You are on page 1of 157

MODELING OF ROAD VEHICLE LATERAL DYNAMICS

by
Joseph R. Kiefer A Thesis Submitted
in

Partial Fulfillment of the Requirements for the MASTER OF SCIENCE m Mechanical Engineering Approved by: Professor - - - - - - - - Dr. Kevin Kochersberger Thesis Advisor Dr. Alan Nye

Professor - - - - - - - - Professor
Dr. Michael Hennessey

_ _
Dr. Charles Haines Thesis Advisor

Professor

DEPARTMENT OF MECHANICAL ENGINEERING COLLEGE OF ENGINEERING ROCHESTER INSTITUTE OF TECHNOLOGY AUGUST 1996

Disclosure Statement

Pennission Granted

I, Joseph R. Kiefer, hereby grant pennission to the Wallace Memorial Library of the Rochester Institute of Technology to reproduce my thesis entitled Modeling of Road Vehicle Lateral Dynamics in whole or in part. Any reproduction will not be for commercial use or profit.

August 16, 1996 Joseph R. Kiefer

ii

Abstract
The lateral dynamics
mathematical model.

of a road vehicle

is

studied through the

development

of a

The

vehicle

is

represented with

two degrees of freedom, lateral and

yaw.

Equations

of motion are

derived for this

vehicle model

from basic

principles of

Newtonian

mechanics.

Both linear

and non-linear models are

developed.

In the linear
relationship.

model

tire lateral forces are


are written

represented

by an approximate linear
for
various control and

Transfer functions

for the

vehicle system

disturbance inputs. Steady-state


response are

and transient response characteristics and

frequency

studied, and numerical simulation of the


non-linear model a

model

is

performed.

In the

detailed

representation of tire

lateral forces known

as tire

data nondimensionalization is

utilized.

Simulation is

performed and compared with

the

linear model

simulation

to determine the range of applicability of the linear modeling

assumptions.

Table

of

Contents

Disclosure Statement Abstract Table List List List


of

ii iii

Contents

iv
vi

of

Tables Figures Symbols Introduction Literature Review Tire Behavior

of

vii

of

Chapter 1

Chapter 2
Chapter 3

5
10 10 11 14
14

3.1 Introduction

3.2 Lateral Force Mechanics


3.3 Linear Tire Model 3.4 Non-Linear Tire Model Chapter 4 Two Degree-of-Freedom Vehicle Model

24
24 25 26 27 28

4. 1 Introduction 4.2 Description of Model 4.2. 1 Assumptions 4.2.2 Vehicle Parameters 4.2.3 Free-Body Diagram

4.3 Derivation
4.4 Derivation

of of

Equations Tire

of Motion

29 31
33 34 35 35 36 37 39 41 42 42 43 43

Slip Angles

4.5 Linear Model 4.5. 1 Additional Assumptions 4.5.2 Vehicle Sideslip Angle 4.5.3 Tire Slip Angles 4.5.4 External Forces and Moments 4.5.5 Equations of Motion 4.5.6 Transfer Functions 4.5.7 Vehicle Sideslip Angle Gain 4.5.8 Yaw Velocity Gain 4.5.9 Front Tire Slip Angle Gain 4.5.10 Rear Tire Slip Angle Gain 4.5.11 Path Curvature Gain

iv

Table

of

Contents

4.5. 12 Lateral Acceleration Steady-State 4.5.13 Steady-State Steer Angle 4.5.14 Understeer Gradient 4.5.15 Stability Factor 4.5. 16 Neutral Steer Point 4.5.17 Static Margin 4.5.18 Tangent Speed 4.5.19 Critical Speed 4.5.20 Characteristic Speed 4.5.21 Characteristic Equation 4.5.22 Undamped Natural Frequency 4.5.23 Damping Ratio 4.5.24 System Poles 4.5.25 System Zeros 4.5.26 Frequency Response 4.5.27 Simulation 4.6 Non-Linear Model 4.6. 1 Model Equations 4.6.2 Simulation
Chapter 5

Step Response Gain

44 45 46 48 48 49

50 51
51 52 53 53

55 55 58 65 89 89 90
100 102

Conclusion

References Appendix A Tire Model MATLAB Programs

105 105 106


107 108 123 123 124

A. 1 MagicFit.m A.2 MagicError.m

A.3 NLTire.m
Appendix B Appendix C Two DOF Model Mathematica Session Two DOF Model MATLAB Programs

C.l DOF2Control.m C.2 DOF2Param.m C.3 DOF2DependParam.m


C.4 SteerAngle.m C5 DOF2LFreq.m

126 127 129


130 132 133 135 136

C.6 DOF2LSim.m
C7DOF2LDE.m

C.8 DOF2NLSim.m

C.9 DOF2NLDE.m
Appendix D Relevant Literature

List

of Tables

Table 3.1: Non-Linear Tire Model Parameters Table 3.2: Experimental Tire Data Table 4.1: Vehicle Parameters
Table 4.2: Linear Tire Model Parameters

17 21 28
36

Table 4.3: Steady-State Response Gains Table 4.4: System Zeros

45 56

vi

List

of Figures

Figure 3. 1 : Tire

Slip Angle

11

Figure 3.2: Tire Lateral Force Versus Slip Angle Figure 3.3: Experimental Tire Data Figure 3.4: Tire

12

21 22 22 23
23 26

Cornering Coefficient

Figure 3.5: Tire Lateral Friction Coefficient


Figure 3.6: Tire Normalized Lateral Force Figure 3. 7: Reconstructed Tire Lateral Force

Figure 4.1: Vehicle Model


Figure 4.2:

Free-Body Diagram

29
32

Figure 4.3: Kinematic Diagram Figure 4.4: Gravitational Side Force Figure 4.5: Natural Frequency Figure 4.6: Poles Figure 4. 7: Figure 4.8:
and vs.

