EPA/ Industry Dynamometer Comparison Study
              Nine Vehicle Fleet
                Submitted to:
   Dynamometer Comparison Study Task Force
                  April 1995
               Martin Reineman
                Richard Nash
  United States Environmental Protection Agency
  National Vehicle and Fuel Emissions Laboratory
             2565 Plymouth Road
          Ann Arbor, Michigan 48105

-------
Abstract

      Between October 1993 and September 1994 a test program was
conducted at the EPA National Vehicle and Fuel Emissions Laboratory (NVFEL)
to evaluate emission and fuel economy differences between tests conducted on a
large single roll electric and a twin small roll hydrokinetic chassis dynamometers.
The principal objective of the program was to compare emissions and fuel
economy results from a twin 8.65 in. roll chassis dynamometer, adjusted per
current EPA practice, to results obtained with the 48 in. electric dynamometer,
which was adjusted to more closely duplicate actual on-road forces over a wide
speed range (70 to 10 mph/hr).  Because of concerns about how to phase-in a
change in dynamometer design, a secondary objective of the study was to
assess the accuracy of a dynamometer manufacturer's attempt to simulate twin
roll hydrokinetic dynamometer characteristics on the electric dynamometer such
that emission and fuel economy results would be similar between the two
dynamometer designs. This report serves as a summary of the results.  It is not
intended to explain reasons for the observed differences or to address the
various issues that may surround a change in dynamometer design for emission
and fuel economy testing.
Background

      This test program is one step in a series of events which occurred since
the late 1980s to investigate the need for an improved method to simulate vehicle
road operation using a chassis dynamometer. The 1990 Amendments to the
Clean Air Act directed EPA to make changes to its current test practices, if
appropriate, to be more representative of actual road operation. As a result,
several cooperative studies were conducted between EPA and the American
Automobile Manufacturers Association (AAMA) to determine whether a twin or
single roll chassis dynamometer would best achieve the new test objectives.

      Through mutual agreement with the AAMA members, and with input from
other vehicle manufacturers, equipment suppliers, and other interested parties,
EPA adopted-a 48 in. single roll electric dynamometer design for Cold CO
Emissions Regulations and has proposed that this design be adopted for future
emission and fuel economy compliance testing.

      The results presented in this study follow from an EPA October 1992
public workshop attended by vehicle manufacturers, equipment suppliers, and
other interested parties.  It was agreed at the October meeting to form an
EPA/Industry task force to design a test program to compare the results of
changes in dynamometer designs and associated changes in test procedures.

      The Task Force consisting of representatives of EPA's Engineering
Operations Division and Certification Division met several times with vehicle
manufacturers and other interested parties to develop the test program design.

-------
The basic test program consisted of exhaust emissions and fuel economy
comparisons among an EPA 8.65 in. twin roll hydrokinetic chassis dynamometer,
and two load setting methods using the EPA 48 in. single roll electric chassis
dynamometer. Efforts were made to assure the representativeness of the two
dynamometers, and to reduce the number of uncontrolled test variables.  The
number of tests per vehicle was first estimated using assumptions for typical
emission and fuel economy variability and standard sample size design theory.
In practice, the actual number of tests were based on an examination of an
individual vehicle's emission and fuel economy variability, engineering judgment,
and practical testing resource considerations. The basic test sequence consisted
of a cold start FTP followed by a HFET test and a vehicle/dynamometer coast
down.
Summary

      Tables 1-A, 2-A, 1-B, and 2-B summarize the program results.  Tables 1-A
and 2-A present emission and fuel economy changes obtained using the large
roll electric dynamometer relative to the current twin small roll hydrokinetic
dynamometer. Table 1-A presents normalized results expressed in percent
difference. Table 2-A presents absolute differences expressed in units of g/mi or
mi/gal. These data suggest the changes in loading associated with the electric
dynamometer were vehicle specific.  In general, the data showed that testing a
vehicle on the large single roll electric dynamometer resulted in higher exhaust
emissions, similar FTP fuel economy, and lower HFET fuel economy relative to
the twin small roll hydrokinetic dynamometer.

