-------
FOREWORD
This report is submitted in fulfillment of EPA Contract No. 68-03-2534.
The report has been reviewed by the Emission Control Technology Division,
U. S. Environmental Protection Agency, and approved for publication. Approval
does not signify that the contents necessarily reflect the views and policies
of the U. S. Environmental Protection Agency, nor does the mention of trade
names or commercial products constitute endorsement or recommendation for
use.
11
-------
ABSTRACT
The purpose of this research program was to achieve a quantitative compar-
ison between a conventional double-roll chassis dynamometer and an experimental
prototype of a flat-bed dynamometer as regards tire power losses, fuel economy
and exhaust emissions. Design and fabrication of the special dynamometer,
partially funded under this contract, were based on the use of a power absorp-
tion unit and an inertia unit from a conventional Clayton dynamometer.
Two 1977-model vehicles, one in the 2250-lb. inertia class (a ChevetteJ
and one in the 4500-lb. class (an Oldsmobile 98) were used for test purposes.
Each vehicle was supplied with complete sets of bias-ply, bias-belted and
radial-ply tires. Rear-wheel rims on each vehicle incorporated torque trans-
ducers.
Test operations included: (1) the measurement of wheel torque and wheel
horsepower during steady-state operation of the vehicles on a test track, (2)
the measurement of the rolling resistance of the individual tires on the
Calspan Tire Research Facility and (3) the measurement of emissions, fuel
economy, tire power loss and wheel torque for dynamometer tests using the
Federal Test Procedure, the Highway Fuel Economy Test and steady-state velocity
tests. Dynamometer tests were made only on the 4500-lb. inertia class vehicle.
In terms of fuel economy, the flat-bed dynamometer correctly ranks the
different tire constructions showing the radial-ply tires are superior to the
bias-type tires. The roll dynamometer, however, reverses the order and yields
data that show the bias-type tire affords better fuel economy than the radial-
ply tire. For radial-ply tires, the differences in measured fuel economies on
the two dynamometers were relatively minor for all tests. Some differences in
exhaust emissions were noted between results from identical tests performed
on the two dynamometers. The significance of these differences is questionable
because of the nominally large variability in these measurements, a paucity of
replicate data and the presence of certain small differences in ambient condi-
tions at the dynamometer test sites.
This report was submitted in fulfillment of Contract No. 68-03-2534 by the
Advanced Technology Center, Calspan Corporation under sponsorship of the U. S
Environmental Protection Agency. This report encompasses the time period March
11, 1977 to April 24, 1979, and work was completed as of June 29, 1979.
111
-------
METRIC CONVERSION FACTORS
Approximate Conversions to Metric Measures
Symbol
Wrun You Know
Multiply by
LENGTH
To Find
Symbol
in
ll
yd
mi
in2
h2
yd2
mi2
oz
Ib
isp
Tbsp
II oz
c
pt
1'
gal
II3
yd3
mches
leet
yards
miles
square inches
square leet
square yards
square miles
acres
ounces
pounds
short tons
12000 Ib)
teaspoons
tablespoons
fluid ounces
cups
pints
quarts
gallons
cubic feet
cubic yards
•2.5
30
0.9
1.6
AREA
6.6
0.09
0.8
2.6
0.4
MASS (weight)
28
0.45
0.9
VOLUME
5
15
30
0.24
0.47
0.95
3.6
0.03
0.76
centimeters
centimeters
meters
kilometers
square centimeters
square meters
square meters
square kilometers
hectares
grams
kilograms
tonnes
mitlihters
milliliters
milliliters
liters
liters
liters
liters
cubic meters
cubic meters
cm
cm
m
km
cm2
m2
m2
km2
ha
g
kg
i
ml
ml
ml
1
1
1
I
O!3
m3
TEMPERATURE (exact)
°F
- '' S4 1
Umtb uf Wgiulv
Fahrenheit
temperature
5.'9 (after
subtracting
32}
Celsius
temperature
bleb id* NBSM.bL, Pul.l.
°c
I'Stj.
Approximate Conversions from Metric Measures
Symbol Wh.n You Know Multiply by To fiarf
LENGTH
Synbol
- =i
mm
cm
m
m
km
cm2
m2
km2
ha
millimeters
centimeters
meters
meters
kilometers
square centimeters
square meters
square kilometers
hectares (10,000m2)
0.04 inches
0.4 inches
3.3 feet
1 .1 yards
0.6 miles
AREA
0.16 square inches
1.2 square yards
0.4 square miles
2.6 acres
in
in
It
yd
mi
in2
Yd2
rru2
MASS (weight)
g
kg
t
grams
kilograms
tonnes (1000 kg)
0.035 ounces
2.2 pounds
1.1 short tons
01
Ib
VOLUME
ml
1
1
1
m3
m3
milliliters
liters
liters
liters
cubic meters
cubic meters
0.03 tluid ounces
2.1 pints
1.06 quarts
0.26 gallons
35 cubic feet
1.3 cubic yards
II oz
pt
<)i
gal
It3
yd3
TEMPERATURE (exact)
°C
Celsius
temperature
°F 32
-40 0 40
1 . i . i i • i i 1 .
I r \^ " \ r T
-40 -20 0 2
9/5 (then Fahrenheit
add 32) temperature
°F
96.6 212
BO 120 160 200 I
, . i . . , i • . . . • J
1 1 1 1 ,1 1 i 1
3 40 60 BO 100
37 °C
°F
-------
CONTENTS
Page
Foreword 1X
Abstract
Figures
Tables ix
Acknowledgement xi
1.0 Introduction 1
2.0 Conclusions 4
3.0 Recommendations 6
4.0 Discussion 7
4.1 The Flat-Bed Dynamometer 7
4.2 Test Plan Details 9
4.3 Vehicle Road Tests 11
4.3.1 Test Details 11
4.3.2 Results 12
4.4 Tire Rolling Resistance Measurement 16
4.4.1 Methodology 21
4.4.2 Results 21
4.5 Dynamometer Check Tests 22
4.6 Vehicle Dynamometer Tests 25
4.6.1 Dynamometer 28
4.6.2 Test Procedures 30
4.6.3 Summary of Results 34
4.6.3.1 FTP Tests 34
4.6.3.2 HFET Tests 37
4.6.3.3 Steady-State Tests 39
4.6.3.4 Statistical Analysis53
-------
CONTENTS (Continued)
4.6.4
Comments on the Results ...... 59
References [[[ 61
Appendices
A. A Physical Description of the Flat-bed Dynamometer. . . 62
B. The Test Schedule for the Dynamometer Tests .......... 69
C. Procedures for Performing the Dynamometer Tests ...... 72
D. A Description of the Special Instrumentation
Used for the Vehicle Tests ........................... 74
-------
FIGURES
Number
Page
O
1 Flat-bed dynamometer plan-view sketch .........................
2 Least-squares curves fitted to wheel-torque
data for the 1977 Chevette equipped with bias,
bias-belted and radial tires .................................. ^
3 Least-squares curves fitted to wheel-torque
data for the 1977 Oldsmobile 98 equipped with bias,
bias-belted and radial tires .................................. 18
4 Least-squares curves fitted to wheel -horsepower data
for the 1977 Chevette equipped with bias, bias-
belted and radial tires ....................................... 19
5 Least-squares curves fitted to wheel -horsepower data for
the 1977 Oldsmobile 98 equipped with bias, bias-
belted and radial tires ....................................... 20
6 Flat-bed dynamometer installation at the GM
Proving Ground at Milford ..................................... 26
7 Friction horsepower versus velocity for the flat-bed and
Clayton dynamometers .......................................... 31
8 True absorbed horsepower for the Clayton dynamometer
as a function of velocity ..................................... 32
9 Wheel horsepower versus velocity, Oldsmobile 98 with radial-
ply tires (GR78-15) on the flat-bed dynamometer ............... 33
10 Hydrocarbon emissions as a function of velocity, steady-
state tests on bias-ply tires ................................. 42
11 Hydrocarbon emissions as a function of velocity, steady-
state tests on radial -ply tires ............................... 43
12 Oxides of nitrogen emissions as a function of velocity,
steady-state tests on radial-ply tires ........................ 44
vn
-------
FIGURES (Continued)
Number Page
13 Oxides of nitrogen emissions as a function of velocity,
steady-state tests on bias-ply tires ....................... 45
14 Steady-state tests, fuel economy versus tire construction .. 46
15 Fuel economy as a function of velocity, steady-state
tests on radial-ply tires .................................. 48
16 Fuel economy as a function of velocity, steady-state
tests on bias-ply tires .................................... 49
17 Wheel horsepower versus velocity for radial-ply tires
(GR78-15) , steady-state tests ............................... 50
18 Wheel horsepower versus velocity for bias-ply tires
(G78-15), steady-state tests ............................... 51
19 Wheel horsepower versus velocity for bias-belted
tires (G78-15), steady-state tests ......................... 52
20 Tire horsepower consumption versus velocity, Clayton
Dynamometer ................................................ 54
A-l View of the right half of the flat-bed dynamometer ......... 66
A-2 Flat-bed dynamometer plan-view arrangement ................. 67
A-3 Close-up view of the gearbelt drive arrangement ............ 68
D-l Torque wheel and instrumentation; 1977 Oldsmobile .......... 86
D-2 Distance-measuring equipment; one-foot circumference
wheel mounted on the flat-bed dynamometer .................. 87
E-l Calspan tire research facility (TIRF) ...................... 91
E-2 Tire research machine ...................................... 92
Vlll
-------
TABLES
Number
1 Chronology of Maj or Pro j ect Events
Page
2 Steady-State Measurements of Wheel Torque and
Wheel Horsepower at Various Speeds on a 1977
Oldsmobile 98 Equipped with Bias, Bias-Belted
and Radial-Ply Tires ........................................ 13
3 Steady-State Measurements of Wheel Torque and Wheel
Horsepower at Various Speeds on a 1977 Chevette
Equipped with Bias, Bias-Belted and Radial-Ply Tires ........ 14
4 Summary of Least-Squares Curve Fits to Wheel Torque
and Wheel Horsepower Data for the Chevette and
the Oldsmobile 98 ........................................... 15
5 Summary of Equilibrium Values of the Rolling
Resistance Force for the Test Tires ......................... 23
6 Exhaust Emissions, Fuel Economy and Wheel Torque
Data for FTP Tests .......................................... 35
7 Exhaust Emissions, Fuel Economy and Wheel Torque
Data for the HFET Tests ..................................... 38
8 Exhaust Emissions, Fuel Economy, Wheel Torque
and Tire Power Consumption Data for Steady-State Tests ...... 40
9 Block Number Designations for Statistical Tests ............. 55
10 Summary of the Results of Tests of Significance
Applied to the Differences Between Means of Data
for the FTP and HFET Runs ................................... 56
11 Summary of the Results of Tests of Significance
Applied to the Differences Between Means of Data
for the Steady-State Runs ................................... 58
IX
-------
TABLES (Continued)
Number Page
B-l Dynamometer Test Schedule 71
E-l TIRF Capabilities 93
E-2 Balance System Capability 93
-------
ACKNOWLEDGEMENT
Significant conceptual and design contributions to the flat-bed dyna-
mometer were made by G. R. Duryea, I. Gusakov and S. D. Smith. Dynamometer
checkout tests as well as the maintenance and repair of the unit throughout
the entire program were performed by A. J. LaPres.
General Motors Corporation supplied the test vehicles, the personnel and
the test facilities at its Milford, Michigan Proving Ground free of cost. All
vehicular testing and data acquisition was performed by GM personnel who
supplied reduced data to Calspan. Operations at the Milford Emissions Labor-
atory were directed by D. D. Horchler, J. A. Tysver and C. E. VanAcker.
XI
-------
SECTION 1.0
INTRODUCTION
The validity of laboratory test procedures for the measurement of vehicle
fuel economy and exhaust emissions depends on the accuracy with which the energy
transfer and power losses that occur on the road are reproduced. Current
practice involves the operation of the vehicle on a roll-type of chassis dyna-
mometer. Using inertia fly-wheels coupled to the dynamometer rolls, the trans-
fer of engine power to kinetic energy equivalent to that of the moving vehicle
can be accurately simulated. Energy lost within the vehicle system can also
be readily simulated.
Power losses, such as those caused by aerodynamic drag, are simulated by
the use of various types of power absorption devices that also are coupled to
the dynamometer rolls. A device commonly used for this application is a
water-brake type of power absorber which characteristically approximates a cubic
variation of power with velocity as does the aerodynamic drag loss of automotive
vehicles. Federal Test Procedures require that the dynamometer power dissi-
pation is set to equal the vehicular "road load" at 50 mph*. Except for the
presence of power losses associated with the rolling resistance of tires, losses
which vary approximately linearly with velocity, the dynamometer simulation
of vehicle "road load" at other velocities would be reasonably accurate. A
further complication results from the fact that tires operating on rolls under-
go large deflections of the carcass resulting in excessively high tire tempera-
tures. Federal Test Procedures permit the use of cold tire inflation
pressures to 45 psi to preclude tire damage. Thus the tire power consumption
that is experienced on a roll-type dynamometer is expected to differ from that
experienced on a flat roadway.
The experimental study reported here was undertaken principally to eval-
uate the differences in fuel economy and exhaust emissions as measured for
identical tests performed on a dual-roll chassis dynamometer and special flat-
bed type of chassis dynamometer that duplicated a straight and level roadway
surface. Other objectives included the measurement of vehicular road load
power and tire horsepower consumption and the characterization of the dyna-
mometer frictional and true horsepower loading. Bias-ply, bias-belted and
radial-ply tire constructions were used in tests performed on vehicles re-
presentative of the 2250-lb.**and 4500-lb. inertia classes.
* The use of the English system of units was approved by the Project Officer.
A conversion table is included in the report.
** For road tests this vehicle was ballasted to a weight of 2500 Ib.
-------
The flat-bed dynamometer was assembled as an experimental prototype making
use of the basic loading components of a Clayton-type chassis dynamometer.
Two Calspan-developed simulated roadway units (SRU) were the basic building
blocks of this dynamometer to which the Clayton water brake and inertia units
were connected.
A cost-sharing contractual arrangement was entered into by the EPA, General
Motors and Calspan, with General Motors in a subcontractor role to Calspan. By
this agreement General Motors provided, without cost, the test vehicles, tires,
facilities and personnel required to perform the specified road tests and dyna-
mometer evaluation tests at its Milford, Michigan Proving Ground (GMPG).
Calspan provided, without cost, the engineering, design and development effort
required to adapt the SRU to the basic Clayton dynamometer components.
This report describes the accomplishments of the program, the tests that
were performed, the results that were obtained and the significance of these
results. The tests that were performed and the data analyses that were made
were in accordance with the explicit requirements of the contractual Statement
of Work.
To provide an overview of the sequence of events as they occurred during
the course of the project, which continued over a span of more than two years,
a chronology of the significant milestones is itemized in Table 1.
-------
Table 1
CHRONOLOGY OF MAJOR PROJECT EVENTS
EVENT
1. Program initiated
2. Test vehicles and tires procured (GMPG)
3. Design of the flat-bed dynamometer completed
4. Vehicle road tests completed (GMPG)
5. Delivery of Simulated Roadway Unit (SRU) accepted
6. Fabrication of flat-bed dynamometer and water
system completed
7. Tire rolling resistance tests completed on TIRF
8. Checkout tests of the flat-dynamometer at
Calspan completed
9. Delivery of the flat-bed dynamometer accepted
by GMPG
10. Installation and checkout of the flat-bed dynamometer
completed (GMPG)
11. Dynamometer testing begins (GMPG)
12. All test activity terminated
13. Draft of final report submitted
DATE
March 1977
June 1977
September 1977
October 1977
November 1977
February 1978
February 1978
March 1978
August 1978
November 1978
January 1979
March 1979
June 1979
-------
SECTION 2.0
CONCLUSIONS
The following itemized conclusions are drawn from the test results ob-
tained during the performance of the test program and the analyses to which
these results were subjected. There is no significance to be attached to the
ordered sequence of these conclusions.
• Friction horsepower losses for the flat-bed dynamometer that was used in
this study were significantly larger than those for the Clayton roll
dynamometer. At a speed of 50 mph, the friction horsepower for the
flat-bed unit was 11.1 as compared with 2.8 for the roll unit.
• The absorbed-horsepower/velocity characteristic curve of the flat-bed
dynamometer matches the road-horsepower/velocity character-
istic curve of the 1977 Oldsmobile 98 test vehicle much more closely
than the Clayton dynamometer.
• The flat-bed dynamometer correctly rank orders tires in terms of fuel
economy showing the radial-ply construction superior to that of the
bias-ply.
• The roll dynamometer incorrectly rank orders tires in terms of fuel
economy showing the bias-ply construction superior to that of the
radial-ply.
• In comparing data means, the most significant differences, in a statis-
tical sense, occurred between tests on (1) bias-ply tires on the flat-
bed and the roll dynamometers and (2) radial-ply and bias-ply tires on
the flat-bed dynamometer. The least significant differences occurred
between tests on radial-ply and bias-ply tires on the roll dynamometer.
The foregoing remarks apply to exhaust emissions (FTP tests)*, fuel
economy and positive wheel torque (FTP, HFET and SS tests)*.
• Exhaust emissions data (FTP tests) generally are not conclusive in a
statistical sense although the CO and NO levels are consistently
higher on the flat-bed dynamometer than on the roll dynamometer for
comparable test conditions.