37

Vehicle

Velocity

54 57 V
=

Zeros

Sideslip Angle / Steer Angle Frequency Response,

100 km/hr
V
=

61
61

Sideslip Angle /Aero Side Force Frequency Response, Sideslip Angle /Road Side Slope Frequency Response,
Velocity / Steer Angle Frequency Response,

100 km/hr 100 km/hr

Figure 4.9:

62 62 63 63
64

Figure 4.10: Yaw


Figure 4.11: Yaw

V'= lOOkm/hr

Velocity /Aero Side Force Frequency Response,

V=100 km/hr V
=

Figure 4. 12: Yaw


Figure 4. 13: Figure 4. 14:

Velocity /Road Side Slope Frequency Response,


V
=

100 km/hr

Sideslip Angle / Steer Angle Frequency Response,


Sideslip Angle / Steer Angle Frequency Response,

30 km/hr

49.84 km/hr

64 67
71

Figure 4.15: Simulation Steer Angle Inputs


Figure 4.16: Linear Step Steer Lateral

Velocity Response

vn

List

of Figures

Figure 4.17: Linear Step Steer Yaw

Velocity Response

71 72

Figure 4.18: Linear Step Steer Sideslip Angle Response


Figure 4. 19: Linear Step Steer Front Tire Figure 4.20: Linear Step Steer Rear Tire

Slip Angle Response

72
73 73
74

Slip Angle Response

Figure 4.21 : Linear Step Steer Lateral Acceleration Response

Figure 4.22: Linear Ramp Step Steer Lateral


Figure 4.23: Linear Ramp Figure 4.24: Linear Ramp Figure 4.25: Linear Ramp Figure 4.26: Linear Ramp Figure 4.27: Linear Ramp

Velocity Response

Step Steer Yaw Velocity Response Step Steer Sideslip Angle Response
Step Steer Front Tire Slip Angle Response Step Steer Rear Tire Slip Angle Response

74

75
75 76 76

Step Steer Lateral Acceleration Response


Velocity Response

Figure 4.28: Linear Ramp Square Steer Lateral Figure 4.29: Linear Ramp Square Steer Yaw

77
77

Velocity Response

Figure 4.30: Linear Ramp Square Steer Sideslip Angle Response Figure 4.31: Linear Ramp Square Steer Front Tire Figure 4.32: Linear Ramp Square Steer Rear Tire

78
78

Slip Angle Response

Slip Angle Response

79 79 80 80
81 81 82 82 83

Figure 4.33: Linear Ramp Square Steer Lateral Acceleration Response Figure 4.34: Linear I Hz Sine Steer Lateral Figure 4.35: Linear 1 Hz Sine Steer Yaw

Velocity Response

Velocity Response

Figure 4. 36: Linear I Hz Sine Steer Sideslip Angle Response Figure 4.37: Linear I Hz Sine Steer Front Tire Figure 4.38: Linear I Hz Sine Steer Rear Tire

Slip Angle Response

Slip Angle Response

Figure 4.39: Linear 1 Hz Sine Steer Lateral Acceleration Response


Figure 4.40: Linear Step Aero Side Force Lateral Figure 4.41: Linear Step Aero Side Force Yaw

Velocity Response

Velocity Response

83

vui

List

of

Figures

Figure 4.42: Linear Step Aero Side Force

Sideslip Angle Response Slip Angle Response

84
84 85

Figure 4.43: Linear Step Aero Side Force Front Tire Figure 4.44: Linear Step Aero Side Force Rear Tire

Slip Angle Response

Figure 4.45: Linear Step Aero Side Force Lateral Acceleration Response Figure 4.46: Linear Step Road Side Slope Lateral Figure 4.47: Linear

85 86 86 87
87

Velocity Response

Step Road Side Slope

Yaw

Velocity Response

Figure 4.48: Linear Step Road Side Slope

Sideslip Angle Response Slip Angle Response

Figure 4.49: Linear Step Road Side Slope Front Tire Figure 4.50: Linear Step Road Side Slope Rear Tire

Slip Angle Response

88 88
94 94

Figure 4.51: Linear Step Road Side Slope Lateral Acceleration Response Figure 4.52: Non-Linear Step Steer Lateral

Velocity Response

Figure 4.53: Non-Linear Step Steer Yaw

Velocity Response

Figure 4.54: Non-Linear Step Steer Sideslip Angle Response

95 95 96
96 97

Figure 4.55: Non-Linear Step Steer Front Tire


Figure 4.56: Non-Linear

Slip Angle Response

Step Steer Rear Tire Slip Angle Response

Figure 4.57: Non-Linear Step Steer Lateral Acceleration Response Figure 4.58: Non-Linear Ramp Square Steer Lateral Figure 4.59: Non-Linear Ramp Square Steer Yaw

