United States
Environmental Protection
Agency
Motor Vehic'e Emission Lab
2565 Plymouth Rd
Ann Arbor, Michigan 48105
EPA-460,'3-81-009
April 1981
Air
SEPA
Light-Duty Vehicle
Driveability Procedure
Investigation
-------
EPA-460/3-81-009
LIGHT-DUTY VEHICLE DRIVEABILITY
PROCEDURE INVESTIGATION
By
W.C. Williams
Amoco Oil Company
Research & Development Department
Box 400
Naperville, Illinois 60566
Contract No. 68-03-2875
EPA Project Officer: J.P. Whitehead
Prepared for:
ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF AIR, NOISE AND RADIATION
OFFICE OF MOBILE SOURCE AIR POLLUTION CONTROL
EMISSION CONTROL TECHNOLOGY DIVISION
CHARACTERIZATION AND APPLICATIONS BRANCH
ANN ARBOR, MICHIGAN 48105
APRIL, 1981
-------
This report is issued by the Environmental Protection Agency to disseminate technical
data of interest to a limited number of readers. Copies are available free of charge to
Federal employees, current contractors and grantees, and nonprofit organizations—in
limited quantities—from the Library, Motor Vehicle Emission Laboratory, Ann Arbor,
Michigan 48105, or, for a fee, from the National Technical Information Service, 5285
Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by Amoco Oil Company,
Box 400, Naperville, Illinois 60566, in fulfillment of Contract No. 68-03-2875. The con-
tents of this report are reproduced herein as received from Amoco Oil Co. The opinions,
findings, and conclusions expressed are those of the author and not necessarily those of
the Environmental Protection Agency. Mention of company or product names is not to be
considered as an endorsement by the Environmental Protection Agency.
-------
TABLE OF CONTENTS
age
INTRODUCTION 'l •
SUMMARY l
CONCLUSIONS 2
RECOMMENDATIONS 2
EXPERIMENTAL 3
Literature Search 3
Car Screening and Selection 3
Instrumentation Installed on Car ^
Data Collection System ^
Driveability Tests Conducted 7
RESULTS 11
Trained Rater Observations 11
Data Editing by Computer 14
Stumble Measurement by Computer 15
Hesitation Measurement by Computer 19
Engine Stall Measurement by Computer 25
Engine Idle Roughness Measurement by Computer 25
Hard Starting Measurement by Computer 28
Total Demerit Measurement — Phase I Tests 28
Demerit Measurement — Phase II, III and IV Tests 28
Consolidation of Computer Methods 35
LIST OF REFERENCES . 40
APPENDIX A
I CRC Cold Start and Driveaway Test Procedure A-I-1
II Motorist Driving Cycle A-II-1
APPENDIX B
Instrumentation Descriptions B-l
APPENDIX C
Trained Rater Observed Demerits by Driveability Problem Type C-l
APPENDIX D
Unsuccessful Attempts to Measure Stumble and Hesitation D-l
APPENDIX E
Data Used for Developing Hesitation Measurement Method E-l
APPENDIX F
Rater-Observed and Computer-Calculated Demerits
(Test Phases II-IV) F-l
APPENDIX G
Results of Analyzing Phase I Tests with Consolidated
Program G-l
-------
LIST OF TABLES
Table Page
I. CANDIDATE TEST CARS 5
II. CAR SREENING TEST RESULTS 6
III. RATER - OBSERVER DEMERITS BY TEST PHASE 12
IV. STUMBLE DEMERITS BY TEST FUEL (TEST PHASE I) 21
V. HESITATION DEMERITS BY TEST FUEL (TEST PHASE I) 24
VI. STALL DEMERITS BY TEST FUEL (TEST PHASE I) 27
VII. IDLE ROUGHNESS DEMERITS BY TEST FUEL (TEST PHASE I) 29
VIII. HARD STARTING BY TEST FUEL (TEST PHASE I) 31
IX. TOTAL DEMERITS BY TEST FUEL (TEST PHASE I) 33
X. TOTAL DEMERITS BY TEST FUEL (TEST PHASES II-IV) 34
-------
LIST OF FIGURES
Figure Page
1. Driver's Data Entry Keyboard 8
2. Raw Data Printout 9
3. Trained Rater - Observed Demerit Summary 10
4. Hesitation and Stumble(s) During an Acceleration 16
5. Stumble Demerits in Test 73 (Phase I) 18
6. Stumble Demerits for Phase I Tests 20
7. Hesitation Demerits for Phase I Tests 23
8. Stall Demerits for Phase I Tests 26
9. Idle Roughness Demerits for Phase I Tests 30
10. Total Demerits for Phase I Tests 32
11. Total Demerits for Phase II Tests 36
12. Total Demerits for Phase III Tests 37
13. Total Demerits for Phase IV Tests 38
14. Stumble Demerits Using Consolidated Computer Program 39
( Phase I Tests )
-------
INTRODUCTION
The automotive and petroleum industries have long recognized that vehicle
driveability is an important consideration in product design and manufac-
turing. Over the years companies in both industries have independently and
cooperatively gathered research data to use in setting product specifi-
cations that insure desired driveability performance in customer service.
One of the predominate cooperative organizations is the Coordinating
Research Council (CRC). Since 1970 CRC has conducted several research
programs (1,2,3,4)* to evaluate driveability variations among vehicles and
to evaluate the influence of gasoline volatility upon vehicle drive-
ability. In most of these tests trained raters subjectively evaluated
vehicle driveability as they drove cars through a specified test cycle.
Test repeatability was often poor partially because many of the ratings
were subjective.
The Environmental Protection Agency's (EPA) interest in driveability stems
from evidence that adjustments of some vehicle engine settings to values
other than those recommended by the manufacturer can improve driveability
during cold start and warmup driving but often exhaust emissions and/or
fuel economy suffer as a result (5). Because of this, EPA may eventually
consider issuing driveability guidelines or standards. Such regulations
must be based on quantitative test methods but current industry test
procedures are primarily subjective. Consequently EPA awarded contract
68-03-2875 to Amoco Oil Co. to determine whether an objective procedure
could be developed" for assessing vehicle cold stare and warmup driveability.
SUMMARY
A research program was conducted by the Amoco Oil Co. under contract with
the Environmental Protection Agency to develop instruments and computer
programs for objectively measuring vehicle cold start and warmup drive-
ability. After a series of screening tests on 15 candidate cars, a 1979
Chrysler was selected for the extensive driveability testing required to
accomplish the research objective. The car was equipped with several
instruments which the investigators judged capable of detecting and
measuring the severity of driveability problems. Nearly 200 driveability
tests were conducted with this car on chassis dynamometers using various
driving cycles and ambient test temperatures. Throughout each test the
instrument output signals and the trained raters' evaluations of perfor-
mance were computer recorded at the rate of five times per second. From the
data gathered, a series of computer programs were developed to identify and
measure the severity of several driveability problems. Most of the
computer programs were designed specifically for the one car used in the
test work — they generally are not to be considered universally applicable
to other vehicles.
Numbers in parentheses indicate references listed at the end of the
report.
-------
CONCLUSIONS
Computer methods were developed for measuring the severity of five
driveability problems. They were :
• stumble (sudden loss of power followed by a resumption of power)
• hesitation (lack of response to opening the throttle)
• engine stalls (any time the engine quits running with the ignition
key in the "on" position)
• engine idle roughness
• hard starting (excessive cranking time during startup)
In CRC testing, hard starting and engine stalls are rated objectively and
computer methods were developed which nearly duplicate these results.
The other driveability problems are subjectively evaluated and attempts
were made to develop computer correlations that accurately match the raters '
observations. In some cases this was successful but in other cases addi-
tional developmental work is needed. It appears feasible that, with further
effort, the computer methods described in this report can be improved.
No attempt was made to develop methods of measuring backfire (an
explosion in the induction or exhaust system), extension (an abnormally
slow or sluggish acceleration), or surge (cyclic pulses of power).
Previous driveability testing by Amoco and CRC indicated that these are
relatively minor problems. For the car tested in this program, backfire
was a major problem, but surge and extension rarely occurred. With proper
instrumentation methods can probably be developed to measure any of these
problems.
In general, the instruments used and the computer programs developed
during this investigation provide a means to measure driveability of the
only car tested. The specific correlations developed for calculating
demerits probably cannot be directly applied to other cars, but the general
analytical procedures and the basic measurement methods should be valid.
RECOMMENDATIONS
In developing the computer procedures described in this report, only one
car was tested, all work was done on chassis dynamometers and driveability
evaluations were obtained from only two trained raters. To develop a
universal objective driveability system, many more cars need to be
included in the data base, the methods should be verified by on-the-road
driveability tests, and subjective opinions of driveability performance
should be gathered from additional trained raters.
In the long term we suspect test repeatability on chassis dynamometers can
be further improved by using mechanical/electrical "automatic drivers" in
place of human drivers to manipulate the throttle. Auto drivers are more
consistent than humans in throttle opening rates and throttle positions.
-------
We recommend that driveability testing not be attempted using the Federal
Test Procedure cycle nor any other cycle in which a driver is forced to
drive the car according to a predetermined vehicle speed-versus-time
schedule. These types of cycles purposely allow the throttle to be
manipulated to overcome and thereby mask driveability problems. Further-
more, we found the FTP cycle does not require rapid enough accelerations to
highlight driveability problems of the car we tested and we therefore
suspect that differences between cars cannot be found by using this cycle.
Our last recommendation is that if further development work is conducted on
the problem of objective driveability measurement, the engine/vehicle
operating parameters studied should not be limited to the few used in the
system finally developed for this project (primarily engine speed,
throttle position, and starter engagement). Driveability of other cars
may correlate better with other operating parameters.
EXPERIMENTAL
Literature Search
A literature search was conducted early in the program to identify
potential instruments and methods for objectively measuring vehicle
driveability. Based on this search and Amoco's previous driveability
measurement experience, ten engine/vehicle operating parameters were
identified as candidates. They were:
• vehicle speed
• vehicle acceleration
• engine speed
• engine intake manifold vacuum
• engine vibration
• engine rotational movement relative to the car frame
• throttle position
• drawbar pull
• driveshaft torque
• starter voltage
Car Screening and Selection
Screening tests were conducted on several cars to select one which
displayed a variety of driveability problems. For this work the CRC cold
start and driveaway test procedures were used. They are described in
detail in Appendix A. Basically a rater drives the car through a specified
cycle and evaluates the severity (slight, moderate or heavy) of any
driveability problems which occur. These ratings are then translated into
numerical demerit values for analyses. For reference the most serious
driveability problem, engine stall while driving, is assigned 32 demerits
per occurrance.
-------
The screening tests were conducted in an all-weather, large-roll (7 ft.
diametLerlj, chaaai.s dynamometer at 20°F (-7°C). Two trained ra_ters.. each
.pvaliiat-pd pprfo-nnance of all candidate cars three times: .twice us.ing_a_
Low....volatilitY..fuel_.and once.using a high volatility fuel. Pertinent
inspections on these fuels were:
Fuels for Car Screening
Distillation, °F (°C)
10% Evap.
50% Evap.
90% Evap.
Reid Vapor Pressure, Ibs.
(Kpa)
Low
Volatility
140 (60)
250 (121)
355 (179)
7.1 (48.9)
High
Volatility
105 (41)
185 (85)
295 (146)
13.5 (93.1)
A description of the 15 cars screened, is "hown in Table I and results of
the screening tests are shown, in Table II. Based on these results the
Chrysler LeBaron (car" ID 9CHY1) was selected for the remainder of the test
work. This car was chosen for two reasons. First, it displayed a wide
range of driveability problems and problem severities with the low
volatility fuel. Second, the driveability of this car was highly sensitive
to changes in fuel volatility as shown by the difference in total demerits
for the two fuels.
Instrumentation Installed on Car
After selecting the test car, it was equipped with instruments to measure
the ten engine/vehicle operating parameters mentioned previously. Description
of each instrument is given in Appendix B. After a few preliminary tests,
four operating parameters were eliminated from further consideration.
These parameters and the reasons for discarding them were:
• vehicle acceleration — the available instrument could not accu-
rately measure accelerations at vehicle speeds below 10 mph and many
problems occurred at low speeds.
• engine vibration — high frequency instrument signals tended to mask
driveability problems
• engine rotational movement — same as engine vibration, and
• draw-bar pull — because car movement was necessary to produce a
signal, it became highly dependent upon the method used to secure
the car to the dynamometer
Data Collection System
A large number of cold start and warmup driveability tests were conducted
on chassis dynamometers using the instrument-equipped test car. Through-
out each test the analog instrument signals were monitored continuously and
-------
TABLE I
CANDIDATE TEST CARS
Make and Model d'
AMC Concord
Buick Century
Buick Century '2;
Buick LeSabre
Chevrolet Chevette
Chevrolet Impala
Chevrolet Malibu
Chevrolet Malibu
Chrysler LeBaron
Ford Fairmont
Ford LTD
Mecury Marquis '^'
Oldsmobile Cutlass
Plymouth Horizon
Pontiac Sunbird
Engine
Displ. , L.
2.0
3.8
3.8
4.9
1.6
5.0
3.3
4.4
5.2
3.3
5.0
5.8
4.3
1.7
2.5
Carburetor
Venturis
2
2
4
2
2
2
2
2
2
1
2 (3)
2 (3)
2
2
2
Car ID
9AM1
9BU1
9BU2
9BU4
9CV1
9CV4
9CV2
9CV3
9CHY1
9F03
9F04
9MER1
90L1
9PLY1
9P01
(1) All were 1979 models equipped with automatic transmission.
(2) Turbocharged.
(3) Variable Venturi carburetor.
(4) Equipped with three-way catalyst system and closed-loop A/F ratio
control.
-------
TABLE II
CAR SCREENING TEST RESULTS
Average Demerits
Car
ID
9AM1
9BU1
9BU2
9BU4
9CV1
9CV4
9CV2
9CV3
9CHY1
9F03
9F04
9MER1
90L1
9PLY1
9P01
Fuel*
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
L
H
Hard
Starting
9.8
1.5
1.4
1.0
1.2
0.8
2.2
0.5
14.4
1.0
5.8
0.5
2.4
10.5
4.4
1.8
4.5
0.5
2.2
1.0
3.8
0.5
0.0
0.0
3.3
1.0
5.3
1.2
1.5
1.0
Stalls
Idle
4.0
0.0
0.0
0.0
8.0
0.0
10.0
0.0
4.0
0.0
22.0
0.0
12.0
12.0
14.0
4.0
8.0
0.0
8.0
0.0
12.0
0.0
0.0
0.0
7.0
0.0
16.0
4.0
2.0
0.0
Driving
136.0
64.0
32.0
0.0
56.0
0.0
144 . 0
0.0
352.0
48.0
104.0
0.0
32.0
16.0
120.0
0.0
248.0
0.0
104.0
16.0
0.0
0.0
280.0
0.0
176.0
0.0
10.6
16.0
48.0
0.0
Hesita-
tion
120.0
21.0
37.5
30.0
96.0
9.0
54.0
6.0
96.0
27.0
43.5
9.0
60.0
3.0
43.5
9.0
103.5
36.0
42.0
9.0
9.0
0.0
27.0
9.0
29.4
1.2
22.0
33.0
52.5
6.0
Stumble
42.0
6.0
42.0
9.0
36.0
18.0
76.5
3.0
72.0
15.0
64.5
3.0
66.0
12.0
75.0
9.0
159.0
6.0
61.5
3.0
0.0
0.0
46.5
0.0
33.8
6.0
14.0
12.0
16.5
0.0
Surge
1.0
2.0
19.0
16.0
8.0
6.0
5.0
0.0
9.0
0.0
2.0
0.0
12.0
8.0
7.0
0.0
18.0
2.0
2.0
0.0
0.0
0.0
5.0
4.0
0.0
0.0
2.6
10.0
3.0
0.0
Exten-
sion
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
6.0
0.0
0.0
0.0
0.0
0.0
9.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
1.3
8.0
0.0
0.0
Back-
fire
8.5
0.0
0.0
0.0
-0.0
0.0
1.5
0.0
0.0
0.0
13.5
0.0
21.0
0.0
21.0
0.0
3.0
6.0
16.5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Idle
Rough .
1.8
1.5
0.8
2.0
3.0
1.5
2.8
0.0
2.8
0.0
1.8
0.0
7.2
6.0
5.0
1.0
1.5
2.5
5.5
4.5
2.5
0.0
0.2
1.5
1,0
0.0
1.7
3.0
0.0
0.5
Total
323.1
96.0
132.7
58.0
208.2
35.3
296.0
9.5
550.2
91.0
263.1
12.5
212.6
67.5
289.9
24.8
554.5
53.0
241.7
33.5
27.3
0.5
358.7
14.5
250.5
8.2
73.5
87.2
123.5
7.5
* L - Low Volatility Fuel, H - High Volatility Fuel.
