LDTP  78-05
            Technical Support Report for Regulatory Action
               Effects of Dynamometer Variability on the
                     Confirmation of Requests for
                Alternate' Dynamometer Power Absorption
                                  by

                           Glenn D. Thompson
                             March, 1978
                                NOTICE

Technical support reports for regulatory action do not necessarily
represent the final EPA decision on regulatory issues.  The are in-
tended to present a technical analysis of an issue and recommendations
resulting from the assumptions and constraints of that analysis.  Agency
policy constraints or data received subsequent to the date of release of
this report may alter the conclusions reached.  Readers are cautioned to
seek the latest analysis from EPA before using the information contained
herein.
               Standards Development and Support Branch
                 Emission Control Technology Division
             Office of Mobile Soure Air Pollution Control
                  Office of Air and Waste Management
                 U.S. Environmental Protection Agency

-------
I.   Purpose

This report discusses the variations which have been observed among the
EPA light-duty vehicle dynamometers during vehicle-dynamometer coast-
downs.  The purpose of this investigation is to recommend a tolerance
value for the allowable variation between the "target" coastdown time
interval, supplied by the manufacturer and the time interval obtained
from the EPA vehicle-dynamometer confirmatory test.  This tolerance is a
measure of the acceptable error between the road experience of the
vehicle as reported by the manufacturer and the simulation by the dyna-
mometer during the EPA tests.  As such the tolerance is a direct limit
on the road load simulation error which EPA will accept and still consi-
der the test results valid.

II.  Background

The revised version of Advisory Circular 55, AC/55/A, states that if a
manufacturer requests an alternate dynamometer adjustment, then the man-
ufacturer will submit a "target" dynamometer coastdown time interval
which reflects the road experience of the vehicle and a requested dyna-
mometer adjustment which the manufacturer believes will result in a
coastdown of the vehicle-dynamometer system matching the "target" time
interval.  The vehicle, and related test data are to be considered
unrepresentative if the confirmatory vehicle-dynamometer coastdown time
intervals are not within some tolerance of the reported "target" coast-
down time.

The probelm is to choose an acceptable tolerance limit.  This tolerance
should be large enough that a manufacturer is unlikely to be penalized
by the random yariations among the EPA dynamometers, yet sufficiently
stringent to avoid acceptance of consistently biased dynamometer power
absorption requests.

III. Discussion

The data which are presented were collected since December 1, 1977.  All
data were obtained by using an electronic stop watch which automatically
indicated the time interval required for the vehicle-dynamometer to
coastdown from 55 to 45 mph.  All speed measurements were obtained from
the front roll of the dynamometer.  This is the dynamometer roll which
is connected both to the dynamometer power absorber and the mechanical
flywheels used for the simulation of the vehicle inertia.  This is the
type of instrumentation and sensing location which is currently required
in the EPA Recommended Practice for Vehicle-Dynamometer Coastdowns.

Using these techniques, data have been collected from all four dyna-
mometers currently in service at EPA for light-duty vehicle exhaust
emission certification and fuel economy measurements.  Data were ob-
tained from five vehicles, a 1969 Plymouth routinely used as an EPA
warm-up vehicle, a Ford Cougar, an LTD, a Fiesta, and a Mercury Marquis.
The last three vehicles were rental vehicles provided by the Transpor-
tation Research Center of Ohio.

-------
                                      -2-
   The test procedures used and the data obtained from each of the vehicles
   are presented in the following subsections.

        A.   EPA Plymouth

   The vehicle used in this test series was a 1969 Plymouth used at EPA for
   dynamometer warm-up.  The data presented here were obtained by the
   Laboratory Branch personnel in their recent investigation of dynamometer
   variability.  During this investigation the dynamometer simulated inertia
   was 5000 Ibs. for all coastdowns.  To maintain approximately the same
   load on the rear axle, the same driver was used for all tests, and the
   fuel tank was refilled after each test.  Prior to the tests on each
   dynamometer, the vehicle was operated over two EPA Highway Fuel Economy
   Driving Cycles to insure warm-up of the vehicle and the dynamometer.

   Three dynamometer coastdowns were conducted at each of seven nominal
   actual horsepowers.  These data, and the mean of the three coastdown
   times are presented in the tables of Appendix A.  The summary of these
   data, the mean coastdown times, by dynamometer are presented in Table 1
   and plotted in Figure 1.
   The range of these coastdown intervals are approximately 6 percent of
   the mean values.  In the "worst" case, the range of the data at the
   nominal 15.5 actual horsepower point was 7.7 percent of the mean value
   at this point.  The parallel nature of the dynamometer results should be
   noted.  This strongly indicates that much of the observed variations
   among the dynamometers is a consistent effect which might be removed by
   improved dynamometer calibration or improved dynamometer technology.
   These variations might, however, also be the effect of changes in the
   tire rolling resistance caused by variations in the vehicle warm-up or
   variations in the drive axle load.
                                       Table 1

                                    1969 Plymouth

  Nominal          Mean Coastdown Times
  Actual                   (sec)
Power (Hp)  Dyno //I  Dyno #2  Dyno //3  Dyno #4

  15.5

  14.5

  13.5

  12.5

  11.5

  10.5

   9.5
Average Coastdown   Range of Dyno
Time for all Dynos Coastdown Times
	(seconds)       (Min-Max, sec)
13.15
13.77
14.46
15.17
15.91
16.83
17.83
13.35
13.96
14.64
15.44
16.21
17.05
17.84
12.50
. 13.12
13.80
14.42
15.29
16.20
17.10
12.36
13.38
14.13
14.94
15.81
16.59
17.35
     12.84

     13.56

     14.26

     15.00

     15.80

     16.67

     17.53
0.99

0.84

0.84

1.02

0.92

0.85

0.74

-------
                                        -3-
                                  1969 Plymouth

                     Mean Vehicle-Dynamometer Coastdown Time
                                     versus
                   Nominal Actual Dynamometer Absorption Power
a
01
co
CO
flj
0
o
    18
    17
16
    15
14
    12
    11
                                           = Dynamometer 1
                                           = Dynamometer 2
                                           = Dynamometer 3
                                           = Dynamometer 4
                    —»—
                     10
                            •4-
                               •4-
                                                   -f-
                        11     12     13      14      15

                          Nominal Actual Power  (Hp)
16
                                    FIGURE 1

-------
                                   -4-
     B.   Ford Cougar

The Ford Cougar was a vehicle which the EPA Certification Division
tested at the request of Ford.  The vehicle was supplied and prepared by
Ford.  Rear axle load information was not obtained by EPA, however rear
axle load was believed to remain constant since the same driver was used
in all tests and since an auxiliary fuel tank was used during the tests.