      Tables 1-B and 2-B summarize results of the secondary objective of the
program, the attempt to modify the electric dynamometer load such that it would
mimic the performance of the hydrokinetic dynamometer. These tables show the
differences of the electric dynamometer relative to the hydrokinetic
dynamometer, expressed as percent in Table 1-B, and as absolute differences in
Table 2-B.  These differences were vehicle specific, but in general still showed
slightly higher emissions  on the large roll electric, and higher FTP fuel economy
relative to the twin small roll dynamometer. The Honda  vehicle, the only manual
transmission vehicle in the vehicle fleet, was unable to be tested on the electric
dynamometer due to instability in the dynamometer's control circuitry when
operated in the simulation mode. In addition to the uncertainty whether accuracy
could be improved, the issue associated with  modeling each vehicle's twin roll
behavior using generic tire/roll slip characteristics was determined to be too large
to warrant additional development.
Test Design

      Test Site Description - Two dynamometers were used: D006, a Clayton
Model ECE-50 with twin rolls, 78 in. long and a nominal 8.65 in. diameter, with
8875 Ib inertia simulation capability.  Dynamometer D005, immediately adjacent
to D006, is a Horiba Model LDV-48-86-125HP-AC electric chassis dynamometer,

-------
 with 24 in. wide single rolls with a nominal diameter of 48 in. and 6000 Ib inertia
 capability.

      Vehicle restraint on the twin roll dynamometer consisted of the EPA
 standard non-drive wheel chocks, cable restraint, and cross straps for front wheel
 drive vehicles on Dynamometer D006. The restraint system for D005 consisted of
 the GM designed front and rear bumper straps, with cross straps for front wheel
 drive vehicles. The EPA dynamometer video drivers' aids (VDA) were identical,
 as were the nominal 5300 cfm fixed speed cooling fans and cell supply air flow.

      Exhaust gas measurements were made using a shared analyzer site,
 A003, and a common, moveable, standard CVS, No. 29C, with nominal 350 cfm
 flow capacity.  With exception of the 48 in. dynamometer, and the 8875 Ib inertia
 capability of the 8.65 in. dynamometer (versus 6875 Ib inertia capability for the
 most common test site at NVFEL), all equipment used for this study are fully
 representative of an EPA standard test cell used for compliance testing.

      Test Vehicles - The nine vehicle test fleet was selected by the Task Force
and is described in Attachment A. These vehicles represent a compromise of a
desire to  select vehicles with a wide range of drive wheel configuration, road
load, axle and inertia weights, and the desire to select vehicles representative of
the current in-use fleet.

      Test Fuel - Certification test quality 96 RON test fuel was used for all
emission tests. Test fuel analysis reports were provided to task force members
when requested.

      Drivers - One driver operated a particular vehicle for all tests on both
dynamometers. Driver training on the electric dynamometer for the four drivers
used in the program consisted of practice tests to become familiar with the
vehicle response when using the electric dynamometer.

      Driving excursions from the FTP and HFET cycles were monitored using
standard  FTP testing requirements as summarized by EPA's VDA summary,
which is published with every emission and fuel economy test.

      Measurements - Exhaust emission measurements were reported for THC,
CH4, NMHC, CO, NOx,  and COg. Carbon balance fuel economy was calculated
using standard procedures and measured test fuel properties.  A test report
summarizing these results was made available to the interested parties after the
data were inspected using normal EPA quality control guidelines.

      The nine test vehicles were equipped with instrumentation for measuring a
 maximum of 16 channels of dynamometer and vehicle parameters. All vehicles
were equipped with a volumetric fuel measuring system, and additional
 instrumentation at the vehicle sponsor's option. On several vehicles, this
 equipment included wheel torque measuring systems, and instrumentation for
 measuring wheel speed, engine speed, throttle position, and other vehicle
 specific parameters including multiple vehicle temperatures.

-------
      Dynamometer data collected for most, but not all tests, included the
 hydrokinetic dynamometer front and rear roll speeds, hydrokinetic dynamometer
 load cell force, electric dynamometer roll speed, and electric dynamometer load
 cell force.

      The vehicle/dynamometer parameters were measured and recorded at 10
 Hz frequency by using Macintosh based LabView hardware and software.

      Vehicle Preparation - Vehicles were supplied with fuel tank drains,
 standard 2.5 in. ID exhaust connectors, and in some cases, a slave canister
 system to avoid evaporative emission influence on the emission test.

