EPA=AA-SDSB-80-^06
                         Technical.Report
              Quantitative Effects  of Acceleration Rate
                        on Fuel Consumption
                                 by


                            Randy Jones


                            April 1980
                              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 deve-
lopments which may form the basis  for a final EPA decision, posi-
tion 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

-------
Introduction

     One  factor  which has  a significant effect  on vehicle  fuel
consumption is  the rate at  which the  vehicle  is  accelerated.
Quantitative and  qualitative  studies  have  shown  that  more rapid
and/or more  frequent  accelerations  result in increased fuel  con-
sumption.JL/  The EPA has conducted a study to quantify  the  effects
of operating a vehicle at different acceleration rates.

     The "test  involved accelerating  the vehicle  at  a constant
acceleration rate to a speed of  55 mph, and  maintaining the 55 mph
speed  until  total  distance  traveled  equaled one  mile.    Concep-
tually,  this may be viewed  as entering a freeway system with
different  acceleration  rates.   The  acceleration  rates varied in
increments of one from 1  to 5 mph/sec.   Fuel  consumption was
measured with a Fluidyne flow  metering system.

Discussion/Procedure

     The  test  was conducted  on a Clayton twin roll dynamometer.
The  dynamometer  rolls were coupled with a motorcycle chain to
prevent  tire  slippage at  the  higher acceleration rates.  The
vehicle was a  1979  Nova, 250  CID,  1  bbl, with an  automatic trans-
mission.

     The vehicle was first warmed  to stable operating conditions by
being  driven  over 2  HFET  driving cycles.   The  fuel  consumption
tests  were conducted starting the flow meter with  the  engine
running, allowing five seconds of  idle and then accelerating at the
desired constant  rate  until the vehicle  reached 55 mph.  A 55 mph
cruise was  then maintained until  a  distance  of one mile was
travelled from the initial start.

     The  acceleration rate  and  cruise  speeds were  accurately
followed by using a "drivers aid" strip chart on which  the  desired
acceleration ramps and cruise speed had been previously drawn.  At
the  higher acceleration rates,  if  the vehicle could not match the
acceleration trace, the accelerator pedal was depressed  fully until
the  55 mph cruise mode was reached.

     Five  repeat  tests  were  conducted  for each  acceleration rate.
The  order  of  the  tests were  randomized to minimize  any systematic
fuel consumption effects  which  might have occurred  because of
increasing tire or lubricant temperatures.  An example  strip chart
recording  for  a  1.0 mph/sec.  acceleration trial is shown  in  Fig-
ure  1.

     The  data  from the 20  test  trials were corrected  for  actual
distance travelled and fuel  temperature to yield a  fuel  consumption
figure  in  terms  of cc/tnile.   The  distance  travelled was based on
the  dynamometer roll  revolutions  recorded  during  each test.
An SAE  fuel  temperature correction for Group  3  test  fuel was  used
to correct  fuel  volume  measurements.2/   All  data  are tabulated in
.Appendix A.

-------
                                      Figure I
  ro
  tn
   o-
   to
   CO

 m
   ST
   Ul
   o
   t/7
   O
    O
START :
                               Speed Time Trace For A "

                               1.0 mph/sec Acceleration Trial

                                    !           I          t
                                      istance =
1.0 Mile
                                                                 ST(b


                                                          fe_ -I
                                 SPEED  (mph)

-------
Results

     Fuel  consumption  increased  approximately  linearly  with  in-
creasing acceleration rate.  The results are summarized in Table 1
and plotted in Figure II.   A linear  regression line fit to the data
yielded  a  correlation  coefficient of  .93,  and,  as  illustrated in
Figure  2,  nearly passes through  the  standard deviation  limits of
all the acceleration trials.

