PROCEEDINGS OF THE
EPA/INDUSTRY
EMISSIONS TESTING PRACTICES
SYMPOSIUM
DYNAMOMETER CALIBRATION
AND CHARACTERISTICS
AUGUST 16, 1979
-------
PROCEEDINGS OF THE EPA/INDUSTRY
EMISSIONS TESTING PRACTICES SYMPOSIUM
DYNAMOMETER CALIBRATION AND CHARACTERISTICS
August 16, 1979
9:00 a.m. - 4:00 p.m.
HELD AT
EPA Laboratory
2565 Plymouth Road
Ann Arbor, Michigan 48105
Written and Compiled
by Don Paulsell
-------
Proceedings of the Dynamometer
Testing Practices Symposium
Introduction: In 1977, personnel from the EPA Motor Vehicle Emission
Laboratory and the auto industry formed an ad hoc committee to conduct
a series of symposia on emissions testing quality control practices*
Five sessions were held to discuss dynamometers, lab correlation, gas
traceability, interlab test site diagnostics, and curve generation/
data validation* Each symposium was summarized by its chairman in a
document similar to this one*
A new committee has been convened this year to plan, coordinate and
conduct a new series of symposia* The first symposium was held at the
EPA Laboratory on Thursday, August 16, 1979 to discuss areas related
to dynamometer calibration and characterization.
In preparation for the symposium, a questionnaire was mailed with each
announcement to gather information related to dynamometer characteris-
tics and acceptance parameters for calibrations. The responses
received have been summarized in Attachment III.
Synopsis of Discussions: The general outline in Attachment I summa-
rizes the areas discussed during the symposium. This synopsis does
not represent a transcript of exactly what was said, but will attempt
to summarize in general terms the highlights as noted during the
meeting. Item 1 in the outline was an introduction and discussion of
the symposium program. The chairman of this symposium, Don Paulsell
from EPA, briefly described what the committee planned to achieve by
holding a new series of symposiums. It was emphasized that this
program would be a series of informal technical information exchanges
dealing with topics of interest to both EPA and the industry. The new
program would be called Emission Testing Practices rather than Quality
Control Symposiums to reflect that emphasis would be on standardiza-
tion of testing practices. However, quality control is a significant
part of any standardization practice; therefore, quality control
techniques will still be emphasized. It was also reiterated that
participation is the key note of these symposia. If they are to be
effective and achieve their stated goal, informed and prepared parti-
cipation is a very necessary condition. This particular symposium had
approximately 60 attendees, but it is estimated that only 15 actually
participated with information. While nonparticipatlng attendees may
have learned a great deal, it seemed that the presence of a large
number of people was somewhat inhibiting to others who might have been
tempted to participate in a smaller group of people. The purpose of
these proceedings is to provide a mechanism for everyone to share in
the discussions at the symposium without making their attendance
mandatory. The committee hopes that future symposia can also reflect
this philosophy.
-------
-2-
The second area discussed during the symposium was calibration prac-
tices used for calibrating the Clayton ECE50 dynamometer with 125 lb.
inertia flywheels. EPA recently finalized and distributed two proce-
dures dealing with the subject of calibration practices; one on the
electronics circuitry setup, and one on the frictional horsepower
calibration. A third procedure dealing with weekly verification and
quality control practices will also be distributed.
In January, 1977, EPA had published a draft procedure and recommended
practice discussing the technically correct method of subtracting 155
lbs. of inertia for the horsepower calculations. This practice offers
significant quality control assessment capabilities, but at the same
time introduces a nonuniform bias in dynamometer horsepower settings.
Therefore, the industry contended that a notice of proposed rulemaking
would be required to implement this change. However, the benefits of
this technique were judged to be so significant that EPA continues to
utilize the technique for quality control purposes, but reprocesses
the data in the manner that it has always used for the certification
purposes.
Don Paulsell discussed the acceptance criteria and the validation
process which is used to verify the data collected in Test Procedure
202. A copy of the data sheet was discussed and EPA offered to
provide these to any manufacturer who would collect the data speci-
fied. EPA would process the data using the EPA program to provide a
comparative analysis between EPA and another facility. In response to
a question regarding the availability of the EPA computer program, it
was ptated that EPA's policy was not to release software because, at
times, the documentation is not complete and routines that are incom-
patible with other computer programs frequently cause problems and
consume considerable time in trying to get the program to work. The
information, however, in the Test Procedure 202 should be sufficient
for other facilities to model their computation practices around the
EPA procedure, if they so wish to do this.
All of the acceptance criteria used to validate Test Procedure 202 are
included in the procedure. The following paragraphs will briefly
summarize the areas discussed and some of the information provided by
the participants at the symposium regarding related criteria used
throughout the industry.
The data manipulation routines used for Test Procedure 202 are speci-
fied through an interactive program. First of all, one can specify
whether full inertia weight or (IW-155) is to be used in the computa-
tion of AHp. The initial QC processing normally uses the (IW-155)
routine. The technician can also specify a regression of thumbwheel
(TW) vs AHp or can use the integrated data (T*S) vs AHp. In both
cases, no certification table is produced. This information is
-------
-3-
printed out in the upper right hand corner of the first page of the
printout. All of the general descriptive information concerning the
dynamometer and its calibration are listed in the upper left hand
corner of the first page of the printout. The last item in that
listing is the bearing friction as calculated from the average of five
rear roll coastdowns. EPA performs the coastdowns of the rear roll
during several flywheel warmup periods in the procedure. The average
coastdovm time for this rear roll is usually 50-60 seconds; the
variation is +.5 seconds from the average time.
At this time, a discussion of bearing friction took place. The
requirement for stable and constant bearing friction during a calibra-
tion and dynamometer operation is very critical to the correlation of
dynamometers. EPA typically uses a break-in period for new bearings
of approximately eight hours. Some participants at the symposium
mentioned that they have found it necessary at times to use 24 hours
of vehicle operation to break in the bearings and at times General
Motors has used up 72 hours before they considered the bearings
stable. The most critical parameter related to bearing friction
seemed to be the alignment of the bearing housings and the shaft
couplings. The use of laser equipment or measurement devices capable
of detecting +.005 inches of misalignment were mentioned as standard
techniques. Although the quantification of bearing friction values
has not become a well established practice, limited data seems to
indicate that the frlctional values for the new C3 bearings used In
the Clayton dynamometers do not differ significantly from those values
obtained with standard bearings. Attachment IV Illustrates the
frlctional values obtained across all EPA dynamometers for these new
bearings. Also shown on that table are the frlctional values obtained
two years ago with the standard 250 lb. configuration.
Only the trim friction appears to be lower for the new units. As a
final aside to the discussion of bearing friction, we briefly touched
on the use of machine rolls. It was noted that machined rolls pro-
duced better driveability and seemed to have lower noise level, and
for obvious reasons, more consistent circumference. This may be
reflected in the statistics associated with the actual distance
measurements made on EPA dynamometers as compared to General Motors
dynamometers, which have the machined rolls. These data are attached
in the survey. EPA will be installing the machined rolls late in 1979
on all of their dynamometers, so the assessment can then be made as to
whether machined rolls provide better precision for distance measure-
ments.
The discussion turned to the description of the validation process of
EPA's calibration data. The top three sections on the printout
illustrate the speed calibration data, load cell calibration data, and
data collected to quantify the exponent of the power absorber unit.
-------
-4-
The tach generator signals are compared to an absolute reference speed
as detected by the frequency of a 60-tooth digital output from a gear
and magnetic pickup. Other laboratories are using similar techniques
such as high resolution pulse encoders and micro processor data
acquisition to perform the calibrations on the tach generators. EPA
performs an 11 point tach generator calibration, an 11 point dead
weight calibration on the load cell, and an 11 point power absorber
curve. However, on the power absorber curve, only the data from 25-60
mph are used in the log-log regression to obtain the exponent. EPA's
criteria on exponent control is 1.85 to 2.15. Generally, the expo-
nents are much closer to the value of two and can be adjusted,
although the circuit cards have sealed adjustments. The participants
from VW agreed that the value of two was an appropriate exponent for
track data. However, GM noted that 2.0 may not be the right value for
the dynamometer simulation, which differs because of the tire roll
interface characteristics. EPA's quantification of the PAU exponent
comes from steady state type of data collection. General Motors
explain that they also employ an XY plotter to look at the relation-
ship under dynamic conditions. They generally see that the torque
speed relationship is slightly low during the accelerations and comes
back a little high during a deceleration. This characteristic was
noted during the last symposium, but it is generally accepted that the
automatic control mode reduces the historesis band significantly, as
compared to the manual control mode.
The next section of the computer printout used at EPA shows the three
point regressions used to fit the thumbwheel (TW) vs total horsepower
(AHp). When the (IW-155) calculation is used, the slope of these
curves all approximate 1.00 and should be very uniform across all of
the inertia weights. The occurrence of a slope of 1.00 means that the
friction does not change as a function of a PAU or TW setting. A
slope that is less than one indicates that friction is increasing with
increasing thumbwheel setting and a slope greater than one means that
friction is decreasing. Theoretically, the slope could also be
affected if the inertia weight used in the calculation was not physi-
cally equivalent to the value assumed in the calculation. EPA has
verified the dimensional parameters of all of the inertia weights and
has discussed this aspect with Clayton. Deviations from assumed
inertia values are not considered to be a significant problem.
EPA has experienced some deviations in slopes from the theoretical
value of 1.00. Special investigations have not revealed the cause of
these differences, although, shaft misalignment and dead band adjust-
ments are prime candidates for causing this characteristic. In
particular, the deviations from theoretical behavior appear to be
associated with the engagement of the 2,000 lb. flywheels. From
another viewpoint, the shifting in average slopes can also be asso-
ciated with thumbwheel values set for the PAU, since the values change
at 3000 and 5000 pounds, coincident with the engagement of 2K fly-
wheels .
-------
-5-
EPA also prints out the average speed during the coastdown as a
measure of uniformity of data. Although one would assume that the
average speed of a 55-45 coastdown would be 50 mph, the integrated
speed is more on the order of 49-7~ The representative of Ford
mentioned that their microprocessor data collection method indicates
the instantaneous derivative of the speed time trace is higher at 50
than would be approximated by the average computation of i5v/ bt over
the 55-45 range.
The frictional horsepowers determined in the individual inertia
computations -are also summarized in a table of flywheel frictions.
This information has become very useful in detecting significant
changes lti flywheel friction and Ln assessing the uniformity of the
values obtained through the selective inertia subtractions shown in
the third area of the printout* These subtractions determine the
flywheel friction contributed to the total friction by the addition of
the single flywheel. When large amounts of scatter are observed in
the statistical summary, this table is extremely useful in identifying
which inertia weights a re associated with the abnormal data. The
final printout on the (1W-155) processing is a regression of integra-
ted horsepower T*S and IEtp <3 50 versus t"he thumbwheel value set. This
information is useful in verifying the stability of Hp control across
thumbwheel values of 3 to 22 horsepower.
All of the techniques previously discussed and the acceptance criteria
that are employed in validating the calibration data set are specified
in Test Procedure 202. Furthermore, a comprehensive exanrple o£ each
afea is printed in the attachment to 202. Once the d$ta set has
passed this validation procedure, the data can be reprocessed by
requesting a certification table. This option automatically defaults
to TW vs AHp and the Full IW computation for total horsepower.
Although this example was not attached to Test Procedure 202, an
example of this printout for the data set used in 202 is i*> Attachment
V of the appendix.
EPA Test Procedure 207 was not discussed in detail since it is primar-
ily a reiteration of what Clayton specifies in their calibration
procedure for the electronics. However, there are some functional
checks that EPA performs and there is a log sheet in Test Procedure
207 that is filled out for diagnostic purposes which may be of some
interest to people establishing new calibration procedures.
Test Procedure 302, which is tVie a&d monitoring
procedure, was briefly discussed, although this procedure is in its
final review phase. A copy of the data sheet that is used with Test
Procedure 302 and the control charts, which are plotted to assess the
stability and trends in verification data, were distributed. The use
-------
-6-
of control chart techniques for data of this type cannot be emphasized
too strongly. While many facilities employ a go, no-go criteria to
verification coastdown times, the use of control chart can show trends
in data long before the criteria are exceeded. EPA's experience
generally indicates that it is shifts in the control circuit, primar-
ily the dead band, which cause the coastdown times to deviate from
their theoretical values. This indicates that the dynamometer fric-
tion obtained during the initial calibration is probably a very
stable, consistent parameter.
Closely related to the discussion of the calibration practices was the
review of the data collection and instrumentation methods which are
employed to collect the coastdown data. Most laboratories are utiliz-
ing an automated type of triggering device to measure the coastdown
times. EPA uses an analog comparator which has a filtered tach signal
as an input, which is compared to precision reference voltages set at
4.50(0 and 5.500. The EPA circuit can control and trigger to within
+.02 mile per hour of the given set point. General Motors stated that
they use an optical encoder on the front roll shaft and a digital
comparator circuit to trigger the 55 and 45 set points. Although the
precision of their timing circuit was not mentioned, they did say they
had a repeatability of about .2 horsepower over an eight-month period
across two dynamometers. Ford uses a single pulse per rev input to a
microprocessor which can time the period of each pulse compared to a
set point. At 50 mph, each period would translate to 30 milliseconds.
However, since Ford obtains coastdown times that are quite long, in
excess of 30 seconds, this uncertainty translates to a triggering
accuracy of +.01 mile per hour.
Many laboratories stated that they perform multiple coastdowns to
minimize the uncertainty in the coastdown value. While repetitions
are a valid technique to reduce uncertainty, it becomes very time
consuming when calibrating the 125 lb. inertia configuration because
of the large number of flywheel inertia weights needing calibration.
The utilization of a highly precise triggering circuit and the quanti-
fication of its precision can eliminate the need for these multiple
coastdowns. The data validation techniques discussed earlier can also
eliminate the need for time-consuming data collection by taking
advantage of the characteristics of the complete data set.
Another technique which was discussed for reducing the amount of time
required to calibrate a dynamometer was the use of what is referred to
as "prime" inertia weights. This method can utilize either the
successive addition or the successive subtraction of individual
flywheels. Since there are seven flywheels, one can determine the
frictional horsepower of each individual flywheel by the successive
subtraction of the total friction of each of the "prime" combinations.
-------
-7-
Prime combinations can be formed starting from the 1,000 lb. trim and
adding individual flywheels one at a time, or can be derived by
starting at 6,875 and successively dropping each flywheel from the
total combination. EPA has elected to use the former method because
it offers the ability to get diagnostics data on each particular
flywheel. Ford has had considerable experience using the latter
method and has found very good agreement between the prime calibration
technique and the total calibration technique. Their investigations
are still in the developmental phase and other laboratories are
encouraged to try this technique and to report their results at a
later time. The use of prime flywheel friction determinations
requires the subtraction of the rear roll 155 lbs. However, this
manipulation can be retransformed by using (IW-155)/IW to generate the
current certification type equation for each inertia weight. As more
laboratories begin to collect and analyze individual flywheel fric-
tions and calibration slope characteristics for the (IW-155) type of
analysis, more data will be gathered to quantify whether the prime
inertia calibration method is, in fact, an acceptable equivalent.
The third area discussed during the morning session dealt with dyna-
mometer characteristics. Some of these have already been discussed,
but two additional parameters are worth noting. First is the sensi-
tivity of the load cell to changes in barometric pressure. Represen-
tatives of Clayton have noted that the hermetically sealed load cell
displays this characteristic. They have instituted a modification to
minimize the sensitivity. This has been done by venting the internal
cavity of the load cell through a sealed type vent. General Motors
stated that they had simply drilled a hole and put a filter into the
hole on the side of the load cell cavity. Clayton personnel said this
would be an acceptable practice, but any metal chips that might fall
inside during the drilling operation could be detrimental to the
operation of the load cell.
The second area discussed related to the slip characteristics of the
dynamometer during a horsepower verification or setup procedure. Slip
is defined as the difference in speed between the front roll and rear
roll caused by the difference in loads imposed on the vehicle tire by
the two rollers. This slip characteristic will not affect the control
of the power absorber unit in the automatic mode, but can affect the
indicated horsepower readout when observed in the rear roll speed
position. EPA has standardized its practice of verifying all dynamom-
eter load control functions at the PAU reference speed of 50 mph.
Laboratories that use the manual mode of load control should be
particularly aware of the effects of slip and should avoid setting any
calibration or test setup PAU values in the rear roll mode. In other
words, front roll (PAU) speed should always be used for calibration
and test setup. The driving trace is generated by the rear roll
signal.
-------
-8-
After a one-hour break for lunch, the symposium reconvened to discuss
the fourth topic on the general outline, calibration monitoring and
correlation. The topic of site-to-site variability was discussed
and areas were highlighted which affect the variability from one
dynamometer to another. The tire roll interaction is considered by
most to be the largest source of variability related to loaded vehicle
coastdowns. Although not too many laboratories within the industry
perform quick checks, EPA's data tends to indicate that repeatability
of a quick check on one particular dyno is on the order of +.2 seconds
or approximately 1% of the coastdown time. However, when determining
alternate power absorber settings across different dynamometers, the
variability can be much larger, because the amount of friction in the
vehicle drive train considerably dominates the frictional component of
the dynamometer. Therefore, the sensitivity of coastdown time to
horsepower setting is approximately one to two or 7% change in coast-
down time would equate to 15% change in dynamometer horsepower set-
ting. The participants from Mercedes Benz noted that the restraints
that are used during the test can also contribute to significant
variability in dynamometer loading. This was quantified in a sub-
mission made during the 1977 symposium on dynamometers.
Load control and transient response characteristics are another source
of variability. EPA has instituted a monitoring practice to verify
that the high rate load and unload response times are typically less
than eight seconds for each dynamometer. This represents a change of
5 to 15 horsepower and from 15 to 5 horsepower. Plugged valves due to
water deposits can frequently make the response time much longer than
this. Some laboratories have gone to closed water systems for the
power absorber unit, while others employ the standard Clayton plug for
removing these deposits. EPA briefly discussed its experience with
using a grease gun to unplug clogged ports. The lower temperatures
now being used in the power absorber units do not allow the grease to
be purged from the system. Therefore, we are recommending that water
soluble hand cleaner be used at all times. This technique works just
as well and does not cause the problems of grease deposits and the
almost certain reocurrencce of plugged ports.
The use of the lower temperature setting specified for the Clayton 125
lb. dynos was briefly discussed. Clayton previously recommended a
power absorber heat exchanger setting of 135°, but now specifies
90-100°. Some of the laboratories had collected data on this change
and found a negligible difference in coastdown times for the two
temperature settings.
Other techniques that can be used to assess calibration stability
are the use of repeatable cars and dynamometer quick-check results.
The statistical analysis of the quick-check results from a regularly
scheduled repeatable vehicle can provide data for highlighting signif-
icant differences in dynamometers. These differences can then be
diagnosed by other special tests. EPA's repeatable car has an average
-------
-9-
quick-check coastdown time of about 12 seconds* The overall variation
from day to day is on the order of +2%. The average difference from
dyno to dyno is less than 1.5%;. It is recommended that a rear-wheel
drive, medium-size tire, vehicle be used for doing these repetitive
coastdowns. The most Important aspect of this procedure is performing
it in a standardized, uniform manner so that the variability can be
minimized and the data collected become more meaningful.
A lot of the specific details and data relative to the areas discussed
in this synopsis are summarized in the composite response to the
questionnaire survey sent out prior to the symposium. These are shown
in Attachment III of the appendix. EPA's response to the question-
naire Is also shown. Even if a laboratory did not respond to the
questionnaire, they are encouraged to pursue the quantification of
some of the parameters discussed In the questionnaire. This informa-
tion could be submitted to the Quality Assurance Staff at EPA for a
follow-up report to the symposium, if enough Information can be
compiled.
The last topic discussed in the general outline was a report on three
special projects. The first was a dynamometer calibration project
performed by a European group. Dr. Shurmann from Volkswagen reported
on this project and a copy of the calibration technique Is in Attach-
ment V of the appendix. The report differentiates the various compo-
nents of the total road load force required for vehicle load simula-
tion and discusses the techniques which can be used to simulate this
load on both the water brake and electric chassis dynamometer.
The second project reported was a technique for determining the power
absorber curve by the use of torque wheel measurements. Wolfgang Berg
from Mercedes Benz presented the results of their development activi-
ties using torque wheels and some of the instrumentation problems that
can occur in the use of such a system. A copy of his report is in
Attachment VI of the appendix for your reference. The third report was
made by Glen Thompson of the Emission Control Technology Division of
the EPA Laboratory. This report summarized the results obtained by
comparing load/speed data from the track to that obtained on a Clayton
dynamometer using three modes of speed reference; one using rear roll,
the second using the front roll, and the third using the rolls coupled
together with a synchronizing chain drive. This Investigation con-
cluded that the coupled rolls provided the best simulation to the
on-road data. The elimination of the tlre-to-dyno slippage has an
associated Increased net loading on the vehicle with a resultant
decrease in fuel economy. A copy of this report is also attached in
the appendix for your reference.
Following the symposium, a short tour of the EPA test facility was
conducted for those visitors who had not seen the laboratory before.
-------
Index of Attachments
I. General Outline
II. Attendance Roster
III. Survey Questionnaire Summary
IV. Flywheel Friction Table (EPA)
V. Certification Printout (TP 202 Example)
VI. VW Report on Dyno Calibration
VII. Mercedes Benz Torque Wheel Report
VIII. EPA Report on Dyno Velocity Simulation
IX. EPA Verification Procedure - TP 302 (separate cover)
X. Quick Check Timer Circuit
-------
Attachmeiot X
EPA/INDUSTRY TESTING PRACTICES SYMPOSIUM
DYNAMOMETER CALIBRATION AND CHARACTERISTICS
August 16, 1979
9 a.m. - 4 p.m.
GENERAL OUTLINE
I. Introduction and Discussion of Symposium Program
II. Calibration Practices
A. Describe EPA TP-202, 207, 302
B. Data Collection and Instrumentation
C. Analysis and Acceptance Testing
III. Dynamometer Characteristics
A. Speed, Load, PAU Curves
B. Flywheel Frictional HP
C. Slip Characteristics and Effects
LUNCH 12-1
IV. Calibration Monitoring and Correlation
A. Site to Site Variability
B. Quick Check Test Results
C. Control Charts/Statistics
V. Special Projects
A. CEC Dyno Calibration Project
B. Alternate PAU by Torque Wheels
C. Roll Coupling
-------
/iTTACHMEislT X
EPA/INDUSTRY TESTING PRACTICES SYMPOSIUM
DYNAMOMETER CALIBRATION AND CHARACTERISTICS
August 16, 1979
ATTENDANCE LIST
Andrew Z. Sarosy
General Motors Proving Ground
Milford, Michigan 48042
685-5134
Daniel F. Sougstad
General Motors Proving Ground
Milford, Michigan 48042
685-5176
John M. Lang
Northrop Services
P. 0. Box 12313
Research Triangle Park,
North Carolina 27709
Arthur W. Fornell
American Motors Corporation
14250 Plymouth Road
Detroit, Michigan 48232
493-2959
Henry Marschner
Fiat N.A. R&D
S. 1210 Parklane Towers, West
One Parklane Blvd.
Dearborn, Michigan 48126
336-3515
Hank Bernstein
Alfa Romeo, Inc.
250 Sylvan Avenue
Englewood Cliffs
New Jersey 07632
(201) 871-1234
Taka Suzuki
Toyota Motor Co., Ltd.
1012 Pontiac Trail
Ann Arbor, Michigan 48105
769-1350
Robert Yeates
New Jersey DEP
380 Scotch Road
Trenton, New Jersey 08628
(609) 292-7640
Klaus W. Drexl
Mercedes-Benz of N.A.
One Mercedes Drive
Montvale, New Jersey 07645
(201) 573-2641
Arlo Gifford
EPA - Quality Assurance
2565 Plymouth Road
Ann Arbor, Michigan 48105
668-4239
Wolfgang Groth
Volkswagen of America
7111 E. 11 Mile Road
Warren, Michigan 48090
588-5779
Dieter Schurmann
Volkswagen Werk AG
318 Wolfsburg, Germany Abt. 1786
05361-226928
Heinrich Orlet
Volkswagen of America
3560 Marine Road
Toledo, Ohio 43609
(419) 382-0210
C.B. Tracy
Amoco Oil Co.
Box 400
Naperville, Illinois 60540
(312) 420-5925
George F. McCaskey
Chrysler Corporation
Dept. 7220, CIMS 417-30-31
12800 Oakland Avenue
Highland Park, Michigan
956-4959
Arnold Welbel
Chrysler Corporation
12800 Oakland Avenue
Highland Park, Michigan
956-2036
-------
-2-
Wiplove Juneja
General Motors Proving Ground
Milford, Michigan 48042
685-6084
David Gourley
Environment Canada
Air Pollution Tech Centre
River Road Labs
Ottawa, Ontario, Canada
(613) 998-9590
David Lawrence
EPA
2565 Plymouth Road
Ann Arbor, Michigan 48105
668-4347
Dale Turton
EPA - ECTD
2565 Plymouth Road
Ann Arbor, Michigan 48105
668-4237
Wolfgang Berg
Daimler-Benz
7000 Stuttgart
(West Germany)
(711) 302-3717
- 60
Robert Gilkey
EPA - QC
2565 Plymouth Road
Ann Arbor, Michigan 48105
668-4397
Paul M. Reece
EPA - LAB Correlation
2565 Plymouth Road
Ann Arbor, Michigan 48105
668-4254
Tom Chamberlain
U.S. Department of Energy
P. 0. Box 1398
Bartlesville, Oklahoma 74003
(918) 336-2400
Stephen A. Riggs
Olson Engineering, Inc.
3901 Varsity Drive
Ann Arbor, Michigan 48104
973-0310
Virgil Woodard
IHC TEC
2911 Meyer Road
Fort Wayne, Indiana
(219) 461-5257
46815
Carl Paulina
EPA - Correlation
2565 Plymouth
Ann Arbor, Michigan 48105
668-4489
Alan Pearson
IHC
2911 Meyer Road
Fort Wayne, Indiana 46815
(219) 461-6878
Emmet Arndt
EPA MSED (En-340)
401 Main Street, S.W.
Washington, D.C. 20460
Roy Reichien
EG&G Automotive Research, Inc.
6204 Gravel Avenue (Franconia)
Alexandria, Virginia 22310
(703) 971-8823
(VTL)
Joseph Hammel
EPA - Quality Assurance
2565 Plymoth Road
Ann Arbor, Michigan 48105
(313) 668-4355
Aaron R. Martin
EPA MSED (EN-340)
401 Main Street, S.W.
Washington, D.C. 20460
(202) 473-9430
-------
-3-
Rlchard Morgan
Allied Chemical Corporation
353 Cass Avenue
Mt. Clemens, Michigan 48043
469-2900
Takasui Yonemoto
ERI
3995 Research Park Drive
Ann Arbor, Michigan 48104
665-2044
Kenlchl Sasabe
ERI
3995 Research Park Drive
Ann Arbor, Michigan 48104
665-2044
Ken Takahashl
Nissan Motor Co., Ltd*
c/o ERI
3995 Research Park Drive
Ann Arbor, Michigan 48104
665-2628
Loy Morrison
U.S. DOE
P. 0. Box 1398
Bartlesville, Oklahoma 74003
(918) 336-2400
George V. Shaw
Chrysler Corporation
Chelsea Proving Ground
Box 387 Chelsea, Michigan 48118
(313) 475-8651, Ext. 435
Wllhelm Hall
BMW - MUHICH
Betueiring
8 Munich 40,
North Germany
(089) 3895-4630
Robert H. Gower
Chrysler Corporation
12800 Oakland Avenue
Highland Park, Michigan
956-1640
William H. Drew
Chrysler Corporation
12800 Oakland Avenue
Highland Park, Michigan
956-3610
Steve Whittier
Ford Motor Company
Scientific Research Lab
Room E-2212
P.O. Box 2053
Dearborn, Michigan 48121
323-3301
Bruce P. Gardner
Ford Motor Company
Garrison Place, West - 3rd Fl. 4
19855 W. Outer Drive
Dearborn, Michigan 48124
594-2234
Robert E. Rice
Chrysler Corporation
Chelsea Proving Grounds
Box 387
Chelsea, Michigan 48118
475-8651, Ext. 455
T. A. Rhoad
Ford Motor Company
AEFEO - ETL, Room G028
21500 Oakwood
Dearborn, Michigan
322-3167
Michael H. Goemer
Ford Motor Company
Allen Park Test Laboratory
1500 Enterprise Drive
Allen Park, Michigan
322-3458
-------
-4-
Steve Rossi
SAAB - Scania of America
SAAB Drive Orange Court 06477
(203) 795-5671
James Kautzer
American Motors Corporation
5626 - 25th Avenue Exem. Boulevard 45
(414) 658-6641
Daniel A. Reis
Jeep Corp. (AMC)
940 North Cove Boulevard
Toledo, Ohio 43609
(419) 470-7453
Daniel R. Wamboldt
American Motors Corporation
5626 25th Avenue 53142
(414) 658-6332
Michael Crawford
American Honda Motor
100 West Alondra Boulevard
Gardena, California 90243
(213) 327-8280
Takesiumi Hosaka
American Honda
3947 Research Park Drive
Ann Arbor, Michigan 48104
(313) 994-8441
John Glaser
Ford Motor Company
AADGO Room D2010
17000 Oakwood Boulevard
Allen Park, Michigan
427-7128
Shinkichi Nakashita
Toyo Kogyo Company Limited
23777 Greenfield Road, Suite #462
Southfield, Michigan 48075
(313) 559-5040
Larry Flegal
Clayton Manufacturing Company
24750 Swanson Road
Southfield, Michigan
(318) 354-2220
Jack Wilson
Clayton Manufacturing Company
4213 N. Temple City Boulevard
El Monte, California
(213) 443-9381
Norman D. Pranger
Chrysler Corporation
P.O. Box 1118
Detroit, Michigan 48288
CLMS CODE 417-30-31
(313) 956-1239
Richard A. Middleton
Chrysler Corporation
37005 Chelsea - Manchester Road
475-8651 Ext. 381
Allen White
Chrysler Corporation
Chelsea Proving Grounds
3700 South Chelsea - Manchester Road
Chelsea, Michigan 48118
(313) 475-8651 Ext. 381
Marvin Frinkle
Chrysler Corporation
Chelsea Proving Grounds
3700 South M52
Chelsea Michigan 48118
(313) 475-8651 Ext. 462
N. E. March
Chrysler Corporation
P.O. Box 1118
Detroit, Michigan 48288
(313) 956-4892
-------
Attachment JET
DYKAhOMETER CHARACTERIZATION QUESTICMiAIRE
This questionnaire ha9 been developed to characterize the various
physical components of the dynamometer system. It has been written
around a Clayton ECE-50 (125$) dyno, but if you have data or informa-
tion that can be supplied for other dynamometers it may be valuable
for comparitive purposes* Many questions relate to how tests and
analyses are done; if the space provided is insufficient to give an
abbreviated answer, please use separate pages or send a copy of your
company's procedure* If you do not have data readily available that
is pertinent to a question or can't interpret the question as stated,
mark it as N/A* If you have a particular question that you would like
discussed at the symposium, please include it on the last page o£ the
ques tionnaire•
The responses to these questions will be tabulated for discussion at a
dynamometer symposium which will be held at the EPA Laboratory on
Thursday, August 16, 1979* Your response to this questionnaire is
voluntary* However, the generation of a more complete tabulation will
facilitate a more thorough discussion of this subject* Please send
your response to:
Quality Assurance Staff
EPA Laboratory
2565 Plymouth Road
Ann Arbor, Michigan 48105
If you have questions, please call us at 668-4342, 668-4239, or
668-4355.