The vehicle warm-up for these tests consisted of two EPA Highway Fuel
Economy Cycles.  This warm-up procedure was performed prior to each
series of dynamometer coastdowns.  In addition, the front roll speed
calibration was checked before and after each series of coastdowns.

The individual coastdown data obtained from each dynamometer at each of
the seven nominal adjustments are given in the tables of Appendix B.
The summary of these data and the mean coastdown times are presented in
Table 2 and plotted in Figure 2.

The range of these data are, on the average, 3 percent of the mean
values.  In the worst case, the range of the data at the nominal power
of 12.5 horsepower was approximately 4 percent of the mean value at this
point.

Again the parallel nature of the coastdown time versus actual power
lines indicate much of the observed variations are systematic.  The
reduced range of the variations for this vehicle are probably indicative
of the greater care taken in both the vehicle and dynamometer preparation
prior to these tests compared to the care taken in the Plymouth tests,
the first vehicle in this series.

     C.   Ford LTD (TRC rental vehicle)

This vehicle was a rental vehicle supplied by TRC for a recent project.
This test series was designed to duplicate the vehicle and dynamometer
conditions which occur during the normal certification process.  In this
case, the vehicle was allowed to equilibrate for at least four hours
between any dynamometer tests to minimize any possible effects of syste-
matically increasing tire temperatures.  In each case the vehicle-dyna-
mometer warm-up consisted of one EPA Urban Driving Cycle followed by two
EPA Highway Fuel Economy Driving Cycles.  The load on the drive axle was
maintained at approximately the same level by refilling the fuel tank of
the vehicle between each dynamometer sequence.  No unusual dynamometer
calibrations or calibration checks were performed during the test sequence.
The dynamometer inertia simulation was adjusted to 5000 pounds for all
tests.

Again coastdown data were obtained at each of seven nominal actual
dynamometer power adjustments.  These are given in the tables of Ap-
pendix C.  The summary of these data and the mean coastdown times at
each dynamometer adjustment are given in Table 3 and plotted in Figure
3.

-------
Table 2
Ford Cougar
Nominal
Actual
Power
(HP)
15.5
14.5
13.5
12.5
11.5
10.5
9.5

Dyno #1
13.22
13.97
14.55
15.08
15.94
16.67
17.80
Mean
Dyno #2
13.42
13.95
14.56
15.20
16.08
16.96
17.69
Coastdown
Dyno #3
12.93
13.56
14.06
14.62
15.62
16.52
17.17
Times (sec)
Dyno #4
13.00
13.64
14.16
14.78
15.73
16.66
17.15

Dyno #4*
13.17

14.40
15.14
15.91
16.86
17.74
Average Coastdown
Time for all
Dynamometers
(sec)
13.15
13.78
14.35
14.96
15.86
16.73
17.51
Range of
Dynamometer
Coastdown Times
(Max-Min, sec)
0.49
0.41
0.50
0.58
0.46
0.44
0.65
* During the time period required for the collection of these data a routine monthly calibration
  of Dynamometer #4 was conducted.   The tests were subsequently rerun on Dynamometer #4 to inves-
  tigate the effects of this recalibration.

-------
                                     -6-
                                 Ford Cougar

                   Mean Vehicle-Dynamometer Coastdown Time
                                   versus
                Nominal Actual Dynamometer Absorption Power
o
a)
CO
o
u
    18
    17
Ji   16
H
    15
    14
    13
    12
Dynamometer 1

Dynamometer 2

Dynamometer 3

Dynamometer 4
Dynamometer 4 after
recalibration
                -f	1	'       '        '	
                 10     11      12     13     14

                          Nominal Actual  Power (Hp)
     15
16
                                  FIGURE 2

-------
                                       -7-


                                       Table 3

                            Ford LTD (TRC Rental Vehicle)

  Nominal          Mean Coastdown Times             Average Coastdown   Range of Dyno
  Actual                   (sec)                    Time for all Dynos Coastdown Times
Power (Hp)  Dyno //I  Dyno #2  Dyno #3  Dyno #4          (seconds)       (Min-Max, sec)
13.5
12.5
11.5
10.5
9.5
8.5
7.5
15.
15.
16.
17.
18.
19.
21.
12
96
61
67
63
96
23
15.34
16.18
16.74
17.80
18.71
20.22
21.49
14.
15.
16.
16.
18.
19.
20.
41
54
25
74
05
57
58
15.
16.
16.
17.
18.
20.
21.
12
06
88
86
64
19
30
15.
15.
16.
17.
18.
19.
21.
00
94
62
52
51
98
15
0
0
0
1
0
0
0
.93
.64
.63
.06
.66
.65
.91

-------
                                     -8-
                       Ford LTD (TRC Rental Vehicle)


                  Mean Vehicle-Dynamometer Coastdown Time
                                  versus
                Nominal Actual Dynamometer Absorption Power
o

-------
                                    -9-
The range of the data was approximately 5 percent of the mean values.
In the worst case, at a nominal 10.5 horsepower, the range of the data
was 6 percent of the mean coastdwon time at this power.  Again the
coastdown time versus the dynamometer power adjustments are approxima-
tely parallel lines indicating there was a systematic effect between
various dynamometer tests.

     D.   Ford Fiesta (TRC rental vehicle)

This 1978 Ford Fiesta was a rental vehicle supplied by the Transporta-
tion Research Center of Ohio.  The data from this vehicle were collected
in the same manner as were the data from the Ford LTD.  These data are
presented in Appendix D.  The summary data are presented in Table 4 and
plotted in Figure 4.

These data show greater variations among the dynamometers than have been
observed for any of the previous vehicles.  These variations were parti-
cularly severe at the lowest test powers.  At these powers, 4.3 and 3.2,
actual dynamometer horsepower the range of the variations were almost 14
percent of the coastdown times.  It should be noted that these powers
are, however, lower than currently used for exhaust emission certifica-
tion or fuel economy testing at EPA.  The lowest power adjustment pro-
vided by the weight based table was 5.9 Hp for a vehicle of less than
1,125 pounds.  The lowest power adjustment which has been accepted as an
alternate dynamometer adjustment has been 5.4 horsepower requested by
Ford for the Fiesta.  In the test range above 5.4 horsepower the varia-
tions were approximately 10 percent of the coastdown times.  The worst
case variation was 11 percent, which occurred at the nominal actual
power of 5.4 horsepower.

