting only 11
vehicles are less demanding. Also, inspection of carburetor adjust-
ments was more comprehensive. Percent CO readings were taken from
strip charts for various diagnostic modes to detect serious discrepancies
in the carburetor metering circuits.
Engi ne tuneups were performed to ensure a good base 1 i ne for the
main experiment and to. prevent undetected or incipient malfunctions from
influencing test results. The tuneup procedure was based entirely on
recommended shop procedures and specifications; there were no deviations
3-34
-------
from this criterion. By following this procedure, the test results
were more indicative of the exhaust emission reduction which could
result from field repairs and maintenance. Only malfunctions which
could be detected during inspection and corrected by engine tune-up
and normal maintenance were determined. No vehicles were rejected
from the experiment for mechanical problems.
To establish a good experimental base line, tuneups always in-
cluded new spark plugs, ignition points and condenser, PCV valve, air
cleaner element, fuel filter, and carburetor kit. Additional parts
were replaced when indicated by the diagnostic inspection. Typical parts
replaced were ignition wires, distributor cap, and rotor. Occasionally
there were insufficient data available for carburetor adjustment, par-
ticularly for idle setting procedures. Also, the technique to be used
for carburetor mechanical adjustment was not always clear. In examining
several shop data sources, we often found disagreement both on pro-
cedures and specifications. Whenever possible, the manufacturer's
shop manual was used as the primary source of information. Idle"
mixture setting data were particularly limited and usually stated in
terms of air-fuel ratio (as measured with a thermal conductivity in-
strument), best lean idle, or a prescribed rpm drop caused by mixture
screw settings. The idle %CO resulting from the idle setting procedure
was used as the null point for the idle %CO settings in the parameter
screening experiment. As in the other tuneup specifications, the
idle %CO was set using only the available shop data. No attempt was
made to use sophisticated methods for balancing the carburetor side-
to side or to obtain an optimum air-fuel mixture for minimum hydro-
carbon emissions.
This type of procedure was used to best represent the type of
work to be done in the field by competent mechanics.
3.4.2 Measurements and Special Test Equipment
Exhaust emission analyses were performed with basic,nondispersive
infrared equipment as specified in the Federal Register, June 4, 1968,
as well as with equipment for analysis of oxides of nitrogen and com-
posite dilute bag samples. The equipment was constructed so th~t b~g
3-35
-------
samples could be taken concurrently with direct concentration readings. Mass
emissions were obtained from the mass sampler system" which is effectively
a temperature controlled constant volume through-flow device. Bag samples
were collected of proportional dilute gas compositions in either a "hot bag"
or a "cold bag," and a "background bag," as shown in Figure 3-8, and were
analyzed as described below. The combined system was constructed for oper-
ator efficiency, and was capable of both complete and semiautomatic operation.
Additional details of the mass emission management technique are described
in the Federal Register.
Figure 3-8 shows a schematic of the entire emission instrument-
ation system. Several special considerations ensured data reliability.
All direct concentration samples analyzed for HC, CO, C02' NO, and N02
were returned to the total exhaust duct without transit delays through
the duct and analysis system. Because of the return of direct con-
centrated samples from the analysis system, uniform pressures in the
sample cells had to be assured. This was accomplished by maintaining
regulated flow through all instruments during all operational modes.
If either sample or calibration gas were not flowing through an instru-
ment, nitrogen was employed to maintain the correct total flow and
sample cell pressure.
Direct concentration readings were made of HC, CO, C02' and NO
using NDIR instrumentation. N02concentrations were analyzed using
nondispersive ultraviolet detection. Bag sample analyses were accom-
plished by flame ionization detection for total hydrocarbon, low range
CO (0-10%) ,low range NO (0-500 ppm), and normal range N02 (0-200 ppm).
The cold bag was collected during 7-mode cycles one through four, and
the hot bag was filled during cycles five through nine. The background
bag was filled over all nine cycles to sample the actual dilution air
for the constant volume system.
To produce controlled ignition misfire and point resistance,it
was necessary to design a misfire generator. The point signal was
periodically shorted to induce misfire, and a variable resistance was
put in series with the primary ignition circuit. The misfire generator
was constructed to provide several percentages of misfire from 5 to 2%.
A 2.5% rate was finally selected and the misfire simulator set to
3-36
-------
BACK-
GROUND
AIR
FIL TER
TOTAL
EXHAUST SAMPLE
BACK-
GROUND COLD
HC/CO/C02
(EXHAUST)
Figure 3-8.
MASS SAMPLER
SYSTEM
CONCENTRATE
SAMPLE
RETURN LINE
HOT
(NO/N02 EXHAUST)
BAG SAMPLE LINE
HC/CO/C02
SYSTEM
CALIBRATION GAS
Flow Schematic for Simultaneous
Composite and Mass Sampling
3-37
SAMPLE BAGS
NO/N02 SYSTEM
-------
produce one misfire in every 39 firings (2.56%). The odd number of
firings was used to ensure that misfirings would be equally distributed
among all spark plugs in the engine and no one cylinder would be unduly
deteriorated. Figure 3-9 shows a schematic of the misfire generator and
vehicle installation circuits. Development problems with the misfire
simulator were caused by radio frequency interferencet both internal (caused
by misfires themselves) and external (caused by normal spark ignition).
Improved radio interference shielding and grounding were required to
eliminate these problems. The misfire generator operated properly under
all engine speeds and loads encountered during the test series and
without affecting ignition timing.
3.4.3 Orthogonal Test Procedures
The main, or orthogonal, experiment of this program consisted
of systematically detuning (adjusting to an out-of-specification
condition) the 11 representative engine parameters/settings as
specified in the test matrix. Nine Federal emission cycles (7-mode)
were run from a hot ~tart and the exhaust emissions continuously moni- .
tored and bag samples measured as described in section 3.4.2. Imme-
diately after the 7-mode cycles and bag analyses, six short, or
diagnostic, cycles were run and direct concentration exhaust emissions
measured. The engine was then set to the next detune state and the
emission tests repeated until all 32 orthogonal tests had been completed
on that vehicle.
The following sequence and adjustment procedures were used to
set the mistuned parameters for each orthogonal test:
1. The carburetor float level was set either 1/8 inch high
(rich) or 1/4 inch low (lean).
2. On vehicles with engine modification emission control,
the hose from the PCV valve was either normally open
or completely blocked off. On vehicles equipped with
air injection systems, the PCV valve system was not
adjusted, but the air pump was allowed to operate normally
or the air distribution hoses were disconnected and the
distribution manifolds sealed.
3. A calibrated orifice was installed in a centrally located
vacuum accessory hose when manifold leakage was called for,
and the normal installation was used for the standard con-
dition. Figure 3-10 illustrates a typical installation.
3-38
-------
-
-
VEHICLE
DISTRIBUTOR
POINTS
BATTERY
Figure 3-9.
COUNTER
RESET
GATE
DECODER
POINT
SIGNAL
VEHICLE
IGNITION
VEHICLE
COIL
PRIMARY
IGNITION POINT
RESISTANCE
SIMULATOR
Ignition Misfire Generator
3-39
~
-
-------
TO POWER BRAKE
BOOSTER
Figure 3-10.
Vacuum Leak Rate Simulator
3-40
CALIBRATED
SHARP-EDGED
ORIFICE
TO VACUUM
OPERATED
ACCESSORIES
-------
4.
Either a new air filter element or an element taped partially
closed (with five equal area, vertical openings) produced the
desired restriction.
5.
The ignition misfire generator was turned either off or on as
specified.
6.
The ignition point resistance switch was turned off or on as
specified.
The distributor vacuum advance hose was either connected or
disconnected as specified. If it was disconnected, the hose
was plugged to prevent leaks.
7.
8.
The ignition timing was set to the
manufacturer's specification. The
temporarily disconnected if it was
test.
desired deviation from
vacuum advance hose was
to be connected for this
9.
The idle rpm was set to the desired deviation from manufactur-
er's specification.
10.
The idle %CO was set to the desired deviation from the refer-
ence %CO determi ned after engi ne tuneup. The i dl e rpm was
maintained during the setting, and the procedure described in
Section 3.4.5 was used to maintain engine temperature consistency.
One of the 11 vehicles in the experiment was a 1968 Volkswagen which
required some test design modifications before successfully undergoing the
experiment. It was found that under mistuned conditions, the vehicle's
speed and acceleration capabilities were inadequate for both 7-mode and
short-cycle tests. Thus the vehicle speed was consistently reduced to
45 mph for the 7-mode tests from the normally specified 50 mph. For the
short cycles the following changes were made:
1.
2.
0-60 mph quick acceleration - changed to 0-33 mph.
0-60 mph acceleration - changed to 0-45 mph.
3.
4.
ACID cycle 30 cruise - third gear instead of fourth gear.
Performance Cycle 30 cruise - third gear instead of fourth,
30-60 mph wide open throttle acceleration was 30-40 mph, and
60 mph cruise was reduced to 45 mph cruise.
5.
Keymode Cycle loading was changed from 10 road horse power
at 24 mph to 2 hp. The fourth gear 40 mph cruise at 30
road horse power was changed to third gear 40 mph at 5 road
ho rs e power.
3-41
-------
6.
For the Synthetic Cycle the 60 mph cruise was eliminated
and the e.ngine manifold vacuum settings at 45, 30 and
15 mph were changed to 9, 14 and 16 inches Hg,respectively,
in second gear.
Certain other modifications to the Volkswagen were also required with
respect to the parameter maladjustments. The vacuum advance maladjustment
was eliminated from the test because the vehicle had only a vacuum controlled
spark advance. The PCV system consisted of a single crankcase-to.-air cleaner
hose and when a test called for a plugged PCV system, the hose was removed
and the air cleaner fitting plugged. The air cleaner was of the oil bath
type so the air cleaner restriction parameter was deleted from the experi-
ment. The spark timing adjustment for retard was reduced from 10 to 5
degrees. The float level was not adjustable at the float and was changed
by varying the thickness of the shim gasket under the needle seat until the
effective float setting was 0.043 inch rich and 0.179 inch lean.
3.4.4 Cold Engine Parameter Tests
A subexperiment, run concurrently with the orthogonal experiment,
involved maladjusting engine parameters that influenced cold engine exhaust
emissions. While running each vehicle on the test schedule and at the end
of each work shift, engine parameters were returned to normal specifications
and the choke relief (vacuum kick, vacuum break, pull-down) and heat riser
valve, or air cleaner hot air inlet door control, varied according to test
requirements. The vehicle was allowed to soak overnight and a cold start
test was run the next day with the same procedures used for the main experi-
ment, including short cycles. Table 3-19 shows the six combinations of choke
pull-down and heat riser settings tested. Only one vehicle (Car 505, 1966
Ford Mustang with 200-cubic inch displacement six cylinder engine) had
neither heat riser valve nor hot air inlet door.
3.4.5 Test and Measurement Quality Control
Efforts were made to ensure that the condition of every vehicle was
as consistent as possible for each test. The idle rpm and % CO were
always measured and set with a warm engine. During the actual idle setting,
the engine was never idled for prolonged periods of time, and a 0 to 50 to 0
3-42
-------
Table 3-19. Choke Relief and Heat Riser Test Schedule
Test No. Choke Relief Deviation Heat Riser Setting
A 0 Operating
B + 1/16" (Ri ch) Operating
C -1/8" (lean) Operating
*
D 0 Stuck Open
*
E + 1/16" (Rich) Stuck Open
(lean) *
F -1/8" Stuck Open
*
On cars equipped with air inlet temperature control, secure
air mixing door to receive underhood air only (no exhaust
heat) .
3-43
-------
clean out was made intermittently with the desired idle and %CO being con-
firmed immediately after clean out. The instrument system was then spanned
and set at zero with nitrogen. Two mid-point calibration gases were recorded
as a curve check for the CO, C02' and HC instruments.
The composite bag samples were analyzed immediately after completion
of the 7-mode cycles and the low-range instrument calibration. Any undue
delay in taking these readings was considered unacceptable (because of
excessive thermal conversion of NO and HC in the sample bags) and the run
repeated. After each test all sample bags were purged with nitrogen to
eliminate residuals. The nitrogen purge gas was left in the bags overnight
as an extra precaution.
To assure compliance with the orthogonal test design, an Engine
Parameter Deviation Check list was used to provide the mechanic with the
desired adjustment for each experiment run. Before a vehicle was started
on the main experiment, the "desired" settings were listed for each run
by the project engineer. These settings were based on the manufacturer's
specification and the designed experimental deviation. The test driver
filled in the columns labled "Achieved" as the mechanic made the setting.
This provided a permanent record of that parameter setting and served as a
check that the setting was actually made or could not be made.
3.5 DATA REDUCTION AND ANALYSIS PROCEDURE
Data for all of the above tests were manually reduced. Templates were
fabricated to facilitate the manual reading of the emissions strip chart
time traces. The Federal and diagnostic mode emissions and the pre- and
post-instrument calibration check data were read and the data were converted
to punched cards. Mode-emissions averaging was accomplished by approximately
balancing areas on both sides of the mode emission mean level. The 1966
Federal data reduction procedures were used (i.e., the dilution factor was
calculated from the expression, 15/(6 HC + CO + C02)'
A computerized data reduction program was written to convert these
data to concentration measurements, both mode, cycle, and composite.
3-44
-------
The instrument calibration curve was fitted with the law of optics govern-
ing its operation, Reference B. Typical computer program data output for
the full Federal emissions test and the diagnostic modes are shown in
Table 3-20.
An optional subroutine to the above described data reduction program
was written to perform the statistical calculations pertinent to the inter-
pretation of the data. The following data were calculated:
.
Emissions response surfaces for mass, composite and mode
emissions
.
Analysis of variances for emission response statistical
significance
. Analysis of variances for postulated engine parameter
interactions
3-45
-------
Table 3-20.
Reduced Test Data Printout
,if...'" (f'" ""/I"'''' ,'7"'/1'\ Awl e.
,if 't;Jj - fC ~ f:71J ~
6Ql UZN156
565 8 CHEVROLET WAGON lq61 54111 321 ROCH 4 ~ 1
2
CO
8. OnCjQOOOQE-C 3
12-04-69
W
I
~
0"1
++++ ENGINEERING UNITS ++++
. .. '~. . ,
., -""-,.~"""~""""",,.,.,,.--,.,.,
,. . ''''~~~''''. ,',-""",,,,,'""v' ',,,..-M', ''''>~~'''''''''_'_''~-''¥' ,. ,.
, "-""""'.v.w.,,_...~.- ",.._-_.~,. ,.~.' ,,- ..,.,
""""-""'~"'~'--'~' "..-~..,,,~-....~-,-- "~'~"'~"~"'''--'''''''''W'-' ~"". ",-.,....,. .'-~" '. "-'''''''''~''~'~''-' '~'~'-~ --- .'-' ..
. ",,""".""-'O"-"""'-"<""""'~
"'. ~".""" ""0"'-"" "
'. ""...-"",~"__,,,,,_-,-,,,,,,-,,,,_,-,_~,,,,,,,,,,,'..V'',..
-------
Table 3-20.
Reduced Test Data Printout (Cant.)
S "5... '11'" 1=1l¥~jJiWq')~
."'~.- KJ, " f;~ f., ,"~d}'j;).. .'
,.-,.. w .. ...,-, ,¥";:,$"~ '"'" '<'" .
HC
327.281
4.153
0.)00
W
I
.J::>
-....J
MEANS PER MODE
HOT CYCLE ONLY
'-. ~ ',--,,~,"-'. 'N'-""""'__'_"",-"",,,,,-,,,,,,",""""'''''~',,,,"',,,,,,_,,,,<,",',»-~,,,,,-,,,,,,,,,,,,,,v-,~'=--.,--,,,,.~'.,,.,'v..;..H_'''-''~'''-'''''-'.~'b~.,,..'.H,''~.~,",~~',,,,,_w_,,",,,,,,,,,,,~,,,-,,,,~,,,-,,,,~..~-,~.,.~.,,'-~'.,,,,,,,,,.-'-_~""'-""'-~'~" ,.
.. ,,""~.,~,~ ,. - ".~....,_.,...- - ."
-------
4.
DEFINITIVE EXPERIMENT
The results of the screening experiment clearly pointed to seven engine
parameters as having the most leverage in a mandatory maintenance program.
This determination was based on the frequency and extent of the malfunction
as well as the emission sensitivity to malfunction. Some of these malfunc-
tions are known to have nonlinear effects over the range of variables evaluated.
The design of the definitive experiment was directed at three areas:
. Number of levels of the variables to be studied to
characterize nonlinearities
. Test controls for minimizing residual error
. Statistical power of the experiment to characterize parameter
interactions
The results of the experimental program are given below.