*FTP - Federal Test Procedure, HFET - Highway Fuel Economy Test, SS -
Steady Speed
4
-------
• Differences in tire power consumption (SS tests) among radial-ply,
bias-belted and bias-ply constructions operating on the roll dynamometer
generally are not statistically significant.
• The flat-bed dynamometer requires further development to improve (1)
belt life, (2) water-bearing seal durability and (3) belt tracking.
Friction horsepower losses need to be reduced also.
-------
SECTION 3.0
RECOMMENDATIONS
Based on the conclusions resulting from this experimental^study, the
desirability of conducting additional programs to further quantify the differ-
ences between fuel economy and exhaust emission data as measured on conven-
tional roll-type chassis dynamometers and flat-bed-type dynamometers would
seem to be justified. It is recommended that follow-on effort should include:
(1) The design and fabrication of a production-type flat-bed dynamometer
that would incorporate such features as:
• a programmable, electrical-type power absorption unit that would
overcome the relatively high frictional power loss and simulate,
to a high level of accuracy, a wide spectrum of road load power
versus velocity functions
• a direct means for determining friction horsepower
• roadway belts with improved durability
• improved water-bearing seals
• a convenient and rapid way to vary the track width
• a compact and efficient heat exchanger to control water tempera-
ture
(2) A test schedule with a significantly increased level of replication.
The wide variability in HC and CO emissions measurements necessitates
more than the three replicate tests that were conducted in this pro-
gram if meaningful statistical analyses are to be performed. The
confidence with which the population mean can be defined for a set
of data is improved by approximately a factor of two when replication
is increased from three tests to ten.
(3) Tests on other sizes and brands of vehicles as well as on low-
inertia class vehicles to ascertain how different size tires affect
the differences in fuel economy, emissions and tire power loss be-
tween the flat-bed and roll-type dynamometers.
-------
SECTION 4.0
DISCUSSION
4.1 THE FLAT-BED DYNAMOMETER
The flat-bed dynamometer consists of two distinct systems, the dynamometer
itself and its associated water supply system. A description of the salient
features of the entire assembly will be presented in this section of the report.
Additional detailed information is included in Appendix A.
The description of the dynamometer that follows will be rendered more
meaningful by reference to the simple sketch of Figure 1 which shows a plan
view of the unit. Basically, the system can be considered as a chassis-type
dynamometer similar to one manufactured by Clayton whose rolls have been re-
placed by simulated roadway units (SRU) (1)* which provide flat and level
surfaces for the vehicle's driving wheels to operate on. Each SRU consists
of two parallel, vertical frame members that support two crowned drums on
which the endless belt operates. One drum, the lower one shown in the sketch,
operates in bearings that are fixed in the frame of the SRU. The other drum
is so mounted that belt-tension and belt-tracking adjustments can be made by
a repositioning of the drum. Each drum is 19.25 inches in diameter and 15
inches wide while the belt is 15.75 inches wide and 0.024 inches thick.
A hydrostatic water bearing is located midway between the axes of rotation
of the drums and provides the physical support for the belt and its load.
Spring-loaded, wiper-type seals are used to control water leakage at the bear-
ing which operates at a supply pressure of approximately 200 psi. The water
bearing is designed to accommodate tire contact pressures up to 100 psi and
has an active bearing size that is about 10 inches in width and 12 inches in
length. Except for differences in shaft lengths and shaft termination details,
the two SRU units are identical in all other respects.
The nonadjustable drums are coupled directly through a pair of flex-type
rubber couplings joined by a. floating shaft. With this arrangement, the dyna-
mometer accommodates the rear-track of the larger test vehicle. By removing
the floating shaft and translating the left-hand SRU**and the power absorption
*Numbers in parentheses designate references listed at the end of the report.
** As shown in the sketch.
-------
SIMULATED ROADWAY
UNIT(SRU)
WATER
BEARING
WATER BEARING
GEAR BELT
CLAYTON
INERTIA
UNIT
(DIRECT
COUPLED)
CLAYTON
PAU
FLOATING
SHAFT
(NO SCALE)
Figure 1
Flat-bed dynamometer plan-view sketch.
-------
unit (PAU), the spacing between the SRU's will satisfy the rear-wheel track
requirements of the smaller test vehicle. To facilitate this adjustment, both
the SRU and the PAU are mounted on a common base plate which can translate as
well as rotate. This latter feature permits setting a small slip angle at the
vehicle's right rear wheel, if necessary, to stabilize the vehicle's lateral
position at the belts*.
A direct-coupled, five-wheel, Clayton inertia unit was coupled to the
SRU's through a step-up speed ratio using a gearbelt-pulley combination.
Because of the large difference in the diameters of the SRU drums and the
Clayton rolls (8.65 inches, diameter), the rotational speed of the SRU drums
was much slower at any given vehicle speed than the Clayton rolls. As a con-
sequence, the effectiveness of the flywheel inertia in simulating vehicle
weight at the roadway surface had to be augmented by a step-up ratio (1:1.875).
All dynamometer components were mounted on a steel-beam support base
allowing the system to be transported as a single unit with the component
parts in an aligned condition. Dynamometer design envelope was dictated by
the requirement that the unit fit a standard (3 ft x 8 ft x 15 ft) pit at the
GM Proving Ground.
The water supply system consists of a high-pressure, high-volume pump
(ps200 gpm at 200 psi), a sump, a low-pressure, high-volume pump (300 gpm), a
500-gallon reservoir (polyethylene), particulate filters and sundry valves,
switches, motor controls, etc. With the exception of the reservoir, all of
the water system components were mounted on the dynamometer support base.
Following the assembly of the complete dynamometer system, the belts on
the SRU's were covered with a medium-grit, adhesive-backed abrasive material
sold under the tradename of "Safety-Walk"**. This same material is used on the
Calspan TIRF machine and satisfactorily simulates the texture and skid resis-
tance of typical highway surfaces when a "stoning" operation is performed on
the as-delivered material.
4.2 TEST PLAN DETAILS
The overall test plan required that two vehicles be obtained for test
operations; one in the 5000-lb. inertia class and one in the 2500-lb inertia
class. Three sets of tires were to be provided for each vehicle representing
bias-ply, bias-belted and radial-ply constructions. Vehicle rear wheels were
* It was not necessary to use this feature during this test program
** 3-M Company
-------
to be instrumented to permit the measurement of rear-wheel torque during all
testing operations.
Testing operations included the measurement of rear-wheel torque during a
series of steady velocity runs on an outdoor test track with the vehicles
equipped with each of the three sets of tires. From these data the wheel
horsepower variation with velocity could be established. Subsequent to these
road tests, laboratory measurements were to be made to determine the rolling
resistance of each of the tires on the Calspan flat-bed tester (TIRF) under
the same normal load conditions as each tire experienced on the test vehicle.
Data from these tests were to be used in establishing road load power settings
for the dynamometer tests. Dynamometer tests were to be conducted on both the
flat-bed and roll-type units for the purpose of measuring fuel economy, exhaust
emissions, wheel horsepower, tire power loss and other dependent variables.
Vehicles were to be operated in accordance with the Federal Test Procedure
(FTP), the Highway Fuel Exonomy Test (HFET) and a series of steady velocity
tests (SS). The foregoing comments represent a statement of intentions. The
succeeding comments present the events that actually transpired.
Vehicle down-sizing for the 1977 model year did not permit fulfillment of
a 5000-lb. inertia class requirement with a high-volume production vehicle.
As a compromise, General Motors selected an Oldsmobile 98 as a 4500-lb. inertia
class vehicle and the Chevrolet Chevette as a 2250-lb inertia class vehicle.
These choices still preserved the desired factor-of-two ratio between vehicle
inertias. To minimize variation in tire dimensions with construction differ-
ences, all tires in a given size were obtained from one manufacturer. Bias-
belted tires in the 13-inch size used by the Chevette are not normally fabri-
cated, however, the General Tire and Rubber Company did supply GM with a set
of these tires.
Road tests were performed on each vehicle following a test schedule that
exceeded in scope the requirements of the contract as tabulated below:
ROAD TEST SCHEDULE
Run No. Tire Construction Tire Pressure (cold) Speed, mph
1-10 Radial MRP* 20,30,40,50,60
60,50,40,30,20
11-13 Bias-belted MRP* 50,50,50
14^-16 Bias MRP* 50,50,50
*Tire manufacturer's recommended pressure (24 psi)
10
-------
The full set of runs specified on the radial-ply tires was completed for the
bias-belted and bias-ply tires as well. An in-depth discussion of the road
tests and the resultant data is included in Section 4.3.
Following the completion of the road tests, Calspan performed laboratory
tests to measure the equilibrium rolling resistance force of the tires at a
constant velocity of 50 mph (rolling resistance is essentially independent
of speed up to about 65 mph). Since data on each of the tires were not re-
quired in performing the dynamometer tests, only 20 of the 24 tires were
tested in the interests of minimizing costs. Section 4.4 elaborates on the
details of these tests and presents a summary of the measured data. Since
only the front tires are free-rolling (°n a rear-wheel drive vehicle), the
applicability of free-rolling data to rear tires may be questioned. Test data
show that for even sizable levels of steady wheel torque, tire rolling
resistance remains substantially constant. (2) In fact there is evidence
that minimum rolling resistance can occur at a positive (driving) torque
and not in the free-rolling condition.
The bulk of the testing was performed on the dynamometers, faithfully
following a test plan which was outlined in the contractual statement of work.
FTP, HFET and steady-speed tests were performed, with replication, on both
radial-ply and bias-ply tires. Bias-belted tires were only tested at steady-
speed conditions. Replication was limited to only two repeat tests as a maxi-
mum. In some instances, only one repeat test was made. For all tests, wheel
torques and exhaust emissions were measured. The effects of changes in infla-
tion pressure and dynamometer load settings on measured data were investigated
for the case of the flat-bed dynamometer only. Specific details of the dyna-
mometer test schedule are given in Appendix B. Test results are presented in
Section 4.5 which also includes the results of dynamometer calibration tests.
4.3 VEHICLE ROAD TESTS
4.3.1 Test Details
The purpose of the road tests on the Chevette and the Oldsmobile 98
vehicles was to obtain rear-wheel torque measurements at various steady-state
velocities on a flat and level roadway. Tests were performed on each vehicle
which was equipped alternately with sets of radial-ply, bias-belted and bias-
ply tires. Both test vehicles were equipped with automatic transmissions and
used an oxidizing catalytic converter as part of its emission control system.
Each vehicle was fitted with torque rims on the rear wheels to permit the
measurement of wheel torque. These rims were specially built to maintain the
same vehicular wheel track as existed with standard rims. The reason for this
requirement was that the flat-bed dynamometer had been designed to accommodate,
within a small tolerance, the standard rear track of the Chevette and the
Oldsmobile 98.
11
-------
Electronic equipment was carried on board the vehicles to log and to
display the data. This equipment, designated by GM as the Total Torque Tester
(TTT), totalizes and digitally displays time, torque and distance. Distance
measurements were made using a trailing-wheel device which incorporated an^
optical encoder permitting an accuracy of 0.1 ft. to be realized. A descrip-
tion of this test instrumentation is included in Appendix D. Velocity, there-
fore, is a calculated quantity based on time and distance measurements.
Tests were conducted in dry weather only and test procedures employed for.
each vehicle were identical. Prior to each test day, the fuel tank was filled,
the vehicle weight at each wheel was recorded and the tires were set to the manu-
facturer's recommended pressure (24 psi). The vehicle was operated for 30
minutes at 50 mph to stabilize tire temperatures. Wind velocity and direction,
barometric pressure, weather conditions and ambient temperature were recorded.
The vehicle was then operated for two laps, at each test speed, on the Proving
Ground north-south straightaway which is a flat, level test track that incor-
porates high-speed turn loops at either end.
Data were collected for one mile in each direction with four miles of data
taken at each speed on each day. A minimum of three days of test data were
collected for each vehicle and tire-type combination. Wheel torque and wheel
horsepower measurements were made at average steady speeds of 20, 30, 40, 50
and 60 mph. Test data were logged manually from the digital displays.
Road tests took glace during September/October, 1977. Ambient temperature
varied from 48 F to 67 F for these tests with barometric .pressure ranging from
28.40 to 29.70 in Hg. Data shown in the next section were not corrected for
these changes in ambient conditions which would affect the retarding forces
experienced by the vehicles due to aerodynamic drag and tire rolling resis-
tance.
4.3.2 Results
A tabular summary of the numerical data resulting from the road tests on
the Oldsmobile is shown in Table 2. Velocity, wheel torque and wheel horse-
power are tabulated for each tire construction. Details of the wheel horsepower
calculation are shown as step 4.1.12 of Appendix C. Corresponding data
for the Chevette are shown in Table 3.
2
Least-squares curves of the form (a + bV ) were calculated to fit the
wheel torque data for each vehicle as tested with each of the three sets of
tires. Similarly, least-squares curves of the form (cV + dV ) were fitted to
the wheel horsepower data. A listing of the least-squares equations that apply
to the test data appears in Table 4. A measure of the goodness of fit of the
data to these equations is indicated by the magnitude of the quantity "s"
12
-------
Table 2
STEADY-STATE MEASUREMENTS OF WHEEL TORQUE
AND WHEEL HORSEPOWER AT VARIOUS SPEEDS ON A
1977 OLDSMOBILE 98 (TEST WEIGHT = 4561 LB) EQUIPPED
WITH BIAS, BIAS-BELTED AND RADIAL-PLY TIRES
(ROAD-TEST DATA)
G78-15
G78-15
GR78-15
TIRE
CONSTRUCTION
Bias Ply
Bias Belted
Radial Ply
ROAD
SPEED, MPH
19.86
29.91
40.07
50.15
60.29
19.99
30.05
40.05
50.23
60.40
19.86
29.87
39.95
50.19
60.29
WHEEL
TORQUE, FT-LB
96.26
112.64
134.44
166.47
201.99
104.27
124.41
148.05
181.96
218.87
80.62
94.02
114.11
142.98
177.48
WHEEL
HP
4.64
8.14
12.98
20.06
29.19
4.99
8.92
14.09
21.65
31.19
3.86
6.78
10.95
17.19
25.68
NOTE:» Vehicle operated for a distance of 10 miles at
each constant speed.
• Values shown represent the average of three test
points.
13
-------
Table 3
STEADY-STATE MEASUREMENTS OF WHEEL TORQUE
AND WHEEL HORSEPOWER AT VARIOUS SPEEDS ON A
1977 CHEVETTE (TEST WEIGHT = 2491 LBf EQUIPPED
WITH BIAS, BIAS-BELTED AND RADIAL-PLY TIRES
(ROAD-TEST DATA)
TIRE ROAD WHEEL WHEEL
CONSTRUCTION SPEED, MPH TORQUE, FT-LB HP
P155/80D13 Bias Ply 19.99 46.97 2.72
29.92 59.11 5.11
40.06 75.73 8.74
50.11 98.76 14.21
60.24 127.34 21.97
P155/80D13 Bias Belted 19.93 49.73 2.91
29.86 61.87 5.41
39.72 78.00 9.04
49.82 101.12 14.67
59.77 126.23 21.91
P155/80R13 Radial Ply 20.18 43.27 2.54
30.04 55.73 4.87
40.04 72.07 8.39
49.92 93.87 13.62
59.67 121.55 21.08
NOTE:* Vehicle operated for a distance of 10 miles
at each constant speed.
• Values shown represent the average of three test
points.
* Vehicle was ballasted to =2500 Ibs. for these tests.
14
-------
Table 4
SUMMARY OF LEAST-SQUARES CURVE FITS TO
WHEEL TORQUE AND WHEEL HORSEPOWER DATA FOR
THE CHEVETTE AND THE OLDSMOBILE 98
(ROAD-TEST DATA)
TIRE
CONSTRUCTION
Bias Ply
Bias Belted
Radial Ply
TIRE
CONSTRUCTION
Bias Ply
Bias Belted
Radial Ply
CHEVETTE T (t°T^ = £(V) OLDS 98
T = 36.61 + 0.0249V2
s = 0.54 ft-lb
T = 40.23 + 0.0242V2
s = 0.51 ft-lb
T = 33.01 + 0.0247V2
s = 0.58 ft-lb
T = 83.09 + 0.0327V
s = 0.82 ft-lb
T = 91.59 + 0.0352V
s = 1.29 ft-lb
T = 67.54 + 0.0300V
s = 1.02 ft-lb
2
2
2
CHEVETTE HP = f '^ OLDS 98
HP = 0.1033V + 7.209 x 10"5V3
s = 0.065 hp
HP = 0.1208V + 6.899 x 10~5V3
s = 0.079 hp
HP = 0.0943V + 7.245 x 10~5V3
s = 0.074 hp
HP = 0.2019V + 7.782 x
s = 0.088 hp
HP = 0.2236V + 8.069 x
s = 0.13 hp
HP = 0.1602V + 7.287 x
s = 0.088 hp
io-V
io-V
io-V
s =
s =
Unbiased estimate of the standard deviation of the
test data.
n - 1
X = measured value of T or HP
m
X =
calculated value of T or HP
from equation
n = number of test points (speeds) = 5
15
-------
associated with each equation. The letter "s" denotes the unbiased estimate
of the standard deviation of the test data with respect to the least-squares
curve. The equation used to calculate "s" is shown at the bottom of Table 4.
