2

Vehicle Energy and Fuel Consumption

Basic Concepts
This chapter contains three main sections: First the dynamics of the longitudinal motion of a road
vehicle are analyzed. This part contains a discussion of the main energy-consuming effects
occurring in the "vehicle-to-miles" part and some elementary models that describe the longitudinal
dynamics and, hence, the drivability of the vehicle.
The influence of the driving pattern on the fuel consumption is analyzed in the second
section. The main result of this analysis is an approximation of the mechanical energy
required to make a road vehicle follow a given driving cycle. The sensitivity of the energy
consumption to various vehicle parameters or the potential for the recuperation of kinetic
energy when braking is derived from that result.
The third section briefly introduces the most important approaches used to predict the fuel
economy of road vehicles, the main optimization problems that are relevant in this context, and the
software tools available for the solution of these problems.
2.1 Vehicle Energy Losses and Performance Analysis
2. /. / Ener gy Los s es I nt r oduc t i on
The propulsion system produces mechanical energy that is assumed to be momentarily stored in
the Vehicle. The driving resistances are assumed to drain energy from this reservoir. This
separation might seem somewhat awkward at first glance. However, it is rather useful When one
has to distinguish between the individual effects taking place.
The energy in the Vehicle is stored:
In the form of kinetic energy When the Vehicle is accelerated; and
In the form of potential energy When the Vehicle reaches higher altitudes.
In the form of potential energy when the vehicle reaches higher altitudes
The amount of mechanical energy "consumed" by a Vehicle
1
When driving a pre-specified
driving pattern depends on three effects:
The aerodynamic friction losses;
The rolling friction losses; and
The energy dissipated in the brakes.
The elementary equation that describes the longitudinal dynamics of a road Vehicle has the
following form
M2 dtv(t) = F. (t) (F. (t) + F* (t) + Fs (t) + Fd (t)) (2.1)
Where Fa, is the aerodynamic friction, Fr the rolling friction, Fg the force caused by gravity
When driving on non-horizontal roads, and Fd the disturbance force that summarizes all
other not Yet specified effects. The traction force F. Is the force generated by the prime
mover minus the force that is Used to accelerate the rotating parts inside the Vehicle and
minus all friction losses in the powertrain. Figure 2.1 shows a schematic representation of
this relationship. The following sections contain more information about all of these forces.
Fig. 2.1. Schematic representation of the forces acting on a vehicle in motion.
Aerodynami c Fri c t i on Loss es
The aerodynamic resistance Fa, acting on a Vehicle in motion is caused on one hand by the Vis cous
friction of the surrounding air on the Vehicle surface. On the other hand, the losses are caused by
the pressure difference between the front and the rear of the Vehicle, generated by a separation of
the air flow. For
1
To be thermodynamically correct, the first part of this sentence should read: The amount of
exergy transformed to useless anergy ...
2
This force can be negative, for instance during braking phases.

Idealized Vehicle shapes, the calculation of an approximate pressure field and the resulting force
is possible With the aid of numerical methods. A detailed analysis of particular effects (engine
Ventilation, turbulence in the Wheel housings, cross-Wind sensitivity, etc.) Is only possible With
specific measurements in a Wind tunnel.
For a standard passenger car, the car body causes approximately 65% of the aerodynamic
resistance. The rest is due to the Wheel housings (20%), the exterior mirrors, eave gutters,
window housings, antennas, etc. (approximately 10%), and the engine Ventilation (approximately
5%) [100].
Usually, the aerodynamic resistance force is approximated by simplifying the Vehicle to be a
prismatic body With a frontal area Af. The force caused by the stagnation pressure is multiplied by
an aerodynamic drag coefficient cd that models the actual flow conditions
1
FQ(v) = 2 'PQ 'Af cd v
2
. (2.2)
Here, v is the Vehicle speed and Pa the density of the ambient air. The parameter cd must be
estimated Using CFD programs or experiments in Wind tunnels.
Rol l i ng Fri ct i on Losses
The rolling friction force is often modeled as
FT(v,pt, ... ) = ct(v, pt, ... ) m g cos(a), v > 0 , (2.3)
Where mv is the Vehicle mass and g the acceleration due to gravity. The term cos(a) models the
influence of a non-horizontal road. However, the situation in Which the angle a Will have a
substantial influence is not often encountered in practice.
The rolling friction coefficient cr depends on many Variables. The most important
influencing quantities are Vehicle speed v, tire pressure pt, and road surface conditions. The
influence of the tire pressure is approximately proportional to 1/,/pt. A Wet road can
increase ct by 20% and driving in extreme conditions (sand instead of concrete) can easily
double that value. The Vehicle speed has a small influence at lower values, but its influence
substantially increases When it approaches a critical value Where resonance phenomena
start.
A typical example of these relationships is shown in Fig. 2.2. Note that the tires reach their thermal
equilibrium only after a relatively long period (a few ten minutes). Figure 2.2 includes examples of
typical equilibrium temperatures for three speed values. For many applications, in particular When
the Vehicle speed remains moderate, the rolling friction coefficient ct may be assumed to be constant.
This simplification Will be adopted in the rest