      Test Sequence - The basic test sequence for a typical vehicle using either
 dynamometer was as follows:

      Day 1   Drain fuel, fill to 40% of tank volume with pre-conditioning fuel.
              Prep LA-4
              Purge slave evaporative canister if required
             Soak 12-24 hours

      Day 2   Push on dynamometer - no heat build
              FTP
              HFET (warm-up and sample)
             Coast downs -  Hydrokinetic 55-45, Electric 70-10
             Add test fuel to makeup for daily consumption (vehicle specific)
              Purge slave canister if required
              Soak 12-24 hours

      The typical test sequence began Day 1 activities on Monday, and
 repeated Day 2 events Tuesday through Friday. Emission, fuel economy, and
coast downs on Tuesday through Thursday served as pre-conditioning for the
following test day.  Day 1 pre-conditioning for all test vehicles was performed on
each Monday, or at the outset of the test program, or if there was a disruption in
the week long test series for any given vehicle.

      The dynamometer test  location was alternated between  individual tests on
 D005 and D006 to cancel vehicle emission and fuel economy changes with time.
Assignment of the vehicle/dynamometer test configuration and  the daily testing
order was based on site availability and the schedules of the personnel from the
vehicle suppliers who often monitored their independent data acquisition systems
 or chose to witness the daily tests.
Track Data

      Road Coast Down Data - Coast down tests on the nine vehicles were run
at the Ford Motor Company Michigan Proving Grounds using instrumentation
described in a draft SAE procedure, J1743. New tires were installed prior to the

-------
track coast downs and stabilized with 1000 miles of track mileage accumulation.
Coast down tests where run using a 75 to 5 mph velocity range and nominal
ambient conditions of zero wind speed, 68 +/-10 F, and 29.0 +/- 2.0 in. Hg.  It
was the opinion of the Task Force that the tight range of ambient coast down
temperature eliminated the need for an ambient temperature correction The
single roll three term road load force equation was developed using methodology
of draft SAE procedure J1743 and the 70 to 10 mph data, while the two term road
load force equation and resulting 55-45 coast down time was derived using the
methodology of EPA Advisory Circular No. 55C and the 60 to 20 mph data.
Track and dynamometer load settings for the nine vehicle fleet are presented in
Attachment C.  None of the load settings for the single or twin roll dynamometers
included an increase in loading for air conditioner simulation.
Dynamometer Adjustments

      Hvdrokinetic - Load setting on the Clayton dynamometer was based on
the EPA guidelines published in Advisory Circular No. 55C. Given the "target"
55-45 mph dynamometer coast down time, and Attachment B (an EPA procedure
for adjusting loading on a twin roll dynamometer), a 50 mph actual power
absorption setting was determined for each test vehicle. The 50 mph
horsepower value for each vehicle was a two or three test average determined as
a function of the repeatability of the procedure for the individual vehicle.

      Dynamometer tire pressure for passenger cars and trucks was set to 45
psi.

      Electric (True Road Load Simulation) - Standard Horiba dynamometer
software was used to derive the dynamometer coefficients for each test vehicle
from the track coefficients. This technique is described in the Horiba
Dynamometer Operation Manual, under the section Coast  Down - Derivation of
dyno-setting road load parameters. Protocol for the load setting included: the
dynamometer inertia was defined as the equivalent test weight (ETW) multiplied
by 1.015, the target and actual force curves were produced by coasting down the
vehicle from 70 to 10 mph with 5 mph speed intervals, the actual and target force
curves were matched to within two pounds, and two verification runs were
required to obtain final dynamometer coefficients.

      Dynamometer tire pressure was set at the same pressure used for track
coast downs, which was based on vehicle manufacturers' recommended inflation
pressure.

      Electric (Hydrokinetic Load Simulation) - Tests were performed using an
electric dynamometer that was modified by Horiba Instruments to produce
loading comparable to the hydrokinetic dynamometer. Modifications included
hardware and software (version 1.34C TR) changes. Horiba experimented with
two versions of the twin roll simulation during the program.  The first version of
the twin roll simulation was judged to be unsatisfactory after the first three
vehicles were tested due to poor correlation between the twin roll and the twin roll

-------
 simulation results, and was replaced by a second version of the simulation, which
 was used to test eight of the nine vehicle fleet.  A description of the second twin
 roll dynamometer simulation was not published.

      The electric dynamometer coefficients for the twin roll simulation tests
 were derived from a three run average of consecutive 60 to 10 mph coast downs
 on the hydrokinetic dynamometer following a HFET driving schedule (warm-up
 and sample). This was done to match the loading characteristics of the twin roll
 hydrokinetic dynamometer. The EPA VDA system was used to measure the
 speed/time data. These data were reduced to provide average coast down times
 for 5 mph speed intervals from 60 to 10 mph. Standard Horiba software was
 used to derive A, B, and C target force coefficients from the hydrokinetic
 dynamometer coast down data. These target force coefficients were used with
 the Horiba automated approach to determine dynamometer coefficients (over a
 60 to 10 mph coast down range) to yield a second set of coefficients for the
 electric dynamometer.