4.1 TEST RESULTS AND CONCLUSIONS
Summary tables of the results of the definitive, statistically designed
parameter experiments are presented in Tables 4-1 through 4-3 for composite
and mass HC, CO, and NO emissions, respectively. These data were abstracted
from more complete data sheets in Appendix B. Data summarized are the emis-
sions responses, de/dp, for the individual parameters evaluated on 11 power
trains. Power train attributes (make, model, etc.) can be found on Table
4-14 by keying through vehicle number.
Emission response sign conventions are based on emission change in
going from low to the high value of an adjustment or from the failed to the
nominal state of a malfunction. For example, if the emission change in
going from timing retard (-1) to advance (+1) is positive, the emission
value increases, and vice versa.
4.1.1
HC Emissions Response
The HC composite emission response to timing maladjustment (Table 4-la)
is seen to be fairly consistent across power trains. Vehicles (605, 607 and 608)
4-1
-------
TABLE 4-1a
HC COMPOSITE EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine POiameters
Definitive Experiment
Engine Parameter Response, a c./ ape
Car I I
No. Timing Idle RPM PCV Air Cleaner Idle CO Air Pump
ppmjdeg ppmjrpm ppmjin.H20 ppmjdeg . ppmj%
601 10.65* -0.527* 0.05 18.6/.44 -155.0*
602 10.3* -0.383* 0.12* -1.9/13.5 -112.2*
603 8.6* -0.712* 50.4* 0.078 -0.4/26.5*
604 .9.91* -0.354* 49.3"" .0.516* -52.4/16.0
605 2.9 -0 . 672* -119.8 -0.185 -132.3/-45.3*
606 11. 9* -1. 007* 71.6* 0.147 -44.4/-5.4*
607 4.9* -0.965* - 66 . 2 0.211 -26.9/-29.07
608 5.5* -0.128* .74.9* 0.068 10.2/7.4
609 10.2* -0.589* 84.0 -0.188 39.9/-33.7
610 7.1* -0.695* -0.303 59.9/-30.1 216.91:
611 8.1* -0.665* 138.0* 0.406* 5.2/2.07
*Statistically significant at the 90 per cent confidence level.
4-2
-------
TABLE 4-lb
HC MASS EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine Parameters
Definitive Experiment
Engine Parameter Response, ae./ap.
I I
Car
No. Timi ng Idle RPM PCV Air Cleaner Idle CO Air Pump
g/m/deg g/m/rpm g/m/in.H20 g/m/deg g/m/%
601 0.062* -0.0077* 0.00175 0.85/-0.13* 3.14* .'
602 0.089* -0.0054* 0.00218 -0.13/0.19 1.02*
603 0.046* -0.0087* 1.02 -0.00069 -0.02/0.25
604 0.12* -0.0067* 1 . 09* 0.0083* -0.68/0.33*
605 0.016 -0.009 - 1. 67 -0.00177 -1. 76/-0 .49
606 0.052* -0.011* 1.56 0.00266 -0.71/0.62
607 -0.0082 -0.0038* 0.07 -0.00177* -0.046/0.123
608 0.027 -0.0021 1 .42* 0.00012 0.20/0.087
609 0.072* -0.0042* 0.27 -0.00127 0.31/-0.20
610 0.039* -0.0058* 0.00091 -0.07/0.167 3.25*
611 0.028* -0.0028* 0.36 0.00305* 0.04/-0.02
*Statistically significant at the 90 per cent confidence level.
4-3
-------
TABLE 4-2a
CO COMPOSITE EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine Parameters
Definitive Experiment
Engine Parameter Response, a e./ ap.
Car I I
No. Timi ng Idle RPM PCV Air Cleaner Idle CO Air Pump
%/deg %/rpm %/in.H£O %/de9 %CO/%
601 0.0187* 0.00001 0.00163 0.15/0.30* 1.765*
602 0.0025 -0.0031* 0.0049* -0.064/ 1. 26"7*
0.254*
603 -0.0004 0.0015* 0.727* 0 .0006] 5* 0.287/0.19*
604 0.0076 0.001 .2.507* 0.0103* -0.89/0.405
605 -0.020* 0.0018* 0.151 0.00107 0.396/0.211*
606 0.0043 0.00074* 0.938* 0.00006 0.215/0.141*
607 -0.0100* 0.00053 -0.267 - 0.00262* 0.179/0.082*
608 +0.007* 0.00048 1 .551 * 0.00208* 0.415/0.295*
609 -0.0202* 0.00095* 0.709* -0.00059* 0.088/0.197*
610 -0.012* 0.00055 0.00132* -0.153/.335* 1. 325*
611 0.0004 0.0035* 0.989 0.016* 0.321/0.7.48
*Statistically significant at the 90 per cent confidence level.
4-4
-------
TABLE 4-2b
CO MASS EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine Parameten
Definitive Experiment
Engine Parameter Response, 3 e./ 3p.
I I
Car
No. Timing Idle RPM PCV Air Cleaner Idle CO Air Pump
g/m/deg g/m/rpm g/m/in.H20 g/m/deg g/m/%
601 -0.579* 0.0018 .0.013 3.46/-5.26 22.64*
602 -0.235 0.0086 0.112* 1.5/5.67* 22.84*
603 -0.241* 0.052* 2.29 0.0085 6.23/4.73*
604 -1.18 0.030 61.24* 0.242* 0.176/16.14
605 -0 . 71* 0.042* 1.22 .0.0332* 8.06/4.21*
606 -0.294* 0.058* 16.24* -0.0095 9.11/6.84*
607 -0.31* 0.028* -3.73 .0.0536* 4.60/2.55*
608 -0.091 0.031* 27.04* 0.0522* 14.94/10.4*
609 -1. 39* 0.00088 16.89 -0.0069 17.06/.187
610 -0.841* 0.040* 0.0451* -9.78/12.14* 39.39*
611 0.0018 0.067* 9.31 0.2026* 5.56/3.05*
*Statistically significant at the 90 per cent confidence level.
4-5
-------
TABLE 4-3a
NO COMPOSIT EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine Parameters
Definitive Experiment
Engine Parameter Response, a e./ ape
Car I I
No. Timing Id1eoRPM PCV Air Cleaner Idle CO Air Pump
ppm/deg ppm/rpm ppm/in.H20 ppm/deg ppm/%
601 30.7* 0.122 -0.425 -14/-86.7* - 29
602 29.8* 0.358* -1. 248* 66/-50.5 - 181*
603 53.6* -0.510* -456* 0.538* -130/51.1*
604 20.1* 0.220 - 318* -1. 627* 13/-53.3
605 33.1* -0.426* 104 0.256 -51/44.3
606 63.9* -0.525 . 1602* 0.133 -220/225
607 39.3* 0.331 222* -0.717* -142/64.3
608 35.2* -0.07, . -711* - 0.236 ":1. 9/-87.5*
6,09 34.8* -0.141 - 240* 0.439 37/-11.3
610 31.0* 0.04 - 1. 640* 189/-105.7 70.4
611 27.1* -0.963 - 356 - 1. 044* -239/49.5
*Statistically significant at the 90 per cent confidence level.
,4-6
-------
TABLE 4-3b
NO MASS EMISSIONS RESPONSE SURFACE COEFFICIENTS
Emission Response to Engine Parameters
Definitive Experiment
Engine Parameter Response, a e./ ape
Car I I
No. Timing Idle RPM PCV Air Cleaner Idle CO Air Pump
g/m/deg g/m/rpm g/m/in.H20 g/m/deg g/m/%
601 0.121* 0.0013 -0.0014 -0.02/-0.05 0.11
602 0.066* -0.0005 -0.0041* 0.51/-0.10 - 0.75*
603 0.162* -0.0018 -2.16* -0.00053 -0.90/0.62*
604 0.091* 0.0019* -0.82* -0.0079* 0.36/-0.15*
605 0.09* -0.00009 -0. 16 -0.0019 0.10/-0.16
606 0.202* 0.0018 -2.00 0.00018 -0.16/0.013
607 0.055* -0.00027 0.22 - 0.0018* -0.21/0.15
608 0.128* -0.00056 -2.67* -0.004 0.10/-0.187
609 0.156* 0.0018 -2.04 -0.00277 0.09/0.65
610 0.096* 0.00085 -0.00205 0.46/-0.35 - 0.02
611 0.226* -0.0016 -0.20 -0.00161 -1.4/0.41
*Statistically significant at the 90 per cent confidence level.
4-7
-------
appear to be less sensitive than other vehicles. The extremely low sensi-
tivity of the 1970 vehicle (605) does not appear to have a reasonable
physical interpretation. Additional data on this vehicle are given in the
subsection on diagnostic modes. Aside from this vehicle, the remaining
power trains lie between responses of 4.9 to 11.9 ppm of HC per deqree of
timing.
Similarly with idle rpm, only vehicle 608 had an HC response
cantly less than the test population. Responses of the remaining
were between -0.35 and -1.0 ppm per rpm.
All statistically significant responses to a plugged PCV were positive,
i.e., HC emissions were increased as flow restriction (crank case pressure)
increased. The average increase was approximately 84 ppm per inch of water.
Similarly, ~nissions increases resulting from restricted air cleaner are
positive, an average value being 0.347 ppmo per degree for those vehicles
with statistically significant responses.
signifi-
vehicles
The effect of idle CO was evaluated at three levels to assess non-
linear influences. The change in emissions in going from the low-to-
nominal and nominal-to-high values of the parameter are both shown. Only
a few responses were statistically significant. Six of the vehicles had
negative responses at the low range of the idle CO setting indicating that
the manufacturer's idle F/A ratio specifications are near the minimum value
for HC emissions.
Air injection system malfunctions have a significant influence on
emissions; HC increases by 112 to 217 ppm. The effectiveness index indi-
cates that even when emissions are adjusted to the 10% failure rate indica-
ted on the parameter survey study, the reduction is a substantial 15 ppm.
The cost of repair, although large, may be cost effective because of the
high leverage in reduction (see Volume II).
Mass emissions (Table 4-lb) show the same general trends as composite
HC. There are some relative changes in parameter effects. For example,
timing has nearly twice the percentage effect on composite emissions as on
mass emissions. Idle rpm has approximately twice the percentage effect on
composite emissions relative to mass emissions. Air injection system
equipped vehicles mass emissions, 601, 602 and 610, were as sensitive to
this system failure as were composite.
4-8
-------
4.1.2 CO Emissions Response
CO composite emissions are relatively unresponsive to idle rpm and
only slightly so to basic timing (Table 4-2a). Increasing idle rpm
slightly increased composite emissions. On approximately one-half of the
vehicles, CO decreased with spark advance; 4 out of 6 at the statistically
significant emission responses decreased with spark advance.
PCV had a consistently large influence,
to 2.5% per inch of water when restricted as
significant responses.
Significant CO emission responses to air cleaner blockage were mea-
sured on eight of the vehicles. Responses to this malfunction showed the
largest range in response, 0.00006 to 0.016%, of any of the parameters
selected. This malfunction will be discussed in a subsection on air
cleaner test results.
increasing emissions by 0.71
indicated by the statistically
Idle CO effects on composite CO emissions are again shown as two
slopes connecting the three levels measured in the experiment. As demon- .
strated, the results indicate nonlinear effects as the slopes of the two
line segments usually differ by factors of two or greater. Five vehicles,
603, 605, 606, 608 and 611, deviate from this statement. The preferred
curve fit through these data was a single straight line. Three vehicles
(602, 604 and 610) have negative slopes at the low level of idle CO
setting. Two of these vehicles (602 and 610) were air injection system
vehicles from two different manufacturers. Low level negative responses
at lean idle F/A ratios for these vehicles may be due to an optimum mixture
of combustibles in the exhaust manifold.
CO response to a disconnected airpump was always substantial, 1.27 to
1.76% increases were typical.
4.1.3 NO Emissions Response
The predominant parameter affecting composite NO emissions is basic
timing although changing the fue1-to-air ratio by large deviations in in-
duction system parameters can also affect these emissions (Table 4-3a).
4-9
-------
The NO response to timing ranged between 20 and 64 ppm per degree of
crank. The timing effect is largely one of suppressing combustion gas
residence time and peak temperature. Vehicle 605, which showed signifi-
cantly less He response to timing, also had one of the lower NO responses.
All vehicles had emission responses to timing which were statistically
significant.
The next most significant and prevalent parameter affecting NO were
pev valve restruction.This influence is attributable to rich carburetion
and the attendant reduction in peak combustion temperature in the highly
loaded modes. These are precisely the modes which predominate in weighting
the composite NO emissions. If this is the case, we should anticipate that
there would be a one-for-one correlation between CO and NO emissions. It
was shown that those vehicles whose CO emission responses were statistically
significant also had NO emission responses of opposite sign which were
statistically significant. The exceptions to this statement are vehicles
603, which had relatively small, positive air cleaner restriction responses
to both CO and NO, and 608, which was extremely sensitive to CO, but not
to NO. The above trends are generally substantiated in their mass emission
responses as well. The most significant exception is vehicle 611 which had
the strongest CO mass emission response to air cleaner blockage, but only
a small non-significant NO mass emission response, Figure 4-3b. The demon-
strated inverse relationship will be used to develop a correlation between
composite CO and NO emissions for each power train using the air cleaner
experimental data (see Volume II).
On an average, timing effects
mass emissions than on composite.
HC emission responses.
had a slightly stronger influence on
The opposite was true with regard to
4.1.4 Emission Diagnostic Signatures
The diagnostic mode (signature) emission responses developed from the
definitive orthogonal experiment are presented in Tables 4-4 through 4-9
4-10
-------
Table 4-4.
HC Diagnostic Mode Emission Sensitivity to Malfunction
*
Change in Emission, AE, Low to High Parameter Level (ppm)
Vehicle 601 (AIR)
90 Percent Confidence Level and Above
Mode Timing Idle rpm Air Pump
Idle F 166. 14 -111. 42 - 2 1 4. 88
Idle C 1 65. 88 -112.91 - 244. 06
30-15F 302. 49 -369. 03 - 518.86
50-20F 491. 42 -1435. 56 - 1499. 39
~ 1500 rpm 143.49 - 94. 34
I
.....
..... 30 F 104. 28 -164.93
33 C 82. 1 7 -158. 16
0-25 F 144. 65 - 38. 38 - 1 3 5. 5 1
15-30 F 1 29. 64 - 114. 2 9
50 C 92. 05 - 40. 22
WOT 87. 34 -157. 99
Air Cleaner
Idle CO
55. 1 5
49. 99
354. 96
64. 45
C = Clayton
F = Federal
* see Table 3-1 for parameter levels set
-------
Table 4-5. CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 601 (AIR)
90 Percent Confidence Level and Above
*
Change in Emission, 6E, Low to High Paramet~r Leve 1 (%)
Mode Timing Idle rpm Air Pump Air Cleaner Idle CO
Idle F -4.310 2. 145
Idle C - O. 777 - 4. 981 2.070
30-15 F - 5. 543 1. 378
50-20 F - O. 539 0.443 - 5. 509 1. 107
1500 rpm -0. 463 - 5. 421 1. 934
~
I
~ 30 F 0.246 - 2. 1 28 - O. 3 1 8 0.408
N
33 C - 1. 439 - o. 609
0-25 F O. 351 - 1. 257 0.390
15-30F - 1. 066 - O. 268 0.364
50 C O. 373 - O. 298
WOT -1. 780 - O. 361 0.440
C = Clayton
F = Federal
* see Table 3-2 for parameter levels set
-------
Table 4- 6.
NO Diagnostic Mode Emission Sensitivity to Malfunction
x
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 rpm
~
,
-' 30 F
w
33 C
0-25 F
15- 30 F
50 C
WOT
Vehicle 601 (AIR)
90 Percent Confidence Level and Above
Change in Emission, L\E, Low to High Parameter Level (ppm)
Timing Idle rpm Air Pump Air Cleaner
Idle CO
22.80
19. 86
- 1 9. 05
57. 01
- 36. 45
182. 79
- 64. 83
- 29. 95
327. 02
-107.77
82. 04
- 148. 11
627. 58
-247.43
270. 13
509. 67
-87.86
569. 46
-181.98
1 275. 76
- 204. 04
550. 53
-181.84
117.30
C = Clayton
F = Federal
-------
Table 4-7. HC Diagnostic Mode Emis sion Sensitivity to Malfunction
Vehicle 603 (EM)
90 Percent Confidence Level and Above
Jf
Change in Emis sion, 6 E, Low to High Parameter Level (ppm)
Mode Timing Idle r~ PCV Air Cleaner Idle CO
Idle F 178 -95 67
Idle C 154 -75 73
30-15 F -431
50-20 F 431 -1604 _206 235
1500 rpm 142 -110 85
~
I
--' 30 F 111 -18
~
33 C 77 - 31 - 23 23
0-25 F 154 -46 -19 25
15- 30 F 118 -23 - 18 15
50 C 52 _15
WOT C 94
C = Clayton
F = Federal
~ see Table 3-1 for parameter levels set
-------
Table 4- 8.