Velocity Response

Velocity Response

97 98 98

Figure 4.60: Non-Linear Ramp Square Steer Sideslip Angle Response Figure 4.61: Non-Linear Ramp Square Steer Front Tire Figure 4.62: Non-Linear Ramp Square Steer Rear Tire

Slip Angle Response

Slip Angle Response

99
99

Figure 4.63: Non-Linear Ramp Square Steer Lateral Acceleration Response

IX

List
l/R
a

of

Symbolst

Path

curvature

(1/m)
center

Distance from mass


Acceleration
of

to front axle

(m)
(m/s2)

a0

the origin of the

vehicle-fixed coordinate system

ay

Acceleration Acceleration

of vehicle mass center

in y-direction (m/s in y-direction in

)
"g"

Ay
b

of vehicle mass center

units of

(g)

Distance from

mass center

to rear axle

(m)

Bj

Magic Formula curve fit parameter Tire cornering


coefficient

B3 B5
c

intercept (N/deg/N) intercept force (m)

Tire lateral friction

coefficient

Distance from front

axle to aerodynamic side

C} C3

Magic Formula curve fit parameter


Tire cornering
coefficient slope

(N/deg/N2)

C5
Ca

Tire lateral friction Tire cornering

coefficient slope

stiffness

(N/rad)
-

Cf
Cr
Cc
d

Front tire cornering Rear tire cornering Tire cornering

stiffness

two tires

(N/rad)

stiffness

two tires

(N/rad)

coefficient

(N/deg)

Distance from front

axle

to neutral steer point

(m)

D} Ej
/
F

Magic Formula curve fit parameter Magic Formula curve fit parameter
Front
axle weight

fraction

External force

(N)

By convention,

scalar variables are

italicized

and vectors are

boldfaced.

List

of

Symbols

F^

Fictitious lateral force for finding neutral


Tire lateral force (N) Normalized tire lateral force

steer point

(N)

Fy Fy
Fya

Aerodynamic

side

force disturbance (N)


-

F^

Front tire lateral force Gravitational


side

two tires

(N)

Fyg
F

force disturbance (N)


-

Rear tire lateral force Tire


vertical

two tires

(N)

Fz
g

load (N)

Acceleration due to gravity (m/s2)


Linear
momentum

G
H

(kg-m/s)
(about
mass

Angular Unit

momentum

center)

(kg-m2/s)

vector

in x-direction

of vehicle-fixed coordinate system

Izz
j
k K

Total Unit Unit

vehicle yaw mass moment of

inertia (kg-m2)

vector

in y-direction
in

of vehicle-fixed coordinate system

vector

z-direction of vehicle-fixed coordinate system

Stability factor (s2/m2)


Understeer
Wheelbase Total
gradient

Kus
L
m

(rad)

(m) (kg)
mass

vehicle mass

External Yaw

moment

(about

center)

(N-m)

Nr

damping derivative (N-m-s/rad)


(N-m/rad)

/Vp
Ns
r

Directional stability derivative

Control

moment

derivative (N-m/rad)

Yaw velocity Path


radius

(rad/s)

(m)

XI

List

of

Symbols

Rf Rr
s

Position Position

vector

from vehicle
from

mass center to

front tire (m)


rear

vector

vehicle mass center

to

tire

(m)

Laplace-domain variable
Static
margin

SM
u

Forward velocity (m/s)


Lateral velocity (m/s) Magnitude
of vehicle

velocity

(m/s)

Vchar

Characteristic
Critical
speed

speed

(m/s)

VcHt Vf V0 Vr
V[an Vx

(m/s)

Velocity of front tire (m/s) Velocity of vehicle-fixed coordinate system (m/s)

Velocity of rear tire (m/s)


Tangent
speed

(m/s)
velocity in x-direction velocity in y-direction

Component Component

of

(m/s) (m/s) (N-s/rad)

Vy
Yr

of

Lateral force/yaw coupling derivative

Tp
Y6
a

Damping-in-sideslip derivative (N/rad)


Control force derivative
Tire slip Tire
angle

(N/rad)

(rad)
slip
angle

normalized

af
ar

Front tire slip


Rear tire slip

angle

(rad)

angle

(rad)
(rad)

(3

Vehicle sideslip

angle

5
^Acker

Front

steer angle

input (rad)

Ackerman

steer angle

(rad)

xn

List

of

Symbols

8/
8r

Front
Rear

steer angle

(rad)

steer angle

(rad)
for
given path radius

8M
yr

Steady-state

steer angle

(rad)

Magic Formula intermediate variable


Road
side slope

(rad)
variable

Magic Formula intermediate

Damping ratio
liy
oo
Q.

Tire lateral friction Undamped


natural

coefficient

frequency (rad/s)
(rad/s) (rad/s)

Angular velocity Angular velocity

of vehicle-fixed coordinate system

Qz

of vehicle-fixed coordinate system about z-axis

xin

Introduction

The first practical


Daimler.1

automobiles were

built in 1886 by Karl Benz


only
about

and

Gottlieb hour. With

The top

speeds of these vehicles were

fifteen

miles per

much of

the automotive industry's early engineering

effort

devoted to

developing faster
and

vehicles, production car


miles per

top

speeds reached

forty-five

miles per

hour by 1900

eighty

hour by 1915. This year,


speed of sound

Craig Breedlove will attempt to be the first man to travel

faster than the

in

a ground vehicle.