-------
DYNAMOMETER CALIBRATION PROCEDURE
1) Describe in general terms the calibration procedure used for your
dynamometers, such as thumbwheel values, inertias used, repetitions,
warmups, etc.
GM
Dynamometer warm-up is performed by driving a vehicle at 50 mph for 15 minutes
on the dynamometer with the 6875 pound inertia test weight engaged and an
indicated horsepower setting of 10.0 hp.
All inertia test weights used are calibrated at the three indicated
horsepowers listed in Table 1.
Three coast downs are performed at each horsepower setting. These coast down
times are averaged to calculate actual horsepower using the following equation:
Actua, horsepower - 0.06°73 « (pounds)
Table 1
Indicated Horsepower Table
Inertia Test Weights
(pounds)
Verification
Indicated
Horsepower
Calibration
Indicated
Horsepowers
6875, 6750, 6625, 6500
12.0
18.0, 12.0,
6.0
6375, 6250, 6125, 6000
25.0
30.0, 25.0,
20.0
5875, 5750, 5625, 5500
11.0
16.5, 11.0,
5.5
5375, 5250, 5125, 5000
10.5
15.8, 10.5,
5.2
4875, 4750, 4625, 4500
10.0
15.0, 10.0,
5.0
4375, 4250, 4125, 4000
9.4
14.1, 9.4,
4.7
3875, 3750, 3625, 3500
8.8
13.2, 8.8
4.4
3375, 3250, 3125, 3000
8.0
12.0, 8.0,
4.0
2875, 2750, 2625, 2500
7.2
10.8, 7.2,
3.6
2375, 2250, 2125, 2000
6.4
9.6, 6.4,
3.2
-1-
-------
Coastdowns at 5 thumbwheel settings per inertia calibrated - 22
inertias; waraiup min. 30 minutes at start, other warmups as needed,
- 50 MPH; 3 repetitions at first thumbwheel per inertia repeat
within 0.2 sec., one coastdown at remaining 4 thumbwheels -
Evaluating "Prime" wheel technique.
CHRYSLER
Complete calibration of each dynamometer performed each
month:
Inertia Wt. Thumbwheel values
2250-2500 4, 6, 9, 12
2625-3500 5, 8, 12, 16
3625-4500 6, 10, 14, 18
4750-5500 6, 11, 15, 20
Coastdowns are performed over the range of 2250-5500
lb.IW, at the increments required for certification testing.
Three repetitions at each IW setting must agree within 0.2
sec. Warmup for 10 min. at 50 mph at 5500# IW before start-
ing coastdowns.
jpjQ Warm up with all inertia weights engaged for
20 minutes, select test inertia, strobe rear and front rolls at
1,800 rpm and observe speed on digital readout. Adjust Thumbwheel
Settings every A hi.p. from A to 20 horsepower. Repeat each setting
a minimum of 2 times. Calculate actual horsepower.
T0Y0 KOGYO
50mph set Power; U-15 HP(U-5 points) Inertia; 2000-3500 lb
Repetitions; 2 or 3 times each power set
Warm up; 15 minutes at 30mph
-------
NISSAN
wanjiup : ids. at lease naiiiite.s
Inertia used and thumbwheel values:
At rape.
Inertia,.
Thumbwheel value
2250-1125
Repetition : repeat to obtain 3data
(every 125 lbs)
5, 7, 9, 11
within ±0.05 AHP
3250-3750
fevp.fv 250
•7, 9, 11, 13
3750 lbs
17, 19, 21, 23
m.
I ions, varirujjs, etc- Warmups: 4875
1
Inertia used and thumbwheel values;
a'ceitan if the indicated HP is within
- .
+ 0.1 HP of thumbwheel value at each
point. Repetition: repeat to obtain 3
30 minutes at L;0 mph^
INERTIA THUK£V?HEEL YALU£ ir.??. '
3750 lbs. 5.0, 10. 6~T5TbT~2• 0
3500 ~ 3000 {every" 5.0, 10.0, 15.0 T j
125 lbs) !
2875 ~ 2000 (every 2.0, 6.0, 10.0 j
125 lbs) ! I
data within + 0.05 AHP.
EPA 15 min @ 6875 TW = lOHp
1000 - 2875 3,6,10 TW 3 repeats on warmup
33 S 3000 - 4750
5,10,15
TW
1 coastdown @ each
IW's I
\5000 - 6500
10,16,22
TW
TW value
Read & record indicated HP @ FR50 for
each TW. & integrate torque
& speed. Data analysis with computer. See TP202.
2) Describe the curve fit method used to reduce the data.
GM
A least squares straight line is fitted through the three points of actual
versus indicated horsepowers. From these coefficients, a look-up table of
actual versus indicated horsepowers is tabulated in 0.1 hp steps. The
deviation of any data point from the straight line must not be greater than
0.3 hp.
-------
FORD, CHRYSLER, IHC, TOYO KOGYO, NISSAN, and ERI
Linear least squares fit of thumbwheel values versus the
actual horsepower obtained.
EPA
LLSQ QC Validation
T*S vs AHp (IW-155), TW vs AHp (IW-155) and RRFhp
TW vs AHp (FUUjIW) for Cert Table. No RRFHp used.
3) What other analyses are performed as part of the calibration data
reduction, such as prime inertias, average slopes, or repeatability
measurements?
GM
A slope check is performed monthly on the 6875 pound inertia test weight to
compare current slope to the slope at the time the look-up table was
generated. The equation used in this comparison is:
i Difference., s1°Pe (currentj^slope (old) x m
If the slope is witiiin +3%, a spot-check is performed using the verification
indicated horsepowers (Table 1) for all inertia test weights used. The .
spot-check requires the actual horsepower to be within +0.4 hp of the current
look-up table value. If a spot-check fails, only that Tnertia test weight
needs to be recalibrated. If the slope check fails, complete recalibration of
all inertia test weights is required.
The three coast down times at each indicated horsepower during a spot-check,
slope check or calibration must be within 0.3 second of each other, or the
coast downs are repeated until a group of three repeat within 0.3 second.
-------
FORD
FHPn^w vs- FHPni^ 1 °-5 hP> sl°Pe = 1.000 ± 0.015, actual vs.
calculated + .3
CHRYSLER .
The following additional calculations and tolerances are
performed.
a) difference between actual coastdown HP and calculated
HP from curve fit must be < 0.15.
b) slopes (IW -155) are printed out in a summary and aver-
aged. No tightly defined tolerance, but average slope
should be .950 to 1.050, and individual slopes should be
reasonably close to average.
c) Prime inertia method has been looked at/ but is not being
used to generate coastdown data. Frictional contribution
is calculated from 7 weights (^2250, 3000, 3125, 3250,
3500, 4000, 5000). Data has bpen- summarized for informa-
tional purposes, but has not been used to generate horse-
power settings at different inertia weights.
IHC
Calibration curves are compared to previous calibration and
changes noted.
NISSAN and ERI Check the average slopes and the frictional'
horsepower of each incrtio coinbinations.
EPA
PRIMES, SLOPES, o, %C.V., 5 RR Ats, speed curves, torque curves,
PAU exponent, T*S vs TW, IHp @ 50 vs TW< FW FHp analysis.
Toyo Kogyo replied N/A.
-------
COASTDOWN. TIMER FUNCTIONS
1) Describe your coast down timing circuit.
GM
The coast down timer is a microprocessor based system that utilizes a 160
pulse per revolution optical encoder coupled to the dynamometer front rolls.
The instrument is programmed to determine when a coast down is in progress.
At the completion of a test, the coast down time is displayed. Speed is also
continuously displayed for the purpose of checking the indicated horsepower
setting at a 50 mph front roll speed.
FORD
Describe your coastdown timing circuity • Microprocessor
operating from Front Roll one pulse/rev, signal, measures period
of revolution vs. period_for 55 & 45 mph.
CHRYSLER
The front rolls tach generator voltage is monitored by a
computer which triggers a timer at voltages corresponding to
55 and 45 mph. The computer prints out the elapsed time
between the two speeds.
IHC
Speed tachometer and stop watch.
NISSAN.
Describe your coastd^vn tiding circuit. a .pulse disc 13 mounted on the
front roller shaft(60 pulse/rev). The pula'e 9ignal is compared witn design
pulse-wide(55™ph .and 45moh) in the comparator a^d timer eate Is driven.
The gate open time is neasured with 1 MHZ cldck signal automatically and
ri-Upl flr-qd on the digital counter.
ERI
Describe your coastdown tiding circuit. The computer scans the
front roller tach-aenerator signal every 1/40 sec. and calculates
the coastdown time comparing it with the stored values of S5mph & 45 mph.
-------
TOYO KOGYO
PIG 1 Coast down timer circuit
60 tooth-
nagnetic
pickup
Coast down timer
Speed set
: © ©
Start signal
Stop signal
Pulse out
Analogue out
Universal counter
Coast down time
0 0 0 0 0 (ms)
Universal counter
Pen record
-er
Pulse integrate
0 0 0 0 0
©-
In put
S»trigger
circuit
Analogue out
set
EPA
Describe your coastdovn tiding circuit. Analog comparator.
See detailed circuits in Attachment 6.
-------
2) What is the precision of this circuit using a clean ramp signal?
GM
The units' specifications are:
Display Resolution
Time - 0.01 second
Speed - 0.1 mph
Accuracy
Time - within one roll revolution;
e.g., one revolution at 50 mph equals
30.8 milliseconds.
Speed - trimmed to display exactly 50.0 mph
FORD
Tins + 0.01 sec; 4 MHz clock; 0.2% accuracy for speed detection.
CHRYSLER
+ 0.1 mph, and + 0.2 sec.
TQYO KQG¥Q_
TA3LE 1 "Inspection results
j.n cut
mDh
Kz
Analogue
out
Vorts^fe.
V
trj*.
. t*
100(388610
908497 ~
80(3109
70^720
602392
55^137
501943
450.7^9
4oa.551<
•301166
20
_1Q
777
JM
OOlp-o.Ol
003^0 *03
. 002*0.02
-OOlk-O.Ol
.C03*f0.03
.501jf0.01
.C02j*0.02
.503*0.03
• QQO! 0
~ 000| 0 -
- 999(-0.01
,o,o.i!f.n.ni.i
±0.03?
Set
inch
Start sli
ou
;nal
*
Stop signal
out
Kz
Speed
mph
Erj.
r»
M-z
moh
Erj.
to
100
2°
80
70
60
55
50
^5
40
30
20
10
3878
b*91
3104
2720
2332
?137
1945
1750
1555
1168
780
"*87l
99-80
09.84
79.88
70.00
60.01
55.00
50.05
45.04
40.02
30.06
20.07
9 - 96
-0.20
-0.16
-0.12
0
l-O.Ol
0
50.05
~•0.04
+0.02
*0.06
*0.07
=0.04^
^886
3498
3107
2720
2331
2138
Jl944
1749
5.556
1168
778
387
LOO.01
90.02
79.96
70.00
59.99
55.02
50.03
45.00
40.04
30.06
20.02
9.96
>•0.01
1-0.02
-0.04
0
-0.01
KJ.02
*¦0.03
0
+0.04
+0.06
1-0.02
-o.ofc
fhS
±0.072 (Less than 7
Ornph)
-------
NISSAN
Delay time : les9 than 0.0006 sec.
ERI
Using 12 bits for A-D conversion, system readability is 0.02 mph & 1/40 sec.
EPA
±.005 sees
IHC replied N/A.
3) What are your set point accuracy or hysteresis criteria?
GM
Since the timer uses a digital pulse train, no hysteresis criteria is
required. The optical encoder has a gray code output, which eliminates any
jitter in the output. The microprocessor program is loaded for each
dynamometer roll diameter to trigger on the exact digital comparison for 55 to
45 mph interval. The timer system runs off a 5 Megahertz crystal clock.
FORD
53.00 + 0.01, 45.00 + 0.01
CHRYSLER
+ 0.1 mph
TOYO KOGYO
-0.03 mph (55-^5 mph)
-------
NISSAN
Using 60 pulse/rev signal, less than 0,001 sec.
ERI
The speed set point accuracy is +0.02 mph because of its, readability.
The speed calibration has been made in steady state using the strobe.
EPA
+.002 VDC 4.500/5.500 HYS <.005
IHC replied N/A.
4) What types of integration or sampling techniques are used for torque,
speed, or horsepower?
GM
Speed is continuously updated in a shift register until it agrees with the
preset start and stop coast down speed values. Torque and horsepower are not
integrated over the coast down.
FORD No "integration
IHP checked vs. dial @ 50 mph. must be within + 0.2 hp, speed
indication checked vs. microprocessor - ^ 0.1 mph.
CHRYSLER .
The front rolls tach generator voltage is monitored by a
computer which triggers a timer at voltages corresponding to
55 and 45 mph. The computer prints out the elapsed time
between the two speeds.
-------
EPA
torque, speed, or horsepower? torque VCO ->• TOTALIZED
SPEED -*¦ VCO -*¦ TOTALIZE. Could use 60 tooth gear.
IHC, Toyo Kogyo, Nissan, and ERI replied N/A.
5) What is the repeatability of dynamometer coast downs at TW = 10,
IW = 6875?
GM
Repeatability was within +0.25 hp as observed for two sites over an 8-month
period.
FORD
Better than 0.2 sec.
CHRYSLER
Coastdown times are repeatable within +0.2 seconds
TOYO KOGYO
Coastdovrn time^_lU.358,lU.305,1^• 376 sec. (IW=3000 lb,Power=10 HP)
NISSAN
At TW = 10,TW *= 5375 lb9,n = 10, coastdown time : ¦ 0.13
(AHP :(T- 0.06)
EPA
Coastdown time: 31 ±0.1 sees
-------
IHC and ERI have not collected any data on this.
6) Have you assessed the set point stability for the coast down timer?
GM
Digital comparator does not have any instability.
FORD
No detectable.set point drift over 6 months.
CHRYSLER
Tach generator voltages (front rolls speed deflections) have?
been found to stay constant over periods of six months or
more, unless the tach generator is changed.
NISSAN
i 3 x 10 ^ sec/day
ERI _ .
-J^^ccmputer_j^_gi.yen_thfi_trig g.e n_sp_e ed_y_a luas_fxo.nL ^_spf!sd_cjLLibxa t i on.
-1 ine.-_—Th.is._g3JLibration @ 46. 32mph has been made within +0.1% for a year
EPA
.Set points 4.5 - 5.5 Volts, with a drift of ±2 mv/month, checked
monthly reset if > ± 5 mv, variation : 3 mv, usually less than 1 mv
IHC and Toyo Kogyo replied N/A.
-------
7) What effect do the uncertainties associated with your triggering function
have on the coast down time?
GM
A digital system does not have any uncertainties in triggering the start and
stop function.
CHRYSLER
They give variations in coastdown time of up to +0.2 seconds.
EPA
The calibration of tach generation is held within 0.1% @ 50 mph.
Ford, IHC, Toyo Kogyo, Nissan, and ERI have no data on
uncertainties associated with triggering function.
8) Do you apply any adjustment of speed data for triggering errors?
GM
The design characteristics of this digital coast down timer do not require any
adjustment to be performed.
ERI
The 6 points speed data are averaged for comparison with the trigger
value to avoid the fluctuation error of the tach-generator signal.
Ford, Chrysler, Toyo Kogyo, IHC, Nissan, and EPA make no
adj ustment.
-------
RLPC CIRCUITS
1) Describe the signal readouts incorporated with your dynamometer?
GM
The two digital meters above the RLPC display front roll speed in mph and
torque in ft-lbs. The Driver's Aid has a digital horsepower display and
selectable digital front or rear roll speed display.
FORD
Digital displays of rear roll speed, front.Toll speed,
indicated horsepower, and torque.
CHRYSLER . .
Front/Rear Rolls Speed and Horsepower is displayed on Drivers Aid
Stand, digital meters with a resolution of .1 MPH and •! HP.
IHC '
Digital read out.
ERI
The CLAYTON ECE-50 RL?C standard type (Model No. A-32206) without
reconstruction is used.
EPA
Digital speed, selectable front or rear roll hp always based on
front roll speed. Vehicle factor pot not used, hp is varified at
PAH
Toyo Kogyo, and Nissan replied N/A.
-------
2) Do you have control limits for the power absorber exponent? How is it
assessed?
GM
At installation, the exponent module was set at 2.0 using Clayton's
procedure. Monthly an X-Y-Y plot of torque and horsepower versus speed are
compared to the theoretical plot (see Figure 1) when driving the second cycle
of the FTP schedule. The RLPC electronics are adjusted, if necessary, to
minimize hysteresis and exponent level changes.
—CHRYSLER
No special control limits. The Load and Unload functions and dead-
band are calibrated monthly per RLPC procedure in manual R-8713.
ERI
No. It was confirmed that the*power absorber
exponent on the condition of trangient and steady state is 3.00 +
0.0 8 at a speed of 20 to 60 mph.
EPA Control limits: 1.85-2.15. Tvplrallv 2.0 in
automatic mode. 11 pt. Steady State PAU curvj, torque and speed
data used in log-fog
Ford, IHC, Toyo Kogyo, and Nissan have no control limits.
3) Are any tests performed to validate the accuracy of the thumbwheel signal?
GM
Yes, thumbwheel span is calibrated at one span value of 39.0 hp. New
thumbwheel switches have been installed on the Driver's Aids that have the
outputs of the three decades summed and weighted. These digital switches have
had their output signals verified for linearity at installation.
FORD Drive 50 mph front roll speed 5 observe IHP -
must be within + 0.2 of thumbwheel; regress thumbwheel vs.
AHP ("corrected for rear roll) - slope within 1.000 + 0.015
-------
CHRYSLER
Lo Span and High Span adjustment of Thumbwheel per RLPC procedure
monthly.
IHC
Thumbwheel signal is calibrated per Clayton procedure.
The agreement on the thumbwheel set value and the
digital readout is checked to be within + 0.1 HP at 50 mph.
EPA
Only_ during electronics calibration as specified Clayton.
Toyo Kogyo and Nissan replied N/A.
4) How is the indicated horsepower signal calculated?
GM
For the purpose of dynamometer calibration, the indicated horsepower is
defined to be equal to the thumbwheel setting.
For the purpose of defining the theoretical torque and horsepower versus speed
plot (Figure 1), the indicated horsepower is calculated as follows:
Indicated horsepower = ^or{lg?t.? rPm
FORD
Front rolltorque times front roll speed.
CHRYSLER
Board 8 of RLPC electronics calculates indicated horsepower. Torque
cell is calibrated and the linearity of torque cell is verified per
RLPC monthly calibration procedure.
-------
hc—iiee
ee
50
LlI
a.
Ck
UJ
cc
GQ
40
3d
H
LJ
ZD
O
oc
o
I—
II
X
20
1 0
Figure 1
CLAYTON DYNAMOMETER ABSORBER CHARACTERISTICS
TECHNICIAN
SITE NUMBER
DATE/TIME X ^ • .
inertia wt.-d&fia_Las
ACTUAL. HP
INDICATED
MODE CAUTO/MAN5
COMMENTS
e-3r
X 7
7
7
7
G>
Cv
1 a
.20 25 30 3E 40
FRONT ROLL SPEED IN M.P.H.
/
X CAL
+ CAL
"7
7
7
7
/
"7"
. 30
> 25
* r
j \
• 20
t 0
45
ESQ
65
ee
r\
2
LJ
Cl
LJ
r>
_i
CO
cm
LJ
3
O
CL
LU
CO
en
o
X
Cl
LJ
h-
<
O
H
Cl
2
H
II
+
i
vo
I
Z
o
-------
IHC
Composite of front roll speed and torque signal.
ERI
Clayton's RLPC standard type without reconstruction is used.
EPA
How is the indicated horsepower signal calculated? EPA rewired
circuit to always use front roll speed and torque for hp calculation.
(Standard Clayton circuit uses front or rear roll speed as selected
on control.J
Toyo Kogyo, and Nissan replied N/A.
5) What limits do you have for deadband control in the automatic mode and
what problems have you encountered in meeting these limits?
GM
The deadband adjustments are performed monthly under static conditions. The
low rate deadband is adjusted to be centered exactly at 3.0 hp and 39.0 hp and
to be no wider than +0.2 hp. The high rate deadband is adjusted to be
centered at 39.0 hp and to be no wider than +1.5 hp.
The problem with adjusting these deadbands is that they interact and several
iterations are sometimes required. The driver cut-off module is being
adjusted prior to the deadband adjustments to eliminate another potential
interaction (more appropriate than Clayton's procedure).
Thumbwheel +_ 0.2 hp at all thumbwheef settings.
(1) Precise centering of deadband is difficult.
(2) Deadband width below 6 hp tends to be less than 0.2
CHRYSLER
.4+0 H.P. Differences in calibration techniques existed when equipment
was first installed. No problems in obtaining .4 HP deadband. A
procedure modification was made to Clayton low range adjustment and
driver cut off adjustment to meet .4 HP deadband. See Procedural Attachment.
-------
CHRYSLER
Modification to Clayton Calibration Procedure
This change will make possible the .4 HP difference tolerance specified
in the Thumbwheel low range adjustment.
A. Recommend adding the following procedure to the Thumbwheel Low Range
Adjustment after Para 4-16-e-2.
NOTE: If unable to obtain a .4 difference in the thumbwheel low
range adjustment then perform the following steps.
a. Adjust P6/B8 to obtain a reading of 2.8 HP on the Drivers' Aid
Power Meter.
b. Adjust P1/B7 until low rate load light just comes on.
i -
c. Adjust P6/B8 to obtain a reading of 3.0 HP on the Drivers' Aid
Power Meter and observe that low and high rate load and unload
lights are out.
d. Adjust P6/B8 to obtain a reading of 3.2 HP on the Drivers' Aid Power
Meter.
e. Adjust P2/B3 unitl low rate unload light just comes on.
f. Adjust P6/B8 clockwise until the low rate load light just starts
to pulse. Record the power reading.
g. Adjust P6/B8 counter-clockwise until the low rate.unload light just
starts to pulse. Record the power reading.
h. If unable to obtain a .4 difference then repeat steps 4-16-a through 4-
B. Recommend Para 4-19-e of the Driver Cut-Off Adjustment to read as
stated below: .
Adjust P2/B3 clockwise to jsut stop low and high rate unload lights.
This adjustment should vary no more than 1/8 turn from its original
setting.
-------
CHRYSLER
Modification Co Clayton Calibration Procedure
This change will make possible the .4 HP difference tolerance specified
in the Thumbwheel low range adjustment.
A. Recommend adding the following procedure to the Thumbwheel Low Range
Adjustment after Para 4-16-e-2.
NOTE: If unable to obtain a .4 difference in the thumbwheel low
range adjustment then perform the following steps.
a. Adjust P6/B8 to obtain a reading of 2.8 HP on the Drivers' Aid
Power Meter.
b. Adjust P1/B7 until low rate load light just comes on.
c. Adjust P6/B8 to obtain a reading of 3.0 HP on the Drivers' Aid
Power Meter and observe that low and high rate load and unload
lights are out.
d. Adjust P6/B8 to obtain a reading of 3.2 HP on the Drivers' Aid Power
Meter.
e. Adjust P2/B3 unitl low rate unload light just comes on.
f. Adjust P6/B8 clockwise until the low rate load light just starts
to pulse. Record the power reading.
g. Adjust P6/B8 counter-clockwise until the low rate unload light just
starts to pulse. Record the power reading.
h. If unable to obtain a .4 difference then repeat steps 4-16-a through 4-
B. Recommend Para 4-19-e of the Driver Cut-Oirf Adjustment to read as
stated below:
Adjust P2/B3 clockwise to jsut stop low and high rate unload lights.
This adjustment should vary no more than 1/8 turn from its original
setting.
-------
IHC _Dc;ad band _cont_rol l_imits are j^_0-2_HP._ We sometimes
have experienced the fluctuation of torque at steady state.
ERI
i + .2 hp — no problems.
EPA * - "r - TM. g;Wl3- 30
39 points. Use a little wider 6 30.
Toyo Kogyo and Nissan replied N/A.
6) Do you utilize the load cell shunt calibration check function on the
Clayton dyno?
GM
Yes, to verify static horsepower calibration. If this check is within +0.1 hp
of the tagged reading (reading after last complete deadweight calibration), a
deadweight calibration is not required. If the difference is greater than
+0.1 hp, a complete static torque cell and horsepower calibration (deadweight
calibration) is performed according to the Clayton procedure.
FORD
Yes - used as diagnostic to assess load cell
calibration when quality control performance check is failed.
CHRYSLER
Calibration check function is recorded as part of monthly calibration
but is not used by operational personnel for comparison check.
The calibration check button is used every
week and the indicated value is recorded as reference.
-------
EPA Performed:at cal. and logged on data sheet.
Checked for diagnostics or during monthly maintenance.
IHC, Toyo Kogyo, and Nissan do not.
7) Describe the failures, retrofits arid general problems that you have
encountered with respect to the Clayton road load power control circuits.
GM
The two types of components that have failed the most frequently are the
multifunction generator (4302) and the triacs.
Any component failure on boards 8, 7 or 6 requires reoptimizatton of all three
boards, because of circuit loading.
P0RD (1) Found some factory supplied circuit boards
to be out of calibration, (2) Frequent failure of SCR's on boards
2 5 3, (5) Drift of front roll speed reference signal, (4) Circuit
boards #6, 7 § 8 come in two configurations.
CHRYSLER
No significant failures. Procedural change as noted above.
EPA
Added R/C filters; IHp is based on FRVDC. Tach" brush tension
can cause high A/C ripple. B7,B8 mods., Load cells digital meters.
IHC, Toyo Kogyo, Nissan, and ERI have had no problems'.
8) highlight any solutions that you have developed to these problems.
GM
The following modifications have been added to the RLPC system to increase its
reliability and maintainability:
• Added zero horsepower switch to facilitate checking the horsepower
meter electrical zero level.
-------
GM
t Replaced all original triacs with a device which has higher voltage
and current ratings.
• Replaced original indicated horsepower thumbwheel switches with a
unit that has only one analog output and performs decade weighting
internally.
• Modified all RLPC units to be current with Clayton's latest circuit
changes. These changes are as follows:
(1) Increased resistor R14 on board 7 from 50K ohms to 200K ohms.
(2) Removed R18 (10K ohm resistor) from board 7.
(3) Removed R35 (10K ohm resistor) from board 6.
General Motors is currently working on an improved method of optimizing the
Clayton RLPC system.
FORD
(11 Have developed calibration and repair procedures for all
Clayton circuit boards, (2) Alternate source for SCR's.
CHRYSLER
See Procedural Attachment.
EPA
Exponent control could be improved.
IHC, Toyo Kogyo, Nissan, and ERI have no solutions to these
problems.
,TACH GENERATORS
1) What method is incorporated to assess the relationship between tach
voltage and true speed?
GM
A strobotac is used monthly to synchronize the dynamometer rolls with the
front and rear roll tachometer voltage outputs. These voltage levels are
adjusted to read exactly what the calculated roll surface speed is for that
roll diameter at 1800 rpm.
-------
FORD ...
Tach. voltage vs. microprocessor (i pulse/rev.j
CHRYSLER -
Utilize a photo cell pickup into a crystal controlled counter for direct
speed readout.
IHC
Strobe light_per Clayton procedure.
TOYO KOGYO
nn gt.ro-hotach or digital tach generator
NISSAN Strobe light which syncronlzes
to line frequency (50±0.2Hz) and pulse signal from pulse gear
fitted on the roller shaft are used.
Strobe light which uses line
current (60 Hz) is used. Model type: Clayton's D-6784
EPA
Count 60 tooth vsfVDC» vm -»• TOTALIZED/lQ]
2) What tolerance do you place on this relationship?
GM
The speed signal is trimmed to read exactly the roll surface speed at 1800 rpra.
FORD
Speed meters: + 0.1 mph, RLPC: +_ 0.002 VDC
-------
CHRYSLER
With a resolution of .01 MPH, the relationship between true speed and
tachometer speed is less than .1 MPH.
TOYO KOGYO
Less thea +1?
NISSAN
±0,3 cile/h
ERI
4.632 + 0.002 volts at 46.32 mph aftd O.ODO + 0.002 at 0 mph.
Boiler speed is synchronized with strobe light signal {1800 rpmj
EPA
± l.Z & random 50 = 5.000 ± .010 VDC.
IHC has no tolerance.
3} Do you have limits or control for tach generator A/C ripple?
GM
RLPC system has low pass filter to reduce effects of A/C ripple. Maintenance
is performed if effects are visually observed on Driver's Aid pen.
CHRYSLER
AC ripple is recorded at monthly cal on both Tach generators. Runs
between 100 to 300 MV - RMS
TOYO KOGYO
Less than + 1?
NISSAN
So adjust amplifier as less thau +0.13 mile/h on the Driver Aid.