The curves fit the data quite well with the largest percentage deviations
occurring at the lowest velocity where the torque and horsepower levels are
the smallest. For wheel torque, one standard deviation corresponds to a de-
viation of about 1% while for horsepower this deviation is about 3%.
Power absorption units on Clayton dynamometers are designed to absorb
horsepower in proportion to the velocity raised to the 2.8 power (approximate-
ly) . This implies that the velocity/torque relationship for a passenger car
is characterized by velocity to the 1.8 power. To test whether th^ g
road data were better described by an equation of the form (a + SV ), the
data obtained for the Chevette equipped with radial-ply tires were employed
to calculate the coefficients a and g. The "s" value obtained for this
equation was 1.34 and compares unfavorably with the corresponding value of
0.58 shown in Table 4. It was concluded that measured wheel torque data are
better described by a second order equation in velocity.
The road test data show that wheel horsepower required to sustain a con-
stant speed is least when the vehicles are operating on radial-ply tires and
largest when operating on bias-belted tires. For the Chevette the maximum
difference among tire constructions is less than one hp while for the Olds-
mobile it is 5.5 hp (both at 60 mph) reflecting the large disparity in tire
rolling resistances. To the extent that all other test conditions remained
constant, these horsepower differences can be ascribed to differences in the
tire power losses.
Assuming that the velocity-cubed term in the wheel horsepower equations
(Table 4) is entirely the result of aerodynamic losses, the coefficients of
this term should be independent of the tire construction use in the tests.
The tabulated results support this assumption reasonably well.
Graphical presentations of the test results are given in Figures 2 through
5. On all figures, the curves drawn represent graphically the least-squares
equations of Table 4 while the plotted symbols represent measured data. Plots
were made on an expanded vertical scale to emphasize the excellent conformity
of the fitted curves to the measured data.
4.4 TIRE ROLLING RESISTANCE MEASUREMENT
A measurement of the tire rolling resistance was required so that the road
load horsepower settings for certain of the dynamometer tests could be deter-
mined. Specifically, these tests required that the dynamometer setting of road
load horsepower, as measured from vehicle road tests on radial-ply tires, be
augmented by the difference in power loss (at 50 mph) between the front bias-
16
-------
SYMBOLS REPRESENT TEST DATA
O BIAS TIRES
A BIAS-BELTED TIRES
O RADIAL TIRES
20 30 40
V = ROAD SPEED, mph
50
60
Figure 2 Least-squares curves fitted to wheel-torque data for 1977 Chevette equipped with
bias, bias-belted and radial tires.
17
-------
220
SYMBOLS REPRESENT TEST DATA
O BIAS TIRES
A BIAS-BELTED TIRES
O RADIAL TIRES
20 30 40
V= ROAD SPEED, mph
Figure 3 Least-squares curves fitted to wheel-torque data for 1977 Oldsmobile 98
equipped with bias, bias-belted and radial tires.
18
-------
21
20
19
18
17
16
15
CC 14
UJ
O
&• 13
in • °
in
CC
O
X 12
ui
UJ
11
10
Figure 4
H =0.1 33V +
SYMBOLS REPRESENT TEST DATA
O BIAS TIRES
A BIAS-BELTED TIRES
O RADIAL TIRES
HP = 0.
208V
7.209
x 10
10
20 30 40
V= ROAD SPEED, mph
50
60
Least-squares curves fitted to wheel-horsepower data for 1977 Chevette equipped
with bias, bias-belted and radial tires.
19
-------
SYMBOLS REPRESENT TEST DATA
O BIAS TIRES
A BIAS-BELTED TIRES
O RADIAL TIRES
0.2236V+ 8.069x10
0.2019V+ 7.782x10
20 30 40
V = ROAD SPEED, mph
Figure 5 Least-squares curves fitted to wheel-horsepower data for 1977 Oldsmobile 98
equipped with bias, bias-belted and radial tires.
20
-------
ply and front radial -ply tires. This difference is calculable from a knowledge
of the rolling resistance of these tires.
Tests were performed on all of the bias-ply and the radial -ply tires.
Since rolling resistance data were not required for the bias-belted tires, only
the rear tires in this type construction were tested to provide comparative
results.
4.4.1 Methodology
The measurement of tire rolling resistance on the Calspan Tire Research
Facility (TIRF) (see Appendix E) is routine procedure (3) . In the present
case, the tires were received from GM mounted on individual rims following the
completion of the road tests. Each tire was marked according to its location
on the vehicle. Each tire was demounted and then mounted on TIRF rims of the
same width as the original rims*. Each tire/rim assembly was mounted on the
shaft of the TIRF metric balance which measures the three orthogonal forces
and moments produced by the tire. The tires were operated on the textured
surface of the flat roadway at a speed of 50 mph. Tires were free rolling and
operating at zero degrees camber and slip angle. GM supplied data on the
actual normal load experienced at each tire position on each vehicle in its
test configuration. Within the tolerance limits (<10 Ibs.) of the machine,
each tire was tested at its designated load. Cold inflation pressures of 24
psi (165 kPa) , the same as used in road and most flat-bed dynamometer tests,
were used.
Each test was of a 30-minute duration which was sufficient to attain an
equilibrium state in tire temperature, tire pressure and rolling resistance
force. TIRF machine operation and data sampling are computer controlled.
4.4.2 Results
Rolling resistance force (FR, Ib.) is calculated from measured values of
longitudinal force (FX) , bearing friction torque (BFT, ft-lb) and tire loaded
radius (RL, in.) according to the following equation.
FR -FX
RL
* Tires and rims were carefully marked and the tires were remounted as
originally received.
21
-------
The FX and BFT values are corrected for any instrumentation zero shifts ex-^
perienced during the course of the tests and represent the average of the final
20 data points* taken during a test. A summary of the test results is pre-
sented in Table 5. A perusal of the data will show that the rolling resistance
for a given size tire correlates with the normal force (rolling resistance
varies approximately linearly with normal force). One replicate run was made
and the data are typical of the repeatability achieved on TIRF tests.
These rolling resistance data confirm the results obtained from the road
tests presented in Section 4.4.2 which showed that in order of increasing power
loss the tire constructions rank in the following order: radial, bias and bias-
belted.
4.5 DYNAMOMETER CHECK TESTS
By contractual requirement, the operational integrity of the flat-bed
dynamometer was to be demonstrated at Calspan by performing steady-state runs
to 65 mph and duplicating the accelerations/decelerations of the FTP test cycle
using a 4500-lb. inertia class vehicle. In addition to demonstrating the
structural adequacy of the dynamometer, the tests would answer the question of
vehicle stability on the flat surfaces of the SRU's and the criticality of
vehicle/dynamometer alignment. On a dual-roll type of dynamometer, the cradling
of the drive wheels automatically aligns the vehicle laterally and provides
a fore-and-aft centering stability.
Lacking a pit which would place the belt surfaces at floor-level, it was
necessary to perform the tests with the dynamometer base resting on the floor
(belt surfaces about 3 feet above floor level). Testing was performed using a
1978 Chevrolet Impala sedan which was placed on the dynamometer with a fork-
lift truck. After aligning and leveling the vehicle, the front wheels of which
rested in hollows formed in massive wooden blocks, the rear of the vehicle was
tethered loosely to the dynamometer base with chains to provide lateral re-
straint.
A series of steady-speed tests to 75 mph and several complete FTP velocity/
time schedules were completed without difficulty. For all tests, the vehicle
exhibited no perceptible lateral movement. Also, there was no evidence of any
belt motion relative to the drums. Water retention at the bearing seals was
satisfactory with only minor seepage being observed.
*Data were sampled at a rate of one per second.
22
-------
Table 5
SUMMARY OF EQUILIBRIUM VALUES OF THE ROLLING
RESISTANCE FORCE FOR THE TEST TIRES
TIRE
SIZE
GR78-15
G78-15
G78-15
P155/80R13
P155/80D13
P155/80B13
TYPE OF
CONSTRUCTION
Radial Ply
Bias Ply
Bias Belted
Radial Ply
Bias Ply
Bias Belted
POSITION
ON VEHICLE
LR
RR
RF
LF
LR
RR
RF
LF
LR
RR
LR
RR
RF
RF*
LF
LR
RR
RF
LF
LR
RR
TIRE LOADED
RADIUS, IN.
12.72
12.74
12.58
12.57
13.10
13.16
13.04
13.03
13.30
13.31
10.69
10.70
10.70
10.69*
10.61
11.00
10.97
11.00
10.90
10.97
10.91
NORMAL
FORCE, LB.
-1073
-1077
-1192
-1239
-1081
-1078
-1192
-1236
-1078
-1077
-603
-602
-601
-600*
-676
-601
-600
-597
-673
-600
-607
ROLLING RESISTANCE
FORCE, LB.
11.69
11.53
12.68
13.56
13.87
13.97
15.61
16.26
14.74
14.58
7.96
7.81
8.00
7.94*
8.87
8.47
8.16
8.64
10.17
9.49
9.57
Repeat Run
All 15-inch tires are by Uniroyal,
All 13-inch tires are by General.
-------
Following these tests, personnel from the GM Proving Ground requested the
opportunity to perform coastdown tests on the dynamometer at Calspan. The
purpose of these tests was to obtain a true horsepower versus indicated horse-
power calibration of the unit. These tests are performed by driving the dyna-
mometer up to a speed of a 65 mph, rapidly lifting the vehicle free of the
unit and measuring the velocity-time coastdown curve. Excessively large
quantities of water leaked past the water-bearing seals as soon as the vehicle
load was removed from the bearings. Even if the water leakage problem were re-
solved, the validity of such a coastdown calibration would be questionable
since the viscous losses in the bearing may be a function of load.
As an alternative approach, GM personnel performed vehicle-on-dyno tests
on a Clayton unit at Milford using an Impala sedan which was equipped with the
same drivetrain and tires as was the Calspan vehicle. Coastdown tests were
made with the vehicle's transmission in neutral and the vehicle resting on the
dynamometer. GM personnel then visited Calspan and a similar series of coast-
down tests was performed on the flat-bed unit using the Calspan Impala vehicle.
Equivalent loads of 2250 and 4500 Ib. were used.
Correcting for measured vehicle power loss from the tests on the roll
dynamometer, a frictional horsepower loss of approximately 10 hp was found for
an equivalent load of 4500 Ib. at a velocity of 50 mph and the water bearings
under load. Comparable friction losses for the roll dynamometer are about 3 hp.
Other findings from these tests were:
• With dynamometer horsepower settings identical at 50 mph, the flat-bed
unit imposed larger power loadings at the lower speeds and smaller
loadings at the higher speeds than the roll dynamometer.
• At an equivalent weight setting of 4500 Ib. and dynamometer load equal
to vehicle road load at 50 mph, the flat-bed unit exhibited a horse-
power/velocity characteristic that more closely duplicated the road-
measured* characteristic for the Oldsmobile than did the roll dynamo-
meter.
• At an equivalent weight setting of 2250 Ib. and a velocity of 50 mph,
the flat-bed unit provides a power loading, with the PAU completely
unloaded, equal to the measured road load horsepower for the Chevette
at 50 mph.
• Repeatability of the test data for the flat-bed/vehicle combination
was very good.
Vehicle road tests (Section 4.3) preceded these tests in a chronological
sense.
24
-------
These data demonstrated the inherent differences in the horsepower/
velocity characteristic for each type of dynamometer.
Following these preliminary checks at Calspan, the flat-bed dynamometer
system was shipped to th^ Vehicle Emissions Laboratory at the Milford Proving
Ground. Since there was no pit available in which the dynamometer could be
installed, it was necessary to operate the unit above floor level as had been
done at Calspan. With the aid of a hydraulically actuated platform, large
enough to accommodate the test vehicle, the vehicle could be raised to the
level of the SRU belt surfaces and either pushed or driven onto the dynamometer.
Figure 6 shows an annotated picture of the flat-bed dynamometer installa-
tion at the GM Proving Ground. The aluminum channel tracks lying on the belts
are required to span the unsupported lengths of the SRU belts between the water
bearings and the drums whenever a vehicle is moved on or off the dynamometer.
The unsupported belts cannot sustain the loads imposed by the test vehicles.
In concluding this section of the report, it is appropriate to identify
two operational factors related to the PAU used with the flat-bed dynamometer.
This PAU was an early Clayton model not equipped with the automatic road load
power control used on later models. The early models are known to have sizable
hysteresis characteristics which can adversely affect measurements made in
cyclic type tests (4). Such a unit was not available for this test program.
As stated earlier (Section 4.1), the PAU was coupled directly to the SRU
drum shaft despite the fact that at any speed the PAU rpm would be lower than
that of the PAU used on the roll dynamometer. This was done for reasons of
economy after it had been demonstrated, using Clayton data, that the same PAU
horsepower versus dynamometer velocity characteristic could be obtained over
different ranges of the shaft rpm (within limits). Clayton personnel verified
this fact*.
4.6 VEHICLE DYNAMOMETER TESTS
At the time the flat-bed dynamometer was received at the GM Proving Ground,
the Emissions Laboratory did not have a dyno pit available for the installation.
Consequently, the dynamometer was operated above floor level in the manner des-
cribed in Section 4.5 and Appendix A. Ambient temperature and humidity were
not controlled in this area as they were in areas where the permanent dynamome-
ter installations are located.
In order to avoid biases in the test data which might result from making
all the tests in sequence on the flat-bed dynamometer and then on the Clayton
dynamometer, tests between the two units were performed in a randomized manner.
Telephone conversation with. Max Moore, Clayton Manufacturing Co., El Monte,
California, May 2, 1977.
25
-------
to
FLYWHEEL
INERTIA UNIT
POWER
ABSORPTION
UNIT
SIMULATED
ROADWAY
UNITS
Figure 6 Flat-bed dynamometer installation at the GM Proving Ground at Milford.
-------
As a result, tests were made on whatever Clayton dynamometer site was available
at the time. As many as four different dynamometers were used in completing
the sequence of tests whose results are discussed in this section. Each dyna-
mometer was calibrated and maintained to EPA certification standards.
Testing began on the Oldsmobile 98 because the differences in wheel
torque/horsepower among the tire constructions used on this vehicle were
larger than for those used on the Chevette (see Section 4.3). These larger
differences would make it easier to detect any differences in fuel economy and
emissions, measured on the two types of dynamometers, ascribable to tire
construction. All tests adhered to current certification rules requiring a re-
peat of the entire test if any vehicle or equipment malfunction occurred.
About 60% of the test runs were repeated for this and other reasons. A con-
siderable number of the runs on the flat-bed unit had to be repeated because
of an incorrect inertia setting*.
Operational difficulties developed with the flat-bed dynamometer. With
useage, the leakage at the water seals increased to the extent that it caused
a curtailment of testing. A major reason for the seal problems appeared to be
abrasive wear caused by rust and dirt particles,from the chassis of the test
and dyno warm-up vehicles, getting onto the inner surface of the SRU belts.
New seals were fabricated and installed to correct this problem. Also belt-
tracking instabilities were encountered at times with the belts tending to drift
off the drums. During steady-speed tests, the direction of the drift changed
with speed. A belt-tracking adjustment on the moveable drum was used to
stabilize the belt drift.
A total of five belt failures occurred during the test period from
December 1978 to March 1979 with three of these experienced in the last month.
In all cases, lateral cracks developed in the belt to a length of several
inches. The first four cracks occurred in belts used on the same SRU and in
each case the crack started at the edge of the belt. One of these failures
was attributable to damage sustained by the belt during installation when
rapid migration during belt run-in and adjustment caused one edge to come in
contact with the SRU frame. Despite no visible damage, this belt shortly
thereafter developed a series of lateral cracks along the damaged edge of the
belt. The last belt failure was experienced on the remaining original belt.
This crack was located in the central portion of the belt. With this belt
failure the available stock of replacement belts was exhausted.
Belt fatigue failures are felt to be related to manufacturing problems and
not to design deficiencies. Belts for TIRF and the SRU's have been obtained
from two sources. To date, all belt fatigue failures on TIRF and the flat-bed
The inertia unit selector switch for the Oldsmobile vehicle was set for 4500-
Ib. instead of 5500-lb. (see Appendix A).
27
-------
dynamometer have occurred in belts made by the same manufacturer*.
A review of the project status from the standpoint of time schedule, costs,
objectives and test results obtained, led to the mutual decision that all
experimental work would cease at this point. The major consideration was
that the test results that had been acquired were sufficient to demonstrate
the differences in measurements made on the two types of dynamometers. Thus
all dynamometer tests on the Chevette vehicle were eliminated as well as two
steady-speed tests on the flat-bed unit with the Oldsmobile equipped with the
bias-ply tires.
To place the cited operational problems experienced with the flat-bed
dynamometer in perspective, it is necessary to point out that the two SRU's
were the initial prototypes and that these were expected to see limited use.
Even though no tests were performed on the Chevette vehicle, dynamometer
operating time considerably exceeded original estimates.
4.6.1 Dynamometer Calibration
Different procedures were used in calibrating the two types of dyna-
mometers. That used for the Clayton dynamometers will be described first.