It should also be noted that the variations in the case of the Fiesta
were much more random in nature than the variations with the previous
vehicle.  This may indicate greater dynamometer variability in the very
low power absorption regions or may  indicate that variations in tire
rolling resistance are more predominant.  It is also possible that
slight variations in the position of the drive wheels would have a
greater effect in the case of front wheel drive vehicles than in the
case of rear wheel drive vehicles.  Since this positioning could be
readily changed by test driver motion of the vehicle steering wheel,
these effects might appear as random variations.

     E.   Ford Mercury Marquis (TRC rental vehicle)

This vehicle was also a rental vehicle supplied by the Transportation
Research Center of Ohio.  The vehicle test program was conducted in the
same manner as the test program for the Ford LTD and the Ford Fiesta.
The data are presented in Appendix E, summarized in Table 5 and plotted
in Figure 5.

The variations observed in the case of the Mercury Marquis were very
similar to those observed for the Cougar, and the Ford LTD.  The average
observed range of coastdown times, maximum minus the minimum time, was
about four percent of the mean coastdown time.

-------
                                     -10-


                                    Table 4

                        Ford Fiesta (TRC Rental Vehicle)

  Nominal          Mean Coastdown Times             Average Coastdown   Range of Dyno
  Actual                   (sec)                    Time for all Dynos Coastdown Times
Power (Hp)  Dyno #1  Dyno #2  Dyno #3  Dyno #4          (seconds)       (Min-Max, sec)
9.7
8.6
7.6
6.5
5.4
4.3
3.2
8.
8.
9.
10.
12.
12.
14.
24
89
98
45
48
60
26
8.
8.
9.
10.
11.
11.
12.
39
56
57
20
70
60
97
8.43
9.06
9.22
10.69
11.28
11.95
13.60
8.
9.
10.
11.
11.
13.
14.
71
38
17
02
87
33
89
8.
8.
9.
10.
11.
12.
13.
44
97
73
59
83
37
93
0
0
0
0
1
1
1
.47
.82
.95
.82
.2
.73
.92

-------
                                     -11-
                      Ford Fiesta (TRC Rental Vehicle)



                   Mean Vehicle-Dynamometer Coastdown Time

                                   versus

                 Nominal Actual Dynamometer Absorption Power
o
01
CO
•u
CO
a
o
     15  •
     14
     13
     12
11
     10
                                    •= Dynamometer 1

                                    D = Dynamometer 2

                                    •fc= Dynamometer 3

                                    f= Dynamometer 4
                                                            10
                            Nominal Actual Power (Hp)
                                  FIGURE 4

-------
                                       -12-



                                    Table 5

                       Mercury Marquis (TRC Rental Vehicle)

  Nominal          Mean Coastdown Times             Average Coastdown   Range of Dyno
  Actual                   (sec)                    Time for all Dynos Coastdown Times
Power (Hp)  Dyno //I  Dyno #2  Dyno //3  Dyno #4          (seconds)       (Min-Max, sec)
13.0
12.0
11.0
10.0
9.0
8.0
7.0
15.
15.
16.
17.
19.
20.
21.
25
93
93
76
30
17
58
15.42
16.28
16.98
17.98
18.75
20.37
21.73
14.
15.
16.
17.
18.
19.
20.
66
61
22
31
81
49
96

15.
16.
17.
18.
19.
21.

87
61
70
76
81
14
15.
15.
16.
17.
18.
19.
21.
11
92
69
69
91
96
35
0.
0.
0.
0.
0.
0.
0.
76
67
76
67
55
88
62

-------
                                   -13-
               Ford Mercury Marquis (TRC Rental Vehicle)



               Mean Vehicle-Dynamometer Coastdown Times

                                versus

              Nominal Actual Dynamometer Absorption Power
a
0)
co
4J
CO
n!
O
o

a
«
a)
     22
     21
     20
     19   "
     18-   "
17   ••
     16
     15
                                      = Dynamometer 1

                                      = Dynamometer 2


                                      = Dynamometer 3

                                      = Dynamometer 4
                             t        I       I      ".•

                            9      10      11       12


                         Nominal Actual Power  (Hp)
                                                               13
                               FIGURE 5

-------
                                  -14-


Again, the lines of dynamometer power absorption versus coastdown times
for the different dynamometers are generally parallel.  The exception
seems to occur at a monimal power of 9 horsepower, where some conver-
gence of the data occurs.

IV.  Analysis

This section is divided into three subsections.  The first subsection
analyzes the observed dynamometer variability directly.  The second sub-
section investigates the effect this dynamometer variability has on the
acceptance or rejection of requests for alternate dynamometer power
absorption.  The third subsection summarizes the previous two.

     A.   Analysis of Dynamometer Variability

The vehicle-dynamometer coastdown variations observed with the Plymouth,
Cougar, LTD and Marquis were similar.  The variations observed with the
Fiesta were greater.  These results are summarized in Table 6.

In general, the range of variation for heavy conventional drive vehicles
appears to be about 5 percent.  The "worst case" range of variations for
these vehicles was about 6%.  These results are increased somewhat by
the data obtained from the Plymouth which had somewhat greater varia-
bility than the data from the other rear wheel drive vehicles.  The
greater variability of the Plymouth data probably occurred since this
was the first vehicle tested, and somewhat less care was taken with this
test series.

                                Table 6

                   Summary of Dynamometer Variations
                      (Expressed As Percentage)

                        Average of the Observed        "Worst Case"
                          Range of Variations      Range of Variations
     Vehicle           [(max - min)/(mean)]100% [(max - min)/(mean)]100%

     Plymouth

     Cougar

     LTD

     Fiesta

     Marquis
Average for heavy                  ,  _                      _
conventional drive vehicles

     Fiesta                       10.0                     14.0*
5.9
3.3
4.5
10.0
4.1
7.7
3.9
6.2
14.0*
5.1
*  This extreme range of variability occurred at dynamometer power
absorptions lower than any value which has currently been used for
EPA testing.