-------
tPA
40 mv! @ 50 mph.
100 mf/5k ohms
4) What is the stability of the tach generator amplifier signal?
GM
The analog speed signal is checked against the digitally derived speed from
the optical encoder prior to every highway fuel economy coast down test.
These measurements must agree within the +0.1 mph resolution of the speed
meter. No discrepancies have been noted.
FORD
Poor - generally the cause of failed performance checks.
CHRYSLER
Stable D.C. output into RLPC amplifier and to Process Computer System (PCS)
NISSAN
± 0.5Z / 8hra
EPA
Good.
IHC, Toyo Kogyo, and ERI have no data on the stability.
5) Describe the maintenance problems you have had with tach generator
c i rcu i ts.
GM
No maintenance problems have been noted with the tach generator circuit. Some
instability problems have been noted on the Driver'^'Aid pens and were found
to be caused by tach generator pulley wear and misalignment.
FORD
None - stability problems are in RLPC.
-------
CHRYSLER
Normal wear out as is expected of rotating devices. Replace very seldom
IHC, Toyo Kogyo, Nissan, ERI, and EPA have had no problems.
6) Describe in general the procedure used to zero and set the gain for the
tach generator (reference signal).
GM
The output of the tachometer circuit is zeroed with the rolls stationary.
Span is adjusted by substituting precision voltage supply set to the output
level of the tach generator which was read at the strobe speed.
for the tach generator (reference signal)* Rolls stopped,
brake off, adjust B8-P3 to 0.000 + 0.002 VDC; drive 50.0 mph
by microprocessor - adjust B8-P4 to 5.000 +_ 0.002 VDC
CHRYSLER
Run .vehicle at 50.0 MPH (after rolls warm up) and set voltage on
Fluke DVM to 5.000 VDC + .002 (Both front and rear tachometers) Monitored
zero must be 0.000 + .002
IHC
Clayton Calibration Procedure — Adjust zero and set gain
for A.63 volts 01.800 rpm.
NISSAN
for tha tach jjeneracor (rtfcccr.ce signal). Comparing the Speed
signal with pulse generator signal or the speed determined by strobe
light,ad1u9t the constant of the amplifier.
ERI
Clayton's manual has been followed.
-------
ZERO, 1800 rpm Ext. SYNC STROBE TO SET TACH @ 4.632 VDC, then ZERO
AGAIN & RESPAN
Toyo Kogyo replied N/A.
7) Have you assessed the linearity of the tach generator signal?
GM
Yes, tachometer output was checked at installation and periodically as a
diagnostic. The linearity is checked at 5 mph steps from 0 to 60 mph.
FORD . .
Rear tach only in driver's aid calibration.
chrys le r
Linearity front and rear tachometers are measured monthly from 0 to 60 MPH
in 5 MPH increments. IE the band of tach linearity exceeds +.1 MPH the
tachometer is replaced.
TOYO KOGYO
Less than +1$
NISSAN „
yep .we have. 20 £ U.3 mph, 38 £ 0.3 tnph,58 i 0.3 ®P".
ERI
23.2 mph (900 rpm) _+ 0.1 mph and 69.5 mph (2700 rpm) 0.1 mph
are checked monthly, as well as 0, 46.32 mph, with strobe light.
EPA
Yes, very straight ± .3%
IHC replied N/A.
-------
8) What problems have you experienced with the speed signal readouts?
GM _
A couple of meters (Weston current type) failed and had to be replaced.
EPA
Some failures and excessive noises.
Ford, Chrysler, IHC, Toyo Kogyo, Nissan, and ERI have had
no problems.
LOAD CELLS
1) Generally describe your calibration procedure, i.e., number of dead weight
values and data recorded.
GM
The load cells are deadweight calibrated at a simulated 50 mph using the
following sequence and tolerances:
Conditions
Horsepower
Meter
Readings
Torque
Meter
Three weights & arbor
Two weights & arbor
One weight & arbor
Arbor
None
40.7 jO.l hp
27.8 +0.1 hp
14.8 +0.1 hp
1.9 +0.1 hp
0.0 +0.1 hp
110.0 +0.3 ft-lbs
75.1 +0.3 ft-lbs
40.0 +0.3 ft-lbs
5.1 +0.3 ft-lbs
0.0 +0.3 ft-lbs
FORD
Arbor plus 3-35 lb. weights - record torque 5 IHP
(with 50.0 mph synthetic signal) at 5 points.
CHRYSLER
Weights of 0, 5, 10, 20, 35, 70 & 105 lbs with indicated HP recorded
within + .1 HP.
-------
. inc.
L^®£o.sej_plus ^dead^wei^ht;. values. . Record and adjust, h.p.
to +_.l at each dead weight value.
TOYO KOGYO
TOTAL WEIGHT DEAD WEIGHT INDICATE
5 10 10 35 35 ^WFT-LB)
5 x 5.0
15 xx 15.0
25 xxx 25.0
i+0 x x Uo.O
50 x x x 50.0
*£L> c:~t~ Fufldaaeotally Follow "the
Clayton aanual. Dead weight ft(Arbor,10,10,35lb)»data recorded
record the IHP in condition of increasing fir decreasing dead weights.
EPA
dead weight values and data recorded • Q & 110 St-lbs. 11 points
for zero & span.
5-90 data
2) Have you assessed the barometric sensitivity of the load cell? If so,
what is the relationship?
The load cells on all Clayton dynamometers have been vented, which makes them
insensitive to changes in barometric pressure.
TOYO KOGYO
About 0.5 HP at 30mnHg
EPA
Yes, barometric pressure changes shift the zero up and down. ( 31bs/p9i)*
Ford, Chrysler, IHC, Nissan, and ERI have not
-------
3) What data do you have describing the stability of the load cell
calibration?
GM, FORD, IHC, TOYO.KOGYO, NISSAN UEF » .anj EPA
The calibration stability nas not been determined.
CHRYSLER
Monthly calibrations show no abnormal drift or non linearity-
4) Do you have zero and span tolerances for load cell signals?
GM
Yes, the zero signal must be within +0.1 hp of zero and the span must repeat
within jK).l hp.
FORD
i 0.002 VDC for both
CHRYSLER
Yes - i -I HP
IHC
Yes, + .1 h.p. at eachdead weight. (Torque Cell Calibration")
TOYO.KOGYO
Less than 0.1 ft-Ib
NISSAN and ERI
Zero tolerance 0.0 £ 0-0 IHP
Span tolerance 40.7 ±0.1 IHP
EPA
± .002 VDC for zero and span. Zero is set as specified by Clayton
to.
-------
5) Please describe any failures and maintenance procedures which you have
encountered with the dynamometer load cell.
GM
The universal couplers can wear and stick causing mechanical hysteresis.
Ford, Chrysler, IHC, Toyo Kogyo, Nissan, ERI, and EPA have
had no failures.
PAU CHARACTERISTICS
1) Please describe the methods you use to characterize the power absorber
unit.
GM
The X-Y-Y plot of speed versus torque and horsepower with the vehicle driven
over the second cycle of the FTP schedule is used monthly to characterize the
PAU. These signals are compared to theoretical curves. The plots are also
examined for hysteresis to determine if tighter deadband or pulser adjustments
are required.
CHRYSLER
a. At a thumbwheel setting of 10.0, IW setting of
4 000 lbs., indicated horsepower is recorded at 5 mph incre-
ments up to 60 mph.
b. Speed is held constant at 50 mph with a horsepower
setting of 10.0 in manual mode. Indicated horsepower is
observed to verify that the horsepower setting stays constant.
TOYO KOGYO , v
Coastdown.Steady state(10-100Km/h),Instrumented test vehicle
(shaft-mounted torque transducers)
, The relation" of front roller "speed and indicated
NISSAN J —
t-n-rnup r>f tlie dynamoaeter in steady state is used.
-------
The relation of front roller speed and indicated
torque of the dynamometer in steady state is used.
11 point 10-60 mph 5 mph AV - integrate torque/
speed then do log-log fit 25-60.
Ford and IHC replied N/A.
2) What torque and speed data are used in the characterization and how are
they measured?
GM
The torque, horsepower, and front roll speed measurements are all derived
internally from the RLPC system. The plots are scaled using synthetic
measurement levels.
CHRYSLER
a. At a thumbwheel setting of 10.0, IW setting of
4000 lbs., indicated horsepower is recorded at 5 mph incre-
ments up to 60 mph.
b. Speed is held constant at 50 mph with a horsepower
setting of 10.0 in manual mode. Indicated horsepower is
observed to verify that the horsepower setting stays constantj
TOYO K0GY0
Indicate torauefpen recorder) >
Load roll revolution(digital tach),(F-V cortVerter-pen recorder)
NISSAN ,
how Era they neasungiLL . AHP=10HP. at RP. WliLVl
Speeds for steady state are 10,20,30,40,50,60,70,80,90 kq/h.
We evaluate the PAU characteristic reading the torque value at each speed.,
-------
DIM .
how are they rseasured? IW = 2750 lbs., IHP = 7.5 HP at 50 mph.
Speeds for steady state are 20, 30, 40, 50, 60 mph. The data is
recorded and measured by strip chart recorders.
EPA
how are they measured? Eleven data points from 10-60 mph are
collected. A log-log fit of torque versus speed from 25-60 mph vill
show the exponent as the slope of the torque vs. speed.
Ford and IHC replied N/A.
3) Has "a comparison of the manual vs automatic control exponent function been
performed?
GM
Manual mode has excessive hysteresis which makes this comparison extremely
difficult.
CHRYSLER
A brief comparison was done, which showed that indicated
horsepower v.s. speed was the same in the manual mode as
in the thumbwheel mode.
ERI
function been performed? Manual; 2.74 —**2.87. Automatic; 2.96'^'
3.08 Under the conditions of trangient and steady state.
EPA
function "been performed? Yes. H 9?2.2 - 2.3 A ~ 1.9 2.1
This difference has negligible effect on 55-45 coastdown.
Ford, IHC, Toyo Kogyo, and Nissan have not.
-------
4) What is the effect of the exponent value on vehicle loadina?
GM
No actual data, but theory indicates that lower exponent would result in
increased vehicle loading during FTP schedule.
EPA
Very small effect from 20-60 for same set point @ 50.
Ford, Chrysler, IHC, Toyo Hogyo, Nissan, and ERI replied N/A.
5) Does the power absorber exponent differ under transient vs steady state
response?
Steady state; 3.04. Acceleration (90 mph/30 sec.);
2.96- Deceleration (90 mph/30 sec.); 3-05. Average 3 times each
on AUTO.
EPA • 1
>...r inertia logins dominates transient operation.
GM, Ford, Chrysler, IHC, Toyo Kogyo, and Nissan have no data
6) Has the effect of temperature setting on PAU exponent been assessed?
GM
No effect has been observed when the system is operated in the automatic;
mode. In the manual mode, increasing temperature results in increasing ;
exponent.
EPA KPA uses 95 ± used to be 13S<>F. Havm't
analyzed or quantified affects.
-------
TOYO KOGYO
yig 2 Load Water Temp.-Indcate Power
13
Indicate
Power 12
At 50mph
(HP)
11
10
25
30 35 10 15
Load Water Temp. (*C)
Ford, Chrysler, IHC, Nissan, and ERI have not.
7) What is the response time of your dynamometers in changing the thumbwheel
from 5 to 15 horsepower or 15 to 5 horsepower?
FCTRD
5 to 15 hp - 3 sec; 15 to 5 - 2 sec (observed with vehicle
being driven on rolls)
CHRYSLER
We do not do this. We check for leaks as described in 1) b.
IHC .
7 seconds
-------
from 15 HP to 5 HP 2.5 Sec.
from 15 HP to 5 HP 2.5 Sec..
EPA
8 seconds or less.
ERI
( 95% Response Time )
£1 c/D; from 5 HP to 15 HP 3.8 Sec.
#2 C/D; from 5 HP to 15 HP 2.3 See.
GM, Toyo Kogyo, and Nissan have no data available.
FLYWHEEL FRICTlONALHORSEPOWER
1) Do you have frictional horsepower values for each flywheel as determined
by the subtractive inertia method described in the Clayton manual?
FORD
Yes - at least one run on_each of 7 cells.
CHRYSLER ...
The data we have on each flywheel shows the following range of
frictional horsepower values:
125 and 250# wheels 0.25 up to 0.5
500# wheel 0.3 up to 0.6
trim and 1000# wheels 0.5 up to 1.0
2000# wheels 0.75 up to 1.5
Values above the maximum value listed indicate increas-
ing bearing wear. Repeatability from month to month is
approximately + 0.1 HP for the low weights and about + 0.2 HP
for the higher weights.
T0Y0 K0GY0
FLYWHEEL FRICTIONAL HORSEPOWER l) & 2)
TW H AVG . SIGMA.
250 3 0.271HP 0.033
500 2 0.382 0.028
1000 2 0.U93 0.011
2000 1 0.690
TRIM 8 1.082 0.026
-------
ERI aResults_from ftl c/nr _____
£45_Ibs.—ifxim-flyvflieeL 0.829 HP) . 125 lbs. (0.15S HP) . ?sn 1HC
10..J6.0. HP),. .500 lbs. J0,218__HJP;,.. 100p_ Ibs. J0^03 H?)L 2000 lbs. (0.636 HP)
EPA
Yes, see Attachment VIII
GM, IHC, and Nis san do not.
2) If multiple determinations of flywheel frictions have been made using
various inertia combinations, what is the repeatability of those values?
FORD SIGMA value between 0.04 and 0.15-
Prime wheel method (back-to-back tests): SIGMA equals
0.02 to 0.05.
NISSAN and ERI
Repeatability ( Msx. - Min. ) : 0.2 HP.
Generally < .05 Hp. ' Sometimes around .10 Hp for 2K wheels.
GM, Chrysler, and IHC replied N/A.
3) What are the transient warm-up characteristics associated with flywheel
friction?
Annn FHP "decreases with warmup time until
rUKLJ -—¦ - ———___—_
bearing temperature stabilizes.
-------
EPA
Seem to stabilize after 10-15 minutes.
GM, Chrysler, IHC, Toyo Kogyo, Nissan, and ERI have not
determined.
4) Once the dynamometer is stable, how does the warmup time constant
EPA differ with various flywheel combinations? A new flywheel combination
will reach a stabilized coastdown time in about 4 minutes.
GM, Ford, Chrysler, IHC, Toyo Kogyo, Nissan, and ERI have
not determined this.
5) What is the warm-up or cool-down time constant for frictional horsepower?
ERI
30*minutes at 50 mph warmup.
EPA Limited data Indicates that fritional horsepower re-
mains stable longer than it takes to reach that stable value, ie
£wu < £ cd.
GM, Ford, Chrysler, IHC, Toyo Kogyo, and Nissan have not
determined this.
6) What is the effect of warm-up speed and time on achieving stable
fricitonal characteristics?
CHRYSLER
We have no data on this, but feel that 10 minutes at 50 mph
is minimum needed for stability.
-------
TOYO KOGYO
Xj pig 3 Effect of warmup speed and time
Torque
(Kg-m)
3.0
Transmission
out put shaft^
2*0
•' JjOKm/h
*-¦*—* 60Km/h
¦ v *»!_
10 20 3d ^0
Time .(minj
50
6a
NISSAN
IW-5250lbs,ABP-lQfiP ; 30mln. at 50nph>
6Qmin. at- 37mph,90n>in. at 25mph.
EPA
frictional characteristics? Has not "been quantified, but FHp in-
creases with the square of velocity, so higher speed should warm-
up faster.
GM, Ford, IHC, and ERI have no data on this.
7} Do you detect frictional horsepower changes as a function of power
absorber thumbwheel setting? If so, what magnitude?
FORD
No. Our coastdown data shows frictional horsepower is
independent of thumbwheel setting, with the average slope
of indicated v.s. actual horsepower very close to 1.00 over
time and across all of our dynamometers. (After subtracting
rear rolls inertia weight.)
-------
CHRYSLER
No apparent change in FHP has been observed with
in thumbwheel setting.
EPA
Analysis by ( 1W-155) method generally shows increasing FHp with
increasing TW. New 125// dynos show this moreso than older units.
GM, IHC, Toyo Kogyo, Nissan, and ERI detected no changes.
8) What are the effects of coupling lubrication and alignment on flywheel FHP?
GM
No data available. However, all dynamometer shafts are leveled and aligned
with each other to within +0.005 inches. At the present time, the couplings
are lubricated every three to six months.
CHRYSLER
No effects have been identified although in one case, mis-
alignment caused bearing failure within a short period of
time. The coupling itself failed also.
EPA _
flywheel FHP? Coupling misalignment could cause increased FHp as a
function of PAU load, but has not been verified.
Ford, IHC, Toyo Kogyo, Nissan, and ERI have no-data.
9) How long does it take to stabilize the frictional value for a new bearing?
GM
This has not been defined by any engineering study; however, our experience
indicates that a minimum of 72 hours of steady state or test cycle driving is
sufficient to stabilize the friction of new bearings.
-------
FORD
All npw hparinpc; - Ifl tn IS hmir.S
CHRYSLER
No data available. We use an arbitrary time of 24 hours at
50 mph.
NISSAN
Roughly , 50 hra at 31 mile/h > . 32 hra at 50 mile/h
EPA EPA operates dyno for 8 hours at 30-50 mph. Spot-
of the 6875 1W coastdown time are made.
IHC, Toyo Kogyo, and ERI have not measured this.
10) What problems or FHP differences have been detected for C3 vs standard fit
bearings?
GM
We have not experienced any C3 bearing failures. The C3 bearings have less
friction than the standard fit bearings.
EPA
C3 bearings do not have significantly lower FHp. Failure rates are about equal.
Ford, Chrysler, IHC, Toyo Kogyo, Nissan, and ERI replied N/A.
11) How does flywheel friction vary with dynamometer speed (function and
intercept)?
GM
Limited data indicates flywheel frictional horsepower to be a function of
speed raised to a power of 2.2. Torque intercept may range from 0.3 to 1.5
ft-lbs.
-------
1976 study shows friction torque linear with speed
NISSAN
Equivalent friction force i
f = 0.28334 V -f- 1.7740 V = taph, f = lb
ERI
and intercept)? Flywheel friction (torque) is a linear curve with
EPA
and intercept)? Flywheel FHp increases with roll velocity squared..
Static torque is less than 0.2 FT-LB.
Chrysler, IHC, Tcyo Kogyo replied N/A.
12) What percentage of total fractional horsepower can be attributed to
aerodynamic drag?
GM
Theoretical analysis indicates that the aerodynamic drag contribution to total
frictional horsepower ranges from 25 to 70% depending upon selection of
inertia wheels.
theoretical analysis indicated about 60-75% of total FHp comes
from aerodynamic drag on the flywheel.
Ford, Chrysler, IHC, Toyo Kogyo, Nissan, and ERI had no data
on this.
-2
dynamometer speed. T = 8.089 X 10 S + 0.7603^ (
IW=27501bs
-*—S—mprr
(T — lb. ft)
EPA
Data from the bearing manufacturer and a
-------
ROLL DIAMETER AND SPACING
1) What measurement techniques do you have for measuring roll diameter?
installation the diameter of all rolls were measured at five locations
using a micrometer and the concentricity (run-out) was measured at three
locations- The rolls are checked periodically using a fifth wheel and digital
odometer. The rolls are rotated 100 revolutions and the distance traveled Is
measured. From this measurement, the diameters can be calculated and compared
to the previously determined values. Periodically the rolls have been
examined with an indicator which is slid along the roll surface to determine
if any wear pattern can be detected.
FORD, NISSAN, and ERI
Rear roll diameter was measured with micrometer.
CHRYSLER
Nominal diameter is used,
2) What is the wear rate for machine rolls?
GM
No measurable wear has been detected over an 18-month period.
NISSAN
Ma*. 5^092 (toller dia. ) / 5 yeara
IHC> T°y° KOgy°' ERI- a"d ¦<* uSe
3} What statistics do you have for actual distance measurements?
GM
Table 2 contains statistics of actual distance measurements for 1978
Certification tests at the General Motors MiVford-Vehicle Emission Laboratory.
CHRYSLER
No statistics were ever tabulated, although correlation
testing with EPA shows very close agreement.
-------
GM
Table 2
Actual Distance Measurements
from 1978 Certification Tests at M-VEL
FTP
Phase 1
Mean % Diff* Std. Dev. Min Max
3.592 0.06 0.016 3.53 3.65
Theoretical Distance - 3.59 Miles
Phase II
Mean % Diff Std. Dev. Min Max
3.860 0.0 0.022 3.79 3.93
Theoretical Distance - 3.86 Miles
Phase III
Mean % Diff Std. Dev. Min' Max
3.593 0.08 0.016 3.55 3.66
Theoretical Distance - 3.59 Miles
HWFE
Mean % Diff Std. Dev. Min Max
10.255 0.1 0.038 10.04 10.39
Theoretical Distance - 10.242 Miles
_ Actual-Theoretical „
*Dlff Theoretical x 100
-14-
-------
EPA
x S.D. % CV
Bag 1 3.5934 .0241 .67% Bag 3 3.5947 .0221 .61%
Bag 2 3.8899 .0349 .89%HWFE 10.259 .049 .48%
Ford, IHC, Toyo Kogyo, Nissan, and ERI have no statistics.
4) What tolerance do you have on distance measurements for phases 1, 2, 3,
and HWFET?
GM
The actual distance traveled tolerances for an FTP schedule are listed below:
o Phases I & III - 3.591 mi +2%
(3.519 to 3.663 mi)
o Phase II - 3.859 mi +2%
(3.782 to 3.936 mij
The actual distance traveled tolerance for a HWFET schedule is +2% of 10.242 mi
(10.037 to 10.447 mi).
FORD Phases 15 3: 3.51 to 3.69 miles
Phase 2: 3.77 to 3.96 miles
HWFET: 9.99 to 10.50 miles.
CHRYSLER
We use the tolerances EPA uses which are:
Phase 1 - 3.606 + 2%
Phase 2 - 3.901 + 2%
Phase 3 - 3.607 + 2%
HWFE - 10.295""+ 2%
IHC
Calculated distance from roll revs must be within + 2%.
TOYO KOGYO
Approximately 2/» of the theoretical mileage
-------
NISSAN
i 0.2 mile for phase 1 , 3
i 0.3 raile for phase 2 , HWY
ERI + 2.0% for each phase value, CT: 3.6 mile,
CS: 3.9. mile, H.T. : 3.6 mile, HWYFET: 10.242 mile.
EPA
Bag 1 3.52 - 3.66 Bag 3 3.52 - 3.66
Bag 2 3.83 - 3.-99 HWFE 10.OA -10.45
5) What is the effect of roll spacing on actual loading and tire slip rate
(both steady state and transient) and on quick check coast downs?
GM
No data available on new Clayton dynamometer systems.
IHC
Vehicle coastdown on 20" spacing vs. 17" spacing is within
7% criteria.
EPA
N/A All EPA rolls are 17.25" spacing.
Ford, Chrysler, Toyo Kogyo, Nissan, and ERI have no data.
6) What value for rear roll inertia and frictional horsepower do you obtain
in your laboratory? Describe the roll configuration.
GM
See Figure 2 for M-VEL's roll and inertia weight configuration.
Frictional horsepowers were calculated using Clayton's nominal rear roll
inertia of 77.5 pounds. Frictional horsepower per split roll.for all sites
ranges from 0.07 to 0.17 hp. The total rear roll frictional horsepowers range
from 0.16 to 0.32 hp.
-------
CLAYTON CTE
8.65" DIA. ROLLS^r
50 HP
ABSORBER
NEW RLPC
w-
17.25" ROLL SPACING
WITH LIFT GATES—
FIGURE 2
-50 CONFIGURATION
00
-I
o
in
(VJ
5
CD
_J
KS
"33
03
1000 TO 8875 LB.
(125 LB. INCREMENTS)
-------
FORD
155 lb. for all rear roll inertias
3 cells are ECE-50; 4 cells are CTE-50.
CHRYSLER
Configuration is Clayton CPE-50. Rear rolls inertia is
155 lbs. (per Clayton).
TOYO KOGYO .
Inertia; 1551b.- FHP; about 0.1 HP at 50mph
Roll configuration; Fig.U
JT
8.65J-
-77.75*-
8.65?
!
17.25"
1
NISSAN and-EJU
Rear inertia is 155 lbs, (from Clayton's manual) Average frlctloaal
horsepower is 0.253 HP (including roll revollution coupler) on
ECE-50 model arrangement B.
EPA has Clayton configuration( B) with a single roll. Assigned
1W = 155 lbs.
IHC replied N/A,
7) Describe the coast down repeatability and statistics for rear roll
friction across all dynamometers.
GM
Rear roll coast down technique repeatability, as measured on two dynamometers,
was within +0.006 hp.
-------
FORD
CLVT'-V [^".OMETER REAR F:ILL FRICTION STUDY
CHL PJ:1SER I
Fw:.;£R Cf REAR ROLLS 2
55 TG <5 TIME HDRSEPCHER
A3.29
41.20
5i.ro
5.45
1.70
1.37
0.19
0.23
0.16
AVE6A3E
STD L£V
1.51
0.17
0.20
0.02
Ci'll .-¦.JfSER 3
nJ r::L5 i
55 TG ?.£IA7j FCSCt HCfcSEPC.ER
4 1 .CO
1 .6?
0.23
40.25
1.72
0.23
4 f. 20
1.50
0.20
<3.Ci
1.58
0.21
3 6.20
1.91
0.26
a. 50
1.49
0.20
A JC. 1.5i 0 r 21
C-rr I:£Y 0.13 0.02
-------
FORD
cat fH.'JlSER
4
Or *£Aft
ROLLS 2
55 TO 45 TIME
(GEO f.£IASD FORCE
(LB)
HOfcSEPCUER
38.30
1.81
0.24
32.00
2.1?
0.29
39.42
t.78
0.24
37.50
1.87
0.25
42.50
1.65
0.22
44.30
1.53
0.21
33.20
1.84
0.24
34.95
1.90
0.25
37.40
1.83
0.25
AVERAGE 1.S3
0.24
STD nv 0.17
0.02
CFU. MM5E*
5
i < J11K CF f c. i-
f:CLLS 1
s-j to 45 Tine
J5L';> S5TA.-.S FORCE
-------
.CHRYSLER
Range of values across 8 dynamometers for four months was
.095 to .203, with a mean value of .139. Maximum range of
variation for one dynamometer during this time was .150
to .203. With the exception of two data points, frictional
horsepower was within .15 + .05.
ERI
Results of dynamometers were 0.252 HP and 0.254 HP. Repeatability at
each dynamometer is + 0.8%.
friction across all dynamometers* Repeatability during calibration
should be ± .5 seconds fron the average of about 50-60 seconds.
Longterm changes in the nominal coastdown time shoud be less than
5 seconds.
CALIBRATION MONITORING PRACTICES
1) What pre-test horsepower check procedure and tolerance are used in your
laboratory?
GM
Prior to each Certification test, a non-Certification vehicle is used to set
the dynamometer horsepower. Using the displayed indicated horsepower and the
calibration curves, the actual horsepower must be within jp.5 hp of the
desired setting.
FORD
Drive front roll 50 mph 5 read IHP: must be within + 0.3 of thumbwheel.
CHRYSLER
Not formally done. Any discrepancy exceeding +0.2 horse-
power between thumbwheel and indicated readings at 50 mph
will be resolved before the rolls is used for further testing.
IHC
Vehicle is operated at 50 mph and horsepower checked to
+ .5 h.p.
-------
TOYO KOGYO
A) Select the inertia. B) Check the power meter (zero,span)
C) Set the warm up car. D) Warm up (l5min. at 30mph)
E) Set the proper horsepower for the test vehicle at the front roll
speed 19^3rpm(50rph) "with manual sw.
F) Accelerate the dynanometer to 60mph and return to 50mph.
Verify that the indicate power within _+0.5HP of the set power.
NISSAN and ERI _pleated horsepower is checked.gt.
50 nph before each test. Tolerance ¦ i 0.1 HP
EPA The thumbwheel value is verified at PAU 50.
Acceptance criteria are ± .5Hp, but properly adjusted dyno
circuits can maintain ± .3 Hp.
2) If multiple quick checks are performed to confirm road load horsepower,
what tolerance do you place on repeatability?
FORD
3 repetitions: highest-lowest times less than 0.3 sec.
CHRYSLER
Three repetitions within +0.2 seconds are required. A
fourth repeat is performed if the first three are more than
0.2 seconds apart.
NISSAN
+ 0.1 HP
EPA
Data show that the range (max—min) of 3 loaded quickcheck times
will be less than .1 seconds on the average. Range values
greater than .J are abnormal.
GM, IHC, Toyo Kogyo, and ERI do not perform multiple quick checks.
-------
3) If a weekly dynamometer verification procedure is performed, please
describe it in general (IWs checked, warm-up procedure and tolerance).
GM
Only monthly dynamometer calibrations are being performed.
FORD 30 minntp wa-rmnp- aJ- thrpp
•i np-rti a wpightc - rmct he within + 1 sec, of* Q.5 hn
whichever is smaller: slope at highest inertia 1.000 + 0.015.
IHP within + 0.3 of thumbwheel.
CHRYSLER
3) and 4) A weekly vehicle crosscheck is performed at 4000 lbs.
inertia weight and 13.2 horsepower. Duplicate hot 505 tests
are perforemd on each rolls. Any significant changes in
dynamometer loading would be identified by changes in the
COj emissions.
T0Y0 K0GY0.
A) Set the vehicle lifting device and the coastdown timer.
B) Check the power meter. C) Set the check car.
D) Select the inertia 28751h or (Trim+500+250+1251b)
E) Warm up • F) Set the horsepower 8HP at 50mph (front roll speed)
F) Accelerate the dynamometer to 60mph and lift up the car from the roll
Measure the coastdown time of 55-^5mph
G) Verify that the coastdown time is within +1 second of the latest
calibration data.
H) Set the inertia 30001b or (Trim+lOOOlb). Repeat step E) to G).
I) If the coastdown times differ by more than j\Lsec.,
a new calibration is required.
NISSAN Torque meter check ; vamnp—30mph x 40oin.check 0 &
501b-Ct point,dead weight tester 5,15,25,40,50, lb-ft(tolerance;0.3lb-ft)
Deter check j by strobe, tolerance ±0.3 mile/h.