CO Diagnostic Mode Emis sion Sens itivity to Malfunction
Vehicle 603 (EM)
90 Percent Confidence Level and Above
J(.
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 rpm
~
I 30 F
-'
U'1
33 C
0-25 F
15-30 F
50 C
WOT C
Change in Emission, ~E. Low to High Parameter Level (%)
Timing Idle rpm PCV Air Cleaner
Idle CO
- O. 257
2. 507
-0. 177
o. 535
- O. 179
1. 952
1. 662
O. 671
O. 295
1. 785
O. 1 53
2. 860
0.036
O. 114
-0. 192
O. 126
O. 133
-0. 439
O. 196
-0. 385
O. 41 2
O. 144
- O. 410
O. 1 72
-0. 250
O. 098
- O. 31 3
O. 11 6
O. 1 51
C = Clayton
F = Federal
* see Table 3-2 for parameter levels set
-------
Table 4-9. NO Diagnostic Mode Emis sion Sens itivity to Malfunction
x
Vehicle 603 (EM)
90 Percent Confidence Level and Above
Change in Emis sion, ~ E, Low to High Paramc:ter Level (ppm)
Mode Timing Idle rpm PCV Air Cleaner Idle CO
Idle F 37 32
Idle C 19 43
30-15 F 214 -48
50-20 F 157 -63
1 500 r pm C 17 7
~ 30 F 1098
I
~
33 C 1076 -80 276 74 - 12
0-25 F 928 202
15- 30 F 1024 -47 233 -51 - 45
50 C 1039 247 -116
WOT C 1091 164 - 106
C = Clayton
F = Federal
-------
for representative AIR and EM vehicles and in Appendix B, organized by
generic classes of power train loading. General conclusions from the
screening experiment, Section 3, are still valid. The following addi-
tional comments are significant;
. An indication of the precision of the orthogonal experiment is
exhibited by the response of Federal mode idle CO (engine
modification vehicles) to idle CO settings; values ranged be-
tween 2.15 and 2.65. The actual test value range was always 2.5%.
. The Clayton mode idle position in their cycle (i.e., last in
the sequence of key modes) usually correlated within 10% of
the Federal idle mode CO, indicating little carry-over effects
from the transition deceleration mode.
. A failed air injection system slightly i.ncreased NO on two of the
three AIR vehicles, possibly because no exhaust dilution occurred.
. Rich carburetion, as indicated by CO caused by PCV or air cleaner.
restrictions, always reduced NO in the highly loaded modes.
. The fast idle (1500 rpm) is essentially performed with the idle
circuits of the carburetor as indicated by the mode CO response
to idle CO (i.e., the fast idle response is approximately equal
to the idle CO response).
4.2 EMISSIONS STABILITY TESTS AND EXPERIMENTAL ERROR
Data summaries for the emissions stability tests are presented in
Tables 4-10 and 4-11 for composite and mass emissions, respectively. These
tests represent the chronological history of the base-line vehicle emissions
over the period of the statistically designed experiment. Therefore, they
should reflect the general noise in the statistically designed experiment
because of vehicle emission instability and procedural errors (i.e.,
correlate with the residual error of the orthogonal experiments).
The residual error, SRE' from the orthogonal test correlated well
with the standard error of the stability tests. There were four instances,
4-17
-------
Table 4-10. Emissions Stability Tests
Composite Emis sions
STABILITY TESTS (Table 4-13) - SST SRE
A B C D u
601 HC 247 264 274 328 278 35 49
CO 2.24 1. 93 2.21 2.08 2. 11 0.14 0.35
NO 908 920 784 991 900 86 117
602 HC 412 416 409 440 419 14. 1 20
CO 1. 65 1. 70 1. 59 2.42 1. 84 .389 .23
NO 1313 1315 1416 1231 1319 75.7 7H
603 HC 339 215 297 296 287 51. 9 27
CO 1. 15 0.99 0.97 0.97 1. 02 .087 . Ok
NO 1585 836 1615 1774 1453 41<; 71
604 HC 178 168 176 182 176 5.Hb 49
CO 2.21 1. 27 I. 76 1. 95 1. 7c,18 .397 O. 99
NO 607 867 974 666 779 171. 4 126
605 HC 208 194 210 662 318 229 163
CO 0.87 1. 25 1. 29 1. 22 1. 15 .194 0.28/
NO 1012 832 786 710 835 128 1 15
606 HC 355 444 407 453 418 44.5 34
CO 1. 28 0.81 0.70 1. 35 1. 04 .328 O. 16
NO 2782 4265 2886 2670 3151 798 128
607 HC 333 331 553 438 414 105.4 95
CO 3. 0 2.9 3. 2 2.7 2.95 .208 0.22
NO 1058 1294 1236 1575 1297 215 129
608 HC 250 223 222 204 225 19. 0 19
CO 1. 06 O. 76 0.77 0.80 .85 . 143 O. 17
NO 1638 2068 1640 1567 1728 229 106
609 HC 173 173 149 158 163 11. 8 77
CO 1. 33 0.46 0.56 0.37 .68 .44 O. 08
NO 800 1009 1182 1222 i 053 192 130
610 IIC 163 208 208 251 208 35.9 96
CO 0.87 0.69 O. 78 1. 06 .85 . 158 O. 26
NO 1235 1163 g95 1214 1 151 109 200
611 HC 183 231 209 214 209 19. 9 60
CO 1. 02 O. 75 0.76 0.79 .83 . 128 0.68
NO 15qg 1514 1592 1589 1574 39.g 255
SST denotes standard error from stability test.
SRE denotes residual er ror from orthogonal test.
~ denotes IlH'an of runs A, 13, C, and D.
4-18
-------
Table 4 -11. Emis sions Stability Tests
Mass Emissions
STABILITY TESTS (Table 4-13) SST SRE
A B C D u
-
601 HC 4.06 4.85 4.56 4.61 4.52 .332 0.50
CO 49. 7 63. 2 54. 1 55.6 55.6 5.62 8. 13
NO 2.55 1. 09 2.20 1.77 1. 903 .628 O. 53
602 HC 2.34 3.29 2.96 2.77 2. 84 .397 0.61
CO 22. 5 19. 3 25. 1 32.9 24.95 5.81 6. 10
NO 3.37 3.72 1.97 1. 23 2.57 1. 17 0.90
603 HC 4.61 4.32 4.75 4.50 4.55 . 182 O. 79
CO 24. 0 19. 0 19.3 19.4 20.4 2.39 3.6
NO 6. 18 4.29 6.30 5.78 5. 64 .925 O. 56
604 HC 5.05 3.79 3.90 4.79 4.38 . 631 0.57
CO 61. 1 43.5 67.3 61. 6 58.4 10.3 30.0
NO 3. 16 3.70 2.69 2.27 2.95 . 615 O. 29
605 HC 3.76 3.55 3.89 8.92 5.03 2.59 2.72
CO 18. 1 23.5 23.2 22.8 21. 9 2.55 6.05
NO 4. 13 3.27 3.54 2.90 3.46 .518 0.54
606 HC 6.39 7.98 6.77 6.81 6. 987 .688 1. 04
CO 36.5 23. 7 20. 6 20. 7 25.4 7.55 3.42
NO 9.26 9.80 11. 3 10.9 10.32 ,945 1. 46
607 HC 3. 0 2.9 3. 2 2. 7 2.95 .208 0.41
CO 39. 5 19.4 19. 1 24. 8 25. 7 9.56 4. 15
NO 1.7 1.9 2.6 1.8 2.00 .408 0.28
608 HC 4. 13 3.72 3. 71 3.28 3. 71 .347 O. 70
CO 20.6 25.3 22.5 23. 5 22.98 -1. 96 3.67
NO 6.23 7.64 5.45 6.34 6.415 .908 1. 01
609 HC 3.29 2.33 1. 63 1. 46 2. 18 .832 0.63
CO 32. 0 16.0 18.4 13.4 19.95 8.29 17. 8
NO 3.75 3.89 4. 19 4. 18 4.003 .218 1. 32
610 HC 3.55 3.54 4.82 3. 975 .601 O. 75
CO 33. 9 21. 0 27.2 26. 1 27.05 5.31 8.51
NO 3.44 3.34 3.54 4.00 3.58 .292 0.50
611 HC 1. 98 1. 63 1. 86 1. 82 1. 825 . 145 0.24
CO 13. 0 14. 9 17. 7 17.8 15.85 2.32 8.25
NO 4.21 5.64 6.66 3.44 4.99 1. 442 1. 65
SST denotes standard deviation from stability test.
S denotes residual error from orthogonal test.
RE
u denotes mean of runs A, B, C, and D.
4-19
-------
mass emissions for vehicles 604, 606, 607 and 609, where vehicles had large.
standard errors in the stability tests although they were stable after the
first test. Residual errors were nominal indicating that test IIAII may
have been anomalous. One case existed where residual error for NO mass
emissions was high even though vehicle base line stability was indicated.
There was only one instance where an extremely high residual error was not
correlated with base line stability emissions, NO composite for vehicle 611.
In general, residual errors are greater than vehicle stability error.
This is because other errors related to vehicle setup are also present.
These stem from the precision and accuracy with which the adjustments in
the orthogonal test are made.
Pooled estimates of the error for basic vehicle composite emissions
stabiTity are 27 ppm, 0.20%, and 128 ppm for HC, CO, and NO, respectively
(Reference 8). Values exceeding these numbers by a factor of two must be
considered high. On this basis, standard errors for four HC, two CO, and
one NO composite emissions stability tests must therefore be considered
high. The stability data show frequent increases of HC emissions with time,
indicating that deterioration is occurring. This deterioration may be caused
by engine deposit buildup. High average values of emissions occur because
the levels of malfunction simulated tend to bias emissions to the high side
(i.e., PCV, air cleaner cannot have negative levels of their parameter
settings, only nominal or failed).
In summary, the experimental precision appears to be good, if not a
substantial improvement, over the scteening experiment. Most error anomalies
are explained by extreme vehicle instability.
4.3 AIR CLEANER TESTS
Air cleaner test results are shown in Figures 4-1 through 4-3 plotted
against AC air cleaner tester measurement, calibrated to read in degrees.
Only those vehicles indicated by the orthogonal experiment to have respon-
ses to this parameter were evaluated. Two of the vehicles were found to
be unresponsive (Nos. 603 and 606). One vehicle, 611, had extreme composite
CO sensitivity to air cleaner blockage indicated by a factor of 4.5 increase
in CO over its nominal value of 0.8%. This sensitivity was reflected in
the mass CO emissions as well.
4-20
-------
8
~
'-
a<
~
I
w
Q
X
0
Z 6
o
~
z
0
II>
a<
«
v
4
I ]
I
I 66)4
. ...-
6 AIR ClEANER ESTS I
. ORTHOGONJ L TESTS I II
I
I
I
6 II
I
0 I
I
I I
,
I
~
I
I
i --
o I /
, /
/
~~/1
~ I
~~
-- ! ,....- 601
-- -- ~-
L.------ -- ---
0 ---- --
-
,.. ---' - I
~ ~ ~ 61'
~ ' I.
, t
i ~ ,1
:-.610
o ~ '
I I <
I
L Ie <)..
... .5 ~/ 0 I ...'
... ~~ 1 -- .0608
, ' ............,..
't-- . - 606
..................~..... .... .. ......, -- .1 ~- _""V y
~ - C
o ~ :
/ 6CY/
Q
12
10
2
20
40
Figure 4-1.
60 80 100 120
AIR CLEANER TEST ANGLE - DEGREES
140
160
180
Effect of Air Cleaner Restriction
on CO Mass Emissions
4-21
-------
..,.,~ 605
......,..... ,..,............ ...... .... ,Q.... .'. ......... ~ : .~:' ~~~'~.: .:' ,,,~'~:' '-~ --.--.- - -
--- ~~608
~V606
I V
. ---- - t-
4.
-0
0-
w
o
x
o
z
o
~
z
o
en
Ot:
<
V
---
.....
I.
o
I
, .
I . I
l> AIR CLEANER TESTS -
. ORTHOGONAL TESTS
,
I
I
I
20
60
8 100 120 140
AIR ClEANER TEST ANGLE - DEGREE~
40
Figure 4-2.
,
i
I
I
+-
I
I
I I
------- r-
--.----
611
l>
...-1
....~ """'0
...... .....,...:-::-t.::-=-..'T"-
o .
I .
I
610
9
6()Y
160
180
Effect of Air Cleaner Restriction
0" CO Composite Emissions
4-22
-------
~
I
N
W
6
I
;
I
i I
, -.
AIR CLEANER TESTS I I
A , i
I I
4 ORTHOGONAL TESTS ' i
I
i
00 ! T
, I
I 611
I
! ; 0
,
00 ! I 606
! /6g2
A }11
o ~ $ 601 /
~!.-- I L7
~ --
'"
,"'- I I
604
~ ~ ;-608 I
-
00 ~ 610-
,./ 609
.
00
,
0 I i
800
700
~
~ 5
I
z
o
co
Dii:
« 4
u
o
Dii:
o
>-
J:
3
2
o
80
180
100
120
140
160
20
40
60
Figure 4-.3.
AIR CLEANER TEST ANGLE, DEGREES
Effect of Air Cleaner Restriction on HC Composite Emissions
-------
There appears to be no correlation with engine size or number of car-
buretor venturis. For example, both the largest (400 CIO) and the smallest
(91 CIO) engines had extremely sensitive responses. The CO mass emissions
measurements correlated very well with the composite values in both sensi-
tivity and degree of linearity. Non1inearities in both composite and mass
emissions occurred much earlier for vehicle 604 than for the other vehicles
with similar manifestations.
HC Emissions Response
HC emissions responses to air cleaner restriction showed considerably
less sensitivity, Figure 4-3. For three vehicles small decreases in compos-
ite emissions were recorded. In only one instance (vehicle 611) was the
emissi?n response substantial in terms of percent increase over the nominal
(unrestricted) condition. Non1inearities did not occur in the emissions
response until air cleaner test angles greater than 140 degrees were
exceeded.
Correlation with Orthogonal Experiment
The emission level from the emissions stability tests and the emissions
response from the statistically designed experiments were combined to pro-
vide an estimate of the total emissions resulting from a restricted air
cleaner. The stability base line emission selected was that test which was
the closest in chronological time to the air cleaner tests. Where the
stability test indicated abnormal emissions variations, the mean value of
the four stability runs was taken. These data are designated by a closed
symbol in Figures 4-1 through 4-3. Agreement between the two tests
was generally good, although vehicles 606 and 610 had considerable devia-
tions in CO response and vehicles 601, 604 and 611 had large differences in
HC. Most of these deviate vehicles had extremely large variances about
their base line emissions which may partially explain these anomalies. In
general, emissions data points from the statistically designed experiments
lie on the linear portion of the air cleaner test data. The exceptions .to
this are vehicles 602 and 604 whose power trains comprise a small fraction
of the population. In addition, after the first maintenance treatment, the
tail of the parameter distribution curve, Figure 2-1, should be significantly
reduced because of periodic, systematic maintenance. Therefore, to simplify
4-24
-------
this parameter model in the economic effectiveness study, the linear respon-
ses from the statistically designed experiment were used.
The following conclusions were drawn:
. Strong, nonlinear responses occurred relatively infrequently
and then, only in vehicles comprising a low percent of the
population (i.e., air injection controlled or si~ cylinder
power trains).
.
Both composite and mass emissions show similar respopses and
nonlinearities to air cleaner restriction.
. A number of power trains, particularly Fords,
no sensitivity to this malfunction, indicating
possible to design around this influence.
showed 1 i ttle to
that it may be
.
Significant emissions data scatter occurred in two power trains
indicating possible local flow instabilities at critical pressure
sensing points in the carburetor.
Air Cleaner Difficulties and Anomalies
CO emission response to air cleaner restriction did not occur on a
large number of vehicles. This failure to respond was observed in both
the screening and definitive experiments. It was also manifested in the
initial attempts to set up the air cleaner restriction at the 50 mph road
load condition.
It was determined that the restrictions created were of small magnitude
as far as air density effects are concerned (i.e., only a few inches of
water pressure drop). Thus, the mixture effects caused by the change in
air density were insignificant. The air cleaner restriction action on
fuel/air mixture is presumed to be caused by the variation of the fuel level
because of the induced pressure in the fuel bowl. Fuel bowl pressure is
determined by the pressure signal transmitted via the bowl vents, both inter-
nal and external. When external bowl vents are incorporated, they are usually
open to ambient atmosphere at idle and off-idle modes. Thus, the majority
4-25
-------
of carburetor operating modes are dependent on the effect of the internal
bowl vents.