     Fuel  consumption  increased by  10.4  percent  between acceler-
ation rates of  1  mph/sec  and 4 mph/sec.  An apparent exception to
the trend  of  increased  fuel  consuption with increased acceleration
rate occurred when vehicle acceleration rate changed from  4 mph/sec
to  5  mph/sec.   Here,  fuel consumption decreased.   This apparent
decrease  in  fuel consumption is  most likely  explained  by  the
difficulty  the vehicle  encountered  in  achieving these  constant
acceleration rates.   The  acceleration rates of  3.0,  4.0, and  5.0
mph/sec exceeded the maximum vehicle acceleration capability at the
higher speeds of  30  to  45 mph.  The vehicle speed-time traces were
particularly  similar for  the  acceleration rates of 4.0  and  5.0
mph/sec.   Driver  comments indicated that the accelerator pedal was
fully  depressed  for approximately  80 percent  of the 4.0 mph/sec
acceleration and virtually 100  percent of the 5.0 mph/sec  ac-
celeration.   Therefore, the  actual  vehicle accelerations  for these
two  test  acceleration modes differed very little.   The standard
deviation  values  of fuel  consumption for the 4.0 and 5.0 mph/sec
trials  overlapped,  indicating  little significance  in a difference
in  fuel  consumption between the  tests  for  these two acceleration
modes.

     It  is interesting to  note the greatest  increase  in  fuel
consumption for any  1 mph/sec increment occurred when  the  accelera-
tion  rate changed  from  3.0 to  4.0 mph/sec.   The   increase  was
probably caused  by an anonmalous delay in  the  speed at which the
transmission  1st  -  2nd gear shift  occurred when the acceleration
rate changed  from 3.0 to 4.0 mph/sec.  This  transmission effect was
discussed  in an  earlier  EPA technical report  involving  the same
vehicle.jl/   The  report  predicted  a  significant  increase in fuel
consumption when  the vehicle acceleration  rate exceeded the maximum
acceleration rate on EPA test cycles, 3.3  mph/sec.

     A model to calculate the energy demand  on the vehicle for each
different  acceleration  trial was derived,  and  actual fuel energy
expended  was compared  to the theoretical energy  demand.   The
derivation,  calculations,  and  results  are  discussed  in detail  in
Attachment I.  The results are  summarized  in Table 2.

     The energy  model  indicates that  most of  the increase in fuel
consumption with  increased acceleration rates occurs  because of the
increased  energy demand.   For  acceleration rates between 1 and  3
mph/sec, the  energy demand  increased  by about 6  percent as did the
fuel  consumption.   Only  when  the acceleration  exceeded  3 raph/sec
did  the fuel  consumption  increase  more  rapidly than  the energy
demand.                  .     .

-------
Table 1
Rate of
Acceleration
(mph/sec)
1.0
2.0
3.0
4.0
5.0
No.
of Trials
4
4
4
4
4
Average Fuel
Consumption
(cc/mile)
259.4
270.9
274.1
286.4
282.4
Standard
Deviation
(cc/mile)
4.83
1.39
2.87
6.88
6.79

-------
                                  Figure II



                   Fuel Consumption vs. Acceleration Rate
  290 _.
6 280
o
u
c
o
1-1
o-i


§270

tn
c
o
o
3




  260
  250
H	h-	H


 2.0         3.0         4.0

 Acceleration  Rate (mph/sec)
                  1.0
5.0

-------
                              Table  2
Trial
Acceleration
Rate
1 mph/sec
2 mph/sec
3 tnph/sec
4 mph/sec
Theoretical
Energy
Demand
1.479 x 106J
1.550 x loAj
1.573 x 106J
1.582 x 106J
Fuel
Energy
Expended
8.307 x 106J
8.657 x 10& J
8.778 x 106J
9.171 x 106J
% Eff.
Energy Demand
Energy Expended
17.8
17.9
17.9
*
17.3
Note:  The theoretical energy demand  assumes  the vehicle accurately
followed the acceleration-cruise  trace.   This was reasonably  true
for acceleration rates up to 4 mph/sec.   The  5 mph/sec  acceleration
significantly exceeded the  vehicle acceleration capability,  there-
fore the  theoretical  energy demand  for  this acceleration was  not
included in the table.

-------
     This may  also  be expressed  in  terms of  a simple  energy
efficiency  ratio  (theoretical  energy demand/fuel energy con-
sumed).   The efficiency  remained  almost constant  for  all tests
except for a notable  decrease at the 4.0 mph/sec acceleration.  The
decrease  may  be explained  by the  transmission  effect  discussed
earlier.

