`SOCIETY OF AUTOMOTIVE ENGINEERS,
`Two Pennsylvania Plaza, New York, N.Y. 10001
`
`Development of the
`Federal Urban Driving Schedule
`
`Ronald E. Kruse and Thomas A. Huls
`Environmental Protection Agency
`
`Automobile Engineering Meeting
`Detroit, Mich.
`May 14-18, 1973
`
`730553
`
`1 of 8
`
`FORD 1343
`
`
`
`730553
`
`Development of the
`Federal Urban Driving Schedule
`
`Ronald E. Kruse and Thomas A. Huls
`Environmental Protection Agency
`
`MOTOR VEHICLE EXHAUST is a complex mixture of
`nitrogen (N), carbon dioxide (CO2), carbon monoxide (CO),
`oxygen (O2), hydrocarbons (HC), oxides of nitrogen (NOx),
`and other materials. The concentration of these exhaust
`components is variable and dependent upon the vehicle's
`operating mode. The measurement of the gaseous pollutants
`from motor vehicles, usually CO, HC, and NOx requires the
`use of sophisticated, complicated, and delicate instrumenta
`tion. It is extremely difficult to use this instrumentation in
`conjunction with motor vehicles as they are being operated
`in the urban traffic environment A motor vehicle can be
`operated in a laboratory on a chassis dynamometer and
`exhaust emissions can be properly measured. Basic questions
`which must be answered when vehicle emissions are mea
`sured in the laboratory are, how is the vehicle to be operated
`and how does one decide upon the driving conditions which
`will be represented by this operation9 These are questions
`which are most difficult to answer (1)*. Many groups have
`over the years, conducted research projects which have added
`
`*Numbers in parentheses designate References at end of
`paper.
`
`to the understanding of motor vehicle traffic-related, operat
`ing characteristics.
`
`HISTORY
`
`Among the early contributors to this understanding of
`urban traffic characteristics were personnel of the Los
`Angeles County Air Pollution Control District. They worked
`in cooperation with, or considered the work of,other groups
`in their efforts to identify driving habits or vehicle operating
`characteristics which typify Los Angeles traffic.
`In the late 1950s, Los Angeles County personnel in
`association with members of the Automobile Manufacturer's
`Association Traffic Survey Panel identified 11 characteristic
`driving modes which could be used in combination with
`appropriate weighting factors to characterize Los Angeles
`driving. This work culminated in the development of the
`California 7-mode cycle. The 7-mode cycle was based on the
`1956 (2) survey and was intended to represent average driving
`conditions throughout Los Angeles County during both
`peak and off-peak traffic conditions
`Early in the 1960s, the California air pollution personnel
`re-evaluated their criteria for determining average driving
`
`ABSTRACT
`
`This paper reviews the development of the LA 4 road
`route, and discusses efforts directed toward development of a
`short repetitive dynamometer cycle based upon the road
`route Also described are the instrumentation, methods, and
`selection process used to obtain a speed profile of a typical
`drive over the 12 mile long route The methods used to
`shorten the speed profile to 7.5 miles, and to shorten the
`average trip length, while preserving trip description such as
`Copyright ©Society of Automotive Engineers, Inc. 1973
`All rights reserved.
`
`average speed, idle time, number of stops, etc., are explained.
`A measure of the correlation of emissions from vehicles
`driven over the EPA Urban Dynamometer Driving Schedule
`(UDDS) and over the full LA 4 driving schedule is provided
`The UDDS is a speed-trace consisting of 18 profiles,
`separated by idle periods of 0-39 s duration The schedule
`covers 746 miles in 1372 s for an average speed of 19.6
`mph.
`
`2 of 8
`
`FORD 1343
`
`
`
`conditions and decided that the cycle should not represent
`24 h county-wide vehicle operation They felt that it was
`necessary to concentrate on the kind of driving which con
`tributes to the peak primary pollutant levels in the Los
`Angeles basin (3).
