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 .

-------