-------
recorded on strip charts. The instrument signals were also converted to
ditigal form and stored by computer on magnetic disk as the rate of five
times per second. Throughout each test the trained rater entered his
driveability ratings into the computer via a keyboard located in the car-
the keyboard was also used to indicate which maneuver was being attempted
A photograph of the driver's keyboard is shown in Figure 1, and examples
of computer printouts available for each driveability test are shown in
Figures 2 and 3. Figure 2 shows the value of each engine/vehicle parameter
and the status of all keyboard buttons every 0.2 seconds during a 10-second
segment of Test Number 26. Figure 3 shows a tabulation of rater-observed
demerits for Test 26. After an elapsed test time of 265.6 seconds the
driver opened the throttle to attempt the 2nd, 0-35 mph wide-open-throttle
acceleration of the test. The engine stalled at 268.0 seconds (manifold
vacuum was 0.0 in. Hg). At 269.0 seconds the rater depressed the "Stall"
button on the keyboard and then engaged the engine starter at 271.4 seconds
The engine started at about 272.4 seconds. Figure 3 shows a tabulation
of" rater-observed" demerits in"TestT2b.
Driveability Tests Conducted
Driveability tests conducted in this program have been segregated into six
p.hase_§_ described, by the chassis dynamometer used, the. ambient test tempera-
ture, and the driving cycle used. They are:
Ambient Driving
Phase Dynamometer Temperature, °F (°C) Cycle
I Large Roll 20 (-7) CRC
II Large Roll 70 (21) CRC
III Large Roll 20 (-7) Motorist
IV Small Roll 20 (-7) CRC
V Small Roll 20 (-7) FTP
VI Small Roll 70 (21) FTP
The CRC and motorist driving cycles are described in Appendix A; the FTP
cycle is the cycle currently specified by EPA to use for measuring light-
duty vehicle exhaust emissions and fuel economy (6). The Phase I tests
were conducted because the investigators felt this combination of dyna-
mometer, test temperature, and cycle held good potential for objective
driveability measurement. Test Phases II thru VI were conducted to
determine whether driveability could be measured either by raters or
instruments on other dynamometers, at different test temperatures or using
various driving cycles.
In all test phases, each of two trained raters evaluated car performance
several times on three test fuels. Two of the fuels were the same
throughout the entire program but the third fuel was inadvertently changed
between Phases I and II. Pertinent inspections on all test fuels are:
-------
Figure 1
Driver's Data Entry Keyboard
*9
Cycle Indicator
Driveabillty
Problem Keys
M«*
•
oo
Driving Maneuver
Indicator Lights
Problem Severity Keys
-------
Figure 2
Raw Data Printout
FUEL-
SEC.
264. 4
264. 6
264. 3
265. 0
265. 2
265. 4
265. 6
2.65.
2*6.
266.
.2*6.
266.
266.
267.
267.
267.
26 '••'
267.
'263.
263.
263.
263.
263.
269.
269.
269.
269.
269.
2V 0.
270.
2V 0.
270.
2VO.
27 1 .
271.
271.
2V 1 .
271.
•~- ' / —f
272.
2V2.
272.
2/2.
273.
2V '~<
273.
2/3.
273.
O
0>
03
•»
Ol
E
'o
(.1
2
4
4.
3
0
2
4
;'-.
;-;
(.1
2
4
6
3
0
'2
4
6
3
(j
CYC. MNVR.
ENG 2 0-3SWO
ENG 2 0-35WG
ENG 2 0-35WO
ENG 2 0-35WO
ENG 2 0-3SUO
ENG 2 0-35WG
ENG 2 0-35UIO
ENG 2 0-35WO
ENG 2
ENG 2
ENG i
ENG 2
'..I-35WO
0-35WG
0-35WO
0-35UG
ENG 2 0-35WO
ENG i
ENG 2
ENG 2
ENG 2
ENG 2
ENG 2
ENG 2
ENG 2
ENG 1
ENG 2
0-35WO
(.I-35UO
0-.55WO
(.I-35WO
0-35WG
O-35WO
0-35WO
0-35WO
0-35WO
0-35WQ
ENG 2 0-35WO
ENG 2
ENG 2
ENG 2
ENG 2
ENG 2
0-35WO
0-35WG
0-35WO
0-35WO
0-35WO
2 ENG 2 0-35WO
4
6
3
0
2
4
6
8
0
•^
4
6
3
0
2
4
6
3
_^
ENG 2
0-3SWO
ENG 2 0-35WG
ENG 2
ENG 2
U-35WO
0-35WO
ENG 2 0-35WO
ENG 2 0-35WO
ENG 2
ENG 2
0-35WO
0-35WO
ENG 2 U-35UIO
ENG 2
ENG 2
ENG 2
0-35WO
U-35WO
0-35WO
ENG 2 0-35UO
ENG 2 0-35WO
ENG 2
0-35WO
ENG 2 0-35WO
ENG 2 0-35WO
ENG I
4
Ol
1— 1
o
^
u
00
0-35UO
k
01
^
3
0)
C
CO
•H C 03 S
H
•H C
> O U-l
T>
01
03
a.
CO
•H-f
1-1 4.
c •*•
4.
U-l c
-i O
j
H 0)
j a
u >•
MPH
1. 1
1. 1
0. 3
0. 3
u. 3
0. 4
U. 4
0. 4
(.1. 4
0. 4
M. 4
0. 4
U. 4
0. 0
O. O
0. 0
U. I.1
0. 0
o. o
0. 0
0. O
0. 0
I.I. I.I
0. 0
o. o
0. 0
O.
,__)
o:
U-l
• H
C
CO
s:
0)
.*
CO
4-1
C
f-H
7
C
O
TURQ
155.
142.
145.
148.
145.
140.
148.
3 102.
3 131.'.
3
3
3
3
3
3;
3
3
;'.;
3
0
5
0
1.1
0
C.I
0
(.1
5
i.i
5
U
0
i.i
0
5
0
5
0
i.l
0
7
3
4
5
3
9
3
. — '
C
01
a.
1
-
-
-
-
-
-
-
-
-
-
_
-
-
_
-
-
-
-
-
-
-
-
-
—
-
-
1 8.
2O.
-'5.
(.1.
18.
15.
15.
15.
15.
15.
15.
15.
15.
13.
15.
15.
15.
15.
15.
15.
15.
13.
13.
13.
15.
15.
15.
15.
13.
13,
-10.
-5.
0.
i.i.
-5.
-3.
-3.
-
w*
0
C
•HO U
e"
3
U
CO
>
4J
03
O
cu
Ol
r-4
| I
4-1
O
t-l
J=
^s
O
H
4J
U-l
CO
.C
03
Ol
>r.
^
Q
10.
10.
V—1
\
XI
VH
1
U-4
1-4 OS > H
Ol
S
3
S
DEC. F". , RI_ltM 2• ^ ^
Reset +• -n-n
Trace (Not Used)— -*• -n-n
Sliohl Ifc -n-n
' F F F
""* O" " *" ~~ '
Moderate ^ -n -n
Heavy 1» ^ ^
Very-Heavy (Not used) -n-n
CH^I i
F F F F F F F
F F F F F F F
F F F F F F F
F F F F F F F
F F F F F F F
F F F F F F F
Idle Roughness -*. -n-n
Stumble fr -n -n
Extension ft» -mi
Hesitation -fr •" ~n
Surge •*> T, -n
Backfire . -» -n-n
Keyboard Button^Status
T ^ True (Button Depressed)
F =^ False (Button Not Depressed)
-------
10
Figure 3
Trained Rater-Observed Demerit Summary
•S»CMV—
1.9 DEO.
LE
0
1
1
1
1
1
1
TOTAL
2
2
2
2
2
2
MANEUVER
START
0-25PT
CRUISE
2S-35D
O-35WO
HJ-23P
IDLE
0-23F-T
CRUISE
2S-35D
U-35UIO
10-2SP
IDLE
O 1 .
TIME
7. 4
IJ. U
0. 0
u. u
o. o
IJ. IJ
0. 0
7. 4
0. 0
ij. tj
0. 0
0. 0
O. 0
IJ. U
IDLE
16.
U.
O.
0.
O.
IJ.
o.
16.
0.
i.i.
0.
1.1.
0.
IJ.
ACCEL.
0.
U.
0.
U.
32.
32.
0.
64
0.
U.
O.
32.
64.
U.
DECEL.
0.
O.
0.
U.
O.
O.
0.
0.
0.
IJ.
o.
u.
0.
u.
11M_C
ROUGHNESS
0.
U.
O.
u.
O.
i.i.
0.
II
0
1.1.
o.
I.I.
0.
2.
STUMBLE
0.
U.
O.
6.
6.
12.
0.
24.
6.
U.
72.
12.
12.
ii
EXTENSION
O.
U.
O
u.
u.
(J
O.
tl
o.
1.1.
n
(.1
, ,
ij
HESITATION
0.
U.
0.
i.i.
ij
O.
0
i.i.
0.
I.I
0.
I.I
0.
I.I
SUPOE
BACKFIRE
O.
U.
0.
0.
i.i.
0.
0.
12.
0.
TOTAL
U. U
O. O
6. ij
as. o
44. U
O. 0
111. 4
6. U
B8. O
36. U
TOTAL
3
3
3
3
3
3
TOTAL
4
4
4
4
4
4
U-2SPT
CRUISE
2S-35D
0-35UO
UJ-23P
IDLE
0-45CD
CRUISE
23-35D
U-3SUO
IDLE
TOTAL
b IJ-4SCD
5 CRUISE
5 23-35D
3 0-3SWO
S 1O-25P
IDLE
TOTAL
0. 0
U. U
O. O
U. U
0. O
O. IJ
0. 0
0. 0
U. 0
o. o
O. 0
0. 0
U. O
0. 0
IJ. IJ
0. 0
IJ. IJ
0. 0
o. o
o. o
6
6
6
6
6
'OTA
0-45CD
CRUISE
23-35D
U-3SWO
10-25P
L
O. O
IJ. 0
0. O
0. u
0. 0
0. 0
0.
o.
0.
0.
0.
u.
0.
IJ.
0.
IJ.
0.
o.
102.
36.
0.
24.
12.
0.
34.
12.
U.
24.
U.
0.
36.
U.
0.
12.
6.
6.
24.
30.
U.
0.
30.
U.
0.
i.i.
0
0.
IJ.
o.
16.
O
i.i.
O
0.
IJ.
IJ.
0
24
34.
7S.
IS
34.
0.
i.i.
72.
223. O
63. U
O. O
36 U
l^O. O
12. U
-------
11
Driveability Test Fuels*
135 (57)
240 (116)
360 (182)
105 (41)
185 (85)
305 (152)
115 (46)
220 (104)
330 (166)
128 (53)
220 (104)
330 (166)
Driveability,°F (°C) 1 2A 2B
10% Evap.
50% Evap.
90% Evap.
Reid Vapor Pressure 8.4 (57.9) 13.5 (93.1) 10.5 (72.4) 8.3 (57.2)
Ibs. (Kpa)
* Fuels 1, 2A and 3 were used for Phase I testing and Fuels 1, 2B and 3 were
used for Phases II thru VI.
Fuels 1, 2A, and 2B are the same fuels as the low, high, and average
volatility fuels, respectively, used in the 1980 CRC program on cold start
and warmup driveability. Fuel 3 was similar to Indolene used for EPA
certification testing.
RESULTS
Trained Rater Observations
Trained rater observed demerits tabulated by test number and driveability
problem type are shown in Appendix C. Total demerits for each test are
shown in Table III. A few general conclusions can be made based on the
information for Fuels 1 and 3 (the third fuel was changed after Phase I).
Perhaps most noticeable is the large standard deviations shown in the
table. They range from about 10 percent of the mean value to over
100 percent of the mean. Based on past experience, these standard
deviations are unusually high and we suspect some of the variability is due
to vehicle performance inconsistencies in addition to changes in rater
severity between runs. Comparing Phase I and Phase II results it appears
that driveability of this car at 70°F (21°C) is generally not as poor as at
20°F (-7°C) and, therefore 20°F (-7°C) would be a better temperature to use
for developing an objective driveability system. Comparing Phases I
and III it is interesting to note that use of the motorist driving cycle
appears to have markedly improved the standard deviations without large
changes in mean demerits.
Phases I and IV are identical except that they were conducted on
dynamometers of different design. Because the means and standard
deviations are nearly the same in both phases, we conclude that either
dynamometer could be used equally well for driveability testing. In
Phases V and VI the FTP driving cycle was used and ambient test
temperatures were 20 and 70°F respectively. Recall that the car used for
this test work had the poorest driveability of all those screened and yet
average driveability demerits when using the FTP cycle are very low.
Therefore, this cycle should not be used for driveability measurement
-------
12
TABLE III
RATER-OBSERVED DEMERITS BY TEST PHASE
Fuel:
Test Phase Rater:
I
Large Dyno
CRC Cycle
20°F (-7°C)
** Mean
Std. Dev. ,
% of Mean
II
Large Dyno
CRC Cycle
70°F(21°C)
Mean
Std. Dev. ,
% of Mean
III
Large Dyno
Motorist Cycle
20°F (-7°C)
** Mean
Std. Dev. ,
% of Mean
1
A
577.6
489.4
613.4
1374.4
90.0
700.0
823.4
717.6
275.6
295.2
444.0
332.2
230.2
508.8
167.6
509.3
63.2-
650.2
208.4
152.0
88.6
35.0
226.8
108.2
349.4
365.6
495.2
326.4
414.8
191.4
369.8
358.9
25.7
B
617.6
572.4
543.4
577.8
398.0
311.6
439.4
268.0
303.6
515.8
446.4
270.8
-
-
-
438.7
29.1
381.4
126.0
122.0
71.0
-
175.1
79.8
400.0
422.4
401.0
344.2
345.4
-
—
382.6
9.3
2A or
A
149.4
177.0
326.8
32.0
25.2
10.0
42.0
18.6
25.0
38.4
18.0
7.2
13.0
-
-
67.9
139.1
6.0
15.2
6.0
0.0
6.0
6.6
81.8
37.0
80.0
79.4
93.2
-
-
-
72.4
33.8
2B *
B
34.8
18.0
106.0
13.0
24.0
37.0
31.0
31.0
10.0
12.0
12.0
44.8
28.0
25.0
-
30.5
82.8
37.0
47.0
41.0
59.0
-
46.0
20.9
49.0
92.0
52.0
93.0
120.0
-
-
81.2
37.2
3
A
335.0
188.8
188.8
183.6
293.2
152.0
75.6
28.2
40.0
28.2
113.0
-
-
-
-
147.9
70.1
18.0
25.0
6.0
15.0
0.0
12.8
77.3
111.8
89.4
126.0
120.8
67.4
-
-
103.1
23.6
B
199.2
237 . 2
175.0
70.0
72.2
54.2
30.0
81.2
74.0
58.2
-
-
-
-
-
105.1
67.6
44.0
51.0
56.0
85.0
-
59.0
30.5
89.4
68.0
125.0
93.0
92.0
117.0
-
97.4
21.2
-------
13
TABLE III
- Continued -
RATER-OBSERVED DEMERITS BY TEST PHASE
Fuel:
Test Phase Rater:
IV
Small Dyno
CRC Cycle
20°F (-7°C)
** Mean
Std. Dev. ,
% of Mean
V
Small Dyno
FTP Cycle
20°F (-7°C)
** Mean
Std. Dev. ,
% of Mean
VI
Small Dyno
FTP Cycle
70°F (21°C)
** Mean
Std. Dev. ,
% of Mean
]
A
284
670
293
592
460
43
43
99
127
44
78
53
8
0
-
4
142
.0
.4
.0
.6
-
—
.1
.6
.6
.0
.8
.2
—
.7
.1
.0
.0
.0
.5
L
B
157
562
432
333
596
416
43
196
174
204
31
151
53
148
42
42
77
79
.0
.6
.6
.6
.0
-
.4
.0
.2
.6
.4
.8
-
.7
.3
.0
.0
.0
.3
.2
2A
A
151
258
293
244
102
88
189
45
12
42
36
15
26
56
42
1
32
25
84
or
.0
.4
.4
.4
.2
.6
.7
.9
.2
.2
.4
.2
-
.5
.6
.0
.2
.0
.1
.9
2B
B
275
136
138
129
169
41
0
6
10
18
8
88
22
53
-
37
58
*.