      Dynamometer tire pressure was set using the same approach as
described earlier for the true road load simulation tests.
Quality Control

      Test Vehicles - Vehicle sponsors conducted emission and fuel economy
tests to verify stability before delivery of the vehicles to EPA.

      D006  Representativeness - The representativeness of the D006
hydrokinetic dynamometer was based on EPA intra laboratory comparisons of
0006 diagnostic data, and emission and fuel economy results from EPA and
AAMA repeatable test vehicles on dynamometer D006.

      EPA Intra laboratory repeatable vehicle (REPCA) tests consisted of
weekly hot 505 second emission and fuel economy tests on dynamometers D001
through D006.  D006 results were examined using standard control charting
practices comparing it with D001-D004.  These data were made available to task
force participants before and during the test program.

      CVS. Analyzer Bench - The common analyzer bench and CVS for the test
program were verified using CFR and EPA laboratory specific criteria. These
included weekly TGI tests, daily CVS flow count checks, monthly analyzer curve
verifications,  daily analyzer checks using the EPA sample analysis correlation
(SAC) checks, and monthly NCJ2 to  NO converter efficiency checks.

      Hydrokinetic Dynamometer - The quality of the hydrokinetic dynamometer
was maintained by weekly REPCA tests and CFR and EPA standard diagnostic
checks.

      Electric Dynamometer - The operational integrity of the electric
dynamometer was based on adherence to Horiba's recommended practices and

-------
from suggestions from Task Force members. Attachment D describes the EPA
diagnostic checks for the electric dynamometer during the comparison study.
Weekly EPA REPCA tests were also run to assess emission and fuel economy
repeatability.

       Precautions were exercised during the dynamometer comparison study so
that the timing of calibration adjustments had a minimal impact on the results of
thp etnrtv
the study.
Results

      The tabular and graphical data specific to the electric dynamometer are
frequently described as "Road Simulation" or "Twin Roll Simulation".  Road
simulation refers to data obtained when the electric was used to simulate actual
on-road loading conditions. Twin roll simulation describes the mode of operation
of the electric when it was adjusted to simulate the twin roll hydrokinetic
dynamometer loading.

      Tables 1 -A and 2-A present emission and fuel economy changes of the
large roll electric dynamometer relative to the current twin small roll hydrokinetic
dynamometer. Table 1-A presents normalized results expressed in percent
difference. The Ford F-150 and Ranger trucks exhibited some of the greatest
changes in emissions and fuel economy. Table 2-A presents absolute
differences expressed in units of g/mi or mi/gal. Tables 1-B and 2-B present
similar results for the twin roll simulation data.

      Figures 1-5 are displays of the percent difference between the electric and
hydrokinetic dynamometers.  Figures 1-5 compare the differences in HC, CO,
NOx emissions, and FTP and HFET fuel economy, respectively.  All differences
are presented with respect to the hydrokinetic dynamometer.

      Figures 6-10 are similar presentations for results using the hydrokinetic
simulation loading on the electric dynamometer versus the  hydrokinetic baseline
condition.

      Appendices 1-4 present the individual composite data and the individual
bag 1, 2, and 3 values. Appendix 1 summarizes individual  test results for the
FTP and HFET road simulation results versus the hydrokinetic tests.  Appendix 2
presents the data for the tests used to compare the hydrokinetic simulation to the
hydrokinetic baseline.  Note that the hydrokinetic baselines for individual vehicles
in Appendices 1 and 2  are sometimes different because twin roll simulation tests
were run after the road simulation tests, and therefore the hydrokinetic baseline
was rerun to prevent confounding effects due to changes over time in the
vehicle's emission and fuel economy performance. Appendices 3 and 4 are the
individual test bag data.

      Appendices 5 and 6 summarize coast down data. Appendix 5 presents
55 to 45 mph coast down times for road simulation tests versus-the hydrokinetic

-------
 dynamometer data. Appendix 6 displays similar data for twin roll simulation tests
 on the electric dynamometer compared to the hydrokinetic data. Coast downs on
 the hydrokinetic dynamometer were the standard 55 to 45 mph quick check,
 while coast downs on the electric were from 70 to 10 mph for the road simulation
 tests on the electric, and from 60 to 10 mph for coast downs utilizing the
 simulation mode of the electric.