-------
                                  -15-


It should be noted that for all of the heavy conventional drive vehicles,
the curves of the coastdown times versus the dynamometer power absorption
were approximately parallel.  In addition, the dynamometers were usually
ranked in the same order.  The sequential dynamometer-to-dynamometer
relationships were not the same for all dynamometers, therefore, there
is no reason to believe that any change in vehicle warm-up or weight
variations could induce this systematic effect.  The systematic nature
of the dynamometer ranking strongly indicates that even the small re-
maining dynamometer variations can be reduced by improved calibration
techniques.

In the case of the Fiesta, the variability was much greater than for the
heavy conventional drive vehicles.  In addition, the variability appeared
more random in nature.  Much of this increased variability, when evalu-
ated as a percentage of the coastdown time, occurred because the coast-
down times were shorter for this vehicle.  This explains most of the
increased variability observed at the larger dynamometer power absorp-
tions used with the vehicle.  In the very low power absorption ranges,
however, the variations in absolute magnitude of the coastdown times
were greater for this vehicle than for the remaining vehicles.  This is
believed to be caused by two effects.

First, this power absorption range is below the normal dynamometer power
absorption range for certification tests and is not used in the dynamo-
meter calibration.  Therefore, greater dynamometer variability and
potential error is expected in this region.  In addition, small changes
in dynamometer power absorption will have very significant changes in
the coastdown time of the vehicle-dynamometer system.

Secondly, it is possible that this variability could result from random
driver induced variations in the tire-roll alignment for front wheel
drive vehicles.  Again, small changes in the power requirements of the
system could have significant effects on the coastdown times because of
the lower total system power absorption.

     B.   The Effects of Dynamometer Variability with Respect to
          Acceptance or Rejection of a Request for an Alternate
          Dynamometer Power Absorption

The purpose of this investigation is to recommend a tolerance value for
the allowable variation between the "target" coastdown time interval
supplied by the manufacturer and the time interval obtained from the EPA
vehicle-dynamometer confirmatory test.  Tolerance values which have been
suggested are 5, 7 and 10 percent of the EPA coastdown time.  The poten-
tial results of each of these values will be investigated in this section.

In the first situation it is hypothesized that a manufacturer submits a
true, correct value, for the dynamometer power absorption for his vehicle.
The number and percentages of the times this request would be challenged
is investigated.  The second situation hypothesizes that a manufacturer
conducts the original dynamometer power absorption determination on a

-------
                                  -16-
dynamometer randomly selected from a dynamometer population equivalent
to the EPA dynamometers.  The question is again the number of times
this request would be challenged at each of the proposed tolerance
levels when the confirmatory tests were conducted on randomly selected
EPA dynamometers.  The third situation is somewhat the opposite of the
first two.  In this case it is hypothesized that a manufacturer submits
a dynamometer adjustment request which is inappropriate by approximately
one horsepower.  The question is, again, how often this request would be
challenged.

     1.   The probability that a request for a true, correct value for
the dynamometer adjustment would be rejected because of EPA dynamometer
variability.

For this evaluation it is assumed that each nominal dynamometer adjust-
ment is a true, correct dynamometer adjustment for the mean observed
coastdown time.  The question for this case is; if each nominal dyna-
mometer power absorption were requested as the dynamometer adjustment
cooresponding to the mean observed dynamometer coastdown time, how often
would this request be rejected assuming random selection of the EPA
dynamometer for confirmation testing.   For example considering the first
set of data from the 1969 Plymouth, Table 1, the question is would a
request for a dynamometer adjustment of 15.5 horsepower submitted with a
dynamometer coastdown time of 12.84 seconds be rejected when confirma-
tory tested on any of the EPA dynamometers.

Since it is proposed that the tolerance be unidirectional the only
rejections would occur when a dynamometer yields a coastdown longer than
the mean value plus the tolerance.  Therefore, the sum of the mean
values for the dynamometer coastdowns at each of the nominal dynamometer
adjustments plus each of the proposed tolerances were calculated.  The
occasions when any observed dynamometer coastdown times exceeded these
values were recorded.  These are the instances when a true, correct mean
value would be rejected in an EPA confirmatory test.

Table 7 summarizes the number of instances, based on all of the observed
data, where such a rejection of the dynamometer request would occur.  In
the range of dynamometer power absorptions which have been used at EPA
only one rejection of such a dynamometer request would occur.  This re-
jection case was observed for the Ford Fiesta at the 5 percent toler-
ance value.

     2.    The probability of rejection when a dynamometer power absorp-
tion is determined from a dynamometer randomly selected by a manufac-
turer and confirmatory tested on an EPA dynamometer randomly selected.

This case assumes a manufacturer randomly selects a single dynamometer
to perform the vehicle-dynamometer coastdown part of the recommended
practice for Determination of Vehicle Road Load and submits the results
from that dynamometer.   It is further assumed that the population of the
dynamometers are the same for both EPA and the manufacturer.  With this

-------
                                  -17-
                              Table 7

            Confirmations and Rejections of Dynamometer
              Requests Which are True Correct Values

Tolerance      Number of      Number of      Observed Probability
  Value      Confirmations   Rejections          of Rejection+

                                                       2.1%

                                                       0.7%

                                                       0.0%
*   Two of the three observed failures were at power absorption
    settings lower than have yet been used in EPA certification
    or fuel economy testing.

**  This failure occurred at a power absorption setting lower than any
    power adjustment which has yet been used in EPA certification or
    fuel economy testing.

+   The observed probability of rejection is th computed ratio of the
    number of rejections to the total number of observation,  145.
5%
7%
10%
142
144
145
3*
1**
0

-------
                                  -18-
assumption it is possible to use the results of the EPA tests to predict
the probability a request for an alternate dynamometer adjustment would
be rejected under these conditons since this is equivalent to the
hypothetical case in which a road load determination would be performed
on one randomly selected EPA dynamometer and then the confirmation test
performed on a second, randomly selected dynamometer.

The test data were examined to ascertain the number of possible combin-
ations which would result in acceptance or rejection at each tolerance
level if a requested dynamometer adjustment were determined on one
randomly selected EPA dyanmometer and then confirmatory tested on a
second EPA dynamometer.  For example, the data of the 1969 Plymouth,
Table 1 first line, was examined to see if a request for 15.5 horsepower
submitted with the coastdown of any of the dynamometers 1 through 4
would be rejected if confirmatory tested on any of the dynamometers
1 through 4.