.0
.0
.0
.0
-
-
.5
.5
.0
.0
.0
.0
-
.5
.2
.0
.0
.5
.4
3
A
24
60
18
112
224
87
96
36
14
16
30
6
20
60
24
12
-
18
47
.0
.2
.2
.4
.4
-
.8
.9
.0
.0
.0
.0
.0
.4
.2
.0
.0
.0
.2
B
130.2
139.0
101.4
101.4
-
-
118.0
16.5
9.0
21.0
•7.0
-
-
12.3
59.2
8.0
23.0
-
15.5
68.4
* Fuel 2B was used for all tests except Phase I.
** Means and Standard Deviations calculated from all available observations.
-------
14
because it likely will not yield measurable driveability differences
between cars. There are two possible explanations for the low demerits.
First, the driver is forced to manipulate the throttle so vehicle speed
follows an established speed versus time chart. This tends to mask
driveability problems. Second, the FTP cycle is not severe enough
(acceleration rates are low) to disclose driveability problems.
The computer routines that were developed are based solely upon the data
collected in the Phase I tests; they are then applied to Phases II thru IV
data. Because the FTP cycle is not well suited to driveability testing, we
have not attempted to computer analyze the data from Phases V or VI.
Data Editing by Computer
An enormous amount of data was collected from the instruments and raters
and stored by computer. To correlate the raters' evaluations with the
engine/vehicle operating parameters, it was necessary to devise ways of
eliminating data collected when specific driveability problems could not
or should not occur. For measuring hard-starting (excessive cranking)
demerits, the computer only considers data collected during the start
maneuver. Recall that one of the keyboard entries by the rater was type of
maneuver being made. To find idle stalls the computer "looks" during the
start maneuver and all the idle maneuvers during the test. Conversely,
driving stalls could occur during any maneuver except start or idle. To
identify stumble and hesitation, the computer searches only those pieces
of the data collected while an acceleration was being attempted. This
means that first the computer completely ignored the start, cruise, and
idle maneuvers. Second, because the CRC driving cycle is a series of
maneuvers made at constant or continually increasing throttle opening, the
beginning, duration, and end of each maneuver is computer-defined by
throttle movements and position. The computer was programmed to recognize
that an acceleration begins when the throttle opens by 2 percent (of full-
throttle opening) or more in 0.2 seconds and continues as long as the
throttle is open more than 18 percent or until the throttle closes by
2 percent or more in 0.2 seconds. Third, the computer was also programmed
to ignore data collected during a transmission shift. This was done
because the investigator's previous experience indicated that some engine/
vehicle operating parameters respond to transmission shift and stumble in
much the same manner. Transmission shift is not considered a driveability
problem. A transmission shift occurs when all the following conditions are
met:
1. Vehicle speed is above 10 mph and increasing by 1 mph during a
1.0 second interval.
2. Engine speed declines by 50 or more rpm over the same time interval
as in "1".
3. Intake manifold vacuum declines during a 1.0 second interval which
begins 0.4 seconds later than the interval in "1".
-------
15
For any time interval in which these three requirements are met simultane-
ously, ._a.ll. data are excluded between the times when engine speed is maximum
and minimum (inclusive).
By using these editing rules, the amount of data to be searched for
driveability problems was reduced to a manageable size.
Stumble Measurement by Computer
Stumble is a sudden loss of power followed some time later by'a recovery of
power. By inspecting strip chart recordings of the engine/vehicle
operating parameters, engine speed and driveshaft torque were initially
selected for objectively measuring stumble. However, later analyses
showed that excluding torque from consideration improved the system's
stumble measurement ability. A brief description of this and several other
unsuccessful attempts to objectively measure driveability problems (in-
cluding stumble) are described in Appendix D.
Figure 4 graphically shows engine speed as a function of time during an
acceleration in which the rater noted one heavy stumble occurred. The
fluctuations or dips in engine speed are used for stumble measurement.
This is not a novel idea; engine speed was also used extensively for
objectively measuring stumble in a 1973 CRC program on driveability
instrumentation (7). Attempts were made to correlate various charac-
teristics of these dips with the severity (demerits) assigned the stumble
by the trained rater. The correlation finally selected is:
Rater Observed Stumble Demerits = bg + bj(At) + b2(Aa) + b3(At)(Aa)
where: b^'s are constants to be determined by regression analysis,
t and a are the time duration and amplitude, respectively, of a
dip (see Figure 4).
Two questions presented themselves, however:
1. Should At be limited only to the time interval between maximum and
minimum values of engine speed?
2. If the number of dips during an acceleration is greater than the
number of stumble evaluations by the rater, which dip should be
paired with the evaluation for the regression analysis (Figure 4
shows a case like this)?
To provide flexibility on the time interval question, part of the
"shoulder" immediately prior to the dip is included in the At calculation.
To do this, an RPM slope cut-off was established which effectively allows
the dip to begin when the RPM slope drops below this value. The value of
the slope cut-off (variable S) was not known and had to be determined.
-------
16
Figure 4
Hesitation and Stumble(s) During an Acceleration
Phase I test #10, 4th 25-35 mph Detent
1500
1400
E 1300
Q.
•o
03
-------
17
The problem of a rater identifying fewer stumbles than dips was handled
with three rules. First, if the end of one dip and the beginning of the
next dip were close to one another the dips were grouped together and the
At and the Aa values of the individual dips were summed together. However,
the amount of time to allow between dips without grouping them was unknown
and had to be determined. This variable was named association interval
"I". Second, through trial and error, it was discovered that many small
dips not noticed by the raters could be eliminated from consideration by
imposing some minimum requirements. Dips are ignored if the following
three statements are all true:
1. A t < 0.4 sec
2. A a < 68.5 rpm
3. (At)(Aa) < 207
Third, of those dips remaining after applying these rules, it was necessary
to find which ones to "pair" with the raters' evaluation of stumble (if
any). To be paired with a dip, the drivers' keyboard entry of stumble could
not preceed the beginning of a dip and both the keyboard entry and the dip
had to occur within the same driving cycle maneuver. Of those dips which
satisfy all these rules, the drivers' stumble evaluation was paired with
the dip having the largest product of At and Aa and the remaining dips were
paired with an assumed stumble demerit rating of zero.
A series of regressions were conducted to find the "best" values for the
association interval (I), the slope cutoff (S), and the regression co-
efficients (b^'s) in the equation above. The Phase I driveability data
were used for the regressions and included over 500 dips that potentially
could be associated with rater observations of stumble. To run the
regressions, values for S and I were manually entered into the computer.
Next, the computer went through the data determining all the dip group-
ings and stumble pairings and last conducted a linear least squares
regression to determine the stumble equation coefficients (bj_'s). The
best values for these variables are:
S = 98 rpm/sec
I = 0.2 sec
Computer Calculated Stumble Demerits = 9.107-1. 76(At) + 0.0048( Aa) + 0.00377(At)(Aa)
This demerit equation and these values for "S" and "I" were next used to
calculate stumble demerits for each driveability test conducted. Figure 5
is a comparisons of the trained rater-observed and computer-calculated
stumble demerits for all engine speed dips in Phase I Test Number 73.
Typically the computer and rater seldom agree exactly on the demerits to
assign a given dip or stumble. This is partly because the rater is forced
to put his evaluations into one of four categories having demerits of 0, 6,
12 or 24 (severities of none, slight, moderate and heavy, respectively).
The computer on the other hand assigns demerits using a continuous
function. Another part of the rater versus computer discrepancy is caused
by inconsistent severity assignments by the rater whereas the computer,
given a set of rules, is very consistent in assigning demerits.
-------
18
Figure 5
Stumble Demerits in Test 73 (Phase I)
30
25
.§
0>
CD
•o
•o
0)
+rf
(0
20
15
(0
o
1_
0)
a
E
o
O
10
I »
I
I
I
5 10 15 20 25
Trained rater observed demerits
3G
-------
19
Figure 6 is a comparison of rater and computer stumble demerits for
Phase I tests. Each data point represents total stumble demerits for
one driveability test. In this case the two are in good general agree-
ment but the computer tends to underpredict at high demerit levels and
overpredict at lower levels. Another regression of computer-versus-rater
demerits could have been conducted to improve the agreement between the
computer and rater but this was not done because the result certainly
would only apply to the car tested in this program and the general con-
clusions would remain unchanged.
Compared with the rater evaluations, the computer calculations yield a
narrower range of average demerits between low and high volatility fuels.
This does not mean the computer method is less able to measure performance
difference between fuels. Because the standard deviations of the computer
averages are lower than those of the trained rater averages (Table IV),
the difference between fuels is measured with greater confidence by the
computer than by the raters. These analyses show that the computer system
developed can adequately measure stumble demerits for this car; whether
this system can be used for other cars cannot be determined without further
testing beyond the scope of this contract.
Hesitation Measurement by Computer
Hesitation is a momentary lack of response to opening the throttle. Again,
after inspecting strip charg recordings, engine speed was selected as the
best parameter to use for detecting and measuring hesitation. Following
several futile attempts, described in Appendix D, a method was developed
which correlates rater-observed hesitation with the rates of throttle open-
ing, and engine speed increase during the initial 1.0 second of an acceleration,
and with the vehicle speed immediately before the start of an acceleration. It
was theorized that the raters' opinion of hesitation was primarily influenced
by how rapidly the engine speed initially responded to the throttle movement
and less influenced by the response later in the one-second interval. The
relationship form is:
WTTL
Rater Observed Hesitation Demerits = bn + bi (-— - — ) + bo WRPM
fir ri|". T U . -5o
where: MPHQ is vehicle speed 0.2 seconds prior to the acceleration start
WRPM = 0.5(ARPM1)+0.25(ARPM2)+0.13(ARPM3)+0.06(ARPM4)+0.06(ARPM5)
WTTL = 0 . 5 (ATTLi ) +0 . 25 (ATTL2 ) +0 . 1 3 ( ATTL3 ) +0 . 06 ( ATTL4 ) +0 . 06 ( ATTL5 )
i is the engine speed (rpm) increase during the ic^ 0.2-
second time interval following the start of an acceleration
i is the throttle opening (expressed as percent of wide-
open) increase during the ic" 0.2-second time interval
beginning 0.2 seconds before the start of an acceleration
This demerit calculation scheme places strong emphasis upon the first
0.2 seconds of an acceleration and progressively less emphasis upon
following time intervals. The data used for this regression analysis is
shown in Appendix E. It consists of the 27 accelerations during the Phase I
-------
20
Figure 6
Stumble Demerits for Phase I Tests
350
300
.§250
*Z
0)
E
0>
•O
•O 200
0)
+rf
JO
3
O
150
0)
<-*
3
a
100
50
I
I
50 100 150 200 250
Trained rater observed demerits
300
350
-------
21
TABLE IV
STUMBLE DEMERITS BY TEST FUEL
Fuel 1
(Test Phase I)
Fuel 2A
Stumble
Test
Number
1
2
3
6
9
10
12
13
19
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
•>»Mean
Std. Dev. ,
% of Mean
Demerits
Rater
168
180
330
336
738
306
312
90
294
300
270
354
150
138
96
144
228
162
114
90
162
108
180
66
192
138
217.2
63
Comp.
_ *
-
-
316.9
'
215.3
288.5
123.2
242.0
-
165.6
-
109.2
81.3
116.2
135.9
202.7
113.5
129.6
81.9
133.2
123.8
98.5
84.1
175.2
133.1
153.5
44
Test
Number
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58
61
64
66
69
72
74
80
81
Stumble
Demerits
Rater
18
18
126
162
24
210
12
18
30
30
30
30
6
6
6
12
0
6
6
24
12
6
18
24
6
18
6
32.0
158
Comp .
71.2
61.7
135.6
188.7
193.3
62.8
76.3
57.1
58.1
41.2
49.8
38.4
21.7
33.0
42.0
45.3
7.3
34.7
23.6
28.2
36.0
19.9
16.9
39.4
14.9
46.3
19.6
54.2
87
Fuel 3
Stumble
Test Demerits
Number
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
Rater
102
204
150
132
102
120
210
102
54
60
18
30
12
18
6
18
48
36
42
36
113,
171,
111.
95.
71.2
89.
124.
109.
65.
51.
30.3
38.0
45.9
33,
43,
61,
153.9
40.5
45.1
75.0 78.6
83 54
* Dashes indicate that the test data could not be computer analyzed".
** Means and Standard Deviations are calculated from all available observations.
-------
22
tests in which the trained raters said hesitation occurred and another 36
accelerations made when the car was fully warmed-up and no hesitations were
recorded by the rater. The demerit calculation equation resulting from the
linear-least squares regression on this data is:
WTTL
Hesitation Demerits = 7.5 + ^S'^MPH—+ Q 58^ ~ °'033 WRPM
To improve the agreement between rater-observed and computer-calculated
hesitation demerits, a few empirical requirements were established which
must be met before hesitation demerits are calculated for an acceleration.
Demerits are calculated if:
1. the engine was still running 3.0 seconds after the start of the
acceleration (intake manifold vacuum >0.0" Hg) or
2. WRPM is less than 140 or
3. a. WRPM
WTTL
is 7 or less and
b. MPHQ is 0.2 or less
In general, the equations and rules assign hesitation demerits to those
accelerations when the engine speed increase was "abnormally slow". The
correct engine speed increase is defined by the rate and final amount of
throttle opening and by car speed at the start of the acceleration. For
example, an acceleration with slow engine speed increase will be assigned
fewer demerits if the throttle opening-rate is slow rather than fast.
Results of applying this method to the Phase I data are shown in Figure 7
and Table V. Each data point in the figure represents the total hesitation
demerits for one driveability test. The figure shows that total hesitation
demerits calculated for each test do not agree with the rater observed
values. One obvious explanation for this is that other independent
variables should be included in the prediction equation. Another possible
reason for the discrepancy is that raters may have mis-named some
hesitations as stumble and vice-versa — often these driveability problems
are difficult for a rater to distinguish.
In studying Table V, there is another fact which comes to light. Average
computer-demerits are considerably higher than the rater-observed values
but, the standard deviations (as percent of the mean) of the computer
values are much less than for the rater values. Additionally, the computer
method, like the rater, recognizes a difference in performance of low-
volatility Fuel 1 and high-volatility Fuel 2A.
-------
23
Figure 7
Hesitation Demerits for Phase I Tests
100
' 80
-------
24
TABLE V
HESITATION DEMERITS BY TEST FUEL
Fuel 1
(Test
Hesitation
Test
Number
1
2
3
6
9
10
12
13
19
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
*"'Mean
Std. Dev. ,
% of Mean
Demerits
Rater
36
66
102
0
54
18
12
0
0
0
0
12
0
0
0
0
0
0
0
0
0
0
0
6
0
0
11.8
215
Comp
_ *
-
62
75
99
73
63
12
76
38
77
76
65
47
37
83
67
46
57
34
56
57
76
49
50
51
59.6
31
Test
Number
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58
61
64
66
69
72
74
80
81
Phase I)
Fuel 2 A
Hesitation
Demerits
Rater
0
0
18
12
12
0
0
0
6
0
0
0
0
0
0
0
0
0
12
0
0
0
0
0
0
0
6
2.2
223
Comp
30
27
25
28
97
42
28
57
41
49
36
12
30
11
26
18
62
48
32
45
39
57
29
43
0
19
18
35.1
55
Fuel 3
Test
Number
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
Hesitation
Demerits
Rater Comt
0
6
0
6
6
0
0
0
0
0
6
0
0
0
0
6
0
0
0
0
78
69
74
53
40
69
57
52
37
44
68
52
71
54
53
47
24
57
52
33
1.5 54.2
178 26
* Dashes indicate that the test data could not be computer analyzed.
** Means and Standard Deviations are calculated from all available observations
-------
25
Because of the large discrepancy between rater and computer demerits, the
hesitation method developed in this investigation is inadequate and
additional effort needs to be expended on this problem, first using the
data collected in this program and then with data from other cars.
Engine Stall Measurement by Computer
An engine stall is any time when the engine quits running with the ignition
switch in the "on" position. Developing a method of detecting engine
stalls was relatively simple. The computer searches the data to:
1. find when the starter was engaged,
2. determine the type of driving maneuver being attempted when the
engine stalled,
3. read engine speed following each starter engagement.