       Appendix 7 contains plots of coast down road load force curves for the
 nine vehicle test fleet.  Each figure includes a plot derived from the 75 to 5 mph
 track force loading (labeled "Track"), the 70 to 10 mph force loading on the
 electric used for road simulation tests (labeled "Single  Roll"), and the hydrokinetic
 dynamometer load from 60 to 10 mph (labeled "Twin"). The plots are least
 squares fits of the reduced track and dynamometer coast down data extrapolated
 to 0 mph. The two dynamometer force  curves include dynamometer load and
 friction at the tire/roll interface. The lower half of each  figure also includes a plot
 of the difference in loading between the two dynamometers.

      Ninety-five percent confidence intervals around the absolute means for
 emissions and fuel economy are displayed in Appendices 8 and 9. Confidence
 intervals in Appendix 8 are specific to the road simulation mode of the electric
 versus the hydrokinetic baseline. Appendix 9 refers to similar summaries for the
 simulation mode of the electric versus the hydrokinetic baseline. The plots in
Appendices 8 and 9 are interpreted as showing statistically significant differences
 at the 95 percent confidence level if the confidence intervals do not overlap.
Discussion

      Road Simulation Results - Tables 1-A, 2-A, Figures 1-5, and Appendix 8
all suggest that, in general, when adjusted to reproduce on-road loading
conditions, vehicles tested on the electric dynamometer showed higher exhaust
emissions, similar FTP fuel economy, and lower HFET fuel economy relative to
the hydrokinetic dynamometer.  An examination of the data shows that the
dynamometer influence is quite vehicle specific. Individual vehicles showed
higher, lower, or no change in exhaust emissions and fuel economy.

      Twin Roll Simulation Results - Tables 2-A, 2-B, Figures 6-10, and
Appendix 9 again suggest a vehicle specific response to twin roll simulation on
the electric dynamometer. In general,  CO emissions were biased higher on the
electric, as was FTP fuel economy. As previously stated, the manual
transmission Honda vehicle could not be tested on the electric dynamometer due
to feedback stability in the twin roll simulation control circuitry which caused the
vehicle to shake severely in the 15-20  mph speed range as the transmission was
shifted into second gear.

      An informal consensus of the Task Force was that  significantly more
development work was necessary, without certainty of success, to have
confidence that the twin roll simulation could produce emissions  and fuel
economy similar to the hydrokinetic dynamometer on a vehicle specific basis.

-------
      Coast Down Data - An inspection of the 55-45 mph coast down data in
Appendices 5 and 6 show that the loaded coast down times had lower
coefficients of variation for seven of nine vehicles in road simulation mode, and
lower variation in twin roll simulation mode for six of eight vehicles, relative to the
twin roll results.

      Force Curve Plots - Track and dynamometer road force curves are plotted
in Appendix 7.  They show that the coast downs performed on the electric
dynamometer more accurately matched the measured track coast down force
versus velocity data than coast downs performed on the hydrokinetic
dynamometer.

      When reviewing only the hydrokinetic coast down data, it can be noted
that the  rear wheel drive vehicles more accurately matched the track road load,
while the front wheel drive vehicles tended to overload at low  speed and
underload at speeds above about 50 mph.
Conclusions

      Three conclusions are developed from the data in this study:

1)    When the 48 in. single roll electric dynamometer was operated in road
simulation mode, exhaust HC, CO, and NOx emissions were generally higher,
FTP fuel economy results were similar, and HFET fuel economy results were
lower relative to the 8.65 in. twin roll hydrokinetic dynamometer.

2)    The emissions and fuel economy differences of the nine vehicle fleet
attributable to the change in dynamometer design and loading were vehicle
specific.

3)    It was the Task Force's opinion that the twin roll simulation did not
accurately estimate the loading of the twin roll hydrokinetic dynamometer, and
that significantly more development work would not guarantee better accuracy.