With the knowledge that several different Clayton units would be used in
the course of the test program, the performance of the conventional coastdown
tests to measure friction horsepower and true horsepower as a function of
velocity on any one unit was unreasonable. Instead, a partly experimental
and partly analytical method was used to determine an average true absorbed
horsepower. Three calibrated Clayton units were selected randomly and the
Oldsmobile vehicle** was operated on each at a speed of 50 mph. The road load
power on each unit was adjusted until the measured wheel torque equaled that
measured during the road tests at this speed (142.98 ft-lb.). The true dyna-
mometer hp in each case was noted and the three values averaged to yield 11.60
hp with the measured values within -0.2 hp of the mean. Note that the corres-
ponding indicated hp corresponds to the setting PR specified in Table B-l
of Appendix B. From prior coastdown tests on these dynos with the PAU's
disconnected, the average friction hp values were calculated:
Speed, mph Friction, hp
20 0.581
30 1.104
40 1.814
50 2.830
60 4.167
*The only exception involved the belt that was damaged in installation.
** Equipped with radial tires inflated to 45 psi.
28
-------
Thus the PAU average absorbed power at 50 mph is 8.77 hp (11.60-2.83) and since
the hp/velocity relationship for the Clayton automatic road load power control
unit is described by the following equation: hp = k [v(mph)] ., the value of k
is found to be 7.016 x 10~ . With the value of k known, the PAU hp at other
speeds can be calculated. Thus summing the experimentally-determined friction
hp and the calculated PAU hp at each speed, the true dynamometer horsepower
at each speed was obtained. The Clayton dynamometer data are summarized below:
Average Average Average
Speed, mph Friction hp PAU hp True hp
20 0.581 + 0.561 = 1.142
30 1.104 + 1.894 = 2.998
40 1.814 + 4.489 = 6.303
50 2.830 + 8.768 = 11.598
60 4.167 + 15.151 = 19.318
As discussed in Section 4.5, coastdown tests on the flat-bed unit could
not be performed because the unloaded water bearing leaked water excessively.
Consequently, steady-speed tests were performed with the vehicle on the
unit and the PAU disconnected. The Oldsmobile was equipped with the radial
tires inflated to 24 psi. At each steady speed the vehicle was driven for four
miles. Data were collected during the last two-mile portion of each run. Two
sets of tests were made, one with the speeds in increasing sequence and one
with the speeds in decreasing sequence. Resultant data were averaged and used
to calculate wheel horsepower according to the following equation.
HP.
Wheel HP = 2* x R. ^ x
int t.
i
Where: HP = integrated HP measured with the TTT (HP-sec.)
int = integrated time (sec.)
int = integrated distance* (ft.)
j^
int = integrated revolutions, vehicle wheel
R = constant (1.1484) built into the TTT corresponding
to a tire rolling radius (ft.)
* Measured with a calibrated wheel operating on the roadway, see Appendix D.
29
-------
A summary of the average wheel horsepower data and friction_horsepower for
the flat-bed dynamometer is tabulated below. Note that the friction horse-
power data were obtained by subtracting tire power consumption from the wheel
horsepower using TIRF-measured rolling resistance for the two rear radial-ply
tires*.
Average
Speed, mph Wheel hp Tire hp Friction hp
20 3.570 1.238 2.332
30 6.010 1.858 4.152
40 9.081 2.477 6.604
50 14.231 3.096 11.135
60 19.572 3.715 15.857
A plot of the friction horsepower as a function of velocity for the flat-
bed and the Clayton dynamometers is shown in Figure 7 while the true absorbed
horsepower as a function of velocity for the Clayton unit is shown as Figure 8.
Wheel horsepower versus velocity for the Oldsmobile equipped with radial-ply
tires operating on the flat-bed dynamometer, with the PAU disconnected, is
presented in Figure 9. The plots of the dynamometer friction loss dramatically
illustrate the large differences that characterize these two particular types
of dynamometers.
A fuller account of the detailed procedures employed in performing these
tests is included in Appendix C.
4.6.2 Test Procedures
The test design (Appendix B) specified tests on both dynamometers using
each of the three different types of tire constructions. In general a test
day began with a cold start FTP (non-evaporative) test followed by a HFET test
and then by the steady speed (SS) tests with tests between the dynamometers
types performed on a randomized basis. Vehicle and dynamometer preparation
for the nonevaporative FTP or HFET tests followed standard certification
procedures.
FTP tests followed the 1976 Federal Emission Testing Procedure without
evaporative testing. Fuel drain, fuel system pressure checks and diurnal heat
build prior to each test were not required. Measurements of wheel torque and
total integrated values of the positive and negative torque were obtained.
23.22 Ibs. total for two tires, see Table 5 in Section 4.4.2.
30
-------
FLAT-BED DYNAMOMETER
CLAYTON DYNAMOMETER
30 40
VELOCITY, mph
60
Figure 7
Friction horsepower versus velocity for the flat-bed and Clayton dynamometers
(PAU disconnected).
31
-------
22
on
£\J
1st
16
CC 14
£ ™
1
S 12
CC
0
X
rn in
\BSORBI
» c
g
4
2
0
1C
I
-A
2(
-_,
)
^
3
/
.—^.
<
3
^. ...--
^
/
/
41
,
/
/
/
3
/
/
/
5
., _-..
/
/
/
/
/
'-
0
1
y.
/
./.
/
60
VELOCITY, mph
Figure 8 True absorbed horsepower for the Clayton dynamometer as a function of
velocity; dyno indicated HP = PR.
32
-------
22 t-
(PAU DISCONNECTED)
20
30 40
VELOCITY, mph
50
60
Figure 9 Wheel horsepower versus velocity, Oldsmobile 98 with radial-ply tires
(GR78-15) on the flat-bed dynamometer.
33
-------
The HFET tests were performed in accordance with the "EPA Recommended Practice
for Conducting Highway Fuel Economy Tests". Thus if the HFET followed within
three hours of an FTP, a preconditioning HFET cycle was performed. Fuel econ-
omy as well as the integrated values of the positive and negative wheel torques
were measured. Although not required by the Statement of Work, emissions were
also measured. Just prior to the preconditioning HFET cycle, the vehicle was
operated for one minute at 50 mph and the dynamometer indicated hp was recorded
and, if necessary, adjusted to the proper setting. Steady speed tests were
made at the following speeds and in the order listed: 60, 50, 40, 30, and 20
mph. At each speed the run duration was ten minutes with wheel torque and
exhaust emissions (HC, CO, CO- and NO ) measured. The original test plan
specified a ten-mile run at each speea but a waiver was obtained from the
Project Officer to use ten-minute runs to permit the completion of an FTP/HFET/
SS sequence within one day's testing.
All tests were performed with the dynamometer indicated horsepower set at
either PR or PRA. PR corresponds to that indicated horsepower that was obtained
when the Oldsmobile vehicle, equipped with radial-ply tires inflated (cold)
to the specified pressure*, was operated on the dynamometer at 50 mph and the
PAU was adjusted so that the rear-wheel torque was equal to that measured at
50 mph during the road tests on these same tires on the same vehicle. The
value of PRA was obtained by adding to PR the difference in front-tire power
loss at 50 mph between the bias-ply and the radial-ply tires as calculated
from TIRF measurements of the rolling resistance of the individual tires. For
the Oldsmobile, this difference amounted to 0.75 hp. PRA settings were only
used for tests on the bias-ply tires on the flat-bed dynamometer.
Detailed step-by-step procedures which were used in performing the dyna-
mometer tests and obtaining the test data are included in Appendix C.
4.6.3 Summary of Test Results
4.6.3.1 FTP Tests
The principal purpose of the FTP tests was to measure exhaust
emissions (HC, CO, NO and CO- on a grams-per-mile basis),fuel economy (mpg)
and integrated (positive as well as negative) wheel torques for the radial-ply
and bias-ply tires on each dynamometer at an indicated hp setting equal to PR.
In addition, tests on the bias tires on the flat-bed dynamometer were repeated
using an indicated hp equal to PRA. Each test was performed three times for
replication purposes. A summary of the data means and standard deviations is
presented in Table 6 where the Block number designation for each data set
corresponds to that shown in the contractual Statement of Work. Block numbers
24 psi on the flat-bed unit and 45 psi on the Clayton unit
34
-------
TABLE 6. EXHAUST EMISSIONS, FUEL ECONOMY AND WHEEL
TORQUE DATA FOR THE FTP TESTS
1977 OLDSMOBILE, 4500 Ib INERTIA WEIGHT
Block
No.
1
2
11
14
15
Dynamometer
Flat Bed
Flat Bed
Flat Bed
Clayton
Clayton
Type
Test
FTP
FTP
FTP
FTP
FTP
No. of
Tests
3
3
3
3
3
Tire
Type
Radial
Bias
Bias
Radial
Bias
Inflation
Pressure
MRP
MRP
MRP
45 psi
45 psi
Indicated
hp
PR
PR
PRA
PR
PR
HC
(gm/mi)
JL--S
0.632
0.046
0.849
0.087
0.678
0.015
0.654
0.124
0.689
0.024
CO
(gm/mi)
^^^
5.31
0.248
9.11
0.600
5.81
0.640
3.70
0.318
5.23
0.472
NOX
(gm/mi)
*^-^C
1.527
0.091
1.669
0.091
1.809
0.081
1.189
0.0093
1.190
0.057
CO2
(gm/mi)
*^-^;r~
631.5
4.52
693.2
15.4
706.9
14.2
611.5
13.7
599.4
21.3
F.E.
(mpg)
*^^^r
13.82
0.085
12.49
0.272
12.31
0.266
14.33
0.315
14.70
0.531
Avg. + Torq.
(Ib-ft)
^^—-7^"
165.30
4.85
178.03
4.84
185.42
5.44
150.67
4.10
144.26
2.34
Avg. • Torq.
(Ib-ft)
J^, -^^
76.49
1.42
73.59
2.53
71.71
2.92
78.95
1.52
80.61
2.21
en
x = SAMPLE MEAN
s = SAMPLE STANDARD DEVIATION
MRP = MANUFACTURER'S RECOMMENDED PRESSURE (24 psi)
-------
provide a convenient way to designate data sets when the differences between
pairs of data means are analyzed for statistical significance in Section 4.6.3.4,
The wheel torque data are presented as average torque values rather than inte-
grated values since ft-lb-sec is not a meaningful quantity. An average torque
for the FTP test cycle was obtained by summing the average for the first 1372
seconds of the test with that for the last (hot-transient portion) 505 seconds
of the test.
Fuel economy data show that for a given tire construction, tests conducted
on the Clayton dynamometer yield a higher level of fuel economy than do
similar tests on the flat-bed unit. Further, on the roll dyno the bias-ply
tires yield a larger miles-per-gallon figure than do the radial-ply tires.
On the other hand, the flat-bed dynamometer ranks the tires in the reverse
order in terms of fuel economy and in agreement with fuel economy data measured
for vehicles operated on test tracks where the radial-ply tires show a pro-
nounced and consistent advantage over bias-ply tires.
Fuel economy data correlate inversely with CCL emissions levels. This
result is inevitable as C0_ is the major factor in the carbon-balance equation
employed to calculate fuel economy from exhaust emissions measurements. With
the indicated horsepower set at PRA, an increase of 0.75 hp, the fuel economy
is slightly lower than that at the PR condition. This result is consistent
with expectations.
There are no obvious trends apparent in the HC and CO emission levels.
Since the measurements of these two parameters are known to be characterized by
a large variability (5), the inconclusiveness of data means calculated from
only three tests is not surprising. Data variability is illustrated by the
results shown for Blocks 2 and 11 where the only difference in test conditions
is a 0.75 hp change in indicated hp. HC and CO levels show large differences
which cannot be explained on the basis of the small change in the hp setting.
Comparable tests (Blocks 1-14 and Blocks 2-15) do show that the NO levels
measured on the flat-bed dynamometer are consistently larger than fhose measured
on the roll dynamometer. In all cases, these data correlate directly with the
levels of the positive wheel torque which is a measure of the power expended
in driving the dynamometer. In Section 4.6.4, reference is made to ambient
environmental factors that also could have affected these data.
The positive-torque data are seen to be smaller for the roll dynamometer
than for the flat-bed unit. This is the result of the differences in the hp/
velocity characteristics between the two units. At speeds below 50 mph, the
flat-bed dynamometer has a consistently higher hp loading. Dynamometer test
data illustrating these differences are included in Section 4.6.3.3.
36
-------
Negative-torque data are a measure of the power expended in decelerating
the dynamometer. They are inversely correlated with the positive torque data.
This relationship seems logical. Since the same test is being repeated and the
system inertias remain the same, it follows that the more power that is ex-
pended in accelerating a system, the less power is required in decelerating it.
Consider data Blocks 1 and 2 in which the only test difference is in tire con-
struction. The larger average positive torque for the bias-tire case compared
to that of the radial-tire case is the result of the higher rolling resistance
of the former relative to the latter. When braking traction is required, the
higher rolling loss of the bias-ply tire necessitates a lower negative torque
for a given braking traction force than for a radial-ply tire. Hence, the
negative torque level for Block 2 data is smaller than that for Block 1.
4.6.3.2 HFET Tests
HFET tests were performed to measure fuel economy and the positive
as well as negative torque. These data are shown in Table 7 as the means
for three tests and the associated sample standard deviations. Emissions
data were measured and these are also included.
Fuel economy data show the same trends as were found for the FTP
tests. On the flat-bed dynamometer the radial-ply tires yield a better fuel
economy than do the bias-ply tires. On the roll dynamometer, the reverse trend
is found. In addition, the fuel economy data are higher for the roll dynamo-
meter tests than they are for tests on the flat-bed unit. To a large extent,
this difference is the result of the fact that, on the average, the flat-bed
unit imposes a larger horsepower loading on the vehicle than does the roll-
type unit.
Positive torques correlate inversely with fuel economy as do the FTP
tests. On the other hand, the negative torques do not correlate as well with
the positive torques as was the case previously. Because of the large differ-
ences between the FTP and the HFET cycles in terms of accels/decels, the abso-
lute magnitudes of the negative torques are much smaller and hence experimental
variability factors could affect the data to a much greater extent. As before,
the average torque values are listed. Integrated torque can be calculated by
multiplying the data by 765 seconds.
The emissions data again are characterized by considerable scatter as
evidenced by the ratio of the standard deviation to the mean*. Emissions data
for CO especially show a large scatter. For comparable test conditions, CO and
NO emissions show consistently higher levels for tests performed on the flat-
bed dynamometer when compared with those performed on the roll dynamometer.
* Also referred to as the coefficient of variation
37
-------
TABLE 7. EXHAUST EMISSIONS, FUEL ECONOMY AND WHEEL
TORQUE DATA FOR THE HFET TESTS
1977 OLDSMOBILE, 4500 Ib INERTIA WEIGHT
Block
No.
3
4
12
16
17
Dynamometer
Flat Bed
Flat Bed
Flat Bed
Clayton
Clayton
Type
Test
HFET
HFET
HFET
HFET
HFET
No. of
Tests
3
3
3
3
3
Tire
Type
Radial
Bias
Bias
Radial
Bias
Inflation
Pressure
MRP
MRP
MRP
45psi
45 psi
Indicated
hp
PR
PR
PRA
PR
PR
HC
(gm)
*^^?
0.818
0.033
1.041
0.091
0.953
0.039
0.883
0.113
0.873
0.019
CO
(gm)
~*^^T
8.510
4.88
17.28
6.37
17.88
2.80
1.314
0.997
4.057
2.75
NOX
(gm)
^-^~^*~
14.30
1.43
18.47
0.860
18.31
0.491
12.79
0.789
13.32
0.509
CO2
(gm)
JL— -~ — ^
4543.3
54.02
5037.4
66.90
5094.0
46.44
4470.2
46.01
4349.0
83.49
F.E.
(mpg)
~*^^Z~
19.92
0.201
17.93
0.254
17.72
0.146
20.27
0.159
20.85
0.406
Awg. + Torq.
(Ib-ft)
JL- — •""'«"'
160.89
2.71
183.59
2.69
187.80
4.15
160.00
9.41
149.68
3.84
Avg. - Torq.
(Ib-ft)
JL— - • — *~"
18.29
1.34
19.10
1.64
16.93
0.345
17.51
1.28
18.45
0.502
O4
00
x = SAMPLE MEAN
s = SAMPLE STANDARD DEVIATION
MRP = MANUFACTURER'S RECOMMENDED PRESSURE
-------
4.6.3.3 Steady-State Tests
As in all of the earlier tests, emissions and wheel-torques were
measured at each of the five steady state velocities that consituted a test
run. In contrast to the other tests, the schedule provided for only two runs
at each condition and, in some cases, only for one run. When possible, data
means and standard deviations were calculated. From these measured data, such
other parameters as fuel economy, wheel horsepower, true dynamometer absorbed
horsepower and tire horsepower consumption were calculated for each velocity.
A summary of the numerical results appears in Table 8. Emissions of CO were
so low as to be on the threshold of detection and therefore do not appear in
the Table. A discussion of the results follows together with graphical pre-
sentations of selected data.
Emissions data generally are in agreement with those obtained
during the FTP and HFET tests. With even fewer replicate data available, the
significance that can be attached to the numerical results is questionable.
A series of computer-prepared plots, Figures 10 through 13, shows the variation
of HC and NO emissions, on a total weight basis, as a function of velocity for
the radial-ply and bias-ply tires tested on the flat-bed and roll-type dyna-
mometers. In all these plots, the curves represent second-order, least-squares
fits to the data. HC emissions tend to decrease with velocity for the roll
dynamometer tests and show a peaked response for the flat-bed tests with a
maximum at about 45 mph. The NO emissions data show that the levels are con-
sistently higher for the flat-bea unit for both tire constructions. The
difference between the emissions levels for comparable tests on the two dyna-
mometers appears to be in the form of a bias.