`They determined that the largest contributor to Los Angeles
`smog was the morning home-to-work trip. They felt that if
`they could identify the prevalent operating modes, they
`could reproduce those modes with vehicles on chassis
`dynamometers and measure exhaust emissions which result
`from these modes. As a part of their efforts, they instru
`mented several automobiles with the Du Pont driving habits
`recorder which recorded time in each driving mode. This
`device determined the amount of time spent in each of the
`16 to 19 engine speed-manifold pressure categories. The data
`were displayed on elapsed time clocks, one clock for each
`mode, with one additional clock to measure total elapsed
`time This was the state-of-the-art in the early 1960s(3).
`The Engineering Research & Development (ER&D) Section
`of the Division of Air Pollution of Health, Education and
`Welfare(HEW), the predecessor to the Office of Mobile
`Source Air Pollution Control (OMSAPC) of the Environ
`mental Protection Agency (EPA), built a modified version
`of the Du Pont driving habits recorder in the early
`1960s. This recorder functioned in the same way as the Du
`Pont recorder, except that it had 36 engine speed-manifold
`pressure categories. Evaluation of the HEW driving habits
`recorder indicated that the data gathered by that device
`were difficient in the basic respect that the chronology of the
`mode events was lost. Subsequently, a 22 channel strip
`chart event recorder was used to collect the same type of
`data. While this approach allowed analysis of the chronology
`of each trip, the data reduction problem was formidable.
`The next step was to use a fifth wheel and tachometer
`generator, attached to the back of the car, combined with a
`Potentiometric strip chart recorder. While this approach
`provided a good visual display of a trip, it also suffered data
`reduction problems, as had been noted by others (2).
`Attempts were made to reduce the strip chart speed re
`cordings to key punch cards, but even with the use of
`
`2
`mechanical curve readers this method was impractical. While
`this method of recording data provided useful and interesting
`information about road routes as they were driven, it could
`not be used, at this stage of development, to determine
`typical or representative routes. Because of these data
`reduction problems, the project was set aside at this point
`of development, in 1963.
`
`REPRESENTATIVE ROAD ROUTES
`
`Meanwhile, Los Angeles County and California personnel
`were gathering traffic and driving habits data, evaluating the
`data, and finding street segments which represent various
`driving habit patterns (3). Having classified many of the
`urban Los Angeles street segments by driving type, they
`undertook to develop a continuous road route which con
`tained segments such that the total route represented the Los
`Angeles morning trip to work. This was a tremedous task
`and involved developing and discarding many routes before
`the Los Angeles road route designated LA 4 was decided
`upon (Fig. 1). HEW participated in the latter stages of this
`development by providing technical staff and equipment for
`proportional exhaust sampling over the route (February 1966).
`In September 1966, HEW personnel used the PHS-Ethyl
`driving modes analyzer(4) to develop driving mode data
`from the LA 4 road route These data were analyzed and
`a similar route (CLA 4) was constructed in Cincinnati, Ohio.
`This route was the local basis for comparison with cycles
`being developed for replacement of the 7-mode cycle for
`certification testing. The candidate cycles were compared to
`the CLA 4 route by both proportional sampler emission tests
`and by driving modes analyzer tests To produce a driving
`cycle which met both the emission and the driving mode
`criteria, and which was of the same general format as the
`7-mode cycle, was a very difficult and time-consuming task
`This work was conducted through the summer of 1968, and
`the most likely short cycles were carefully compared to the
`CLA 4 road route.
`In August 1968, a large part of the mobile source activity
`was moved from Fairfax, Ohio, to Ypsilanti, Mich., in
`
`3 of 8
`
`FORD 1343
`
`
`
`Coast Laboratory and laboratory personnel were used as
`drivers. Six different drivers were used. Two of the drivers
`drove the route twice each because of equipment malfunctions
`during their first drives. The six complete tapes were brought
`back to Ypsilanti and used to generate strip chart recordings
`for visual comparison.