Since the top little


concern with

speeds of the

first

automobiles were

relatively low, there


as cars

was

initially

the dynamic behavior of the vehicles.


vehicle

However,
an

quickly became
concern

capable of

achieving higher speeds,

dynamics became

important

for

automotive engineers.

Of primary importance from a safety

standpoint was

the behavior of
since

vehicles

in

maneuvers such as

turning
quality

and

braking as top

speeds

increased. Also,

early

roads were of

very

poor

by today's

standards, isolation of the driver

and

passengers

from road disturbances became


of vehicle

increasingly important.
three basic modes of vehicle
refers to the vertical response

The field
performance.

dynamics

encompasses

Vertical dynamics,

or ride

dynamics, basically

of

the

vehicle

to road disturbances. Longitudinal dynamics

involves the

straight-line

acceleration and

braking of the vehicle. Lateral dynamics is concerned with the vehicle's


acceptable performance

turning behavior. Achieving


order

in

each of

these modes is necessary in

for

a vehicle

to

meet

the requirements of the consumer, and the government, with


and safety.

regards

to comfort, controllability,

Vehicle dynamics

needs to

be

considered

throughout the entire design and development process

from initial

conceptualization

through

production

if performance

goals are to

be

met.

Mathematical modeling is

an

excellent

tool for engineers to use to design and

develop vehicles that meet performance

goals.

Chapter 1

Introduction

Traditionally there has been a relatively long cycle in the design and development
process

from the initial

concept

for

a vehicle

to its
can

production.

With

such a

long time from

initial

concept to

production,

vehicle

designs

be

out of style and obsolete

by the time

they reach production. Increasing competition from a globally expanding industry has
driven
automobile manufacturers to reduce the

length of the design


in
market

cycle.

This

allows

manufacturers to respond more

quickly to

changes

demand. In addition, reducing


new vehicle.

the length of the design cycle reduces the cost of developing a

One way that


possible

manufacturers can reduce

design

cycle

length is
of

by achieving the best

design before any

prototypes are

built. The development

the digital computer

and

the techniques of computer-aided engineering such as solids modeling, finite element

analysis, computational fluid

dynamics,

and

multibody dynamics

simulation

have greatly
software

facilitated this
becomes
more

effort.

As

computer speeds

continually increase
can

and

engineering

powerful, better

vehicle

designs

be

obtained

before

prototypes are

built,

resulting in fewer

prototypes and reduced

development time dynamics helps

and cost.

Mathematical modeling
to achieve a design

of vehicle

engineers reduce

the time it takes

which will meet performance requirements

for the

consumer and

for

government regulations.

A proposed design

can

be

studied to

determine if it
be

can meet goals

before any

prototypes are

built. The

effects of

design

changes can

evaluated without

building costly prototypes.


with

Development

engineers can use mathematical models to assist

the

tuning

of prototypes

by identifying the changes which should be made to produce

desired ride

and

handling characteristics.
simulation offers a

Computer
of

controlled, repeatable environment where the effects


without

individual

parameters can

be isolated

the influence of the variations in the

environment.

Simulation

can remove

the performance of the test driver from the picture to

isolate the
maneuvers

performance of

the

vehicle.

In addition,

simulation can

be

used to

study

that

could result

in costly damage to the

vehicle or

danger to the test driver. An

Chapter 1

Introduction

example would

be

a maneuver

resulting in

roll over.

Real time

driving simulators can be

used to train

drivers

and to evaluate

driver performance in

crash avoidance maneuvers or

when

drowsy or under the influence of alcohol. Many of the new safety and comfort
technologies
such anti-lock

related

brakes,

traction control,

stability control,

and variable

damping would be very difficult, if not impossible, to develop without the use of
simulation.

Computer

simulation

has many

useful applications

in the field

of vehicle

dynamics.
Vehicle dynamics
models can

have

a wide range of complexity.

Models

can

consider just a single mode of performance

(vertical, longitudinal, lateral)


include

or a combination

of modes.

Depending upon its purpose,


the
vehicle.

a model must

representations of

appropriate systems of

Effects
or

of

the suspension system, steering system,

powertrain

system,

braking system,

tires may need to be modeled. Representations of

these systems can be linear or non-linear, quasi-static or dynamic

depending upon the

accuracy

required.

Control

of

the vehicle can be open-loop or it can be closed-loop if an

appropriate representation of the

driver is

available.

The

vehicle model used must

be

suitable

for the

maneuvers

it

will simulate.

The

subject of

this thesis is the modeling of road vehicle lateral dynamics. As such,

it is

concerned with

the

turning behavior of the vehicle in response to control

and

disturbance inputs. A literature


as

simple

two degree-of-freedom vehicle model popularly known in the

the bicycle

model

is

used

in this

study.

Despite its simplicity, the two

degree-

of-freedom model can

be very

useful

in

demonstrating the interaction of major parameters


location,
wheelbase, and forward

such as

tire properties, inertia properties, mass center

speed.

Chapter 2 is

a review of vehicle

dynamics literature relating to lateral dynamics.

Since the majority


of

of

the forces acting on a vehicle are developed

by the tires,

an overview

tire lateral force

mechanics

is

provided

in Chapter 3 along

with a

description

of the

two

Chapter 1

Introduction

tire models used in this thesis. Both a linear tire

model and a non-linear

tire model are used

in this

study.