EPA
See TP-302.
IHC and ERI replied N/A.
-------
4) Describe any special correlation test performed to assure dynamometer
correlation and stability.
GM
As a part of the dynamometer monthly calibration, hysteresis checks are made
in the automatic mode and stability checks in the manual mode.
The stability check consists of an 18-cycle FTP schedule driven at an
indicated horsepower of 12.0. The starting and ending indicated horsepower
readings must agree within jO.5 hp.
The hysteresis check is performed by driving a vehicle at 50 mph front roll
speed while the thumbwheel settings are cycled to obtain the following
sequence of indicated horsepower readings, within the tolerance of +0.4 hp:
10.0, 12.0, 10.0, 8.0, and 10.0. The middle and last'reading at the 10.0 hp
setting must agree within +0.2 hp.
FORD
moaieter correlation and stability. Loaded vehicle coastdowns
weekly on every cell; breakaway force: front S rear rolls:
rear roll coastdown.
Refer to Chrysler's answer to question #3 in this section.
EPA
mometer correlation and stability. Rpppat-ahip veh^io -fc ,1Qorf
and loaded quickcheck times are also analvzprl hPfwPPn b-H-pq Ai-
ternate PAU deterainatipns have been performed across dvnns at- i-Wc.
IHC, Toyo Kogyo, Nissan, and ERI do not perform any special tests.
5) What control charts or trend analyses are performed using the verification
data?
GM ,
Tolerances are described in 4) above.
FORD
Plot all of above.
-------
CHRYSLER
Weekly crosscheck data is kept for six months back, but no
formal trend analysis is performed since any trends would
likely be caused by vehicle changes.
EPA
See TP-302.
IHC, Toyo Kogyo, Nissan, and ERI replied N/A.
6) What techniques have you developed to correlate road load data to quick
check coast down times?
GM
Quick check coast downs are not performed.
CHRYSLER
Road load data is correlated to rolls coastdown times as
required by EPA in A/C #55B.
EPA
quick check coastdown tines? If AHp is calculated from the quick
check coastdown time and plotted versus PAU TW value, a tight band
of data is observed and can be used to spot outliers.
Ford, IHC, Toyo Kogyo, Nissan, and ERI replied^(N/A.
SITE CORRELATION TESTS
1) If your facility uses a repeatable car, what is the variation in the coast
down time across dynamometers?
GM
Repeatable car is not used to gather coast down data.
FORD
+ 5.35%
-------
CHRYSLER
Coastdowns are not normally done. Our tolerance is a
coefficient of variation of 2.0% or less on CO emissions.
(This is sometimes exceeded.)
IHC
Maximum variation at 50 raph across dynamometers is 5 to 6%.
T0Y0 K0GY0
55-U5mph coastdown time(repeatable car)
IW=27501b AHP=9-9HP N=1+
AVG TIME=11.36sec. SIGMA=0.075
ERI
the coastdown time across dynamometers? At IW = 2750 lbs. AHP = 9.9 HP,
?1 C/D; x = 12.765 sec. QT = 0.088 at recent 12 data.
{j2 C/D; = 12.842 sec. Q~ = 0.067 at recent 9 data.
-EpA
the coastdown time across dynamometers? The vehicle used has an
average timp nf 17.1 n„OT-an ,roT^^?^-^on from day to daY ls
about ± 2%. Average differences dyno to dyno are normally less than
+ 1.5%
Nissan replied N/A.
2) What is the variability in the road load determination from dyno to dyno?
GM
Investigations have indicated that the variability, defined as 2.77 times the
standard deviation divided by the mean road load horsepower, is approximately
10 per cent.
-------
CHRYSLER
The dynamometers vary slightly depending on the last
calibration performed. Since they are recalibrated each
month, the offset of a particular dynamometer is temporary,
and they are all treated as being equal. Several dynamometers
are averaged to generate road load horsepower settings.
TOYO KOGYO
5.C
Torque
(Kg-m)
n.cf
roll
surface
3.0"
2.C-
Pig 5 . Dyno road load variability
10
0
20
£
e
JL
G
25T
O
o
s
o Clayton-A
O * B
A ¦» c
'' D
30 40 50
Car speed Km/h
60
70
80
ERI The. difference between two dynamometer^ 3$, _
When IW = 2500 3000 lbs., AHP = 9' '11 HP Class vehicles are used.
Ford, IHC, Nissan, and EPA replied N/A.
-------
3) What is the sensitivity of the quick check coast down time to the
horsepower setting?
CHRYSLER
This varies from vehicle to vehicle.
Indicate
Power
At 50mph
(HP)
Pie 6 Coast down t1tne-THP
2750 lb
13 1* 15
Coast down time (sec)
ERI
0.09 sec./O.1 HP at IW
= 2750 lbs., AHP = 9.9 HP.
EPA A 7% change in coastdown time equates to a
15% change in PAD, since the FHp of the drive train Is the dominant
part of total FHp.
-------
GM, Ford, IHC, and Nissan had no data on this.
4) How much friction in the quick check procedure is attributable to the
vehicle dyno combination (i.e., dyno bearing friction vs total frictional)?
ERI About 15 % (i.e., dyno bearing friction is 2. 0 HP, total
friction is 13.2 HP).
EPA tor one vehicle the total friction"was 10 Hp ¦ 2 FHp
dyno + 8 FHp vehicle.
GM, Ford, Chrysler, IHC, Toyo Kogyo, and Nissan have no data on
this subject.
SPEED SIGNAL FOR DRIVING TEST
1) Please describe any studies performed to characterize the effect of tire
slip on test results.
No response to this question.
2) Has test variability been assessed using coupled rolls vs front
or rear roll driving trace signal? *
IHC !
Front roll driving trace introduces large efror due to
tire slippage.
GM, Ford, Chrysler, Toyo Kogyo, Nissan, ERI, and EPA have not
done any assessment of test variability.
-------
1
2
3
4
5
6
7
a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
29
26
27
28
29
30
31
Fa ic t/omal /-/o&zzpoujefi
AHP t/s T+$
ft f/P us
ZoooCz)
zZooo6)
*/O0O 5 S'OO *260 /zs
r&*
TA/Sf f&A**oil
0.9VV
0.9ZJ
.57
0.67*
O. 68/
0.6sy
.63
o>^>/
a 6//
o.699
o.s/s
o.(o/V
.60
O.90Z
0,6*6
0.66/
.*7
0.99&
0.9//
O.68Z
.62-
0,90*/
o,ss£
0*537
.H6
O.SS8
O.V6?
.51
o.y&f
0.*/9
O.f/39
.*S
o,**/8
O.V38
O.VSf
0^3*
.y-o
O.sog
o,*sy
0.Y35
.¥7
O.&Of
o,*/3£
o.vtz
.<+1
o,ys&
o^S6o
o,so/
Mz
O.Z9/
O.Z9/
.3*
O.ZZ8
O.Z93
O.ZZ&
.2?
0.239
O.ZS*
O.ZSO
O.ZiZ
O.Z3*
.23
O.Z90
O.ZVO
O.ZV
.31
O.ZZg
0,389
a*3z
.2.7.
0,2ZS
Or3V9
O.Z/9
.23
O.tSS
o./sz
.11
0.Z08
O./SZ
0./89
.25
0./S9
O.ZOj
0./9Z
o./te
0./9/
J6
O.ZZ&
O./SV
o.z/6
.17
O.ZS9
o. zn
o.zzs
.20
o./ss
O.20*
o. z/z
.11
0./9/
o.zos
O.ZO/
0./69
0./B9
0,/Z8
0//3S
O./S/
0,/S0
o,/SO
O./07
O>09?
O.Z39
a/36
0./3S
0./*/8
0,//9
0,/3*
o.//y
0.8/6
/./*
Q.9Z2
1.3i
o,6y/
0.&09
hto
N/A
1.36
0.9/S
1.H
0,?S9
A 06
0.26 3
o.99z
O.Ce//
0.909
0.93V
0,83V
0,9*6
O.C08O
o.(,o9
0,609
/. ovo
O.SO/
O,?#*
0,(o96
0,690
0.9Z/
o.rsS'
0.8*3
A0/3
0,//9
O./Ofo
0./S9
0./9Z
O.Z6C,
0.203
O. /OS
o./zs
0.099
0.096
0.098
o./**
0./S9
0./S9
0./3?
o.tzo
a /8 3
J7S<*
zso
-SSOO
luJ CoaIF/6,
3
0
x
5
H
y~°
f asu
-------
X
2
3
4
5
6
7
8
9
10
11
12
13
14
13
16
17
18
19
20
21
22
23
24
29
26
n
r
29
30
31
bzF- F- !LjlSr.fv)C^ V LiI'nJE i\JO e L:. ~1
08' Zl -79 JW
f&cr/OAMU. //of&efkuieA
t/s
Amp c/s too
£oooQ)
*/ooo
soo
zso
/ZS"
772//*
TX/*f
jggqg toll
0.4,41,
0.938
O.fc?
0.93Z
Of8/9
0,63/
O.b/9
o,u>/
O.G9Y
0£9(>
0*S9(?
0.6/f
0.8/1
0, (ofo'f
o,o92*f
.56
O.V98
O.SZ/
0,Y9o
0,8
0.03V?
.n
0.3>0L>
az/&>
O.238
O, ZS/
O.ZU
o,/9(?
0,/9Z
o.3*4
O.ZS*
0.239
O.Z5S
0./9Z
O.V3Z
0.0628
O.Otf/S
o.8?Z
/. zog
A3VZ
O.G09
A3VZ
0,89i
0.2S6Z
hzo
/, 3i/(=>
A3 Of
0.9 9Z
/.OoZ
o.?//
/.3V?
/.ZSS
0.920
0.86>9
/.Z3<*
A39
-------
DYNAMOMETER SITE! 0003
OTf.'O EPA PIDJ
CALIBRATION MODE!
CALIBRATION DATtt
CALIBRATION TImC:
OPERATOR IO:
BAPO-ETER:
SPEED RANGE:
CATA LIKE NO.:
CAL CHECK HPS
REah roll FhPi
AVG. MR DELTA Ti
N/A
AUTO
07-12-79
04:00 < 00
309U0
29.15
55.00-45.00 HPH
2736.000
0.0
0.124 HP
73.32
•••• DYNAMOMETER CAllBMATlON ANO ••••
COASTOOWN DATA ANALYSIS ••••
o*»»0*«*a4«4*e*o»ttaatt*«eo«4oa>»*e*««»*o«a»»
oooooo
oooooo
33333333
00 00
00 00
3 33
0 0
0 0
33333
00 00
00 00
3 33
oooooo
oooooo
33333333
DOOOOO
00 00
00 DO
00 00
000000
PROCESSED! AUG 9, 1979 08l44!35
VERSION! 1.23
OPTIONS! TW VS AHP.FULL IW.CERT TBL
COMMENTS! GENERAL 125LH. CALIB TO US FOR CERT WITH 600 CFM-CVS FRtRR RESET7-13 6500REDON
•»•• SPEEO CALIBRATION ••••
•• TOROUE CALIBRATION •»
• O*O»«e04»«*»OO4»O*0*4O»»O
•••• POWEH absorber curve oata ••••
SPO
FRNT
REAR
TRUE
FRNT
FRNT
DEAD
LOAD
TORO
TORO
REAR
AVG
AVG
PAU
PAU
MTR
TACH
TACH
SPEED
SPO
VDCl
WGHT
CELL
cms
VDC*
SPO
SPEED
TOROUE
TIME
SPEED
TORO
CALC
CURVE
CURVE
MPH
VOC
VDC
PPM
MPH
DIFF
FTLB
VOC
HZ
0 IFF
MPH
COUNTS
COUNTS
SECS
MPH
FTLU
I HP
FTLU
*DIFF
10.
1 .00
1.00
387.
10.0
0.29
5.0
0. 189
189.
2. 17
10.0
3971.
693.
10.250
9.97
1.87
0.14
0.99
89.47
15.
1 .50
1.50
S83.
15.0
-0.14
10.0
0.367
366.
-0.51
15.0
5828.
1219.
10.099
14.85
3.31
0.36
2.24
47.72
20.
2.00
2.00
776.
20.0
. 0.03
15.0
0.551
551.
-0.52
*0.0
7764.
1776.
10.048
19.88
4.82
0.71
4.08
18.31
25.
2.50
2.50
9 f>*.
24.9
"0.24
20.0
0.739
738.
0.01
25.0
9692.
2332.
10.064
24.78
6.31
1.16
6.41
-1.52
30.
2.9V
3.00
1160.
29.9
0.04
25.0
0.919
917.
-0.53
30.0
11886.
3510.
10.279
29.76
9.28
2.04
9.33
-0.53
35.
3.40
3.SO
1350.
34.7
0.05
30.0
1.109
1108.
0.01
35.0
13664.
4819.
10.160
34.61
12.87
3.30
12.72
1.20
40.
3.98
4.00
1540.
39.6
0.31
40.0
1.485
1483.
0.41
40.0
15706.
6371.
10.237
39.48
16.87
4.93
16.67
1.24
45.
4.46
4.SO
1733.
44.6
-0.11
50.0
1.84S
1843.
-0.21
45.0
17594.
8091.
10.203
44.37
21.49
7.06
21.18
1.44
sO.
4.95
S.00
1V2J .
49.4
0.01
60.0
2.225
2223.
0.27
50.0.
19473.
9935.
10.165
49,30
26.48
. 9.66
26.29
0.71
55.
5.44
5.SO
2114.
S4.4
-0.12
75.0
2.768
2766.
-0.21
55.0
21184.
11767.
10.087
54.04
31.59
"12.63
31.75
-0.50
60.
5.91
6.00
2295.
59.1
-o.os
90.0
3.331
3329.
0.06
60.0
23454.
14049.
10.238
58.95
37.15
16.21
37.95
-2.10
FR01T VDC
REAR VOC
0.1001'MPH ( 5.006 *50)
0.1000'MPH ( 5.000 ffSD)
FT-L8= 0.0270»HZ ~ 0.0450
VOC = 0.037CFTLM-0.0011
PAU CURVE FOR 4000. LB INERTIA WEIGHT
UETWEEN 25 ANU 60 MPH
TORQa
K <
M o
K • (N««M»
0.8B17E-02
2.OS
C0A5T00WN OATA
Iw
LciS.
OELTA
T
TW
HP
CALC
ArtP
CALC
FHP
LLSO
TW
ftOIFF
TW
LLSO TW
M
oM«AHP«8
S
1000.
14.462
7.639
4.616
3.0
6.0
10.0
4.20
7.95
12.61
1.20
1.95
2.61
2.96
6.08
9.96
1.513
-1.328
0.361
0.6333
-O.b440
1125.
15.854
8.636
5.480
3.0
6.0
10.0
4.31
7.91
12.47
1.31
1.91
2.47
7.97
6.06
9.97
1.132
-0.991
0.266
0.8589
-0.7350
1250.
17.684
9.781
6.176
3.0
6.0 .
10.0
4.24
7.76
12.29
1.24
1.76
2.29
2.98
6.04
9.98
0.739
-0.649
0.1/1
0.8705
-0.7171
C-ERTIFIcATtOiO
PRINTOUT PO«.
EXAMPLE OS£t>
IN TP - 2.0 2- DATED
-------
r4
IK
LBS.
DELTA
T
Tw
HP
CALC
AMP
CALC
f MP
LLSO
TV
HOIFF
TW
LLSO TW
M
¦M'AHP.B
H
1375.
IV.021
10.671
6. 766
3.0
6.0
10.0
4.39
7.83
12.31
1.39
1.83
2.31
2.99
6.03
9.99
0.476
-0.*19
0.109
0.8848
•0.8985
1500.
21.265
11.833
7.472
3.0
6.0
10.0
4.28
7.70
12.19
l.?8
1.70
2.19
2.99
6.02
9.99
0.285
-0.249
0.065
0.885S
-0.8017
1625.
22.<>82
12.536
6.075
3.0
6.0
10.0
4.39
7.84
12.22
1.39
1.84
2.22
2.97
6.06
9.98
1.061
-0.930
0.249
0.8947
/
-0.9587
17S0.
24.129
13.694
6.769
3.0
6.0
10.0
4.40
7.76
12.12
1.40
1.76
2.12
2. 98
6.03
9.99
0.567
-0.497
U * 131
0.9078
-1.0IS4
1875.
2S.120
14.377
9.318
3.0
6.0
10.0
4.S3
7.V2
12.22
1 .53
1.92
.2.22
2.97
6.06
9.98
1.054
-0.925
0.247
0.9115
-1.1629
2000.
27.948
15.804
10.109
3.0
6.0
10.0
4.35
7.69
12.02
1.35
1.69
2.02
2.98
6.03
9.99
0.603
-0.5«!9
0.139
0.9133
-0.9871
2125.
26.769
16.545
10.674
3.0
6.0
10.0
4.49
7.80
12.09
1.49"'
1.80
2.09
2.98
6.03
9.99
0.6J8
-0.558
0.147
0.9211
-1.1508
2250.
30.427
17.532
11.290
3.0
6.0 '
10.0
4.49
7.77
12.09
1.49
1.77
2.09
2.99
6.01
9.99
0.257
-0.225
0.059
0.9209
-1.1431
237S.
31.244
16.250
11.890
3.0
6.0
10.0
4.61
7.90
12.13
1.61
1.90
2.13
2.98
6.04
9.99
0.795
-0.6V7
n. 1*4
0.9319
-1.3228
2500.
33.225
19.295
12.547
3.0
6.0
10.0
4.57
7.87
12.10
1.57
1.87
2.10
2.98
6.04
9.98
0.832
-0.729
0.193
0.9302
-1.2755
2625.
33.888
20.032
13.102
3.0
6.0
10.0
4.70
7.96
12.17
1.70
1 .96
2.17
2.98
6.03
9.98
0.6S1
-0.571
0.151
0.9365
-1.4345
2750.
35.597
20.960
13.750
3.0
6.0
10.0
4.69
7.97
12.15
1.69
1.97
2.IS
2.97
6.05
9.93
0.957
-0.640
0.224
0.9399
-1.4379
2875.
35.976
21.533
14.186
3.0
6.0
10.0
4.85
8.11
12.31
1 .85
2.11
2.31
2.98
6.04
9.98
0.710
-0.623
0.165
0.9396
-1.S81S
3000.
26.307
15.507
10.750
5.0
10.0
15.0
6.44
11.75
16.95
1.44
1.75
1.95
4.98
10.04
14.98
0.35H
-0.359
0.121
0.9S1-3
-1.140S
-------
IW
L8S.
DELTA
T
TW
HP
CALC
AMP
CALC
FMP
LLSO
TV
*OIFT
TW
LLSO TW
*4
om»AHP«B
a
3125.
2d.6*8
15.984
11.105
5.0
10.0
15.0
6.62
11.87
17.09
1.62
1.87
2.09
4.99
10.01
14.99
0.101
-0.102
0.034
0.9555
-1.3352
3250.
29.728
16.616
11.567
5.0
10.0
1S.0
6.64
11 .88
17.06
1.64
1 .88
2.06
4.99
10.02
14.99-
0.173
-0.173
0.058
0.9593
-1.3777
3375.
30.063
17.031
11.928
S.O
10.0
15.0
6.82
12.03
17.18
1.82
2.03
2.18
4.99
10.02
14.99
0.219
-0.219
0.074
0.9647
-1.5881
3500.
31.628
17.774
12.419
5.0
10.0
15.0
6.72
11.96
17.12
1.72
1.96
2.12
4.99
10.03
14.99
0.2bl
-0.261
0.088
0.9620
-1.4781
3625.
31.921
18.141
12.764
5.0
10.0
15.0
6.90
12.14
17.25
l.VO
2.14
2.25
4.98
10.04
14.98
0.404
-0.406
0.138
0.9661
-1.6826
3750.
32.974
18.801
13.182
5.0
10.0
15.0
6.91
12.11
17. ?.H
1.91
2.11
2.28
4.99
10.01
14.99
0.138
-0.139
0.047
0.9643
-1.6671
3B75.
33.560
19.134
13.501
5.0
10.0
15.0
7.01
12.30
17.43
2.01
2.30
2.43
4.93
10.05
14.97
0.491
-0.495
0.168
0.9598
-1.7545
<.000.
35.503
20.114
14.149
5.0
10.0
15.0
6.84
12.04
• 17.17
1.84
2.04
2.17
4.99
10.02
14.99
0.191
-0.192
0.064
0.9684
-1.6354
*250.
36.888
21.204
14.916
5.0
10.0
15.0
7.00
12.17
17.30
2.00
2.17
2.30
4.99
10.01
14.99
0.142
-0.142
0.04A
0.9702
-1.7957
4500.
38.464
22.268
13.721
S.O
10.0
15.0
7.10
12.27
17.38-
2.10
2.27
2.38
4.99
10.02
14.99
0.182
-0.184
0.062
0.9729
-1.9215
4750.
40.061
23.116
16.463
5.0
10.0
15.0
7.20
12.44
17.52
2.20
2.44
2.52
4.97
10.05
14.97
0.511
-0.514
0.175
0.9688
-2.0013
5000.
24.879
16.615
12.455
10.0
16.0
22.0
12.21
18.28
24.38
2.21
2.28
2.38
10.01
15.99
22.01
-0.055
0.069
-0.025
0.9856
-2.0244
5253.
2S.824
17.263
12.992
10.0
16.0
22.0
12.35
18.47
24.54
2.35
2.47
2.54
9.99
16.02
21.99
0.083
-0.105
0.038
0.9841
-2.1579
5500 .
26.697
17.967
13.529
10.0
16.0
22.0
12.51
lb.59
24.69
2.51
2.59
2.69
10.00
15.99
22.00
•0.0)2
0.039
-0.014
0.9854
-2.3258
-------
-=f
IV
DELTA
Tw
CALC
CALC
LLSO
%OIFF
LLSO TW
sN*AHP«B
LBS.
T
HP
*hp
KMP
TW
TW
M
8
6000.
2d.771
10.0
12.66
2.66
9.99
0.08*
0.9896
-2.5412
19.430
16.0
10.7b
2.75
16.02
-0.105
14.69b
22.0
24.79
2.79
21.99
0.039
6S00.
30.679
10.U
12.87
2.87
10.00
-0.021
0.9882
-2.7124
20.8S0
16.0
10.93
2.93
16.00,
0.026
15.783
22.0
25.01
3.01
22.00
-0.009
•••t••••••••«•••#•••««««••••««*o«•
•••• SLOPt/SPfcCO STATISTICS ••••
N HIN MAX ' AVG SIGHA *CV
IMP slopes 33 0.83J3 0.9896 0.9387 0.0404 4.3047
AVC SPEEOS 99 49.60 49.67 49.63 0.0109 0.0220
-------
kn
oooou
0 0
D
D
0
0 0
OOOOO
00000
0 0
0 0
0 0
0
0
OOOOO
0 0
0 0
0 0
33333
3 3
3
333
3
DYNAMOMETER CALIBRATION TABLE FOK
oeriveo Farm amp versus t« data
CALIBRATION OATE» 07-12-79
APPROVED HYI.
INERTIA
• • M ••
• • B ••
0 0 0
3
3
OOOOO
00000
33333
EFFECTIVE OATE»_
/ /_
1000
1125
1250
1J75
0.833
0.859
0.871
0.885
-0.
54
-0.
73
-0.
72
-0.
90
AHP
TW
AHP
TW
AMP
TW
AHP
TW
3.0
2.0
3.0
l.rt
3.0
1.9
3.0
1.8
3.1
2.0
3.1
1.9
3.1
2.0
3.1
1 .8
3.2
2.1
3.2
2.0
3.2
2.1
3.2
1.9
3.3
2.2
3.3
2.1
3.3
2.2
3.3
2.0
3.4
2.3
3.4
2.2
3.4
2.2
3.4
2.1
3.5
2.4
3.5
2.3
3.5
2.3
3.5
2.2
3.6
2.5
3.6
2.4
3.6
2.4
3.6
2.3
3.7
2.5
3.7
2.4
3.7
2.5
3.7
2.4
3.8
2.6
3.8
2.5
3.8
2.6
3.8
2.5
3.9
2.7
3.9
2.6
3.9
2.7
3.9
2.6
<•.0
2.8
4.0
2.7
4.0
2.8
4.0
2.6
4.1
2.9
4.1
2.8
4.1
2.9
4.1
2.7
4.2
3.0
4.2
2.9
4.2
2.9
4.2
2.8
<>.3
3.0
4.3
3.0
4.3
3.0
4.3
2.9
4.4
3.1
4.4
3.0
4.4
3.1
4.4
3.0
4.5
3.2
4.5
3.1
4.5
3.2
4.5
3.1
4.6
3.3
4.6
3.2
4.6
3.3
4.6
3.2
4.7
3.4
4.7
3.3
4.7
3.4
4.7
3.3
4.B
3.5
4.8
3.4
4.8
3.5
4.8
3.3
4.9
3.5
4.V
3.5
4.9
3.5
4.9
3.4
5.0
3.6
S.O
3.6
5.0
3.6
S.O
3.5
S.l
3.7
5.1
3.6
5.1
3.7
5.1
3.6
5.2
3.8
5.2
3.7
5.2
3.8
5.2
3.7
5.3
3.9
5.3
3.H
S.3
3.9
5.3
3.8
5.4
4.0
5.4
3.9
5.4
4.0
5.4
3.9
5.5
4.0
5.5
4.0
5.5
4.1
5.S
4.0
5.6
4.1
5.6
4.1
5.6
4.2
5.6
4.1
5.7
4.2
5.7
4.2
5.7
4.2
5.7
. 4.1
5.8
4.3
5.8
4.2
5.8
4.3
5.8
4.2
5.9
4.4
5.9
4.3
5.9
4.4
5.9
4.J
6.0
4.S
6.0
4.4
6.0
4.5
6.0
4.4
6.1
4.5
6.1
4.5
6.1
4.6
6.1
4.5
6.2
4.6
6.2
4.6
6.?
4.7
6.2
4.6
6.3
4.7
6.3
4.7
6.3
4.U
6.3
4.7
6.4
4.B
6.4
4 .H
6.4
4.9
6.4
4.8
6.5
4.9
6.5
4.8
6.5
4.9
6.5
4.9
6.6
s.o
6.6
4.9
6.A
5.0
6.6
4.9
6.7
5.0
6.7
5.0
6.7
S.l
6.7
5.0
6 . 8
5.1
6.8
5.1
6.8
5.2
6.8
5.1
6.9
5.2
6.9
5.2
6.9
5.3
6.9
5.2
7.0
5.3
7.0
b.3
7.0
5.4
7.0
5.1
7.1
5.4
7.1
5.4
7.1
5.S
7.1
5.4
7.2
S.S
7.2
5.4
7.2
5.6
7.2
5.S
7.3
5.5
7.3
5.5
7.3
5.6
7.3
5.6
7.4
5.6
7.4
5.6
7.4
5.7
7.4
5.6
7.5
5.7
7.5
5.7
7.5
5.8
7.5
5.7
7.6
5.B
7.6
5.8
7.6
5.9
7.6
5.8
7.7
5.9
7.7
5.9
7.7
6.0
7.7
5.9
7.6
6.0
7.8
6.0
7.8
6.1
7.8
6.0
7.9
6.0
7.9
6.1
7.9
6.2
7.9
6.1
0003 PROCESSED! AUG 9. 1979 0BI44t3S PAGE 1 OF 4
VERSION! 1.23
NOTElFOR AHP VALUES OTHER THAN THOSE GIVEN
IN THE TABLE• USE THE FOLLOWING FORMULA*
HOUNOINO THE RESULT TO THE NEAREST TENTMJ
TW« M • AMP ~ B
1500
1625
17S0
1875
0.865
0.895
0.908
0.911
-0.
80
-0.
96
-1.
02
-1.
16
A HP
tw
A HP
TW
AHP
TW
AHP
TW
3.0
1.9
3.0
1.7
3.0
1.7
3.0
1.6
3.1
1.9
3.1
1.8
3.1
1.8
3.1
1.7
3.2
2.0
3.2
1.9
3.2
1.9
3.2
i.a
3.3
2.1
3.3
2.0
3.3
2.0
3.3
1.8
3.4
2.2
3.4
2.1
3.4
2.1
3.4
1.9
3.5
2.3
3.5
2.2
3.5
2.2
3.5
2.0
3.6
2.4
3.6
2.3
3.6
2.3
3.6
2.1
3.7
2.5
3.7
2.4
3.7
2.3
3.7
2.2
3.8
2.6
3.8
2.4
3.8
2.4
3.8
2.3
3.9
2.7
3.9
2.5
3.9
2.S
3.9
2.4
4.0
2.7
4.0
2.6
4.0
2.6
tt.O
2.5
4.1
2.8
4.1
2.7
4.1
2.7
4.1
2.6
4.2
2.9
4.2
2.8
4.2
2.8
4.2
2.7
4.3
3.0
4.3
2.9
4.3
2.9
4.3
2.8
4.4
3.1
4.4
3.0
4.4
3.0
4.4
2.8
4.5
3.2
4.5
3.1
4.5
3.1
4.5
2.9
4.6
3.3
4.6
3.2
4.6
3.2
4.6
3.0
4.7
3.4
4.7
3.2
4.7
3.3
4.7
3.1
4.8
3.4
4.8
3.3
4.8
3.3
4.8
3.2 -
4.9
3.5
4.9
3.4
4.9
3.4
4.9
3.3
5.0
3.6
S.O
3.5
5.0
3.5
5.0
3.4
5.1
3.7
5.1
3.6
5.1
3.6
S.l
3.5
5.2
3.8
S.2
3.7
5.2
3.7
5.2
3.6
S.3
3.9
S.3
3.8
5.3
3.8
5.3
3.7
5.4
4.0
5.4
3.9
5.4
3.9
5.4
3.«
5.5
4.1
5.5
4.0
5.5
4.0
5.5
3.9
5.6
4.2
5.6
4.1
5.6
4.1
5.6
3.9
5.7
4.2
5.7
4.1
5.7
4.2
5.7
4.0
5.8
4.3
5.8
4.2
5.8
4.2
5.8
4.1
5.9
4.4
5.9
4.3
5.9
4.3
5.0
4.2
6.0
4.5
6.0
4.4
6.0
4.4
6.0
4.3
6.1
4.6
6.1
4.5
6.1
4.5
6.1
4.4
6.2
4.7
6.2
4.6
6.2
4.6
6.2
4.5
6.3
4.8
6.3
4.7
6.3
4.7
6.3
4.6
6.4
4.9
6.4
4.B
6.4
4.a
6.4
4.7
6.5
5.0
6.5
4.9
6.5
4.9
6.5
4. R
6.6
5.0
6.6
4.9
6.6
5.0
6.6
4.9
6.7
S.l
6.7
5.0
6.7
S.l
6.7
4.9
6.8
5.2
6.B
5.1
6.8
5.2
6.8
S.O
6.9
5.3
6.9
5.2
6.9
S.2
6.9
5.1
7.0
5.4
7.0
5.3
7.0
5.3
7.0
5.2
7.1
5.5
7.1
5.4
7.1
S.4
7.1
S.3
7.2
5.6
7.2
5.5
7.2
5.5
7.2
5.4
7.3
5.7
7.3
5.6
7.3
5.6
7.3
5.5
7.4
5.8
7.4
5.7
7.4
5.7
7.4
5.6
7.5
5.8
7.5
5.8
7.5
5.8
7.5
5,7
7.6
5.9
7.6
5.B
7.6
5.9
7.6
5.8
7.7
6.0
7.7
5.9
7.7
6.0
7.7
5.9
7.8
6.1
7.8
6.0
7.B
6.1
7.8
5.9
7.-;
6.2
7.9
6.1
7.9
6.2
7.9
6.0
-------
vO
00000
OOOOO
OOOOC
i 33333
dynamometer
CALIBRATION TAdLE
FOR 0003
PROCESSED*
AUG 9« 1979 08:44I3S PAGE 2 OF 4
0 0
0
0 0
0 3
3
DERIVED
FROM
> AHP VERSUS TW DATA
VERSION!