`In addition, idle time, average route
`speed, maximum speed, and number of stops per trip were
`calculated, and the average was determined for each measure
`ment. The total running time of the drives averaged 37.6 min,
`with a range of 35.3-40.0 mm. The actual time spent driving
`(eliminating idle periods) averaged 31 mm with a range of
`29.8-32.4 min. One of the six good traces exhibited obviously
`excessive throttle movement. The remaining five traces had
`remarkably similar traces. Since the five traces were so similar,
`the drive with the actual driving time closest to the average of
`the six runs was selected as the basis for the driving schedule.
`This trip (trip 2 of Table 1) is represented pictorially in Fig.
`2 which is a reproduction of the chart recording made from
`the original tape. This recording runs from right to left and
`from top to bottom. The beginning of the trip is at the up
`per right and the end of the trip is at the lower left. The
`speed trace is in the lower portion of the chart. The manifold
`pressure recording is located in the upper portion of the chart.
`The actual driving time for this drive was 31.2 min and the
`total running time was 37.7 min.(See Table 1.)
`The LA 4 route is 12.0 miles long, which is longer than the
`average trip length, 7.5 miles(5), so a short version of the
`LA 4 recording was needed. Several weeks were spent decid
`ing how the route could be shortened with the least amount
`of tampering. The LA 4 speed profile chart consists of a se
`ries of individual speed profiles connected by various length
`idle periods (Fig. 2). For identification purposes, these pro
`files were lettered A-Z plus AA and BB; H was further di
`vided into Ha and Hb. The characteristics-top speed, average
`speed, elapsed time, and distance-were calculated for each of
`the profiles and then the like profiles were grouped together.
`(See Table 2.) It was decided that, where groups of like pro
`files existed, one complete profile of every three would be
`deleted. Fig. 3 presents one example of like profiles. Profiles
`G, Ha, and E are compared, and profile Ha was deleted from
`the LA 4 short route. The freeway portion of the route con
`
`3 a
`
`nticipation of the construction of a new mobile source
`emissions laboratory in Ann Arbor. This move disrupted
`most of the mobile sources research and development
`activities. While some cycle development activities continued
`at the Fairfax facilities, the Procedures Development Branch
`personnel at Ypsilanti began to contribute as much as possible
`to the project.
`The equipment available during the transition included an
`automobile, a fifth wheel with tachometer-generator, a strip
`chart recorder, and a d-c to a-c inverter. The combination
`of these items allowed the recording of actual road speed
`during trips on the road. Since it was known that the Fair
`fax facility would ultimately be closed and that the CLA 4
`would become inaccessible, it was decided to investigate the
`possibility of developing an Ypsilanti version of the CLA 4,
`which was to be termed YCLA 4. It was realized, of course,
`that the genuine LA 4 route in downtown Los Angeles was
`the standard route, but it was felt that the experience which
`was acquired during the development of the YCLA 4 would
`aid the analysis of LA 4 data and the development of a rep
`resentative cycle.
`Engineers working on the project drove over several roads
`and streets in the Ypsilanti area and recorded route speed
`profiles with the same basic equipment that had been used
`five and more years earlier. They struggled with the same
`data reduction problems as before to get useful mode data
`from the long speed profile chart.
`It was noted, as was
`noted earlier, that the speed profile chart, taken in its en
`tirety, was an excellent description of the trip. If an average
`trip could be driven over the route, then the speed profile
`chart would display all of the useful characteristics of the
`route.
`
`TRANSFER OF SPEED PROFILE CHART TO
`DYNAMOMETER DRIVING
`
`Once it was realized that the average speed profile chart con
`tained not just the major modes of the route but also in
`cluded all of the little mode changes which characterize real
`driving, consideration was given to how this information
`might be transformed to dynamometer driving. Of course,
`the simplest, and most straightforward method would be to
`use the route speed profile chart as a driving aid and drive
`the car to match its speed to the previously recorded speed
`profile. This was tried and it was found that a driver could
`indeed follow the speed profile with reasonable precision and
`without greater difficulty or more fatigue than was associated
`with driving an equivalent number of 7-mode cycles
`The next step was to obtain a good speed profile of a typi
`cal drive over the LA 4 road route To accomplish this task,
`an engineer and an electronics technician were sent to Los
`Angeles with an instrumented car. The car used was a 1969
`2-door Chevrolet with a 327 CID engine and 2-V carburetor.