The

non-linear model

is based on

a method called tire

data

nondimensionalization.

Chapter 4
of the

presents the main

focus

of this

research, the development and application

two degree-of-freedom vehicle model. Equations of motion are derived from basic

principles of

Newtonian

mechanics.

The

model

is then developed in two forms, linear


used

and

non-linear.

In the linear

model

transfer functions are written and

to derive various

measures of steady-state and transient response and to examine the

frequency response of

the vehicle to control and disturbance inputs. In addition, the response of the linear model
to various inputs is simulated

by integrating the differential equations of motion. Also,


is
performed and the results are compared with those of

simulation of the non-linear model

the linear model. Conclusions are drawn regarding the range of applicability of the linear
model.

Literature Review

As early

mentioned

in Chapter 1
vehicle

as the

top speeds

of automobiles

increased rapidly in the


to
engineers.

part of

this century

dynamics became

an

important

consideration

A large

body of vehicle dynamics literature cunently exists covering all aspects of vertical,
and

longitudinal,

lateral performance. Fundamental to understanding lateral vehicle


of the mechanism of

dynamics is knowledge
been
written on

tire lateral force generation, and much has

this topic.
papers road vehicle

One William

of

first

concerning
work

lateral dynamics

was written

in 1908

by

Lanchester.2

In this

Lanchester discussed the steering behavior


of

of

automobiles.

However,

complete

understanding

turning behavior was hampered in the


In 1925 George Broulhiet

early

years

by a lack of understanding of tire mechanics.

published a paper

titled "The Suspension and the Automobile


generation

Steering

Mechanism"

which

described tire lateral force

in terms

of

the slip angle concept

which

is

still used

today

and

forms the basis for nearly

all

lateral

vehicle

dynamics

models.

Following this

development tire dynamometers


tire
under various conditions.

were

built

which could measure

the forces generated

by a

These

advancements paved

the way for others to

develop

detailed

explanations and models of

turning behavior.

One
responsible

of

the early pioneers in vehicle dynamics research was Maurice Olley. He was
of

for the introduction described the

the independent front suspension in the United States for the system in the 1934 SAE paper "Independent

Cadillac

and

operation of

Wheel Suspension: Its Whys


"Suspension
Handling"

Whererfores."

and

report written

in 1937 titled

and

reviewed

the

research

in lateral dynamics

during the preceding

years and covered much of what

is

today.3

understood

Olley was active in vehicle


in 1955.

dynamics from the early 1930's through his

retirement

During the period of the

Chapter 2

Literature Review

early 1960's he
summarized

published a series of

documents know today

"Olley'

Notes"4'5

as

which

his

extensive

knowledge

of suspension systems and

handling. dynamics
was written

One

of the most significant works


Segal.6

concerning lateral

vehicle

in 1956 by Leonard

Segal,

who worked at

Cornell Aeronautical Laboratory,


which

applied to the road vehicle

many

of

the

analytical

techniques

had been developed for


Experimental

aircraft

dynamics. In this work,


of

entitled

'Theoretical Prediction

and

Substantiation

the Response of the Automobile to

Control,"

Steering

Segal developed
roll)
model of

equations of motion

for

linear three degree-of-freedom (yaw, lateral,


available

and

vehicle

turning behavior. Since digital computers were not


a

for his research, it

was

necessary to have

linear

model

for

which

transfer functions could be written and

closed-

form

solutions

found. Segal

used

the stability derivative technique in the derivation of the


concepts of

equations of motion and

described the

stability factor,

neutral steer

point, and

static margin.

Segal backed his modeling


was

efforts with experimental

testing of a vehicle and

concluded

that a linear model

sufficiently

accurate

for lateral

motions of a reasonable

magnitude.

A
and

second paper

"Design Implications

of a

General

Theory of Automobile Stability

Control"

written

by David Whitcomb and William Milliken, also of Cornell


was presented at

Aeronautical
paper series

Laboratory,

the same time as Segal's paper as part of a five


Performance."7

"Research in Automobile

Stability and Control and in Tyre

This

paper studied

the vehicle as a linear two degree-of-freedom system. This enabled the

authors

to utilize a large
systems.

body of established techniques

for the

analysis of second order

dynamic

These

papers preceded a great

deal

of research

in lateral

vehicle

dynamics

which

has

provided engineers

today

with a comprehensive

understanding

of the subject.

Several

textbooks have been written covering the subject of vehicle

dynamics.

Among these are Car

Suspension

and

Handling (Bastow and Howard,

1993),8

Elementary Vehicle Dynamics

Chapter 2

Literature Review

(Cole,

1972),9

Tyres, Suspension

and

Handling (Dixon,
1989),12

199

1),10

Vehicle Dynamics (Ellis,

1969),11

Road Vehicle Dynamics (Ellis,


1992),1

Fundamentals of Vehicle Dynamics


and

(Gillespie,

Race Car Vehicle Dynamics (Milliken


1969),14

Milliken,

1995),13

Fundamentals of Vehicle Dynamics (Mola, Principles (Reimpell


and

The Automotive Chassis:

Engineering
and

Stall,

1996),15

Mechanics of Vehicles (Taborek,


n

1957),16

Theory of Ground
freedom vehicle
Another
Leffert.18

Vehicles (Wong, 1993).