After the initial startup, 8 demerits were assigned whenever the starter
was engaged if it occurred during the start or idle maneuvers, and 32
demerits if an acceleration or cruise was being executed. Engine speed
was used simply to ensure that the car restarted between starter engage-
ments. If engine speed did not reach 500 rpm between attempted starts,
then the computer treated the first engagement as a false start and
assigned no demerits. The computer and rater comparisons of total stall
demerits by test is shown in Figure 8 and in Table VI. From the figure it
is easy to see that the computer and trained rater stall demerits agreed
very closely. The cases of disagreement were caused by various factors
but primarily they resulted from improper data input by the trained rater
or instrumentation failure during the test. These procedures for detecting
engine stalls are perfected for future use.
Engine Idle Roughness Measurement by Computer
Engine idle roughness is the degree of smoothness perceived by a driver
while the engine is idling. The method developed for measuring engine idle
roughness is also based upon engine speed fluctuations. In this case only
the start and idle maneuvers are computer-inspected for idle roughness.
Each of these maneuvers is divided into a series of concurrent five-second
intervals. Within each interval the computer searches for the minimum and
maximum value of engine speed. From these speed ranges the computer selects
the broadest one for the start and each idle maneuver. A least squares
regression was run using the speed range and rater observed demerits from the
few maneuvers when the rater noticed idle roughness. The resulting
equation is:
Idle Roughness Demerits = -1.0 + 0.038 (Max speed range)
-------
26
Figure 8
Stall Demerits for Phase I Tests
325
300
275
| 250
o
j= 225
0)
"° 200
•o
0)
** 175
JO l/0
J 150
CO
2 125
0)
100
o
O
75
50
25
0
I
* Indicates number of observations
I I I I l I I I
25 50 75 100 125 150 175 200 225 250 275 300 325 350
Trained rater observed demerits
-------
27
TABLE VI
STALL DEMERITS BY TEST FUEL
Fuel 1
(Test Phase I)
Fuel 2 A
. Stall
Test
Number
1
2
3
6
9
10
12
13
19
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
-•-Mean
Std. Dev. ,
% of Mean
Demerits
Rater
272
224
136
208
272
208
176
0
320
304
272
240
208
144
168
112
144
200
112
104
112
296
264
80
208
104
188.0
43
Comp.
312
232
136
216
256
208
176
0
320
304
272
240
240
144
168
104
144
200
112
104
112
296
264
80
208
104
190.5
43.
Test
Number
4
5
7
8
11
14
17
18
22
. 28
29
33
39
41
46
54
54
56
58
61
64
66
69
72
74
80
81
Stall
Demerits
Rater
0
0
0
0
64
32
0
0
0
0
0
0
0
0
0
0
0
32
0
0
32
32
0
0
0
0
0
7.1
228
Comp .
0
0
0
0
64
32
0
0
0
0
0
0
8
0
0
0
0
0
0
0
32
0
0
0
0
0
0
5.0
292
Test
Number
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
Fuel 3
Stall
Demerits
Rater
96
104
32
96
64
32
64
32
0
0
0
32
0
32
0
0
32
32
64
0
96
112
32
96
64
40
64
32
0
8
0
32
8
32
0
0
32
40
64
0
35.6 37.6
98 93
•'•'" Means and Standard Deviations are calculated from all available observations.
-------
28
Rules are applied to eliminate data collected during engine stalls or when
the transmission is placed in gear because either one creates very large
speed _rangesa_nd abnormally large _idle roughness demerits. Comparisons^
between computer and trained rater demerits for idle roughness are shown in
Table VII and Figure 9. Average demerits by the computer compare favorably
with the trained rater evaluations (Table VIII), but totals for individual
tests do not agree well as shown in Figure 9. Because idle roughness
contributes very little to total demerits, the 'method developed is
adequate for this car but more effort is required using additional cars.
Hard Starting Measurement by Computer
Hard starting is excessive cranking during start-up. To measure this..prob-
lem, only jhe start maneuver is investigated by the computer. The computer
records the amount of time (seconds) the starter is engaged and after
subtracting 2 seconds, the result is hard starting demerits. Because this
is such a straight-forward measurement, the raters only measured crank
times during the first few Phase I tests to ensure the computer was making
the proper calculations. Hard starting demerits are listed in Table VIII.
No comparison between rater-observed and computer-calculated demerits is
possible, but this method is perfected for future use.
Total Demerit Measurement - Phase I Tests
Figure 10 and Table IX show comparisons of computer-calculated and rater-
observed demerits totaled across the five driveability problems for which
objective measurement methods were developed. The figure shows that total
rater and computer demerits follow the same trend but the computer demerits
are generally higher. Table IX supports this conclusion for Fuels 2A and 3,
but not for Fuel 1 because of the 6 tests listed in the table which are not
shown on the figure (tests 1, 2, 3, 9, 26, and 31 could not be computer-
analyzed). Excluding these would lower the mean rater demerits from 425.6
to 359.6 and reduce the standard deviation from 48 to 41 percent of the
mean.
As expected from results on the 5 individual driveability problems, the
computer calculation methods yield lower standard deviations than the
rater observations although the reduction is negligible for Fuel 1.
Furthermore, the computer methods apparently can distinguish between
performance differences of fuels.
Demerit Measurement - Phase II, III and IV Tests
In Appendix F the rater-observed and computer-calculated demerits are
tabulated for each driveability test conducted in Phases II, III and IV.
In general, these data further support the above conclusions about
adequacy of the computer methods. For brevity, only total demerits in
Phases II-IV will be discussed; these data are presented in Table X.
-------
29
TABLE VII
IDLE ROUGHNESS DEMERITS BY TEST FUEL
Fuel 1
(Test Phase I)
Fuel 2 A
Idle Roughness
Test
Number
1
2
3
6
9
10
12
13
19
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
-"-Mean
Std. Dev
Demerits
Rater
7
9
1
5
18
13
6
0
5
4
3
0
1
3
3
1
9
10
7
0
7
1
6
0
8
5
5.3
• >
% of Mean 81
* Dashes
indicate that
Comp .
*
-
2.6
-
-
5.5
9.5
10.0
1 76.7|
-
5.8
-
3.6
2.3
8.5
5.4
10.9
6.7
10,2
4.7
8.4
0.1
4.6
4.8
6.3
6.3
6.1
48
the test
Test
Number
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58
61
64
66
69
72
74
80
81
Fuel 3
Idle Roughness
Demerits
Rater
1
0
1
3
0
0
1
2
1
1
1
2
1
0
0
0
0
0
0
1
0
0
0
0
0
1
0
0.6
133
data could not
Comp .
0.4
2.5
6.4
0.9
-
1.3
1.9
2.1
1.1
0.4
1.8
1.6
1.1
3.0
1.9
2.6
2.0
1.2
1.3
0.8
2.9
0.8
1.1
1.4
1.5
7.8
2.5
2.0
83
be computer
Test
Number
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
analyzed .
Idle Roughness
Demerits
Rater
1
6
0
3
3
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0.8
194
Comp.
3.2
9.1
2.1
118.91
3.1
I12.3J
3.1
0.7
1.7
1.4
0.6
1.7
0.6
6.8
1.6
1.3
0.8
0.0
1.7
1.0
2.2
104
** Means and Standard Deviations are calculated from all available observations
except boxed values which appear to be outliers.
-------
30
Figure 9
Idle Roughness Demerits for Phase I
25
20
CD
0)
0>
15
3
_O
CO -ft
O 10
CD
Q.
I B
O
12
"Indicates number of observations
I I I
5 10 15 20
Trained rater observed demerits
25
-------
31
TABLE VIII
HARD STARTING DEMERITS* BY TEST FUEL
(Test Phase I)
Fuel
Test
Number
1
2
3
6
9
10
12
13
19
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
••'•Mean
Std. Dev. ,
% of Mean
1
Hard
Starting
Demerits
4.6
4.4
2.4
16.6
4.4
1.4
3.4
0.0
3.0
7.4
2.8
13.6
3.0
2.6
4.6
2.2
2.4
2.0
3.0
2.2
2.6
1.8
1.8
3.6
2.4
1.8
3.8-
95
Fuel
Test
Number
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58
61
64
66
69
72
74
80
81
2A
Hard
Starting
Demerits
15.8
0.0
4.4
0.0
0.0
0.8
0.0
0.0
0.0
0.0
0.0
0.0
14.2
0.0
0.0
0.0
0.0
0.0
0.6
0.0
0.8
0.4
0.0
0.0
1.2
0.0
1.0
1.4
286
Fuel
Test
Number
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
3
Hard
Starting
Demerits
0.2
3.0
0.8
0.2
0.0
7.6
0.2
0.0
0.0
15.6
0.2
0.2
16.0
0.2
0.2
0.0
0.2
0.0
1.0
0.2
2.3
215
* Starting time used for calculating hard starting demerits was not
measured by the trained rater — only by the computer. The hard
starting demerits shown are included in the total for both the trained
rater and computer.
** Means and Standard Deviations are calculated from all available
observations.
-------
32
Figure 10
Total Demerits for Phase I Tests
700
600
0)
-------
33
TABLE IX
TOTAL DEMERITS* BY TEST FUEL
(Test Phase I)
Fuel 1
Test
Number
1
2
3
6
9
10
12
13
19
. 26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
***Mean
Std. Dev. ,
% of Mean
Total
Demerits
Rater
487.6
483.4
571.4
565.6
1086.4
546.4
509.4
90.0
622.0
615.4
547.8
625.6
362.0
287.6
271.6
259 . 2
383.4
374.0
236.0
196.2
283.6
406.8
451.8
155.6
410.4
248.8
426.1
48
* Total demerits
starting.
** Dashes indicate
Com
630
503
540
145
647
523
420
277
334
330
427
368
311
226
312
478
444
221
441
296
394
34
for
that
Test
p. Number
_**
-
-
.6
-
.2
.4
.2
.1
-
.2
-
.8
.2
.3
.5
.1
.2
.8
.8
.2
.7
.9
.5
.5
.3
.1
stumble ,
the test
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58
61
64
66
69
72
74
80
81
Fuel 2 A
Total
Demerits
Rater
34.8
18.0
149.4
177.0
100.0
242.8
13.0
20.0
37.1
31.0
31.0
32.0
21.2
6.0
6.0
12.0
0.0
38.0
18.6
25.0
44.8
38.4
18.0
24.0
7.2
19.0
7.0
43.4
133
hesitation,
data
Com
117
91
171
217
356
138
106
116
100
90
87
52
75
47
69
65
71
83
57
74
110
78
47
83
17
73
41
97
Test
p . Number
.4
.2
.4
.6
.3
.9
.2
.2
.2
.6
.6
.0
.0
.0
.9
.9
.3
.9
.5
.0
.7
.1
.0
.8
.6
.1
.1
.8
67
stalls, idle
could not be
computer
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
Fuel
3
Total
Demerits
Rater
199
323
182
237
175
159
275
134
54
75
24
62
28
50
6
24
81
68
107
36
115
79
roughness
.2
.0
.8
.2
.0
.6
.2
.0
.0
.6
.2
.4
.0
.2
.2
.0
.2
.0
.0
.2
.2
and
Comp.
290.5
364.6
220.6
246.8
178.5
208.4
249.2
193.8
103.9
120.1
99.1
-
133.6
138.9
88.0
91.5
118.3
250.9
159.2
79.3
175.5
45
hard
analyzed for one or
*** Mean and Standard Deviations are calculated from all available observations".
-------
TOTAL
Test
Test Phase Number
II 4
5
8
11
13
15
17
20
23
""Mean
Std. Dev. ,
% of Mean
III 28
31
34
36
40
43
46
49
50
53
58
59
*Mean
Std. Dev. ,
% of Mean
IV 61
65
68
70
72
77
80
83
86
'"Mean
Std. Dev. ,
% of Mean
Fuel 1
34
TABLE X
DEMERITS BY
TEST FUEL
(Test Phases II-IV)
Fuel
Total
Demerits
Rater
590.2
377.4
202.4
152.0
102.0
90.0
78.6
31.0
39.0
184.7
100
339.4
245.6
338.0
412.4
387.2
305 . 0
302.2
148.4
222.4
309.4
177.4
307.8
291.3
27
127.0
482.0
224.0
566.4
318.6
249.6
179.0
370.6
350.0
318.6
45
Com
390
338
290
165
123
135
107
222
52
312
234
321
278
133
255
30
162
602
272
549
407
293
257
413
329
365
39
P •
_
.5
.3
.4
.6
.9
.1
.9
-
.0
_
.2
.7
-
-
-
.4
-
.1
-
.2
-
.9
.5
.8
.4
.7
.7
.5
.8
.0
.8
.5
Test
Number
1
2
7
9
14
18
21
25
27
29
32
38
41
44
47
51
54
56
60
62
64
66
71
73
75
78
81
85
2B
Total
Demerits
Rater
6.
92.
97.
108.
23.
0
6.
29.
39.
44.
97
37.
45.
84.
74.
46.
63.
81.
92.
57.
64.
30
79.
190.
253.
212.
309.
94.
90.
96.
76.
93.
149.
57
0
0
0
0
0
0
0
0
4
0
0
0
0
0
4
0
0
2
4
0
4
4
4
0
0
2
0
6
0
4
Com
103
127
51
54
57
63
123
60
56
77
40
120
48
131
133
111
140
143
140
121
26
46
201
232
175
266
130
121
153
130
133
159
39
.8
.6
.8
.1
.3
.1
.0
.6
.2
.5
.7
.3
.6
.3
.6
.9
.8
-
.2
.3
.1
.3
.3
.5
.9
.7
.5
.1
.5
.7
.2
Test
Number
3
6
10
12
16
19
22
24
26
30
33
35
37
39
42
45
48
52
55
57
63
67
69
74
76
79
82
84
Fuel 3
Total
Demerits
Rater
18.0
36.0
24.0
23.0
6.0
15.0
0
36.0
65.0
24.8
78
79.8
89.4
64.0
111.0
57.4
80.0
93.0
88.8
82.0
101.0
39.4
80.5
25
24.0
60.2
18.2
106.2
90.4
89.0
146.4
84.4
77.4
55
Comp.
215.8
90.2
107.1
33.3
143.1
115.2
-
30.4
60.5
99.5
62
125.2
-
114.6
112.8
145.7
133.9
135.4
140.3
125.4
-
112.1
126.3
10
75.5
83.6
60.1
1.40 . 8
157.5
186.8
167.4
236.9
138.6
48
Means and Standard Deviations are calculated from all available observations.
-------
35
Figures 11, 12 and 13 are comparisons of rater and computer total demerits
for Phases II, III and IV, respectively. As with the Phase I results
(Figure 10) each of these show that total demerits calculated by computer
generally increase with rater-observed values and at low demerit levels
the computer tends to calculate higher demerits than observed by the rater.
Insufficient data are available for Phases II and III to draw firm
conclusions about the relative magnitude of computer and rater demerits at
higher demerit levels. However, the Phase IV data agree with Phase I in
that generally the computer demerits are larger than rater demerits over a
broad range of demerits.
Consolidation of Computer Methods
The five computer methods described thus far (one for each driveability
problem) were developed independently. The data were first analyzed
for stumble then reanalyzed for hesitation and so on. This requires
about 1-1/2 hours'" of computer time to completely analyze each drive-
ability test of only 20 minutes duration. To reduce the needed computer
time- and to compile the various computer schemes into a single "package,"
the individual analyses were consolidated into one large computer program
capable of analyzing an entire driveability test with one pass through
the data. This reduces the analysis time to only about 20 minutes per test.
Only the Phase I data were analyzed using the consolidated program —
results are tabulated in Appendix G. The one major drawback of the con-
solidated program is that relative to the independent analysis method,
computer-calculated and rater-observed stumble demerits do.not agree well,
as can be seen by comparing Figures 6 and 14. When the data were inde-
pendently analyzed for the five driveability problems, it was possible for
the computer to assign both hesitation and stumble demerits to a short
interval of data at the start of an acceleration. It was decided that this
should be prohibited in the consolidated program by giving preference to
hesitation. Stumble demerits were only assigned to segments of data where
hesitation was not detected. Compared with the independent analysis methods,
it was expected that the consolidated program would calculate fewer stumble
demerits, but the magnitude of the reduction (nearly 50 demerits average)
was larger than expected. A similar reduction in total demerits also re-
sults from using the consolidated program. Additional time is necessary to
perfect the consolidated program.
- The computer being used for these analyses has memory capability of only
64K bytes. Analysis time could be reduced by at least one order of
magnitude if a computer with more memory (e.g. an IBM 370) were used.
-------
36
Figure 11
Total Demerits for Phase II Tests
500
0)
•£ 400
0)
0)
ID 300
(0
[J 200
0)
Q.