`It had a gvw of 4000 lb. The instrumentation consisted of a
`4-channel tape recorder, a fifth wheel with tachometer-gen
`erator, and transducers to measure engine speed and manifold
`pressure. This project was operated out of the EPA West
`
`4 of 8
`
`FORD 1343
`
`
`
`sisted of one long profile. About one-third of this profile was
`deleted. 1.125 miles were removed out of 3.065 miles of op
`eration (96.7 s out of 263.5 s) in such a way that the original
`average speed of the profile was maintained. Fig. 4 shows the
`changes which were made to the original freeway profile. Af
`ter each trial selection of profiles, the short schedule parame
`ters, such as average speed, were calculated and compared to
`the parameters for complete route. Several different combin
`ations of profiles were tried before the combination which
`was to become the short version of the LA 4 route was found.
`To aid in the selection of profiles for the shortened route, the
`actual number of idle periods was compared to the average
`of all of the road runs multiplied by the ratio of short segment
`distance to the total route distance. The expected number of
`idle periods was 17.6. The actual number of idle periods used
`in the shortened route, including an initial idle, was 18.
`It
`was also necessary to know how to apportion the 248 s of idle
`time among the 18 idle periods. A total of 167 idle periods
`were recorded during all of the trips. These idle periods were
`ranked according to duration, divided into 18 groups, and the
`average duration for each group was determined. The sum
`of the averages was 266 s, so each value was reduced slightly
`to give a total of 248 s. The resulting idle lengths were 0, 0,
`0, 2, 3, 5, 8, 10, 12, 14, 16, 18, 20, 21, 24, 26, 30, and 39 s.
`The 20 s value was selected for the initial idle period. The re
`mainder were distributed between the various driving segments
`
`4
`In
`using the actual idle times recorded on trip 2 as a guide.
`Fig. 2, the shaded areas represent the parts of the LA 4 road
`route which were retained in the EPA Urban Dynamometer
`Driving Schedule (UDDS). (See Table 3.) The route selected,
`and indeed each route tried, maintained the original sequence
`of profiles as they were recorded. The driving schedule, m
`this form, was presented as proposed rule making in the Fed
`eral Register (6). Comments received from vehicle manufac
`turers and from dynamometer manufacturers indicated that
`some of the low-speed acceleration rates exceeded the dy
`namometer design rate of 3.3 mph/s. This primarily occurred
`during deceleration. Where this occurred, the acceleration
`rate was cut back to 3.3 mph/s.
`
`COMPARISON OF EMISSIONS
`
`The final step in the development of the UDDS was to com
`pare emissions from vehicles driving the UDDS to the emissions
`from the same vehicles driving the full LA 4 driving schedule.
`Statistical analysis of the results indicated that the UDDS does
`represent the LA 4 route. (See Table 4, and Figs. 5-7).
`The 1372 data points which are listed in the Federal Re
`gister (6,7) were transcribed by hand from the original speed
`profile chart so that the UDDS could be accurately described,
`curately described.