Most

of

these books utilize a two degree-of-

model when

explaining turning behavior.


literature
was made

significant contribution to the

in 1976

by Bundorf and

In this

work

the cornering compliance

concept

is described. With this technique


to
understeer are

the contributions of various vehicle systems and


and added

characteristics

determined

to estimate the total

understeer of the vehicle.

This

allows engineers

to see the

effects of

steering

and suspension

compliances, roll steer, tire cornering stiffnesses, tire


and

camber

stiffnesses, tire aligning torque,

lateral load transfer

on understeer without

developing the detailed vehicle models that would be necessary to simulate these effects
directly. Since the computing hardware
models was not and software needed

to analyze sufficiently detailed

readily

available at

the time, this concept was a significant advancement.


vehicle models

Since the
expand on

work of

Segal in 1956 many

have been developed

which

his

model.

The dynamics

of other systems such as

the steering system have

been integrated into the

vehicle models.

Lateral dynamics
of

models

have been

expanded to

include longitudinal

and vertical

degrees

freedom. Non-linearities, particularly in tire


models.

force generation, have been included in the


found in
works

Some

examples

in the literature

can

be

by

Allen,19

Heydinger,20

Xunmao.21

and

With the 1970's

came

the development of multibody

dynamics

codes.

These

software programs allow

the parts of mechanisms, or in this case vehicles, to be modeled

individually
individually

and connected

using joints.

By modeling each suspension component


be
obtained.

very

accurate

kinematic

representation of a complete vehicle can

Chapter 2

Literature Review

Examples

of the application of

multibody

codes to vehicle

dynamics

can

be found in the

literature.22'23

As

computer

processing

speeds

increase,

the use of multibody codes for


with

vehicle

dynamics

simulations

becomes

more practical.

The biggest disadvantage

the

use of

these codes is the large amount

of

information that is in

required to construct the

models.

The dimensions,

mass properties, and

some cases stiffnesses of each relevant

component must

be known to build
simulation

an accurate model.

Commercial multibody
and

codes used

for

vehicle

dynamics

include ADAMS, DADS,

Mechanica Motion.
models

Paramount to the development

of successful vehicle

dynamics

has been the

development

of accurate representations of

tire behavior. Much effort has been devoted to

this task and the results can be found in the literature. One the first attempts at a theoretical
model of tire

behavior was done

by von Schlippe and Dietrich in

1941.

They represented

the tire

by a massless taut string on an elastic foundation and predicted forces based on the
and material properties of the tire.

geometry have

Most of the digital

major advancements

in tire

models

occurred within the

last fifteen years

as

computers

have become readily

available.

comprehensive analysis of tire mechanics was performed under a government


1981.24

contract

by Clarke in

In 1990

detailed theoretical tire

model was

developed by

Gim

and

Nikravesh.25

However,

most of

the popular tire models in existence


curve

today

are

based primarily
measured

upon empirical

data. These models involve

fitting of experimentally

tire data. One of the most popular empirical tire models known as the "Magic
was published

Formula"

by Bakker, Nyborg,
and
Allen.28

and

Pacejka in

1987.26

Other useful tire


tire models has

models

include those

by

Radt27

The development
modeling.

of accurate

been

critical

to the success

of vehicle

dynamics

There
years

exists a

large

body

literature regarding
significant

vehicle

dynamics. The last forty


the topic. Models of

in

particular

have

seen

many

developments

on

vehicles and

tires have been developed to the point where very accurate simulations of
response can

lateral dynamic

be

performed.

The

advent of the

digital

computer

has greatly

Chapter 2

Literature Review

enhanced the

ability

of engineers

to

develop and utilize these models for practical gains. A


dynamics literature reviewed during this
research

list of relevant
provided

sources

from the

vehicle

is

in Appendix D.

Tire Behavior

3.1 Introduction
With the
exception of gravitational and aerodynamic

forces,

all of

the forces acting


vehicle

on a road vehicle are applied

to the

vehicle

through its tires. In supporting the


vehicle changes speed or

the

ground applies vertical

forces to the tires. When the

direction

as a

result of control

inputs,

the forces and moments which produce these

accelerations

are, in

general, applied to the

vehicle

by the ground through the tires. Thus, to model the


a suitable representation of tire

dynamics
Two tire widely

of road vehicles

it is necessary to have
a simple

behavior.

models are used

in this thesis:

linear model

and a more

accurate, more

applicable non-linear model.

The

requirements of a

tire

model

vary

depending upon the aspects of vehicle


In general, there
are

performance which are

being modeled and the accuracy required.

three

force

components and

three moment components acting on a tire due to its interaction with

the ground. In a complete model of vehicle dynamics where the


vertical motions of the vehicle are

longitudinal, lateral,
be

and

being studied,

all six of these components must

included to accurately model the

effect of the

tires

on

the dynamics of the

vehicle.

However,

this thesis is concerned only with the lateral dynamics of the vehicle. The simple

vehicle model which

is

studied

has only lateral

and rotational

degrees

of

freedom in the
the vertical

horizontal
axis of

plane.