            /
Acknowledgments

EPA wishes to express its appreciation to the many individuals in the motor
vehicle and dynamometer industries who supported this project; it could not have
succeeded without the high level of cooperation which was received.  All aspects
of this endeavor, from planning to the preparation of this report, benefited from
the cooperation received. While naming specific individuals runs the risk of
inadvertently omitting someone, EPA wishes to thank those members of the task
force listed below as well as other unnamed members of their firms who assisted
in this effort:

-------
            Individual
Firm
            Giedrius Ambrozaitis
            Ken Barnes
            Peter Benkmann
            Bob Bisaro
            Frank Buckley
            Ken Burt
            Tommy Chang
            Charles Cowham
            Severino D'Angelo
            Todd Fagerman
            Anastas Farjo
            Todd Fronckowiak
            Robert Gower
            Mark Guenther
            Steve Hunter
            Paul Karas
            John Keefe
            Chuck Kizlauskas
            Dianna Korduba-Sawicki
            Alan Kuge
            Brooke Lament
            Dave Luzenski
            Ted Malachowski
            Bill Mears
            Richard Pearl
            Dave Perkins
            Dave Pruess
            David Robertson
            Bob Slater
            Dan Sougstad
            Kim Waggoner
            William Watkins
            Allen White
Mercedes-Benz
Froude
Mercedes-Benz
Ford
Buckley Associates
General Motors
Honda
Burke-Porter
Horiba
Ford
Ford
Ford
Chrysler
Ford
Ford
Ford
General Motors
Ford
Ford
Honda
Clayton
General Motors
Suzuki
Horiba
Toyota
Froude
Chrysler
Ford
Mercedes-Benz
General Motors
Nissan
Schenck-Pegasus
Chrysler
Acknowledgment is also made to the efforts of all the EPA staff who contributed
to the test program, including Bob Gilkey and the test site technicians:  Phil
Conde, Ken Lesage, Dave VanAmburg, and Paul Velandra.

-------
                                     Attachment A
                         Dynamometer Comparison Study Test Fleet
Vehicle
Model
cubic inches
Ford F-150
Mercedes 300E
Cadillac DeVille
Toyota Truck
Ford Ranger
Nissan 240SX
Dodge Caravan
Chevrolet Lumina
Honda Civic
1992
1987
1991
1991
1993
1991
1992
1992
1992
300
159
300
183
140
146
201
189
92
A-4
A
A-4
L-4
A-4
A-4
A-4
A-4
M-5
Rear
Rear
Front
Rear
Rear
Rear
Front
Front
Front
pounds
1861
1761
2427
3122
1140
1409
2345
2232
1442

Firestone
Pirelli
Michelin
Bridgestone
Firestone
Toyo
Goodyear
Goodyear
Goodyear
1 M W VXI*_Wr
P235/75R15
I95/65VR16
P205/70R15
185/R14LT
P195/70R14
195/60R15
P205/70R15
P215/60R16
P175/70R13

-------
                                 Attachment  B
          EPA/Industry   Dynamometer  Comparison   Study
      Procedure   for  8.65  in.  Twin  Roll   DPA  Determination

 1.     Drain tank and fill to 40% of volume. Record the time and date on the
       Dynamometer Power Absorption Determination data sheet.

 2.     Weigh the front axle, rear axle, and total vehicle.  Record this on the data sheet
       and attach the weigh scale print out. Notify test requester of weight results
       before proceeding.

       2.1   Total weight must be within 100 Ibs of production vehicle weight.
       2.2   Drive axle weight must be within  50 Ibs of production axle weight.

 3.     Fill the drive axle tires to 5 psi above the vehicle manufacturer's dynamometer
       tire test pressure. Record tire data on the data sheet.

 4.     Park the vehicle in the soak area  for a minimum of 4 hours to stabilize the tire
       temperature. Record the initial and final soak  times on the data sheet.

 5.     Prepare the vehicle and dynamometer.

       5.1    Drive the vehicle onto the dynamometer.
       5.2   Position front cooling fan and driver's  aid, adjust the tie-down cable to
             normal  tension.
       5.3   Reduce the tire pressure to 45 +/- 1 psi for passenger cars, or the
             vehicle  manufacturer's  dynamometer  tire test  pressure +/- 1 psi for
             trucks.
       5.4   Position a side cooling fan  if used by the manufacturer for conventional
             HFET tests.
       5.5   Set the  dyno inertia weight  and  manufacturer AHP from the DPA data
             sheet.  Verify proper flywheel engagement.
       5.6   Select the rear roll position on  the dynamometer speed/power meter.
       5.7   Select the Automatic and Count positions on the quickcheck timer.  (Steps
             5.6 and 5.7 will  permit measurement  of front  roll quickcheck time).

