EPA Report No.:
EPA-AA-CD-84-01
Technical Report
Road-Load Coastdown Testing
of Selected 1981 thru 1984
Model Year Light-Duty Vehicles
and Light-Duty Trucks
April 1984
NOTICE
Technical Reports do not necessarily represent final EPA
decisions or positions. They are intended to present technical
analysis of issues using data which are currently available.
The purpose in the release of such reports is to facilitate the
exchange of technical information and to inform the public of
technical developments which may form the basis for a final EPA
decision, position, or regulatory action.
U.S. Environmental Protection Agency
Office of Air and Radiation
Office of Mobile Sources
Certification Division
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
-2-
I. Abstract
Twenty-four 1981 through 1984 model year light-duty
vehicles and light-duty trucks were tested to determine the
force required to overcome the sum of aerodynamic drag, tire
rolling resistance and other frictiorial losses. The primary
purpose of this testing was to compare the results from produc-
tion vehicles with information submitted by the manufacturers
for EPA's emission certification and fuel economy programs.
Reevaluation of EPA's entire alternate road-load procedure was
a secondary goal.
To the extent possible, test vehicles were obtained from
private owners and rental fleets. Those vehicles supplied by
the manufacturer had been used for general transportation and,
from the records available, had not received any maintenance
which would make them unrepresentative of production. Vehicles
were tested under the procedure set forth in EPA's Office of
Mobile Sources Advisory Circular No. 55B. At least 14 runs
were made with the vehicle coasting (transmission in neutral)
from 60 to 20 mph. Alternate runs were made in opposite direc-
tions; vehicle speed was recorded periodically. After mathemat-
ical processing, a final weather and mass corrected 55 to, 45
mph "coastdown time" was calculated. (Coastdown time is a func-
tion of force and vehicle mass. Because of its usefulness in
other testing, it is adopted as a convenient expression for
force.)
Generally, production vehicles did not perform as well
as specified in the manufacturer's application. While many of
the differences were rather small, a few indicated that the
manufacturer's application did not accurately describe the
production vehicle. No explanation for these discrepancies is
available. In addition, EPA's present procedure does not
specifically address the testing of vehicles with retractable
headlights.
II. Purpose
This test program had two main goals. First, and
foremost, was the assessment of how well production vehicles
compare to the information submitted by the manufacturers for
the emission certification and fuel economy programs. In order
to perform dynamometer testing for emissions and fuel economy,
it is necessary that the actual road-load be known so that the
dynamometer can be adjusted. Road-load is normally determined
by manufacturer testing of prototype vehicles prior to produc-
tion. Several years ago EPA discontinued the routine confir-
matory coastdown testing of manufacturer prototype vehicles.
-------
-3-
Instead, limited testing of production vehicles would be used
to verify the representativeness of the manufacturers submitted
information. This program was EPA's first such verification
effort. The second purpose was to give EPA current, firsthand
experience with the coastdown procedure. It has been several
years since the coastdown procedure was developed; changes in
vehicles and test equipment have occurred. A fresh examination,
with an eye toward corrections and improvements, should be
beneficial.
III. Road-Load Definition and Discussion
In this report the term "road-load" will be used to
denote the sum total of aerodynamic, tire rolling resistance
and drivetrain frictional losses. Road-load can be thought of
as a force that the vehicle must overcome to maintain a con-
stant velocity on a level road. The principle components'are
aerodynamic drag and tire rolling resistance; drivetrain fric-
tional losses are much smaller. Road-load increases with vehi-
cle speed. Although tire rolling resistance and drivetrain
friction do not change significantly with speed, aerodynamic
losses increase with the square of the speed. Therefore, at
low speeds tire rolling resistance and friction are the most
important while at high speeds aerodynamic drag predominates.
While somewhat vehicle dependent, the crossover point is about
40 mph.
Since road-load is a force, the conventional units would
be those of pounds-force (English) or Newtons (metric). Road-
load force is not constant, therefore, a reference speed must
be specified. However, instead of using force, road-load is
typically expressed in terms of a 55 to 45 mph "coastdown time".