Thus only forces in the lateral direction

and moments about

the

vehicle need

to be considered. The moment acting on the tire itself about its


moment.

vertical axis

is

called

the tire aligning


of

The

effect of the

aligning

moments of the

tires on the

overall

dynamics

the vehicle is generally small compared to the effect of the


vehicle model which

lateral forces
moments of

of the tires.

In the

is

presented

here the aligning


modeled

the tires are

neglected.

Thus the only

aspect of tire

behavior which is

is lateral force

generation.

10

Chapter 3

Tire Behavior

3.2 Lateral Force Mechanics


The
complete

mechanics of the

lateral force

generation of a

tire is a complex process. A this thesis.

discussion

of

this

process

is beyond the

scope of

Many thorough
The lateral

discussions
force

of the mechanics of

force

generation exist

in the

literature.1'1317

Fy generated by a pneumatic tire depends upon many variables including road surface
carcass

conditions, tire

construction, tread

design,

rubber

compound, size, pressure,

temperature, speed,
given tire on

vertical

load, longitudinal

slip, inclination angle, and slip angle. For a


vertical

unchanging,

dry road surface conditions,

load

and

slip

angle are

the

variables

having the largest effect and are the variables considered for the tire models used

in this thesis.
The tire slip
angle angle

is

represented

by the symbol a and is defined by S AE as


of the center of

"the This

between the

X'

axis and

the direction of travel

tire

contact."29

X'

Figure 3. 1: Tire

Slip Angle

11

Chapter 3

Tire Behavior

definition

references the

S AE tire
X'

system.29

axis

The

origin of

this system is at the center of

the tire contact patch. The


of

axis

is the intersection

of the plane of the wheel and the plane


Z'

the ground and is positive in the forward direction. The

axis

is

perpendicular

to the

plane of the road and

is

positive

in the downward direction. The

Y'

axis

is in the

plane of

the road and oriented to form a right-hand Cartesian coordinate

system.

The tire slip angle,


angle and

lateral force,

and tire axis system are shown

in Figure 3.1. A

positive

slip

lateral

force

are shown.

Simply stated, the slip angle is the angle between the direction the wheel
direction it is traveling
produced at a given

is pointing

and the

instant in time. among


other

The lateral force

by a tire is a non-linear function of,


versus

variables, vertical load and slip angle. A typical lateral force


single vertical

slip

angle curve

for

load is

shown

in Figure 3.2. At low slip

angles the curve

is approximately

linear. Here the lateral force

generated

depends primarily

on

the tire construction, tread


and ground

design,
within

and tire pressure.

There is little sliding occurring between the tire Lateral force is developed
as a result of

the contact

patch.

deformation

of

the tire.

\.<h.j...\....^r.

v>r
<D o

2
0)
CO

10

12

14

16

Slip Anlge (deg)


Figure 3.2: Tire Lateral Force Versus

Slip Angle

12

Chapter 3

Tire Behavior

The initial
tire. The

slope of lateral

force
is

versus

slip

angle curve

is the cornering

stiffness

Ca of the

cornering

stiffness

often used as a

linear

approximation

to the relationship
stiffness can

between lateral force


normalized

and

slip

angle

(see Section 3.3). The cornering

be

by dividing by the vertical load. This quantity is the cornering coefficient Cc of


coefficient

the tire. The cornering stiffness and the cornering

both vary

with

the vertical
while

load

on

the tire. In general, the cornering stiffness increases with


coefficient

vertical

load,

the

cornering

decreases. Both

of

these quantities are used in the tire models used in

this thesis.

As the slip

angle

increases the

slope of

the lateral force curve decreases until the

lateral force reaches

a maximum.

At this

maximum the

lateral force divided by the


coefficient

vertical

force is the tire lateral friction


as

coefficient \iy.

The lateral friction

usually decreases

the vertical load on the tire increases. Beyond the slip angle at which the peak lateral
angles a

force occurs, the lateral force begins to decrease. At high slip


contact patch

larger portion

of

the

is sliding than

at

low slip

angles.

Here the lateral force

produced

depends

largely upon the tire rubber compound,


The
curve shown

the road surface, and the

interface between them.

in Figure 3.2

represents steady-state

tire lateral force

characteristics.

Because

of

the elasticity and

damping inherent in a pneumatic tire it is


a change

actually

dynamic

system within

itself. When

in slip

angle

occurs, the change in


modeled

lateral force lags behind. Although the


an additional

effects of tire

dynamics
for
are

can

be

by including
are

differential equation in the


of

vehicle model

each

tire, the effects

generally

small

below input frequencies

Hz.6

Tire dynamics

typically modeled when


lane
change.

simulating emergency

crash avoidance maneuvers such as a sudden

Tire

dynamics

are neglected

in the

models of this thesis.

13

Chapter 3

Tire Behavior

3.3 Linear Tire Model


As
mentioned above the

initial

slope of the

lateral force

versus

slip

angle curve

for

single vertical

load is the cornering


be

stiffness

Ca of the tire at that load.

Under

certain

conditions this characteristic can

used as a reasonable representation of

tire behavior.

Inspection

of Figure

3.2

reveals

that at small slip angles, the lateral force curve is nearly

linear. Thus

at

sufficiently

small

slip angles, the lateral force

produced

by a tire can be

approximated

by the expression

Fy
where

Caa

(3.1)

c.-Z
a

da

(3.1)
a=0

When lateral force


system.

combined with other assumptions

regarding the vehicle, linearization modeling


of

of

the

versus

slip

angle

relationship
variety
of

permits

the vehicle as a linear

Since there is

a wide

powerful, well-developed analysis techniques for

linear systems,

much can

be learned

about vehicle

lateral dynamics from the study

of a

linear model. The


of simulations of

range of

applicability

of

the linear tire model is examined

by comparison

linear

and non-linear models

in Section 4.6.2.