     The increase in  energy demand with an increase in acceleration
rate is shown in the model  to  be a  function of the  increased  time
the vehicle  is  operated at the  55  mph cruise speed.   Since the
aerodynamic  drag force,  simulated by  the dyno power absorber,  is
proportional to  the square of vehicle speed,  more  power is absorbed
when  the vehicle  is accelerated  to the cruise speed  quickly.

Conclusions

     The test vehicle for this experiment  exhibited an increase  in
fuel consumption  with  an increase in  acceleration rate.  Fuel
consumption increased by  10.4 percent  when  the acceleration  rate
increased from  1.0  mph/sec to 4.0 mph/sec.

     Computation of  the  energy demand  indicates  that most  of the
effect  occurs because  the increased  acceleration rates  result  in
more vehicle operation  at  the cruise speed inducing more dyno power
absorber work on the  vehicle.

     However, the maximum fuel consumption effect  occured when the
acceleration rate  changed  from 3.0  to  4.0  mph/sec.   This  result
probably occurred because of an anomaly in transmission shift
characteristics  of  this  vehicle when the acceleration rate exceeded
the maximum acceleration rates  of the EPA test cycles.

-------
                            References

lj   "Passenger Car Fuel Economy:  EPA  and  Road  - A Report to Che
     Congress," Draft, U.S. Environmental Protection Agency, April
     1980.

2/   "Fuel Economy Measurement  Road Test Procedure - Cold Start and
~~    Warm-Up  Fuel  Economy,"  SAE J1256,  Society  of Automotive
     Engineers, Warrendale, PA, May 1979.

3/   "The  Effect  of  Acceleration  Rate on  Automatic  Transmission
     Shift-Speeds  for Two  1979 Novas," R.  Jones, EPA  Technical
     Report, January 1980.

-------
                            Appendix A




                       Fuel Consumption Data




Vehicle:  1979 Nova, silver




Date:  3/1/80




Dyno Cell:  D207, rolls coupled, d = 8.65 in.




Barometric Pressure:  29.48 in. Hg




Temperature:  WB = 62.5   DB = 74.0




Inertia Weight:  3750 Ib.   AHP = 12.9   IHP




Test Tires:  Bridgestone bias .






                               Data
= 10.4

Trial
Acceleration
Rate
(mph/sec)
5
1
2
5
3
4
1
4
5
4
3
1
3
2
3
2
4
2
1
5



Roll
Revolutions
2325
2348
2220
2291
2356
2239
2348
2319
2302
2322
2334
2360
2341
2216
2349
2238
2333
2239
2373
2350


Fuel
Consumed
(cc)
266.7
251.0
252.4
271.2
272.6
265.3
260.3
280.7
278.9
284.5
268.3
256.7
265.3
250.8
265.8
252.6
270.0
251.5
252.1
274.1


Fuel
Temperature
(°c)
40.0
40.0
41.0
41.0
41.0
41.0
41.0
41.0
42.0
41.0
41.0
41.0
41.0
41.0
42.0
41.0
42.0
41.0
42.0
42.0
Fuel Consumption
Corrected for Fuel
Temperature and
Distance
(cc/mile)
274.8
256. f
272.5
283.7
277.3
284.0
265.7
290.1
291.0
293.6
275.8
260.7
271.6
271.3
271.8
270.5
278.0
269.2
255.2
280.2

-------
        Table 1-A




Fuel Consumption (cc/mile)
Trial
1
2
3
4
"x
s
% s/3c
1.0
mph/sec
256.1
265.7
260.7
255.2
259.4
4.83
1.86%
2.0
mph/sec
272.5
271.3
270.5
268.2
270.9
1.39 _.
.50%
3.0
mph/sec
277.3
275.8
271.6
271.8
274.1 "
2.87
1.05%
4.0
mph/sec
284.0
290.1
293.6
278.0
286.4
6.88
2.4%
5.0
mph/sec
274.8
283.7
291.0
280.2
282.4
6.79
2.4%

-------
                           Attachment I

               Calculation of Vehicle Energy Demand

     The total energy demanded from the vehicle over any  cycle  is
the time integral of the instantaneous power requirement:

         Tf

     E =J  Pdt                                                   (1)



Where:

     E = the total energy demand,

     P = the instantaneous power requirement,

     T£ = the initial time of the beginning of the cycle,

     Tf = the final time at the end of the cycle.