The conclusion is that the air cleaner restriction-created pressure
drop is transmitted to the fuel bowl through the internal bowl vents with
varying degrees of effectiveness on fuel/air in different vehicles. The
variability of fuel/air response to air cleaner-induced pressure drop is
inherent in the carburetor design. These design effects are the location
of the internal bowl vents and the fuel metering response to fuel bowl
pressure changes. The resultant fuel metering sensitivity to the design
effects vary extensively across carburetor models and, to some extent, with
major calibration differences on the same models.
4.4
EXPERIMENTAL DESIGN
The following approach was used in designing the definitive experiment.
A literature search was first performed to determine which, if any, para-
meters have nonlinear effects over the level to be tested. Those parameters
indicated to affect nonlinear emission responses were idle CO and air pump
flow restriction (Figure 4-4). Of the two, idle CO affected CO slightly
and HC significantly. With regard to air injection systems, a nearly
linear response of emissions to air flow is indicated in Reference 9. How-
ever, earlier work, Reference 10, is somewhat at variance with this statement.
Our best interpretation of the failure data is that air injection systems are
either performing nominally or are essentially flowing zero air. In this
case, testing at two levels (nominal and disconnected) will yield results
which directly apply without requiring interpolation for an actual flow rate
condition. Therefore, the two levels of air injection system selected were
its nominal and disconnected states.
In addition to the above parameters, the screening experiment indicated
that two parameters investigated in this program were nonlinear. These
were air cleaner blockage and misfire which were highly nonlinear in CO
and slightly nonlinear in HC emissions, respectively (Figure 4-5).
Misfire was eliminated from further consideration for the following
reasons:
4-26
-------
i :: -~ J~J__I_-
~ I I I !
i ,l, I I
:I: 20~ '
v.
Z
o
V>
V>
:E 10::1.
ILl
l-
V>
::>
<
:I:
X
ILl
o
---- -,
1~ 10 5 0 -5
SPARK TIMING - DEGRE'[S
-10
i~ 300 I II
ViZ
SO' I .
~ ~ 200 _-1,- --- ~ I ;Vi"J
v>U .",
:J q -....'- :~_. ~
~ a I - - -.-. -~'.-'i~
~ fico f' e'
12345678
IDLE CARBON II.ONOXIDE (%)
Figure 4-4.
:E
n..
n.. .
;, 40J
Z
O.
d'J
co:
-
:I:
~ "',J I
Q 200 ., I
~ ----.;:-, ~l'~
~ . I '-'~-.:::--_.,
-------
1.0
0.2
cfi. 0.8 - AIRCLEANER BLOCKAGE
o EFFECT ON CO
u
-------
.
The resulting increase in HC emissions was causing NDIR
instrument hangup ands therefores a substantial increase in
experimental error.
.
Its effect is so strong that it was felt to be adequately
characterized in the screening test.
Statistical Design
The remaining parameterss their indicated non1inearities, and the
corresponding candidate experimental levels are shown in Table 4-12.
Three fractional factorial experiments were considered which reflected
varying mixes of parameters and levels. These were:
. Mixed level 24 32 designs where all highly nonlinear effects
are evaluated and all parameters investigated.
. Mixed level 24 31 designs where the known idle CO non1inearities
are investigated at three levels. On vehicles equipped with air
injection systems, the PCV is not to be tested.
. A two-level 26 where all parameters are to be investigated.
The 24 31 statistically designed experiment was selected on the basis
that the air cleaner response was probably nearly linear over a large range
of the variable. In addition, a substantial reduction in the number of
tests could be achieved. The experimental design was obtained from Refer-
ence 11 and is shown in Table 4-13.
The diagnostic emission modes selected for Orthogonal Experiment No.2
were a subset of those performed in the first experiment. It was deter-
mined from the first experiment that, when emission modes are generically
classified to reflect the degree of engine loadings their diagnostic con-
tent is essentially identical. Therefore, the Federal modes, the key modess
a 1500 rpm no-load mode, and a wide-open throttle at 60 mph mode were
selected as representative classes of diagnostic emissions modes.
Air Cleaner Experiments
To characterize the nonlinear effects of air cleaner on emissions, a
subexperiment was designed using three levels of blockage with all other
4-29
-------
Table 4-12.
Selection of Parameters d.nd Their Levels
Definitive Experiment
L
H L
H
L
Candidate Experimental Designs
Levels Evaluated 6
2432 2432 2
2 2 2
2 2 2
3 3 2
3 2 2
2 2 2
,'.
2 ',' 2
x
24 24 26
32 31
2/9 1/2 1/2
32 24 32
Parameter Po pulation Reduction
Timing 10. 8 ppm
I rpm 9.4 ppm
lCO O. 61 ~Ij
A/C O. 1 8 2 0/0
PCV O. 062 ~o
~
I Air Pump O. 048 0;0
w
a
Nonlinear Res pons e
HC CO
Total Tests at Two Levels
Total Tests at Three Levels
Fractional Factor ial
Total Tests
L
~1
low nonlinear
moderately nonlinear
H
highly nonlinear
,'.
"'On vehicles equipped with air injection systems PCV is not to be tested.
-------
Table 4-13. Orthogonal Test Matrix
PCY Yalve
Idle or Air Idle
Run Timing rpm Air Pump Cleaner CO
""
A -r
1 +70 +1 010 Open Open +0.510
2 +7° -2010 Open Restricted +2. 010
3 +7° -2010 Closed Open - 0.510
4 -10° +1 010 Open Restricted -0.510
5 -10° +1 010 Open Re stricted +0.510
6 +7° -20% Closed Open +2. 010
7 -10° -2010 Open Open +2. 010
8 +7° +1 010 Open Open +2. 010
-'-
-,'
B
9 +70 -2010 Open Re stricted -0.5%
10 -10° -2010 Closed Restricted +0.510
11 +7° -2010 Closed Open +0..510
12 +7° +1 010 Closed Restricted +2. 010
13 -10° -20% Closed Re stricted +2. 010
14 -10° -2010 Closed Restricted -0.510
15 +7° +1 010 Closed Restricted -0.5%
16 +7° +1 010 Open Open -0.5%
.,.
C -,'
17 -100 - 2 010 Open Open +0.5%
18 +7° +1 010 Closed Restricted +0.510
19 -10° +1 010 Open Re stricted +2. 010
20 -10° +1 010 Closed Open -0.510
21 -10° +1 010 Closed Open +2. 010
22 -10° -2010 Open Open -0.510
23 -10° +1 010 Closed Open +0.510
24 +7° - 2 010 Open Restricted +0.510
~-
D -,'
.'-
-"Yehicle Stability Tests at Manufacture rs I Settings
Run A is Cold Start Test After Tuneup
4-31
-------
parameters set at manufacturer's specification. The air cleaner was blocked
while operating the vehicle at 50 mph and the Federal load such that increases
in mode CO of 0.25, 0.5, and 1.0 were obtained. The purpose of this proce-
dure is to ensure that the composite emissions will respond with a finite
change. An additional emissions test point was selected for the air cleaner
blocked to register 60 degrees on the AC tester. This level was selected on
the basis that it lies close to the parameter survey mean and yet is large
enough to effect ~n emissions response. Blockage was measured using the AC
tester for test settings made on the basis of CO at 50 mph.
Vehicle Emissions Stability Test
The screening experiments indicated that a combination of accumulated
mileage and high average emissions were significant contributors to experi-
mental error. To assess this contribution, vehicle emissions stability
tests were performed at the end of the 8th, 16th, and 24th test. In each
orthogonal series, these tests were run with the vehicle returned to manu-
facturer1s specification. Both hot bag and equivalent 6th and 7th cycle
mode emissions were obtained during these tests.
Power Train Selection
Eleven generic power trains were again studied in this experiment. At
the recommendation of the Steering Committee, two of the power trains in
the moderate sized V-8 classification were of 1970 vintage. The power
train matrix is presented in Table 4-14. The power trains have been selec-
ted to: (1) reflect national sales of make, power train size, and model
year, (2) reflect reasonable proportions of air reactor and engine modifi-
cation controlled vehicles and, (3) duplicate, where possible, those power
trains previously tested in the screening experiment.
4.5 DATA REDUCTION PROCEDURES
Data reduction procedures were upgraded for the definitive
reflect the more precise measurement requirements and new power
matrix. These revisions were:
tests to
train test
.
Application of the 1968 composite emissions calculation procedure
(i.e., dilution factor equal to 14.5/(10.5 HC + 0.5 CO + C02)'
4-32
-------
Table 4-14.
Definitive Orthogonal Experiment Power Train Attributes
~
I
W
W
Composite Emissions
Vehicle Pre- Post-
-
No. Make Model- Yr. CID Carb Mileage HC CO NO HC CO NO
601 Chev Wagon 1967 327 4 54058 193 1.94 430 248 2.24 908
602 Ford Mustang 1966 289 2 37690 576 2.24 859 412 1. 65 1313
603 Ford LTD 1969 429 2 18067 415 2. 19 1118 339 1. 15 1585
604 Pontiac Firebird 1968 400 4 29246 489 3. 78 468 178 1. 56 668
605 Dodge Challenger 1970 383 2 10818 178 .938 1236 208 .870 1012
606 Ford Torino 1969 302 2 21418 537 .430 2928 330 1. 21 2865
607 VW 1300 1969 91 1 15033 293 2.48 1122 333 3.02 1058
.608 Plym Barracuda 1968 318 2 32096 277 1.64 1703 213 1. 06 1638
609 Chev Impala 1970 350 2 13378 235 1. 18 997 156 . 913 858
610 Chev Nova 1969 230 1 16480 314 .498 1255 163 .87 1235
611 Chev Chevelle 1969 250 1 22424 154 .58 1748 f83 L02 1599
-------
.
All deceleration modes read and averaged in equal 20-second
increments.
The former modification was adopted to reflect the fact that 1970
vehicles were evaluated. The modified data reading procedure was adooted
as a reasonable approximation to mode integration, and to reflect more
accurately the nonlinear aspects of the CO calibration curves.
The data analysis, as with the screening experiment, consisted of a
statistical analysis of the experimental data. Specifically, the following
were calculated:
I
Emissions response surfaces for mass, composite and diagnostic
mode emissions
I Analysis of variances for emission response statistical sig-
nificance
I Analysis of va ri ances for parameter interactions.
A revised subroutine was written to perform the above analysis for a 31 24
statistical design. A typical data output display from this subroutine is
presented in Table 4-15.
The data field in the right column of Table 4-15 displays the basic
emission measurements in concentration units. Its primary use is to
quickly evaluate the validity of the input data. The top, left data field
is a computation of the mean emissions levels at the levels of the parameter
settings, A through E. These data were calculated applying the orthogonality
condition of the experimental design. A detailed analysis of the parameter
evaluated at three levels (i.e., idle CO) is reported in the field immedi-
ately below. The interactions are shown for the two component parts of the
main effect, El and E2' (i.e., the two linear slopes connecting the three
emission levels) and the remaining parameter main effects.
The second page summarizes the pertinent data from an analysis of
variances calculation. The final two columns are of most importance for
engineering considerations. The "F" ratio column and the degrees-of-freedom
combine to form a statement of statistical confidence or the main effect.
The "slopell column summarizes the respOnse surfaces for parameters I\-E,
4-34
-------
Table 4-15.
Federal Tests
Car Number 602 CD
Mode 8
Data Input Summary
Low Level
-1
~
I
W
CJ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
73
24
1. 585
7. 876
3.298
1.993
2.220
3.457
2.092
2.447
3.144
4.313
3.183
3.960
4.379
4.411
3.701
1.481
2.193
3.861
2.723
2.233
3.233
2.375
2.535
2.229
A
B
C
D
E
A = Timing
B = Rpm
C = Air Pump
D = Air Cleaner
E = Idle CO
2.892
3.162
3.547
3.317
2.829
E1
E2
y =
Emission at Parameter Setting
Means
Mid Level
o
2.765
Regression Slopes
Interactions With E(E1, E2)
A
0.089
-0.127
B
0.160
0.263
C
0.097
-0.127
High Level
+1
2.935
2.664
2.280
2.509
3.146
D
O. 052
O. 092
0.137 ':' E
2.765
Curve Fits For rco
The Linear Fit Is
Y = 2.822 +
The Quadratic Fit Is
0.128 >:, E>:":'2 +
- o. 065 >:< E +
-------
Table 4-15. Federal Tests (Continued)
Car Number 602 CD
Mode 8
Analysis of Variances
D.F. SS MS F Slope
Source Degrees of Sum of Mean Ratio ~ Emi'ssion
Variation Freedom Squares Sum ~ Parameter
Main Effects
A 1 1. 134E-02 1. 134E-02 0.218 -2.557E-03
B 1 1. 49 1. 49 28.6 -1. 660E-02
C 1 9.63 9.63 185 -1. 27
D 1 3.92 3.92 75.4 1. 62
E 2 O. 665 0.333 6.40 0.137
~ Two-Factor
I
(N Inte raction
0'1
AB=CD 1 0.562 0.562 10.8 0.153
AC=BD 1 1. 514E-03 1. 514E-03 2.914E-02 - 7. 944E - 03
AD= BD 1 4.671E-02 4.671E-02 0.899 4.412E-02
AE 2 6.776E-02 3.388E-02 0.652
BE 2 0.728 0.364 7.00
CE 2 7.044E-02 3.522E-02 0.678
DE 2 8.524E-02 4.262E-02 0.820
Re siduals 6 0.312 5. 198E-02
Totals 23 17.6
E1 1 1. 675E-02 1. 675E - 02 0.322 0.137
E2 1 0.581 0.581 11. 2 0.254
SS(L) 1 5.910E-02 5.910E-02 1. 14
SS(Q) 1 0.606 0.606 11. 7
Effective Coefficients Are
A B C D E1 E2 BYE
-0.022 0.149 -0.151 64. 012 -0.241 0.945 0.508
-------
(i.e., change in emission level per unit change in parameter setting). The
slope coefficients for those component failures which were simulated by
making them inoperative (i.e., PCV and air pump) are the emission changes
which result in going from the nominal to failed state. All other emission
response surfaces are based on unit changes in the parameter. For example,
the response surface for parameter A, timing deviation from manufacturers
specification, is-0.0025% change in composite CO per one degree of advance
of basic timing. This statement can only be made at a low level of confi-
dence as reflected by a 0.218 value of "F". An "F" value of 2.5
is required to make a statement on slope at the 90% confidence level. That
is, one out of ten times the measured effect may not be statistically sig-
nificant.
4.6 TEST PROCEDURES
4.6.1
Vehicle Preparation
Preparation for the definitive experiment was basically the same
as for the screening experiment, with each vehicle receiving an oil and
filter change, lias received" cold start emission test, diagnostic inspec-
tion, engine tune-up, and after tune-up cold-start emission test. The
vehicle was then instrumented for the orthogonal test series. For this
series of tests, however, the idle percent CO was specified more defini-
tively. Except for the Volkswagen, a member of each vehicle manufactuers'
engineering staff was contacted to obtain the recommended idle percent
CO setting. For General Motors Corporation vehicles, the emission idle
setting specification decal procedure was the recommended procedure.
This required only setting the carburetor mixture screws both with the
aid of an engine tachometer to determine a required rpm drop and with a
specific number of screw turns. The resulting idle percent CO on General
Motors vehicles and the specified idle percent CO supplied by the other
manufacturers were used as the null point for the idle percent CO deviations.
Section 3 describes the procedures used for the inspection and tune-up.
As in the screening experiment, no special procedures for balancing the idle
mixture screws were used. Occasionally some components, such as distribu-
tors and carburetors needed replacement, but new ones were sometimes
4-37
-------
impossible to obtain on short notice. In these cases, the defective units
were repaired in-house to maintain the close time scheduling of the program.
This procedure was probably in the best interest of the program in that it
provided more realistic data in relation to what could be expected in the
field.
Efforts were made to ensure that air pump performance on air injection
vehicles had been properly diagnosed. As part of the after tuneup inspec-
tion procedure, actual flow rates were measured in the manufacturer's
recommended idle speed and at 50 mph using rounded entrance orifice meters.
The system pressure was also measured at these engine speeds.
4.6.2 Me?surement and Special Test Equipment
The exhaust analysis equipment for the definitive experiment was the
same as used for the screening experiment. The procedures used were nearly
identical; only the "cold bag" constant volume, dilute sample was deleted
since only two hot 7-mode emission cycles were required.
Two pyrometers were installed in each vehicle to monitor the engine
temperature. One thermocouple was placed in the engine crankcase oil pan
and one installed in the upper radiator hose. The misfire generator was
not required.