6.     Warm-up  the vehicle and dyno by  driving an HFET (warmup and sample).

7.     Begin coastdown measurements within  1 minute of the end of the HFET,
                /
       7.1    Accelerate at an  approximate rate of 2  mph/sec to about 65 mph and hold
             that speed for about 2 seconds.
       7.2   Verify that the quickcheck  timer has reset to zero.
       7.3   Shift the vehicle transmission to  neutral.
       7.4   Allow the vehicle to coastdown to  40 mph.
       7.5     Repeat Steps 7.1 through  7.4  until 3 coastdowns (not necessarily
             consecutive) have been obtained within a range of 0.3 seconds.  Do not run
             more than 5 coastdowns. Record all values on the data sheet.

 8.     When 3 coastdowns are completed at the load setting in Step 5.5, increase or
       decrease  the actual horsepower settings, and repeat Steps 7.1-7.5. In order, set
       actual horsepowers equal to +0.5 HP, -0.5,  +1.5, -1.5, +1.0, and -1.0 above
       and below the manufacturer's DPA value. Record the thumbwheel settings on the
       data sheet.

 9.     Complete the DPA data  sheet and  submit to the test requester for processing.

-------
                                                Attachment C
                                Track and Dynamometer Loading Conditions
                                                                                                       Twin Roll
Vehicle
Ford F-150
Mercedes 300E
Cadillac DeVille
Toyota Truck
Ford Ranger
Nissan 240SX
Dodge Caravan
Chevrolet Lumina
Honda Civic
Track Coefficients
A B . C
21.8
40.4
15.9
54.9
33.3
48.1
27.1
26.9
15.4
0.9315
0.3529
0.4716
0.5983
0.1642
0.0096
0.4804
0.4216
0.1384
0.03266
0.01636
0.02444
0.03058
0.02903
0.01809
0.02494
0.01637
0.01960
Electric Dynamometer Coefficients Taraet Time
ABC 55-45 mnh. sec
12.11
13.41
4.91
16.28
13.77
23.71
3.16
8.73
4.60
0.1108
0.2089
-0.0538
0.2812
0.11 13
-0.3065
0.1921
0.0392
0.0105
0.04090
0.01688
0.02467
0.03152
0.02842
0.02001
0.02662
0.01888
0.01982
13.97
17.55
17.96
15.13
14.25
15.43
16.50
19.07
16.30
Test Weight.
pounds
4500
3750
3875
5250
3500
3125
4000
3625
2500
Actual Hp @
50 mph
14.1
6.9
6.6
11.8
11.0
6.5
7.8
4.6
6.9
Note:  A, B, and C track and dynamometer coefficients are expressed in units of Ibf, Ibf/mph, and lbf/(mph)(mph), respectively.

-------
                  Attachment D
       Electric Dynamometer Operational Checks
Frequency        Description

Daily:            Warm-up Procedure
                 Auto Calibration Test
                 Vehicle-Off Coast Down at Test Condition
Weekly:          Parasitic Loss Check
                 Timing Test
                 1500 and 5500 Ib Vehicle-Off Coast Downs
                 Zero Load Speed Check
Monthly:          Lubricate Roll Cover Tracks
                 Inspect Roll Brake Lines
                 Clean Air Filters on CDC-900 and Power
                       Converter Cabinets

-------
                                Table 1-A
                   Dynamometer Comparison Study

    Emissions and Fuel Economy Differences of 48 in. Single Roll Electric
  Dynamometer Relative to 8.65 in. Twin Roll Hydrokinetic Dynamometer, %
Ford F-150

Mercedes 300E

GM Cadillac

Toyota Truck

Ford Ranger

Nissan 240SX

Dodge Caravan

GM Lumina

Honda Civic
N
8
6
6
7
8
4
7
8
7
HQ
25
1
12
5
9
18
-2
25
7
QQ
46
20
19
0
70
58
9
35
17
NOx
40
7
0
7
17
5
22
10
15
FTPFE
-0.6
1.4
1.1
0.1
-2.0
-0.1
2.7
1.1
-0.2
HFETFE
-5.1
-0.6
-4.5
-2.2
-7.2
-0.9
-4.3
-4.8
-1.9
     Notes:      N equals number of FTP tests on the single roll.  Number of FTP
                tests on the twin roll are approximately the same.

                Electric dynamometer matches actual road force from 70 to 10 mph.
             '   Hydrokinetic dynamometer uses standard load curve.