Again, a unidirectional tolerance was assumed so that those vehicles
with conservative dynamometer adjustment requests would not be rejected.
Therefore, in this example if the coastdown time was 12.5 seconds, cor-
responding to 15.5 horsepower on dynamometer 3, this request would be
rejected if confirmatory tested at a 5 percent tolerance value on dyna-
mometers 1 and 2.  The request would be accepeted in confirmatory tested
on dynamometers 3 or 4.  Likewise, a request for 15.5 horsepower corres-
ponding to 12.36 seconds as obtained from dynamometer 4 would be rejected
if confirmatory tested on dynamometers 1 or 2 at the 5 percent tolerance
level, and accepted on dynamometer 3 and 4.  Requests based on dynamo-
meters 1 and 2 would be accepted if confirmatory tested on any dynamometer
at the 5 percent tolerance level.  Therefore, for the nominal power of
15.5 for the sixteen possible combinations of dynamometers which could
be used for power absorption and subsequent confirmation, four combina-
tions would result in rejection of the 5 percent tolerance value.

Table 8 summarizes the number of possible combinations which would
result in acceptance or rejection of the submitted request at each of
the three suggested tolerance values.  In this instance, considering
only the range of dynamometer adjustments which have yet been used at
EPA rejections were observed in 3.6 and 0.9 percent of the observations
at tolerance levels of 5 and 7 respectively.

     3.   The probability of accepting inappropriate dynamometer adjust-
ment requests.

The previous sections discussed the possiblity of rejecting a request
for an alternate dynamometer adjustment in the cases when that request
was a true correct value and in the case when any inaccuracies resulted
from random dynamometer variability.  Naturally, the probability of
rejection diminishes as the allowable tolerance increases.  For this
reason it may appear desireable to adopt a rather large tolerance.
However, a large tolerance also means that EPA will accept data from
vehicles which will have been inappropriately tested.  This section
considers the probability that an incorrect dynamometer adjustment
request would be accepted at various tolerance levels.

-------
                                        -19-
                                   Table 8

             Confirmations and Rejections of Dynmometer Requests
       Assuming Random Selection of both the Test and the Confirmation
                                Dynamometers
          Possible Combinations    Possible Combinations         Percent Rejected
         Resulting in Acceptance   Resulting in Rejection   Based on Total Observations
                                             36*                      5.9%

                                             13**                     2.1%

                                             52***                    0.8%
*    8 of the observed 36 failures occured at dynamometer power adjustment
     settings lower than any yet used for certification or fuel economy
     testing.

**   6 of the observed 13 failures occured at dynamometer power adjustment
     settings lower than any yet used for certification or fuel economy
     testing.

***  3 of the observed 5 failures occured at dynamometer power adjustment
     settings lower than any yet used for certification or fuel economy
     testing.
5%
7%
10%
571
594
602

-------
                                  -20-
In this section it is assumed that a manufacturer submits a correct
dynamometer coast down time, but requests a dynamometer power absorption
which is approximately one horsepower less than the true correct power
for that coast down time.  For example, in the case of the 1969 Plymouth,
first line of Table 1, the assumption is that the manufacturer would
submit a coast down time of 12.84 seconds as corresponding to 14.5
horsepower.

Table 9 summarizes the probabilities that such requests would be accepted
because of the dynamometer variability and the tolerance values.  At the
10 percent tolerance value there is greater than 90 percent probability
that EPA would accept a dynamometer adjustment which was in error by
approximately one horsepower.  This prabability decreases as the toler-
ance value decreases, to 32 percent probability at a 5 percent tolerance
value.

     C.   Summary

The choice of an acceptable tolerance value must be a compromise between
the possiblity of rejecting a legitimate alternate dynamometer request
versus the risk of accepting inappropriate dynamometer adjustments.
Alternately stated, an undue burden should not be imposed on the manu-
facturers, yet the representativeness of the EPA tests must be maintained.

Table 6 demonstrates the average variability in vehicle-dynamometer
coastdown times is about 4 percent for conventional vehicles and is
somewhat higher, approximately 10 percent for the Ford Fiesta, a very
small front wheel drive vehicle.

Table 7 gives the probability that a submitted true correct mean value
for a dynamometer power absorption request and coastdown time interval
would be rejected because of EPA dynamometer variability.  Rejection
under these conditions implies the manufacturer has performed in the
best technically correct manner which can be expected, without supplying
a conservative request, and is rejected because of the variability in
the EPA dynamometers.

Rejection under these conditions is considered inappropriate, therefore,
the 5 percent tolerance criteria is considered inappropriate, even
though this would result in only about 1 percent of the currently obser-
ved test vehicles under these conditions.  In the dynamometer power
absorption regions currently used for EPA testing, no rejections of true
mean value requests would occur at either the 7 or 10 percent tolerance
value.                             '

Table 8 may be considered as the probability that a dynamometer power
absorption request and the associated coastdown time would be rejected
considering the dynamometer variability of both the original test dyna-
mometer and the confirmation dynamometer.  This table indicates that, in
the region of dynamometer power absorption currently used for EPA
testing, there is about 1 percent probability of rejection with the
7 percent tolerance value.  Furthermore, for conventional vehicles, only

-------
                             -21-
                           Table 9

Acceptance or Rejection of Requested Dynamometer Adjustments
          Differing by Approximately One Horsepower
                   From the Correct Value
Tolerance
Value
5%
7%
10%
Number of
Acceptances
40
89
115
Number of
Rejections
85
36
10
Percentage of
Acceptances
32%
71%
92%

-------
                                  -22-
two instances of a rejection were observed, indicating a probability for
rejection of only 0.33 percent.  The probability of rejection is of
course, even less with the 10 percent tolerance value.

Table 9 provides the probability that EPA will accept an inappropriate
dynamometer adjustment with each of the tolerances.  This table demon-
strates that if a tolerance value of 10 percent is chosen there is more
than 90 percent chance of accepting dynamometer adjustment requests
which are in error by approximately 1.0 horsepower.  It is considered
that this is an unacceptable risk for EPA.  From the standpoint of test
accuracy a tolerance value of 5 percent is the value which appears most
desireable.  Only at this level is the probability of rejecting a re-
quested dynamometer setting which is incorrect by approximately 1.0
horsepower greater than the probability of acceptance.