4.6.3 Orthogonal Test Procedures
As in the screening experiment, the test procedure consisted of simul-
taneously and systematically setting five engine parameters to several
levels and measuring the 7-mode exhaust emissions over two cycles, as well as
collecting composite dilute bag samples. Emissions were also measured
during several diagnostic short cycles which were found to be most effectiv{~
for sensing malfunctions. Twenty-four tests were run on each vehicle, as
specified in the statistical design, to quantify the effect of each parameter.
The test matrix is shown in Table 4-13 and the short cycles in Figure 4-6.
Three additional base line tests were run with all the parameters set to
normal specifications. These tests served as an evaluation of the basic
engine condition during the test series.
4-38
-------
TABLE 4-16. Orthogonal Test
Air Injection System--Pump
Flow Rates and System Pressure
610*
Pump Flow Rates System Pressure
Idle 50 mph Idle 50 mph
37.81bs/hr 117.9 lbs./hr] 0.4 in. Hg 4.6 in. Hg
23.51bs/hr 72.31bs/hr 2 . 1 in. Hg 6.1 in. Hg
30.8 lbs/hr 85.5 lbslhr 1 . 3 in. Hg 5. 9 in. Hg
Car No.
601
602
*Flow rates checked at free flow.
I
j
4-39
-------
ro4
mph co
] ...
4.J
-70 ='
60 mph(3) ...
-------
The five parameters were set in a sequence so
setting was not affected by the subsequent setting
sequence and procedure were as follows:
that one particular
of another. The
1.
The ignition timing was set as specified in the test matrix, to either
7 degrees advanced or 10 degrees retarded from the manufacturers'
specifications.
2.
On engine modification controlled vehicles, the PCV hose was either
normally open or completely closed. On air injection-controlled
vehicles, the air pump was allowed to operate normally or the air
distribution hoses disconnected and the distribution manifolds
plugged immediately upstream of the check valves.
3.
Either a new or restricted air cleaner element was installed in the
air cleaner housing. The restricted elements were taped with five
equally spaced vertical openings so that the observed percent CO
at 50 mph road load was 0.5% CO greater than with the new element.
4.
The idle rpm was set with the carburetor speed screw at either 10%
above or 20% below the manufacturer1s specification.
5.
The idle percent CO was set to the desired deviation from the speci-
fication. The idle rpm was maintained at the desired setting and
the engine temperature was kept constant. Deviations were +0.5%
+2.0%, and -0.5% CO from specification.
Other than the modifications in short cycles shown in Figure 4-6, the
Volkswagen did not require any additional modifications to the test plan.
The restriction in the Volkswagen air cleaner was produced by draining the
oil from the housing and taping closed a sufficient number of holes in the
bottom of the filter media container. This produced the,desired increase
in percent CO.
4.6.4 Air Cleaner Restriction Experiment
Four auxiliary air cleaner restriction tests were performed after com-
pletion of the main experiment. The vehicle was set back to manufacturers'
specifications for these tests and air cleaner restrictions set at four
different levels. The vehicle was then run on the same test cycle used for
4-41
-------
the orthogonal experiment, two hot 7-mode cycles and the short cycles. The
idle percent CO and rpm were reset to specification before each test. The
restrictions used were based on the percent CO observed at 50 mph road load.
For three of the tests, the air cleaner element was taped partially closed to
produce 0.25%,0.50%, and 1.00% CO increases over the percent CO obtained
with a standard, unrestricted element. The fourth test was with the element
restricted to produce a 60-degree reading (angular) on the AC air cleaner
tester.
Air cleaner restriction \'as adjusted by taping the element so that there
were five equal-area, vertical openings. The air cleaner element was always
installed in the same position in the housing, and CO readings were made
with the air pump disconnected on vehicles so equipped. Fifty mph CO read-
ings were taken alternately with the standard and taped element during the
setup to ensure a consistent increase in CO.
Several problems were encountered during these tests as the percent CO
values proved unstable. Also, some vehicles did not show any response to
the restriction up to lBO-degree readings on the air cleaner tester. In
these cases the restriction was set on the AC tester. When the desired
higher restrictions '-'Jere obtained on six cylinder engines, the idle CO was
adversely affected by the restriction so that the desired percent CO could
not be set.
4.6.5
Test and Measurement Quality Control
Each vehicle's condition was maintained as consistently as possible
throughout the program. Engine temperatures were held nearly constant for
each test and while setting the idle CO. Both the engine coolant and oil
. temperature were monitored so that consistent warmup was obtained. The oil
temperature proved to be most variable and was held to ~10 degrees Fahrenheit
whenever possible before each test. The target oil temperature was deter-
mined by driving consecutive 7-mode cycles until the temperature stabilized.
Before each test, the vehicle was driven at 40 mph until the desired oil
temperature was reached. The idle CO settings were made by first warming
up the vehicle to the desired temperature and then setting the mixture and
speed without prolonged idle periods.
4-42 .
-------
I~
Table 4-16. System Cross Check Results
Orthogonal Expe riment
Labeled Observed Observed
Gas Value Value
Bottle No. Concentration System No.1 System No.2
2 57.6 ppm N02 58 ppm 55 ppm
3 77.5 ppm N02 70 ppm 69 ppm
4 3460 ppm NO 3270 ppm 3274 ppm
5 890 ppm NO 865 ppm 935 ppm
~:::
6 8596 ppm Hexane 8536 ppm 8613 ppm
-,.
4243 ppm ','
7 4332 ppm Hexane 4429 ppm
8 1 004 ppm Hexane 1003 ppm 989 ppm
8 7.91'70 CO 7.37'70 7. 62'70
8 13.9% C02 13.510 13. 4'70
9 294 ppm Hexane 295 ppm 298 ppm
9 2.00% CO 1. 97'70 2.03%
-..
9 7. 32'70 C02 7. 61 '70 (1 ) 7. 00'70"-
:::::
Indicates that correction was made for instrument drift.
4-43
-------
The same precautions taken with the instrument systems on the screen-
ing experiment were used on this experiment. In addition, with the elimin-
ation of the misfire generator, the hydrocarbon hangup was always maintained
to within one chart division on the low hydrocarbon recorder. Weekly pre-
ventive maintenance schedules were instituted to assure system consistency
throughout the experiment. A monthly system crosscheck (Table 4-16) was
also performed to assure close agreement between the two dynamometer sys-
tems used in th~ experiment.
Precautions were taken to ensure proper setup and maintenance of the
parameter deviations. In addition to the idle setting procedure described
above, the idle consistency was assured for the short cycles by running
them immediately after the 7-mode cycles. The intervening short cycles
required only 4 minutes, allowing sufficient time to obtain emissions
measurements from the bag samples. Parameter setting checklists recorded
actual values of the settings required and provided space for the driver
to list the attained values. Check sheets were used for both the ortho-
gonal experiment and the air cleaner restriction experiment. Quality of
the engine tune-up was assured by repeating the diagnostic inspection
after tune-up.
4-44
-------
REFERENCES
1.
California Air Resources Board, Project M-190.
motive Exhaust Emissions vs. Engine Adjustment
1969 Model Automobiles. Feb. 1969.
Variation in Auto-
Variables - Three
2.
California Air Resources Laboratory, Project Report M-166,
Variation in Automotive Exhaust Emissions vs. Engine Adjustment
Variables for Air Injection and Engine Modification Controls,
Nov. 21,1967.
3.
John N. Pattison. Clark Fegraus, A. J. Andreatch, and John C. Elston.
"New Jersey's Rapid Inspection Procedures for Vehicular Emissions."
SAE Paper No. 680111.
4.
C. L. Cline and L. Tinkham, "A Realistic Vehicle Emission Inspection
System," APCA Paper No. 68-152.
5.
TRW Proposal to the CRC, Sales No. 11608.000, IIVehic1e Emissions
Program,1I September 1968.
6.
M. P. Sweeney, and Brubacher, M. L., Exhaust Hydrocarbon Measure-
ment for Tuneup Diagnosis, SAE, #660105 January 10-14, 1966.
G. P. Box, J. S. Hunter, The 2k-t Fractional Factorial Design.
Part 1, Technometrics, 1. pp. 311-351. August 1961.
7.
8.
TRW Report, Vehicle Emissions Surveillance Study, Contract CRC,
APRAC CAPE-13-68. October 1970.
9.
W. K. Steinhagen. G. W. Neepoth, and S. H. Mick, Design and Develop-
ment of the General Motors Air Injection Reactor System, SAE Paper
660106, January 1966.
10.
R. F. Stebar and D. A. Brownson, Factors Influencing the Effective-
ness of Air Inspection in Reducing Exhaust Emissions, SAE Mid Year
r1eeting, Chicago, May 1965.
R-1
-------
11.
W. S. Conner and S. Young, Fractional Factorial Designs for Experi-
ments with Factors at Two and Three Levels, NBS Applied Mathematics,
No." 58, September 1961.
R-2
-------
APPENDIX A
SCREENING EXPERIMENT
DAH, SUMMARY
":L
-------
1966 Chevrolet Impala
327 cubic inch engine
air reactor
Car 501
Emissions Change (AE)
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
PC V
Air Pump
Leak
HC (ppm)
137.4
6.1
108.6
487.8
-158.4
-145.7
-15.5
148.2
53.4
-108.3
°88.7
CO (<}10)
0.0
0.32
0.07
0.17
0.57
-0.12
-0.83
0.37
0.23
0.55
-0. 65
Parameter Adjustment Range (AP)
-1 +1
-
Timing -10 7
Vac. Adv. 0 norm
Pt. Res. nom high
Mi sfire 0 5%
ICO -0.5 3.0
IRPM -100 100
Float -118 in. 1 14 in.
AIC 0 180
PCV 0 10010
Air Pump 0 100%
Leak 0 3. 0 cfm
Effectiveness Coefficient
Timing
Misfire
ICO
IRPM
Alc
Air Pump
Float
Leak
7.276*
66.345****
-168.3527**
6.558*
32.613*
0.61586***
0.08096**
0.05322**
-0.044***
-0.015***
1r 90% confidence level
~~* 95% confidence level
~:c*~:c 99% confidence level
~:c~:c~:~~:c 999% confidence level
I
-
A-l
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
A/c
PCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
A/c
PCV
Leak
Timing
Misfire
ICO
A/c
PCV
Float
Leak
:!:>
~
1968 Ford Custom
302 cubic inch engine
engine modification
Car 502
Emissions Change (AE)
HC (ppm)
247.6
-132.1
-55.2
472.3
-137.3
-93.9
81. 9
-54.3
63.3
176.5
Parameter Adjustment Range (~P)
-1
-10
o
o
o
-0.5
-100
-118 in.
o
o
o
Effe cti vene s s Coefficients
13.107*~,<*
64.238****
-204.345*
12.2:~:~
A-2
CO (10)
0.06
0.12
0.26
-0.04
0.42
0.09
-0.56
1. 15
0.47
-0.15
+1
7
norm
1
510
2.0
50
114 in.
180
10010
1. 0 cfm
0.63746****
0.25190****
0.05638****
-0.030***
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
PCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
AIC
PCV
Leak
Misfire
Alc
PCV
Timing
lCO
Leak
Float
1968 Dodge Charger
318 cubic inch engine
engine modification
Car 503
Emissions Change (~E)
HC (ppm)
40.0
178.0
61. 6
454.0
-55.0
-16.0
61. 7
-122.5
-2.6
142.8
Parameter Adjustment Range (~P)
-1
-10
o
o
o
-0.5
-100
-118 in.
o
o
o
Effectiveness Coefficients
61 . 744 ':: '::::: '::
-26.951 '::
9.8'::
A-3
CO (10)
-0.83
0.10
-0.11
0.06
0.29
O. 08
-0.62
1.74
1. 33
-0.47
+1
7
norm
1
510
2.0
50
1 14 in.
180
10010
1 . 0 dm
0.38311 '::,::,::,::
O. 15775 '::,::,::,:'
0.426'::
-0.032'::'::
-0.033'::':'
-------
1967 Mercury Cougar
289 cubic inch engine
air reactor
Car 504
Emissions Change (6E)
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
A/C
PCV
A. P. Res.
Leak
HC (ppm)
221..8
24.3
34.2
335.0
-17.0
-76.5
-8.5
-49.8
-30.8
14.1
-40.6
CO (10)
O. 06
0.17
0.10
0.13
0.43
-0.58
-0.30
0.16
0.24
0.29
-0.19
Parameter Adjustment Range (6P)
-1 +1
Timing -10 7
Vac. Adv. 0 norm
Pt. Res. 0 1
Misfire 0 2.3810
lCO -0.5 2.0
lRPM -100 50
Float -1/8 in. 1 /4 in.
A/C 0 180
PCV 0 10010
Air Pump Res. 0 10010
Leak 0 1 . 0 cfm
Effecti vene s s Coefficients
Timing
Misfire
lCO
lRPM
PCV
A. P. Res.
Float
Leak
11.742""':<>:<*
95.700>:":<>:<>:<
4.591':<>:<
O. 64 63 9 ':< >:<>:<>:":<
O. 02863>:":<
0.02815>:<>:<
-0. 016':<>:<
-0.013':<
A-4
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Fl oa t
A/c
PCV
AP. Res.
Leak,
Timing
Vac. Adv.
Mi sfire
ICO
IRPM
Float
A/c
PCV
AP. Res.
.~
.>
..
1966 Ford Mustang
ZOO cubic inch engine
air reactor
Car 505
Emissions Change (AE)
HC (ppm)
103.1541
0.3471
-6.6966
314.2517
-8.9888
-36.0851
2.5048
-37.7114
58.3604
36.1094
-5.9232
co (10)
-0.2681
0.1391
0.1081
-0.1638
0.2494
0.1350
-0.2863
- O. 0502
0.2089
0.2907
-0.1292
Effectiveness Coefficients
5.46>''<*
-0.01
83.47**>:'*
-13.38
2.17
0.13
-8.30
6.94
3.47
-0.0142>',<>',<
-0.0037
-0.0435
O. 371 >:'>:'
-0.0081
-0.0153**
-0.0110
O. 0249>:'
0.0279**
Parameter Adjustment Range (A P)
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
A/c
PCV
AP. Res.
Leak
-1
-10
o
o
o
-0.5
-100
-1 18 in.
o
o
o
o
A-5
+1
7
norm.
1
2.5610
2.0
50 .
114 in.
180
10010
10010
0.66 CFM
NOX (ppm)
888.8230
243.2337
-298.7855
282.2089
129.2298
162. 1784
196.7137
195.7326
-53.1603
-225.3394
15.0771
47.06****
-6.49***
74.96****
192.29*
-9.73**
10.49**
43. 06**
-6.33
-21. 63>,"**
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
rco
IRPM
Float
A/C
PCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
A/C
PCV
Leak
1968 Oldsmobile 98
455 cubic inch engine
engine modification
Car 506
Emissions Change (~E)
HC (ppm)
116.0798
107.2582
-9.4096
303.9252
64.0126
-140.3498
-8.3228
-23.9320
-47.0052
-72.1054
CO (<70)
-0.3360
-0.2844
-0.0328
-0.0402
-0. tt14
0.0802
-0.8246
0.6038
0.3102
-0.5918
Effectiveness Coefficients
6. 14>:'>:'*>:'
- O. 0178*>:'
0.0047>:<>:<
-0.0306
-0.0115
0.1658
-0.0048
-0.0440>:<>:<*>:<
0.1431>:<*>'0<>'0<
O. 0369>:<>:<
-0.0271>:<*>:<*
-1" 78~:~:!'::~:::!'::
-8.78
86.83>:<>:<*>:'
95.25>:'>:<
8.42>:<>:<>'0<>:<
-0.44
-5.68
-5.59>:<
-3.30>:<>:<>:'
Parameter Adjustment Range (~P)
-1 +1
Timing -10 7
Vac. Adv. 0 24
Pt. Res. 350<70 250%
Misfire 0 2. 38<70
ICO -0.5 2.010
IRPM -100 50
Float -1/8 in. 114 in.