     The power required can, of course, be expressed as the  product
of the  instantaneous force  times  the  velocity  of  the vehicle.   For
the  simple  cycles  of  this  project;  that  is, ramp  accelerations
followed by a steady speed cruise, the forces acting on the  vehicle
may be expressed as:

     F = m ^- + f0 + f2V2                                         (2)

Physically the  first  term represents  the inertial effect,  the
second  term primarily  represents  tire losses,  while the  third  term
represents  the  aerodynamic  drag  or  dynamometer  power  absorption.
It  should be noted that  equation (2) does  not  contain any  term
representing  energy dissipated  in the vehicle brakes;  therefore,
this  equation is only  applicable to  cycles  in which the  vehicle
brakes are not used.

     Combining equations (1) and  (2)  and integrating:
     E =  /   Fvdt
       = J  (m |£ + f0 + f2v2)vdt

-------
        Tf               Tf           Tf

        / (m -g£ vdt)  +  / f0vdt  +   / f2v3dt
                         T
         V(Tf)          Tf          Tf

        m /   vdv  +  fo/vdt  +  f2/v3dt

         V(Ti)          T£          Ti
               V(Tf)       Tf          Tf
     =  1/2 mv2
               V(Ti)
                +  f o / vdt  +  f2/v3dt                     (3)
     The simple  cycles of  this  report start with  the vehicle  at
rest; that  is:

     V(Ti)  =0                                                    (4)

Using,

     V(Tf)  = Vf                                                   (5)

The first term of (3) becomes:

            V(Tf)

     1/2 mv2       =  1/2 mVf2                                    (6)

            V(Ti)
     This term is, as expected the kinetic energy of the vehicle  at
the final steady speed cruise.

     The  second  term,  the  time  integral  of the  velocity is,  by
definition the distance traveled:
/
       vdt = D                                                    (7)
     Therefore, using equations (6) and (7) in equation (3):

-------
E =  1/2 mVf2  +  f0D  +  f2 /*
                                    3dt                            (8)
For these  cycles,  or in fact  for  any cycles in which  the vehicle
brakes are  not  used, only the  aerodynamic  drag term dependent on
the detailed  velocity versus  time characteristics  of  the cycle.

     Because of  the simple characteristics  of  the cycles used in
this  program,  even  the  aerodynamic  term  of equation  (8)  can be
integrated  in closed form.   First,  the  integral  can be  separated
into  two  components, the acceleration segment and the  cruise  seg-
ment .

     Tf         Tc         Tf                                    •

      fv3dt  =   fv3dt  +  fv3dt                                 (9)

     T£         T£         Tc


Where:

     Tc = the time at which- the cruise segment is  initiated.

Since the acceleration is constant:

     Vc = aTc

or,                                                             (10)

     Tc - Vc/a                     -

Where:

     Vc = the constant cruise velocity

     In terms of a and Vc equation (9) becomes:
     Tf         Tc            Tf

      I v3dt  =  T(at)3dt  +   / Vc3dt

     T          T         .    T
             Tc          Tf
        -4*
                         Tc

-------
     Choosing to call the initial  time T£ to be zero:


     Tf

     J v3dt  =  l/4a3Tc4  +  Vc3(Tf  - Tc)
                                                           (12)
     For  the particular cycles under  investigation the  final
velocity is the cruise velocity.
Vf
Therefore:
fo
fo
«0
fo
D H
D ^
D M
D H
^ f2 [l/4a3Tc4 +
- f2[l/4 VC3TC +
H f2Vc3(l/4 Tc +
K f2vc3^Tf ~ 3/4
vc
vc
Tf
Tc
3(Tf -
3Tf
-Tc]
]
Tc)l
VC3TC]
                                                                (13)
     E = 1/2 mVc2 + f0D + f2 [l/4a3Tr;4 + Vr3(Tf - Te)]            (14)