The use of coastdown time as an expression for road-load devel-
oped as an outgrowth of EPA dynamometer testing. Dynamometer
coastdown time is measured as a quick-check of the dynamometer
adjustment. The dynamometer time must match the road time (with
certain tolerances).
IV. Background
EPA's Office of Mobile Sources tests passenger cars and
light trucks for exhaust emissions and fuel economy. These
tests are conducted indoors on a chassis dynamometer with the
vehicle held stationary. The chassis dynamometer allows the
vehicle to be "driven" over a prescribed driving schedule while
emissions and fuel economy are measured. The function of the
dynamometer is to reproduce the engine load conditions that the
vehicle would encounter on an actual street or highway.
-------
-4-
The dynamometer must reproduce the mass and road-load
that the vehicle encounters in real operation. The mass is
very easy to simulate; a series of different size flywheels can
be engaged or disengaged as necessary. The only information
needed in making the flywheel selection is the vehicle's
weight/ easily measured on a scale.
Unfortunately, determining road-load is not as simple.
Road-load is comprised of three major components: aerodynamic
drag, tire rolling resistance and drivetrain frictional losses.
These items are more difficult to measure (either singularly or
combined) than vehicle weight, primarily because they only exist
under actual operating conditions. Further, road-load is af-
fected by wind and weather conditions. During the early days
of emission and fuel economy testing, most manufacturers did
not measure road-load.
.EPA's original approach to dynamometer adjustment was to
establish a general table of adjustment values. These values
were a function of vehicle weight; a manufacturer only had to
know the weight of a vehicle in order to establish its dynamom-
eter adjustment. While not as sophisticated as the current
procedure, the general table was adequate at that time. While
not directly correlated with road-load, weight is an indirect
predictor for many vehicles. Heavier vehicles tend to be larger
with increased aerodynamic drag (tire and friction losses also
increase). Vehicle manufacturers could elect to determine road-
load for any vehicle in lieu of using these average adjustment
factors. Whether due to the complexity and difficulty of mea-
suring road-load or whether the average adjustment factors were
generally adequate for existing vehicles, manufacturers rarely
availed themselves of this opportunity.
Beginning with the oil embargo of 1973, the topic of
dynamometer adjustment became much more important. EPA began
publishing fuel economy results which were obtained as a by-
product of emission testing. While dynamometer adjustment does
affect emission results, fuel economy measurements are much more
sensitive. This is especially true for "highway" fuel economy.
The "city" (EPA Estimated MPG) number is derived from operation
over a stop and go driving schedule. During the city cycle a
large part of the engine output is used to overcome inertia;
this energy is ultimately dissipated in the vehicle's brakes as
heat. The highway driving sequence is much different as it is
run at a higher and much more constant speed, both factors
increasing the sensitivity to road-load.
In response to public demand for more fuel efficient
vehicles, manufacturers began making various improvements to
reduce road-load. They also sought credit for vehicles which
were already more efficient than indicated by the general
table. To obtain the appropriate fuel economy benefit for
-------
-5-
vehicle efficiency improvements, manufacturers began testing
their vehicles for road-load.
Initially, the technique used to determine vehicle
road-load was based on engine manifold vacuum. For gasoline
engines, the amount of power produced is inversely related to
the manifold-vacuum. Higher levels of power correspond to
lower levels of vacuum. (As the engine throttle is opened to
demand more powe"r, the amount of pressure drop across the
throttle plate decreases, therefore, decreasing the manifold
vacuum.) In order to employ this principle, the vehicle was
equipped with a vacuum gauge and then driven on a level road at
50 mph constant speed. The manifold vacuum reading was noted.
The vehicle was then placed on a chassis dynamometer and oper-
ated at 50 mph. The dynamometer was adjusted until the same
manifold vacuum reading was obtained. This technique was not
as accurate as desired and was also effected by barometric
pressure changes; with the advent of certain exhaust gas circu-
lation systems it was difficult to use. (These systems were
sensitive to engine manifold vacuum and, while in operation,
would change the vaccum reading. Engine vacuum was not the
reliable predictor of engine power that it had been.) Further,
engine vacuum could not be used to determine the power output
of diesel engines. (For diesels fuel flow can be measured, but
this procedure has its own limitations.) An improved method
was needed.