Alc 0 167
PCV 0 10010
Leak 0 1. 5 CFM
A-6
NOX (pprn)
255.3430
235.5440
182.6096
85.4532
107.6542
54. 3650
151. 2648
-120.0864
-23.4012
110.3088
13.51*>:<*
-3.91***
170.54*>:<
24.41
160.19
-3.26
8. 06>:<
-28.48
-2.78
5.05
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
lCO
lRPM
Float
A/c
PCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
lCO
IRPM
Float
A/c
PCV
Leak
.,
.::
1968 Chevrolet Mali bu
327 cubic inch engine
engine modification
Car 507
Emissions Change (LlE)
HC (ppm)
246.1237
21. 5766
51. 4207
273.5788
-13.3314
-212.3738
-37.3743
23.5502
62.2087
132.8349
CO (ppm)
0.1215
0.0389
-0.0305
0.1825
0.6232
-0.1258
-0.5219
0.2684
0.3825
- O. 1463
NOX (ppm)
524.3278
130.8063
17.1489
6.2426
-90.5837
81.8173
104.9185
-99.3739
-190.7897
84.9297
Effectiveness Coefficients
13. 03::~:::::~:::~
0.0064
-0.0010
-0.0197
O. 0485 >:<>:'
o. 9 2 7 3 >:<>:<>:":'
0.0075
- O. 0278>'0<>:'>:'0:'
O. 0695>:":<>:'
O. 045 5 >:' >:<>:<>:'
-0.0093>:<
27.760:'0:'*>:<
-3.27':<*
11. 09
1. 66
-134.79>:<
-4.91
5.60':<0:'
-25.72
-22.70':<0:<>:<
5.41
-0.54
33.25
7 2. 67 0:":":<>:'
-19.84
12. 74':'>:":<
-1. 99
6.10
7.40
8. 46':":<
Parameter Adjustment Range (Ll P)
-1 +1
Timing -10 7
Vac. Adv. 0 norm
Pt. Res. 40010 250%
Misfire 0 2.5610
lCO -0.5 2.010
IRPM -100 50
Float -1/8 in. 114 in.
AIC 0 1530
PCV 0 10010
Leak 0 1.08 CFM
A-7
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
AIC
PCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
AIC
PCV
Leak
~
C"\
1966 Dodge Polara
383 cubic inch engine
engine modification
Car 508
Emissions Change (~E)
HC (ppm)
180.30
107. 18
- 31. 94
296.13
-97.57
-71. 57
-17.00
-22.69
28.35
-10.55
CO (10)
-0.6875
-0.4802
0.0030
-0.0373
0.3329
0.1266
-0.8237
0.5069
0.3146
-0.0546
Effectivenes s Coefficients
9.55*"'<*>:'
-1.59*>:''''<
1. 92
78.66>:<>:<"'<*
-145.18>:<>:<>:<
4.29>:<>:<
-0.91
-5.45
3.37
-0.57
- o. 0364 >:<>:<>;< ",<
0.0071>:<>:<>:<
-0.0002
-0.0099
0.4954*':<
-0.0076
- O. 0439 ':<>:< >:":'
O. 1216':<>:<>:<
o. 0374':<>:<
-0.0030
Parameter Adjustment Range (~P)
-1 +1
Timing -10 7
Vac. Adv. 0 27
Pt. Res. 300 250
Misfire 0 2.56
ICO -0.5 2.0
IRPM -100 50
Float -1 18 in. 114 in.
Alc 0 165
PC V 0 100
Leak 0 1. 27
A-8
NOX (ppm)
5 64. 15
530.27
191. 44
256.74
-93.72
-41. 74
178.30
-113.02
-171. 09
36.81
29.87>:<>:<",<*
- 7 . 8 6"'<>:<*>:<
-11.49"'<*
68.20>:<>:<*
-139.46
2.50
9.51':<"'<
-27.12
-20.36':<*
2.01
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
PCV
Air Pump
Leak'
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
AIC
PCV
Air Pump
Leak
-.l\
1969 Chevrolet Nova
230 cubic inch engine
air reactor
Car 509
Emissions Change (AE)
HC (ppm)
280.40
58.62
11.53
2 64. 5 6
-35. 82
-256.86
-17.33
29.28
31.64
59.31
23.51
CO (10)
-0.2461
-0.0875
-0.0454
0.0734
0.2928
0.2844
-0.4437
0.1106
0.7059
0.5079
-0.0417
Effectiveness Coefficients
14.84*>'0<**
-1.02*
7.36
70. 27***~<
-53.30
15. 41~<***
-0.92
6.55
3.77
5.69*
2.13
-0.0130
0.0015
-0.0290
0.0195
0.4356>'0<
-0.0171*
-0.0237**
0.0247
O. 0840**>'0<*
0.0488***
-0.0038
NOX (ppm)
411. 93
120.41
-32.37
72.79
-3. 09
-64.15
154.47
18.51
-247.13
-86.80
40.36
21. 81*>'0<**
-2. 09~<*
20. 67
19.33
-4.60
3.85
8. 24~<>'0<
4.14
-29.41****
-8.33
3.66
Parameter Adjustment Range (A P)
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
PCV
Air Pump
Leak
-1
-10
o
250
o
-0.5
-100
-1/8 in.
o
o
o
o
A-9
+1
7
16
167
2.56
2.0
50
1/4 in.
177
100
100
0.76 CFM
-------
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
FCV
Leak
Timing
Vac. Adv.
Pt. Res.
Misfire
ICO
IRPM
Float
Alc
PCV
Leak
-
-
';)
1968 Ford Ranchero
390 cubic inch engine
engine modification
Car 510
Emissions Change (AE)
HC (ppm)
308.83
267.66
142.68
220.59
-288.31
-217.28
92.41
97.67
99.38
247.35
CO (%)
0.0034
0.0279
-0.0084
0.0891
0.5648
0.0942
-0.3214
0.0692
0.4758
-0.0236
NOX (ppm)
571. 41
677.43
-18.83
-27.40
-70.23
10.47
128.52
-26.73
-161. 69
48.71
Effectiveness Coefficients
16. 35>:'~oJ,<
-5. 63*~'>',<
-83.60
58.59>:'>:<
-429. 01 ~o:<~,
13. 04*~'
4.93
21.49
11. 83
13. 21>:'~'
0.0002
-0.0006
0.0049
0.0237
O. 8405*>:'*~'
-0.0057
-0.0171***
0.0152
0.0566
-0.0013
30.25**>:'>:'
-14.26****
11. 03
-7.28
-104.50
-0.63
6.85**
-5.88
-19.24*>',<*
2.60
Parameter Adjustment Range (AP)
-1 +1
-
Timing -10 7
Vac. Adv. 0 19
Pt. Res. 378 250
Misfire 0 2.56
ICO -0.5 2.0
IRPM -100 50
Float -1/8 in. 114 in.
AIC 0 180
FCV 0 100
Leak 0 1. 29
A-10
-------
APPENDIX B
DEFINITIVE EXPERIMENT
DATA SUMMARY
(j.)
-------
YEAR:
1967
A Timing
B Idle rpm
C Air Pump
D Air Cleaner
E Idle CO
Residuals
Timing
Idle rpm
Air Pump
** Air Cleaner
Idle CO
Interaction
AC = BD
AD = BC
BEl
BE2
COMPOSITE EMISSIONS
Car 601
MAKE:
Chevrolet
CID: 327
J
EMISS. DEV: Air
Emission Changes
HC (ppm)
CO(%)
NO (ppm)
181* 0.32*
-95* 0
-155* -1.765*
-8 -0.26
18.6/0.66 0.15/0.45*
522*
22
29
68
- 14/ - 130*
49
0.35
117
Parameter Adjustments
-1
4° ATDC
o
+1
130BTC
660
480
Plugged
Norma 1
Plugged (160°)
Norma 1
5.5%
6.5%
8.0%
Interactions
HC
CO
.till
O. 162*
-0.176*
0.404*
-0. 107:*
51.3*
*5tatistically significant at 90 per cent confidence level.
** Parameter level change, 6P = -1 to provide computer usage sign convention
campa t i b i 1 i ty .
')-
B-1
-------
MASS EMISSIONS
Car 601
YEAR: 1967
MAKE: Chevrolet
CID: 327
EMISS. DEV: Air
Emission Changes
A Timing
B Idle rpm
C Air Pump
D Air Cleaner
E Idle CO
Residuals
HC (grams/mile) <;0 (gram~mile)
1. 05* -9.85*
-1.38* -0.32
-3.14* -22.64*
-0.281 -2.10
0.852/-0.201* 3.46/-7.89
0.50 8.13
NO (grams/mile)
2.05*
0.23
-0 .11
0.23
-0.02/-0.08
0.53
Parameter Adjustments
-1
o
+1
-
Timing
4° ATDC
13° BTDC
Idle rpm
480 rpm
660 rpm
**
Air Pump
Plugged
Normal
Air Cleaner
Plugged (160°)
Normal
Idle CO
5.5%
6.5%
8%
Interactions
Interaction
HC
CO
NO
-
-
-
. AD = BC
0.344*
0.23*
AE1
12.46*
AE2
-0.043*
BEl
0.716*
BE2 -0.520*
CE1 0.149* 8.35*
CE2 -0.818* 0.325*
.Statistically significant at 90 per cent confidence level.
** Parameter level change, 6P = -1 to provide computer
compati bi 1 i ty. B-2
usage sign convention
-------
COMPOSITE EMISSIONS
Car 602
YEAR: 1966
CID: 289
EMISS. DEV: Ai r
MAKE:
Ford
Emission Changes
HC (ppm) CO(OJo)
A Timing 175.5* 0.043
B Idle rpm -61.3* -0.498*
C Air Pump -112.2* -1.267*
D Air Cleaner -19.8* -0.808*
E Idle CO -1.9/20.3 -0.064/0.381*
Residuals 20 0.228
Parameter Adjustments
-1 0
Timing
100ATDC
440
Idle 'pm
** Ai r Pump
Plugged
Plugged (1650)
Air Cleaner
Idle CO
3.1%
4.1%
Interactions
Interaction
HC
CO
AB = CD
AD = BC
AC = BD
AEl
AE2
BEl
BE2
CEl
CE2
DEl
DE2
*Statistically significant at 90 per cent confidence level.
** Parameter level change ~P = -1 to provide computer usage sign
compatibility.
..10.3*
12.7*
-18.1*
27.2*
0.160*
0.263*
B-3
NO (ppm)
506.2*
57.4*
180.9*
206.2*
66.1/-75.7
78
+1
70 BTDC
600
Normal
Norma 1
5.6%
~
31.2*
40.4*
-94.2*
112.1*
-34.5*
-120.3*
..141.5*
52.9*
45.9*
-99.1*
convention
-------
MASS EMISSIONS
Car 602
YEAR: 1966
MAKE: Ford (Mustan~)
CI D: 289
EMISS. DEV: Air
Emission Changes
HC (grams/mile) <;0 (grams/mile) NO (grams/mile)
A Timing 1. 51* -4.00 1.12*
B Idle rpm -0.87* -1. 37 0.08
C Air Pump -1.02* -22.84* 0.75*
D Air Cleaner -0.36 -18.27* 0.67*
E Idle CO -0.13/0.29 1. 50/8.51* 0.51/-0.15
Residuals 0.61 6.1 0.90
Parameter Adjustments
-1 0 +1
-
, Ti m i ng 10° ATDC 7 ° BTDC
Idle rpm 440 rpm 600 rpm
**Air Pump Plugged Normal
Air Cleaner Plugged (165°) Normal
Idle CO 3.1% 4.1% 5.6%
Interactions
Interaction HC CO NO
AB = CD 0.440* 5.21*
AC. = BD -0.336*
AD = BC 0.452* -2.63*
AEl -0.715*
AE2 1.096*
BEl -0.748*
BE2 0.61*
CE1 1.26*
CE2 -9.24*
*Statistically significant at 90 per cent confidence level.
** Parameter level change, 6P = -1 to provide computer usage sign convention
compatibility. 8-4
-------
,-~
YEAR: 1969
A Timing
B Idle rpm
C PCV
D Air Cleaner
E Idle CO
Residuals
Tim ing
Idle rpm
PCV
Air Cleaner
Idle CO
Interaction
AB = CD
AD = BC
AC = BD
BEl
BE2
DEl
DE2
COMPOSITE EMISSIO NS
Car 603
MAKE:
Ford
CID: 429
EMISS. DEV: E.M.
Emission Changes
HC (ppm)
146.5*
CO(%)
-0.007
NO (ppm)
913.2*
- 11 3 . 9* 0.246*
-22.7* 0.327*
-10.1 -0.080*
-0.4/39.7* 0.287/0.286*
-81.7*
205.2*
-70.3*
87
0.082
-130.5/76.6*
71
Parameter Adjustments
-1
40ATDC
o
+1
440
13° BTDC
600
Plugged
plugged (130°)
2.3%
Norma1**
(-0.45 inches of water)
Normal
3.3%
4.8%
Interactions
HC
12.1*
CO
.3.938 x 10-2*
3.780 x 10-2*
NO
59.1*
0.068*
0.058*
112.1*
-93.4*
*Statistically significant at 90 per cent confidence level.
* Parameter change in inches of water-crankcase pressure suction.
B-5
-------
MASS EMISSIONS
Car 603
YEAR: 1969
MAKE: Ford - (LTD)
CID: 429
EMISS. DEV: E.M
Emission Changes
HC (grams/mile) <:;0 (gramv'mile)
A Timing 0.79* -4.10*
B Idle rpm -1. 39* 8.25*
C PCV -0.46 -1.03
D Air Cleaner 0.09 -1.10
E Idle CO -0.02/0.38 6.23/7.09*
Residuals 0.79 3.6
NO (grams/mile)
2.76*
-0.29
0.97*
0.07
-0.90/0.93*
0.56
Parameter Adjustments
-1 0
Timing 40 ATDC
Idle rpm 440 rpm
PCV Plugged
Air Cleaner Plugged (1300)
Idle CO 2.3% 3.3%
Interactions
+1
-
130 BTDC
600 rpm
Norma1**
(-0.45 inches of water)
Normal
4.8%
Interaction HC CO NO
-
AC = BD 0.40*
BEl 1.2*
BE2 -0.7*
*Statistically significant at 90 per cent confidence level.
** Parameter change in inches of water-crankcase pressure suction.
8-6
-------
COMPOSITE EMISSIONS
Car 604
YEAR:
1968
MAKE: Ponti ac
CID: 400
EMISS. DEV: £OM.
Emission Changes
HC (ppm) CO(%) NO (ppm)
A Timing 168.5* 0.130 341.5*
B Idle rpm -69.0* O. 192 42.9
C PCV -22.2 - 1. 128* 143.3*
D Air Cleaner -65.0* -1.296* 204.8*
E Idle CO -52.4/24.0 -0.089/0.608 12.7/-80.0
Residuals 49 0.99 126
Parameter Adjustments
-1 0 +1
Timing 1° ATDC 16° BTDC
Idle rpm 520 715
PCV Plugged Normal**
(1260) (-0.45 inches of water)
Air Cleaner Plugged Normal
Idle CO 1.6% 2.6% 4.1%
Interactions
Interaction HC CO t!Q
AB = CD -22.7*
DEl 31.3*
DE2 33.9*
.
~
*Statistically significant at 90 per cent confidence level.
** Parameter change in inches of water-crankcase pressure suction.
B-7
-------
MASS EMISSIONS
Car 604
YEAR:
MAKE: Pontiac (Firebird) CID: 400
1968
Emission Changes
HC {grams/mile} <;:0 {grams/mile}
A Timing 2.04* -20.00
B Idle rpm -1.30* 5.86
C PCV -0.49* -27.56*
D Air Cleaner -1. 04* -30.49*
E Idle CO -0.68/+0.50* 0.176/24.21
Residuals 0.57 30.01
Parameter Adjustments
-1
o
Ti m i ng
1° ATDC
Idle rpm
520 rpm
PCV
Plugged
Air Cleaner
Plugged (126°)
Idle CO
1.6%
2.6%
Interactions
Interaction
HC
CO
AB = CD
AD = BC
-0.32*
0.27*
AEl
AE2
CE1
CE2
DEl
DE2
0.21*
0.62*
i\'
* Statistically significant at 90 per cent confidence level.
.....
8-8
EMISS. DEV: E.M.
NO {grams/mile}
1. 54*
0.37*
0.37*
1.00*
0.36/-0.22*
0.29
+1
16° BTDC
715 rpm
Normal
Normal
4.1%
NO
-0.18*
0.19*
0.73*
-0.12*
-0.68*
0.11*
-0.32*
0.37*
-------
COMPOSITE EMISSIONS
Car 605
YEAR: 1970
MAKE:
Dodge
CID: 383
EMISS. DEV: E. M.
Emission Changes
HC (ppm) CO(%) NO (ppm)
A Timing 49.6 -0.348* 562.2*
B Idle rpm -137.8* 0.364* -87.5*
C PCV 53.9 -0.068 -47.2
D Air Cleaner 33.3 -0. 194 -45.7
E Idle CO -132.3/-67.9* 0.396/0.317* -51.0/66.4
Residuals 163 0.28 115
Parameter Adj ustments
-1 0 +1
Tim i ng 7.5 ATDC 9.5 BTDC
Idle rpm 520 725
PCV Plugged Normal
Air Cleaner Plugged (1800 +) Normal
Idle CO 0.5% 1.5% 3.0%
Interactions
Interaction HC CO NO
AD = BC 55.4*
BEl 0.46*
BE2 -0.11*
*Statistically significant at 90 per cent confidence level.