Tests on the bias-belted tires were scheduled only in the steady-
state sequence of tests. Fuel economy results are plotted against the type of
tire construction involved in Figure 14 at each of the constant velocities.
Data at the 20 mph speed could not be used since the vehicle was operated in
second gear at this speed. This figure dramatically illustrates the reversal
in tire ranking, according to fuel economy, that occurs between the two dyna-
mometers. For the flat-bed unit tests, the tire ranking in order of increas-
ing fuel economy is: bias, bias-belted and radial ply. For the roll dyna-
mometer the order is reversed. The intermediate position occupied by the bias-
belted tires is surprising, especially for the flat-bed tests, considering
the fact that this construction produced the largest wheel horsepower levels
in the road tests and the highest rolling-resistance levels in TIRF test.
Consequently, it would be expected that this construction would produce the
lowest fuel economy data in the flat-bed tests.
39
-------
TABLE 8. EXHAUST EMISSIONS, FUEL ECONOMY, WHEEL TORQUE AND
TIRE POWER CONSUMPTION DATA FOR STEADY SPEED
TESTS -- 1977 OLDSMOBILE, 4500 LB INERTIA WEIGHT
Block
No.
5
6
7»
8
9
10
Dynamometer
Flat Bed
Flat Bed
Flat Bed
Flat Bed
Flat Bed
Flat Bed
Type
Test
SS
SS
SS
SS
SS
SS
No. of
Tests
2
2
1
1
1
1
Tire
Type
Radial
Bias - B
Bias
Radial
Bias - B
Bias
Inflation
Pressure
MRP
MRP
MRP
45 psi
-
45 psi
45 psi
Indicated
hp
PR
PR
PR
PR
PR
PR
Velocity
(mph)
60
50
40
30
20
60
50
40
30
20
60
50
40
30
20
60
50
40
30
20
60
50
40
30
20
HC
(gins)
*^^C
0.494
0.0028
0.560
0.0064
0.558
0.0559
0.450
0.0120
0.354
0.0396
0.622
0.0417
0.668
0.0113
0.662
0.0233
0.634
0.0205
0.447
0.123
0.543
0.720
0.770
0.601
0.396
0.523
0.631
0.791
0.626
0.646
0.603
0.675
0.642
0.657
0.440
NOX
(gms)
*^^
18.72
0.354
8.008
0.130
4.117
0.0523
2.546
0.315
1.604
0.261
20.08
0.757
8.136
0.0926
4.386
0.139
2.662
0.170
1.682
0.0672
22.24
8.479
4.227
2.818
1.766
17.56
6.839
3.159
2.123
1.297
18.74
7.143
3.751
2.589
1.589
CO2
(gms)
T^^T
4423.1
54.1
3380.8
20.00
2597.0
17.61
1857.7
15.41
1480.2
12.87
4692.0
102.2
3620.1
17.25
2753.2
1.34
2024.2
2.90
1581.0
33.09
4980.8
3778.9
2819.4
2106.1
1629.9
4868.5
3487.5
2508.4
1855.1
1471.3
4637.4
3518.2
2646.5
1932.5
1525.8
F.E.
(mpg)
x^--V
20.12
0.205
22.13
0.467
22.82
0.247
23.70
0.0566
19.88
0.304
18.87
0.488
20.33
0.156
21.40
0.0141
21.74
0.0707
18.60
0.382
17.76
19.47
20.89
21.08
18.19
19.60
21.26
23.69
24.27
19.57
18.97
20.88
22.18
22.89
19.29
Wheel
hp
^^^
25.24
0.636
16.82
0.0778
11.42
0.276
7.10
0.0212
3.88
0.0212
26.30
0.290
17.86
0.0990
11.84
0.0990
7.64
0.269
4.26
0.0071
27.67
18.44
12.54
8.155
4.350
24.69
16.23
11.02
6.44
3.32
25.59
16.96
11.16
6.96
3.98
This Run Not Completed - Belt Failure
Note: All 20 mph runs were made in 2nd gear; CO emissions were too low to measure
*A repeat run was not completed - Belt Failure
(Continued)
-------
TABLE 8. (Continued)
Block
No.
13
18
19
20
Dynamometer
Flat Bed
Clayton
Clayton
Clayton
Type
Test
SS
SS
SS
SS
No. of
Tests
2
2
2
2
Tire
Type
Bias
Radial
Bias - B
Bias
Inflation
Pressure
MRP
45psi
45 psi
45psi
Indicated
hp
PRA
PR
PR
PR
Velocity
(mph)
60
50
40
30
20
60
50
40
30
20
60
50
40
30
20
60
50
40
30
20
HC
(gms)
^-"f"
0.554
0.0184
0.666
0.0219
0.629
0.0255
0.598
0.0170
0.552
0.140
0.673
0.0910
0.814
0.109
0.984
0.0926
0.986
0.0976
0.766
0.0191
0.549
0.0240
0.721
0.0509
0.963
0.119
1.408
0.0177
1.005
0.0346
0.519
0.0601
0.672
0.0262
0.827
0.109
1.236
0.0651
1.108
0.129
NOX
(gms)
JL^-^'T'
24.10
1.14
9.254
0.0622
4.698
0.115
2.818
0.0778
1.580
0.0156
17.09
2.14
6.494
0.659
2.910
0.0849
1.956
0.262
1.121
0.126
17.96
1.237
6.600
0.0544
3.114
0.0742
1.966
0.156
1.038
0.0933
18.33
0.113
6.700
0.350
3.198
0.426
1.734
0.211
0.976
0.0856
CO2
(gms)
*^-^Z'
5004.7
3.96
3779.6
26.94
2845.1
5.74
2101.0
25.10
1627.9
22.34
4573.1
80.0
3417.5
119.4
2562.8
20.72
1832.4
18.10
1456.0
20.64
4462.5
14.2
3351.2
16.33
2490.2
5.02
1803.1
19.80
1436.9
6.08
4331.9
74.60
3150.1
74.67
2401.4
99.07
1707.4
23.26
1402.2
38.33
F.E.
(mpg)
^^^
17.64
0.0849
19.36
0.290
20.71
0.283
20.97
0.0141
18.02
0.389
19.33
0.368
21.60
0.757
23.02
0.184
24.31
0.156
20.13
0.212
19.86
0.0282
22.06
0.0283
23.78
0.184
24.56
0.0636
20.56
0.0495
20.50
0.438
23.18
0.827
24.60
1.01
25.86
0.364
21.08
0.785
Wheel
hp
j^---^r
28.67
0.523
18.77
0.198
12.49
0.0566
8.32
0.177
4.43
0.0566
25.58
0.651
16.66
0.120
10.18
0.389
5.76
0.247
3.04
0.346
24.34
0.799
15.75
0.283
9.67
0.226
5.78
0.325
2.84
0.134
24.32
0.198
15.67
0.156
9.62
0.120
5.41
0.0849
2.60
0.0212
Tire Power
Consumption
^-— ' — 1~~
_
-
-
-
-
-
-
-
-
-
6.262
0.651
5.067
0.120
3.872
0.389
2.757
0.247
1.903
0.346
5.017
0.799
4.152
0.283
3.367
0.226
2.782
0.325
1.693
0.134
5.002
0.198
4.072
0.156
3.322
0.120
2.412
0.0849
1.453
0.0212
Note: All 20 mph runs were made in 2nd gear; CO emissions were too low to measure
-------
1.2
1.1
C/5
2 1.0
s.
V)
o
0.9
O
o
cc
> 0.8
X
0.7
0.6
BIAS-PLY TIRES, G78-15
25
+ = CLAYTON DYNAMOMETER (45 psi)
X = FLAT-BED DYNAMOMETER (24 psi)
45 55
VELOCITY (mph)
65
Figure 10 Hydrocarbon emissions as a function of velocity, steady-state tests on bias-ply tires.
42
-------
RADIAL-PLY TIRES, GR78-15
CO
O
0.7
O
cc
> 0.6
x
0.5
0.4
= CLAYTON DYNAMOMETER (45 psi)
X = FLAT BED DYNAMOMETER (24 psi)
ORD:2
25
35 45 55
VELOCITY (mph)
65
Figure 11 Hydrocarbon emissions as a function of velocity, steady-state tests on radial-ply tires.
43
-------
17.5
15
12.5
O
X 10
7.5
RADIAL-PLY TIRES, GR78-15
I
-I- = CLAYTON DYNAMOMETER (45 psi)
X = FLAT-BED DYNAMOMETER (24 psi)
L
2.5
ORD:2
25
35 45 55
VELOCITY (mph)
65
Figure 12 Oxides of nitrogen emissions as a function of velocity, steady-state tests
on radial-ply tires.
44
-------
22.5
20
17.5
15
12.5
10
7.5
BIAS-PLY TIRES, G78-15
-h = CLAYTON DYNAMOMETER (45 psi)
X = FLAT-BED DYNAMOMETER (24 psi)
x
o
2.5
ORD:2
25
35 45
VELOCITY (mph)
55
65
Figure 13 Oxides of nitrogen emissions as a function of velocity, steady-state tests
on bias-ply tires.
45
-------
FLAT-BED (24 psi)
CLAYTON (45 psi)
RADIAL
BIAS BELTED
TIRE CONSTRUCTION
BIAS
Figure 14 Dynamometer steady-state tests, fuel economy versus tire construction.
46
-------
A comparison of the fuel economy data for the radial-ply tires
obtained from tests on the two dynamometers is presented in Figure 15. These
results show that differences in the measured fuel economy are quite small; a
fact borne out by the FTP and HFET fuel economy data as well. Figure 16 shows
similar data for the bias-ply tires where the fuel economy data obtained from
the two dynamometers show larger differences with the roll-dynamometer tests
yielding the higher numerical values.
Wheel horsepower was calculated for each velocity of each steady-
state run using the relation given in Section 4.6. Figures 17 through 19 show
plots of wheel horsepower as a function of steady-state velocity for each tire
construction on both the flat-bed and roll-type dynamometers. Also shown on
each plot are the data obtained during vehicle road tests. Since the dynamo-
meter indicated hp, in each case, was set with respect to the wheel torque
at 50 mph as measured for the radial-ply tires during the road tests, all the
wheel horsepower data plotted for the radial-ply tires at 50 mph should be
substantially identical. Figure 17 bears out this fact. For the other tire
constructions, the data at 50 mph diverge as expected. The graphical presen-
tations show that tests on the flat-bed dynamometer closely duplicate the wheel
horsepower curve established by vehicle road tests on both the radial-ply and
bias-ply tires. For the bias-belted tires, the road test data lie well above
those measured on the flat-bed dynamometer.
The wheel horsepower data measured on the roll dynamometer are consis-
tently lower than those measured on the flat-bed unit for a number of reasons
peculiar to these tests. Firstly, all tests were performed using one dynamome-
ter indicated hp that was established by duplicating the road-test wheel
torque as measured with the vehicle operating on radial-ply tires. It is a
proven fact that the rolling loss of a radial-ply tire is less than that of
a bias-ply tire when both are operated on a flat surface, see data in Table
5, for example. Hence, for tests on the flat-bed unit, the wheel horsepower
for the bias constructions would be larger than for the radial construction.
On the other hand, the rolling loss for the radial-ply tire is larger than
that for the bias tire when both are operated on a dual-roll configuration*
(see Tire Power Loss Data, Table 8). Thus, for the case of the Clayton
dynamometer tests, the wheel horsepower for the bias constructions would be
smaller than for the radial construction. Secondly, the Clayton dynamometer
absorbs less horsepower at velocities below 50 mph than the flat-bed unit if
both are set to the same indicated hp at 50 mph.
A close examination of the wheel horsepower data measured during tests
on each dynamometer shows that the results for the bias-belted tires lie
between those of the other two constructions. This fact explains why this
One reason for this result is the use of increased inflation pressures
for the Clayton dynamometer tests and the relatively high sensitivity
of bias-tire rolling loss to inflation pressure.
47
-------
26
25
24
3
a.
§23
I
O
o
22
*^
x__ "^
RADJAJ--PLY T
+ = CLAYTON 1
•^ '
XN
^»
"V
\
RES, GR78-15
DYNAMOMETER (45 psi) |
DYNAMOMETER (24 psi) |
w.
N\.
\,
ORD:2
Ul
3
LL.
21
20
19
18
25
35 45
VELOCITY (mph)
55
65
Figure 15 Fuel economy as a function of velocity, steady-state tests on radial-ply tires.
48
-------
26
25
24
23
BIAS-PLY TIRES, G78-15
+ = CLAYTON DYNAMOMETER (45 psi)
X = FLAT-BED DYNAMOMETER (24 psi)
O
O 22
O
o
LU
-J 21
20
19
18
,
ORD:2
25
35 45
VELOCITY (MPH)
55
65
Figure 16 Fuel economy as a function of velocity, steady-state tests on bias-ply tires.
49
-------
o
a.
ill
CO
cc
o
X
_l
LU
LU
X
• CLAYTON DYNAMOMETER (45 psi)
O FLAT-BED DYNAMOMETER (24 psi)
A ROAD-TEST DATA
Figure 17
30 40
VELOCITY, mph
Steady-state tests - Oldsmobile 98 Wheel horsepower versus velocity for
radial-ply tires (GR78-15); indicated HP = PR.
50
-------
100
90
80
70
60
50
40
30
20
O
a.
LLJ
00
cc
O
I
i
10
9
7
6
5
4
CLAYTON DYNAMOMETER (45 psi)
BED DYNAMOMETER (24 psi)'
-TEST DATA
O FLAT-
ROAD
—.i
10
20 30 40
VELOCITY, mph
50
60
Figure 18 Steady-state tests - Oldsmobile 98 Wheel horsepower versus velocity
for bias-ply tires (G78-15); indicated HP = PR.
51
-------
• CLAYTON DYNAMOMETER (45 psi)
O FLAT-BED DYNAMOMETER (24 psi)
A ROAD TEST DATA
30 40
VELOCITY, mph
Figure 19
Steady-state tests - Oldsmobile 98 wheel horsepower vs velocity for bias-belted
tires (G78-15); indicated HP = PR.
52
-------
tire construction also ranks in the intermediate position in terms of fuel
economy. It thus appears that bias and bias-belted tires operating on the
flat-bed dynamometer exhibit different relative rolling losses than they do
when operating on a test track or on the TIRF machine.
For dynamometer tests in which the true absorbed horsepower and the
wheel horsepower at each velocity are known, the tire power consumption re-
presents the differences in the corresponding two data points. Since it was
not possible to directly determine the true absorbed power for the flat-bed
dynamometer, tire power consumption could only be calculated for the roll-type
dynamometer. The data appear in Table 8 and are shown in graphical form in
Figure 20 where linear regression lines have been fitted to the test results.
Radial-ply tires are seen to show substantially higher rolling losses on the
roll dynamometer relative to both the bias-belted and bias-ply tires.
As described in Appendix C, the dynamometer indicated hp-was set
properly prior to each of the test sequences (FTP, HFET and SS) and then checked
at the completion of each test sequence following one minute of operation at
50 mph. In all cases, for both dynamometers the differences were so small
(<0.3 hp) that these data are not tabulated in this report.
4.6.3.4 Statistical Analysis
Standard statistical tests were performed to identify differences
in means between specific data-block pairs at the 90, 95, and 99 percent levels
of confidence. The analyses were made to the extent that the available data
permitted, recognizing that not all tests were completed and that only single
runs were made for some test conditions. The "t" test of significance between
two sample means for unpaired variates (6) was employed.
For convenience of reference, Table 9 contains a concise identifi-
cation of each data Block Number in terms of the following set of descriptors:
dynamometer used, type of test performed, number of tests involved, tire con-
struction used, tire inflation pressure (cold) and dynamometer indicated horse-
power setting used. A tabular summary of the results of the analysis for the
FTP and HFET tests is presented in Table 10 where the code used to denote
levels of significance, if any, is explained in the footnote.
An examination of the results shows that there is a consistency
between the FTP and HFET tests and that the following general observations
apply:
• The most significant differences in data means occur between
tests performed on (1) bias-ply tires on the flat-bed and the
roll dynamometers and (2) radial-ply and bias-ply tires on the
flat-bed dynamometer.
53
-------
5'
Q.
X
z"
o
V)
2
O
U
cc
LLJ
I
o.
Ill
cc
O RADIAL-PLY TIRES (GR78-15)
BIAS-BELTED TIRES (G78-15)
O O BIAS-PLY Tl RES (G78-15)
1977OLDSMOBILE
INERTIA WEIGHT =
10
20 30
VELOCITY, mph
40
50
60
Figure 20 Tire horsepower consumption versus velocity; Clayton dynamometer;
tire pressure = 45 psi, dyno indicated HP = PR.
54
-------
Table 9
BLOCK NUMBER DESIGNATIONS FOR
STATISTICAL TESTS
en
cn
BLOCK
NO.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
DYNA-
MOMETER
FB
FB
FB
FB
FB
FB
FB
FB
FB
FB
FB
FB
FB
CLAY.
CLAY.
CLAY.
CLAY.
CLAY.
CLAY.
CLAY.