       = 1/2 mVc2

       = 1/2 mVc2

       =1/2 mVc2 + f0D

     The time at which  cruise  condition  is reached Tc is given by
equation (10).   The condition  for  Tf is  that  the total distance
traveled remain  constant for  the-different accelerations.   That
is:
    -•C       T

D =  / atdt + |vc

    T-       T •
    Ll       *•!
                      dt
                                                                 (15)
       = 1/2 at2
                  Vrt
Calling the initial time T^ = 0,

     D = 1/2 aTc2  +  Vc(Tf - Tc)

Using equation (10) and solving for Tf,
                                                            (16)
     D = l/2a -4   +  VcTf - VcVc/a
               *-

-------
                       V 2
         -  VcTf - 1/2 -|

                    V 2                                          (17)
     V-Tf = D + 1/2 —
                     a


          = 2aD + vc2
            2a
     Tf - 2aD + Vc2
             2aV
     Equations  (14,  (17),  and  (10)  provide the complete algrebraic
expression  for  the energy demand  for  the vehicle over  the  simple
cycles used in this study.

Dynamometer Representation

     In the  case of dynamometer  tests  it is  reasonable  to  assume
that the power  absorbed  by the PAU is  proportional to the velocity
cubed, and that the remainder of the power absorbed by the vehicle-
dynamometer system occurs in the tires and dynamometer bearings and
is linear in velocity.

     The  indicated  dynamometer power  absorption at  50 mph  may,
therefore, be equated to:

     IHP(V) = f2V3
                                                                 (18)

     f  =   IHP(50)
      2
          (50 mi/hr)3
     The  vehicle-dynamometer  coastdown is  a  measure of  the  total
dissipation forces  acting  on  the system and may be  used  to deter-
mine f():
     fO
                   At
            AW
            ~

-------
gxample Calculation

     In  this  program, the  following  parameters were used  for  all
dynamometer tests:

     IHP = 10.4

     I = 3750 Ibs. =  1705 kg

     At = 12.42 (vehicle-dynamometer coastdown time)

     Vc = 55 mi/hr =  24.59 m/sec

Consequently from equation  (18):

     f  =    10.4 hp                                             (20)
      2 " (50 ni/hr)3


     _  10.4 hp hr3   ,     mi     .3 ,3600 secv3
       125 x 103 mi3  1.609 x 103m       hr
     = 0.695
Computing fo from equation (19),
Therefore:
     f0 = 613 nt - f2V2
                                 2
          = 613 nt - 0.695 nt      (22.35
                               m          sec

          = 613nt - 347

          = 266 nt                                               (22)

     Considering  the  specific  case  of  an  acceleration of  1  mi/
hr-sec  the  time  to reach  the  55 mi/hr cruise  speed is  given by
equation (10) .

-------
     Vc = 55 mi/hr = 24.59 ra/sec

     a = 1 mi/hr-sec = 0.447 mi/sec'2

     Tc = Vc/a

          = 24.59 m/sec                                          (23)
             0.447 m/sec2

          = 55 sec

     The  final  time at  the completion of the  cruise  is,  from
equation (17):

        = 2(0.447 m/sec2)(!,.609'm)  +  (24.59 m/sec)2
      f         2(0.447 m/sec2)(24.59 m/sec)

     T  = 1438 m2/sec2 + 604 m2/sec2
      f            22 m/sec3

        = 93 sec

     The total energy may now be  calculated using  equation  (14) and
the  values  for  fo,  f2»  Tc,  and Tf  from  the  previous equations:

     E = 1/2 mVc2 + f0D + f2Vc3 (Tf - 3/4 Tc)


       ." 1/2(1705 kg)(24'6 m)  +  (266nt)(1,609m)
                       S6C
          + 0.695 nt ISfl (liiljE)3 (93 - 3/4 55)sec
                      m2    sec
       = 0.516 x 106 nt m + 0.428 x 106 nt
                                           m
          + o.695 nt      (i      (51.75)sec
                      m2    sec