E
o
100
_L
100 200 300 400
Trained rater observed demerits
500
-------
37
Figure 12
Total Demerits for Phase III Tests
500
(0
'= 400
0)
0)
•o
"2 300
5
(0
2 200
0)
+*
a
o 100
I
I
I
100 200 300 400
Trained rater observed demerits
500
-------
38
Figure 13
Total Demerits for Phase IV Tests
600 -
Trained rater observed demerits
-------
39
Figure 14
Stumble Demerits Using Consolidated
Computer Program (Phase I Tests)
300
03
^
a
o
O
250
V
§ 200
•o
03
JS
3 150
"(0
O
100
50
@696
Indicates number of
observations
I
I
50 100 150 200 250 300
Trained rater observed demerits
-------
40
REFERENCES
1. "Evaluation of a High Temperature Driveability Test Procedure - 1971
Yuma Program," Coordinating Research Council Report No. 455, June,
1973.
2. "Driveability Performance of 1975 Passenger Cars at Intermediate.
Ambient Temperatures - Paso Robles," Coordinating Research Council
Report No. 486, May, 1976.
3. "Driveability Performance of 1975 Passenger Cars at High Ambient
Temperatures," Coordinating Research Council Report No. 490, November,
1976.
4. "Driveability Performance of 1977 Passenger Cars at Intermediate
Ambient Temperatures - Paso Robles," Coordinating Research Council
Report No. 499, September, 1978.
5. "Light Duty Vehicle Driveability Investigation," Environmental Pro-
tection Agency Report 460/3-78-012 , Toulmin, H. A., Jr., Suntech, Inc.,
December, 1978.
6. "Code of Federal Regulations, Title 40 - Protection of Environment,
Parts 81-99," Office of the Federal Register, U.S. General Services
Administration, July 1, 1980.
7. "1973 Driveability Instrumentation Tests," Coordinating Research
Council Report No. 489, November, 1976.
8. "Comparison of Vehicle/Fuel Test Procedures with Customer Driving,"
Klen, D. S. , Amoco Oil Company, SAE paper 810491, presented at SAE
Congress and Exposition — Detroit, Michigan, February 23-27, 1981.
-------
APPENDIX A
I. CRC Cold Start and Driveaway Test Procedure
II. Motorist Driving Cycle
-------
A-I - 1
I. CRC COLD START AND DRIVEAWAY TEST PROCEDURE
TEST PROCEDURE AND DATA RECORDING
A. Start engine per Owner's Manual Procedure. Record start time.
B. If engine fails to start after 15 seconds of cranking, stop cranking
and depress accelerator pedal to the floor once and release. Repeat
procedure until engine starts. Record total cranking time.
C. Record idle quality in "Neutral" or "Park" immediately after start;
foot should be removed from accelerator pedal.
D. If engine stalls, repeat Steps A and B. Record number of stalls.
E. Allow engine to idle 15 seconds. Apply brakes (right foot), shift
transmission to normal drive range, and record idle quality. If
engine stalls, restart immediately. Record number of stalls. Idle
5 seconds in "Drive".
This completes the start-up portion of the procedure. Note that space
on only three restarts at idle are to be noted. After the third stall,
manipulate throttle to keep engine running. Proceed to next
maneuver.
F. Drive through the cycle shown in Figure A-I-1.
G. During each maneuver observe and record the severity of any of the
following malfunctions (see definitions):
1. Hesitation
2. Stumble
3. Surge
4. Stall
5. Backfire
6. Idle Roughness
DEFINITIONS AND EXPLANATIONS
A. Maneuver
A specified single vehicle operation or change of operating con-
ditions (such as idle, acceleration or cruise) that constitutes one
segment of the driveability driving schedule.
-------
A-I - 2
B. Cruise
Operation at a prescribed constant vehicle speed with a fixed
throttle position on a level road.
C. Wide Open Throttle (WOT) Acceleration
"Floorboard" acceleration through the gears from prescribed starting
speed. Rate at which throttle is depressed is to be as fast as
possible without producing tire squeal or appreciable slippage.
D. Part Throttle (PT) Acceleration
An acceleration made at any defined throttle position, or consistent
change in throttle position, less than WOT. Several PT accelerations
are used. They are:
1. Light Throttle (Lt Th) - All light throttle accelerations are
made by holding throttle position constant throughout the ac-
celeration. The throttle position selected is one which allows
the car to accelerate 0-25 mph in 0.1 mile when car engine is
warm.
2. Crowd - An acceleration made at a constant intake manifold
vacuum. To maintain constant vacuum, the throttle opening must
be continually increased with increasing engine speed. Crowd
accelerations are performed at the manifold vacuum which ini-
tially exists for the light throttle acceleration.
3. Detent - All detent accelerations are made at constant throttle
position. The throttle opening is the downshift position.
E. Malfunctions
1. Stall - Any occasion during a test when the engine stops with the
ignition on. The three types of stall, indicated by maneuver
being attempted, are:
a. Stall; idle - Any stall experienced when the vehicle is not in
motion, or when a maneuver is not being attempted.
b. Stall; maneuvering - Any stall which occurs during a pre-
scribed maneuver or attempt to maneuver.
c. Stall; decelerating - Any stall which occurs while decele-
rating between maneuvers.
-------
A-I - 3
2. Idle Roughness - An evaluation of the idle quality or degree of
smoothness while the engine is idling.
3. Backfire - An explosion in the induction or exhaust system.
4. Hesitation - A temporary lack of vehicle response to opening of
the throttle.
5. Stumble - A short, sharp reduction in acceleration after the
vehicle is in motion.
6. Surge - Cyclic power fluctuations occurring during acceleration
or cruise.
F. Malfunction Severity Ratings
The number of stalls encountered during any maneuver are to be noted
by the rater. Each of the other malfunctions must be rated by
severity. The following definitions of severity are to be applied in
making such ratings:
1. Slight - A level of malfunction severity that is just discernible
to a test driver but not to most laymen.
2. Moderate - A level of malfunction severity that is probably
noticeable to the average layman.
3. Heavy - A level of malfunction severity that is pronounced and
obvious to both test driver and layman.
The rater enters his evaluations into computer storage by depressing
the appropriate keys on the keyboard.
DEMERIT CALCULATIONS
Driveaway malfunctions rated during this program and the manner in which
total demerits were calculated are as follows:
Demerits for Poor Starting:
Demerits = starting time(s) - 2
Demerits for Stalls: (decelerating stalls are not assigned demerits)
Demerits = (no. of idle stalls) x 8 + (no. of maneuvering stalls)
x 32
-------
A-I - 4
Demerits for Subjective Ratings
Trace = 1
Moderate = 2
Heavy = 4
Weighting Factors for Each Malfunction
Idle Roughness = 1
Surge = 4
Backfire, Stumble, Hesitation = 6
Weighted Demerits = Demerits x Weighting Factor
Calculation:
Total Demerits = Weighted Demerits + Demerits for Stalls +
Demerits for Poor Starting
Demerits for each run were summed, counting all malfunctions that
occurred.
-------
Figure A-l-1
CRC driving cycle
Q.
E
•o
0>
0)
Q.
O
!E
Q)
60
40
20
25-35 mph detent
0-35 mph WOT
0-25 mph
light throttle
25 mph cruise
Start
-per owners
manual
L
10-25 mph
light throttle
Stop
30 sec.-
, idle
Repeat 3 times then
continue with cycle
below
I •
.1
.2 .3 .4 .5
Miles traveled
.6
.7
60
40
20
0
0-45 mph crowd
45 mph cruise 0-35mphWOT
25-35 mph
detent
10-25 mph
light throttle
Stop,
idle
30 sec.
0
.1
.2 .3 .4 .5
Miles traveled
.6
.7
Ul
-------
A-II - 1
II. MOTORIST DRIVING CYCLE
The motorist driving data collected by Amoco and used for developing the
cycle, are detailed in SAE paper, 810491 (8). The motorist cycle was
adapted from the current CRC cycle by redefining the throttle positions to
use for the various maneuvers. The CRC cycle uses four types of
accelerations; light-throttle, crowd, detent, and wide-open-throttle. The
throttle positions used for each of these are described in Appendix A-I.
For the motorist cycle, the CRC accelerations are replaced by the following
motorist accelerations (definitions are given in SAE paper 810491):
CRC Acceleration Motorist Acceleration
Light-Throttle Random and cold-start driving — 50ch percentile
acceleration** (Random-50)
Crowd Random and cold-start driving — modified 50tn
percentile acceleration
Detent Random and cold-start driving — 90t" percentile
acceleration (Random-90)
Wide-Open-Throttle Toll plaza driving — 90C^ percentile acceleration
(Toll-Plaza-90)
Figure A-II-1 shows the relationship between vehicle acceleration and
vehicle speed for "Random-50, "Random-90", and "Toll-Plaza-90" accele-
rations. Throttle positions are selected which yield vehicle accele-
rations closely matching the appropriate profile in Figure A-II-1. For
example, the Chrysler LeBaron being used in this program required
25 percent, 40 percent, and 50 percent throttle openings to approximate the
motorist Random-50, Random-90, and Toll-Plaza-90 accelerations, respec-
tively. The modified Random-50 acceleration replaces the crowd accele-
ration and is made by initially opening the throttle to 25 percent followed
by a very slow increase to 30 percent.
In addition to changes in throttle positions, the rate of throttle movement
is slower for the motorist cycle than for the CRC cycle. For example, when
making the wide-open-throttle CRC acceleration, the rater is instructed to
open the throttle as rapidly as possible without causing tire squeal or
slippage. When using the motorist cycle, however, raters are to take about
1 second to open the throttle to the desired position.
The last changes between the CRC and motorist cycles are that a "stabi-
lization" period is allowed between maneuvers and that speed ranges for the
maneuvers are slightly different. A graphic description of the motorist
cycle is shown in Figure A-II-2. When using the motorist cycle, the start-
up procedure and driveability problem ratings are the same as when using
the CRC cycle.
The 50tfl percentile acceleration was exceeded in 50% of the motorists'
accelerations.
The 901-" percentile acceleration was exceeded in 10% of the motorists'
accelerations.
-------
A-II - 2
Rgure A-ll-1
Motorist Driving Data Used for Cycle
Development
10
8
CM
o
® 6
I 5
2
_«
0 4
o
Toll-plaza driving - 90th percentile
acceleration
Random & cold start driving
' 90tn percentile acceleration
— Random &• cold start driving -
SQth percentile acceleration
10 20 30 40
Car speed, mph
50
60
-------
Figure A II 2
Motorist driving cycle
First three cycles
40
20
25-40 mph
Random 90
20-35 mph
Random 50
25 mph
cruise
~ 0-25 mph
Random 50
a
E
*
•a
0)
0)
a.
W
o
• ••
0)
0-35 mph
toll-plaza 90
30 sec. idle
60
4t
20
0
.1
.3
.4
.6
.7
Last three cycles
0-45 mph
Modified x.
random 50
25-40 mph
Random 90
20-35 mph
Random 501
0-35 mph
toll-plaza 90
20 mph
cruise
30 sec. idle
0
.1
.2
.3 .4 .5 .6
Miles traveled
.7
.8
.9
-------
APPENDIX B
Instruments for Measuring Engine/Vehicle
Operating Parameters
-------
B - 1
INSTRUMENTS FOR MEASURING
ENGINE/VEHICLE OPERATING PARAMETERS
1. Start Time - This was measured by activation of a Potter-Brumfield KRP
110 6V DC relay. The activation voltage for the relay was provided by
connecting two wires across the starter solenoid and then to the relay
coil. When the car is initially started, the relay coil closes the
relay contacts and activates the event marker on the Gould 6-channel
recorder (Model 15-6367-00). A 6V DC relay was used because there may
not always be a full 12V DC at the solenoid.
2. Oil Temperature - This was sensed by use of a 24" long, 1/8" sheathed,
Type K thermocouple, inserted into the oil dipstick hole. The
thermocouple signal was then fed into a Type K digital pyrometer
(Newport Model 268).
3. Intake Manifold Vacuum - This was sensed by use of a Robinson-Halpern
P61-995-31, 0-30" Hg vacuum, 1000 ohm potentiomatic transducer.
5V DC Supply >* »(2J ( + ) Signal Out
(-) (3) . £ • (-) IV = 10" Hg
4. Throttle Position - This was sensed by mechanically linking a New
England Instruments Company Model #78CBA102-C1B, 1 turn, 1200 ohm
potentiometer, to the throttle linkage on the carburetor.
(+) CD*-
5V DC Supply J-*-
, x Signal Out
•"; 1.89V = 100% Throttle
5- Vehicle Speed - This is driveshaft speed and is taken, as a pulsetrain,
from the Himmelstein Torque/RPM head. This pulsetrain is then
converted, by electronics on the cart built by ESD, to a voltage of
1.33V per 50 mph, or .0266V/mph, and displayed on the appropriate
digital meter.
6- Driveshaft Torque - This parameter is measured by use of the S.
Himmelstein Torquemeter and speed pick-up. This unit is located in the
vehicle's driveshaft. The calibration and installation of the unit is
as per the manual.
-------
B - 2
7. Engine Speed - The tachometer signal from the vehicle distributor was
converted to a voltage output via a Gould Model 13-4618-30 converter.
The resulting output is Imv/rpm to a maximum engine speed of
10,000 rpm. Engine speed was then monitored with a Gould strip chart
recorder.
8. Vehicle Acceleration - This was monitored by an ESD accelerotneter.
This device is mechanically linked in series with the speedometer
cable. The sine wave output of this transducer is then fed into ESD's
accelerometer. This signal conditioning device converts the sine wave
signal into a 0-5V analog signal which is proportional to the change in
vehicle speed. This analog signal is then recorded by a Gould 260
recorder.
9. Drawbar Pull - This condition was sensed by using a Transducer, Inc.
strain gauge, Model BTC-FF63H-CS-500. The transducer was bolted to
the bed plate soak room floor, the other end of this assembly was
attached to the frame of the vehicle. The transducer output was fed
into a Transducer, Inc. signal conditioner, Model 75C-42-0003E. The
analog signal produced was then recorded by a Gould 260 recorder.
10. Engine Vibration - This condition was monitored by a P.M.C. standard
vibration velocity transducer, Model 260C. The output of this unit was
recorded by a Gould 260 recorder.
11. Engine Rotation - A 1000 phm potentiometer was mounted to the fender-
well, the other end was connected to the engine block by means of a
moment arm. The mechanical advantage of this moment arm and associated
gears was approximately 10 to 1. A 5 volt signal was placed across this
potentiometer. The resulting output was fed to a Gould 260 recorder.
12. Signal Conditioning Cart - All signals were converted from analog form
to digital form for computer recording. The cart utilizes a
microprocessor for actual data transmission.