At the current time a tolerance value of 7 percent is considered most
appropriate.  This value provides a minimal probability that EPA will
reject an appropriate dynamometer request.  Within the range of cur-
rently used dynamometer adjustments, no instances were observed where a
true mean dynamometer adjustment would be rejected.  Within the range of
currently used dynamometer adjustments there is only about 1 percent
probability that a dynamometer adjustment would fail a confirmation
test.  The very few questions which may arise can be expected to occur
for very small, probably front wheel drive vehicles.  Minor additional
care for these small number of vehicles of this type should not impose a
significant burden on either EPA or the manufacturers.  For example, a
manufacturer can eliminate even this small risk by determining a dyna-
mometer adjustment from several dynamometers and requesting the mean
value.  It is particularly appropriate that additional care be taken in
assuring the representativeness of the dynamometer test of these vehicles
since the test results from such vehicles are very sensitive to variations
in the dynamometer adjustment.

V.   Conclusions

It is concluded that a 7 percent allowable tolerance between the submit-
ted coastdown time and the coastdown time obtained during the EPA confir-
matory test is most appropriate at this time.

It is' also concluded that a 5 percent tolerance would be most desireable
from the standpoint of rejecting inappropriate dynamometer requests.
Several improvements in the dynamometer and in the dynamometer calibra-
tion are currently planned by the Laboratory Branch.  These improvements
should increase the dynamometer system accuracy sufficiently that the
tighter, 5 percent tolerance value can be adopted.

VI.  Recommendation

It is recommended that the tolerance value of 7 percent be adopted for
the 1979 model year.  It is further recommended that a tolerance value
of 5 percent be tentatively proposed for the 1980 model year.

-------
APPENDICIES

-------
                              APPENDIX  A
                               Table A-l

                            1969 Plymouth
                             Dynamometer 1
                           IWC = 5000 pounds
                               12/12/77

Nominal        Indicated      Computed                                Mean
Actual        Dynamometer      Actual                              Coastdown
Power      Power Adjustment  Dynamometer    Coastdown Times (sec)     Time
 (Hp)      	(Hp)           (Hp)      Run #1   Run #2   Run #3    (sec)
15.5
14.5
13.5
12.5
11.5
10.5
9.5
12.6
11.6
10.6
9.6
8.7
7.7
6.7
15.52
14.50
13.47
12.44
11.52
10.49
9.46
13.16
13.77
14.46
15.16
15.93
16.85
17.76
13.17
13.78
14.46
15.15
15.90
16.80
17.85
13.12
13.77
14.45
15.20
15.91
16.85
17.89
13.15
13.77
14.46
15.17
15.91
16.83
17.83
                        Dynamometer Calibration
                          INHP = M * ACT + B
                              M = 0.9730
                              B = -2.5051

-------
       Table A-2

    1969 Plymouth
     Dynamometer 2
   IWC = 5000 pounds
       12/12/77
Nominal Indicated Computed Mean
Actual Dynamometer Actual Coastdown
Power Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) (Hp) Run #1 Run #2 Run #3 (sec)
15.5
14.5
13.5
12.5
11.5
10.5
9.5
12.3
11.4
10.4
9.4
8.5
7.5
6.6
15.56
14.61
13.57
12.51
11.57
10.52
9.57
13.29
13.95
14.63
15.43
16.24
17.15
17.96
13.36
13.97
14.60
15.50
16.14
16.97
17.77
13.40
13.96
14.70
15.40
16.26
17.03
17.78
13.35
13.96
14.64
15.44
16.21
17.05
17.84
Dynamometer Calibration
  INHP = M * ACT + B
      M = 0.9526
      B = -2.5242

-------
       Table A-3

    1969 Plymouth
     Dynamometer 3
   IWC = 5000 pounds
       12/12/77
Nominal Indicated Computed Mean
Actual Dynamometer Actual Coastdown
Power Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) (Hp) Run #1 Run #2 Run #3 (sec)
15.5
14.5
13.5
12.5
11.5
10.5
9.5
13.1
12.2
11.2
10.3
9.3
8.3
7.4
15.45
14.52
13.48
12.54
11.50
10.46
9.53
12.45
13.10
13.79
14.44
15.26
16.17
17.12
12.51
13.12
13.78
14.46
15.30
16.24
17.08
12.55
13.15
13.83
14.37
15.32
16.20
17.09
12.50
13.12
13.80
14.42
15.29
16.20
17.10
Dynamometer Calibration
  INHP = M * ACT + B
      M = 0.9617
      B = -1.7620

-------
       Table A-4

    1969 Plymouth
     Dynamometer 4
   IWC = 5000 pounds
       12/12/77
Nominal Indicated Computed Mean
Actual Dynamometer Actual Coastdown
Power Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) (Hp) Run #1 Run #2 Run #3 (sec)
15.5
14.5
13.5
12.5
11.5
10.5
9.5
12.6
11.6
10.6
9.7
8.7
7.7
6.8
15.53
14.50
13.46
12.53
11.50
10.46
9.53
12.13
13.38
14.15
14.93
15.84
16.56
17.30
12.40
13.36
14.03
14.86
15.82
16.50
17.37
12.56
13.40
14.22
15.01
15.78
16.70
17.39
12.36
13.38
14.13
14.94
15.81
16.59
17.35
Dynamometer Calibration
  INHP = M * ACT + B
      M = 0.9670
      B = -2.4182

-------
          APPENDIX B
           Table B-l

          Ford Cougar
         Dynamometer 1
       IWC = 5000 pounds
           12/13/77
Nominal
Ac tual
Power
(HP)
12.5
9.5
13.5
15.5
11.5
10.5
14.5
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run //I Run #2 Run #3 (sec)
9.6
6.7
10.6
12.6
8.7
7.7
11.6
12.44
9.46
13.47
15.52
11.52
10.49
14.50
15.06
17.78
14.59
13.21
15.94
16.62
14.04
15.07
17.82
14.51
13.21
15.95
16.64
13.91
15.11
17.81
14.55
13.25
15.92
16.74
13.96
15.08
17.80
14.55
13.22
15.94
16.67
13.97
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.973
          B = -2.5051
Vehicle Warm-Up Started at 2:49