       = (0.516 + 0.428 + 0.535) x 106 J

       = 1.479 x 106 J

     The total energy for the remaining four acceleration rates  are
shown below:

     2 mph/sec  E = (0.516 + 0.428 + 0.606)  x 106  J

                E = 1.55 x 106 J

     3 mph/sec  E = (0.516 + 0.428 + 0.629)  x 106  J

                E = 1.573 x 106 J

-------
     4 mph/sec   E =  (0.516 + 0.428 + 0.638) x 106 J

                E =  1.582 x 106 J

     5 mph/sec   E =  (0.516 +0.428 + 0.649) x 106 J

                E =  1.593 x 106 J

     Comparison to Test  Results

     The actual fuel energy expended in this experiment is  compared
to the theoretical energy demand  in Table  I and  Figure I.  A
fuel-energy efficiency value is also  shown  in Table  I.   The  cal-
culations assume a value of 32023 J/cc of fuel.I/

     A linear  regression  line fit to  the data  yielded a 0.926
correlation coefficient  indicating  a  linear relation between
theoretical energy demand and fuel energy consumed.  The efficiency
ratio of theoretical energy demand  to  fuel energy expended remains
nearly constant, around 17.7 percent.   This also  indicates a  good
correlation between  the two  parameters.   However,  the efficiency-
value  for  the highest  acceleration  rate, 5.0 mph/sec, is mis-
leadingly high.   Since  the higher acceleration  rates exceeded  the
vehicle  acceleration capability,  the  actual vehicle work  was
significantly less than the theoretical  energy demand.  The energy
efficiency  for the 5.0 mph/sec  acceleration  mode was omitted  since
the actual work, as  discussed earlier, was  nearly  the same as  the
4.0 mph/sec acceleration trials.

     It  is  interesting  to  note from the  theoretical  energy model
that the increased  energy demand at the  higher  acceleration  rates
are not  a  result of  the increased  acceleration  rates themselves.
Mathematically  the increase in energy with an  increase in  ac-
celeration rate results solely  from a  change in  the f2V^, or
aerodynamic drag  term,  while  the  .inertial  and  rolling resistance
terms remain constant.  The increased  energy demand is a result of
the vehicle  reaching a cruising speed sooner and  maintaining  the
speed  for  a  longer  period  of time.  Thus, the maximum aerodynamic
drag  force simulated by  the dynamometer power  absorber, which
increases with  the square of vehicle speed, acts on the vehicle for
a longer period of time when the vehicle is accelerated to a cruise
speed quickly.

     The model  is a physical approximation  of  the energy  required
to operate  this vehicle over the given  cycles.   The actual  fuel
consumption correlates very well to  the theoretical energy demand.
However, the model  does not treat  all  vehicle factors which  may
influence  fuel  consumption  and which  might change  with  changing
acceleration rates.    Examples include  automatic transmission  shift
speed and peak combustion temperature.
JY   D.E. Foringer, "Gasoline Factors Affecting Fuel Economy," SAE
     Paper No. 650427.

-------
                                                      Figure A-I
    9.20  -L
->   9.00

x
•o
oi
E  .

w
c


<->  8.80

s*.
oC
m
0)

w
   8.60 4-
(9

3
   8.40 JL
   8.20
                    1.46
1.48
                                              Theoretical Energy Consumed  Vs

                                                 Actual Fuel Energy  Consumed
1.50        1.52        1V54  .       1.56



    .  Theoretical Energy Consumed  (x 10 )
1.58
1.60

-------
        Table I
Theoretical
Average
Trial
Acceleration Rate
1 mph/sec
2 mph/sec
3 mph/sec
4 mph/sec
Energy Demand for
a 1.0 Mile Trial
1.479 x 106 J
1.550 x 106 J
1.573 x 106 J
1.582 x 106 J
Fuel Energy Expended
for a 1.0 Mile Trial
8.307 x 106 J
8.675 x 106 J
8.778 x 106 J
9.171 x 106 J.
Percent
Efficiency
17.8
17.9 "
17.9
17.3

-------