Fig. 2.2. Tire friction coefficient as a function of vehicle speed.
Uphill Driving Force
The force induced by gravity When driving on a non-horizontal road is conservative and
considerably influences the Vehicle behavior. In this text this force Will be modeled by the
relationship
Fg(a) = m g sin(a) , (2.4)
Which, for small inclinations a, may be approximated by
Fg(a) ,: m g a (2.5)
(angle a in radians).
Inertial Forces
The inertia of the Vehicle and of all rotating parts inside the Vehicle causes fictitious (d'Alembert)
forces. The inertia force induced by the Vehicle mass is included in (2.1) by the term on the left
side. The inertia of the rotating masses of the powertrain can be taken into account in the
respective submodels. Nevertheless, sometimes for rapid calculation, it may be convenient to add
the inertia of the rotating masses to the Vehicle mass. Such an analysis usually considers a prime
mover and a transmission With a total transmission ratio y The total
3
inertia torque
Of the Wheels is given by

0 60 v m/s
3
The inertia O A includes all wheels and all rotating parts that are present on the wheel side of
the gear box. The speed of all wheels -A is assumed to be the

"
T!,3(F) = E)3 83(F) (2.6)
"F
And it acts on the Vehicle as an additional inertia force F!,3 = T!,3EA3, Where A3 is the Wheel
radius. Usually, the Wheel slip is not considered in a first approximation, i.e., J = A3 83. In this
case
E)3
!3(F) = _____
A3
" J(F) ; (2.7)
"F
Consequently, the contribution of the Wheels to the Vehicle overall inertia is given by the
following term
3 * , 3 =
E)3
A2
3
; (2. 8)
Similarly, the inertia torque of the engine is given by
"
"
T!,e(F)= E)e . "f8e(F) =E)e . "F (). 83(F)) =E)e w "FJ(F)!(2.9)
Where E)e is the total moment of inertia of the powertrain
4
and 8! Its rotational speed. Again,
assuming no Wheel slip, this torque is transferred to the Wheels as a force
F!~e(F) = ) . T-.,e(F) = E)e r2 " dt J(F) ; (2.10)
A3 W
Note that this expression is only Valid if the gear box has an efficiency of 100%. Since the
powertrain inertia is added to the larger Vehicle inertia, the errors caused by that simplification
are usually small and often may be neglected.
Assuming a constant gear ratio ), the force (2.10) corresponds to an additional Vehicle mass of
3*~~ =
)2
__ E),2 e ;
__ (2.11)
A3
In summary, the equivalent mass of the rotating parts is approximated as follows
3* = 3*, 3 + 3*, e =
1 2
2 E ) 3 + 2 E ) e . ( 2 . 1 2 )
A3 A3
And it should be added to the Vehicle mass 32 in (2.1). The total gear ratio )EA3 appears
quadratically in this expression. A ccordingly, for high gear ratios (lowest gears in a standard
manual transmission), the influence of the rotating parts on the Vehicle dynamics can be
substantial and may, in general, not be omitted.
4
The inertia Oe includes the engine inertia and the inertia of all rotating parts that are present on
the engine side of the gear box.