EPA developed a new general method to establish correct
dynamometer adjustment. This method relies on frontal area,
vehicle shape, tire type, weight and size of protuberances.
The equation relating these factors was derived empirically;
data was gathered using the coastdown test procedure described
below. Most manufacturers do not use the general method for
calculating road-load. It is to their advantage to establish
specific road-load values for each vehicle. The coastdown
technique is almost universally used.
V. Coastdown Test Procedure
The coastdown test has become the standard industry
technique for determining vehicle road-load for purposes of EPA
emissions and fuel economy testing. This procedure is set
forth in EPA's Office of Mobile Sources Advisory Circular No.
55B. This technique is also set forth in the Society of Auto-
motive Engineers (SAE) Recommended Practice J1263. The SAE
practice has statistical screens for removing variable data
(outliers). The reader is referred to those documents for a
more detailed discussion; following is a brief outline.
-------
-6-
The test vehicle is accelerated to a speed above 60 mph,
the transmission is placed in neutral, and the vehicle is
allowed to decelerate to a speed below 20 mph. This process is
then repeated with the vehicle traveling in the opposite
direction on the same test road. This road must be straight
and quite level. Vehicle speed is obtained from a calibrated
fifth wheel and is recorded periodically. Wind speed, wind
direction, ambient temperature and barometric pressure are also
recorded.
Results from each run are fit into a two term equation
in which "acceleration" is predicted as a function of a constant
and velocity squared term, A = -(ao + a2V2). Various sta-
tistical tests performed, temperature, barometric pressure and
wind corrections are made, and average coefficients are estab-
lished. From these, a dynamometer coastdown time from 55 to 45
mph is calculated, appropriate corrections being made for the
dynamometer inertia simulation actually used (as opposed to test
vehicle mass) and the nonrotation of the two wheel assemblies
not on the dynamometer. This 55 to 45 mph coastdown time is
referred to as a "target" time and is used as a "quick-check"
after vehicle dynamometer testing.
The reader is urged to consult Advisory Circular 55B and
SAE recommended practice J1263 for further information.
VI. Vehicle Selection
«-,
Twenty-four test vehicles from the larger foreign and
domestic manufacturers were selected for this program. Two
criteria were employed in making the selections. First, an
attempt was made to select the "best performing" vehicles, i.e.
low road-load relative to their competition. To a lesser
extent, a random cross section of vehicles was also selected.
However, because of the importance of vehicle road-load in
establishing fuel economy values, and because of the emphasis
most manufacturers place on obtaining high fuel economy values,
selections were biased toward smaller vehicles with high fuel
economy.
Attempts were made to procure vehicles with approxi-
mately 4,000 miles from rental fleets or private owners.
Occasionally, a particular vehicle would be difficult to
obtain. In such case, the vehicle specification might be
changed or the desired 4,000-mile odometer reading would be
relaxed. In several instances, the manufacturer supplied
vehicles from a "transportation fleet" or from a manufacturer
employee. (Because the purpose of this program was to asess
performance on production vehicles, an engineering or test
-------
-7-
vehicle might have received unrepresentative maintenance and
would not be acceptable.)
A complete list of test vehicles and their specifica-
tions can be found in the Appendix.
VII. Test Program
Most of the testing in this program was performed (under
EPA contract) by the Transportation Research Center (TRC) of
East Liberty, Ohio. One test was performed by a manufacturer
at its test facility. Several manufacturer correlation tests
.were also run. TRC's testing was done on a 1.9 mile straight
segment of their 7 1/2 mile oval track. This straightaway is
aligned approximately northwest and southeast with a 0.25 per-
cent downgrade towards the southeast. Testing was conducted
during the end of the summers of 1981, 1982, and 1983.