-------
           Table B-2

          Ford Cougar
         Dynamometer 2
       IWC = 5000 pounds
           12/13/77
Nominal
Ac tual
Power
(HP)
12.5
9.5
13.5
15.5
11.5
10.5
14.5
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run //I Run #2 Run #3 (sec)
9,4
6.6
10.4
12.3
8.5
7.5
11.4
12.44
9.52
13.48
15.46
11.50
10.46
14.53
15.14
17.64
14.53
13.40
16.05
16.91
13.91
15.22
17.67
14.54
13.42
16.05
16.94
13.95
15.23
17.77
14.60
13.45
16.15
17.02
13.99
15.20
17.69
14.56
13.42
16.08
16.96
13.95
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.9568
          B - -2.5242
Vehicle Warm-Up Started at 12:55

-------
           Table B-3

          Ford Cougar
         Dynamometer 3
       IWC = 5000 pounds
           12/14/77
Nominal
Actual
Power
(HP)
12.5
9.5
13.5
15.5
11.5
10.5
14.5
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
10.3
7.4
11.2
13.1
9.3
8.3
12.2
12.54
9.53
13.48
15.45
11.50
10.46
14.52
14.59
17.11
14.07
12.90
15.58
16.49
13.55
14.64
17.22
14.03
12.94
15.62
16.55
13.59
14.63
17.19
14.08
12.95
15.66
16.53
13.54
14.62
17.17
14.06
12.93
15.62
16.52
13.56
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.9617
          B = -1.7620
Vehicle Warm-Up Started at 9:45

-------
           Table B-4

          Ford Cougar
         Dynamometer 4
       IWC = 5000 pounds
           12/13/77
Nominal
Actual
Power
(HP)
15.5
14.5
13.5
12.5
11.5
10.5
9.5
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
12.6
11.6
10.6
9.7
8.7
7.7
6.8
15.53
14.50
13.46
12.53
11.50
10.46
9.53
12.95
13.64
14.16
14.70
15.72
16.66
17.16
13.00
13.63
14.15
14.73
15.71
16.64
17.06
13.06
13.60
14.17
14.90
15.76
16.67
17.22
13.00
13.64
14.16
14.78
15.73
16.66
17.15
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.9670
          B = -2.4182
Vehicle Warm-Up Started at 8:50

-------
                      Table B-5

                     Ford Cougar
Dynamometer 4 - Repeat Test After Routine Calibration
                  IWC = 5000 pounds
                      12/14/77
Nominal
Actual
Power
(HP)
12.5
9.5
13.5
15.5
11.5
10.5
15.0
Indicated . Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run //2 Run #3 (sec)
9.7
6.8
10.6
12.6
8.7
7.7
12.0
12.53
9.53
13.46
15.53
11.50
10.46
14.91
15.03
17.85
14.38
13.09
15.90
16.76
13.59
15.18
17.66
14.40
13.18
15.90
16.88
13.65
15.22
17.72
14.43
13.23
15.93
16.94
13.75
15.14
17.74
14.40
13.17
15.91
16.86
13.66
               Dynamometer Calibration
                 INHP = M * ACT + B
                     M = 0.967
                     B = -2.4182

-------
          APPENDIX C
           Table C-l

     Ford LTD (TRC Rental)
         Dynamometer 1
       IWC = 5000 pounds
           12/21/77
Nominal
Actual
Power
(HP)
8.5
7.5
9.5
11.5
10.5
13.5
12.5
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(HpJ (Hp) Run #1 Run #2 Run #3 (sec)
5.7
4.7
6.7
8.7
7.7
10.7
9.7
8.43
7.41
9.46
11.52
10.49
13.57
12.54
20.01
21.12
18.64
16.60
17.61
15.06
15.94
20.02
21.35
18.66
16.63
17.72
15.16
15.97
19.85
21.23
18.61
16.60
17.67
15.13
15.98
19.96
21.23
18.63
16.61
17.67
15.12
15.96
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.973
          B - -2.5051
Vehicle Warm-Up Started at 2:42

-------
           Table C-2

     Ford LTD (TRC Rental)
         Dynamometer 2
       IWC = 5000 pounds
           12/21/77
Nominal
Actual
Power
(HP)
8.5
7.5
9.5
11.5
10.5
13.5
12.5
Indicated Computed . Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
5.6
4.6
6.6
8.6
7.6
10.6
9.6
8.48
7.43
9.52
11.60
10.56
13.69
12.64
20.18
21.43
18.83
16.94
17.76
15.31
16.05
20.23
21.52
18.54
16.61
17.80
15.36
16.26
20.25
21.53
18.77
16.68
17.84
15.34
16.22
20.22
21.49
18.71
16.74
17.80
15.34
16.18
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.9586
          B = -2.5242
Vehicle Warm-Up Started at 10:15

-------
           Table C-3

     Ford LTD (TRC Rental)
         Dynamometer 3
       IWC = 5000 pounds
           12/20/77
Nominal
Actual
Power
(HP)
8.5
7.5
9.5
12.5
10.5
13.5
11.5
Indicated
Computed



Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
6.4
5.4
7.4
10.4
8.4
11.4
9.4
8.49
7.45
9.53
12.65
10.57
13.69
11.60
19.61
20.49
17.95
15.54
16.60
14.22
16.24
19.68
20.55
18.03
15.52
16.77
14.46
16.23
19.41
20.69
18.16
15.56
16.86
14.56
16.28
19.57
20.58
18.05
15.54
16.74
14.41
16.25
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.9617
          B = -1.7620
Vehicle Warm-Up Started at 10:55

-------
           Table C-4

     Ford LTD (TRC Rental)
         Dynamometer 4
       IWC = 5000 pounds
           12/20/77
Nominal Indicated
Actual Dynamometer
Power Power Adjustment
(HP) (HP)
8.5
7.5
9.5
11.5
10.5
13.5
12.5
5.8
4.8
6.8
8.8
7.8
10.8
9.8
Computed Mean
Actual Coastdown
Dynamometer Coastdown Times (sec) Time
(Hp) Run #1 Run #2 Run #3 (sec)
8.49
7.46
9.53
11.60
10.57
13.67
12.63
20.06
21.19
18.47
16.90
18.09
15.16
16.08
20.00
21.32
18.64
16.84
17.77
15.14
16.08
20.50
21.39
18.82
16.91
17.73
15.06
16.01
20.19
21.30
18.64
16.88
17.86
15.12
16.06
    Dynamometer Calibration
      INHP = M * ACT + B
          M = 0.967
          B = -2.4182
Vehicle Warm-Up Started at 2:45