TYPE OF
TEST
FTP
FTP
HFET
HFET
SS
SS
SS
SS
SS
SS
FTP
HFET
SS
FTP
FTP
HFET
HFET
SS
SS
SS
NO. OF
TESTS
3
3
3
3
2
2
2*
1
1
1 **
3
3
2
3
3
3
3
2
2
2
TYPE
TIRE
Radial
Bias
Radial
Bias
Radial
Bias-B
Bias
Radial
Bias-B
Bias
Bias
Bias
Bias
Radial
Bias
Radial
Bias
Radial
Bias-B
Bias
INFLATION
PRESSURE
MRP
MRP
MRP
MRP
MRP
MRP
MRP
45psi
45psi
45psi
MRP
MRP
MRP
45psi
45psi
45psi
45psi
45psi
45psi
45psi
INDICATED
HP
PR
PR
PR
PR
PR
PR
PR
PR
PR
PR
PRA
PRA
PRA
PR
PR
PR
PR
PR
PR
PR
NOTE: All SS Tests @ 20, 30, 40, 50 and 60 MPH
MRP = Manufacturer's recommended inflation pressure.
* only one test completed
** test not completed
-------
Table 10
SUMMARY OF THE RESULTS OF TESTS OF SIGNIFICANCE APPLIED TO
THE DIFFERENCES BETWEEN MEANS OF DATA FOR THE
FTP AND HFET RUNS
BLOCK
PAIR
1-2
1-11
1-14
2-11
2-15
14-15
3-4
3-12
3-16
4-12
4-17
6-17
TYPE
TEST
FTP
FTP
FTP
FTP
FTP
FTP
HFET
HFET
HFET
HFET
HFET
HFET
TESTS PER
BLOCK
3
3
3
3
3
3
3
3
3
3
3
3
TEST PARAMETER
HC
GM/MI
**
-
-
**
**
-
CO
GM/MI
***
-
***
***
***
***
NO
X
GM/MI
-
**
***
-
***
-
co2
GM/MI
***
***
*
-
***
-
ANALYSIS NOT
REQUIRED
FUEL ECONOMY
MPG
***
***
*
-
***
-
***
***
*
-
***
*
AVG+TORQ
LB-FT
**
***
**
-
***
*
***
***
-
-
***
-
AVG-TORQ
LB-FT
-
*
-
-
**
-
-
-
-
*
-
-
tn
CODE:
*
**
***
not significant (<90%)
significant at the 90% level
significant at the 95% level
significant at the 99% level
-------
• The least significant differences in data means occur between
tests performed on (1) radial-ply and bias-ply tires on the roll
dynamometer and (2) bias-ply tires on the flat-bed dynamometer with
the indicated hp changed from PR to PRA.
The results of tests of significance on the difference between means
of the emissions data (HC, CO and NO ) are characterized by inconsistencies
exacerbated by insufficient data replication. For example, the test
conditions applicable to data-block pairs 1-2 and 1-11 were identical
except for a small change in indicated hp between data blocks 2 and 11. This
change should have had only a minor impact on the measured data, yet results
in Table 10 show significance levels at 95% and 99% changing to levels of no
significance between data-block pairs 1-2 and 1-11. Clearly the results of
statistical analyses of the emissions data must be treated with caution.
It is apparent also that the C0_, fuel economy and positive torque
results are closely correlated. Differences in the negative-torque means, in
most cases, were not found to be significant.
The results of tests of significance applied to data means obtained
from the series of steady-state tests concerned with fuel economy and
tire consumption are shown in Table 11. As regards fuel economy, the following
conclusions can be drawn:
• The most significant differences in data means occur between
tests performed on (1) bias-belted tires on the flat-bed and on the
roll dynamometers, (2) radial-ply and bias-ply tires on the flat-bed
dynamometer and (3) radial-ply and bias-belted tires on the flat-
bed dynamometer.
All other designated data pairings fail to show any consistent indications of
significances in means. From the velocity standpoint, the least number of
cases of significance was found for the 20-mph velocity and the largest number
at the 30-mph velocity.
Tire power consumption data as determined on a dynamometer were
available only from the roll-dynamometer tests. These data failed to show,
except for two isolated instances out of 15, any significant differences
between the three tire constructions.
57
-------
Table 11
SUMMARY OF THE RESULTS OF TESTS OF SIGNIFICANCE APPLIED
TO THE DIFFERENCES BETWEEN MEANS OF DATA FOR THE
STEADY-STATE RUNS
01
oo
BLOCK
PAIR
5-6
5-13
5-18
6-19
18-19
18-20
19-20
18-19
18-20
19-20
TESTS PER
BLOCK
2
2
2
2
2
/ 2
2
2
2
2
TEST
PARAMETER
Fuel Economy, MPG
Tire Power Consumption,
HP
1
VELOCITY, MPH
20
*
**
_
_
_
_
_
-
-
30
***
***
**
***
_
**
**
-
-
40
**
**
_
***
*
_
_
-
-
50
**
**
_
***
_
_
_
*
**
60
*
**
_
**
_
_
_
-
-
CODE:
NOTE:
*
**
***
= Not significant (<90%)
= Significant at the 90% level
= Significant at the 95% level
= Significant at the 99% level
Test conditions corresponding to each block number are
identified in Table
-------
4.6.4 Comments on the Results
In drawing inferences and conclusions from the numerical data that
have been presented, it is important to recognize that a number of different
factors were involved in the test program that affected the resultant data.
Some of these have been directly or indirectly alluded to in other sections of
the report but they are gathered here for ease of reference.
At the outset of this program, one of the objectives was to
minimize the differences between the roll and flat-bed dynamometers so that
the test data would reflect only the effects of tires operating on rolls or on
a flat belt. Unfortunately, this desirable condition could only be approached
but not realized. The friction horsepower of the flat-bed unit unavoidably was
considerably larger than that of the roll dynamometer. As a result, in setting
the indicated horsepower the horsepower absorbed by the PAU on each unit was
also considerably different. A further complicating factor was the use of an
automatic road load power control unit on the roll dynamometers and a manual
PAU on the flat-bed unit. Because of known hysteresis effects in the latter,
the transient response would be expected to be different. Data from the FTP
tests would be primarily affected by this factor.
Another operational characteristic that affected all the data was
the difference in the true-horsepower/velocity relationship between the two
dynamometers. The flat-bed unit had a higher power absorption level at all
speeds below 50 mph than the roll dynamometer when the settings were equal at
50 mph.
At a much lower level of effect, it is assumed that the use of
several different roll-dynamometer installations contributed some variability
in the test results.
A difference in tailpipe back pressure was found to exist between
the flat-bed test site and the roll dynamometer sites. This difference amount-
ed to only 0.4 inch of water, well within the one-inch tolerance permitted by
the Federal Register regulations. This fact may account for some of the
increased NO levels obtained for the flat-bed tests relative to the roll-
dynamometer £ests. This effect is considered of secondary inportance.
The flat-bed dynamometer was operated in a laboratory space that was
not subject to the same high level of temperature and humidity control that
is exercised in the space occupied by the regular dynamometer test sites.
What effect, if any, this condition might have had on the data cannot be
evaluated.
59
-------
Finally, it must be recognized that the statistical analyses were
limited by the lack of adequate replication in view of the well-known vari-
ability of such data as exhaust emissions, for example. Undoubtedly there are
instances where significance was found and none was warranted and, conversely,
no significance was found where significance did, in fact, exist.
60
-------
REFERENCES
1. *Cassidy, R. J. (Editor), A Feasibility and Preliminary Design Study
of an Advanced Tire Force and Moment Test Facility, Cornell Aero-
nautical Laboratory, Report No. YD-2639-K-IR, November 1969, pp. 269.
2. Schuring, D. J., Energy Loss of Pneumatic Tires Under Freely
Rolling, Braking and Driving Conditions, Tire Science and Tech-
nology, Vol. 4, No. 1, February 1976. pp. 3-15.
3. Gusakov, I., Measuring Rolling Resistance of Tires in the Labora-
tory, Presented at the International Rubber Conference, Kiev
(Russia), October 10-14, 1978, 20 pp.
4. Leiferman, M. W., Performance and Cost Analysis of Chassis
Dynamometers, EPA Technical Report LDTP 76-1, February 1976.
5. Juneja, W. K., D. D. Horchler, and H. M. Haskew, A Treatise on
Exhaust Emission Variability, SAE International Automotive
Engineering Congress and Exposition, February 28 - March 4, 1977,
Paper No. 770136, 24 pp.
6. Handbook of Chemistry and Physics, Thirty-Sixth Edition (1954-
1955), Chemical Rubber Publishing Co., .Cleveland, Ohio, pp. 215-217,
7. Bird, K. D. and J. F. Martin, The Calspan Tire Research Facility:
Design, Development and Initial Test Results, SAE Automotive
Engineering Meeting, May 14-18, 1973, Paper No. 730582.
This reference describes an analogous development, on a larger scale
using an air bearing.
61
-------
APPENDIX A
A PHYSICAL DESCRIPTION OF THE FLAT-BED DYNAMOMETER
INTRODUCTION
The heart of the flat-bed dynamometer consists of a pair of simulated
roadway units (SRU) which take the place of the dual rolls that are located
between the power absorption unit (PAU) and the inertia wheels in a conven-
tional chassis dynamometer. Built under license from Calspan, the two SRU's
were purchased from Akron-Standard with certain mechanical details tailored
to fit the specific needs of this particular flat-bed dynamometer. Design
and fabrication of the dynamometer were undertaken based on the following
initial considerations:
• The PAU and inertia-flywheel components from a Calspan-owned Clayton
CTE-50 dynamometer were to be used.
• Checkout tests had to be performed at Calspan where a dyno pit did
not exist.
• The dynamometer had to fit a standard dyno pit at the GM Proving
Ground at Milford.
• 2250-lb. and 4500-lb. inertia class vehicles were to be accommodated
(specifically, a Chevrolet Chevette and an Oldsmobile 98).
The effect that these constraints exerted on the design of the dynamometer will
be discussed.
First, the flywheel unit available at Calspan was one in which the wheels
were engaged by belts and pulleys. Since this arrangement is subject to slip-
page problems, a direct-coupled flywheel unit was obtained on a loan basis
from the GM Vehicle Emissions Laboratory at Milford (M-VEL), This unit used
five inertia wheels and had a full scale range of 5500-lbs. equivalent weight*.
The PAU did not have the automatic road load control feature desired, but was
used for reasons discussed elsewhere (Section 4.5).
When incorporated in a Clayton chassis dynamometer.
62
-------
To perform dynamometer check tests in the absence of a pit in which to
anchor the individual component parts, it was necessary to fabricate a support-
ing structure upon which the dynamometer could be assembled, aligned and
rigidly attached. In designing this structure, the critical envelope dimen-
sions had to be compatible with the pit dimensions supplied by GM (3 ft. deep
x 8 ft. wide x 15 ft. long). Finally, the two specified vehicles established
the exact equivalent weights that the dynamometer had to provide as well as the
rear-wheel track widths that had to be accommodated.
Simulated Roadway Units
The SRU consists of two 19.25-inch diameter steel drums, 15 inches in
width, which support an endless belt. This belt is fabricated from 301 stain-
less steel and is 0.024 inches in thickness, 15.75 inches in width and 144.55
inches in length. It is covered with a replaceable textured material to sim-
ulate road surfaces. Support for the tire vertical load on the belt is pro-
vided by a linear hydrostatic bearing located at the midpoint of the unsupported
span of the belt. Water is used as the bearing working fluid. Both drums
operate in bearings located in two, 3/4-inch thick, parallel vertical steel
frames. One drum "floats" longitudinally on mechanical flexures to adjust belt
preload through a spring arrangement. A preload of 10,000 psi is used so that
there is no belt/drum slippage during vehicle accel/decel operation. Belt pre-
load is checked using strain-gaged adjusting bolts.
Belt tracking in the lateral direction is augmented by the crowned con-
tact surfaces of the drum (0.0034 inches taper per inch length). The drum
surfaces are circumferentially grooved (21 grooves) to preclude accumulation
of water in the belt/drum interface. Containment of the hydrostatic bearing
fluid is achieved by a seal/scraper assembly preloaded against the underside
of the belt. The assembly is enclosed by an elastomeric boot so variations
in the vertical position of the seal can be accommodated. Figure A-l shows a
photograph of one SRU installed in the dynamometer assembly. The various
features described above are identified by labels.
Dynamometer Layout
A plan view of the dynamometer is shown in schematic form in Figure A-2.
All rotating shafts can be visualized as lying in the plane of the paper except
for the shaft of the inertia unit which is in a plane above, but parallel to,
the plane of the paper. Placement of the inertia unit was dictated by the
dimensions of the standard dyno pit. All of the dynamometer hardware was
mounted on a welded steel-beam supporting base and resulted in an overall
envelope having the following dimensions: 51.5 inches in height, 99.5 inches
in width and 183.5 inches in length.
63
-------
The right-side SRU* and the PAU are attached to a common base plate to
permit adjusting the dynamometer track width (distance between centers of the
water bearings) to that of the Oldsmobile (60.7 inches) and that of the Chevette
(51.2 inches). SRU and PAU shafts are direct coupled. Figure A-2 shows the
configuration used for the Oldsmobile in which the two SRU shafts are joined
by two rubber-type flex couplings with an intervening floating shaft (see
Figure A-l). To accommodate the narrower track of the Chevette, the floating
shaft together with one-half of each flex-coupling is removed and the PAU/SRU
assembly is shifted laterally 9.5 inches permitting the two remaining halves
of the couplings to be mated and form a single unit.
The left SRU is coupled to the inertia unit by a gearbelt/pulley combina-
tion. A 60-tooth gearbelt pulley is mounted on the SRU shaft while a 32-tooth
gearbelt pulley is mounted on a shaft operating in a pair of pillow-block
bearings and is coupled to the inertia unit. Two, three-inch wide gearbelts
operate on these pulleys. A close-up view of the gearbelt drive is shown in
Figure A-3. The smaller diameter pulley is flanged for the purpose of belt
retention.
Inertia Requirements
The equivalent weight provided by the dynamometer inertia at the surface
of the rolls for commercial chassis dynamometers is a calculated quantity based
on roll diameter and the inertia of the rotating components. Since the ro-
tating components are usually regularly-shaped bodies of revolution, the mo-
ments of inertias can be easily calculated if the dimensions and the density of
the materials of construction are known. This same procedure was used in con-
figuring the flat-bed dynamometer to provide exact settings of 2250-lbs. and
4500-lbs. equivalent weight at the tire patch.
The moments of inertia for the PAU and the inertia unit were obtained
from the manufacturer** Moments of inertia were calculated for all rotating
SRU components including drums, belts, shafts pulleys, couplings, etc. Data for
the gearbelt pulleys and the flex-couplings were obtained from the manufacturers.
Using all of the inertia flywheels, it was determined that to obtain an equiv-
alent weight of 4500-lbs. at the SRU belt surface, a step-up ratio of about
1.93 was needed between the SRU's and the inertia unit. Standard gearbelt
pulleys were purchased that achieved a slightly smaller ratio of 1.875. Small,
demountable flywheels were fabricated to be attached to the inertia unit shaft
(see Figure A-2) to achieve exactly 4500-lbs. with the inertia unit selector
switch set to 5500-lbs. and 2250-lbs. with the selector switch set to 2500-lbs.
Despite the significant inertias of the SRU drums, the step-up ratio to the
inertia unit is necessitated because the SRU drums are 19.25 inches in diameter
while the Clayton dynamometer rolls are 8.65 inches in diameter.
From the point of view of the driver of the vehicle.
Telecon: Max Moore, Clayton Manufacturing Company, El Monte, California,
May 2, 1977
64
-------
Water System Description
The hydrostatic bearings were designed to operate at tire contact
pressures up to 100 psi with a water supply pressure of 225 psi and a flow of
110 gpm to each bearing. To meet these requirements, a centrifugal pump with
a capacity of 225 psi and 220 gpm is used. It is driven by a 75 horsepower
motor operating at 3600 rpm on 440 volts A. C. Each hydrostatic-bearing, high-
pressure manifold is supplied by dual lines to reduce pressure losses. The
low-pressure plenum chamber is drained by gravity forces. Large-diameter
plastic piping routes the discharge water to a 65-gallon sump. This sump is
equipped with high-and low-limit electrical switches which control a 7.5 hp sump
pump that has a flow capacity of 300 gpm.
Water from the sump is returned to a 500-gallon reservoir. The reser-
voir is a free-standing, polyethylene cylinder that is open at the top.
This reservoir supplies the high-pressure pump and acts as the dump for the
return flow. Figure A-l shows some of the water lines in the region of a
hydrostatic bearing.
Two particulate filters were installed in the high-pressure water system
to trap contaminants which could obstruct flow through the individual
bearings which are used in a geometrical array to form the composite water
bearing. Reduced flow could result in the steel belt riding on the bearing
faces rather than on the water film. Except for the reservoir, all components
of the water system were mounted on the dynamometer support structure.
At a given pressure, the flow through the hydrostatic bearing varies
inversely with the viscosity of the fluid. The viscosity of water at 130 F
is only one half of that at 68°F. If water temperature is allowed to increase,
the flow will increase and the required pumping power will increase. In-
creased flow will also require active scavenging of the bearing plenum chamber
by a pump. To avoid these difficulties water temperature must be controlled
so as not to exceed Al05°F. Because of the large amount of energy put into the
water in the form of heat by the high-pressure pump, this temperature would be
exceeded in about 20 minutes of continuous dynamometer operation.