`Scott Research Laboratories, Inc., as a part of their CAPE
`
`5 of 8
`
`FORD 1343
`
`
`
`5
`
`Table 2 — LA-4 Road Route Profnles
`
`Table 3 - LA 4-S3 Road Route Profiles
`
`Segment
`
`D1stance,
`m1les
`
`Tlme, s
`
`Average Speed,
`mph
`
`Selected
`
`Segment
`
`D1stance,
`mrles
`
`T1me, s
`
`Average Speed,
`mph
`
`Idle, s
`
`Freeway C selected
`C not selected
`G
`Ha
`E
`I
`P
`B
`A
`Q
`Y
`O
`S
`J
`T
`Hb
`M
`D
`F
`N
`K
`
`AA
`Z
`L
`
`R
`U
`X
`
`‘}:B
`V
`10141
`
`*WIth15%Id1<=
`
`1.941
`1 124
`0.460
`0.470
`0.365
`1.36
`1.374
`0.67
`0.375
`0.376
`0.30
`0.30
`0.22
`0 25
`0.28
`0.21
`0.25
`0.11
`0.14
`0.20
`0 17
`
`0-14
`0.13
`0 14
`
`0.07
`0 06
`0 21
`
`166.8
`96 7
`58.6
`60.5
`49 3
`193.6
`191.8
`103.9
`79.7
`63.0
`57.6
`68.5
`48.0
`51.6
`510
`39.5
`524
`19.5
`25 0
`38 5
`35.0
`
`33 6
`30.0
`40.5
`
`2L0
`“'6
`64 2
`
`41.89
`41.84
`28.26
`27.97
`26.65
`25.29
`25.79
`23 21
`16.94
`21.49
`18 75
`15.77
`16 50
`17.44
`19 76
`19.14
`17.18
`20 31
`20.16
`18.7
`17.59
`
`15~0
`15.6
`12.44
`
`110
`1331
`11.78
`
`3
`0:02
`11 855
`
`12 O
`1813-5
`
`1:
`6'00
`2000*
`
`v
`x
`v
`)1
`v
`x
`v
`v
`x
`v
`v
`x
`v
`v
`v
`v
`x
`x
`v
`v
`v
`
`X
`v
`v
`
`X
`V
`V
`
`1‘
`:
`
`From start
`B
`C
`E
`F
`G
`H
`.1
`K
`L
`N
`P
`Q
`S
`T
`U
`X
`Y
`Z
`
`0.67
`1.941
`0.365
`0.14
`0 46
`0 21
`0.25
`017
`0.14
`0.20
`1.374
`0.376
`0.22
`0.28
`0.06
`020
`0.30
`0.13
`7.486
`
`103.9
`166.8
`49.3
`25.0
`58.6
`39 5
`51.6
`35.0
`40.5
`38.5
`191 8
`63.0
`48.0
`51.0
`16.6
`48.0
`57.6
`300
`1114 7
`
`23.31
`41.89
`26.65
`20.16
`28 26
`19.14
`17.44
`17.49
`12.44
`18.7
`25.79
`21.49
`16.50
`19.76
`13.01
`11.78
`18.75
`15 6
`19.3
`
`20
`39
`14
`5
`18
`0
`21
`26
`12
`0
`0
`2
`30
`3
`16
`10
`8
`24
`248
`
`Total cycle txme = 114 7 + 248 = 1363 s = 22.71 m1n.
`
`
`Table 4 - Correlatnon to EPA Urban Dynamometer Dnvmg
`Schedule to Complete LA 4 Schedule
`
`Correlatron
`Coeffrcrent
`0.99466
`0.943014
`
`0 997157
`
`HC
`CO
`
`C02
`
`Slope
`0.9458
`1 09076
`
`Intercept
`0 070564
`—0.841045
`
`0.967583
`
`7.111
`
`Note Represents 19 data sets from f1ve veh1cles wrth mertra welghts
`of 2000, 3000, 3500, 4000, and 5000 lb
`
`
`Ho
`
`G
`
`E
`
`j_____.__.::__%:__
`
`Frg. 3 — Typlcal l1ke profiles from LA 4 speed profile
`
`HC
`
`EM(§:iAO)NS
`
`*—oELETED ‘‘
`
`y = 0 945ex 9 0 070564
`
`
`URBAN
`DRIVING
`SCHEDULE
`TESTS
`
`'
`
`,
`,
`1
`[
`
`I
`
`[
`
`Frg. 4 - Freeway prof1le from LA 4 speed profile
`
`1’1g. 5 - EPA Urban Dynamometer Dr1v1ng Schedule, HC emlssrons
`compared to LA 4 dynamometer HC emrssxons (hot start)
`
`6 of 8
`
`FORD 1343
`
`HC EMISSIONS (GPM)
`LA'4 DYNAMOMETER TESTS
`
`6 of 8
`
`FORD 1343
`
`
`
`6
`
`Many different approaches were tried. The present ap
`proach of measuring the speed-time relationship on a selected
`typical route and applying this relationship directly to a dy
`namometer driving schedule is straightforward and easily
`understood As long as the drive over the route is controlled
`by the prevailing traffic conditions, and not by limitations or
`capabilities of the test vehicle, the speed-time relationship is
`independent of an individual vehicle used for recording the
`trip The most important characteristics of the long, non
`repetitive driving schedule is that all of the nuances of driving
`characteristics, which are always discarded during the de
`velopment of short, repetitive cycles, are fully represented
`during the drive
`
`REFERENCES
`
`1. John N. Pattison and M. P. Sweeney, "A Study of Los
`Angeles Driving as it Relates to Peak Photochemical Smog
`Formation " Paper presented at APCA National Meeting, San
`
`Francisco,June 1966.