The following is a general outline of the procedure used
for each vehicle tested at TRC:
A. Initial Inspection. The vehicle would be received
and inspected to verify that it met the contract specifications.
Safety and fluid level checks were also performed.
B. Instrumentation. A fifth wheel bracket was
fabricated and installed. The necessary wiring from the fifth
wheel to the passenger's compartment was put in place. Depend-
ing upon test scheduling, the instrumentation package was either
installed at this point or immediately prior to vehicle weigh
in.
C. Wheel Alignment. Wheel alignment adjustments were
measured and, if necessary, the vehicle was adjusted to the
manufacturer's recommended specifications. When recommended by
the manufacturer, rear wheel alignment was also measured and
adjusted. At the conclusion of the wheel alignment, brake drag
was checked and adjusted, if necessary. The tires were infla-
ted to five psi above manufacturer's recommended pressure, and
the vehicle was placed outside.
D. Weight. After a minimum of four hours at ambient
temperature, the tire pressures were reduced to the manufac-
turer's recommendation and the vehicle was weighed. Distance
from the ground to the top of each fender opening was measured
as an indication of vehicle height.
E. Warm-up. The vehicle was operated at 50 mph for a
-------
-8-
distance of approximately 26 miles in preparation for testing.
Fifth wheel calibration was checked against a measured mile.
F. Coastdown Test. A total of fourteen 60 to 20 mph
coastdowns were made, seven in each direction. (Occasionally,
if the driver—noticed or suspected a problem, additional runs
were made.)
G. Weight. An after test weight with driver was
obtained.
H. Data Processing. Test data was processed by
computer in accordance with SAE Recommended Practice J1263
(without the cross-wind correction.)
Based on our experience during the 1981 testing season,
manufacturers' representatives were invited to view the 1982
and 1983 testing. Also, for the 1983 testing, an attempt was
made to perform the coastdown test with extremely low wind
velocity, well below the maximum allowed in Advisory Circular
No. 55B.
VIII. Results
Except for one vehicle, all test results yielded a
shorter coastdown time than stated in the manufacturer's
application for emission certification. Shorter coastdowns
indicate a greater road-load force. A distribution of
comparisons between vehicle tests and the manufacturer's
application is shown in Figure 1. A summary of the test
results is contained in the appendix.
-------
-9-
Figure I
Comparison- of Results to Manufacturers Applications
% Shortfall
(1.0)
0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
10.0
11.0
12.0
13.0
14.0
15.0
X
X
X
XXX
X X
x
X X X X X
XXX
X
X
X X
X
X X
X= One Vehicle
-------
-10-
In addition to the overall results, this test program yielded
some additional information. The 1981 Toyota Celica was inad-
vertently tested with a very small door pillar vent open. After
the first 10 runs, this error was noticed and the vent was
closed. A full 14 run coastdown test was then performed.
Results from the first 10 runs show a lower road-load, 16.00
vs. 14.20 seconds; such a large difference would not be ex-
pected. There is no present explanation, perhaps it can be
partially attributed to tire cool down during the first 10 runs.
During testing of Vehicle Number 10, 1983 Ford Ranger,
winds were within specification but higher than desirable, (5
to 8 mph with gusts from 7 to 12.) The wind direction was
basically in line with the track. While the results did not
pass the statistical screen in SAE Recommended Practice J1263,
they did indicate a higher road-load than claimed by the manu-
facturer. Two tires were taken off this vehicle and were
tested for rolling resistance coefficient on a 67 inch drum.
Results were very low with rolling resistance coefficients of
0.0087 and 0.0091. A second 1983 Ford Ranger, Vehicle Number
12, was tested and agreed quite well with the manufacturer's
submission. While testing Vehicle Number 22, a 1983 Ford
Escort, the initial test sequence was aborted because of rain
and gusty wind. The eight test runs (four pairs) with winds
from 4.5 to 6 mph yielded a coastdown time only slightly
shorter than the complete test with winds from 0.5 to 3.0 mph
(14.28 versus 14.48 seconds).