-------
           APPENDIX D
            Table D-l

Ford Fiesta (TRC Rental Vehicle)
          Dynamometer 1
        IWC = 2000 pounds
            12/21/77
Nominal
Ac tual
Power
(HP)
4.3
3.2
5.4
7.6
6.5
9.7
8.6
Indicated
Computed



Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run //3 (sec)
2.6
1.6
3.6
5.6
4.6
7.6
6.6
4.30
3.22
5.39
7.55
6.47
9.72
8.64
12.32
13.88
12.42
9.92
10.35
8.16
8.72
12.56
14.22
12.59
10.02
10.48
8.26
8.97
12.91
14.69
12.43
10.01
10.53
8.31
8.99
12.60
14.26
12.48
9.98
10.45
8.24
8.89
     Dynamometer Calibration
       INHP = M * ACT + B
           M = 0.9223
           B = -1.3673
 Vehicle Warm-Up Started at 1:15

-------
            Table D-2

Ford Fiesta (TRC Rental Vehicle)
          Dynamometer 2
        IWC = 2000 pounds
            12/21/77
Nominal
Actual
Power
(Hp)
4.3
3.2
3.4
7.6
8.6
6.5
9.7
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
2.4
1.4
3.4
5.4 '
6.4
4.4
7.4
4.32
3.24
5.40
7.57
8.65
6.48
9.73
11.25
12.70
11.64
9.43
8.42
9.98
8.25
11.68
13.05
11.72
9.58
8.65
10.19
8.49
11.88
13.16
11.73
9.69
8.60
10.42
8.44
11.60
12.97
11.70
9.57
8.56
10.20
8.39
     Dynamometer Calibration
       INHP = M * ACT + B
           M = 0.9246
           B = -1.5956
 Vehicle Warm-Up Started at 8:30

-------
            Table D-3

Ford Fiesta (TRC Rental Vehicle)
          Dynamometer 3
        IWC = 2000 pounds
            12/20/77
Nominal Indicated Computed Mean
Actual Dynamometer Actual Coastdown
Power Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) (Hp) Run #1 Run #2 Run #3 (sec)
4.3
3.2
5.4
7.6
9.7
6.5
8.6
3.2
2.2
4.2
6.2
8.2
5.2
7.2
4.33
3.25
5.41
7.58
9.74
6.50
8.66
11.62
13.14
11.18
8.96
8.25
10.61
9.09
12.04
13.68
11.25
9.38
8.52
10.72
9.12
12.18
13.98
11.40
9.32
8.51
10.74
8.96
11.95
13.60
11.28
9.22
8.43
10.69
9.06
     Dynamometer Calibration
       INHP <= M * ACT + B
           M = 0.9238
           B = -0.8007
 Vehicle Warm-Up Started at 1:48

-------
            Table D-4

Ford Fiesta (TRC Rental Vehicle)
          Dynamometer 4
        IWC = 2000 pounds
            12/20/77
Nominal
Ac tual
Power
(HP)
4.3
3.2
5.4
7.6
6.5
9.7
8.6
Indicated
Computed



Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run //I Run #2 Run #3 (sec)
2.4
1.4
3.4
5.4
4.4
7.4
6.4
4.26
3.18
5.34
7.50
6.42
9.66
8.58
13.20
14.72
11.80
10.17
11.18
8.57
9.38
13.39
14.95
11.89
10.16
10.98
8.73
9.37
13.41
15.01
11.91
10.18
10.91
8.83
9.38
13.33
14.89
11.87
10.17
11.02
8.71
9.39
     Dynamometer Calibration
       INHP = M * ACT + B
           M = 0.9255
           B = -1.5435
 Vehicle Warm-Up Started at 8:00

-------
APPENDIX E
Table E-l
Mercury Marquis (TRC Rental Vehicle)
Dynamometer 1
IWC = 5000 pounds
1/3/78
Nominal
Ac tual
Power
(HP)
8.8
10.9
13.0
7.8
9.8
11.9
6.7
Indicated Computed Mean
Dynampmeter Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
6.1
8.1
10.1
5.1
7.1
9.1
4.1
8.84 19.23 19.35 19.31 19.30
10.90 16.86 16.93 16.99 16.93
12.95 15.26 15.28 15.21 15.25
7.82 20.17 20.20 20.15 20.17
9.87 17.76 17.76 17.76 17.76
11.93 15.91 15.91 15.95 15.92
6.79 21.59 21.58 21.58 21.58
    Dynamometer Calibration
      INHP = M * ACT + B
          M - 0.9730
          B = -2.5051
Vehicle Warm-Up Started at 3:05

-------
              Table E-2

Mercury Marquis (TRC Rental Vehicle)
            Dynamometer 2
          IWC = 5000 pounds
               1/3/78
Nominal
Actual
Power
(Hp)
8.8
10.9
13.0
7.8
9.8
11.9
6.7
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
5.9
7.9
9.9
4.9
6.9
8.9
3.9
8.79
10.87
12.96
7.76
9.83
11.92
6.70
18.74
16.76
17.06
15.41
20.37
18.06
16.24
21.69
18.73
17.03
17.04
15.40
20.37
17.91
16.32
21.72
18.79
16.94
17.06
15.46
20.36
17.98
16.29
21.77
18.75
i
16.98
15.42
20.37
17.98
16.28
21.73
       Dynamometer Calibration
         INHP = M * ACT + B
             M = 0.9586
             B = -2.5242
   Vehicle Warm-Up Started at 10:45

-------
              Table E-3

Mercury Marquis (TRC Rental Vehicle)
            Dynamometer 3
          IWC = 5000 pounds
              12/30/77
Nominal
Ac tual
Power
(Hp)
8.8
10.9
13.0
7.8
• 9.8
11.9
6.7
Indicated Computed Mean
Dynamometer Actual Coastdown
Power Adjustment Dynamometer Coastdown Times (sec) Time
(Hp) (Hp) Run #1 Run #2 Run #3 (sec)
6.7
8.7
10.7
5.7
7.7
9.7
4.7
8.80
10.88
12.96
7.76
9.84
11.92
6.72
18.16
16.24
14.63
19.54
17.22
15.62
20.94
18.21
16.22
14.69
19.42
17.35
15.57
21.00
18.30
16.21
14.66
. 19.52
17.35
15.63
20.94
18.22
16.22
14.66
19.49
17.31
15.61
20.96
       Dynamometer Calibration
         INHP = M * ACT + B
             M = 0.9617
             B = -1.7620
   Vehicle Warm-Up Started at 1:47

-------