-------
APPENDIX C
Trained Rater Observed Demerits
by Driveability Problem Type
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
Test
Phase
Test
Number
1
2 -
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Demerits
Rater
A
A
A
B
B
B
A
A
A
B
B
B
A
A
Data
B
B
B
A
A
A
B
B
B
A
A
A
B
B
B
Hard
Stalls
Fuel Starting* Idle Accel Decel
1
1
1
2A
2A
1
2A
2A
1
1
2A
1
1
2A
Destroyed
3
2A
2A
1
3
3
2A
3
3
3
1
3
2A
2A
1
4.6 16
4.4 0
2.4 8
15.8 0
0 0
16.6 16
4.4 0
0 0
4.4 16
1.4 16
0 0
3.4 16
0 0
0.8 0
by Computer
0.2 0
0 0
0 0
3.0 0
3.0 8
0.8 0
0 0
0.2 0
0 0
7.6 0
7.4 16
0.2 0
0 0
0 0
2.8 16
256
224
128
0
0
192
0
0
256
192
64
160
0
32
Malfunction
96
0
0
320
96
32
0
96
64
32
288
64
0
0
256
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
7
9
1
1
0
5
1
3
18
13
0
6
0
0
1
1
2
5
6
0
1
3
3
0
4
1
1
1
3
Stumble
168
180
330
18
18
336
126
162
738
306
24
312
90
210
102
12
18
294
204
150
30
132
102
120
300
210
30
30
270
i
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
36
66
102
0
0
0
18
12
54
18
12
12
0
0
0
0
0
0
6
0
6
6
6
0
0
0
0
0
0
Surge
24
0
0
0
0
16
0
0
180
8
0
4
0
0
0
0
4
0
0
0
0
0
0
0
40
0
0
0
0
Backfire
66
6
42
0
0
36
0
0
108
18
6
30
0
84
0
0
0
78
12
6
18
0
0
24
168
18
0
0
30
Total
577.6
489.4
613.4
34.8
18.0
617.6
149.4
177.0
1374.4
572.4
106.0
543.4
90.0
326.8
199.2
13.0
24.0
700.0
335.0
188.8
37.0
237.2
175.0
183.6
823.4
293.2
31.0
31.0
577.8
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
- Continued -
Demerits
Test Test
Phase Number
I 31
32
(Cont'd) 33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60 **
Rater
A
A
A
B
B
B
A
A
A
A
B
B
A
A
A
B
B
B
A
A
B
B
B
A
A
A
B
A
A
A
Fuel
1
3
2A
1
1
3
3
1
2A
1
2A
1
1
3
1
2A
1
3
3
1
3
2A
1
2A
3
2A
3
2A
1
1
Hard
Starting*
13.6
0
0
3.0
2.6
0
15.6
4.6
14.2
2.2
0
2.4
2.0
0.2
2.2
0
3.0
0.2
16.0
2.2
0.2
0
2.6
0
0.2
0
0
0.6
1.8
0
Stalls
Idle
16
0
0
16
16
0
0
8
0
16
0
16
8
0
8
0
16
0
0
8
0
0
16
0
0
0
0
0
8
0
Accel
224
32
0
192
128
0
0
160
0
96
0
128
192
0
128
0
96
32
0
96
32
0
96
0
0
32
0
0
288
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
6
0
2
1
3
0
0
3
1
1
0
9
10
0
4
0
7
0
0
0
0
0
7
0
0
0
0
0
1
0
Stumble
354
102
30
150
138
54
60
96
6
144
6
228
162
18
168
6
114
30
12
90
18
12
162
0
6
6
18
6
108
0
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
12
0
0
0
0
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
0
0
0
6
12
0
0
Surge
8
0
0
0
0
4
0
4
4
0
4
8
16
4
4
0
8
4
0
4
4
0
8
4
4
4
0
0
0
0
Backfire
84
18
0
36
24
12
0
0
0
36
0
48
54
0
18
6
24
6
12
30
0
0
12
6
18
0
6
0
102
0
Total
717.6
152.0
32.0
398.0
311.6
70.0
75.6
275.6
25.2
295.2
10.0
439.4
444.0
28.2
332.2
12.0
268.0
72.2
40.0
230.2
54.2
12.0
303.6
10.0
28.2
42.0
30.0
18.6
508.8
0
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
- Continued -
Demerits
Test Test
Phase Number
I 61
62
(Cont'd) 63 **
64
65
66
67 **
68
69
70
7 1 **
72
73
74
75 **
76
77
78 **
79
80
Rater
A
B
B
B
B
A
A
A
A
B
B
B
B
A
A
A
B
B
B
B
Fuel
2A
3
9
2A
1
2A
9
1
2A
3
9
2A
1
2A
9
3
1
9
3
2A
Hard
Starting*
0
0.2
0
0.8
1.8
0.4
0
3.6
0
0
0
0
2.4
1.2
0
1.0
1.8
0
0.2
0
St-alls
Idle
0
0
0
0
8
0
0
16
0
0
0
0
16
0
0
0
8
0
0
0
Accel
0
32
0
32
256
32
0
64
0
32
0
0
192
0
0
64
96
0
0
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
1
1
0
0
6
0
0
0
0
0
0
0
8
0
0
0
5
1
0
1
Stumble
24
48
0
12
180
6
0
66
18
36
0
24
192
6
0
42
138
0
36
18
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
0
0
0
Surge
0
0
0
0
16
0
0
0
0
0
0
4
12
0
4
0
4
8
4
0
Backfire
0
0
0
0
48
0
0
12
0
6
0
0
24
0
0
6
18
0
18
6
Total
25.0
81.2
0
44.8
515.8
38.4
0
167.6
18.0
74.0
0
28.0
446.4
7.2
4.0
113.0
270.8
9.0
58.2
25.0
81
2A
1.0
83
84
85
86
87
A
B
B
A
A
2A
2A
2A
2A
2A
Each of these tests was a series of 0-45 mph accelerations
at various throttle openings. The data were used in
determining throttle positions for the Motorist-cycle tests.
13.0
o
I
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
(Phases II thru VI Results)
Trained Rater Observed Demerits
Test Test
Phase Number
II 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
III 28
29
30
Rater
A
A
A
A
B
B
B
A
A
A
A
B
B
B
B
A
A
A
A
A
A
A
B
B
B
B
B
A
A
A
Fuel
2B
2B
3
1
1
3
2B
1
2B
3
1
3
1
2B
1
3
1
2B
3
1
2B
3
1
3
2B
3
2B
1
2B
3
Hard
Starting*
0
1.2
0
2.2
0.4
0
0
2.4
0
0
4.0
0
0
0
0
0
0.6
0
1.0
1.0
0
0
0
0
0
0
0
1.4
1.0
1.8
Idle
0 "
8
0
16
0
0
0
8
0
0
8
0
0
0
0
0
8
0
8
0
0
0
0
0
0
0
0
0
0
8
flails
Accel
0
0
0
320
256
0
0
96
0
0
32
0
0
0
0
0
64
0
0
0
0
0
0
0
0
0
0
128
0
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
0
0
0
6
1
0
1
0
0
1
0
5
6
5
6
0
0
0
0
0
0
0
3
6
5
5
9
0
0
0
Stumble
0
6
18
234
96
36
24
96
6
24
108
18
84
18
84
6
6
0
6
30
6
0
36
30
24
60
30
210
36
42
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
6
0
0
12
24
0
0
0
0
0
0
0
12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
36
Surge
0
0
0
48
4
8
12
0
0
0
0
28
24
24
32
0
4
0
0
4
0
0
32
20
12
20
20
4
0
0
Backfire
0
0
0
12
0
0
0
6
0
0
0
0
0
0
0
0
6
0
0
0
0
0
0
0
0
0
0
6
0
24
Total
6.0
15.2
18.0
650.2
381.4
44.0
37.0
208.4
6.0
25.0
152.0
51.0
126.0
47.0
122.0
6.0
88.6
0
15.0
35.0
6.0
0
71.0
56.0
41.0
85.0
59.0
349.4
37.0
111.8
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
(Phases II thru VI Results)
- Continued -
Trained Rater Observed Demerits
Test Test
Phase Number
III 31
(Cont'd) 32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
IV 60
Rater
A
- B
B
B
B
B
B
B
A
A
A
A
B
B
B
B
A
A
A
A
B
B
B
B
B
A
A
A
A
A
Fuel
1
2B
3
1
3
1
3
2B
3
1
2B
3
1
2B
3
1
2B
3
1
1
2B
3
1
2B
3
2B
3
1
1
2B
Hard
Starting"
4.6
0
0.4
2.0
0
1.4
0
0
1.4
1.2
0
0
0
0
0
1.2
1.4
2.8
4.4
6.8
0
0
1.4
0
0
1.2
1.4
1.4
3.8
0
Idle
16
0
0
8
0
8
0
0
8
16
0
0
8
0
0
8
8
8
16
8
0
0
8
0
0
8
8
8
8
0
Stalls
Accel
96
0
0
32
0
128
0
32
0
160
32
0
64
0
0
96
0
0
32
64
0
0
32
0
32
0
0
0
32
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
3
3
5
8
4
5
3
4
0
0
0
0
5
4
3
5
0
0
0
0
3
4
4
2
3
0
0
0
0
1
Stumble
102
42
84
288
60
246
96
48
30
168
42
66
156
42
90
162
36
72
90
138
66
72
240
90
66
42
24
168
174
30
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
24
0
0
0
0
24
12
0
18
42
0
24
72
0
0
30
18
6
6
6
12
6
24
0
0
6
6
0
90
48
Surge
48
4
0
8
4
4
8
8
8
24
0
0
12
0
0
0
16
8
76
0
0
4
0
4
4
0
4
8
8
36
Backfire
72
0
0
54
0
6
6
0
24
84
6
36
84
6
0
42
0
24
102
192
12
6
36
24
12
36
24
6
54
36
Total
365.6
49.0
89.4
400.0
68.0
422.4
125.0
92.0
89.4
495.2
80.0
126.0
401.0
52.0
93.0
344.2
79.4
120.8
326.4
414.8
93.0
92.0
345.4
120.0
117.0
93.2
67.4
191.4
369.8
151.0
n
I
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
(Phases II thru VI Results)
- Continued -
Trained Rater Observed Demerits
Test Test
Phase Number
IV 61
(Cont'd) 62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
V 88
89
90
Rater
B
A
A
A
B
A
A
A
A
A
B
B
B
B
A
A
B
B
B
A
A
A
A
B
B
B
B
A
A
A
Fuel
1
2B
3
2B
1
2B
3
1
3
1
2B
1
2B
3
2B
3
1
2B
3
1
2B
3
1
3
2B
1
3
1
3
2B
Hard
Starting*
0
0.4
0
1.4
1.6
0.4
0.2
7.0
0.2
2.4
0
2.6
0
0.2
0.2
0.4
2.6
0
0
3.0
0.6
0.4
2.6
0.4
0
2.0
0.4
1.6
0
0.2
Idle
0
0
0
0
8
0
0
8
0
8
0
8
0
0
0
0
8
0
0
8
0
0
8
0
0
16
0
0
0
0
Stalls
Accel
0
64
0
96
288
0
0
64
0
256
96
192
0
32
0
0
96
0
32
96
0
32
128
32
32
128
32
0
0
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
7
0
0
0
11
0
0
1
0
0
5
8
4
2
0
0
5
0
5
0
0
0
4
4
1
0
3
0
0
0
Stumble
120
84
24
138
126
168
6
114
12
174
126
84
66
72
60
78
102
84
84
66
30
78
114
48
60
102
42
18
30
0
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 '
0
0
0
0
0
Hesitation
0
42
0
18
48
54
54
30
6
126
12
24
24
0
30
12
36
12
0
6
36
36
114
0
0
102
0
24
6
12
Surge
0
32
0
16
8
16
0
12
0
8
0
0
0
0
0
4
0
0
0
0
4
0
24
0
0
24
0
0
0
0
Backfire
30
36
0
24
72
6
0
48
0
96
36
114
42
24
12
18
84
42
18
114
18
78
198
24
36
222
24
0
0
0
Total
157.0
258.4
24.0
293.4
562.6
244.4
60.2
284.0
18.2
670.4
275.0
432.6
136.0
130.2
102.2
112.4
333.6
138.0
139.0
293.0
88.6
224.4
592.6
101.4
129.0
596.0
101.4
43.6
36.0
12.2
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
(Phases II thru VI Results)
- Continued -
Trained Rater Observed Demerits
Test Test
Phase Number
V 91
(Cont'd) 92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
VI 112
113
114
115
116
117
118
119
120
Rater
A
B
B
B
A
A
A
A
A
B
B
A
A
B
B
B
A
A
B
B
B
A
B
B
B
A
A
B
B
B
Hard
Fuel Starting*
1
2B
1
1
2B
1
3
2B
3
1
3
2B
3
2B
1
3
1
3
2B
3
2B
1
2B
3
1
2B
3
1
3
2B
3.0
0
0.2
0.6
0.2
3.8
0
0.4
2.0
1.4
0
1.2
0
0
1.8
0
0.2
0
0
0
0
0
0
0
0
0
0
0
0
0
Idle
16
0
0
0
0
16
0
0
8
8
0
8
0
0
8
0
8
0
0
0
0
0
0
0
0
0
0
0
0
0
Stalls
Accel
32
0
32
32
0
0
0
0
0
32
0
0
0
0
0
0
0
0
0
0
0
0
0
0
32
0
0
0
0
0
Decel
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Roughness
0
0
10
8
0
0
10
0
0
5
9
0
0
6
4
15
0
0
4
7
8
2
4
2
6
0
0
6
5
11
Stumble
30
0
138
102
12
18
0
24
6
138
0
0
12
0
18
6
24
0
6
0
6
0
18
6
90
12
6
36
18
42.0
Extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Hesitation
18
0
0
0
30
90
0
12
0
0
0
6
18
0
0
0
12
6
0
0
0
6
0
0
0
30
18
0
0
0
Surge
0
0
16
20
0
0
4
0
0
20
0
0
0
0
0
0
0
0
0
0
4
0
0
0
20
0
0
0
0
0
Backfire
0
0
0
12
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Total
99.0
0
196.2
174.6
42.2
127.8
14.0
36.4
16.0
204.4
9.0
15.2
30.0
6.0
31.8
21.0
44.2
6.0
10.0
7.0
18.0
8.0
22.0
8.0
148.0
42.0
24.0
42.0
23.0
53.0
-------
TRAINED RATER OBSERVED DEMERITS BY DRIVEABILITY PROBLEM TYPE
(Phases II thru VI Results)
- Continued -
Trained Rater Observed Demerits
Test
Phase
VI
Test
Number
121
122
123
124
125
Rater
A
A
A
B
A
Fuel
3
2B
1
1
2B
Hard
Starting*
0
0.2
0
0
0
Idle
0
0
0
0
0
Stalls
Accel
0
0
0
0
0
Decel
0
0
0
0
0
Roughness
0
1
0
10
2
Stumble
0
0
0
24
18
Extension
0
0
0
0
0
Hesitation
12
0
0
0
12
Surge
0
0
0
8
0
'
Backfire
0
0
0
0
0
Total
12.0
1.2
0
42.0
32.0
* Starting time used for calculating hard starting demerits was not
measured by the trained rater — only by the computer.
** These tests were made with the car engine fully warmed-up. The
data were used in developing the hesitation measurement method.
n
i
GO
-------
APPENDIX D
Unsuccessful Attempts to Objectively
Measure Stumble and Hesitation
-------
D - 1
UNSUCCESSFUL ATTEMPTS TO
OBJECTIVELY MEASURE STUMBLE AND HESITATION
STUMBLE
A. General
From past experience the investigators were confident that vehicle
stumble could be related to fluctuations in engine or vehicle
operating parameters. Throughout the program, efforts to measure
stumble dealt with:
1. ways to identify all fluctuations
2. ways to eliminate fluctuations not noticed by the rater
3. ways to associate fluctuations with trained rater observed stum-
ble, and
4. ways to predict trained rater observed stumble demerits from
properties of the fluctuations.
Various attempts to accomplish items 1, 2 and 3 are described below in
section B, and item 4 is discussed in section C.
B. Stumble Identification
1. Cubic - Initially, attempts were made to identify dips by fitting
cubic equations to short segments of engine speed and driveshaft
torque data. The cubic equation form was selected because it has
a local maximum and minimum value. These equations were then
analyzed by computer to find the time duration (At) and amplitude
(Aa) of dips in the data. We next attempted to eliminate
meaningless dips by placing a lower limit on the multiple
regression coefficient squared for each equation; the thought
being that if a stumble occurred and was recognized by the rater,
then the dip should be large enough to provide good equation fit.
The At and A a values were then used in an optimization routine. The
types of optimization routines considered throughout the program
are described in section C. However, regardless of the type of
optimization used, several items were universal. First, rules
were applied to the A t and A a information to see whether
consecutive dips should be grouped together. Second, decisions
had to be made as to which dip or dip-groups should be paired with
the trained rater observed stumble. (Explanations of dip grouping
and pairing are included in the main body of this report.) Third,
demerits had to be calculated for each dip or group, and fourth, a
score had to be calculated for the data being optimized. The rules
for dip grouping and pairing were quite complex when using the
cubic equation approach. It is sufficient to say that some of the
rules included variables to be optimized. When using the cubic
equation approach, the following equation was used for calculating
total stumble demerits for an entire driveability test:
-------
D - 2
Stumble
Demerits =S£ (R£2 - RT2) (b0 + bj (Ati) + b2 (Aai))T + S; (Rj2
- RS2) (b3 + b4 (Atj) + b5 (Aaj))s
where i and j denote the ic^ driveshaft torque dip or jc^ engine
speed dip respectively.
R£2 and Rj2 denote the multiple-correlation-coefficient squared
for the equation fitted to the i^ driveshaft torque
dip or j1-" engine speed dip respectively.
R-j2 and Rg2 denote minimum cutoff values for the multiple-
correlation-coefficient squared on equations fitted
to driveshaft torque or engine speed dips respec-
tively.
bx's are coefficients for calculating computer
demerits of individual dips
After calculating demerits, the computer calculated a score that
described a discrepancy between rater observed and computer calculated
demerits. It was:
Score = DR + Dc + 1/2 DB
where DR = sum of demerits for stumbles observed by the trained rater
that are not paired with a dip.
DC = sum of demerits for dips found by computer which are not
paired with trained rater-observed stumbles.
Dg = sum of the absolute difference in demerits for computer-
observed dips that are paired with trained rater-observed
stumbles .