A heat exchanger could be used to control water temperature and achieve
a closed water system - (except for makeup water required to replenish leakage
losses). As a temporary; economical alternative, temperature control was
maintained throughout the test program by the simultaneous draining of warm
return water from the sump and the adding of an equal volume of chilled
(«50°F) water at rates of several gallons per minute.
65
-------
BEARING
HIGH-PRESSURE
WATER LINE
SRU DRUM
0 (ADJUSTABLE
SRU BELT WITH "SAFETY
WALK"SURFACE
RETUR
WATER
, LINE
VERTICAL
FRAME
SUPPORT STRUCTURE MEMBER
SRU DRUM (FIXED)
FLOATING SHAFT
FLEX COUPLING
Figure A-1 View of the right half of the flat-bed dynamometer.
-------
SIMULATED ROADWAY
UNIT (SRU)
RIGHT SRUn
WATER
BEARING
ROADWAY
(SRU) BELT
CLAYTON
PAU
NONADJUSTABLE
DRUMS
WATER
BEARING
ROADWAY (SRU) BELT
GEAR BELT (6" WIDTH)
CLAYTON
INERTIA
UNIT
5-WHEEL
(DIRECT
COUPLED)
TRIM
FLYWHEEL'
(NO SCALE)
Figure A-2 Flat-bed dynamometer plan-view arrangement.
67
-------
00
' GEAR BELT
i (TWO 3-INCH
I WIDE BELTS)
60-TOOTH PULLEY
ONSRUSHAFT
COUPLING TO
INERTIA FLYWHEELS
Figure A-2 Close-up view of gearbelt drive arrangement.
-------
APPENDIX B
THE TEST SCHEDULE FOR THE DYNAMOMETERS
The schedule followed for the test vehicles on the flat-bed and the Clayton
(double-roll) chassis dynamometers was explicitly defined in the Scope of Work
section of the contract. This schedule is outlined in Table B-l and provides
for measurements of rear-wheel torque and exhaust emissions for the following
test cycles: (1) the non-evaporataive Federal, Test Procedure (FTP), (2) the
Highway Fuel Economy Test (HFET) and (3) steady-state tests (SS).
Steady-state tests were made at the following velocities and in the se-
quence shown: 60, 50, 40, 30 and 20 mph. Each velocity was maintained con-
stant for a period of ten minutes. Exhaust emissions measured included HC,
CO, NO and CO^. The wheel torques were measured directly as integrated values
extending over the duration of the test. For the FTP and HFET cycles, separate
values of the positive and negative integrated torques were recorded.
A perusal of Table B-l shows that three replicate tests of each of the
aforementioned test cycles were made using radial-ply and bias-ply tires on
the test vehicle on each dynamometer. Only the steady-state tests were performed
on the bias-belted tires. The manufacturer's recommended cold inflation
pressure of 24 psi was used for all tire constructions tested on the flat-bed
dynamometer. For each tire construction one steady-state test was performed
on the flat-bed dynamometer using a cold inflation pressure of 45 psi. All
testing on the Clayton unit was conducted with tire cold inflation pressures
of 45 psi.
Road-load power setting on each dynamometer was established in the follow-
ing manner. The test vehicle was fitted with the radial-ply tires and operated
on the dynamometer for 20 minutes at 50 mph to warm up the unit. With the
vehicle at 50 mph, the power absorption unit was adjusted until the rear wheel
torque was the same as the mean torque value measured during road tests on that
vehicle equipped with the same radial tires. The corresponding reading of dyna-
mometer indicated horsepower is identified as PR in the Table and was used for
tests on all three types of tires.
A repeat set of test runs (substantially so) was made only on the bias-
ply and only on the flat-bed dynamometer with the sole exception that the road-
load power setting was changed (increased). Using rolling-resistance force data
measured for each of the vehicle's two front bias-ply and radial-ply tires,
69
-------
the difference in the total power loss between the two pairs of tires was
calculated. This increment of power loss was added to the previously measured
PR reading and the value, as shown in Table B-l, was identified as PRA.
For reasons given in the body of the report, the above schedule of runs
(except for two flat-bed tests) was completed only for the 4500-lb. inertia
class vehicle. To minimize the possibility of introducing an inadvertent bias
into the test data, a randomized test sequence was followed in performing the
tests scheduled for the flat-bed and the Clayton dynamometers.
70
-------
TABLE B-1. DYNAMOMETER TEST SCHEDULE
FB = FLAT BED, C = CLAYTON
Purpose
Ind. hp
Warm-Up®
Data
1
Warm-Up®
Data
T
Warm-Up®
Data
\
Warm-Up®
Data
1
T
Warm-Up®
Data
|
Warm-Up®
Data
|
Warm-Up @
Data
1
Warm-Up®
Data
|
Warm-Up®
Data
T
Warm-Up®
Data
{
No. of
Runs
1
1
5
1
1
5
1
1
5
5
5
5
1
1
5
1
1
5
1
1
5
1
1
5
1
1
5
1
1
Dyno
Type
FB/C
FB
•
Tire
Type
Radial
-
Radial
\
-
Radial
\
•
Radial
I
Bias Belt
I
T
Bias
T
Bias
I
Bias
Bias
r
Bias
I
Bias
t
Cold
Pressure
0
-
®
I
0
\
•
0
45 psi
-
0
f
45 psi
0
1
®
|
0
45 psi
24 psi
1
24 psi
\
24 psi
I
Test
Cycle
SOmph
50 mph
FTP
HFET
SS
50 mph
FTP
HFET
SS
50 mph
FTP
HFET
SS
50 mph
SS
SS
SS
50 mph
FTP
HFET
SS
SOmph
FTP
HFET
SS
50 mph
FTP
HFET
SS
50 mph
FTP
HFET
SS
50 mph
FTP
HFET
SS
50 mph
FTP
HFET
Dyno
Ind. hp
X
Set PR
PR
1
Set PR
PR
\
Set PR
PR
1
Set PR
PR
1
7
Set PR
PR
\
Set PR
PR
1
Set PR
PR
1
Set PRA®
PRA
I
Set PRA ®
PRA
1
Set PRA (5)
PRA
t
Measurements
Wheel
Torque
®
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Exhaust
Emissions
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
® Tire manufacturer's recommended pressure, MRP, of 24 psi used on the flat-bed dyno, 45 psi
used on Clayton Dyno.
(2) Using test vehicle equipped with radial tires, the dyno indicated hp was adjusted so that the
rear wheel torque was the same as that measured on the road tests at 50 mph. This value of
indicated hp equals PR.
@ Dyno was warmed up for 20 min @ 50 mph using non-test vehicle. At end of warm-up,
indicated hp was set to proper value.
® Same as above except that test vehicle was used.
(5) PRA = PR + (difference in rolling tire power loss between the bias and radial tires on the
front of the test vehicle).
71
-------
APPENDIX C
PROCEDURES FOR PERFORMING THE DYNAMOMETER TESTS
A detailed account is presented in this appendix of the test procedures
that were used in performing the dynamometer tests which are identified in
Table B-l of Appendix B. The procedural details are shown in an itemized
format and, except for minor editorial revisions required to achieve consis-
tency with the rest of the report, are reproduced verbatim as supplied by
General Motors.
The step-by-step instructions and procedures are written in the future
tense since they were prepared prior to the initiation of the test effort
itself. All dynamometer testing took place in accord with the explicit details
outlined below.
A tabulation of selected specifications for the engine, driveline and
emission control system for the two test vehicles is shown at the end of this
Appendix.
72
-------
1.0 Preliminary Vehicle Checks
1.1 Prior to testing, torque rims and the radial tires will be in-
stalled on the drive wheels of the vehicle.
1.2 The test vehicle will be serviced to assure that it is mechanically
sound. Adjustments and repairs will be made, as necessary to
assure that the vehicle is operating at the required tune-up speci-
fication.
1.3 A "slave" canister will be installed on the vehicle. A "slave"
canister eliminates variability in emissions tests due to evapora-
tive fuel losses.
1.4 The vehicle will be ballasted so that with a full tank of fuel and
driver the wheel weights are the same as those measured in the
road tests.
NOTE: Two Total Torque Testers (TTTs) will be used for this
project. One unit will record positive torque and horse-
power, negative torque and horsepower, and total distance.
The second unit will record elapsed time and wheel revolu-
tions.
2.0 Preliminary Instrumentation Calibration
2.1 The TTT used with the torque wheels will be calibrated.
2.2 The Driver's Air recorder on the flat-bed dynamometer will be
calibrated at 50.0 mph.
2.3 The power absorption unit (PAU) and horsepower meter on the flat-
bed dynamometer will be calibrated using the dead weight technique
per the Clayton calibration procedure.
2.4 The critical flow venturi (CFV) sampler, used to collect exhaust
in the flat-bed site,will be verified. Critical flow orifice (CFO)
propane injection tests will be run to insure that the flow cali-
bration is correct and that there are no leaks in the sample system.
2.5 The simulated roadway unit's (SRUs) belt tension on the flat-bed
dynamometer will be set to the proper value by a Calspan repre-
sentative.
73
-------
3.0 Determine PR
The flat-bed and Clayton PAUs must be adjusted so that the vehicle's
drive wheel torque, at 50 mph, is the same on both dynamometers. PR, as
measured in the road tests, is the average drive wheel torque, at 50 mph, for
each vehicle when equipped with radial tires at manufacturer's recommended
pressure (MRP), and operating on a flat road surface. The PAU settings to
achieve PR will be determined by the following procedure.
3.1 Set the drive wheel radial tire pressure to the proper value (45
psi Clayton, MRP flat-bed). Set the required inertia weight.
3.2 Warm up the dynamometer using the test vehicle for 30 minutes at
50 mph.
3.3 Let the vehicle coast to a stop in neutral. Adjust the TTT zero
bias and gain to the proper values.
3.4 Drive the vehicle at 50 mph.
3.5 Adjust the absorber load to obtain a rear wheel torque equivalent
to PR. Record the meter's indicated HP (and also the correspond-
ing actual HP on the Clayton dynamometer).
3.6 Let the vehicle coast to a stop in neutral. Check and record the
TTT zero bias and gain readings.
3.7 Repeat steps 3.1 through 3.6 several times (on each dynamometer) to
insure that PR is correct.
3.8 Repeat steps 3.1 through 3.7 on at least two Clayton dynamometers.
Determine the average PR setting for these dynamometers.
4.0 Frictional Power Absorption Characteristics
4.1 Flatbed Dynamometer
4.1.1 Set the drive wheel radial tire pressures to MRP. Set the
required inertia weight.
4.1.2 Disconnect the PAU from the dynamometer rolls.
4.1.3 Warm up the dynamometer for 30 minutes at 50 mph using the
test vehicle.
4.1.4 Let the vehicle coast to a stop in neutral. Adjust the
TTT zero bias and gain to the proper values.
4.1.5 Drive 2 miles at 20 mph; then 2 miles of data collection
at 20 mph.
74
-------
4.1.6 Drive 2 miles at 30 mph; then 2 miles of data collection
at 30 mph.
4.1.7 Drive 2 miles at 40 mph; then 2 miles of data collection
at 40 mph.
4.1.8 Drive 2 miles at 50 mph; then 2 miles of data collection
at 50 mph.
4.1.9 Drive 2 miles at 60 mph; then 2 miles of data collection
at 60 mph.
4.1.10 Let the vehicle coast to a stop in neutral. Check and
adjust, if necessary, the TTT zero bias and gain to the
proper values.
4.1.11 Repeat steps 4.1.5 through 4.1.9. Start at 60 mph, how-
ever, and proceed to 20 mph.
4.1.12 Calculate the average wheel horsepower and plot as a
function of speed.
NOTE:
HP. R,^
wheel horsepower = -1" x ^-— x 2ir x R
v t. d. int
int int •Lnt
where:
HP = integrated HP (HP-sec) as measured by the TTT
in u
t. = integrated time (sec)
d. = integrated distance travelled (ft)
R. = integrated revolutions, vehicle wheel
j.n L>
= rolling radius used in TTT electronics =
1.1484 (ft)
4.1.13 Calculate the average horsepower consumption for the two rear
tires, at each speed, from the TIRF rolling resistance data.
NOTE: (FLR + FRR) x V
average tire HP (2 tires) =
where: F = the average rolling resistance force for the left
L rear radial type test tire (Ib)
F = the average rolling resistance force for the right
RR rear radial type test tire (Ib)
V = average velocity (mph)
75
-------
4.1.14 Calculate dynamometer friction horsepower at each speed
(wheel HP - rear tire HP). Plot friction horsepower as a
function of speed.
NOTE: This technique assumes the tire rolling resistance
on the flat-bed dynamometer is the same as that on
the TIRF tire test facility as both are flat-bed
machines.
4.2 Clayton Dynamometer
4.2.1 Disconnect the PAU from the dynamometer rolls. Set the
required inertia weight.
4.2.2 Warm up the dynamometer for 20 to 30 minutes at 50 mph.
4.2.3 Drive the dynamometer to 65 mph and raise the vehicle off
the rolls. Allow the front and rear rolls to coast down
to at least 15 mph.
4.2.4 Calculate the friction horsepower from the coast down times
at 60, 50, 40, 30, and 20 mph.
NOTE: friction HP = (front roll friction) + (rear roll
friction)
6/"\ T T 1 /~\ "™ ^ /*i T ™ i r " *\ 1 I i W J. ™" .L O O J .LOO I
.073 x 10 (V1 - V9 ) —s — + T—
-L £• ' I" ^^ I
where: friction HP = horsepower at velocity of (V1 + V9)/2
V. = initial velocity (mph)
V_ = final velocity (mph)
IWT = dynamometer inertia weight setting (Ib)
155 = dynamometer rear roll inertia (Ib)
tf = front roll coast down time (sec)
t = rear roll coast down time (sec)
4.2.5 Repeat steps 4.2.3 and 4.2.4 at least once. Determine the
average friction horsepower at each speed.
4.2.6 Repeat steps 4.2.1 through 4.2.5 on at least two Clayton
dynamometers. Calculate the average friction horsepower
at each speed for three dynamometers.
76
-------
5.0 True Dynamometer Absorbed Horsepower (Clayton only)
The true absorbed horsepower vs speed curve on each Clayton dynamometer
can be determined directly from coast down times with the PAU indicated horse-
power adjusted to achieve PR. An alternate method is to calculate the average
true load by the method shown below.
Avg. true absorbed HPy = avg. windage HP + friction HP
= KV3 + friction HPy
where: V = roll velocity (mph)
friction HP = the dynamometer friction HP for the appropriate
v
inertia setting (4.2)
K = (avg. actual HP @ PR) - (avg. front roll friction HP @ 50 mph)
(50.0)3 (mph)3
6.0 FTP, HFET, and Steady Speed Tests
6.1 Test Sequence
NOTE: The test design calls for tests on both dynamometers with three
tire types. In general, a test day will begin with a cold start
FTP test followed by a HFET and steady speed tests. Tests between
the two dynamometer types will be randomized.
6.1.1 Perform the required tests for the radial tires, on both
dynamometers, in a random order.
6.1.2 Perform the tests for the bias-belted tires, on both dynamome-
ters, in a random order.
6.1.3 Perform the tests for the bias-ply tires, on both dynamometers,
in a random order.
NOTE: Steps 6.2 to 6.9 describe in detail how each test will be
performed.
6.2 Vehicle Preparation
6.2.1 Prior to each day's testing, the test vehicle will be
filled with room temperature Indolene clear fuel.
6.2.2 If the vehicle is scheduled for a cold start FTP test, it must
have a dynamometer prep within the previous 12 to 36 hours. A
dynamometer prep (for this project) will be considered a 23
minute LA-4 test (Bag 1 and 2) and/or an HFET or series of
steady speed tests.
77
-------
6.3 Dynamometer Preparation
6.3.1 Select the required dynamometer inertia.
6.3.2 The dynamometer will be warmed-up for 20 to 30 minutes
at 50 mph (for the FTP tests, a non-test vehicle is used).
6.3.3 Adjust the dynamometer indicated HP to achieve PR or PRA,
as required.
NOTE: The flat-bed dynamometer must be set to the proper
indicated HP within 1 hour prior to the start of
either a FTP, HFET, or steady speed test. The
Clayton dyno will be operated in the automatic road-
load mode and therefore will be warmed-up within
2 hours of the start of the test.
6.4 CVS/CFV Preparation
6.4.1 CVS (Constant Volume Sampler)
6.4.1.1 Check that the CVS pump, heater and temperature
controller are on and that the pump inlet tempe-
o o
rature is statilized at 100 F ± 5 F.
6.4.1.2 Operate the automatic sample bag purge and
evacuate sequence.
6.4.1.3 Replace the flip-top filters.
6.4.1.4 Turn on the sample pumps. Set the flow to the
proper values.
6.4.2 CFV (Critical Flow Venturi Sampler)
6.4.2.1 Manually fill and evacuate the bags with dilu-
tion air. Repeat at least 2 times.
6.4.2.2 Replace the flip-top filter.
6.4.2.3 Enter the corrected barometric pressure into the
CFV using the thumbwheel switches.
6.4.2.4 Turn on the ambient and sample pumps.
NOTE: The CFV electronic control module will be
left on 24 hours a day to prevent drift.
6.5 Pre-Test Vehicle Preparation
6.5.1 Move the test vehicle onto the dynamometer (do not start
78
-------
the engine if an FTP test is to be performed).