1.23
D 0
0
0 0
0
3
calibration
DATEt 07-!
12-79
NOTE>FOR AHP VALUES OTHER
THAN THOSE
GIVEN
D 0
0
0 0
0
333
IN THE
; TABLEt USE THE
FOLLOWING
FORMULA.
0 0
0
0 0
0
3
APPROVED BY I
ROUNOINO THE RESULT TO THE NEAREST TENTH!
0 0
0
0 0
0 3
3
00000
OOOOO
00000
i 33333
EFFECTIVE DA
1
TWa M
• AHP ~ 8
INERTIA
2000
2125
22S0
2375
2500
2625
2750
2875
•• M ••
0.413
0.921
0.921
0.932
0.930
0.939
0.940
0.940
• • B ••
-0.
99
-1.
15
-1.
14
-1.
.32
-1.
28
-1.43
-1.
44
-1.5S
AHP
TW
AHP
TW
AHP
TW
AHP
TW
AHP
TW
AHP
TW
AHP
TW
AHP
TW
6.0
4.5
6.0
4.4
6.0
4.4
6.0
4.3
6.0
4.3
6.0
4.2
6.0
4.2
6.0
4.1
6.1
4.6
6.1
4.5
6.1
4.5
6.1
4.4
6.1
4.4
6.1
4.3
6.1
4.3
6.1
4.2
6.2
4.7
6.2
4.6
6.2
4.6
6.2
4.5
6.2
4.5
6.2
4.4
6.2
4.4
6.2
4.2
6.3
4.8
6.3
4.7
6.3
4.7
6.3
4.5
6.3
4.6
6.3
4.5
6.3
4.5
6.3
4.3
6.4
4.9
6.4
4.7
6.4
4.8
6.4
4.6
6.4
4.7
6.4
4.6
6.4
4.6
6.4
4.4
6.S
4.9
6.5
4.8
6.5
4.8
6.S
4.7
6.5
4.8
6.5
4.7
6.5
4.7
6.5
4.5
6.6
5.0
6.6
4.9
6.6
4.9
6.6
4.8
6.6
4.9
6.6
4.8
6.6
4.8
6.6
4.6
6.7
5.1
6.7
5.0
6.7
5.0
6.7
4.9
6.7
5.0
6.7
4.9
6.7
4.9
6.7
4.7
6.8
5.2
6.8
5.1
6.H
5U
6.8
5.0
6.8
S.l
6.8
4.9
6.8
5.0
6.8
4.8
6.9
5.3
6.9
5.2
6.9
5.2
6.9
5.1
6.9
5.1
6.9
5.0
6.9
5.0
6.9 "
4.9
7.0
5.4
7.0
S.3
7.0
5.3
7.0
5.2
7.0
5.2
7.0
5.1
7.0
5.1
7.0
5.0
7.1
5.5
7.1
5.4
7.1
5.4
7.1
5.3
7.1
5.3
7.1
5.2
7.1
S.2
7.1
5.1
7.2
5.6
7.2
S.b
7.2
b.S
7.2
5.4
7.2
S .4
7.2
5.3
7.2
5.3
7.2
5.2
7.3
5.7
7.3
5.6
7.3
5.6
7.3
5.5
7.3
5.5
7.3
5.4
7.3
5.4
7.3
S.3
T.4
5.a
7.4
5.7
7.4
5.7
7.4
5.6
7.4
5.6
T.4
5.5
7.4
5.5
7.4
5.4
7.5
S.9
7.5
s;«
7.5
5.8
7.5
5.7
7.5
5.7
7.5
5.6
7.5
5.6
7.5
5.5
7.6
6.0
7.6
5.8
7.6
5.9
7.6
5.8
7.6
5.8
7.6
5.7
7.6
5.7
7.6
5.6
7.7
6.0
7.7
5.9
7.7
5.9
7.7
5.9
7.7
5.9
7.7
5.8
7.7
5.8
7.7
5.7
7.8
6.1
7.B
6.0
7.8
6.0
7.8
5.9
7.a
6.0
7.8
5.9
7.8
5.9
7.8
5.7
7.9
6.2
7.9
6.1
7.9
6.1
7.9
6.0
7.9
6.1
7.9
6.0
7.9
6.0
7.9
5.8
8.0
6.3
8.0
6.2
8.0
6.2
a.o
6.1
8.0
6.2
8.0
6.1
8.0
6.1
8.0
5.9
8.1
6.4
b.l
6.3
tt.l
6.3
8.1
6.2
8.1
6.3
8.1
6.2
8.1
6.2
8.1
6.0
8.2
6.5
8.2
6.4
8.2
6.4
8.2
6.3
8.2
6.4
8.2
6.3
8.2
6.3
8.2
6.1
8.3
6.6
8.3
6.5
8.3
6.5
8.3
6.4
8.3
6.4
8.3
6.4
8.3
6.4
8.3
6.2
8.4
6.7
8.4
6.6
8.4
6.6
8.4
6.5
(1.4
6.5
8.4
6.4
8.4
6.5
8.4
6.3
8 .5
6.8
8.5
6.7
8.5
6.7
8.5
6.6
8.5
6.6
8.5
6.5
8.5
6.6
8.5
6.4
8.6
6.9
8.6
6.8
8.6
6.8
8.6
6.7
8.6
6.7
8.6
6.6
8.6
6.6
8.6
6.5
8.7
7.0
8.7
6.9
8.7
5.9
8.7
6.8
8.7
6.8
8.7
6.7
8.7
6.7
8.7
6.6
8.8
7.0
8.6
7.0
8.8
7.0
8.8
6.9
8.8
6.9
8.8
6.8
8.8
6.8
8.8
6.7
ft. 9
7.1
fe.9
7.0
8.9
7.1
8.9
7.0
8.9
7.0
8.9
6.9
8.9
6.9
8.9
6.8
9.0
7.2
9.0
7.1
9.0
7.1
9.0
7.1
9.0
7. 1
9.0
7.0
9.0
7.0
9.0
6.9
9.1
7.3
9.1
7.2
9.1
7.2
9.1
7.2
9.1
7.2
9.1
7.1
9.1
7.1
9.1
7.0
9.2
7.4
9.2
7.3
9.?
7.3
9.2
7.3
9.2
7.3
9.2
7.2
9.2
7.2
9.2
7.1
9.3
7.5
9.3
7.4
9.3
7.4
9. J
7.3
9.3
7.4
9.3
7.3
9.3
7.3
9.3
7.2
9.4
7.6
9.4
7.5
9.4
7.5
9.4
7.4
9.4
7.5
9.4
7.4
9.4
7.4
9.4
7. J
9.5
7.7
9.5
7.6
'9.5
7.6
9.5
7.5
9.5
7.6
9.5
7.5
9.5
7.5
9.5
7.3
9.6
7.8
9.6
7.7
9.6
7.7
9.6
7.6
9.6
7.7
9.6
7.6
9.6
7.6
9.6
7.4
9.7
7.9
9.7
7.8
9.7
7.8
9.7
7.7
9.7
7.7
9.7
7.7
9.7
7.7
9.7
7.5
9.6
a.o
9.8
7.9
9.8
7.9
9.8
7.8
9.8
7.8
9.8
7.8
9.8
7.8
9.8
7.6
9.9
8.1
9.9
8.0
9.9
8.0
9.9
7.9
9.9
7.9
9.9
7.9
9.9
7.9
9.9
7.7
10.0
8.1
10.0
8.1
10.0
8.1
10.0
8.0
10.0
R.O
10.0
8.0
10.0
8.0
10.0
7.8
10.1
8.2
10.1
8.2
10.1
8.2
10.1
8.1
10.1
8. 1
10.1
8.0
10.1
8.1
10.1
7.9
10.2
8.3
10.2
8.2
10.2
*.2
10.2
8.2
10.2
8.2
10.2
8.1
10.2
8.1
10.2
R.O
10.3
A.4
10.3
8.3
10.3
8.3
10.3
8.3
10.3
6.3
10.3
8.2
10.3
8.2
10.3
8. 1
10.4
8.5
10.4
8.4
10.4
8.4
10.4
8.4
10.4
n.4
10.4
8.3
10.4
8.3
10.4
8.2
10.5
8.6
10.b
8.5
10.5
a.5
10. 5
8.5
10.5
fl.5
10. 5
8.4
10.S
8.4
10.5
A.3
10.6
8.7
10.6
8.6
10.6
8.6
10.6
8.6
10.6
0.6
10.6
8.5
10.6
8.S
10.6
8.4
10.7
8.8
10.7
8.7
10.7
8.7
10.7
8.6
10.7
H.7
10.7
8.6
10.7
8.6
10.7
8.5
10. A
fl.9
lo.a
8.A
10. H
8.8
10.8
H.7
10.8
M.6
10.8
8.7
10.8
8.7
10.8
8.6
10.9
9.0
10.9
8.9
10.9
a.9
10.9
U.B
10.9
0.9
10.9
0.0
10.9
8.8
10.9
6.7
-------
f-
00000
0 0
0
0
0
0 0
00000
00000
coooo
00000
0 0
0 0
0 0
0 0
0 0
00000
33333
333
oynamometer calibration table row 0003
DERIVED FROM AHP VERSUS Tw DATA
CALIBRATION 0ATEI 07-12-79
33333
APPKOVEO BYI_
EFFECTIVE DATE >.
PROCESSED! AUG 9. 1974 0814*135 PAGE 3 OF V
VERSIONS 1.23
N0TE»F0R AMP VALUES OTHER THAN THOSE GIVEN
IN THE TABLE• USE THE FOLLOWING FOKMULA,
HOUNDING THE RESULT TO THE NEAREST TENTH:
TW3 M • AHP ~ B
INERTIA
• • M ••
• • B ••
3000
3125
3250
3375
0.9S]
0.95ft
*0.9S9
0.96S
-1.
14
-1.
34
-1.
38
-1.
59
AHP
rw
AHP
TW
AHP
TW
AHP
TW
8.0
6.5
a.o
6.3
a.o
6.3
tt.O
6.1
8.1
6.6
B.l
6.4
8.1
6.4
8.1
6.2
8.2
6.7
8.2
6.5
a.2
6.5
a.2
6.3
8.3
6.S
8.3
6.6
8.3
6.6
8.3
6.4
8.4
6.9
8.4
6.7
8.4
6.7
8.4
6.5
8.5
6.9
8.5
6.8
a.s
6.8
a.s
6.6
8.6
7.0
8.6
6.9
8.ft
6.9
8.6
6.7
8.7
7.1
a.7
7.0
8.7
7.0
a.7
6.8
8.8
7.2
8.a
7.1
8.a
7.1
a.a
6.9
8.9
7.3
£.9
7.2
8.9
7.2
8.9
7.0
9.0
7. 4
9.0
7.3
v.o
7.3
9.0
7.1
9.1
7.5
9.1
r. 4
9.1
7.4
9.1
7.2
9.2
7.6
9.2
7.5
9.2
7.4
•*.£
7.3
9.3
7.7
9.3
7.6
9.3
7.5
9.3
7.4
9.4
7.8
9.4
7.6
9.4
7.6
9.«
7.5
9.5
7.9
9.5
7.7
9.5
7.7
9.5
7.6
9 .6
8.0
9.6
7.a
9.6
7.8
9.6
7.7
9.7
a.i
9.7
7.9
9.7
7.9
9.7
7.8
9.6
8.2
9.8
H . 0
9.8
8.0
9.8
7.9
9.9
8.3
9.9
8.1
9.9
8.1
9.9
8.0
10.0
a.4
10.0
8.2
10.0
8.2
10.0
a.i
10.1
8.5
10.1
4.3
10.1
8.3
10.1
a.2
10.2
a.6
10.2
8.4
10.2
' a.4
10.2
a.3
10.3
a.7
10.3
8.5
10.3
a.s
10.3
4.3
10.4
8.8
10.4
8.6
10.4
8.6
10.4
8.4
10. b
8.a
10.5
8.7
10.5
8.7
10.5
8.5
10. ft
a.9
10.6
a.a
10.6
8.8
10.6
a.6
10.7
9.0
10.7
a.9
10.7
8.9
10.7
a.7
10.8
9.1
10.8
1.0
10. 8
9.0
10.8
H.H
10.9
9.2
10.9
9.1
10.9
9.1
10.9
a.9
11.0
9.3
11.0
9.2
11.0
9.2
11.0
9.0
11.I
9.4
11.1
9.3
11.1
9.3
11.1
9.1
11.2
9.5
11.2
9.4
11.2
9.4
11.2
9.2
11.3
9.6
11.3
9.5
11.3
9.5
11.3
9.3
11. 4
9.7
11.4
9.6
11.4
9.6
11.4
9.4
11.5
9.8
11.5
9.7
11.5
9.7
11.5
9.5
11.6
9.9
11.6
9.7
11.6
9.8
11.6
9.6
11.7
10.0
11.7
9.0
11.7
9.8
11.7
9.7
11.0
10.1
1 i.a
9.9
11.8
9.9
11. U
9.a
11.9
10.2
11.9
10.0
11.9
10.0
11.9
9.9
12.0
10.3
12.0
10.1
12.0
10.1
12.0
10.0
12.1
10. 4
12.1
10.2
12.1
10.2
12.1
10.1
12.2
10.5
12.2
10.3
12.2
10.3
12.2
10.2
12.3
10.6
12.3
10.4
12.3
10.4
12.3
10.3
12.4
10.7
12.4
10.5
12.4
10.5
12.4
10.4
12.5
io.a
12.5
10.6
12.5
10.6
12.5
10.5
12.6
io.a
12.6
10.7
12.6
10.7
12.6
10.6
12.7
10.9
12.7
10.a
12.7
10.8
12.7
10.7
12.9
11.0
12.6
10.9
12.8
10.9
-12.8
10.a
12.9
11.1
12.9
11.0
12.9
11.0
12.9
10.9
3500
0.962
-1.48
AHP
TW
11.0
9.1
11.1
9.2
11.2
9.3
11.3
9.4
11.4
9.5
11.5
9.6
11.6
9.7
11.7
9.8
11.8
9.9
11.9
10.0
12.0
10.1
12.1
10.2
12.2
10.3
12.3
10.4
12.4
10.5
12.S
10.5
12.6
10.6
12.7
10.7
12.A
10. 8
12.9
10.9
1.3.0
11.0
13.1
11.1
13.2
11.2
13.3
11.3
13.4
11.4
13.5
11.5
13.6
11.6
13.7
11.7
13.8
11.8
13.9
11.9
14.0
12.0
14.1
12.1
14.2
12.2
14.3
12.3
14.4
12.4
14. 5
12.5
14.6
12.6
14. 7
12.7
14.A
12.3
14.9
12.9
15.0
13.0
15.1
13.0
IS.2
13.1
15.3
13.2
15.4
13.3
15.5
13.4
15.6
13.5
15.7
13.6
IS.8
13.7
IS.9
13.8
3625
3750
3875
0.966
0.964
0.960
-1.
6b
-1.
67
-1
75
AHP
TW
AHP
TW
AHP
TW
11.0
8.9
11.0
8.9
1 1.0
8.8
11.1
9.0
11.1
9.0
11.1
8.9
11.2
9.1
11.2
9.1
11.2
9.0
11.3
9.2
11.3
9.2
11.3
9.1
11.4
9.3
11.4
9.3
11.4
9.2
11.5
9.4
11.5
9.4
11.5
9.3
11.6
9.5
11.6
9.5
11.6
9.4
11.7
9.6
11.7
9.6
11.7
9.S
u.a
9.7
11.8
9.7
ii.a
9.ft
11.9
9.A
11.9
9.8
11.9
9.7
12.0
9.9
12.0
9.9
12.0
9.H
12.1
10.0
12.1
10.0
12.1
9.9
12.2
10.1
12.2
10.1
12.2
10.0
12.3
10.2
12.3
10.2
12.3
10.1
12.4
10.3
12.4
10.3
12.4
10.1
12.5
10.4
12.5
10.4
12.5
10.2
12.6
10.5
12.6
10.5
12.6
10.3
12.7
10.6
12.7
10.6
12.7
10.4
12.8
10.7
12.a
10.7
12.8
10.5
12.9
10.a
12.9
10.8
12.9
10.ft
13.0
10.9
13.0
10.9
13.0
10.7
13.1
11.0
13.1
11 .0
13.1
10.t
13.2
11.1
13.2
11.1
13.2
10.9
13.3
11.2
13.3
11.2
13.3
11.0
13.4
11.3
13.4
11.3
13.4
11.1
13.5
11.4
13.5
11 .4
13. S
11.2
13.6
11.5
13.6
11.4
13.6
11.3
13.7
11.6
13.7
11.5
13.7
11.4
13.8
11.6
13.8
11.6
13.8
11.5
13.9
11.7
13.9
11.7
13.9
11.6
14.0
11. B
14.0
11.a
14.0
11.7
14.1
11.9
14.1
11.9
lt.l
11.6
14.2
12.0
14.2
12.0
14.2
11.9
14.3
12.1
14.3
12.1
14.3
12.0
14.4
12.2
14.4
12.2
14.4
12.1
14.5
12.3
14. 5
12.3
14.5
12.2
14.6
12.4
14.6
12.4
14.6
12.3
14.7
12.5
14.7
12.5
14.7
12.4
14.8
12.6
14. S
12.6
1*. A
12.5
14.9
12.7
14.9
12.7
1«.9
12.5
15.0
12.a
15.0
12.8
15.0
12. 4
15.1
12.9
15.1
12.9
15.1
12.7
15.2
13.0
15.2
13.0
15.2
12.4
IS.3
13.1
15.3
13.1
IS.3
12.9
15.4
13.2
15.4
13.2
IS.4
13.0
15.5
13.3
IS.5
13.3
15.5
13.1
15.6
13.4
15.6
13.4
15.6
13.2
15.7
13.5
15.7
13.5
15.7
13.3
15.8
13.6
1S.S
13.6
15.0
13. 4
15.9
13.7
15.9
13.7
IS.9
13.5
-------
0&
00000
0 0
o o
0 D
0 0
0 0
00000
INERTIA
• • M ••
•• B ••
00000
0 0
0 0
0 0
0 0
0 0
00000
<•000
0.968
- 1 • 64
00000
0 0
0 0
0 0
0 0
0 0
00000
33333
333
33333
A HP
10.0
10.2
10. 4
10.6
10. 8
11.0
11.2
11.4
11.6
11. 8
12.0
12.2
TW
B.O
8.2
e.4
8.6
ft. 8
9.0
9.2
9.*
9.6
9.8
10.0
10.2
12. * 10.4
12.6 10.6
12.8 10.8
13.0 11.0
13.? 11.1
13. 4 11.3
13.6 11.5
13.8 11.7
14.0 11.9
14.2 12.1
14.4 12.3
1<>.6 12.5
14.8 12.7
15.0 12.9
IS.2 13.1
15.* 13.3
15.6 13.5
IS.8 13.7
16.0 13.9
16.2 14.1
16.
.4 1<>.2
lb.6 14,4
16.4 14.6
17.0 14.8
17.2 1S.0
17.4 IS.2
17.6 15.4
17.8 15.6
18.0 15.8
lb.2 16.0
Id.4 16.2
18.6 16.4
18.8 16.6
19.0 16.8
19.2 17.0
IV.4 17.2
19.6 17.3
19.B 17.5
4250
0.970
-I.BO
AMP TW
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
7.9
H.l
8. J
8.5
'8.7
8.9
9.1
V.3
9.5
9.7
9.8
2.2 10.0
2.4 10.2
2.6 10.4
2.8 10.6.
3.0
3.2
3.4
3.6
10.8
11.0
11.2
11.4
3.8 11.6
4.0 11.8
4.2 12.0
4.4 12.2
4.6 12.4
4.8 12.6
5.0 12.8
5.2 13.0
5.4 13.1
5.6 13.3
5.8 13.5
6.0 13.7
6.2 13.9
6.4 14.1
6.6 14.3
6.8 14.5
7.0
7.2
14.7
14.9
7.4 15.1
7.6 15.J
7.8
8.0
8.2
15.5
15.7
15.V
8.4 16.1
b .6 16.3
8.8
9.0
9.2
9.4
16.4
16.6
16.8
17.0
9.6 17.2
9.8 17.4
DYNAMOMETER CALIBRATION TABLE FOR 0003
0ER1VEU FROM AHP VERSUS TW DATA
CALIBRATION DATEl 07-12-7«
ApptfovEo av«
EFFECTIVE OATEI.
4500
0.973
-1.92
AMP
10.0
10.2
10.4
10.6
10.U
11.0
11.2
11.4
11.6
11.8
12.0
12.2
12.4
TW
7.8
8.0
8.2
6.4
8.6
8.8
9.0
9.2
9.4
9.6
9.8
9.9
10.1
12.6 10.3
12.8 10.5
13.0
13.2
10.7
10.9
13.4 11.I
13.6 11.3
13.8 11.5
14.0 11.7
14.2 11.9
14.4 12.1
14.6 12.3
14.8 12.5
15.0
15.2
15.4
15.6
12.7
12.9
13.1
13.3
15.8 13.5
16.0 13.6
16.2
16.4
16.6
13.8
14.0
14.2
16.8 14.4
17.0 14.6
17.2 14.8
17.4 15.0
17.6
17.8
18.0
18.2
IB.4
18.6
15.2
IS.4
15.6
15.8
16.0
16.2'
18.8 16.4
19.0 16.6
19.2 16.8
19.4 17.0
19.6 17.1
19.8 17.3
4750
0.969
-2.00
AHP
10.0
10.2
10.4
10.6
10.8
11.0
11.2
11.4
11.6
11.8
12.0
12.2
12.4
12.6
12.H
13.4
13.6
15.6
15.8
TW
7.7
7.9
8.1
8.3
8.5
8.7
B.8
9.0
9.2
9.4.
9.6
9.8
10.0
1H.2
10.4
13.0 10.6
13.2 10.8
11.0
11.2
13.8 11.4
14.0 11.6
14.2
14.4
14.6
11.8
11.9
12.1
14.a 12.3
15.0 12.5
15.2 12.7
15.4 12.9
13.1
13.3
16.0 13.5
16.2 13.7
16.4 13.9
16.6 14.1
16.8 14.3
17.0 14.5
17.2 14.7
17.4 14.9
17.6 15.0
17.8 15.2
18.0 IS.4
18.2 15.6
18.4 IS.8
18.6 16.0
18.8 16.2
19.0 16.4
19.2 16.6
19.4 16.8
19.6 17.0
19.8 17.2
5000
0.V86
-2.02
AMP
13.0
13.2
13.4
13.6
TW
10.8
11.0
11.2
11.4
13.8 11.6
14.0 11.8
14.2 12.0
14.4 12.2
14.6 12.4
14.8 12.6
15.
15.
12.8
13.0
15.4 13.2
15.6 13.4
IS. 8
16.0
13. 5
13.7
16.2 13.9
16.4 14.1
16.6 14.3
lb.8 14.5
17.0 14.7
17.2 14.9
17.4 15.1
17.6 15.3
17.3 15.5
18.0 15.7
18.2 15.9
18.4 16.1
18.6 16.3
18.8 16.5
19.0 16.7
19.2 lb.9
19.4 17.1
19.6 17.3
19.8 17.5
20.0 17.7
20.2 17.9
20.4 18.1
20.6 18.3
20.8 18.5
21.0 18.7
21.2
21.4
21.6
21 .8
22.0
22.2
18.9
19.1
19.3
19.5
19.7
19.9
22.4 20.1
22.6 20.3
22.8 20.4
PROCESSED I AUG 9. 1979 08:44135 PAGE 4 OF 4
VERSION! 1.23
NOTEIFOH AHP VALUES OTHER THAN THOSE GIVEN
IN THE TABLEf USE THE FOLLOWING FORMULAt
HOUNDING THE RESULT TO THE NEAREST TENTH1
5250
0.984
-2.16
AHP
14.0
14.2
14.4
16.0
16.2
22.2
22.4
TW
11.6
11.8
12.0
14.6 12.2
14.8 12.4
15.0 12.6
15.2 12.8
15.4 13.0
15.6 13.2
15.8 13.4
13.6
13.8
16.4 14.0
16.6 14.2
16.8 14.4
17.0 14.6
17.2 14.8
W.4 15.0
17.6 15.2
17.8 15.4
18.0 15.6
18.2 15.8
18.4 15.9
13.6 16.1
18.8 16.3
19.0 16.5
19.2 16.7
19.4 16.9
19.6 17.1
19.8 17.3
20.0 17.5
20.2 17.7
20.4 17.9
20.6 18.1
20.8 18.3
21.0 18.5
21.2 18.7
21.4 18.9
21.6 19.1
21.8 19.3
22.0 19.5
19.
19
22.6 20.1
22.8 20.3
20.5
20.7
.9
21.1
23.0
23.2
23.4 20.
23.6
23.8 21.3
TW= M • AMP
5500
0.985
-2.33
AHP TW
15.0
15.2
15.4
12.5
12.7
12.8
15.6 13.0
15.8 13.2
16.0 13.4
16.2 13.6
16.4
16.6
13.8
14.0
16.8 14.2
17.0 14.4
17.2 14.6"
17.4 14.8
17.6 15.0
17.8 15.2
18.0 15.4
18.2 15.6
18.4 15.8
18.6 16.0
18.6 16.2
19.0 16.4
19.2 16.6
19.4 16.8
19.6 17.0
19.8 17.2
20.0
20.2
17.4
17. b
20.4 17.8
20.6 18.0
20.8 18.2
21.0 18.4
21.2 18.6
21.4 18.8
21.6 19.0
21.8 19.2
22.0 19.4
22.2 19.6
22.4 19.7
22.6 19.9
22.8 20.1
23.0 20.3
23.2 20.5
23.4 20.7
23.6 20.9
23.8 21.1
24.0 21.3
24.2
24.4
21.5
21.7
24.6 21.9
24.8 22.1
B
6000
0.990
-2.54
AHP
TW
15.0 12.3
15.2 12.5
15.4 12.7
15.6 12.9
15. R
16.0
13.1
13.3
16.2 13.5
16.4 13.7
16.6 13.9
16.8 14.1
17.0 14.3
17.2 14.5
17.4 14,7
17.6 14.9
17.R 15.1
18.0 15.3
18.2 15.5
18.4 15.7
18.6 15.9
18.8 16.1
19.0
19.2
16.3
16.5
19.4 16.7
19.6 16.9
19.8 17.1
20.0 1/.3
20.2 17.4
20.4 17.6
20.6 17.8
20.8 18.0
21.0 18.2
21.? 18.4
21.4 18.6
21.6 18.8
21.8
22.0
22.2
22.4
22.6
19.0
19.2
19.4
19.6
19.8
22.8 20.0
23.0 20.2
23.2
23.4
20.4
20.6
23.6 20.8
23.8 21.0
24.0 21.2
24.2 21.4
24.4
24.6
21.6
21.8
6500
0.988
-2.71
AHP TW
16.0 13.1
16.2 13.3
16.4 13.5
16.6 13.7
16.8 13.9
17.0 14.1
17.2 14.3
17.4 14.5
17.6 14.7
17.8 14.9
18.0 15.1
18.2 IS.3
18.4 15.5
18.6 15.7
18.8 15.9
19.0 16.1
19.2 16.3
19.4 16.5
19.6 16.7
19.8 16.9
20.0 17.1
20.2 17.2
20.4 17.4
20.6 17.6
20.8 17.8
21.0 1*.0
21.2 18.2
21.4 18.4
21.6 18.6
21.8 18.8
22.0 19.0
22.2
22.4
22.6
19.2
19.4
19.6
22.6 19.8
23.0 20.0
23.2 20.2
.4 20.4
.6 20.6
23.8 20.8
24.0 21.0
24.2 21.2
24.4 21.4
24.6 21.6
24.8 21.8
25.0 22.0
.2 22.2
.4 22.4
.6 22.6
23.
23.
25.
25.
25.
24.8 22.0
25.8 22.8
-------
ATTAcHMtNlT 31
Calibration of Chassis Dynamometers
for Emissions- and Fuel Economy Testing
of Passenger Cars
Recommended by
CEC-CF 22 project group
-------
FOREWORD
A general calibration method for chassis dynamometers is an important
prerequisite for obtaining comparable emissions- and fuel economy mean
test results with reasonable confidence.
Recent studies of several investigators had shown that the methods currently
in use for calibrating the load behaviour of chassis dynamometers are un-
satisfactory in some respects.