The optimization programs always attempted to minimize the score by
adjusting the values of the variables. These variables included:
1) the rules for grouping and pairing dips, 2) the multiple correlation
coefficient cutoffs (Rg and R-p 2) , and 3) the demerit equation
coefficients (bx's). The cubic approach was abandoned because the
number of consecutive data points for equation fitting had to be held
constant for each variable. This damaged the equation's ability to
accurately measure A t and Aa for all dips.
Smoothing - First Try
The only difference between the smoothing and cubic methods are:
1) the system used for finding At andAa, and 2) the demerit calculation
equations. The smoothing method is so named because to find At andAa
-------
D - 3
the engine and vehicle operating parameter data were smoothed slightly
and then searched to find actual maximum and minimum values. The
smoothing function was:
Xis = .ZSXi.! + .5X£ + .25Xi+1
where: X^s is the smoothed value of the ic data point
X£ is the un-smoothed value of the itn data point
^i-1 > ^i+1 are the un-smoothed values of the data points
on either side of the i*1*1 point
Along with this new method of finding dips, it was also decided that
the time interval between maximum and minimum values of the engine
speed or torque may not be the appropriate interval to use. Therefore,
to allow data immediately prior to the maximum to be included a slope
cutoff variable was established. This defined a dip as beginning when
the speed or torque slope (as a function of time) dropped below this
cutoff. The new demerit calculation equation became:
Stumble Demerits = i(bi(Ati) + b2(Aai)) + j(b3
-------
D - 4
C. Optimization Routines
1. General
The function of the optimization routines was to find values for
several variables which resulted in the best possible agreement between
trained rater observed and computer calculated demerits (minimize the
score). Before deciding upon the final optimization method three other
schemes were investigated.
2. Random Lines
When using this method simultaneous optimization of eleven variables
was being attempted. The procedure followed by the computer was to
select two points in the eleven variable space, project a line between
these points, and find the best optimum between them. Next, project a
second random line through this best point, find a better optimum, and
so on until no better optimum could be found. Since this is a random
approach, there was no guarantee that repeat optimizations of identical
data would yield matching results. In fact, they yielded widely
varying results. Another optimization approach was then tried.
3. Partan
The computer projects two parallel lines through the n-space and finds
the best optimum along each line. Through these two best points a third
line is projected. A fourth line parallel to the third is then
projected and the process starts again. As with the random line
approach, this method tended to converge to local rather than universal
optimum and it was therefore abandoned.
4. Exhaustive Search
With this method the computer found the optimum value for each variable
by exhaustively investigating all possible values of each variable. By
the time this third optimization procedure was being considered,
several variables had been discarded from the optimization which
reduced the required computer time to a manageable level. Using the
exhaustive method the computer always converged to a unique optimum but
the optimum values for the variables varied widely between individual
driveability tests.
Because none of these optimization methods were satisfactory, the
optimization described in the report was finally used — linear-least-
squares regression.
-------
D - 5
HESITATION
In developing a system for measuring hesitation, it was theorized that a
lack of "proper" engine speed response to throttle opening could be used.
To do this required defining an ideal rate of engine speed increase and
comparing this rate with the actual rate. A simple equation was developed
to predict ideal engine speed slope during the initial 1 or 2 seconds of an
acceleration as a function of throttle opening. The equation form was:
Ideal Engine Speed Slope = KjCbQ + b^TTL))
where: TTL was the throttle opening 2.0 seconds after the beginning of
the acceleration
bx's are regression coefficients
Next, a relationship to predict hesitation demerits was developed as a
function of the difference between the ideal slope and the actual slope
existing during the acceleration. This provided very poor correlation
between trained rater observed and computer calculated demerits.
The next attempt was to predict hesitation demerits as a function of:
1) time required for engine speed to increase by a constant amount (e.g.
50 or 100 RPM) , and 2) throttle position 2.0 seconds after the beginning of
an acceleration. This too provided poor correlation between observed and
calculated demerits.
The first two hesitation methods may have failed because they placed too
little emphasis on the time immediately after the acceleration began, and
they did not recognize the possible effect of throttle opening rate upon
engine speed increase. Therefore, coefficients were developed for the
following regression equation through linear-least-squares techniques:
WRFM.
Hesitation Demerits = bg + b^ (—„,-,)
where: WRPM and WTTL are weighted engine speed increase and throttle
opening increase, respectively, during the
first 1.0 second of an acceleration:
WRPM = 0.5(ARPM1)+.25(ARPM2)+.13(ARPM3)+.06(ARPM4)+.06(ARPM5)
WTTL = 0.5(ATTL1)+.25(ATTL2)+.13(ATTL3)+.06(ATTL4)+.06(ATTL5)
ARPM£ is the engine speed increase during the i*-*1 •
0.2 second interval of the acceleration
is the throttle opening increase during the itn
0.2 second interval of the acceleration
-------
D - 6
The coefficients for this equation resulted in calculated demerits being
very sensitive to the independent variable (WRPM over WTTL) and again
correlation was poor between observed and calculated demerits. In
addition, this method yielded nearly identical hesitation demerits for all
three test fuels.
The attempts described above cover only the basic equations considered; in
most cases several alternate equation forms were considered but found to be
of little benefit. The final scheme for hesitation demerit measurement is
described in the body of this report.
-------
APPENDIX E
Data Used for Developing
Hesitation Measurement Method
-------
HESITATION DATA
Test
Number
60
63
67
71
75
Maneuver
1st 0-25
2nd 0-25
1st 25-35
2nd 25-35
1st 0-35
2nd 0-35
1st 0-25
2nd 0-25
1st 25-35
2nd 25-35
2nd 0-35
1st 0-25
2nd 0-25
1st 25-35
2nd 25-35
1st 0-35
2nd 0-35
2nd 0-25
1st 25-35
2nd 25-35
1st 0-35
2nd 0-35
1st 0-25
2nd 0-25
1st 25-35
2nd 25-35
1st 0-35
2nd 0-35
Initial
TTL,
% Open
5.8
2.6
6.9
5.3
3.7
2.6
12.7
2.6
9.0
21.2
3.7
4.2
4.2
7.4
5.7
3.2
2.6
3.2
12.2
4.2
2.6
5.8
4.2
3.7
10.6
2.6
2.6
2.6
1
2.1
4.8
4.2
3.2
82.5
7.5
7.4
12.7
26.4
19.5
45.0
2.7
3.7
2.1
1.7
6.3
7.5
24.8
29.6
24.9
4.3
39.2
4.8
1.1
27.5
4.8
10.1
3.2
/
2
4.3
2.7
6.9
7.4
10.6
75.1
.5
12.2
6.4
1.6
48.1
1.6
6.9
6.9
7.4
72.0
68.8
-1.5
5.8
7.4
82.0
50.2
1.6
10.0
9.0
6.9
89.9
82.0
^ TTL^
3
7.4
2.6
7.9
30.1
0
11.6
.6
-1.0
2.6
1.6
0
2.1
3.7
11.6
9.0
15.3
15.3
0
0
5.3
7.9
1.1
1.6
8.0
2.1
11.6
4.2
9.0
4
3.7
3.7
10.6
3.7
0
0
1.0
1.0
0
1.1
0
0.5
3.2
17.5
14.3
0
0
-0.6
0
4.2
0
0.5
2.6
2.1
0
12.2
0
0
5
3.7
3.2
2.1
0
0
0
-1.0
-1.0
.6
0
0
2.1
4.2
2.6
3.8
0
0
-2.6
0
0
0
0
1.1
0
0
3.7
0
0
Initial
RPM
710
645
850
880
550
550
580
590
920
760
550
780
620
910
830
560
540
640
900
800
550
550
950
615
940
940
540
540
A RPMj
1
20
140
50
30
450
425
110
100
350
520
320
30
30
60
70
625
550
180
390
120
375
400
20
50
220
50
725
750
2
50
50
80
130
350
350
390
340
140
90
425
20
60
220
60
175
250
230
140
390
425
375
50
50
190
80
50
-30
3
70
140
210
320
10
0
20
30
50
20
-25
30
120
140
120
-30
-25
90
50
60
20
10
40
120
80
120
60
40
4
200
30
60
30
10
50
0
-60
20
20
20
30
160
100
200
30
50
-50
5
50
25
50
0
330
40
200
10
75
5
100
120
70
70
50
30
-5
-60
-5
-5
50
130
130
20
60
30
70
-40
5
10
40
20
20
10
-5
80
40
20
MPH0
19.5
0
25.1
27.8
1.1
1.1
0
0
25.5
27.0
2.6
0
0
25.5
26.6
1.1
1.1
0
25.9
27.8
1.5
2.3
0.4
0.4
26.2
27.0
1.1
1.5
WTTL
3.5
3.8
3.9
9.3
44.0
24.0
3.9
9.3
15.2
4.4
34.5
2.2
4.5
7.6
5.0
23.0
23.0
11.8
16.3
15.2
24.0
32.0
3.2
4.2
13.0
6.6
28.0
23.3
WRPM
44.6
109.7
160.0
95.1
317.0
3050.0
155.0
182.0
217.4
286.0
11.1
335.0
63.0
110.4
81.2
356.0
341.0
64.3
237.0
169.0
300.0
299.0
28.9
73.5
170.0
77.4
386.0
378.0
Trained
Rater
Observed
Hesitation
Demerits
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
PI
I
-------
HESITATION DATA
Test
Number
78
3
7
8
9
10
11
12
20
22
23
24
31
44
58
81
Maneuver
ist
1st
2nd
1st
2nd
2nd
3rd
4th
5th
6th
2nd
6th
3rd
2nd
4th
5th
4th
4th
3rd
4th
1st
2nd
1st
3rd
1st
4th
6th
2nd
3rd
4th
2nd
0-25
25-35
25-35
0-35
0-35
0-35
0-25
0-45
0-35
0-45
0-35
0-35
0-35
0-25
0-35
0-45
0-45
25-35
0-25
0-45
25-35
0-25
0-35
0-25
10-25
0-35
0-45
0-35
0-25
0-45
0-35
Initial
TTL,
% Open
2.6
10.1
7.9
37.6
10.6
6.3
8.5
6.9
3.2
2.6
7.4
3.7
6.3
10.1
6.3
2.6
3.7
6.9
4.8
6.3
6.3
14.8
11.1
11.1
6.3
10.1
7.4
4.8
6.3
3.2
6.9
ATTLj Initial
1
2.7
8.9
33.9
53.2
77.2
19.6
3.7
2.1
24.3
4.3
53.4
54.0
12.2
2.6
36.0
22.3
6.4
34.8
3.7
12.7
41.3
2.7
27.5
0.5
15.9.
-3.7
-2.6
11.6
2.2
4.2
27.5
2
24.3
23.3
5.3
0
9.0
69.9
5.3
2.1
64.0
1.2
34.4
38.6
77.8
4.5
50.8
65.6
6.3
8.5
10.0
8.5
2.1
1.0
57.2
6.9
0
3.7
5.3
57.1
3.7
6.4
60.8
3
-4.7
1.6
0
0
0
-0.5
2.6
7.4
0
4.8
1.1
0
0
4.2
2.7
5.8
6.4
-0.6
0.5
1.6
-5.3
1.1
0.5
0.5
-0.5
9.0
3.7
15.3
4.7
2.6
1.6
4
-1.1
0
0.6
0
0
0
1.1
6.4
0
2.6
0
0
0
0
0.5
0
-1.1
-4.7
0.6
16.3
-0.5
3.7
0
0.6
0
2.7
3.2
0
7.4
1.6
0
5
-0.5
0
0
0
0
0
1.0
4.7
0
8.0
0
0
0
0
0
0
1.1
0.5
0
-13.2
0
0.5
0
0
0
2.6
2.1
0
4.8
3.2
0
RPM
520
790
930
520
540
890
710
740
825
645
940
1000
850
620
530
510
540
950
690
850
1190
810
770
660
860
630
660
780
680
600
660
1
30
90
450
660
560
-60
60
20
260
70
280
530
320
30
200
110
30
-70
240
160
-130
70
260
130
-30
160
60
250
30
90
360
ARPMi
2
390
380
150
230
250
50
100
160
270
260
40
40
320
5
50
100
140
160
-70
-160
-30
20
-100
-50
-20
160
120
290
90
240
370
3
210
60
10
0
-20
30
-10
170
60
110
180
10
50
-80
50
-30
-40
60
-40
-80
240
110
450
-70
120
20
140
60
20
30
120
4 5
20 -30
0 -20
30 0
20 30
50 20
10 80
-80 -50
-10 -25
-25 -80
20 -20
65 10
-10 -15
-5 -80
100 230
-50 -60
40 -30
-60 -70
-80 -100
-10 5
0 320
50 140
30 50
65 10
-20 0
80 -5
-40 -50
50 -40
0 -5
60 -150
20 -40
10 -20
MPHp
0.4
27.0
27.0
1.9
3.4
0
0
0
1.5
0
0.8
0.4
1.5
0
0.8
1.5
0
27.4
0
0
27.0
0
2.3
0
13.9
0
0
2.3
0
0
0
WTTL WRPM
6.5
10.5
18.3
26.6
10.9
27.0
3.6
3.2
28.0
3.7
35.4
37.0
26.0
2.9
31.0
28.0
4.3
20.0
3.7
8.9
16.0
2.9
28.0
4.5
5.5
2.4
2.1
21.0
3.7
4.3
29.0
139.2
147.0
266.0
390.0
344.0
-8.2
45.9
70.0
199.0
114.3
178.0
275.0
243.0
25.6
112.4
76.7
37.0
2.0
97.0
48.8
-29.9
59.0
168.0
42.2
0.1
117.2
78.8
205.0
34.7
108.0
287.5
Trained
Rater
Observed
Hesitation
Demerits
0
0
0
0
0
24
24
12
24
12
12
6
12
6
12
12
12
12
6
6
12
6
6
6
6
6
6
6
6
6
6
-------
APPENDIX F
Rater-Observed and Computer-Calculated Demerits
(Test Phases II-IV)
-------
F - 1
TEST FUEL 1 DEMERITS
(Phases II thru
IV Results)
Fuel 1
Demerits
Idle
t—l
h-l
Cd
WJ
a:
O.
EH
to
Cd
EH
1— 1
'w
w
a:
D-i
EH
H
>
M
Id
OO
X
cu
H
to
w
H
Test
Number
4
5
8
11
13
15
17
20
23
""•'•'Mean
Std.
%
28
31
34
36
40
43
46
49
50
53
58
59
""-Mean
Std.
%
61
65
68
70
72
77
80
83
86
•'"-Mean
Std.
%
Rater
A
B
A
A
B
B
A
A
B
Dev. ,
of Mean
A
A
B
B
A
B
B
A
A
B
A
A
Dev. ,
of Mean
B
B
A
A
B
B
A
A
B
, Dev. ,
of Mean
Hard
Starting'"
2.2
0.4
2.4
. 4.0
0
0
0.6
1.0
0
1.2
116
1.4
4.6
2.0
1.4
1.2
0
1.2
4.4
6.8
1.4
1.4
3.8
2.5
79
0
1.6
7.0
2.4
2.6
2.6
3.0
2.6
2.0
2.6
72
Roughness
Rtr
6
1
0
0
6
6
0
0
3
2.4
118
0
3
8
5
0
5
5
0
0
4
0
0
2.5
114
7
11
1
0
8
5
0
4
0
4
102
Comp
_
8.2
4.7
28.8
21.0
6.0
3.1
3.3
1.9
9.6
103
1.0
0.0
2.4
4.8
14.3
2.1
3.8
16.0
0.8
0.0
7.0
6.8
7.9
109
4.7
6.5
3.6
8.9
4.5
2.2
1.2
6.2
3.6
4.6
51
Hesitation
Rtr
12
24
0
0
12
0
0
0
0
5.3
164
0
24
0
24
42
72
30
6
6
24
0
90
26.5
110
0
48
30
126
24
36
6
114
102
54
88
Comp
74.8
52.2
83.5
121.1
79.6
44.9
26.2
55.0
-
67.2
43
77.1
30.1
5.3
57.9
86.8
86.0
92.9
84.2
63.9
92.0
25.0
40.0
61.8
48
46.0
144.0
77.9
155.0
99.0
72.0
65.1
130.5
105.9
99.5
38
Stumble
Rtr
234
96
96
108
84
84
6
30
36
86
77
210
102
288
246
168
156
162
90
138
240
168
174
178.5
33
120
126
114
174
84
102
66
114
102
111
27
Comp
_
73.7
79.7
88.5
65.0
73.0
33.2
48.6
39.5
62.7
32.0
_
165.5
185.0
-
-
-
119.5
-
126.6
-
91.8
-
137.7
27
111.8
154.7
111.9
119.4
101.6
112.7
84.5
105.7
82.3
109.4
19
Stall
Rtr
336
256
104
40
0
0
72
0
0
90
138
128
112
40
136
176
72
104
48
72
40
8
40
81.3
61
0
296
72
264
200
104
104
136
144
146.7
64
Comp
320
256
168
48
0
0
72
0
0
96
128
128
112
40
136
168
64
104
48
80
40
8
48
81.3
59
0
296
72
264
200
104
104
168
136
149.3
63
Total
Rtr
590.2
377.4
202.4
152.0
102.0
90.0
78.6
31.0
39.0
184.7
100
339.4
245.6
338.0
412.4
387.2
305.0
302.2
148.4
222.4
3-09.4
177.4
307.8
291.3
27
127.0
482.6
224.0
566.4
318.6
249.6
179.0
370.6
350.0
318.6
45
Comp
^
390.5
338.3
290.4
165.6
123.9
135.1
107.9
—
222.0
52
_
312.2
234.7
-
—
-
321.4
-
278.1
-
133.2
-
255.9
30
162.5
602.8
272.4
549.7
407.7
293.5
257.8
413.0
329.8
365.5
39
Starting time used for calculating hard starting demerits was not measured by the
trained rater — only by the computer. The hard starting demerits shown are
included in the total for both the trained rater and computer.