`2. D. M. Teague, "Los Angeles Traffic Pattern Survey."
`Published in "Vehicle Emissions, Part I," SAE Progress in
`Technology Series, Vol 6. New York Society of Automo
`tive Engineers, Inc., 1964, pp. 17-38, 44.
`3.G. C. Haas, M. P. Sweeney, and J. N. Pattison, "Lab
`oratory Simulation of Driving Conditions in the Los Angeles
`Area." Published in "Vehicle Emissions, Part II," SAE Pro
`gress in Technology Series, Vol. 12. New York: Society of
`Automotive Engineers, Inc., 1968, pp. 317-324.
`4.Ethyl Corp., "Survey of Truck and Bus Operating Modes
`in Several Cities." U.S. Public Health Service Contract No
`PH 86-62-12, Final Report, June 1963
`5.D. H. Kearm and R. L. Lamoureux, "A Survey of
`Average Driving Patterns in the Los Angeles Urban Area."
`TM-(L)-4l19/000/01, Feb. 28, 1969.
`
`10(8,9)vehicle operation's survey conducted a modal anal
`ysis of the LA 4 road route along with a corresponding modal
`analysis of the UDSS. Their analysis of the modal data
`shows a very good correspondence between the road route
`and the UDDS (10). The results of the Scott work show
`that the UDDS is a good representation of the Los Angeles
`road route LA 4.(See Table 5.)
`
`CONCLUSIONS
`
`The UDDS is the result of more than 10 years of ef
`fort by various groups to translate the Los Angeles smog
`producing driving conditions to dynamometer operations.
`
`7 of 8
`
`FORD 1343
`
`
`
`7
`6.Federal Register, Vol. 35, No. 136(July 15, 1970),
`pp. 11,357-11,359.
`7.Federal Register, Vol. 37, No. 221 (Nov. 15, 1972),
`pp. 24,317-24,318.
`8.Scott Research Laboratories, Inc.,"Vehicle Operations
`
`Survey." CRC-APRAC Project No. CAPE 10-68 (1-70), Dec.
`17, 1971.
`
`9. M. Smith and M. Manos, "Vehicle Operations Survey:
`Interim Report, LA-3 Road Route Study." SRL 2922-09
`1071, Oct. 29, 1971.
`10. Scott Research Laboratories, "Construction of Chassis
`Dynamometer Test Cycles." Interim Report SRL 2948
`050671, June 25, 1971.
`
`This paper is subject to revision Statements and opinions
`advanced in papers or discussion are the author's and are
`his responsibility, not the Society's, however, the paper has
`
`been edited by SAE for uniform styling and format. Discussion will be printed
`with the paper if it is published in SAE Transactions For permission to publish
`this paper in full or in part, contact the SAE Publications Division and the
`authors.
`
`8 page booklet
`
`Printed in U.S.A.
`
`8 of 8
`
`FORD 1343