The 1984 GM Fiero was the only vehicle in this program
equipped with retractable headlights. The test procedure does
not specifically discuss how such headlights are to be treated;
the manufacturers application for certification was based on
testing with the headlights off. EPA's testing was conducted
at night with the headlights on. The manufacturer ran correla-
tion tests before and after EPA's testing with headlights both
on and off. (A total of four manufacturer tests were run.)
The manufacturer tests indicated a 0.7 to 0.9 second decrease
in coastdown time with the headlights operating. Had EPA
tested the vehicle with the headlights off, a coastdown time in
the vicinity of 16.4 would have been expected. This would be
10 percent less than the manufacturers application.
As an adjunct to this program, several manufacturer
correlation tests were run. Agreement between TRC and the
manufacturers tests was quite good with TRC indicating
approximately one percent shorter coastdown times.
-------
-11-
-Coastdown Time-
Vehicle _ Mfr; Before TRC Mfr; After
9 Cavalier • 16.39s 15.99s 16.12s
14 Alliance 12.82 12.94
24 Fiero 15.74 15.57 15.81
Vehicles 9 and 24 were tested at the manufacturers test track a
day or two before and after the TRC test. The manufacturer
tested vehicle 14 immediately following the TRC test using its
own driver and equipment.
IX. Acknowledgment
EPA would like to thank the manufacturers who participated
in this program by sending representatives to view testing/ pro-
viding test vehicles/ and conducting correlation testing. The
assistance provided was much appreciated.
American Motors
Chrysler
Ford
^General Motors
'Honda
Nissan
Toyota
-------
Appendix I
Test Vehicles
No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Year
1981
1981
1981
1982
1982
1981
1981
1981
1982
1983
1982
1983
1982
1983
1983
1983
1983
1983
1983
1983
1983
1983
1983
1984
Manufacturer
Chrysler
Chrysler
Ford
Ford
GM
GM
Honda
Toyota
GM
Ford
Nissan
Ford
GM
Renault
GM
Ford
Ford
Nissan
Honda
Chrysler
Ford
Ford
Toyota
GM
Model
Reliant
TC3
Granada
LN7
Cavalier
Citation
Accord
Celica GT
Cavalier
Ranger
Sentra
Ranger
Ciera
Alliance
Cavalier
Ranger
LTD
200SX
Accord
Charger
Ranger
Escort
Camry
Fiero
Body Style
4-door
2-door HBK
4-door
2-door HBK
4-door
2-door HBK
2-door HBK
2-door HBK
2-door HBK
PU Truck
2-door
PU Truck
4-door
4-door
4-door
PU Truck
4-door
2-door
4-door
2-door HBK
PU Truck
4-door HBK
4-door
2-door
Cdometer
9587
7817
4608
1056
397
2406
6033
14698
6419
2285
2979
6886
14233
5153
3942
8203
17127
6392
2818
5396
21000
13463
5453
3905
Height
2,510
2,510
2,940
2,290
2,520
2,695
2,200
2,690
2,630
2,705
1,950
2,620
2,780
, 2,190
2,560
3,060
3,030
2,660
2,400
2,260
2,270
2,490
2,560
Engine
2.2-L
2.2-L
3.3-L
1.6-L
1.8-L
2.8-L
1.8-L
2.4-L
1.8-L
2.3-L
1.5-L
2.3-L
2.5-L
1.4-L
2.0-L
2.2-L
3.3-L
2.2-L
1.8-L
1.6-L
2.2-L
1.6-L
2.0-L
2.5-L
Trans-
mission Tires