2./.2 Performance and Drivability General Remarks
Performance and drivability are very important factors that, unfortunately, are not easy to
precisely define and measure. For passenger cars, three main quantifiers are often Used:
Top speed;
Maximum grade at Whi ch the fully loaded Vehicle reaches the legal top-speed limit; and
Acceleration time from standstill to a reference speed (100 km/h or 60 miles/h are often
Used).
These three quantifiers are discussed below.
Top Speed Performance
For passenger cars, the top speed is often not relevant because that limit is substantially higher
than most legal speed limits. However, in some regions and for specific types of cars that
information is still provided. This limit is mainly determined by the available power and the
aerodynamic resistance.
5
Neglecting all other losses, the maximum speed is obtained by solving
the following power balance
1
Pmax ~ 2 . Pa A f ' Cd ' v, n,ax , (2.13)
Where Pmax is the maximum traction power available at the Wheels. The relevant information
contained in this equation is the fact that the power demand depends on the cube of the
Vehicle speed. In other Words, the engine power must be doubled in order to increase the
top speed by 25%.
Uphill Driving
For trucks and any other Vehicles that carry large loads, a relevant performance metricis the
Uphill driving capability. The relationship between Pmax and the maximum gradient angle amax is
obtained by neglecting in (2.1) all the resistance forces but F9
Pmax , Tnv ' vmin ' g ' sin(amax) , (2.14)
Where vmin is the desired Uphill speed. For a given rated power and a minimum speed, this
equation Yields the maximum Uphill driving angle. This dis cussion is not complete without an
analysis of the influence of the gear box. However, that analysis deserves to be treated in some
detail and this discussion is thus deferred to Chap. 3.
5
At that speed, the power consumed to overcome rolling friction is typically one order of magnitude
smaller than the power dissipated by the aerodynamic friction.

Acceleration Performance
The most important drivability quantifier is the acceleration performance. Several metrics
are Used to quantify that parameter. In this text the time necessary to accelerate the Vehicle
from 0 to 100 km/h is taken as the only relevant information. This value is not easy to
compute exactly because it depends on many Uncertain factors and includes highly dynamic
effects. An approximation of this parameter can be obtained as shown below.
The energy required to accelerate the Vehicle from standstill to any Velo city Jo is given by
~~ = 2 32 J
1 ~ ; (2.15)
If all the resistance forces in (2.1) are neglected, the energy Eo is also the energy that has to be
provided by the powertrain to accomplish the acceleration required. Accordingly, the mean
power P that has to be provided by the engine is P = eoefo, Where Fo is the time available for the
acceleration. An approximation, Which takes into account the varying engine speed and the
neglected losses, is obtained by choosing P ,: P!,,,4E2,
Where
P!,,,4 is the maximum rated power.
Consequently, a simple relation between the acceleration time and the maximum power of the
engine is given by the following expression
Fo
J
2
0 32
P!A4 ; (2.16)
<
e s t i ma t e
d s
O
1 4
1 o
6
2

o 4 8 1 2 1 6
< p u b l i s h e d s
O
Fig. 2.3. Comparison between acceleration times as published by the manufacturers (crosses) and
values calculated with (2.16) for vo - 100 km/h (solid line).

Figure 2.3 shows the comparison between the acceleration times as published by various
manufacturers and as estimated Using (2.16) for vo = 100 km/h. Although this relationship
is based on many simplifying assumptions, its predictions of the acceleration times agree
surprisingly Well With measured data.
2./.3 Vehicle Operating Modes
From the first-order differential equation (2.1), the Vehicle speed v can be calculated as a function
of the force Ft. Depending on the value of Ft, the Vehicle can operate in three different modes:
Ft > 0, traction, i.e., the prime mover provides a force to the Vehicle;
Ft < 0, braking, i.e., the brakes dissipate kineti c energy of the Vehicle, the engine can be
engaged or disengaged (to consider fuel cut-off); and
Ft = 0, coasting, i.e., the prime mover is disengaged and the resistance losses are exactly
matched by the decrease of kinetic energy.
T i me t
Fig. 2.4. Modes of vehicle motion.
For the limit case Ft = 0 on a horizontal road and without disturbances, the coasting Velocity
v(t) of the Vehicle can be computed by solving the following ordinary differential equation
derived from (2.1)