As an example, the power absorption unit is calibrated in only one point
of velocity. In practice, it is quite possible for two dynamometers which
absorb the same amount.of power at the calibration point to differ markedly
at any speed below this point, giving rise to non-comparable emissions- or
fuel economy test results.
In order to provide recommendations for a more suitable chassis dynamometer
calibration procedure a project group was established in January 1977. This
group - called CF 22 - is integrated into the C.E.C. Engine Fuels Technical
Committee.
As a result of our studies within the CF 22 group, we present in this report
a proposal which is in our opinion suitable for a standardized chassis dyna-
mometer calibration procedure.
November 1978
D. Schlirmann
-------
- 1 -
Project group:
W. Berg
Daimler Benz
Germany
J.P. Bowskill
British Leyland Cars
Great Britain
M. Buith&
Ecole Royale Militaire
Belgium
F. Cavallino
Fiat
Italy
D. Dundas
Esso
Great Britain
R. Homann
Schenck
Germany
L. Johnsson
Volvo
Sweden
M. Poillot
Citroen
France
D. SchUrmann
Volkswagen
Germany
J.P. Soltau
Automotive Eng. Cons.
Great Britain
S. Vaccaneo
Fiat
Italy
M. Veron
Empa
Switzerland
J. Zaalberg
Clayton
Belgium
For any comments please
contact:
Dr. D. Schiirmann
Volkswagenwerk AG, Wolfsburg
Mefttechnik - Forschung 5
AbgasmeBtechnik
Tel: 05361/22-6928
Tx : 095 86 533 vww d
-------
- 2 -
Table of Contents
Introduction
1. Description of Method
1.1 General
1.2 Physical Background
2. Types of Chassis Dynamometers
2.1 General
2.2 Characteristic Properties
3. Comparison of Torque Method to other Methods
3.1 Review of Existing Methods
3.1.1 Inertia Weight Classes
3.1.2 Intake Manifold Pressure
3.1.3 Aerodynamical Assumptions
3.1.4 Coast Down
3.2 Torque vs Coast Down Method
4. Apparatus
4.1 Torque Meters
4.2 Electronic Equipment
5. Test Procedure
5.1 Measurements on the Road
5.2 Adjusting the Dynamometer
6. Evaluation and Reporting of Results
6.1 Constant Speed Torque
6.2 Integrated Torque over Variable Driving Pattern
7. Precision of Torque Method
7.1 Constant Speed Torque
7.2 Integrated Torque over Variable Driving Pattern
8. Summary and Conclusion
-------
- 3 -
Introduction
According to e.g. the European procedure the load of the dynamometer is
adjusted at a speed of 50 km/h, in the US at 80 km/h. This does not
furnish any data regarding the behaviour of the power absorber below
50 km/h or 80 km/h, respectively, giving rise to a arave problem, because
the major proportion of the ECE test is run in the speed range below 50 km/h,
that of the US test below 80 km/h (s. Fig. 1).
Cumulative frequency
100
%
80
60
t
AO
20
0
0 20 £0 60 km/h 100
100
%
80
L0
20
0
0 20 AO 60 km/h 100
Car speed —*>
Fig. 1: Cumulative Frequency of Speeds in the European- and US Driving
Cycles.
Nearly 90 % of all driven speeds are below 40 km/h (ECE) or below
60 km/h (US).
1
t
1
|
i
ECE 1
,
1
1 1
! I
i 1
i M
\ w
s
I
US 75
1
!
-------
- 4 -
According to Fig. 1, 90 % of all driving modes involve speeds below 40 km/h
in the case of the European test and below 60 km/h in the case of the US
test.
Therefore, the way in which the power absorption unit behaves in these
ranges is of decisive influence on the exhaust emissions- and fuel economy
test results of the vehicle under test.
Special problems are raised if, according to the European regulation (and also
to the US regulation in the past time) the load resistance of the dynamometer
is adjusted by using the manifold vacuum pressure value measured when driving
on the road at 50 km/h (ECE) or 80 km/h (US). With this method precision and re-
producibility' are insufficient since generally the intake manifold vacuum
values are widely dispersed between 20 km/h and 70 km/h, being strongly de-
pendent on such factors as engine type, emission control system, driver.
A typical example for this situation is displayed in Fig. 2.
200
MM
" IBB
IH0
W I 20
ID
a
[£ I 00
~
00
G0
H2
20
Manifold vacuum
0.6
bar
0.3
Torque
10 20 30 40 S0 S0 70 B0 KM/H 100
CRR 5PEED -
Fig. 2: Manifold Vacuum as a Function of Car Speed.
The spread of the data below 70 km/h is considerable.
For Comparision the torque on the drive shafts of the test vehicle
is also plotted, showing much less dispersion.
-------
- 5 -
As a consequence, the power absorber loading of the dynamometer cannot
accurately be determined by using the vacuum pressure method.
In recognition of the fact that the legal procedures described above
suffer from a variety of defects, several investigators began to
develop an alternative method for calibrating chassis dynamometers
aiming at an improved procedure. Some members of the CEC-CF 22 project
group carried out an extensive measuring program in order to back up
the procedure with more data on various types of vehicles.
1. Description of Method
1.1 General
Starting from the deliberation that a procedure should be developed
which allows the measurement of the forces (driving resistances)
experienced by the vehicle on the road as directly as possible the
CEC-CF 22 group agreed to study torque methods. This decision was
supported by promising test results already available at that time
through some group members. To enable an accurate adjustment of a
dynamometer to road load conditions it is necessary to measure the
load behaviour of the vehicle on the road. This is done by recording
the road torque measured at the drive shafts or in the wheels of the
test car at constant speeds in the range from 20 km/h to 100 km/h in
increments of 10 km/h (s. Fig. 3). If then - for calibration purposes
the car is placed on the dynamometer and the same torque curve over
the above speed range is reproduced as determined on the road, it is
ensured that the power absorption unit is adjusted to actual road
load. This completes the constant speed (stationary) portion of the
adjustment procedure.
-------
- 6 -
200-
NM-
<1
IG0 -
1MB
u
120-
D
~
100-
~:
a
»-
00
E0-
H0 ¦
20-
TYPICRL TDRQUE CURVE
MEASURED ON RDRD
10 20 30
H0 S0 E0 70 B0 KH/H 100
CAR 5PEED
Fig. 3: Torque Measured as a Function of Stepwise Constant Speed
of a Test Vehicle.
The data were reduced to standard ambient conditions and "zero
qrade" of the road. The error bars mean the standard deviations. The
curve is a second order polynominal of the form M = a0 + a2 v2 and
commuted by means of the least squares method giving the numerical
values of a0 and a2.
M = torque, v = vehicle velocity.
-r- can be interpreted as the rolling resistance under the
assumption that the coefficient of rolling resistance is inde-
pendent of velocity which is justified as an approximation in
the velocity ranqe studied here.
Then -p- = f^mg, with r = rolling radius of tire, f^ = coefficient
of rolling resistance, m = vehicle mass, g = acceleration of
gravity.
Under the above assumption for fR, —
—2. y? =
r 2
c, A v2
3
with $
v2 is the air resistance
air density at standard conditions,
c = coefficient of air resistance, A = projected vehicle area.
-------
- 7 -
In practice - due to the lack of flexibility in their characteristics -
for water brake and older eddy current dynamometer types it is not
possible to adjust the dynamometer load curves over the whole speed
range (20 km/h to 100 km/h) to actual road load. In the CF 22 group
it was therefore agreed to recommend - as the best practicable approxi-
mation to the complete road load curve - a compromise for those dynamometers
consisting of adjusting the load curve in two points with defined
tolerances. In this way, the current method of using only one adjustment
point is substantially improved. A perfect reproduction of actual road
load conditions can be achieved with DC-machine and modern eddy current
type dynamometers. Here, the CF 22 group recommends the adjustment of
the whole load curve. In addition, with these types of dynamometers all
influences of the dynamometer geometry such as roll diameters, spacing
of rolls and also the bearing friction can be compensated.
Going back to the basic philosophy that the dynamometer should reflect
"what the vehicle sees on the road" (EPA) it is also necessary to look
at the dynamic behaviour of the vehicle during the driving cycles. This
dynamic behaviour can also be characterized very exactly by using the
torque method. For this purpose the CF 22 group recommends the use of
the integrated (mean) torque determined by the integral
measured over a variable driving pattern, preferably the
first 505 s of the US driving cycle (LA-4) excluding the idle period
M(t) dt (with M(t) > 0)
t
1
at the beginning (tj = 20 s; t2 = 505 s). The first phase of the
European test can of course also be used.
In order to adjust the dynamic behaviour of the dynamometer to actual
road conditions this mean torque, value M first has to be determined
on the roaa.
-------
- 8 -
Leaving the already adjusted load curve unchanged this "dynamic"
M-value is reproduced on the dynamometer by varying its flywheel
masses, the precision of adjustment being only dependent on the
size of the flywheel increments available at the dynamometer.
1.2 Physical Background
In the following a short outline of the physical background of the
torque calibration procedure is given.
We start from the basic equation of motion. To move a vehicle the
tractive force F or torque M acting on the driving wheels is needed
to overcome the various driving resistances.
F=f=Rr + Rg + Ra + R1 ^
Nomenclature
F = Tractive force on driving wheels
M = Torque on driving wheels
Rr = Rolling resistance
Rg = Grade resistance
R, = Air resistance
a
R. = Inertia resistance
r = Distance of drive axle to road surface
Rr = fRmg
fR = Coefficient of rolling friction (in general: fR = fR (v))
m = Vehicle mass, g = Acceleration of gravity
v = Vehicle velocity
Rg = mgsin et
sin«t£ tan oC = grade [%] (valid for small grades)
Ra * I ca A "2
5 = Air density at standard ambient conditions
ca = Coefficient of air resistance
A = Projected vehicle area
= mv+Amv
R. = in ( 1 + X ) v = m* v
dv
v " ar
A - vehK'h* piUMiiipftM- which accounts for rotating masses.
-------
- 9 -
With these expressions and for the case of "zero grade", I.e.
Rg = 0 we rewrite equation (1) as a function of time:
M(t)road = r (fR m9 + 7 ca A y2 W + m* * (*))
This equation governs the adjustment of the dynamometer load
behaviour to that of actual road driving.
The equation of motion for the vehicle on the dynamometer reads:
"(t'dyro * "WPMI + S"'"
with
M(t)dyno = Tor(lue on the driving wheels of the vehicle on the dyno
M(t)pAu = Resistance Torque provided by the power absorption unit
of the dyno.
and
ew (t) =erollu (t) +8s1m k£>(t)
¦ '"roll + k
with
Cj = angular acceleration
0H = Inertia of the dynamometer roll
0 . = Inertia simulated either by discrete flywheels or electrically.
sim
k = Factor accounting for a possible transmission between roll
and flywheel axle.
In most cases k = 1 holds true. Therefore, in the following k = 1
is assumed.
.'i du v
with
R = Radius of dynamometer roll
v = Circumferential velocity at the surface of the roll and at the
rolling radius of the vehicle tires (neglecting slip).
-------
- 10 -
Our recommended procedure is performed in two steps:
(1) Constant speed (stationary) procedure: v = 0, w = 0, i.e. v = const.,
u = constant, and M = constant (acceleration effects are not taken
into account).
M is measured on the road for constant speeds in the range of
20 km/h to 100 km/h, in increments of a v = 10 km/h, resulting
in constant M-values: MgQ, M30, ...M10g.
The dynamometer torque MpAU is then adjusted as good as possible
to these torque values so that Mp^y = Mroacj» thus simulating
rolling and air resistances.
The geometrical dimensions of the dynamometer such as the roll
diameter etc. no longer have an influence. The same holds true
for the friction losses of the dyno. Therefore, coast down runs
for determining these losses are no longer necessary.
(2) Integrated torque (dynamic) procedure: v(t)*0, u (t) * 0, i.e.
acceleration is taken into account (driving cycle).
The torque difference am between the vehicle on the road and on
the dyno reads:
The term r(Rr+Ra) is equal to Mp^j because this part of adjustment
has already been accomplished in step (1).
Then: "road = r(Rr + Ra' and "dyno = MPAU"
*M(t) = M(t)road - M(t)dyn0
Thus,
4M(t) = n«* v(t) - (9roll+6sin])
* (rm. . 9"" * "si") v(t)
R
The only remaining term still to be adjusted, i.e. simulated by
the dyno is therefore rm*.
-------
- 11 -
From this follows the condition for perfect adjustment of the
dynamometer load to road load.
I.e. AM-—0 or M(t)dymp H(t)road:
0
rol 1
+ 0
— = rm* (by adjusting 0sim).
R
This at the same time also simulates the effect of the non-driven
vehicle wheels. In order to simplify the measuring and evaluation
procedures we use the same relation but integrate both sides of the
equation and multiply by-r—as is done in our recommended
t2 - t,
procedure:
with the condition for adjustment that ^oad®* Mdyno*
To sum up the dynamic procedure we note that
(i) the parameter for dynamically adjusting the dynamometer load
behaviour to road conditions is 0s^m , I.e. the inertia of the
flywheel mass of the dynamometer system.
By varying for instance the flywheel mass of the dynamometer so
0 ,, + 0 i
that r0 ft —111 = rm* the difference 4 M between road and
dynamometer vanishes and the dyno is completely matched.
(ii) the driving cycle and the length of the time period, i.e. the
integration limits (tj, t2) can be chosen arbitrarily.
For reasons of a considerable measuring effect it is advisable,
however, to take a driving cycle with rather high accelerations
such as the US cycle or parts of it. Both steps - the stationary
and dynamic one - for calibrating dynamometers must be seen as
a whol<>. taken tooether thev constitute the complete calibration
1
1
1
or
-------
- 12 -
2. Types of Chassis Dynamometers
2.1 General
Motor vehicle emission testing is mainly performed on three types
of dynamometers: Waterbrake dyno, eddy current brake dyno and the
DC-machine dyno. Basically, these three types differ from one
another in their transformation of the mechanical energy of
the vehicle being tested into another form of energy (see table 1).
Dynamometer Conversion of mechanical Simulation of driving simulation of
types energy into: resistance torque inertia resistance
-------
- 13 -
2.2 Characteristic Properties
In the following we discuss the essential properties of the different
types of dynamometers.
(1) Water Brake Dynamometer:
The parameter A (s. table 1) can be varied by the amount of water
within the brake, fi-which should be equal to 2 in order to
simulate the air resistance (s. section 1.2) - is given mainly
by the geometry of the brake. However, ft is not exactly a
constant due to hydraulic losses. To some extent it is possible
to correct the deviation of (? from the required value of 2 by
a control circuit.
(ii) Eddy Current Brake Dynamometer:
Here, the torque is approximately proportional to v . Again M~ v2 is
required in order to simulate the air resistance. This is accomplished
by controlling the current 1 through the exciter coil in such a way
that i^v , resulting 1n M^v with 2 £ f < 3, thus approximating air
resistance. This approximation-which 1s quite common for most
dynamometers currently in use - is not sufficient because of the
fact that the parameter 1s not equal to 2. New designs of this
dynamometer type use electronic control circuits for the exciter current
and enable a more realistic reproduction of the road resistances.
(iii) Direct Current Brake Dynamometer:
With this type of dynamometer the rolling and air resistances aQ and
a2 v2, respectively, can be reproduced nearly perfectly on the dyna-
mometer. In addition, the friction losses (a^v) can be compensated
in some cases. Positive and negative grades can be simulated, too.
Moreover, the inertia resistance can also be simulated electrically
allowing a continuous variation of inertia.
In general, the usefulness of a dynamometer must be measured by the
degree of its capability to reproduce the road driving resistances of a
given vehicle, i.e. what the vehicle sees on the road.
The different types of dynamometers (s. table 1) can be compared by using
the general equation of motion which applies to the system "vehicle on
the roaa" on the one side and "vehicle on the dyno" on the other.
-------
- 14 -
With the nomenclature of the foregoing section we write the equations
of motion:
Mroad = r^Rr + V + rm* * (assuming = 0)
m u ®roll + ®sim .
dyno = PAU + " v
Condition for adjustment: Mroac| = Mdyno
The more important difference between the types of dynamometer refers to
Mp^j, i.e. the power absorption unit.
Most of the water brake dynos currently in use have a fixed shape of the
characteristic curve expressed by torque as a function of car speed. By
chanqing the absorbed power (changing A) in a given point of dyno speed
the whole curve is thereby already established. Fig. 4 shows examples for
different values of A (here: A = a2> see also table 1).
Fig. 4: Torque vs Car Speed of a Water Brake Dynamometer.
An (idealized) functional dependence of M = a0 + a2 v2 was assumed. Only
the parameter a2 can be varied. The three curves represent a2-values of
0.015, 0.008 and 0.005 (from top to bottom); ac = 45 was assumed for each
curve. The disadvantage of the water brake principle is obvious, because
it is not possible to vary the parameter aQ (rolling resistance + internal
friction). Therefore, an acceptable adjustment to road load is not possible
i n man\/ racpc .
-------
- 15 -
In contrast to this, the DC machine dyno and the modern eddy current dyno are
much more flexible. Here, the road torque can be exactly reproduced on
the dynos by adjusting the simulation for the coefficients of rolling
resistance and air resistance resulting in Mp^ = r (R^ + Rfl).
This is especially true for the DC machine type dynamometer. Even when
compared to the eddy current brake dynamometer the DC-dynamometer has
some considerable advantages.
(i) Possibility of continuous inertia simulation.
(ii) Possibility of operating as a motor.
This motor operation is a rather important property for daily use,
because e.g. the warming up of the dynamometer does not require an
additional vehicle or other equipment for driving the dyno.
(iii) In general, the DC dyno has a higher dynamic response to varying
loads.
(iv) The mechanical energy of the vehicle on the dynamometer is converted
into useful electrical energy, which is returned to the main system
during operation of the dyno as a power absorber (generator).
(v) Furthermore, in contrast to the two other types of dynos no
coolinq water is required.
3. Comparison of Torque Method to Other Methods.
The purpose of this section is to describe the various calibration
procedures for chassis dynamometers currently in use and to point
out their advantages and disadvantages compared to the torque method.
3.1 Review of Existing Methods
3.1.1 Inertia Weight Classes
First, we have the well known table values of inertia weight classes for
the dynamometer load at one given speed point. These values are based on
the corresponding loaded vehicle weights. Such tables are used in both
Europe and the US. Necessarily, these table values are only very rough
figures due to the fact that only one vehicle parameter, namely its weight
is considered.
Such an important quantity like the particular shape of a vehicle type
- i.e. its particular air resistance - is not taken into account. This
can easily be seen from the basic equation of motion (with Rg = 0):
-------
- 16 -
F = Rr + F*i + Rg = fRmg + m* v + |- cg A vz
Only two of the three resistance forces - namely the rolling resistance
Rr and the inertia resistance - are proportional to the vehicle mass.
This is not the case for the air resistance R_. Here, the product of
a
the air drag coefficient cfl and the frontal area A is the decisive quan-
tity. Therefore, the classification according to the inertia weight
classes alone is obviously not sufficient.
The calibration of the dynamometer load at the given one point of speed
is carried out by a coast down procedure. From the coast down - which
has to be performed for each individual inertia weight because of varying
amounts of friction losses - the internal friction of the dynamometer can
be calculated.
This internal friction is due to the friction in the bearings of the clutches,
rollers, and flywheels and due to the aerodynamic drag associated with the
spinning rotating parts. The total amount of power - which is required for
the test vehicle at the roller surface to drive the dynamometer - is given
by the sum of the so-called indicated load and the internal friction. The
indicated load means the output of the power meter of the dynamometer.
Here we have the problem that the inertia of the idle roll is not con-
sidered, because the vehicle is lifted from the rolls.
3.1.2 Intake Manifold Pressure
The still remaining problem which is basically inherent in all vehicle
testing on dynamometers, is the question of correlating the vehicle
operation on the dyno with that on the road. The table values are only
a rough approximation for this.
As a first alternative to the table values, the manufacturer was allowed
to take the intake manifold vacuum pressure value of a representative
vehicle driven at constant, speed on the road as a basis for adjusting
the dynamometer at that same speed.
Meanwhile, for 1979 and later model year light-duty vehicles the intake
manifold vacuum procedure was removed from the US regulations as a
requirement. Therefore, this adjustment method is no longer automatically
accepted by EPA but only if a manufacturer requests its use. In Europe,
the manifold vacuum procedure is - as well - no longer considered as the
only adjustment procedure and emphasis was changed to use it as an alter-
native method among others.
-------
- 17 -
The problems arising with this procedure have been already discussed
in section 1. Generally, this procedure is very unsatisfactory and is
therefore no longer considered by the group members as a useful method.
3.1.3 Aerodynamical Assumptions
In the US, a further adjustment procedure for 1979 and later model year
light-duty vehicles is based on calculations of the dynamometer power
absorber setting from vehicle parameters such as frontal area, body shape
and tire type. The assumption about the body shape of the vehicle-including
protruding parts such as mirrors etc. - is only a more or less accurate
approximation to reality. This is due to the fact that the coefficient of
air draq c, is not explicitely considered in the calculations. In this
a
procedure, again the dynamometer power absorber is adjusted in only one
point of speed.
3.1.4 Coast Down
A further attempt to solve the essential problem of reproducing the road
behaviour of the vehicle on a dynamometer is the application of the well
known coast down technique to measurements on the road. This procedure
is valid in the US for 1979 model year light-duty vehicles as an
alternative. The measuring quantity of this procedure is the time versus
speed behaviour of the vehicle when freely decelerating between two
given velocities. The main task which must be accomplished is to extract
from these speed vs time data the deceleration vs velocity function.
From this function the forces actina on the vehicle during deceleration
can be calculated. First, we have to assume a model of the deceleration
(- v) vs speed (v) equation.
-v = a0 + g sin* + a2 (v-w)2
with ae = rollinq resistance term
gsinrf. = grade resistance term
a2 (v-w)z = aerodynamic term, w = wind velocity.
Whenever possible the wind velocity should be constant and the direction
should be the same as the vehicle translation. The most reliable results
are obtained with w=0.
The above equation must be integrated to obtain a function for the
speed vs time behaviour. The result of this integration is of the form:
The measured coast down data v = v(t) have to be fitted to this non
linear equation by a least squares^technique .
; functions of aQ, a2
-------
- 18 -
As a result the coefficients a0 and a2 can be obtained. The total road
force {F) acting on the vehicle can be written as
-F = m* v or -F = m* (a0+a2 v2) = f0 + f2 v2 with m* = total effective
vehicle mass and assuming «C= 0, w=0.
The values f0 and f2 must still be corrected to standard ambient
conditions by the well known equations.
From the corrected fQ and f2 values finally the time interval aT for the
vehicle coasting down from 55 mph to 45 mph can be calculated. This
a T-extracted from the road data as described-is then reproduced on the
dynamometer. So, finally thisaT is the decisive quantity for adjusting
the dyno to road conditions.
3.2 Torque vs Coast Down Method
In the following only the coast down method is compared to the torque
method. The other procedures discussed in the preceding sections are
no more considered here, because in the opinion of the group it is
evident that these procedures cannot compete with the torque- or the
coast down method.
There are several reasons for preferring the torque method rather than
the coast down method.
(i) As was seen in section 1 it is very important to adjust the whole
load curve of the dynamometer or at least two representative points to
road conditions. This is not accomplished - however possible - with
the coast down procedure as recommended by EPA, where again only
one point of velocity is adjusted to road conditions.
(ii) The internal friction losses of the dynamometer are also automa-
tically taken into actount by the torque method. Therefore, coast
down checks for determining the friction losses are no longer
necessary.
{iii) If the driving resistance on the road is measured in the vehicle
drive wheels by the torque method, possible influences on the
measured value - e.g. by sticking brakes, variation of the friction
losses in the drive train (oil temperature variation) - are automa-
tically eliminated. This is not the case with a coast down method.
-------
- 19 -
(iv) A fundamental drawback of the coast down technique is the fact
that it is a very indirect method, i.e. the measured speed vs time data
cannot be taken rightaway to calibrate the dyno. In contrast, some non-
elementary mathematical operations like fitting data to a non-linear
function are necessary. In some cases this might eventually cause
convergency problems in determining the required coefficients a0 and a2,
perhaps, especially if the areodynamic drag coefficient (i.e. a2)
is very small.
These problems are not present with the torque method because the needed
load curve is measured directly. The rolling resistance and aerodynamic
drag can be extracted directly from the measured data by performing a simple
linear least squares fit using a polynomial of second degree (see Fig. 3)
This can be done with the aid of simple calculations. Moreover, the
torque method allows the matching of the dynamometer load curve and
the dynamic behaviour to road conditions, operating the dyno in the
same way as in emissions- or fuel consumption testing.
Finally, the time required to prepare the vehicle with the torque equip-
ment is now considerably reduced due to the fact that the torque wheels
can be easily removed from one vehicle and attached to another using special
adapters. Such adapters can be produced for each type of vehicle.
4. Apparatus
4.1 Torque Meters
The CF 22 group recommends a torque measuring principle which is based
on determining the overall tractive torque. So far, there are experiences
with two different concepts of applying the strain gauges technique.
4.1.1 Torque in the Drive Shafts
This concept makes use of the torsion in the drive shafts when the
vehicle engine is working against the driving resistances. Thus, the
(sum of the) torque in the drive shaft(s) is the measuring quantity. A
certain disadvantage is the fact that the drive shafts have to be
removed and reinstalled for the toroue measurements and at present also
have to be specially machined.
-------
- ?0 -
4.1.2 Torque in the Drive Wheels
In this case the torque is generally measured between the vehicle brake
and the tire. The torque device may consist of two discs which are
welded together at specified locations. One of these discs is provided
with spokes to which strain gauges are applied. The measuring effect
is the flexure of the spokes representing the forces experienced by the
car wheels when running on the road or on a dyno. In order to give a
brief impression of the technique used, a schematic diagram of such a
device is displayed in Fig. 5.
tire
measuring disc
(welded together)
Fig. 5: Schematic Diagram of One Possible Way for Realizing a
Toroue Measuring Device. The torque is measured in the wheel
between tire and brake. Such a torque device is mounted to each
driven wheel.
-------
- 21 -
This concept allows an easy Installation of the measuring device be-
cause it is mounted to the wheels.
An application of this device to the various types of vehicles can be
done be using adapters.
4.2 Electronic Equipment
The signals coming from the torque meter are amplified and processed
into readable quantities. In the case of measuring the load curve
- i.e. torque as a function of stepwise constant car speed - the ampli-
fied signals can be recorded for example on magnetic tape or strip
chart recorder.
In the case of measuring the dynamic behaviour expressed as "M the
torque vs time function has to be integrated over the chosen driving
cycle. One possible method is to store the analog signals on magnetic
tape and integrate the torque function by a computer after the measure
ments.
A more convenient way of obtaining the desired M-value is to use an
electronic device for integrating the torque function. This can be
done by a voltage to frequency converter which converts the torque
signals (voltages) into pulses, the number of which per time unit is
proportional to the height of the torque signals. These pulses are
summed up by a pulse counting device. The final number of pulses is
then divided by the time length of the driving cycle. The result is "R.
This technique has the big advantage that the numerical value of TT is
available immediately after the measurement. A block diagram of the
equipment is shown in Fig. 6.
-------
- 22 -
Voltage to
Strain DC- frequency
gauges amplifier Rectifier converter Counter
Fig. 6: Block Diagram of Typical Electronic Equipment for
Torque Measurements.
For constant speed torque measurements the torque signal coming
from the strain gauges bridge is amplified and directed to a re-
cording device.
For integrated torque measurements the amplified and rectified
torque signal is converted into pulses by the voltage to frequency
converter.
Then the pulses are counted. The total number of pulses devided
by the time interval (e.g. 485 s) represent the integrated torque W.
5. Test Procedure
As described in Section 1 the recommended test procedure consists of
measurements on the road and on the dyno. The driving resistances for
the vehicle on the road should be reproduced on the dyno.
-------
- 23 -
5.1 Measurements on the Road
The road should he straight and the grade - if any - must be known
precisely so that the data can be corrected to zero grade. The grade
must be constant within + .05 % throughout the test section. The road
surface shall be hard, smooth and dry.
During the measurements no interference from other vehicles should
occur.
If any wind is present, the wind velocity should not exceed 2 m/s and it
should be constant within + 0.5 m/s. A reliable correction of the data can
only be performed for steady winds in the direction of the test track.
Nevertheless, the most reliable data are obtained without any wind being
present. The tire pressure should be as recoimiended by the manufacturer.
5.1.1 Constant Speed Torque
A test vehicle which is representative for the vehicle type under study
is driven on the road at constant speeds. Beginning at 20 km/h the
speed is increased in constant intervals of 10 kn/h covering a range
of 20 km/h to 100 km/h. At each speed interval the torque must be kept
constant. The torque pattern at the drive shaft(s) or in the drive wheel(s)
is recorded.
This procedure should be repeated three times alternately in each
direction of the road resulting in 6 sets of data points. Each set has
to be corrected separately for possible grade and wind. The corrected data
are combined and mean values as well as standard deviations of torque
and speed are calculated for each data point. The result is a mean
road load curve of the vehicle under study. This curve should be finally
corrected to standard ambient conditions.
5.1.2 Integrated Torque over Variable Driving Pattern
In this dynamic procedure the mean torque value TT is determined. This is
accomplished by integrating the actual torque values with respect to
time during operation of the test vehicle with a defined driving cycle.
For this purpose the CF 22 group recommends e.g. the first 505 s of the
US 75 cycle excluding the 20 s idle period at the beginning. The total
time interval is therefore 20 s ... 505 s = 485 s. The integrated
torque - obtained by one of the devices described in Section 4 - is
finally divided by 485 s. The result is
505
20
-------
- 24 -
5.2 Adjusting the Dynamometer
Here, It should he pointed out once more that the recommended torque
procedure is applicable to all known types of chassis dynamometers.
This method can substantially Improve the adjustment of a dyno.
5.2.1 Constant Speed Adjustment
The dynamometer is warmed up according to the manufacturers' re-
commendation. The tire pressure of the test car is increased to a
value of 3 bar at ambient test conditions and the car is placed
on the dynamometer. Now we have to distinguish between the two
classes of dynamometers - the one which allows a perfect repro-
duction of the road load curve and the one which does not.