A3 Good Viva
A3
A
M4
M4
A
MB
M5
M4
A3
M4
M4
A3
A3
A3
M4
A3
M5
A4
M4
M4
A3
A4
M4
Fire HPR
UniR Steel
Mich TRX
Fire OCR
Fire OCR
Bridg RD108
Bridg RD116
UhiR RSeal
Fire DCR
Yoko GT
Good Viva II
Fire 721
Good Corsa
Good Viva II
Fire 721
Fire 721
Bridg RD113
Mich X
Good Viva II
Fire
Good Corsa
Bridg R0116
Fire WR12
P185/65R14
P195/60R14
P175/75R14
P165/70R365
j
P175/80R13
P185/80R13
165SR13
185/70SR14
P195/70R13
P185/75R14
155/SR13
P185/75R13
P185/80R13
P175/70R13
P175/80R13
P195/75R14
P185/75R14
185/70SR14
P185/70R13
P175/70R13
P185/75R14
P165/80R13
185/70SR13
P185/80R13
Pressure F/R
35/35
35/35
35/35
30/30
• 35/35
30/30
24/24
24/24
35/35
35/35
26/26
35/35 ^
NJ
35/35 1
30/30
35/35
35/35
30/35
26/26
26/26
35/35
35/35
35/35
30/30
35/35
-------
Appendix II
Coastdown Testing Summary
Vehicle
1981 Testing
3. 1981 Ford Granada
4. 1981 Ford LN7
5. 1982 GM Cavalier
6. 1981 GM Citation
7. 1981 Honda Accord
1982 Testing
9. 1982 GM Cavalier*2
11. 1982 Nissan Sentra
12. 1983 Ford Ran
13. 1982 GM Ciera
_1983 Testing
Model
g
Reliant 4-dr
TC3 *! 2-dr HBK
nada 4-dr
2-dr HBK
ier 4-dr
ion 2-dr HBK
cord 2-dr HBK
elica GT 2-dr HBK
ier*2 2-dr HBK
pretest)
post-test)
ger (#1) Truck
entra 2-dr
ger (#2) Truck
4-dr
g
ult Alliance 4-dr
test)
Inertia Weight
2875 IDS.
2750
3500
2625
2875
3000
2500
3000
2875
3000
2250
3000
3125
2500
Coastdown Time
Vehicle Application
14.035
13.65
14.74
14.57
14.67
14.57
12.57
14.20
15.99
16.39
16.12
11.64*3
12.90
13.24
15.49
12.82
12.92
14.73
15.26
15.08
15.69
16.91
15.63
12.99
16.49
16.91
13.45
13.21
13.45
16.12
12.94
% Difference
Appl icat ion-Vehicle
Application
4.8%
ilO.6
2.3
7.1
13.2
6.8
3.2
13.9
5.4
13.5
2.4
1.6
3.9
0.9
k
15. 1983 GM Cavalier
4-dr
2875
15.23
17.86
14.7
-------
Page 2
Vehicle
16. 1983 Ford Ranger (#3)
Retest
17. 1983 Ford LTD
18. 1983 Nissan 200 SX
19. 1983 Honda Accord
20. 1983 Chrysler Charger
21. 1983 Ford Ranger *2
(Manufacturer test)
22. 1983 Ford Escort
23. 1983 Toyota Camry*2
24. 1984 Q4 Fiero*2
EPA Test
Manufacturer pretest*?
Manufacturer pretest*?
Manufacturer post-test
Manufacturer post-test
Model
Inertia Weight
Coastdown Time
Vehicle Application
% Difference
Application-Vehicle
Application
Truck
4-dr
2-dr
4-dr
2-dr HBK
Truck
4-dr HBK
4-dr
2-dr
3250
3500
3000
2625
2750
3250
2750
2875
13.46*4
13.28*5
16.75
16.28
13.14
16.38
14.48
14.24
15.57*6
15.74*6
16.61
15.81*6
16.53
15.60
17.70
16.31
14.08
16.91
15.16
15.28
14.10
18.23
11.2
12.4
5.4
0.2
J
6.7
3.1
4.8
5.2
(1.0)
i
>-
14.6 '
*1 New tires prior to test, not equipped with rear spoiler.
*2 Manufacturer supplied vehicle.
*3 Results are calculated from nine run pairs with RMS errors and standard deviations greater than permitted by
SAE Recommended Practice J1263. ,
*4 Aftermarket step bumper and modified air dam.
*5 Production step bumper and air dam.
*6 Retractable headlights extended.
*7 Rear alignment possibly slightly misadjusted.
------- |