                % = ((Single roll - Twin roll)/(Twin roll))100
                                                                           2/2

-------
                                Table 2-A
                   Dynamometer Comparison Study

    Emissions and Fuel Economy Differences of 48 in. Single Roll Electric
    Dynamometer Relative to 8.65 in. Twin  Roll Hydrokinetic Dynamometer
Ford F-150

Mercedes 300E

GM Cadillac

Toyota Truck

Ford Ranger

Nissan 240SX

Dodge Caravan

GM Lumina

Honda Civic
N
8
6
6
7
8
4
7
8
7
HC g/mi
0.064
0.003
0.017
0.007
0.014
0.036
-0.004
0.063
0.007
CO g/mi
0.74
0.59
0.37
0.01
3.86
1.72
0.12
1.29 ,
0.16
NOx g/mi
0.206
0.021
-0.001
0.016
0.014
0.024
0.093
0.037
0.028
FTP mi/aal
-0.09
0.28
0.20
0.02
-0.44
-0.01
0.54
0.21
-0.07
HFET mi/<
-1.25
-0.18
-1.45
-0.49
-2.19
-0.33
-1.36
-1.63
-0.98
     Notes:     N equals number of FTP tests on the single roll.  Number of FTP
               tests on the twin roll are approximately the same.

               Electric dynamometer matches actual road force from 70 to 10 mph.
               Hydrokinetic dynamometer uses standard load curve.

               Applicable emission standards -
               Mercedes 300E, GM Cadillac, Nissan 240SX, GM Lumina, Honda Civic
               HC = 0.41  CO = 3.4  NOx = 1 .0

               Ford F-150, Toyota, Dodge Caravan
               HC = 0.80 CO = 10.0  NOx = 1.7
                Ford Ranger
                HC = 0.80  CO = 10.0  NOx = 1.2

-------
                                Table 1-B
                   Dynamometer  Comparison Study

 Emissions and Fuel Economy Differences of Electric Dynamometer Twin Roll
   Simulation  Relative to 8.65 in. Twin Roll Hydrokinetic Dynamometer, %
Ford F-150

Mercedes 300E

GM Cadillac

Toyota Truck

Ford Ranger

Nissan 240SX

Dodge Caravan

GM Lumina

Honda Civic
N
6
5
8
7

7
6
6
6
0
HQ
3
4
8
6
0
-7
4
0
-2

QQ
19
1 1
8
12

-16
10
3
-7

NOx
-1
10
-2
2

-3
0
3
2

FJPFE
1.2
1.9
2.0
3.2

3.6
1.3
0.5
-0.7

HFETFE
-0.3
2.8
1.5
1.8

2.5
0.3
-0.1
-1.3

     Notes:      N equals number of FTP tests on the single roll.  Number of FTP
                tests on the twin roll are approximately the same.  Honda  could not
                be tested on the single roll.

                Electric dynamometer matches hydrokinetic load from 60  to 10 mph.
                Hydrokinetic dynamometer uses standard load curve.
                % = ((Single roll - Twin roll)/(Twin roll)) 100

-------
                                Table 2-B
                   Dynamometer Comparison Study

Emissions and Fuel Economy Differences of Electric Dynamometer Twin Roll
     Simulation Relative to 8.65 in. Twin Roll Hydrokinetic Dynamometer
Ford F-150

Mercedes 300E

GM Cadillac

Toyota Truck

Ford Ranger

Nissan 240SX

Dodge Caravan

GM Lumina

Honda Civic
N
6
5
8
7
7
6
6
6
0
HC g/mi
0.008
0.010
0.011
0.008
-0.010
0.007
0.001
-0.005

CO a/mi
0.27
0.37
0.16
0.17
-0.78
0.30
0.04
-0.23

NOx g/mi
-0.006
0.030
-0.011
0.004 .
-0.003
0.001
0.012
0.009

FTP mi/gal
0.19
0.41
0.35
0.48
0.80
0.31
0.10
-0.15

HFET miA
-0.08
0.86
0.48
0.38
0.74
0.10
-0.04
-0.45

     Notes:     N equals number of FTP tests on the single roll.  Number of FTP
               tests on the twin roll are approximately the same.  Honda could not
               be tested on the single roll.

             ,  Electric dynamometer matches hydrokinetic load from 60 to 10 mph.
               Hydrokinetic dynamometer uses standard load curve.

               Applicable emission standards -
               Mercedes 300E, GM Cadillac, Nissan 240SX, GM Lumina, Honda Civic
               HC = 0.41 CO = 3.4  NOx = 1.0

               Ford F-150, Toyota, Dodge Caravan
               HC = 0.80 CO = 10.0  NOx = 1.7   •

               Ford Ranger
               HC = 0.80 CO = 10.0  NOx =1.2
                                                                           2/27.35

-------