•'"" Means and Standard Deviations are calculated from all available observations.
-------
F - 2
TEST FUEL 2B DEMERITS
(Phases II thru IV Results)
-- Continued -
Fuel 2B Demerits
M
w
Ed
CO
Si
PL,
H
CO
Ed
H
I— i
Ed
CO
PL,
H
CO
Ed
H
I — i
Ed
CO
K
CU
H
CO
H
Test
Number
1
2
7
.. 9
14
18
21
25
27
**Mean
Std.
%
29
32
38
41
44
47
51
54
56
"-Mean
Std.
%
60
62
64
66
71
73
75
78
81
85
Hard
Idle
Roughness
Rater Starting" Rtr
A
A
B
A
B
A
A
B
B
Dev. ,
of Mean
A
B
B
A
B
A
B
B
A
Dev. ,
of Mean
A
' A
A
A
B
B
A
B
A
B
--Mean
Std.
%
Dev. ,
of Mean
0
1.2
0
0
0
0
0
0
0
0.1
400
1.0
0
0
0
0
1.4
0
0
1.2
0.4
152
0
0.4
1.4
0.4
0
0
0.2
0
0.6
0
0.3
148
0
0
1
0
5
0
0
5
9
2.2
149
0
3
4
0
4
0
3
2
0
1.8
100
1
0
0
0
5
4
0
0
0
1
1.1
168
Comp
4.9
6.0
6.9,
21.5
4.1
35.7
33.4
2.7
0
12.8
107
0.0
0.0
0.8
6.6
0.0
0.7
0.0
0.0
3.2
1.3
174
1.8
6.4
4.8
2.3
1.4
0.2
1.5
0.4
0.7
0.8
2.0
101
Hesitation
Rtr
6
0
0
0
0
0
0
0
0
0.7
286
0
0
0
0
0
18
12
0
6
4.0
168
48
42
18
54
12
24
30
12
36
0
27.6
63
Comp
74.8
77.2
13.2
11.7
22.0
27.4
52.9
38.1
7.7
36.1
74
13.7
0
45.8
53.3
67.1
76.2
68.2
83.4
67.0
52.7
54
0
49.0
68.0
59.0
63.0
39.0
52.0
65.6
82.3
38.5
51.6
44
Stumble
Rtr
0
0
96
108
18
0
6
24
30
31.3
133
36
42
48
42
42
36
66
90
42
49.3
36
30
84
138
168
196
66
60
84
30
60
91.6
62
Comp
24.1
35.2
31.7
20.9
31.2
0
36.7
19.8
48.5
27.6
49
98.0
48.3
53.0
41.4
44.5
54.6
75.6
-
60.8
59.5
32
44.3
81.5
94.0
113.8
106.5
91.5
67.8
87.1
46.9
62.4
79.6
30
Stall
Rtr
0
8
0
0
0
0
0
0
0
0.9
300
0
0
32
32
0
8
0
0
8
8.9
152
0
64
96
0
96
0
0
0
0
32
28.8
143
Comp
0
8
0
0
0
0
0
0
0
0.9
300
8
0
32
32
0
8
0
0
8
8.9
152
0
64
64
0
96
0
0
0
0
32
28.8
143
Total
Rtr Comp
6.0 103.8
92.0 127.6...
97.0 51.8.
108.0 54.1
23.0 57.3
0 63.1
6.0 123.0
29.0 60.6
39.0 56.2
44.4 77.5
97 40
37.0 120.7
45.0 48.3
84.0 131.6
74.0 133.3
46.0 111.6
63.4 140.9
81.0 143.8
92.0
57.2 140.2
64.4 121.3
30 26
79.0 46.1
190.4 201.3
253.4 232.2
212.4 175.5
309.0 266.9
94.0 130.7
90.2 121.5
96.0 153:i
76.6 130.5
93.0 133.7
149.4 159.2
57 39
Starting time used for calculating hard starting demerits was not measured by the
trained rater — only by the computer. The hard starting demerits shown are
included in the total for both the trained rater and computer.
Means and Standard Deviations are calculated from all available observations.
-------
F - 3
TEST FUEL 3 DEMERITS
(Phases II thru IV Results)
- Continued -
Fuel 3 Demerits
M
|_|
bl
cn
<;
OS
Oi
H
W
W
H
t— i
i_4
i— i
W
frt
UJ
<
•£
a.
H
en
W
H
>
I-H
W
en
^•t*
Si
pj
E-i
en
W
H
Test
Number
3
6
10
12
16
19
22
24
26
••••Mean
Std.
%
30
33
35
37
39
42
45
48
52
55
57
""'Mean
Std.
%
63
67
69
74
76
79
82
84
-••Mean
Std.
%
Hard
Idle
Roughness
Rater Starting* Rtr
A
B
A
B
A
A
A
B
B
Dev. ,
of Mean
A
B
B -
B
A
A
B
A
B
B
A
Dev. ,
of Mean
A
A
A
B
A
B
A
B
Dev. ,
o f Me an
0
0
0
0
0
1.0
0
0
0
0.1
300
1.8
0.4
0
0
1.4
0
0
2.8
0
0
1.4
0.7
140
0
0.2
0.2
0.2
0.4
0
0.4
0.4
0.2
73
0
0
1
5
0
0
0
6
5
1.9
139
0
5
4
3
0
0
3
0
4
3
0
2
100
0
0
0
2
0
5
0
4
1.4
150
Comp
3.9
5.9
24.4
5.9
3.9
29.4
—
3.7
4.2
10.2
102
0.0
0.1
0.0
1.2
17.3
9.1-
12.3
9.8
0.5
0.7
9.4
5.5
113
7.6
2.3
0.7
0.6
2.4
2.5
5.8
8.1
3.8
80
Hesitation
Rtr
0
0
0
0
0
0
0
0
0
0
0
36
0
0
12
18
24
0
6
6
0
6
9.8
120
0
54
6
0
12
0
36
0
13.5
152
Comp
119.7
51.7
44.4
7.1
60.3
56.6
42.4
5.8
4.8
43.6
83
7.1
6.6
0
49.4
68.7
52.6
52.3
46.6
59.3
56.7
47.0
40.6
59
26.0
52.0
20.0
21.0
52.0
65.6
60.5
130.5
53.5
67
Stumble
Rtr
18
36
24
18
6
6
0
30
60
22
84
42
84
60
96
30
66
90
72
72
66
24
63.8
37
24
6
12
72
78
84
78
48
50.3
64
Comp
60.2
32.6
38.3
20.3
46.9
20.2
20.9
20.9
51.5
34.6
44
108.3
-
114.6
62.2
50.3
72.2
70.8
73.1
65.6
-
46.3
73.7
32
41.9
29.1
39.2
87.0
102.7
86.7
68.7
65.9
65.2
41
Stall
Rtr
0
0
0
0
0
8
0
0
0
0.9
296
8
0
0
0
8
0
0
8
0
32
8
5.6
179
0
0
0
32
0
32
32
32
16
107
Comp
32
0
0
0
32
8
0
0
0
8.0
173
8
0
0
0
8
0
0
8
0
32
8
5.6
179
0
0
0
32
0
32
32
32
16
107
Total
Rtr
18.0
36.0
24.0
23.0
6.0
15.0
-
36.0
65.0
24.8
78
79.8
89.4
64.0
111.0
57.4
80.0
93.0
88.8
82.0
101.0
39.4
80.5
25
24.0
60.2
18.2
106.2
90.4
89.0
146.4
84.4
77.4
55
Comp
215.8
90.2
107. t
33.3
143.1
115.2
- .
30.4
60.5
99.5
62 .
125.2
-
114.6
112.8
145.7.
133.9
135.4
140.3
125.4
-
112.1
127.3
10
75.5
83.6
60.1
140.8
157.5
186.8
167.4
236.9
138.6
44
Starting time used for calculating hard starting demerits was not measured by
the trained rater — only by the computer. The hard starting demerits shown are
included in the total for both the trained rater and computer.
-------
APPENDIX G
Results of Analyzing Phase I
Tests with Consolidated Program
-------
G - 1
DEMERITS CALCULATED USING CONSOLIDATED COMPUTER PROGRAM
Phase I
Fuel 1 Demerits
Test
Hard
Stall
Number Starting* Rtr
1
2
3 •' ,
6
9
10
12
13
19.
26
30
31
34
35
38
40
42
43
47
50
53
59
65
68
73
77
•'"'"Mean
Std. Dev.,
% of Mean
2.8
4.4
2.4
15.6
2.4
0.2
2.0
0
3.0
4.0
1.8
12.4
1.8
1.4
4.6
2.2
1.2
2.0
1.8
2.2
1.8
1.8
1.8
2.2
1.4
1.8
3.1
110
272
224
136
208
272
208
176
0
320
304
272
240
208
144
168
112
144
200
112
104
112
296
264
80
208
104
188
43
Comp
312
232
136
216
288
304
208
0
320
304
272
240
240
144
168
104
144
200
120
104
112
296
264
80
240
104
198
44
Idle
Roughness
Rtr
5
6
1
6
14
12
7
0
5
4
3
5
1
3
3
1
10
8
7
0
8
1
6
0
8
5
5.0
74
Comp
_ '
-
2.6
10.4
10.6
11.5
9.0
9.9
15.5
9.6
5.8
7.2
3.6
2.3
8.5
5.4
11.2
6.5
9.8
4.7
8.7
0.1
4.7
4.8
6.5
5.7
7.3
,48
Stumble
Rtr
168
168
330
336
696
282
306
90
294
300
270
354
150
138
96
148
228
162
114
90
162
108
180
66
192
138
214
61
Comp
_
-
69
80
95
57
55
18
58
128
95
97
48
67
52
67
82
64
45
39
73
64
49
17
79
94
66
39
Hesitation
Rtr
36
60
102
0
42
18
12
0
0
0
0
12
0
0
0
0
0
0
0
0
0
0
0
6
0
0
12
200
Comp
"_
-
62
54
70
75
50
12
52
32
27
48
28
27
37
58
41
40
40
34
41
41
47
57
35
51
44
33
Run
Rtr
484
462
571
566
1026
520
503
90
622
612
547
623
361
286
272
263
383
372
235
196
284
407
452
154
409
249
421
46
Total
Comp
_
-
272
376
465
448
324
40
448
477
401
405
322
242
271
236
279
312
217
184
237
404
366
161
362
256
313
35
Starting time used for calculating hard starting demerits was not measured
by the trained rater — only by the computer. The hard starting demerits
shown are included in the total for both the trained rater and computer.
** Means and Standard Deviations are calculated from all available obser-
vations .
-------
G - 2
DEMERITS CALCULATED USING CONSOLIDATED COMPUTER PROGRAM
Phase I
- Continued -
Fuel 2A Demerits
Idle
Test
Number
4
5
7
8
11
14
17
18
22
28
29
33
39
41
46
52
54
56
58,
61
64
66
69
72
74
80
81
**Mean
Std. Dev. ,
% of Mean
Hard
Starting*
15.8
0
4.4
0
0
0.8
0
0
0
0
0
0
14.2
0
0
0
0
0
0.6
0
0.8
0.4
0
0
1.2
0
1.0
1.6
252
Stall
Rtr
0
0
0
0
72
32
0
0
0
0
0
0
0
0
0
0
0
32
0
0
32
32
0
0
0
0
0
7.4
234
Comp
8
0
0
0
72
32
0
0
0
0
0
0
8
0
0
0
0
0
0
0
32
0
0
0
0
0
8
5.9
267
Roughness
Rtr
1
0
0
3
-
0
1
2
1
1
1
2
1
0
0
0
0
0
0
1
0
0
0
0
0
1
0
.6
134
Comp
0.5
0.7
7.2
0.9
-
1.2
1.9
2.1
1.0
0.4
1.5
1.6
1.0
3.0
4.6
2.6
2.0
1.4
1.3
3.3
2.8
1.1
1.5
1.4
1.4
8.6
2.6
2.2 '
88
Stumble
Rtr
18
18
108
162
24
210
12
18
30
30
30
30
6
6
6
12
0
6
6
24
12
6
18
24
6
18
6
31
162
Comp
28
14
8
10
21
57
23
8
24
13
22
11
13
i
24
33
27
7
66
15
10
18
11
8
22
6
33
11
20
72
Hesitation
Rtr
0
0
18
6
12
0
0
0
6
0
0
0
0
0
0
0
0
0
12
0
0
0
0
0
0
0
6
2
238
Comp
26
27
25
28
57
76
28
57
34
49
29
11
30
11
26
18
62
48
45
45
25
57
29
35
0
19
19
34
52
Run
Rtr
35
18
130
171
-
243
13
20
37
31
31
32
21
6
6
12
0
38
19
25
45
38
18
24
7
19
13
40
139
Total
Comp
78
42
45
38
-
168
53
68
59
62
53
24
66
38
64
48
71
76
62
59
78
69
39
58
9
61
41
59
47
Starting time used for calculating hard starting demerits was not measured
by the trained rater — only by the computer. The hard starting demerits
shown are included in the total for both the trained rater and computer.
Means and Standard Deviations are calculated from all available obser-
vations .
-------
G - 3
DEMERITS CALCULATED USING-CONSOLIDATED COMPUTER PROGRAM
Phase I. .
- Continued -
Fuel 3 Demerits
Idle
Test
Number
15
16
20
21
23
24
25
27
32
36
37
44
48
49
51
55
57
62
70
76
79
**Mean
Std. Dev. ,
% of Mean
Hard
Starting'-
0.8
0.2
1.8
0.8
0.2
0
7.6
0.2
0
0
15.6
0.2
0.2
16.0
0.2
0.2
0
0.2
0
1.0
0.2
2.2
219
Stall
Rtr
32
96
104
32
96
64
32
64
32
0
0
0
32
0
32
0
0
32
32
64
0
35
98
Comp
32
96
112
64
104
64
40
64
32
0
8
0
32
8
32
0
0
32
32
64
0
39
91
Roughness
Rtr
0
1
6
0
3
3
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0.7
217
Comp
1.2
14.2
9.4
2.0
16.2
7.9
12.5
4.0
0.8
-
1.5
0.5
5.9
0.4
9.0
1.6
1.3
0.8
0
1.7
4.3
4.8
105
Stumble
Rtr
210
102
204
150
132
102
120
210
102
54
60
18
30
12
18
6
18
48
36
42
36
81
83
Comp
57
42
13
17
41
40
75
53
41
29
42
16
45
15
34
11
22
21
121
18
20
37
69
Hesitation
Rtr
0
0
6
0
6
6
0
0
0
0
0
6
0
0
0
0
6
0
0
0
0
2
173
Comp
76
58
74
60
46
40
69
57
52
30
44
68
30
71
46
53
47
24
35
37
26
50
32
Run
Rtr
243
199
322
183
237
175
160
275
134
54
76
24
62
28
50
6
24
81
68
107
36
121
77
Total
Comp
168
211
210
143
207
151
204
179
126
-
Ill
85
113
110
121
66
70
78
188
121
51
136
38
* Starting time used for calculating hard starting demerits was not measured
by the trained rater — only by the computer. The hard starting demerits
shown are included in the total for both the trained rater and computer.
** Means and Standard Deviations are calculated from all available obser-
vations .
------- |