Solution (2. 19)
Br a k i n g
Co a s t i n g
Tr a c t i on
D v(t) = dt
.D-A,c.4v
2
(t)gct (2.17)
2 my
= -a
2
. V
2
(t) - o2
1
6 In each case a mass of 100 kg was added to the cur b weight declared by the
manufacturers.
For v > 0, this equation can be integrated in closed form Yielding the result
V(t) = 0 tan { arctan(a v(0)~ a /~ t } (2.18)
A l JJJ
This solution is important because it can be Used to define the three main operating modes of the
Vehicle. As shown in Fig. 2.4, the Vehicle is in:
Traction mode if the speed decreases less than the coasting Velocity v(t) would decrease When
starting at the same initial speed;
Braking mode if the speed decreases more than the coasting Velocity v(t) would decrease
When starting at the same initial speed; and
Coasting mode if the Vehicle speed and the coasting speed v(t) coincide for a finite time
interval.
Therefore, once a test cycle (see Sect. 2.2) has been defined, Using (2.18) it is possible to
decide for each time interval in What mode a Vehicle is operated without requiring any
information on the traction force Ft. Particularly for the MVEG-95 cycle this analysis is
straightforward and Yields the result indicated in the lowest bar (index "trac" for traction mode)
in Fig. 2.6.
2.2 Energy Demand in Driving Cycles
2.2./ Test Cycles
Test cycles consisting of standardized speed and elevation profiles have been introduced to
compare the pollutant emissions of different Vehicles on the same basis. After that first
application, the same cycles have been found to be useful for the comparison of the fuel
economy as Well. In practice, these cycles are often Used on chassis dynamometers Where the
load at the Wheels is chosen to emulate the Vehicle energy losses While driving that specific
cycle (see Sect. 2.1.1). These tests are carried out in controlled environments (temperature,
humidity), With strict procedures being followed to reach precisely defined thermal initial
conditions for the Vehicle (hot soak, cold soak, et c.). The scheduled speed profile is displayed
on a monitor
7
While a test driver controls the gas and brake pedals such that the Vehicle speed
follows these reference values Within pre-specified error bands.
There are several commonly Used test cycles. In the United States, the federal Urban driving
cycle (FUDS) represents a typical city driving cycle, While the federal highway driving cycle FHDS
reflects the extra-Urban driving conditions. The federal test procedure (FTP-75) is approximately
one and a half FUDS cycle, but it also includes a typical Warm-Up phase. The first FUDS cycle is
driven in cold-soak conditions. The second half of the FUDS cycle (hot
7
For vehicles with manual transmission, the requested gear shifts are also signalled on the driver's
monitor.


Fig. 2.5. US test cycle FTP-75 (Federal Test Procedure), length: 11.12 miles (17.8 km), duration:
1890 s, average speed: 21 mph (9.43 m/s).
Soak is driven after a 10-minute engine-off period. The cold start is important to assess the
pollutant emissions (engine and catalyst Warming Up) but the fuel consumption is also influenced
by the cold-start conditions (higher engine friction), though not as dramatically as the pollutant
emissions.
In europe, the Urban driving cycle ECE consists of three start-and-stop maneuvers. The
combined cycle proposed by the motor Vehicle expert group in 1995 (MVEG-95) repeats the
ECE four times (the first With a cold start) and adds an extra-Urban portion referred to as the
EUDC. The Japanese combined cycle is the 10-15 mode, consisting of three repetitions of an
Urban driving cycle, plus an extra-Urban portion. Figures 2.5 and 2.6 show the speed profiles
for the American and the european cycles. Note that for manual transmissions, the MVEG-95
prescribes the gear to be engaged at each time instant.
8