5.2.1.1 Dynos Adjustable to Road Load
Generally, with this class of dynos we can match road load curves.
The (corrected) averaged torque curve measured on the road is repro-
duced on the dynamometer by varying its adjustment parameters (s. Fig. 7).
2B0 ¦
NM"
| !E0 ¦¦
I 1MB-
U J20"
EJ
£ 100 •
a
B0 ¦
SE-
ME-.
20
TORQUE CURVE5 MER5URED
DN RORD RND ON DYNO
-WITH ONE VEHICLE-
DYND
IB 20 30
40 SB EB 7B B0 KM/H IBB
CRR SPEED -
Fig. 7: Adjusting the Dynamometer Torque Curve to the Road Torque Curve.
One and the same test vehicle was used in both cases. The adjustment,
is carried out by varying the rolling resistance + internal friction term
(a.,) and the air resistance term (a2) of the dynamometer PAU control circuit.
Tor PC machine brakes and modern eddy current brakes a perfect adjustment
Oiin ho achieved.
-------
- 25 -
5.2.1.2 Dynos with Limited Adjustment
Generally, with this class of dynamometers we cannot achieve an
adjustment to road load, because only the air resistance term can be
varied. Therefore, the CF 22 group recommends a compromise for this
case. The compromise consists of adjusting the dynamometer in two
points of velocity - at 40 km/h and at 80 km/h. The remaining deviations
between road torque and dyno torque in these two points should not exceed
15 %, i.e. A M ^ 15 % (s. Fig. 8).
Fig. 8: Compromise for Adjusting the Dynamometer Torque Curve to the
Road Torque Curve.
The dyno curve is adjusted in two points of velocity (40 km/h and 80 km/h).
For these points a M should not exceed 15 %. In this case the TT-values
of the two curves (dyno after adjustm. and road) were then in perfect
agreement.
5.2.2 Integrated Torque Adjustment
After the stationary adjustment of the dynamometer the dynamic procedure
is carried out. This means the adjustment to Mrofl(j.
Following the derivations of Section 1
0 . . + 0
52?! = rm* , the adjustment parameter of the dyno
R
-------
- 26 -
Using this equation, 0s^m can be calculated from the effective mass of the
test vehicle and from the dynamometer parameters (9ron» R)- principle,
it Is possible to achieve any precision with respect to Q = ^roaC|
provided that the increments of flywheel masses available at the dynamometer
are small enough and provided that a matching of the dyno load curve to road
load curve was carried out. As discussed in section 2 this is generally only
possible with DC and eddy current dynamometers.
5.2.2.1 Dynos Adjustable to Road Load
In the case of such dynamometers the CF 22 group recommends the following
steps to achieve fldyn0 = *rM<|:
(1) Constant speed adjustment of dyno load curve to road load curve
(s. 5.2.1.1).
(2) Calculation of vehicle equivalent inertia by using
6 + 0 .
roll sim +
rm*
R
(3) Normally, the calculated inertia implies "Wjyn0 = ^roacj " within
measurement precision (»1 %). This, however, should be checked
by the torque method. If the measurement yields a slight deviation
- possibly due to dynamometer tolerances - from ^yno * ^road' a
fine adjustment of the dynamometer inertia should be carried out.
In practice it should be checked whether a better agreement could
be achieved by varying the flywheel masses available. In most
cases a variation of _+ 125 lbs = + 56.7 kg will be sufficient, so
that the flywheel increments available should be 125 lbs. A fault
in the dynamometer performance - e.g. a flywheel disengaged or a
faulty bearing of the non-coupled flywheels - generally causes a
considerable discrepancy between ^dyn0 and ^roa(j- Such a case is
immediately detected by the torque method and that under realistic
conditions, which means that the dynamometer is loaded and operated
(driving cycle) in the same way as during normal exhaust emissions -
or fuel consumption testing, i.e. under equal dynamic conditions.
5.2.2.2 Dynos with Limited Adjustment
In the case of a dynamometer with limited adjustment (e.g. water brake)
the CF 22 group recommends the following procedure:
(1) Constant speed adjustment within 15 % of the dynamometer load
curve to road load curve in two points of velocity (s. 5.2.1.2).
-------
- 27 -
(2) Calculation of vehicle equivalent inertia 0 .m by using
0 ,, + 0 .
roll sun *
rnr.
R
(3) Now, the calculated inertia generally does not imply =
^road ~ ^ue *° more or ^ess 900(' a9reement between the dyno
load curve and the road load curve. In practice, the deviation
between and Mroac| is not very high, provided that the dyno
curve has been adjusted in the two points. In principle, there are
two possibilities for achieving ^jyno = Mroad:
(i) If, for example, J^y^ ^roac|» dyno curve in Fig* 8
could be lowered until fljyn0 = ^roa(j» However, this would
normally result in Inadequate fuel consumption test results
because the load curve setting of the dyno has a considerable
effect on fuel consumption. This is due to the fact that
e.g. the "Highway Driving Cycle" consists mainly of more or
less constant speed driving.
(ii) Therefore, the CF 22 group recommends a final adjustment of
^dyno Mroad ^ varying the inertia of the dyno. Again,
flywheel increments of 125 lbs = 56,7 kg should be sufficient.
6. Evaluation and Reporting of Results
6.1 Constant Speed Torque
According to the stationary procedure the main task in the data analy-
sis of the road measurements is the extraction of the rolling resistan-
ce and the air resistance of the vehicle under test. The three data
sets in each direction of the test track have to be corrected for pos-
sible grade and wind. This must be done prior to combining the 6 data
sets to one mean road torque curve.
(i) If the test track has a non-zero grade the term rmg-sin«C
(s. section 1) has to be added/subtracted to/from each data point
(torque value). If the 6 different curves are combined to a mean
curve prior to correcting each curve separately this would also re-
sult in the same curve but the standard deviation for each data point
would be considerably higher.
-------
- 28 -
(ii) Only steady winds of less than 2 m/s in the direction of the test track
are allowed during data collection (s. section 5). The wind effect
cannot be eliminated by simply combining the data sets for the different
directions. This can easily be seen.
The resulting velocities which are decisive for the air resistances
in both directions are:
v1res = (v + w)2 and v2res = (v ~ w^2» where w is the wind velocity.
Thus, the mean velocity for both directions becomes
7 (v1res2 + v2reP = v* + w*' i,e* not equal t0 v*'
So, each car velocity must be corrected separately for the wind
influence.
The still remaining problem is the constancy of the wind conditions,
i.e. its velocity and direction. Therefore, it is best to perform
the measurements without any wind being present if that is possible.
After having corrected the data for wind and grade a least
squares fit of the form M = a0 + a2v2 is carried out. The
coefficients aOJ a2 and their uncertainties are obtained through
this fit. a0/r represents the rolling resistance and a2v2/r the
air resistance. Here, it is assumed that in the given velocity
range (20 km/h to 100 km/h) the rolling resistance is independent
of car velocity (s. section 1).
Finally, the rolling and air resistances must be still corrected
to standard ambient conditions. Particularly, this is important for
the air resistance R = % ca A v2 because of the temperature
dud
and pressure dependence of the air density § . As standard am-
bient conditions a temperature of T0 = 293.2 K = 20 °C and a barometric
pressure of pQ = 101.32 kPa = 760 Torr is usually presumed. In order to
reduce "arbitrary" ambient conditions - given by T, p - we have to
apply the well known formula:
't-.P.-'t.P ^ ^
I o
$ T = air density at T0 = 293.2 K, p0 = 101.32 kPa
I o > P o
S t p = air density at T , p
-------
- 29 -
For example, let us assume ambient conditions of T = 273.2 K and
p = 103.99 kPa (= 780 Torr) during the measurements on the road.
In this case the correction for air resistance amounts to approxi-
mately 10 %. This example was chosen in such a way that both cor-
rections - for temperature and pressure - go into the same direction.
For the rolling resistance only a temperature correction is necessary.
The correction is carried out by applying the formula for reducing
the rolling resistance j (at ambient temperature T) to the rolling
resistance R0 T at Tc = 293.2 K:
K , I o
Rr To = RR,T ( 1 + °-0018 (T " 293-2))» i-e-
0.18 % per a T = 1 K.
This correction formula is also recommended by EPA (AC 55 A).
For our example of T = 273.2 K we obtain therefore a correction of
approximately 4 % with regard to the rolling resistance R^ y , drawn
from the measured data; i.e. the measured Rp j is 4 % higher than
RR,To-
Having done all the above corrections we obtain the rolling and air re-
sistances and their uncertainties reduced to zero wind, zero grade
and standard ambient conditions.
With these values a reduced average torque curve - which is the mean
of 6 single test runs - can be plotted. This curve is the reference
curve for adjusting a chassis dynamometer to road conditions.
6.2 Integrated Torque over Variable Driving Pattern
The various corrections discussed under 6.1 are much less important for
the dynamic measurements. This can easily be seen from the equation of
motion (s. section 1):
When driving the 505 s cycle R^ dominates over RR and Rfl.
To get an impresssion of the order of magnitude let us assume for example a
typical car with the following technical data:
m = 1 000 kg
m*= 1 040 kg
r = 0.3 m
A = 2 m2
c = 0.4
d
V 0.015
-------
- 30 -
Furthermore, we need the average speed and average acceleration during the
505 s cycle.
The values are: V = 49.9 km/h, J = 0.62 m/s2
(excluding the standstills).
With these data we can now calculate an approximate average value for the
force (or torque) over the 505 s cycle.
As an example let us determine the influence of the ambient conditions - in
a reasonable range. Again, we assume T = 273.2 K and p = 103.99 kPa.
F, „ = 9810 • 0.015 +-^-4^ 0.8 • 13.862 + 1040 • 0.62 (N)
T0,Po 2
= 884 (N)
Ft = 152.7 +14^ 0-8 * 13.862 + 645 (N)
T,p 2
= 899 (N)
The difference between these two values of F for the assumed ambient con-
ditions amounts to A = 1.7 %. The effect is so small that it usually
can be neglected, but if desired the correction can be carried out.
From 5 mean torque values "M an average value and its standard deviation
is computed. This average mean torque value is then used for the dynamic
calibration of the chassis dynamometer.
7. Precision of Torque Method
7.1 Constant Speed Torque
The precision of determining the coefficients a0 and a2 of the rolling re-
sistance and air resistance from the least squares fit applied to the cor-
rected measurement data is better than + 3 % in the case of rolling re-
sistance and better than + 2 % in the case of air resistance.
This situation is displayed in fig. 9 where typical measurement data are
plotted in the usual way. In addition to the "least squares fit" curve
two further curves
M+ = (a0 + a a0) + (a2 * &a2) v2 and
M = (a0 - A a0) + (a2 - a a2) v2 are plotted, where a a0 and a a2 re-
present the uncertainties of the coefficients. In this case the uncertain-
ties amount to & a0 = 2.2 % and a a2 = 1.5 %.
-------
- 31 -
Fiq. 9: Typical Torque Measurement Data.
The curve (M = a0 + a2v2) in the middle of the three curves is drawn
from a least squares fit applied to the corrected data. The two other
curves are:
M+ = (a0 + ^ a0) + (a2 + a a2) v2 and M_ - (a0 - a a0) + (a2 - a a2) v2
Here: a a0 = + 2.2 %, a a2 = _+ 1.5 %, which are resulting from the
least squares fit.
Figure 10 demonstrates the precision which can be achieved when adjusting
an electric brake dynamometer (upper curve) to road load. The two lower
curves are the road load curve and the dyno curve after adjustment.
A difference between the two curves is hardly to be seen (Difference
in a0: 1 % , difference 1n a2: 3.7 %)
Preliminary results for one car in the case of exhaust emissions and for
four different types of cars in the case of fuel consumption Indicate
that the differences in mass emissions and fuel consumption measured
1n the city driving cycle when using the two different adjustments
(upper curve and lower curves) are considerable (a fuel consumption
«8 %, A NOx 18 %). For other types of cars the situation might be
different. Here, still further results are needed.
-------
- 32 -
!
200-
NM
I EE) - -
1MB-
U |20-
n
tr 100 f
~
B0-
E0"
HB-
20
HDJU5TINE ELECTRIC BRRKE
DYNO TD RDRD LORD
IB 2B 30 MB SB GB 7B BB tCM/H IBB
CRR 5PEED -
Fig. 10: Precision of Adjusting Dyno Load Curve.
Upper curve: Before adjustment.
Lower curves: Road load curve and dyno curve after adjustment.
7.2 Integrated Torque over Variable Driving Pattern
The precision of the dynamic measurements can be expressed in form of
the standard deviation of "R-values found in repeated measurements. This
was done for several sets of 10 typical measurements. The average standard
deviation amounted to + 1 %.
-------
- 33 -
8. Summary and Conclusion
It 1s recommended by the CF 22 group that the calibration of chassis
dynamometers for fuel consumption and exhaust emission measurements
should be performed in such a way that the load behavior of the
vehicle on the road is reproduced on the dynamometer.
The best reproduction of actual road load conditions can be achieved
only with electric brake dynamometers. Therefore, if possible, the
use of this type of dynamometer is recommended. In addition, the dyno
should be equipped with discrete flywheels of 125 lbs increments.
In the case of water brake dynos a compromise which consists of
adjusting these dynos in two points of velocity with defined to-
lerances can be performed. In this way, the current method of using
only one point of adjustment is substantially improved.
The current ECE procedure of setting the dyno load according to iner-
tia weight classes has some fundamental disadvantages because it
does not take into account the particular air resistance of a special
vehicle type. The weight classes are only a rough approximation
to reality. Even the driving resistance (i.e. PAU loading) calcu-
lation method,based mainly on the frontal area of the vehicle (as
now used by the US-EPA) cannot really reflect the specific aero-
dynamic conditions of individual vehicle types.
Concerning the coast down procedure recommended by EPA it has to
be pointed out that this method does not yield the desired cali-
bration quantities - i.e. the coefficients a0 and a2 - in a direct
way. Certain non-elementary mathematical operations such as
fitting the measured speed vs time data to a non-linear function
must be carried out. In contrast to this, the torque method yields the
load behavior of the vehicle as a direct result of the measurements.
Finally, the CF 22 group recommends that the use of the torque
measuring principle is widely accepted by government as an inde-
pendent alternative for dynamometer calibration.
Furthermore, the group recommends the acceptance of this proposal
as a basis for a standardized procedure.
-------
ATTACH/HEMT HT
®
DAIMLER - BENZ
SUBMISSION NO. 3 TO EPA
"Detailed description of Daimler-Benz
torque measuring device (torque disc)
with instrumentation for road load
determination in the vehicle's driving
wheels and description of latest on
board measurement evaluation system."
V1MA
8-16-79
-------
Wheel Torque Measuring Device for Adjustment of
Chassis Dynamometer used in Exhaust Emission and
Fuel Economy Testing.
Summary:
Under the impact of steadily increasing demands in
the field of vehicle exhaust emissions and fuel economy
testing, continuous further refinement of existing or
development of new and sophisticated instrumentation
and measurement technology are necessary. In the chain
of measurement uncertainties, chassis dynamometers
have gained increased importance, since they have to
simulate road load conditions, i.e. rolling and air
resistance of vehicles during the above testing.
A special device was developed which allows direct
measurement of the force exerted from the driving
wheels of a vehicle to the road and reproduction of
this force - measured as torque - on a chassis dynamo-
meter.
After a test run on the road, the power absorption
unit of the chassis dynamometer can be so adjusted
to precisely simulate "what the vehicle sees on the
road".
1. Introduction
The steadily increasing demands in the field of
exhaust emission control and fuel economy testing
require the application of more and more qualified
measurement-, instrumentation- and evaluation
techniques in order to obtain reliable test results.
- 2 -
-------
- 2 -
Realizing that the torque in a vehicle's drive train
represents a suitable physical means for realistic
adjustments of chassis dynamometers for exhaust
emission- and fuel economy testing and taking into
account that the precision of torque measurements
in the drive shaft (s) may be adversely affected by
undefined friction losses in the differential or
wheel assembly, Daimler-Benz developed a torque disc
for application to the vehicle's drive wheels. FtJ*:OZ.
The following criteria were set up for the development
of this torque disc:
- contact-free measurement signal transmission
- possibility of electrical calibration
- integrated rpm pick-up
- resolution of low torque values
- high precision
- acceptance of high radial- and side forces
without influencing the torque signal
- compact design with minimal increase of tread
width
- applicability of standard rims
- applicability to non-Daimler-Benz vehicles
- easy installation to vehicle
Daimler-Benz's experience from earlier design work
and testing with similar devices in the field of
brake torques for different kinds of vehicle^ could
be advantageously used in the development work of
the new wheel torque disc.
- 3 -
-------
- 3 -
A prototype of the new device was displayed on a
demonstration test stand on the Hannover Fair in
1978.
2. Torque Measurement Device
2.1 Design and working principle
The torque measuring device mainly consists of the
torque disc and the signal transmitter. The torque
disc itself is formed by two portions, one connected
to the drive shaft, the other connected to the rim.
Both parts are welded together at the circumference
by an electronical beam. The disc portion connected F(g.:o3
to the drive shaft has 5 bores for mounting the device
to the brake disc and 32 measuring bores resulting in
32 webs acting like spokes of a wheel. Of these 32
>
webs, 16 are equipped on both sides with strain
gauges, which are located at the place of maximum
bending force, when forces in circumferential direction
are applied.
In axial direction, the cross section of these webs
is substantially larger than in circumferential
direction. The webs are therefore capable to stand
high radial and axial forces. On the other hand,
the webs are relatively flexible in circumferential
direction as intended.
All 32 strain gauges are electrically combined to
a bridge (R = 2U0 ), the connection of which leads f)j.:
to a rotation transmitter located in the center
of the disc.
- k -
-------
- k -
Axle load, side forces and temperature related
influences are largely eliminated by design-,
construction- and electrical means, so that only
torque related forces are measured and transmitted.
Besides the earlier mentioned length to vidth
ratio of the webs, the number of equally distri-
buted measuring webs should be large and divisible
by h in order to eliminate signal oscillations*
The dimensions and surfaces of the bores/webs
require defined conditions. Further, a precise
application of the strain gauges, optimal arrange-
ment of each individual strain gauge within the
bridge and - if needed - temperature compensation
elements have to be considered.
Due to the very compact design o,f the torque disc Fi
-------
- 5 -
An important part for the electrical calibration
of the measuring chain (sensor -amplifier- recorder,
etc.) is the calibration switch incorporated in the
inductive rotation transmitter. This device, coil 5
(stationary) and Reed contact (rotating) by exciting
of coil 5 (12 V=) switches a certain resistor RK
parallel to a quarter branch of the bridge and thus
electrically decalibrates the full bridge. This
calibration signal in relation to the output signal
stimulates a defined torque (A) when a certain torque
(b) is mechanically applied to the disc. The defined
torque (a), which may be 30 9o to 100 °jo of the nominal
torque (c) (in our case 50 ) , can be achieved by
choosing the value of resistor RK.
For wheel rpm pick-up, a light electric rpm counter,
delivering 60 impulses per wheel revolution, is
additionally incorporated.
2.2 Production
Most of the parts are manufactured out of stainless
steel with the disc itself consisting of high alloy
spring steel which is tempered to a strength of
2200 N/mm2.
Precise production with intermediate checks and
tests of the mechanical and electrical components
guarantees a high quality of the device.
- 6 -
-------
- 6 -
2.3 Calibration
After production,each device is subjected to
a final test and calibration. For this torque
calibration, a specially developed calibration
test bench with aerostatic bearings ranging to
1 kNm vas set up. This test stand is connected
to an automatic data recording system and a com-
and;
puter, which evaluates^immediately plots non-
linearity , sensi tivity and calibration value. All
test data are kept in a protocol and archive.
2.k
Technical Data
Nominal Torque (c)
Overload
Hysteresis
Sensi tivi ty
Calibration Value
Rpm Sensor
Carrier Frequency
Veight
Mt=
I 500 Nm
M, = 100 of (c)
** 0.01 mV/Nm (at 5 V feed
voltage)
50 % of (C)
60 imp/wheel rev.
5 kHz
8.4 kg
3. Measuring and Evaluation System
3.1 Measurement Set-Up
Since the torque discs are designed for application
in chassis dynamometer adjustments by reproduction
of actual road load forces, the complete measurement
set-up has to be mountable into the vehicle with
on board electrical power supply. fy" ^
- 7 -
-------
- 7 -
Since further a subjective evaluation of rpm/
speed or torque strip charts includes many uncer-
tainties, the evaluation should be done by the
system itself. This is possible by application
of a computer.
Fig. 07 shows a set-up vhich has proven high
reliability during many tests on track or dynamometer.
The wheel revolution impulses are transmitted via
a frequency/voltage converter with attached lov-pass ampli-
fier to the inlet of a multiprogrammer. The torque discs
in the vehicle's driving wheels are connected to
5 KHz carrier frequency atnplifiers,the outlet voltages
of which are summed up in an addition-subtraction amplif i
directed via a low-pass amplifier to another inlet
of the multiprogrammer as the sum of 2 torque
values. The multiprogrammer is governed by a
computer which takes care of all required evaluations
of the measured values. Finally, the results of these
evaluations are transmitted to a plotter: An immediate,
complete on board test evaluation has taken place.
3.2 Adjustment of measurement system
The adjustment starts before each measurement with
the vehicle lifted up (tension-free) and consists
of the earlier mentioned defined electrical decali-
bration of the strain gauge bridges on the torque
discs by means of the calibration switch.
The so simulated torque with its known value serves
as a mean for adjusting the whole measuring chain
from the disc through the computer to the plotter.
- 8 -
-------
- 8 -
Correspondingly, a defined frequency is introduced
before the frequency/voltage converter of the rpm
channel,vhich simulates a certain wheel rpm value. Since
in a preliminary test with loaded car and adjusted
tire pressure the amount of impulses over a well
known distance on the road was determined, each
impulse frequency can be correlated to a certain
vehicle speed.
h. Measurement and Evaluation
According to Daimler-Benz's experience,gained over
the last two years during road load determination
testing by constant speed driving and simultaneously
computerized evaluation, it is essential to define
the "constant speed" condition.
It finally was decided to define this term first
as a condition where as well the vehicle speed as
the (sum of) torque do not exceed a certain spread.
This definition represents a desired quality/acceptance
criteria for the measurement points. Second, a limit
was set for the drift of vehicle speed and (sum of)
torque over a certain measurement time.
Both limits were set according to experience as follows:
Max. spread of vehicle speed for tests on the
road/test track:
- 0.2 Kph
max. spread of (sum of) torque for tests on
the road/test track:
- 7-5 Nm.
- 9 -
-------
- 9 -
The latter spread allowance is increased at very
low vehicle speeds.
Max. allowable drift of vehicle speed for tests
on the road/test track:
0.05 Kph/8s
Max. allowable drift of (sum of) torque for tests
on the road/test track
1 Nm/8s
For tests performed on the chassis dynamometer,
these tolerances can be further narrowed.
During the trial, the computer tests each measurement,
and shifts the detected momentary values of vehicle
speed and torque in a pre-programmed slide storage.
With every new pair of data available, the computer
checks whether the above described tolerances are
met for all stored data pairs.
If this is not the case, the data pair, which was
stored longest, will be shifted out of the storage
when another data pair becomes available.
The check for meeting the tolerances then starts
again and this routine continues until all criteria
are met, i.e. until the desired constant speed is
reached. At this moment, the average speed V and
average torque M as well as the standard deviation
s and s are calculated for all data pairs in the
V M
storage, and the results are transferred into the
storage for final results.
- 10 -
-------
- 10 -
The slide storage content is then erased,and a
new search run starts.
In practice, it has proven useful to start this
search run only above 5 Kph constant speed. Further,
there is the possibility to interrupt the search
run for a time, if necessary. The constant speed
points of a road load curve can be approached in
free order.
As soon as enough valid measurements are available
in the result storage, pushing a button terminates
the search runs and switches the program over to
calculation. According to the least square method,
the best fit curve is calculated and plotted through
the measurement points using the formula
b
M = a • v + c • v + d
If needed, the formula
M = a . v^ + c . v + d
can be used as well.
The computer prints the coefficients of these
equations together with the corresponding average
quadratic error.
The plotter delivers a graph of the road load
curve together with the test conditions,which
have to be given into the computer by hand.
If desired,the numeric data of the individual speed
points (V and M) together with their standard devia-
tions are printed out as well.
- 11 -
-------
-11-
An additional program allows for immediate
determination of average road load curves calcu-
lated e.g. from several test runs in one and
several test runs in the other direction of a
test track or road.
Further, all driving resistance curves as plotted
after the test run can be corrected to standard
conditions by giving track grade, barometric
pressure ambient temperature and wind conditions
into the computer. It must be noted, however, that
Daimler-Benz restricts test runs with the system
to practically zero wind conditions, if at all
possible.
In order that this correction can be performed,
knowledge of the vehicle's (c^- A)_term is essential
if not a mere quadratic characteristic of the
driving resistance curve is assumed.
Fig. 08 shows an example for a road load resistance
curve established with the above described system
from driving at constant speed points in two track
directions. An excerpt from the computer print out
is shown as well.
5. Other possibilities of the system
The Daimler-Benz torque disc and evaluat
can be used for further testing purposes
proven in practice, e.g.:
- determination and check of the chassi
inertia mass(es),
- 12 -
ion system
as already
s dynamometer's
-------
- 12 -
- determination of the sum of positive torque
or energy transferred via the vehicle's wheels
during a complete test cycle or portions of it,
- separation of rolling resistance from acceleration/
deceleration portions of driving forces at constant
speed (on board evaluation),
- comparison of dynamic behavior of different dynamo-
meter types or routine check of the same dynamometers
over time (on board evaluation)
Details and results of these investigations will be
published in the near future.
-------
Engine
T
$
Torque Shaft
Differential
Torque Shafts
£
ch -n
inferential
Torque Discs
Possibilities of Torque Measurements
,-n - w*hi-1 - 's n-:vr T~ai"
v. j
V1MA 0879
01
-------
1
8
1 Increase of Tread Width by Installation
of Torque Disc
2 Torque Disc (portion connected to drive shaft)
3 Torque Disc (portion connected to wheel)
4 Bolt Circle with 32 Bores (32 webs, 16 webs
with strain gauges)
5 Measuring Web with Strain Gauge
6 Inductive Rotation Transmitter
7 Standard Rim
8 Protection Cover
9 Brake Disc
10 Wheel Suspension
a-a Flow of Driving Force
Schematic of Torque Disc and its Connection
V1MA 0879
to the Wheel of a Passenger Car
02
-------
Cross Section of
Torque Disc and
Transmitter
V1MA
03
0879
-------
(12V=)
(Measuring Signal)
(Feed)
Light-Electric
rpm Pick-Up
Inlet-(Feed-)
Transmitter
Outlet-(Measurement-)
Transmitter
Calibration Device
Schematic Outlay of the
Inductive Rotation Trans-
mitter with rpm Pick-Up,
Electrical Calibration
Device and Strain Gauge
Bridge
V1MA
04
0879
-------
®
Detail of Torque Disc Installation on Vehicle
V1MA 0879
05
-------
®
View of Vehicle with Torque Disc
V1MA 0879
06
-------
© 1
Block Diagram of Measuring-,
V1MA 0879
ri
<
Instrumentation- and Evaluation System
07
-------
M=A*VtB+CV+D
A
250..
200..
150 J.
¦o
5=
a>
ZD
cr
-1001
50..
0
0
A=
1 .428
E-02
B=
1 .970
E 00
C=
3.103
E-01
D =
5.486
E 01
M=
7.1 76
E 00
M.PKT
97
(tt° o-f metis. points )
MAX. SDDABU. V= max.std.dev,
0*13 for speed
BEI V-QUER= for average
50.11 speed
MAX. SDDABW. M= max.std.dev,
6.99 for torque
BEI M-QUEF<= for average
72.95 torque
venicxe speea
torque
std.dev. for speed
std.dev. for torque
upper measurement points
for uphill driving
lower measurement points
for downhill driving
Ijt^uHU.v IN r\rri
DREHMOMENT IN NM
STAND.ABU. F. V
STAND.ABW. F. M
98.34
213.58
0.1 1
1 .82
98.21
213.39
0.11
1 .04
68.08
134.65
0.05
0.77
68.1 5
133.54
0.03
0.64
50.1 1
88. 13
0.13
1 .71
10
80
20 30 40 50 60 70
Vehicle Speed v in kph
Evaluation Example of Road Load Determination
on Test Track (Average of Uphill and Downhill
90
100
V1MA
08
39.30
72.95
0.11
6.99
29.46
69.64
0.09
1 .25
0879
V i nnl
-------
3E3
0 \ , ~ , , , , . .
0 10 28 38 40 50 68 70 80 93 100
Vehicle Speed v in kph =»-
388
B IB
22 33 40 58 63 70 1
Vehicle Speed v in kph -
n 93 IBS
=*.
®
Influence of Speed Drift
V1MA
^ S
at "Constant Speed"-
09
Driving on Spread of Road
Load Measuring Points
-------
®
View of Electronic Measurement and Evaluation
V1MA .0879
System for Road Load Determination and Chassis
10
Dynamometer Adjustment
-------
Attachment mrr
Technical Report
A Track to Twin Roll Dynamometer Comparison of
Several Different Methods of Vehicle
Velocity Simulation
by
John Yurko
June 1979
NOTICE
Technical Reports do not necessarily represent final EPA decisions or
positions. They are intended to present technical analysis of issues
using data which are currently available. The purpose in the release of
such reports is to facilitate the exchange of technical information and
to inform the public of technical developments which may form the basis
fo a final EPA decision, position or regulatory action.
Standards Development and Support Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
-------
ABSTRACT
The current EPA test procedure for fuel economy and emissions
testing uses a twin roll dynamometer, obtaining a speed signal from the
rear roll and simulating the forces at the front roll. With the rolls
coupled only by the drive wheels of the vehicle, the front roll travels
approximately 2% slower than the rear roll at steady-state 50 mph, re-
sulting in approximately a 4% overprediction of fuel economy. Coupling
the rolls externally equalizes the roll speeds at a value which better
simulates the road velocity and therefore better predicts the fuel
economy. This report describes the test program and data analysis which
led to these conclusions.