6.5.2 Turn on the cooling fan.
6.5.3 Set the wheel chocks. Install the vehicle safety re-
straints.
6.5.4 Open the vehicle hood.
6.5.5 Check to see that the air conditioning is off.
6.5.6 Check the Driver's Aid recorder to be sure the power is
on, the pens are operational, and the chart speed is
6"/minute.
NOTE: On the flatbed dynamometer site, preprinted
driver's traces are used.
6.5.7 Check that the TTTs are on and operating properly.
NOTE: The TTTs will be left on 24 hours a day to minimize
drift.
6.5.8 Connect the tail pipe to the sampler hose immediately
prior to starting the test (cold start FTP).
6.5.9 Make sure the tire pressure is at the proper value.
6.6 FTP Test Dynamometer Procedure
6.6.1 Record the barometric pressure and the wet and dry bulb
temperatures.
6.6.2 Set the TTT zero bias gain to the proper values.
6.6.3 Simultaneously start the engine, the TTTs, and the CVS
or CFV sampler.
6.6.4 When the engine starts, turn on the Driver's Aid recorder,
and drive the test schedule.
NOTE: False starts and engine stalls will be handled
per M-VEL procedures.
6.6.5 At the end of phase I record the TTT data and the CVS/CFV
corrected flow.
6.6.6 At the and of phase II (2 seconds after the end of the
last deceleration) turn off the engine and stop the TTTs.
Five seconds after the engine stops, stop the CVS/CFV
sampler. Start the 9-11 minute soak period.
79
-------
6.6.7 Immediately after the end of phase II, turn off the
cooling fan, disconnect the hose from the tail pipe, and
close the hood.
6.6.8 Record the TTT and CVS/CFV sampler data.
6.6.9 Record the TTT zero bias and gain readings, and reset
the displays.
6.6.10 Immediately before the start of phase III reconnect the
tail pipe hose, turn on the fan, and open the vehicle hood.
6.6.11 Simultaneously start the engine, the TTTs and the CVS/CFV
sampler.
6.6.12 When the engine starts, turn on the Driver's Aid recorder,
and drive the test schedule.
NOTE: False starts and engine stalls will be handled per
M-VEL* procedures.
6.6.13 At the end of phase III, turn off the engine, the TTTs and
the CVS/CFV sampler.
6.6.14 Record the TTT and CFS/CFV sampler data.
6.6.15 Start engine, drive the vehicle at 50 mph and record the
dynamometer indicated horsepower.
6.6.16 Let the vehicle coast to a stop in neutral. Check and
record the TTT zero bias and gain readings.
6.7 HFET Dynamometer Procedure
6.7.1 Record the barometric pressure and the wet and dry bulb
temperature.
6.7.2 Immediately prior to the HFET pre-conditioning cycle,
drive the test vehicle at 50 mph for a minimum of 1 minute.
6.7.3 Adjust the indicated horsepower to achieve PR or PRA, as
required. Record the indicated horsepower.
6.7.4 Let the vehicle coast to a stop in neutral. Adjust the
TTT zero bias and gain, if required, to the proper values.
Record these values.
Milford-Vehicle Emissions Laboratory
80
-------
6.7.5 Drive the HFET pre-conditioning cycle.
6.7.6 Simultaneously start the HFET data cycle, the TTTs, and
the CVS/CFV sampler.
6.7.7 At the end of the HFET data cycle, immediately turn off
the engine and stop both the CVS/CFV sampler and the TTTs.
6.7.8 Record the TTT and CVS/CFV sampler data.
6.7.9 Start engine, drive the vehicle at 50 mph and record the
indicated horsepower.
6.7.10 Let the vehicle coast to a stop in neutral. Check and
record the TTT zero bias and gain reading.
6.8 Steady Speed Test Dynamometer Procedure
6.8.1 Record the barometric pressure and the wet and dry bulb
temperatures.
6.8.2 Drive the vehicle at 50 mph for a minimum of one minute.
Adjust the indicated horsepower to achieve PR or PRA, as
required. Record the indicated horsepower.
6.8.3 Let the vehicle coast to a stop in neutral. Check and
adjust, if necessary, the TTT zero bias and gain to the
proper values. Record these values.
6.8.4 Drive the vehicle at the required steady speed.
6.8.5 After the vehicle speed is stable, simultaneously start
the TTTs and the CVS/CFV sampler.
6.8.6 At the end of 600 seconds, simultaneously stop the TTTs
and the CVS/CFV sampler.
6.8.7 Record the TTT and CVS/CFV sampler data.
6.8.8 Drive the vehicle at 50 mph and record the indicated
horsepower.
NOTE: Steps 6.8.2 through 6.8.8 will be repeated for each
"back-to-back" steady speed test. The steady speed
tests will be conducted at 60, 50, 40, 30 and 20
mph in named order.
6.9 FTP, HFET and Steady Speed Test Exhaust Emissions Analysis
NOTE: All bag analyses will be made on emissions analysis consoles
that meet M-VEL calibration requirements for certification
tests.
81
-------
6.9.1 Zero and span the proper analyzer ranges for bag sample
analysis.
6.9.2 Turn on the bench sample pumps. Analyze the ambient bag
HC, CO, NO and C02. Allow the analyzers to stabilize
before taking a computer reading.
6.9.3 Analyze the sample bag HC, CO, N0x and CO . After the
analyzers have stabilized, take a computer reading.
6.9.4 Conduct a zero and span check after completion of the bag
analyses. If this check shows a change of 1% of full
scale or greater, repeat steps 6.9.1 through 6.9.4.
7.0 Data Quality Assurance Checks
7.1 Clayton Dynamometer
7.1.1 All tests will be made on sites that meet M-VEL's cali-
bration requirements.
7.2 Flatbed Dynamometer
7.2.1 The PAU indicated horsepower meter will be checked
periodically using the Clayton calibration procedure.
7.2.2 The Driver's Aid speed trace will be checked and cali-
brated periodically.
7.2.3 CFO propane injections will be used to check CFV cali-
bration accuracy and sample system integrity.
7.3 Total Torque Tester
7.3.1 The TTT will be re-calibrated each time the tires and
torque wheels are changed.
7.4 Repetitive Tests
7.4.1 The number of tests needed for this project has been defined
in Table B-l of Appendix B. Quality control charts will be
kept on the torque and emissions data. Tests that are in-
validated for equipment and/or procedure problems will
be repeated.
7.4.2 As stated previously, tests between the Clayton and flat-
bed dynamometers will be randomized. Randomized tests
minimize systematic errors in repetitive testing.
82
-------
SELECTED VEHICLE DATA
Year/Body Style
Engine
Carburetor
Transmission
Axle Ratio
Emission Controls
Oldsmobile 98
1977/2 door
403 CID (V-8)
4 bbl. Model M4MC
CBC-350
2.41
• integral backpressure EGR
• oxidizing catalytic
converter (260 in )
• standard GM evaporttive
control
Chevrolet
Chevette
1977/2 door
1.6 litre (4 cyl.)
1 bbl. Model 1ME
THM-200
3.07
• ported vacuum EGR
• oxidizing catalytic
converter (160 in )
• 4 cyl. pulse air
injection
• standard GM evaporative
control
83
-------
APPENDIX D
A DESCRIPTION OF THE SPECIAL
INSTRUMENTATION USED FOR THE VEHICLE TESTS*
The purpose of this appendix is to describe, in somewhat more detail
than was possible in the text of the report, some of the special instruments
employed in the road tests and dynamometer tests performed on the vehicles.
The drive axle on each test vehicle was equipped with a pair of torque
wheels developed at the Milford Proving Ground. Both vehicles on this program
used the normal sensitivity torque wheels (Mod. 10W) wiich provide a range of
-2000 ft-lb with an accuracy of 1%. Strain gages are employed as the basic
sensors. Both sets of torque wheels were made specially for the program in order
to preserve the normal rear-wheel track of the vehicle. This requirement
was a function of the design of the flat-bed dynamometer which matched the
rear-wheel tracks of the Oldsmobile and Chevette vehicles within a small to-
lerable margin. Each torque wheel was fitted with a closed slip-ring assem-
bly (Mod. 2) which is resistant to dust and moderate amounts of moisture and
thus is useful for outdoor testing in dry conditions. The slip-ring assembly
carries power to the strain-gage bridges and permits picking-off the output
electrical signals.
Integral with the mechanical package containing the slip rings is an
optical encoder that provides a measurement of wheel revolutions. This trans-
ducer provides a high degree of resolution in this measurement since it utili-
zes a digital output consisting of 160 electrical pulses per each revolution.
Figure D-l shows the torque wheel and its associated instrumentation
at the right-rear wheel position of the Oldsmobile vehicle which is resting on
the flat-bed dynamometer.
Distance-travelled data, whether in road testing or dynamometer
testing, were measured using a one-foot circumference wheel consisting of a
metal rim and a solid-rubber periphery (tire). This wheel was mounted in a
portable, detachable frame and the wheel shaft was coupled to an optical encoder
of the same type as used with the torque wheels. Since the circumference of the
wheel was known precisely, a measurement of total number of wheel revolutions
was directly relatable to the total distance travelled. In road tests, the
vehicle towed a bicycle-type fifth wheel and the one-foot circumference wheel
The information contained in this Appendix is based on data and descriptive
material supplied by General Motors.
84
-------
operated against it. In dynamometer tests, this small wheel operated directly
against the rear cradle roll of the Clayton unit or the SRU belt in the case of
the flat-bed unit. Figure D-2 is a photograph that shows the one-foot circum-
ference wheel as installed on the left-side SRU of the flat-bed dynamometer.
Power input and signal conditioning for the torque wheels and the
optical encoders were supplied by the Total Torque Tester (TTT). The Total
Torque Tester is a portable electronic instrument specifically designed by the
GM Proving Ground Noise and Vibration Laboratory for use with torque wheels and
optical encoders. This device, operating in conjunction with the torque wheels
and a fifth wheel, functions as a system capable of accumulating vehicle per-
formance data as transmitted or received at the rear wheel interface. The
instrument display is digital in units of work (horsepower seconds), torque x
time (foot pound seconds) and distance (feet). As the torque at the rear wheel
is bidirectional, the tester separates the data with respect to the zero base-
line and provides two separate displays; positive foot pound seconds and nega-
tive foot pound seconds. Since horsepower is the product of rotational speed
and torque, separate displays are also provided of positive and negative horse-
power seconds. In addition, the unit has the capability to make instantaneous
measurements of positive and negative foot pounds and feet per second. This
option can be exercised by the operator.
Two Total Torque Testers were used for this program. The one was a
standard unit and was employed to collect data on wheel torque and distance
travelled. The second, a modified unit, collected data on torque-wheel re-
volutions (using the optical encoder) and elapsed time. These data were used
to calculate wheel horsepower based on wheel torque, wheel speed and effective
tire rolling radius.
85
-------
OPTICAL
ENCODER
(WHEEL REVOLUTIONS)
1 FT
CIRCUMFERENCE
WHEEL
Figure D-1 Torque wheel and instrumentation; 1977 Oldsmobile.
-------
00
OPTICAL
ENCODER
(DISTANCE
TRAVELLED)
1 FT
,' CIRCUMFERENCE
WHEEL
Figure D 2 Distance measuring instrumentation; one-foot circumference wheel
mounted on the flat-bed dynamometer
-------
APPENDIX E
THE CALSPAN TIRE RESEARCH FACILITY (TIRF)
A photograph of the TIRF facility is shown as Figure E-l; a dimensional
view of the facility is shown in Figure E-2. The primary features of the machine
are:*
Tire Positioning System
The tire, wheel, force sensing balance and hydraulic motor to drive
or brake the tire are mounted in the movable upper head. The head provides
steer, camber and vertical motions to the tire. These motions (as well
as vertical loading) are servocontrolled and programmable for maximizing test
efficiency. The ranges of the position variables, the rates at which they may
be adjusted and other information are shown in Table E-l.
Roadway
The 28-inch wide roadway is made up of a stainless steel belt covered
with material that simulates the frictional properties of actual road surfaces.
The belt is maintained flat to within 1 to 2 mils under the tire patch by the
restraint provided by an air bearing pad which is beneath the belt in the
tire patch region. The roadway is driven by one of the two 67-inch diameter
drums over which it runs. The road speed is servocontrolled; it may be programmed
to be constant or varied.
The surfaces usually used are "Safety Walk"**. These surfaces have
excellent microtexture giving a wet skid number*** of about 60 in the untreated
condition. The surfaces are honed to reduce the wet skid number to lower
values (typically surfaces of skid number 50 and 30 are used).
A unique feature of TIRF is the ability to carry out tests under
wet road conditions. A two-dimensional water nozzle spans the roadway. This
nozzle has an adjustable throat which can be set to the desired water depth.
*A more complete description of this facility will be found in Reference 7.
** Manufactured by the 3M Company
*** At 40 mph and 0.020-inch water depth using the ASTME-501 Standard Pavement
Traction Tire
88
-------
The flow through the nozzle is then varied by controlling the water pressure.
At each test condition the water film is laid on tangential to the belt at belt
velocity. The film thickness may be varied from as low as 0.005 inches up to
0.4 inches.
Tire-Wheel Drive
A drive system which is independent of the roadway drive is attached
to the tire-wheel shaft. This separate drive allows full variation of tire
slip both in the braking and driving modes. The tire slip ratio, referenced to
road speed, is under servocontrol.
Balance System*
A six-component strain gage balance surrounds the wheel drive shaft.
Three orthogonal forces and three corresponding moments are measured through
this system. A fourth moment, torque, is sensed by a torque link in the wheel
drive shaft. The load ranges of the basic passenger car and truck tire
balances are shown in Table E-2. Transfer of forces and moments from the balance
axis-system to the conventional SAE location at the tire-roadway interface is
in the data reduction computer program.
System Operation
Data Acquisition Program (DAP) Control
The data acquisition program (DAP) is a software system which controls
machine operation and logs data during tests. DAP controls test operations by
means of discrete setpoints which are generated in the computer by the program.
These setpoints are sent to the machine servos which respond and establish tire
test conditions. After the setpoints are sent to the servos, a delay time is
provided which starts after the machine variables have reached a steady state
value within predetermined tolerances. This allows the system to stabilize
before data are taken. After data are taken, the next set of test conditions
is established and testing continues.
One or two variables can be changed during DAP testing. The other
test parameters are kept fixed throughout the test. Up to twenty data points
can be used for each variable in a run.
More detailed information on the balance systems and their calibration
may be found in Reference 7.
89
-------
A data reduction program is used to operate on the raw data collected
during testing. These new data are reduced to forces and moments in the proper
axis system and all variables are scaled to produce quantities with engineering
units. Raw and reduced data are temporarily stored in a disc file. Both re-
duced and raw data can be transferred to magentic tape and maintained as a
permanent record.
Reduced data points can be listed, plotted and curves can be fitted to
the points. All of the standard Calspan plots can be generated from DAP test
data.
Data lists and plots are displayed on the screen of a CRT console.
Hard copies of this information can be made off this display.
Continuous Sampling Program (CSP) Control
The continuous sampling program (CSP) is a software system which
controls machine operation and continuously logs data during tests. Test
variables can be constant or changed at rapid rates. One or all variables can
be changed during a test. Data can be sampled at rates up to 100 samples per
second. Pauses are used so that data can be logged during desired intervals
of the test.
CSP testing can be conducted quickly which in turn reduces tire wear
during severe tests. The high rate of data sampling also permits limited
dynamic measurements to be made.
Two-parameter plots of data can be made. Carpet and family plots
of test data cannot be made with this program at the present time. CSP data
will also reflect time effects if tire characteristics are a function of the
rate of change of testing variables.
Data reduction is accomplished in a manner similar to that employed
in DAP testing.
90
-------
Figure E-1 CALSPAN TIRE RESEARCH FACILITY (TIRF)
91
-------
1*1
t» ••qW.J^UMX
a^-.^-gpj
II ' ,1 I
!
I
=K'
Figure E-2 TIRE RESEARCH MACHINE
-------
TABLE E-l
TIRF CAPABILITIES
CHARACTERISTIC
TIRE SLIP ANGLE (a)
TIRE INCLINATION ANGLE (y)
TIRE SLIP ANGLE RATE (a)
TIRE INCLINATION ANGLE RATE (y)
TIRE LOAD RATE (TYPICAL)
TIRE VERTICAL POSITIONING RATE
ROAD SPEED (V)
TIRE OUTSIDE DIAMETER
TIRE TREAD WIDTH
BELT WIDTH
RANGE
±30°**
-30°***
10°/sec
-rO ,
7 /sec
2000 Ib/sec
2"/sec
0-200 mph
Up to 46"
24" MAX.
28"
TABLE E-2
BALANCE SYSTEM CAPABILITY
COMPONENT
TIRE LOAD
TIRE TRACTIVE FORCE
TIRE SIDE FORCE
TIRE SELF ALIGNING TORQUE
TIRE OVERTURNING MOMENT
TIRE ROLLING RESISTANCE MOMENT
PASSENGER CAR
TIRE BALANCE
4,000 Ib
-4,000 Ib
-4,000 Ib
-500 Ib ft
TRUCK TIRE
BALANCE
12,000 Ib
-9,000 Ib
is,ooo ib
-1,000 Ib ft
-1,000 Ib ft -2,000 Ib ft
-200 Ib ft
-400 Ib ft
**
***
Can be increased to 90 with special setup,
Can be increased to 60° with special setup,
93
------- |