**Parameter level change~ 6P = -1 to provide computer usage sign convention
compatibility.
B-9
-------
MASS EMISSIONS
Car 605
YEAR: 1970
MAKE: Dodge -
CID: 383
EMISS. DEV: E.M.
Emission Changes
HC (grams/mile) <:0 (grams/mile) NO (grams/mile)
A Timing 0.28 -12.09* 1.53*
B Idle rpm -1.84 8.70* -0.02
C PCV 0.75 -0.55 0.07
D Air Cleaner 0.32 -5.97* 0.35
E Idle CO -1. 76/-0.73 8.06/6.31* 0.10/-0.24
Residuals 2.72 6.05 0.54
Parameter Adjustments
-1
o
+1
-
Timing
7.50 ATDC
9.50 BTDC
Idle rpm
PCV
520 rpm
725 rpm
Plugged
Normal
Air Cleaner
Idle CO
Plugged (1800)
0.5% 1.5%
Interactions
HC CO
- -
-5.04*
Normal
3%
Interaction
NO
AB = CD
AD = BC
0.24*
*Statistically significant at 90 per cent confidence level.
8-10
-------
YEAR: 1969
A Ti m i ng
B Idle rpm
C PCV
D Air Cleaner
E Idle CO
Residuals
Timing
Idle rpm
PCV
Air Cleaner
Idle CO
Interaction
AB = CD
BEl
BE2
COMPOSITE EMISSIONS
Car 606
MAKE:
C 10: 302
Ford
Emission Changes
HC (ppm)
CO(%)
202. 1 * 0.074
-166.2* 0.122*
-32.2* -0.422*
-24.3 -0.010
-44.4/-8.1* 0.215/0.211*
34
0.16
Parameter Adjustments
-1
4° ATDC
o
440
Plugged
Plugged (165°)
1. 7%
2.7%
Interactions
HC
CO
-27.2*
55.5*
17.8*
*Statistically significant at 90 per cent confidence level.
---
8-11
EMISS. DEV: E.M.
NO (ppm)
1084.8*
-86.7
720.8*
-21.9
-220.2/338.1
128
+1
13° BTDC
605
Norma 1
Norma 1
4.2%
.till
-156*
-------
MASS EMISSIONS
Car 606
YEAR:
1969
MAKE: Ford
CID: W.
EMISS. DEV: E.M.
Emission Changes
HC (grams/mile) <;0 (grams/mile)
NO (grams/mile)
A Timing 1. 41* -5.0*
B Idle rpm -1. 77* 9.49*
C PCV -0.70 -7.31*
D Air CI eaner -0.44 1.56
E Idle CO -0.71/0.93 9.11/10.26*
Residuals 1.09 3.42
3.44*
0.30
0.90
-0.03
-0.16/0.02
1.46
Parameter Adjustments
-1
o
+1
Tim i ng
40 ATDC
13 0 BTDC
Idle rpm
440 RPM
605 RPM
PCV
Plugged
Normal
Air Cleaner
Plugged (1650)
Normal
Idle CO
1. 7%
2.7%
4.2%
Interactions
Interaction
HC
CO
NO
AD = BC
-1. 57*
BEl
BE2
DEI
DE2
2.56*
2.35*
5.51*
0~33*
*Statistically significant at 90 per cent confidence level.
B-12
-------
YEAR: 1969
A Timing
B Idle rpm
C PCV
D Air Cleaner
E Idle CO
Residuals
Timing
Idle rpm
PCV
Air Cleaner
Idle CO
Interaction
None*
COMPOSITE EMISSIONS
Car 607
MAKE:
CID: 92
V.W.
Emission Changes
HC (ppm)
83.7*
CO(%)
-0.174*
-246.2* 0.136
29.8 O. 1 20
..38.0 -0.472*
-26.9/-43.6 0.179/0.123*
95 0.22
Parameter Adi~stments
-1
100 ATDC
o
680
Plugged
Plugged (1.9%)
4.0%
5.0%
Interactions
HC
CO
*Statistically significant at 90 per cent confidence level.
8-13
EMISS. DEV: E .M.
NO (ppm)
668.2*
84.4
-99.8*
128.7*
-141.8/96.5
129
+1
70 BTDC .
935
Normal
Norma 1 ( 1. 9% )
6.5%
NO
-------
YEAR: 1969
A Timing
B Idle rpm
C PCV
D Air Cleaner
E Idle CO
Residuals
Ti m i ng
Idle rpm
PCV
Air Cleaner
Idle CO
Interaction
AE1
AE2
I ,
MASS EMISSIONS
Car 607
MAKE: V .w. .
CID: 91
Emission Changes
HC (grams/mile) <;0 (grams/mile)
-0.14 -5.25*
-0.97* 7.04*
-0.03 1.68
0.32* -9.64 *
-0.046/0.185 4.60/3.83*
0.41 4.15
Parameter Adjustments
-1
o
100 ATDC
680 RPM
Plugged
Plugged
4%
5%
Interactions
HC
CO
-
*Statistically significant at 90 per cent confidence level.
8-14
EMISS. DEV: !.:!h.
NO (grams/mile)
0.94*
-0.07
-0.10
0.33*
-0.21/0.23
0.28
+1
-
70 BTDC
.935 RPM
Normal
Normal
6.5%
NO
-
0.12
-0.40
-------
r-
COMPOSITE EMISSIONS
Car 608
YEAR: 1968
MAKE:
Plymouth
CID: 318
EMISS. DEV: E.M.
Emission Changes
HC (ppm) CO(%) NO (ppm)
A Timing 94.6* 0.129* 597.9*
B Idle rpm -25.0* 0.093 -12.8
C PC V -33.7* -0.698* 347. 1 *
D Air Cleaner -11.3 -0.344* 38.7
E Idle CO 10.2/11.1 0.415/0.442* -1.9/-131.2*
Residuals 19 0.17 106
Parameter Adjustments
Idle rpm
-1
-
12.5° ATDC
520
o
+1
4.50BTDC
Timing
715
Air Cleaner
Plugged
Plugged (165°)
Normal
PCW
Normal
Idle CO
0.5%
1.5%
3.0%
Interactions
Interaction
HC
CO
NO
-
AC = BD
AD = BC
AE1
AE2
62. 1 *
6.370 x 10-2*
0.232*
-0.054*
-65.6*
-90.6*
*Statistically significant at 90 per cent confidence level.
B-15
-------
MASS EMISSIONS
Car 608
YEAR:
1968
MAKE: P1ym(Barcda)
CID: 318
EMISS. DEV:
E.M.
Emission Changes
HC (grams/mile) <;0 (grams/mile) NO (grams/mile)
A Ti m i n9 0.46 -1.54 2.17*
B Idle rpm -0.41 6.01* -0.11
C PCV -0.64* -12.17* 1.20*
D Air Cleaner -0.02 -8.61* 0.66
E Idle CO 0.20/0.13 14.94/15.55* 0.10/-0.28
Residuals 0.70 3.67 1.01
Parameter Adjustments
-1
o
+1
-
-
Ti i11 ing
12.50 AIDC
4.50 BTDC
Idle rpm
520 RPM
715 RPM
PCV
Plugged
Normal
Air Cleaner
AC = BD
Plugged (1650)
0.5% 1.5%
Interactions
HC CO
-
-1.85*
2.07*
5.12*
O. 77*
2.24*
-5.12*
Normal
Idle 0)
3%
Interaction
NO
0.54*
AD = BC
BEl
BE2
DEI
DE2
~'Statistically signjfj,:.']"t at 90 per cent confidence level.
8-16
-------
COMPOSITE EMISSIONS
Co r 609
Chevrolet
CID: 350
EMISS. DEV: LM.
YEAR: 1970
MAKE:
Emission Changes
HC (ppm) CO(%)
A Timing 17 4 . 2* -0.344*
B Idle rpm 1 06. 1 * 0.171*
C PCV -37.8 -0.319*
D Air Cleaner 33.9 0.107*
E Idle CO 39.9/-50.5 0.088/0.296*
Residuals 77 0.076
Parameter Adjustments
-1 0
Timing 6° ATDC
Idle rpm 480
PCV Plugged
Air Cleaner Plugged (1800)
Idle CO 0.5% 1.5%
Interactions
Interaction
AB = CD
AC = BD
BEl
BE2
CEl
CE2
DEl
DE2
HC
-
-45.9*
CO
-
-0.138*
-3.886 x 10-2*
-0.125*
0.156*
-0.123*
0.025*
0.133*
-0.086*
*Statistically significant at 90 per cent confidence level.
B-17
NO (ppm)
590.0*
-25.4
107.9*
-78.7
36.5/-17.0
130
+1
11 ° BTDC
660
Norma 1
Normal
3.0%
~
-------
MASS EMISSIONS
Car 609
YEAR: 1970
MAKE:
Chevrolet
CID: 350
EMISS. DEV: .!:1:h
Emission Changes
HC (grams/mile) <;0 (grams/mile) NO (grams/mile)
A Timing 1.22* -23.64* 2.66*
B Idle rpm -0.76* 0.16 0.32
C PCV -0.12 -7.60 0.92
D Air Cleaner 0.23 1.24 0.50
E Idle CO 0.31/-0.30 17.06/0.28 0.09/-0.98
Residuals 0.63 17.78 1.32
Parameter Adjustments
-1
o
+1
-
Ti m i ng
6 ° ATDC
11 ° BTDC
Idle rpm
480 RPM
660 RPM
PCV
Plugged
Normal
Air Cleaner
Plugged (180°)
Normal
Idle CO
0.5%
1.5%
3.0%
Interactions
Interaction
HC
CO
NO
-
-
AB = CD
-3.21
0.589
*Statistically significant at 90 per cent confidence level.
B-18
-------
COMPOSITE EMISSIONS
Car 610
YEAR: 1969
CID: 230
EMISS. DEV: Air
MAKE:
Chevrolet
Emission Changes
HC (ppm) CO(%)
A Timing 120.3* -0.211*
B Idle rpm -146.0* 0.115
C Air Pump -216.9* -1.325*
D Air Cleaner 53.0 -0.231*
E Idle CO 59.9/-75.2 -0.153/0.503*
Residuals 96 .26
Parameter Adjustments
-1 0
Timing 10° AToC
Idle rpm 560
Ai r Pump Plugged
Air Cleaner Plugged (1750)
Idle CO 5.5% 6.5%
Inte roc t ions
HC
Interaction
fQ
-0.191*
Q.281*
-0.317*
-0.194*
-0.240*
AB = CD
AC = Bo
AD = BC
CE1
CE2
*Statistically significant at 90 per cent confidence level.
B-19
NO (ppm)
525.7*
8.4
-70.4
287.3*
188.8/-158.5
200
+1
70BTDC
770
Nonna1
Nonna 1
8.0%
NO
-
116*
-------
MASS EMISSIONS
Car 610
YEAR: 1969 MAKE: Chevrolet CI D: 230 EMISS. DEV: Air
Emission Changes
HC (grams/mile) <;0 (grams/mile) NO (grams/mile)
A Timing 0.66* -14.30* 1. 64*
B Idle rpm -1.22* 8.50* 0.18
C Air Pump -3.25* -39.39* 0.02
D Air Cleaner -0.16 -7.90* 0.36
E Idle CO -0.07/.25 -9.78/18.21* 0.46/-.53
Residuals 0.75 8.51 0.5
Parameter Adjustments
-1 0 +1
-
Ti m i ng 10° ATDC 7° BTDC
Idle rpm 560 RPM 770 RPM
Air Pump Plugged Normal
Air Cleaner P1ugg~d (175°) Normal
Idle CO 5.5% 6.5% 8.0%
Interactions
Interaction
HC
CO
NO
AC = BD
AD = BC
8.4*
-8.4*
AE1
AE2
BEl
BE2
CE1
CE2
1.42*
-12.22*
-0.08*
0.89*
-1. 07*
-0.12*
-11.9*
-1. 9*
*Statistically significant at 90 per cent confidence level.
8-20
-------
COMPOSITE EMISSIONS
Car 611
YEAR: 1969
MAKE:
Chevrolet
CI D:250
EMISS. DEV: E.M.
Emission Changes
HC (ppm) CO(%) NO (ppm)
A Timing 138.6* 0.007 461. 2*
B Idle rpm -109.7* 0.578* - 1 58. 9
C PCV -62.1 * -0.445 159.8
D Air Cleaner -73.1* -2.886* 187.5*
E Idle CO 5.2/3. 1 0.321/0.372 -239.2/74.2
Residuals 60 0.68 255
Parameter Adjustments
Timing
-1
-
6° AT DC
440
o
+1
11 ° BTDC
Idle rpm
PCV
605
Air Cleaner
Plugged
Plugged (180°)
Normal
Norma 1
Idle CO
2.5%
3.5%
5.0%
Interactions
Interaction
HC
CO
NO
-107*
104*
128*
AB = CD
AC = BD
AD = BC
*Statistically significant at 90 per cent confidence level.
8-21
-------
MASS EMISSIONS
Car 611
YEAR: 1969 MAKE: Chev. (Chv11e) CI D: 250 EMISS. DEV: E.M.
Emission Changes
HC (grams/mile) <;0 (grams/mile) NO (grams/mile)
A Timing 0.47* 0.03 3 .85*
8 Idle rpm -0.46* 10.98* -0.27
C PCV -0.16 -4.19 0.09
D Air CI eaner -0.55* -36.47* 0.29
E Idle CO 0.04/-0.03 5.56/4.58* -1.40/0.61
Residuals 0.24 8.25 1.65
Parameter Adjustments
-1 0
Timing 6 ° ATDC
Idle rpm 440 RPM
PCV Plugged
Air Cleaner Plugged (1800).
Idle CO 2.5% 3.5%
Interactions
Interaction HC CO
- -
AD = BC
+1
-
11° BTDC
605 RPM
Normal
Normal
5%
NO
0.647*
*Statistically significant at 90 per cent confidence level.
8-22
-------
;:.;' . I
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 601
90 Pe rcent Significance
Change in Emission, ~E, Low to High Parameter Level
A B C D E
Mode Timing Idle RPM Air Pump Air Cleaner Idle CO
-
Idle F -4. 310 +2.145
Idle C -0.777 -4. 981 +2. 070
30-15 F -5.543 +1. 378
50-20 F -0.539 +0.443 -5.509 +1. 107
1500 RPM C -0.463 -5.421 +1. 934
OJ 30 F +0. 246 - 2. 128 -0. 318 +'). 408
I
N
eN 33 C -1. 439 -0. 609
0-25 F +0.351 -1. 257 +f). 390
15-30 F -1. 066 -0. 268 +0. 364
50 C +0. 373 -0. 298
WOT C -1. 780 -0. 361 +0. 440
-------
~ ,. - 'J
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 602
90 Percent Significar.ce
Change in Emission, 6£, low to High Parameter level
A B C D E
Mode Timing Idle RPM Air Pump Air CleC\.ner Idle CO
Idle F -1. 791 +2.148
Idle C -0. 409 -2. 340 +2. 021
30-15 F -0.279 -0.662 -4. 926 -0. 370 +1.050
50-20 F - 3. 744 -0. 470 +1. 019
1500 RPM C 0.505 -0. 507 -2. 683 + 1. 779
30 F +0.251 -0. 472 -2. 482 -0. 912
c:I
I 33 C -0. 269 -1. 154 -0. 905
N
-'='"
0-25 F -0. 428 -0. 969 -0. 693 +0. 254
15-30 F -0. 362 -1. 026 -0.841 +0. 146
50 C -0. 638 -1. 322
WOT C -0. 585 -2. 296 -1. 023 +'1. 421
-------
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM C
co 30 F
I
N
c.n
33 C
0-25 F
15-30 F
50 C
WOT C
A
Timing
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 604
90 Percent Significance
Change in Emission, 6E, Low to
B C
Idle RPM Air Pump
Hi~h Parameter Level
D
Air Clei;l.ner
0.195
+0.370
+0.411
- o. 969
-0.465
-i. 316 -1.284
-1.033 -1.711
-1.172 -1. 322
-1. 148 -1.546
-0.517 -0.594
-1. 177 -1. 201
E
Idle CO
+1. 682
+2. 285
+2. 104
+1. 192
+ 2. 022
-------
'l..'~
CO Diagnostic Mode Emission Sensitivity to f4alfunction
Vehicle 605
90 Percent Significance High Parameter Level
Change in Emission, 6E, Low to
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
-
Idle F -0.186 +0. 200 +2. 654
Idle C +0. 366 +2. 858
30-15 F +0. 176 -0. 135 +2. 312
50-20 F +0. 204 +2. 200
1500 RPM C +2.763
cg 30 F +0. 323 -0. 299 +0. 303
I
N
0'1
33 C -0.421 +0. 200
0-25 F -0.445 +0. 345 -0. 359
15-30 F -0.428 +0. 306
50 C -to 046
WOTC -0.293 +0. 267 -0. 386
-------
t.-,' .,...