FOREWORD
The EPA has conducted a test program in order to determine the most
representative method for simulating the road velocity of a vehicle on
a Clayton twin-roll dynamometer. The three methods of simulating the
road velocity on the twin-roll dynamometer are:
(1) Using the velocity of the rear roll, which is the current
method,
(2) Using the velocity of the front roll,
(3) Operating with the rolls coupled.
To determine which of these three methods most closely represents
the road experience of a vehicle, steady-state tests were conducted on a
track and compared to dynamometer tests using each speed simulation
method. The same vehicle was used for all phases of the test program.
This report describes the test program, reports the results, and recom-
mends the most appropriate method of velocity simulation on a twin-roll
dynamometer.
SUMMARY
The results of the road to dynamometer comparison show that the
road velocity is best simulated when the front and rear rolls of the
dynamometer are coupled. With the rolls coupled, the simulated velocity
was within 0.025% of actual road velocity. With the rolls uncoupled,
the rear roll velocity over credited the vehicle speed by approximately
1.0% while the front roll under credited the speed by about 1.0%.
Coupling the rolls reduced measured fuel economy by approximately 4% in
comparison with the current method of using the rear roll speed. This
is consistent with the 1% speed errors in each roll, since the force is
proportional to the velocity squared. In conclusion, coupling the rolls
is technically the best method of simulating the vehicle velocity and
should improve EPA fuel economy predictions.
-------
I. INTRODUCTION
When a vehicle is tested for fuel economy and emissions on a
Clayton twin-roll dynamometer, there is a difference between the velo-
cities of the front and rear rolls of the dynamometer.
Therefore, the speed sensor location can have a significant effect
on fuel economy and emissions testing. Steady-state tests have shown
that the rear roll travels approximately 1.0 mph faster than the front
roll at 50 mph (1). This occurs because the drive wheels of the vehicle,
which are cradled between the two rolls, act as the only coupling be-
tween the two rolls when a vehicle is driven on the dynamometer. The
power absorber and inertia flywheels, which simulate the road force
experienced by a vehicle, are connected to the front roll. This causes
a greater tangential force at the tire/front-roll interface than at the
rear-roll interface, resulting in a smaller effective rolling radius in
the tire with respect to the front roll as opposed to the rear roll.
Externally coupling the rolls eliminates the difference in velo-
cities of the two rolls. Therefore, this has been considered as an
alternative method for simulating the vehicle speed. Locating the speed
sensor on the front roll has also been suggested, since the forces and
the velocity would then be associated with the same surface. To de-
termine which method would best simulate the actual road velocity of a
vehicle a test program was conducted. The following discussion de-
scribes the track tests, the dynamometer tests, and the road to dyna-
mometer comparison which were used to determine the optimum method for
measuring the simulated velocity of a vehicle.
II. DISCUSSION
The test program consisted of three portions: 1) track portion 2)
dynamometer portion, and 3) data analysis. The track portion was con-
ducted at the Transportation Research Center of Ohio (TRC). The dyna-
mometer portion was conducted at the EPA laboratory in Ann Arbor. One
vehicle, a 1978 Mercury Montego, was used for all testing. Steady-state
tests were conducted on both the track and the dynamometer, for four
different sets of radial tires which are listed in Appendix A-l.
A. Track Portion
Prior to each test, the vehicle was weighed with a full tank of
indolene test fuel, complete instrumentation, and two operators. After
a 20-minute warm up at 50 mph around an oval track, data were collected
during one lap of the track for approximately 10 minutes at steady state
50 mph. Both left and right rear wheel speeds, left and right rear
wheel torques and a fifth wheel speed were recorded at a once/second
rate. Total fuel flow and distance traveled were also measured. Am-
bient temperature, barometric pressure, wind velocity and wind direction
were monitored during the tests. Tire temperatures were recorded before
and after each test. Immediately following the steady state test, 10
-------
coastdowns were conducted in accordance with the EPA recommended prac-
tice for determination of road load for light-duty vehicles. A detailed
description of all the equipment used is given in Appendix A-2.
B. Dynamometer Portion
The goal was to reproduce the exact road torque and speed condi-
tions for each test on the dynamometer. In order to obtain the necessary
precision, we instead chose to use a 9-point speed/torque test matrix,
and then to interpolate the dynamometer data to the road datum.
For the dynamometer tests it was decided to warm-up the tires so
that they would be at approximately the same conditions as are vehicle
tires during typical EPA tests. This was chosen since the results would
be more representative of conditions during EPA tests than would result
from a 20 minute 50 mph steady-state warm-up and there would be reduced
probability of tire failures. This approach also resulted in tire
temperatures which were closer to the road tire temperatures than would
have occurred with the 20 minute steady-state warm-up.
The test cycle chosen consisted of a tire warm-up of one complete
FTP cycle followed by three consecutive 5 minute steady-state measure-
ments at a single horsepower. At this time a 15 minute cool down period
was provided before the dynamometer adjustment was changed, then the
first 505 seconds (bag one) of the LA4 cycle was driven to precondition
the tires and three more steady-state measurements were obtained. This
cycle of a cool down followed by a preconditioning was repeated until
all data necessary for the 9 point matrix were obtained. The 15 minute
cool down followed by the 505 seconds of preconditioning was chosen on
the basis of tire temperature measurements, to be appropriate to yield
approximately the same tire temperatures as were obtained after one
complete LA-4 cycle starting with a cold tire. No tire failures were
observed in this program, either as a result of the warm-up cycle or the
measurement conditions.
The vehicle was tested with each set of tires at three steady-state
speeds, nominally: 1) 50 mph, 2) 40 mph, and 3) 55 mph. For greater
precision the actual measured velocities were used in the data analysis.
The 55 mph point was chosen instead of 60 mph since, at 60 mph the tire
temperature increased rapidly, indicating possible tire failure prob-
lems. Data were collected during each steady state for 5 minutes at a
once/second rate. As in the track portion, both rear-wheel torques and
rear-wheel speeds were recorded. Instead of a fifth wheel speed, the
front and rear dynamometer roll speeds were recorded. Fuel flow and
rear roll distance traveled were also measured. Each steady state was
followed by a vehicle/dynamometer coastdown from 55 mph to 45 mph and
the coastdown time was recorded. The dynamometer coastdown times were
only used for a fuel economy comparison as described in Section III.
The steady-states and the coastdowns were repeated at each speed for
three different indicated dynamometer power absorber settings: 1) 11.4
HP, 2) 12.4 HP, 3) 10.4 HP, in that order. This test sequence is sum-
marized in the 9-point test matrix shown in Figure 1. The 11.4 HP value
-------
Figure 1
Dynamometer Test Matrix
12.4
Dyno Power
Absorber
Setting (HP)
11.4.
10.4..
40
50
55
Nominal Steady State Velocity (MPH)
-------
approximately represented the road load of the vehicle with the midrange
set of tires, as determined by matching the road and dynamometer coast-
down times. The 10.4 and 12.4 test values were chosen to cover the
range of road loads, observed with different tires. For greater pre-
cision, actual wheel torques, and wheel speeds were used to match the
dynamometer test to the road. This was done by a linear regression
which is described in Section 1IC.
The entire configuration was then repeated, for each tire set, with
the front and rear rolls coupled by a motorcycle chain and sprockets
connected to each roll. A detailed description of the test sequence
including warm-up cycles for the dyno portion is given in Appendix B.
All the equipment used in the dyno portion was the same as the equipment
used in the track portion with the exception of replacement of some
minor damaged components and the additional equipment associated with
the dynamometer. These are included in the equipment list of Appendix
A-2.
C. Data Analysis for Road to Dynamometer Comparisons
For each set of tires, one 50 mph steady-state test was conducted
on the road. For each test, mean rear wheel angular speeds, mean rear
wheel torques and a mean fifth wheel speed were calculated.
Conceptually, the intent was to reproduce the rear wheel torque and
speed conditions of the vehicle which were observed on the road, for
each set of tires on the dynamometer. Under these conditions, the
different possible speed measurements would be sampled, and that method
of measurement which best agreed with the road fifth wheel velocity
would be selected as the most appropriate method of measuring the dyna-
mometer simulated speed.
The conceptual approach could not be used directly because of the
experimental precision considered necessary to resolve the small velo-
city variations among the different methods of dynamometer speed simula-
tion. Therefore, we chose to use the 9-point steady-state speed/torque
test matrix described in Figure 1.
The data obtained at these points uses the interpolated velocity to
obtain a roll velocity corresponding to the conditions observed during
the road tests. The interpolation was conducted by means of a multiple
linear regression using the mean of the data at each point of the test
matrix.
First, as discussed, the mean values of each rear wheel angular
speed, each rear wheel torque, and each dynamometer roll velocity, with
rolls coupled and uncoupled was calculated for every steady-state
test. An example of these data for one of the nine point matrices is
graphically shown in Figure 2. The interpolation of these data to the
observed road point was accomplished by regressing each roll velocity
versus the sum of the mean rear wheel angular speeds and the sum of the
mean rear wheel torques, over each 9-point test matrix, yielding the
coefficients fo the following equations:
-------
Figure 2
Tire 4 With Rolls Coupled
Sum of the Rear Wheel Torques vs. Sum of the Rear Wheel Angular Speeds
-------
V - *R + CR (l)
Vr = aF(QL + V + V\ + V + CF (2)
Vcoup = ac(WL + WR) + bc(TL + TR) + C,, (3)
Where:
VRR = mean rear roll velocity
= mean front roll velocity
FR
V = mean rear roll velocity with rolls coupled
coup J
W = mean left wheel angular speed
L
WD = mean right wheel angular speed
R
T = mean left wheel torque
Jj
T_ = mean right wheel torque
R
* s * s ' s
a , b , c = unique sets of regression coefficients for
each roll condition and each 9-point test
matrix
The road values of mean wheel torques and speeds were inserted into
equations (1), (2), and (3) for each set of tires to obtain the simulated
road velocity for each method of speed measurement interpolated to the
road conditions. The predicted road velocities as given by the above
equations, were then compared to the actual mean road velocity for the
same set of tires:
VRR/Road = + \Voad + bR^L + \\oad + CR
VFR/Road ~ + VRoad + bF(\ + VRoad + CF
-------
Vcoup/Road " aC^L + \Voad + bC(^C + ^Road + CC
Where:
VRR/Road
= Road velocity as simulated by the rear roll at the
road conditions
VFR/Road
= Road velocity as simulated by the front roll at the
road conditions
^coup/Road = ^oa<* velocity as simulated with the rolls coupled
a,b,c = the set of coefficients obtained from the regressions of
the dynamometer data for each tire (different for the rear
roll, front roll, and coupled roll predictions)
Sample calculations and the original data, including the regression
coefficients are given in appendix C.
III. RESULTS
The results of all tests on the radial and bias belted tires are
given in Table 1.
The mean deviation from the actual road velocity for the radial
tires was +1.10% using the rear roll velocity simulation, -1.07% using
the front roll, and -0.22% with the rolls coupled. Where, a positive
deviation corresponds to an observed dynamometer velocity greater than
the road velocity under the same wheel condition.
For the bias-belted tires, the rear roll deviated by +1.23% from
the road, the front roll deviated by -0.04%, and the coupled rolls
deviated by +0.40%.
Overall, the rear roll was in error by +1.15%, the front roll by
-0.71%, while the error with the rolls coupled was only -0.02%. Therefore,
on the average and particularly for radial tires the coupled mode most
closely simulated the road.
Since coupling the rolls improved the vehicle velocity simulation,
the vehicle fuel economy effect of this change was investigated. In the
majority of EPA fuel economy tests, alternate dynamometer adjustments,
obtained by the coastdown technique, are used. Also the coastdown
method is used in dynamometer calibration, and therefore, would account
for the increased friction of the coupling mechanism. Consequently, a
comparison of vehicle fuel economy, obtained with dynamometer adjust-
-------
ments which produced equal coastdown times was considered the most
appropriate approach to evaluate the fuel economy effect of coupling the
dynamometer rolls. This comparison could easily be made since during
the dynamometer portion of this test program vehicle dynamometer coast-
down times were recorded immediately following the fuel consumption
tests.
Figure 3 shows the 50 mi/hr fuel consumption of the vehicle equip-
ped with radial tires plotted versus the coastdown time obtained for
both the uncoupled and coupled tests. This plot indicates that coupling
the dynamometer rolls results in a 2 to 6 percent increase in measured
fuel consumption for the same vehicle-dynamometer coastdown time. For
example, at a coastdown time of 14.0 sec, the fuel consumption was
approximately 7150 cc/km with the rolls uncoupled and about 7450 cc/kra
with the rolls coupled, a difference of approximately 4%.
The fuel economy results obtained in this test program are all from
steady-state measurements. However, the results are consistent are pre-
liminary investigations of the effect on transient cycles. For example,
computer modeling has estimated the transient cycle fuel economy effect
to be 4%.(2) Limited empirical data from transient cycle tests also
indicate the effect to be about 4%.(3)
IV. CONCLUSIONS
Operating with the rolls coupled most closely simulates the road
experience of a vehicle using radial tires, and therefore, provides the
most accurate method of testing for fuel economy. The current EPA
method for simulating the vehicle velocity, using the rear roll speed,
causes an over prediction of steady-state 50 mph fuel economy by approx-
imately 4%. This occurs because the velocity error results in both an
underloading of the energy demand from the vehicle and an overcredit of
the distance travelled.
The same mechanism occurs during transient cycles and in this
instance, inertial forces applied to the vehicle are also inappropri-
ately low because of the velocity error. Computer modeling and limited
empirical data indicate the transient cycle fuel economy errors re-
sulting from this velocity error are also about 4%. It should be noted
that these conclusions are based on data from vehicles equipped with
radial tires, however this is the most important case. It is estimated
that over 70% of the vehicles tested at EPA are equipped with radial
tires.
-------
Table 1
Radial Tires
Tire
No.
Road Velocity
Predicted by
the Front Roll
Cmph)
Road Velocity
Predicted by
the Rear Roll
(mph)
Road Velocity
Predicted with
Rolls Coupled
(mph)
Observed Road
Velocity
(mph)
1
2
3
4
50.05
49.72
50.18
50.26
51.19
50.58
51.43
51.38
50.45
50.00
50.85
50.61
50.81
50.10
50.83
50.62
Mean
% Deviation
50.05
-1.07
51.15
+1.10
50.48
-0.22
50.59
/Predicted - Observed^ ,
Observed
Bias Belted Tires
6
7
50.53
50.55
51.16
51.20
50.82
50.70
50.51
50.60
Mean
% Deviation
50.54
-0.04
51.18
+1.23
TOTALS
50.76
+0.40
50.56
Mean
% Deviation
50.22
-0.71
51.16
+1.15
50.57
-0.02
50.58
Error analysis indicated that on the average, we were 95% confident that
the predicted values were accurate to within +0.23 mph.
-------
References
1. Richard Burgeson, Myriam Torres, "Tire Slip on the Clayton Dyna-
mometer", EPA Technical Support Report, LDTP 78-02, March 1978.
2. John Yurko, "Computer Simulation of Tire Slip on a Clayton Twin
Roll Dynamometer", EPA Technical Support Report, SDSB 79-10,
February 1979.
3. Conversation with Don Paulsell, of the EPA Ann Arbor Laboratory,
March 1979.
-------
Figure 2
Tire 4 With Rolls Coupled
Sum of the Rear Wheel Torques vs. Sum of the Rear Wheel Angular Speeds
-------
Figure 3
Coupled vs* Uncoupled Fuel Consumption
Plotted Against Coastdown Time
(with radial tires)
Tire 1
Tire 2
Tire 3
Tire A
¦ - uncoupled
A - uncoupled
• - uncoupled
~ - uncoupled
~ - coupled
A - coupled
O - coupled-
¦Ar - coupled
7800 "¦
e
a
u
c
o
•H
4J
a
e
a
c
o
c_>
0)
3
fK
coupled
7200
6600 --
upper limit
to uncoupled
lower limit /
to uncoupled
1 In
13.0 14.0 15.0
Coastdown Time (sec)
-------
Appendix A-l
Tire
Tire
Tire
No.
Tire
Type
Size
1
Michelin-X
Radial
GR78xl5
2
Firestone 721
Radial
GR78xl5
3
Firestone 721
Radial
GR78xl4
4
Multimile Supreme
Radial
GR78xl5
6
Uniroyal Fastrak
Bias Belted
G78xl5
7
Uniroyal Fastrak
Bias Belted
G78xl5
1976
Vehicle
Mercury Montego w/ 29,000 accum. miles
-------
Appendix B
OVERVIEW OF TEST SEQUENCE
(eg. using tire no. 3)
1. Tire no. 3 mounted, pressure set to 45 PSI.
2. Tires broken in with 1 FTP.
3. Allowed to cool at least 4 hours.
4. Reset pressure to 45 PSI.
5. Vehicle rear axle weight approximately 2290 lb with driver and
full gas tank.
6. Set dynamometer inertia to 5000 lbs.
7. Set dynamometer horsepower to 11.4 horsepower.
8. Set fixed data to 350114.
9. Conduct 1 FTP, then obtain tire temperatures.
10. Insert tape in techtran, ready for scan at 1 second intervals
(Tape labeled: uncoupled 350114, 3501124).
11. Conduct a 5-minute Steady State at 50 mph, collect data.
12. Conduct a coastdown, collect 55 to 45 mph time only.
13. Record tire temperature during or right after coastdown, reset
fixed data to 340114.
14. Conduct a 5-minute Steady State at 40 mph, collect data.
15. Conduct a coastdown. (NOTE: be sure to collect data only during
the 5-minute Steady State. All data collection devices should
be reset before new Steady State speed is set.) Collect coast-
down time and tire temperatures.
16 Reset fixed data to 355114, conduct a Steady State at 55 mph
for 5 minutes collecting data. Stop data, conduct a coastdown,
record time and tire temperatures. Increase speed above 60 mph.
17. Life vehicle, conduct a dynamometer only coastdown, check zero,
adjust on torque meter. Record 55 to 45 mph time. Reset horse-
power to 12.4, fixed data to 350124. Tires should be allowed to
cool 15 minutes starting from when the vehicle was lifted.
-------
Appendix A-2
Type of Data
Being Collected
Drive wheel torques
(analog voltage output)
Wheel angular velocities
(frequency output)
Conversion of frequency
to analog voltage
Collect and digitize
analog signals for
output to a recording
device
Record data
Record fuel flow
Tire temperatures
Equipment
Lebow torque sensor
Model No. - 7510
Disc/Rotaswitch pulse
Encoders
Anadex frequency to
voltage converter
Fluke datalogger
Model 2240B
Techtran Data Cassette
Model 8400
Fluidyne Flowmeter
Model 1250T
Wahl Heat Spy
Infared thermometer
-------
Appendix B (cont.)
18. Conduct a 505 second warm up, record tire temperature.
19. Repeat 11 and 12.
20. Reset fixed data to 340124.
21. Repeat 14 and 15.
22. Reset fixed data to 355124.
23. Repeat 16.
24. Repeat 17. Reset horsepower to 10.4 after dynamometer coast-
down, fixed data to 350104, allow 15 minutes cooling, rewind
tape and insert new one. (Tape labeled: Uncoupled, 350104).
25. Conduct a 505 second warm up, repeat 11 and 12.
26. Reset fixed data to 340104.
27. Repeat 14 and 15.
28. Reset fixed data to 355104.
29. Repeat 16.
30. Conduct dynamometer only coastdown, recheck zero drift. Rewind tape.
31. Steps 1 through 30 complete a tire for the uncoupled config-
urations. Approximately 3 to 4 hours of testing and 2 cassette
tapes are required. If nothing is done to the vehicle but to let
it set for an hour (say for lunch), you should be able to start at
step 6 with rolls coupled and fuel tank filled, and conduct steps 6
through 30 to complete a tire type.
Steps 1 through 31 will be repeated for each tire set.
-------
Appendix C-l
Tire 3
9-Point Test Matrix Data on Dynamometer
with Rolls Uncoupled
RRV
FRV
(wt-hL)
(tt+t_)
Test
(mph)
(mph)
(rev/sec)
(ftr-iSs.)
350114
50.116
49.170
21.598
155.892
340114
39.979
39.423
17.239
120.743
355114
55.3,25
53.932
23.672
181.191
350124
49.955
48.556
21.402
167.662
340124
40.091
39.183
17.198
125.286
355124
55.074
53.421
23.508
187.087
350104
50.068
48.857
21.415
149.114
340104
40.166
39.375
17.181
112.087
355104
54.881
53.459
23.426
164.744
With Rolls Coupled
351114
50.031
50.075
21.710
178.241
341114
40.109
40.130
17.340
138.261
356114
55.030
55.084
23.941
206.123
351124
49.870
49.902
21.618
177.924
341124
39.936
39.948
17.306
131.293
356124
55.011
55.045
23.894
202.540
351104
49.990
50.041
21.678
159.231
341104
40.029
40.040
17.312
118.504
356104
55.031
55.079
23.821
183.033
Regression Coefficients
(Vi = a (WT + W_) + b
Lt K
-------
Appendix C-2
Example Calculation Using Tire 3
Rear Roll Velocity with the Rolls Coupled
V = a (W_ + Wp) + b (T_ + T_) + C
coup L R L R
from linear regression with RRV from appendix C-l as the dependent variable: :
a = 2.3320, b = -0.46453 x 10~2, c=0.23591
therefore, applying the coefficients to the road data results:
V . . = 2.3320 (W. + 5 ) road + -0.46453 x 10~2 (T. + T ) road + 0.23591
coup/road L R L R
where:
(WT + Wn) road = 22.03, and (TD + T.) road - 162.98
L R K L
therefore:
V , .= 50.85
coup/road
this compares to the actual road velocity:
V , = 50.83
road
(These correspond to the results given in Table 1. Section III of this
report.)
-------
Appendix C-3
Tire 1
9-Point Test Matrix on Dynamometer
vith Rolls Uncoupled
RRV
TRV
(W + W )
-------
Appendix C-4
Tire 4
9— Point Test Matrix on Dynamometer
with Rolls Uncoupled
Test
450114
440114
455114
450124
440124
455124
450104
440104
455104
451114
441114
456114
451124
441124
456124
451104
441104
456104
Vi
Rear roll
Front roll
Coupled roll
RRV
FRV
(W. + 5)
(T. + T_)
(mph)
(mph)
(rev / sec)
(ftV - l&s)
49.960
48.951
20.864
152.446
39.968
39.374
16.623
115.488
55.013
53.765
22.849
176.968
49.990
48.780
20.786
156.114
40.007
39.238
16.591
114.620
55.120
53.680
22.849
178.410
49.956
48.907
20.754
142.134
39.952
39.287
16.575
107.053
55.054
53.873
22.856
165.278
with Rolls Coupled
49.697 49.847 20.999
40.636 40.733 17.021
54.887 55.113 23.223
49.929 50.084 21.154
40.589 40.688 17.048
54.918 55.084 23.268
49.983 50.126 21.086
40.074 40.175 16.810
55.295 55.484 23.359
Regression Coefficients
b x 102 c
a
2.3939
2.3838
2.3781
0.12847
-0.78016
-0.64562
0.071297
0.57721
0.84438
160.798
119.410
189.825
173.993
125.797
200.720
151.774
109.630
175.490
R-SQR
.99973
.99977
.99992
+ road = 21,35 and + TR) road = 155.603
-------
Appendix C-5
Tire 2
9- Point Test Matrix Data on Dynamometer
with Rolls Uncoupled
RRV
FRV
(iiL + SR)
(fL + V
Test
(mph)
(mph)
(rev / sec)
(ft. - lb!
250114
49.880
48.803
20.916
159.274
240114
39.949
39.251
16.773
123.905
255114
54.976
53.764
22.994
182.863
250124
49.921
49.108
20.980
165.863
240124
40.014
39.561
16.863
123.553
255124
55.061
54.018
23.037
189.958
250104
49.945
49.237
21.036
147.109
240104
39.887
39.492
16.820
108.334
255104
55.121
54.222
23.209
165.357
with Rolls Coupled
251114
50.197
50.340
21.404
174.465
241114
40.167
40.251
17.071
130.972
256114
55.121
55.290
23.511
198.727
251124
50.009
50.132
21.347
172.235
241124
39.980
40.056
16.993
124.866
256124
54.958
55.139
23.455
197.955
251104
49.925
50.061
21.243
161.941
241104
40.111
40.194
17.025
123.351
256104
55.061
55.218
23.437
186.192
Regression Coefficients
Vi
a
b * 102
c
R-SQR
Rear roll 2.3000
Front roll 2.3081
Coupled roll 2.4186
1.1332
0.17002
-0.91726
-0.085243
0.40039
0.045152
.99987
.99984
.99996
-------
Appendix C-6
Test
9-Point
Tire 7
Test Matrix Data on Dynamometer
with Rolls Uncoupled
(T + T )
(ftV - iEs)
RRV
(mph)
FRV (W + w )
(mph) (rev / sec)
750114
49.996
49.435
21.102
154.821
740114
39.878
39.562
16.891
114.676
755114
55.081
54.341
23.189
178.088
750124
50.054
49.396
21.114
156.746
740124
39.947
39.576
16.915
112.423
755124
54.917
54.065
23.090
177.521
750104
50.108
49.549
21.136
142.726
740104
40.084
39.773
16.994
106.785
755104
55.158
54.473
23.312
165.109
with Rolls Coupled
751114
50.123
50.281
21.368
166.150
741114
39.894
39.994
17.040
121.278
756114
55.142
55.391
23.513
191.962
751124
50.028
50.168
21.333
167.352
741124
40.034
40.117
17.101
118.877
756124
54.957
55.104
23.415
191.912
751104
49.947
50.100
21.289
157.123
741104
40.041
40.131
17.102
113.344
756104
54.966
55.129
23.372
177.934
Regression Coefficient
s
Vi
a
b x 102
c
R-S0R
Rear roll 2.3246
Front roll 2.3209
Coupled roll 2.4311
0.84300
0.20201
-0.59311
-0.32574
0.12043
-0.83437
.99992
.99994
.99998
CWL + WR) road - 21.59 and (fR + f ) road = 159.165
-------
Appendix C-7
Tire 6
9-Point Test Matrix Data on Dynamometer
with Rolls Uncoupled
RRV
FRV
CWL + V
-------
COASTDOWN TIMER
USED BY EPA
(ATTACHMENT X)
-------
*.
~
SERIAL NO-t
DC TACH
TE8T
-------
*4
FRoa/T ?A//EL LaYOL/T"
cRystal
Sto PWATCL H
S ECO//DS
i&Jjri 4ba£*
hAf| /Wir/«VWl 0£T £~i£c TRaa/>C SysT&ftS) JlsSC.
f c o*/a/£c r*R \
ori
?oW£K (o) st-
op F
STOP
oV*
CO Us/7~
fAf\Nuh(-
(o)S3
A U-To/Kf
-------
S£"Ri/lI> NO.
SET -poiMTS DC T/\CH
5-5 4S TEST"
O O O O O O
<5tfO- TP 3 TP 2 G^D. T?X
O
Dc Tach
XrfpuT
IT 4-
fuse
Bo$ Tovif j
v-/y-7y
5.5® 5£"7~ 4"* 5 (5) S£ 7~
*5 T?Z
V£A% P/irf&L L.Ayocr~
o
o
5" /ozr poujzr soppLy
-------
5 VOLT RcGUIAtElP
P c Boa-rd LAyou-r
-qr
/»c
AC
i: i 5 vo LT T>oW£H S»P?vy
CO/A
+
•B/ "Bos T<9V£\J
V- n- 7 2
-------
h
o
o
i A
^ i
,3S
O
r5
c
•\>
3»
•o
l-f
%
c>
£
—o
<3—
ft
o .,
CT~
*
2J~
4r
.A/SV-AC
r-
_£t
Co/n
-^._—V,5~K0C.
-, 3-*
—\=A/\A4>L*£^r_—
— £T„-f |-^-T7»=J
n __ *\
' ":> •J'i-Cfi on/y
tip- ^
'M_SLufi£ly_
Q-osn
• > [>. T/fC H_JOtp_c/7~ _
-------
C OASTPortrt 7
zg.h£ /nflT/'c.
£.7>.4-
C OASTDOWV TiMSH
"By Tohtf KElLSHy Pov p£pp£H
£ Tio& t2~~tty—i #
ol
rc.2 r
-------
Cofisr- nouM ~Ti/A £¦ R ~Pa>-RTS a; 5 7~
S*"PPLy w 1 -/SVOC, Zoo^/J SOLA £~££lTH'>tt 8f-IS--03.ll oe
TowEiK s^pL/ *" 3 fSVPC. 3a P"~>£&-o*/£, MerntFi "8^-3
Xti
<61/ 1Z£?£R£tJC£ ?A1X 7?£~/?-6t J?A/d>
mop*l- t>s /* ^ v*7T C*ti.a<>*/
Hi
jfio-f*- ^Z'0, y" Q.iVlQot/
u
<1
f?7, ??/o
fl/l
S^J, ^ 9P-l-jol
7?3
/4v\_ ( JO TukfJ 8o Tod. S
C./( C£T, C6
.-55" tAPD. Z-^WvOL. D/J- 0£7?W/n.'C-
O, <17, C *
V-7 /Jl^O /6U/V0C. £l£t-/-XO lyriQ
ta "ztooyioi- oc /y
d?
tMMfb ^ou/yoc_ Et-SfrKclyi-iz. f*\3foo *¦/?!-
-------
VflHTS Li iT C.6*JTiHuE'b
CfiMaiToA-S C 3 ,tZM/*o /oou,^a
/n B/M Gti-tf\k. p»S H b(S!To*t flLCo MpA 'dC.
LiLu& ; Tflc. H XfS ?/ PtPHeA/oi. I3&-OH,
caa jr*/ T/ ftmpH£-*/oL
Li*/ e pio& i; it>#D 11-3 Type- svT ±'-3y^ i.c 7S7y
Fuse /jt/bfA. Buss /V-T/M
TfidKS SO t- o f£>3- & & ' > "XTefWS^/
Ca<3 ,W&T &uD G.0 CJL Type (ft i*/i. B»K SQ.-303O
pi \/£C.T6(L (3cA£b Uf pgy-013.
mote: ^JL use */8Ga. u)A£ Po* Citfitfta*' ft + srvo
------- |