Of course, real driving patterns are often much more complex and demanding (speeds,
accelerations, et c.) Than these test cycles. All automotive companies have their in-house standard
cycles, Which better reflect the average real driving patterns. For the sake of simplicity, in this text
the FTP and MVEG-95 cycles Will be Used in most cases. The methods introduced below, however,
are applicable to more complex driving cycles as Well.
2.2.2 Mechanical Energy Demand Introduction
In the following subsection the mechanical energy is determined that is required to make a
Vehicle follow the MVEG-95 cycle. A key role in these considerations is played by the mean
tractive force Ft,,,,,. The concept of a mean tractive force is particularly useful for the evaluation of
a first tentative
S This restriction is actually not particularly reasonable. Using the additional degrees of freedom
associated with a flexible gear selection can considerably improve the fuel economy. This will
become clearer in the subsequent chapters

Fig. 2.6. European test cycle MVEG-95, gears 1-5, "c": clutch disengaged, "trac": traction time
intervals. Total length: 11.4 km, duration: 1200 s, average speed: urban 5.12 m/s, extrar-urban
18.14 m/s, overall 9.5 m/s. The cycle includes a total of 13 gear shifts.
Value for the fuel consumed by a prime mover and to discuss various aspects of the "Vehicle-to-
miles" energy conversion step.
The mean tractive force ~F.*,,,, is defined as follows
1
Ftrac ______________ F(t) v(t) dt , (2.19)
= .
X. $. Te. *ac
Where x.$. =~ .~~~
~ v(t) dt is the total distance of the cycle and trac is the set
of all time intervals Where F.(t) > 0, i.e., those parts of the cycle in Which the Vehicle drives in
traction modes as defined in Sect. 2.1.3.
Note that in the MVEG-95 cycle the Vehicle is never operated in coasting mode. As mentioned
above, in braking mode the Vehicle does not require any mechanical energy because the
aerodynamic and rolling friction losses are covered by the decreasing kinetic energy of the
Vehicle. The additional Vehicle energy that must be dissipated in these phases is either
transformed into heat by the brakes or converted to another energy form if the Vehicle is
equipped With an energy recuperation device. The last "driving mode," stopped, does not cause
any mechanical energy losses either. However, as Will be shown in the next chapters, this part
leads to an additional fuel consumption caused by the idling losses.
The evaluation of the integral Used to define the mean tractive force is accomplished by a
discretization in time of the drive cycle as follows (more on this approach Will be said in the next
section). The Velocity profile of a test cycle is defined for the given time instants v(ti) = vi, ti = i h,
i = 0, ... , n. Accordingly, the speed Used to compute the mean tractive force is the average
_40 100 200 800 900 1000 1100 1200
5
4
3
2
1
0
c
t
r
o
t

T S

ECE ci ty cycl e
V k m/ h
1 2 0
1 0 0
8 0
6 0
4 0
2 0
EUDC hi ghway c yc l e
(repeat once)
(repeat 4 times)
9
In the MVEG-95 cycle the vehicle is in traction mode approximately 60% of the total cycle time.

Speed
J + J 1
J(F) = J = 2 !V F E [Fi-1! Fi) ; (2.20)
Similarly, the acceleration is approximated by
J ~ J1
(F) = ~~ = + !D F E [Fi-1!Fi) ; (2.21)
Therefore, an approximation of the tractive force (2.19) can be found Using the expression
1
F.*ac P:i _______ &F.*ac,i Ji + ! (2.22)
L. $. Ie. *ac
Where the partial summation I trac only covers the time intervals during Which the Vehicle is in
traction mode.
Case 1 : No Recuperation
According to Sect. 2.1.1, the total tractive force F.*ac includes contributions from three different
effects
F.*ac = F.*ac,a + F.*ac,* + F.*ac,! ! (2.23)
With the three forces caused by aerodynamic and rolling friction and acceleration resistance
defined by
F.*ac,a ~
2 BaApAd &
Ji ' + !
1
L . $ .
1
I e. *ac
1
&~.*~~~* ~ ~ 32 ~ ( ~ A* ~ Ji + ! (2.24)
L.$. Ie.*ac
Ie.*ac
The sums and the distance Lo in the previous equations depend only on the driving cycle and not
on the Vehicle parameters. For the MVEG-95, the ECE cycle, and the EUDC, the following numerical
values can easily be found
1 &
________~ Ji+;::Li1319!82;9!4551!
L. $. Ie. *ac
& Ji + ;::Li {0;856! 0;81! 0;881 ! (2.25)
Ie. *ac
1
& ~ ~i Ji + ,: {0;101! 0;126! 0;0861 ;
L.$.
1
______________ &~.*~~~!~ ~32~ ~iJi+;
L.$.
1 L . $ .
I e. *ac