CO Diagnostic Mode Emission Sensitivity to ~1alfunction
Vehicle 606
90 Percent Significance
Change in Emission, 6E, Low to Hi gh Pa rarre ter Leve 1
A B C D E
Mode Timing Idle RPM FCV Air Cleaner Idle CO
Idle F O. 165 +2. 505
Idle C +2. 530
30-15 F -0.687 +0.644 -0.400 +1.674
50-20 F -0. 383 +0.449 +0. 181 +0. 223 +1. 647
1500 RPM C +2. 237
CJ 30 F +0. 056 -0. 185 +0. 109
I
N
.....,
33 C -0. 356 -0.136
0-25 F -0.460 -0.199
15-30 F +0.140 -0.446
50 C -0.263
WOT C O. 208
-------
J~ t. . .:..'
CO Diagnostic i~ode Emission Sensitivity to ~'alfunction
Mode
Idle F
Idle e
30-15 F
50-20 F
i500 RPM e
tXJ
I 30 F
N
CX)
33 C
0-25 F
15-30 F
50 e
WOT e
A
Timing
Vehicle 607
90 Percent Significance
Change in Emission, 6:',
e
pev
B
Idle RPM
-0.896
+1. 146
-0. 633
+1. 10
-0.210
+0. 11 2
-0. 134
Low to High Parameter Level
D E
Air Cleaner Idle CO
+2. 148
+2. 310
+1. 738
+1. 49
+2. 257
-0. 158
+0. 208
-0. 208
+0. 231
-0. 883
-0.493
-0. 422
+0. 217
-------
I.' :J
co Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 608 F
90 Percent Significance
Change in Emissiont 6Et Low to High Parameter Level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
.
Idle F +0.147 0.195 +2. 621
Idle C +0. 307 +3. 019
30-15 F +0.213 +0.211 +0. 1 21 +2. 465
50-20 F +0. 227 +0. 252 +0. 257 +2. 357
1500 RPM C +0. 203 +2. 297
O::J 30 F +0. 268 -0. 866 -0. 298 +0. 699
I
N
\.0
33 C +0. 182 -0. 988 -0. 424 +0. 430
0-25 F +0.171 -0.665 -0. 396 +0. 637
15-30 F +0. 141 -0. 905 -0. 446 +0. 299
50 C -0. 497 -0. 289
WOT C -0.503 -0. 177 +0. 718
-------
o :~"'d
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 609
90 Percent Significance
Change in Emission, L\E, Low to High Parameter Level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idie CO
Idle F 0.180 +2. 575
Idle C -0.329 +2. 382
30-15 F -0.414 +2. 161
50-20 F -0.984 +0. 284 -0. 226 +2. 188
1500 RPM C -0.621 +2. 239
D' 30 F -0.546 +0.581 +0. 528
I
w
0
33 C +0. 168 -0.499 +0.211
0-25 F -0.538 -0. 304
15-30 F -0.138 -0.577 +0. 243
50 C -0.360
WOTC -1.616
-------
I~ .)
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 610
90 Percent Significance
Change in Emission, 6E, Low to High Parameter Level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F +0. 735 -0.772 -5. 372
Idle C -5.42
30-15 F -4.522 +0.790
50-20 F +0.467 -4. 345
1500 RPM C -5.556
OJ 30 F -0.872 +0.542 -1. 359
I
(.oJ
.......
33 C -1. 238 -1.405
0-25 F -1.164
15-30 F 0.611 -0.197 +0. 144
50 C
WOTC -0. 984 +0.408 -1.530 -0. 344
-------
.. rc.-CJ
CO Diagnostic Mode Emission Sensitivity to t1alfunction
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM C
CtI
I 30 F
w
N
33 C
0-25 F
15-30 F
50 C
WOT C
A
Timing
Vehicle 611
90 Percent Significance
Change in Emission, 6E~
B C
Idle RPM PC V
O. 165
+0. 163
+0. 331
+0.594
+0.056
+0.023
-0.051
+0.769
Low to High Parameter Level
D E
Air Cleaner Idle CO
+2. 526
-0. 199 +2. 396
-1. 021 +2. 004
-0. 300 + 1. 80 2
-0. 453 +2. 905
-4. 642
+0. 070
-2. 493
-2.074
-0. 071
-1. 882
-------
"~ -<.j
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM
CJ
I 30 F
w
w
33 C
0-25 F
15- 30 F
50 C
WOT C
C = Clayton
F = Federal
CO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 601
90 Percent Significance
Change in Emission,6E,
C
Air Pump
A
Ti.ming
Low to High Parameter Level
D
Air Cleaner
B
Idle RPM
+166 -111 -214
+165 -112 -244
+302 -369 -518
+491 -1435 -1499
+143 -94
+104 -164
+82 -158
+144 -38 -135
+129 -114
+92 -40
+87 -157
E
Idle CO
+55
+49
+354
+64
-------
.~ -r;}
HC Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 602
90 Percent Significance
Change in Emission, ~E,. Low to High Parameter Level
A B C D E
Mode Timing Idle RPM Air Pump A ir Cleaner Idle CO
Idle F +125 -49 -157
Idle C +92 -32 -149 +16 +47
30-15 F +127 -51 -320 +24
50-20 F +510 -747 -600 +203
1500 RPM C +109 -90 -177 +73
co 30 F +129 -36 -249 -26
I
(.0.)
~
33 C +82 -145
0-25 F + 151 -47 -140 -15 +17
15- 30 F +126 -27 -160 -21
50 C +65 -13 -83 -14 +19
WOT C +123 -30 -216 +24
C = Clayton
F = Federal
-------
yf ~
HC Diagnostic Mode Emission Sensitivity to Ma1function
Vehicle 604
90 Percent Significance
Change in Emission,6E, Low to High Parameter Leve1
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F -143 -143
Idle C -106
30-15 F + 223 -157
50-20 F +706 -958 -192 -103
1500 RPM C -394 -520
OJ 30 F +110 -30 -34
I
w
<.1'1
33 C +151 -59
0-25 F +143 -42 -69 -40
15-30 F +139 -22 -50
50 C +80 -39
WOT C +129 -45 -45
C = Clayton
F = Federal
I
.
-------
"i~ .v.
Mode
Idle F
Idle C
0-15 F -562 -671
0-20 F -867 -1187
1500 RPM C -600
CtI +22 -19
I 30 F
w
C"I
33 C +21 -14
0-25 F
15- 30 F
50 C
WOT C
HC Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 605
90 Percent Significance
Change in Emission: 6E,
C
PCV
B
Idle RPM
Low to High Parameter Level
D
Air Cleaner
E
Idle CO
A
Timing
+30
+20 -10 -33
+25 -24
C = Clayton
F = Federal
-------
V(. (.
HC Diagnostic Mode Emission Sensitivity to Malfunction
A
Timing
Vehic1e 606
90 Percent Significance
Change in Emi ss ion, 6E, Low to High Parameter Level
BCD
Idle RPM PCV Air Cleaner
Mode
Idle F
Idle C
30-15 F
50-20 F
1 500 R PM C
OJ
I
W 30 F
'-J
33 C
0-25 F +202 -141 -39
15-30 F +118 -54 -32 -16
50 C
WOT C
+341
-258
+373 -444
+807 -1575
+419 -289
+123 -64
-127
-26
-24
C = Clayton
F = Federal
E
Idle CO
-211
-------
g~'~
Mode
Idle F
Idle C
30-15 F
50-20.F
~ 1500 RPM C
I
W 30 F
())
33 C
0-25 F
15-30 F
50 C
WOT C
C = Clayton
F = Federal
HC Diagnostic Mode Emission Sensitivity to Malfunction
A
Timin2
Vehicle 607
90 Percent Significance
Change in Emission, ~E,
C
PC V
B
Idle R PM
Low to High Parameter Level
D
Air Cleaner
E
Idle CO
+120
+34
+103
-146
-89
-1892
+95
-1933
-120
+63
-34
+90
-47
+149
+79
+58
-35
-.736
-62
-------
hi -:~
,
HC Diagnostic Mode Emission Sensitivity to Malfunction
V ehic1e 608
90 Percent Significance
Change in Emissions ~Es Low to High Parameter Level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F +81 +72
Idle C +70 -25 +88
30-15 F +78 -43 +42
50-20 F + 158 -479 +239 -332
1500 RPM C +62 -16 -17 +67
OJ
I
W 30 F +92 -11 -30 -17 +31
\0
33 C +63 -34 -19
0-25 F +92 +5 -28 -17 +33
15-30 F +86 -34 -25 +24
50 C +40 -19 -19
WOT C +84 -27 +23
C = Clayton
F = Federal
-------
"h' :' \
HC Diagnostic Mode Emission Sensitivity to ~~alfunction
Vehicle 609
90 Percent Significanc.;e
Change in Emission, 6:=:, Low to High Parameter Level
A B C D E
Mode Timing Idle R PM PCV Air Cleaner Idle CO
Idle F +278
Idle C +217
30-15 F + 581 -564
50-20 F +959 -1034
1500 RPM C
CD
I
~ 30 F +54 +33 -7 +29
o
33 C +46 -37
0-25 F +85 -37
15-30 F +87 -13 -29 +22
50 C +63
WOT C +32
C = Clayton
F = Federal
~
"
-------
I h-Id
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM C
CP
I
~ 30 F
~
33 C
0-25 F
15-30 F
50 C
WOT C
C = Clayton
F = Federal
HC Diagnostic ~lode Emission Sensitivity to r.1alfunction
A
Timing
Vehicle 610
90 Percent Significance
Change in Emission, 6E,
C
Air Pump
Low to High Parameter Level
D
Air Cleaner
B
Idle RPM
E
Idle CO
+167
-243
-182
-764
-1490
-218
-166
-112
-229 +142
-131
-96
-115
-148
+127
-90
-830
+883
-1653
-116
+134
+61
-36
+94
-44
+93
-48
+54
-50
-------
lh.,q
HC Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 611
90 Percent Significance
Change in Emission, 6E, low to High Parameter level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F +102 -92 +55
Idle C +83 -55 +42
30-15 F +186 -345 -110
50-20 F +448 -1449
1500 RPM C -280
aJ
I 30 F +82 -84 +53
~
N
33 C +53 -38 -24
0-25 +72 -40 -82
15-30 F +116 -84 -61
50 C +26 -11 -23
WOT C +119 -46
C = Clayton
F = Federal
-------
-;:'. j
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM C
OJ
,
~ 30 F
w
33 C
0-25 F
15- 30 F
50 C
WOT C
C = Clayton
F = Federal
NO Diagnostic Mode Emission Sensitivity to Malfunction
A
Timing
Vehicle 601
90 Percent Significance
Change in Emission, AE,
C
Air Pump
B
Idle RPM
+22
+57
+19 -19
-36
-64
-29
-107
-247
-87
-181
-204
-181
+182
+327
+627
+509
+569
+1275
+550
Low to High Parameter Level
D
Air Cleaner
E
Idle CO
+82
-148
+270
+ 117
-------
;~OX Diagnostic Mode Emission Sensitivity to !~a lfunction
V ehic1e 602
90 Percent Significance
Change ~:n Emission, bJ:, Low to High Parameter Level
A B C D E
Mode Tim ing Idle RPM Air Pump Air Cleaner Idle CO
Idle F +17 -15 -24
Idle C
30-15 F +60 +22 +31 -25
50-20 F +60 +60
1 500 R PM C
CD
I
~ 30 F +322 +140
~
33 C +617 -207 +410
0-25 F +492 -84 -140
15- 30 F +531 -69 +247
50 C +499 -223 +636
WOT C +375 + 197
C = Clayton
F = Federal
-------
---~~
NOx Diagnostic Mode Emission Sensitivity to Halfunction
V ehide 604
90 Percent Significance
ChangE! in Emi ss i on, AE, Low to High Parameter Level
A B C D E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F +23 +28
Idle C +46 -40
30-15 F +30 +24
50-20 F + 153
c:J 1 500 R PM C +38 -30
I
~
U'1 30 F +84 +99
33 C +533 +157 +248
0-25 F +392 +188 +223
15-30 F +457 +175 +251
50 C +1321
WOT C +148 +80 +102
C = Clayton
F = Federal
-------
L
NO Diagnostic Mode Emission Sensitivity to Malfunction
A
Timing
Vehicle 605
90 Percent Significance
Change in Emission, 6E,
C
PCV
B
Idle RPM
Mode
Idle F
Idle C
30-15F +34
50-20 F +74
1 500 R PM C -10 +19
CP
I
~ 30 F +253 -168
~
33 C +978 -239
0-25 F -62
15-30 F +692 -105
50 C + 1 061
WOT C +493 -196 -150
+24
C = Clayton
F = Federal
Low to High Parameter Level
D
Air Cleaner
+145
-126
+353
+141
E
Idle CO
-20
+123
-------
Mode
Idle F
Idle C
30-15 F
50-20 F
o::J 1500 RPM C
I
~
-....J 30 F
33 C
0-25 F
15- 30 F
50 C
WOT C
C = Clayton
F = Federal
NOx Diagnostic Mode Emission Sensitivity to ~alfunction
A
Timing
Vehi.cle 606
90 Percent Significance
Change in Emission, 6E,
B C
Idle RPM PCV
+30
+38
+31
+56
Low to High Parameter Level
D
Air Cleaner
E
Idle CO
+38
-40
+368
+97
+35
+1694
+982
+834 +487
+1187 +1063
+872
+1419
+34
-------
ilOx Diagnostic f10de Emission Sensitivity to :~a1function
Vehicle 607
90 Percent Si.gnificance
Change in Emission, 6E., Lo'll to High Parameter Level
A B C 0 E
Mode Timing Idle RPM PCV Air Cleaner Idle CO
Idle F -15 +14
Idle C
0-15F +92
50-20 F +83
0::1 1 500 R PM C -28 +8 +8
I
.;:. -126
(X) 30 F +700
33 C +664
0-25 F + 525 +260
15 - 30 F +874 -117 +137 -42
50 C +977 + 113
WOT C +636 -116
C = Clayton
F = Federal
-------
It. "
j' ."
NO Diagnostic j'1ode Emission Sensitivity to Malfunction
Vehicle 608
90 Percent Significance
Change in Em; s s ; on, 6 E, Low to High Parameter Level
A B C D E
Mode Timing Idle R PM PCV Air Cleaner Idle CO
Idle F -66 -64
Idle C +49
30-15 F +62 +36
50-20 F + 134 +59
1500 RPM C -27
OJ
I +168 -171
~ 30 F +490
1.0
33 C +923 +365 +187 -288
0-25 F +504 +329 -119
15-30 F +813 +469
50 C +1072 +331 +138
WOT C +558 -92 +185 -210
C = Clayton
F = Federal
-------
1°5 8
Mode
Idle F
Idle C
30-15 F
50-20 F
1500 RPM
c:J
I 30 F
<.11
o
33 C
0-25 F
15-30 F
50 C
WOT C
C = Clayton
F = Federal
NO Diagnostic Mode Emission Sensitivity to Malfunction
Vehicle 609
90 Percent Significance
Change in Emi ssion, ~E.
B C
Idle RPM PCV
A
Timing
Low to High Parameter Level
D
Air Cleaner
-65
-57
+91
+125
+223
+102
+102
-53
+33
+610
+440
+136
+863
+130
-84
+1474
+474
E
Idle CO
-54
-59
-30
+1
-------
NO Diagnostic Mode Emission Sensitivity to Malfunction
Mode
Idle F
Idle C
30-15F +42 -51 +45 -23
50-20 F + 158 -101
1 500 RPM C -12 -19
c;o
I 30 F +239 -129 -181 +113 -34
U'1
--'
33 C +964 +515
0-25 F +444 -114 +268
15-30F +614 -336 +244
50 C +2046 -481
WOT C +382 +229
A
Timing
Vehide 610
90 Percent Significance
Change in Emission, ~E,
C
Air Pump
.
Low to High Parameter Level
D E
Air Cleaner Idle CO
B
Idle R PM
C = Clayton
F = Federal
-------
Mode
Idle F
Idle C
30 -15 F
50-20 F
1 500 R PM C
OJ
I
t.11 30 F
N
33 C
0-25 F
15-30 F
50 C
WOT C
C = Clayton
F = Federal
NO Diagnostic Mode Emission Sensitivity to r~alfunction
A
Tim ing
Vehicle 611
90 Percent Significance
Change in Emissian, ~E,
C
PCV
B
Idle RPM
-14
+20
+16
+120
-142
-17 +25
+534
+ 11 08
+689
-1069
+ 521
Low to High Parameter Level
D
Air Cleaner
E
Idle CO
+16
-32
-13
-117
+653
-468
------- |