Once the driving cycle is chosen, (2.22) allows for a simple estimation of the mean tractive force
as a function of the Vehicle parameters 32, Af, Ad, and A*. The mean tractive force is equal to E, the
average energy consumed per distance travelled. When the latter is expressed Using the Units
kj/100km, the relationship between the two quantities is E = 100 ~F.*a,c. In these Units, the
energy consumed in the MVEG-95 cycle is given by
E' MVEG-95-= Af Ad 1;9 . 10
4
+ 32 A* 8;4 . 10
2
+ 32 10 (kj/100 km) !
(2.26)
Where, to simplify matters, the physical parameters air density BQ (at sea level) and acceleration (
have been integrated in the numerical values. Considering typical values for the Vehicle parameters,
the three contributions in this sum are of the same order of magnitude.
Case 2: Perfect Recuperation
Equation (2.22) is Valid for the case that none of the Vehicle's kinetic energy is recuperated When
braking. In the opposite case, With perfect recuperation (recuperation device of Zero mass and
100% efficiency), the energy spent to accelerate the Vehicle is completely recuperated during the
braking phases. As a consequence, the mean force F does not have any contributions ~F! Caused by
acceleration losses, i.e.,
F= FQ + F* ; (2.27)
However, in the full recuperation case, the mean force must include the losses caused by
aerodynamic and rolling resistances also during the braking phases. In this case, (2.24) must be
replaced by
FQ ___________
1 ~
L.$.
1
2 . BaAf Ad
&" IV + ! (2.28)
I =1
1
~~* ~ ~ 32 ~ ( ~ A* ~
L.$.
&"
i =
1
Vi + ! (2.29)
With the following numerical values Valid for the MVEG-95, the ECE, and the EUDC cycles
Vi + ~ ~363!100!515}!
(2.30)
Vi . + = {1!1!1};

With these results, an approximation can be derived that is similar to (2.26) of the mechanical
energy
-
E*~~,MV EG-95 needed to drive 100 km in the MVEG-95 cycle. In the case of full
recuperation this quantity is given by
~E*, ,MV EG-95 -= Af cd 2.2 10
4
+ m2 c* 9.81 10
2
(kj/100 km) .
(2.31)

N
4 0 0
3 0 0
2 0 0
1 0 0
L /100 km
Diesel
1.
0
0.
5
F
<: )
+, 5
F
<: )+
, :
F
<: )+
, )
17: -
+ =
0
17
: - +
=
1
F:
F)
Ft: )+, 5
F
<: )+, :
Ft:)+, )
17: -
+ =
0 17:
- +
=
1
F5,
*
F5,
:



F5,
*

F5,
:
F:

F)

A. +, = 0. 7 m2

A. +, = 0. 4 m2

5> = 15
00
K
g
5> = 75
0
K
g
+: =
0
. 01
2
+: =
0
. 00
8

Fig. 2.7. Comparison of the energy demand in the MVEG-95 cycle for a full-size car (left) and
a light-weight car (right). A mean force of 1 N is equivalent to 27.78 Wh mechanical energy
per 100 km. The lower heating value for Diesel fuel is approximately 10 kwh/l.
Examples
Figure 2.7 shows the mean tractive force and the corresponding specific energy demand in the
MVEG-95 cycle for two different Vehicles. On the basis of the fuel's lower heating value, a full-size
passenger car (without recuperation) requires per 100 km travelled distance an amount of
mechanical energy that corresponds to the heating value of 1.16l of Diesel fuel.
10
This result is
obtained by inserting the Vehicle parameters shown in Fig. 2.7 into (2.26). In the case of perfect
recuperation (a 100% efficient recuperation device With no additional mass), the corresponding
value, obtained by Using (2.31), is 0.89l of Diesel fuel. In the best case, i.e., When equipped With an
ideal recuperation device, a hypothetical advanced light car would require the equivalent of
10 Because of the losses in the engine and in the other components of the powertrain, the actual
fuel consumption is much higher