3.4 Non-Linear Tire Model


When tire slip
angles

become high the linear tire

model

does

not

accurately predict

tire lateral force. At a high slip angle the linear model predicts a force which is higher than the actual tire force. A non-linear tire model is necessary to accurately determine tire lateral

force

at

high slip

angles.

As discussed in Chapter 2, found in the literature. Some


experimentally
measured

several approaches

to modeling tire behavior can be


upon curve

models are

purely empirical, based

fitting of

tire data. Other models are

primarily theoretical,

with some

14

Chapter 3

Tire Behavior

parameters

determined experimentally,

such as

the

stiffness of

the tire. Each type of model

has

advantages and

disadvantages. The type data. This type

of

tire model

used

in this thesis is the former,


because
of its

based entirely

on empirical

of

tire

model

is

used

limited

complexity

and

its suitability to the tire data which is


model chosen

available

to the

researcher.

The tire
was originated

for this study is

called

tire data

nondimensionalization and

by

Hugo

Radt.13,27,30

While this technique is


combined

able to predict tire

aligning
and

moment, longitudinal

force,

and

lateral force for

lateral slip, camber,


of

longitudinal slip, only the lateral force due to lateral slip is


effects of camber on

interest in this thesis. The


assumes a

lateral force
so

are

being ignored, and the vehicle model


necessary to
to have
consider

constant

forward velocity,

it is

not

longitudinal tire force. In this


the
overall

study tire aligning

moments are considered

a negligible effect on

dynamics

of

the vehicle.

There

are

two

main steps

in using the tire data

nondimensionalization

technique.

The first step is preprocessing


model.

experimental

tire data to determine the parameters for the tire

The

second

step is using the


angle.

model to calculate the tire

lateral force for

a given

vertical

load

and

slip

In

a vehicle

dynamics simulation, the first step


second

would

typically

be done before ranning the

simulation.

The

step

would

be done

at each

time step

during the

simulation

based

on

instantaneous

values of

tire vertical load and slip angle.

The tire data


manufacturer

used

for this study is based

on experimental

data provided by the


shown

for

a production passenger vehicle tire.

The tire data is

in tabular form
versus

at

the end of this

section

in Table 3.2 for

and

is

plotted

in Figure 3.3. Lateral force

slip

angle curves are available

vertical

loads

of

2793 N, 4190 N, 5587 N, 6984 N,


each of the vertical

and

8380 N. The slip


at
0

angle varies

from 0 to 15. At

loads the lateral force


and/or

slip

angle

is

not zero as might

be

expected.

This is due to conicity


while

ply

steer

in

the tire.
errors

Conicity

arises

from asymmetries in tire construction,

ply

steer results

from

in the

angles of

the belt cords in the tire. Both conicity and ply steer depend upon

15

Chapter 3

Tire Behavior

quality
to tire.

control

in the manufacturing process


effects are not

and can

be

random

in nature, varying from tire

Since these

important for the

vehicle models under consideration

here, these effects have been eliminated from the experimental data by shifting each of the
lateral force
used curves to the

left until they intersect the

origin of

the plot. This

zeroed

data is

in

all subsequent analysis.

Preprocessing the experimental data is done by normalizing the data and then curve
fitting the normalized data. The first step in normalizing the data is to determine the tire
cornering
coefficient

Cc at each load.
angle

Since the cornering

stiffness

is the initial

slope of the

lateral force
stiffness

versus

slip

curve, and since the cornering coefficient is the cornering

divided

by the vertical load, the cornering coefficient can be approximated at each


1

load

by dividing the lateral force at


data the cornering

slip

angle

by the vertical load. Thus, from the


load is

experimental

coefficient at a single

Cc=^==%1

(3-2)
in Figure 3.4. As
and vertical

The cornering

coefficients at each

load

are plotted

can

be

seen

from the figure, the relationship between cornering coefficient

load is load
can

approximately linear. For this tire the cornering


be
represented as

coefficient as a

function

of vertical

CC=B,
Values
used of

C3FZ
This
expression can

(3.3)
be

the constants

B3 and C3 are listed below in Table 3.1.


an

to predict the cornering coefficient for

arbitrary load

during a simulation.

Next the lateral friction

coefficient

\yy at each load must be found. This is done by


Thus

dividing the maximum lateral force for a given vertical load by the vertical load itself.
for a
single vertical

load,

the lateral friction coefficient is

16

Chapter 3

Tire Behavior

Table 3.1: Non-Linear Tire Model Parameters

Parameter
Tire cornering coefficient intercept Tire cornering coefficient
Tire lateral friction Tire lateral friction
slope

Symbol

Value
0.333

coefficient

intercept

coefficient slope

Magic Formula curve fit parameter

Magic Formula curve fit parameter


Magic Formula curve fit parameter Magic Formula curve fit parameter

B3 C3 B5 c5 B, c, D, E,

1.173

0.5835 1.7166 1.0005 0.2517

Py-

..

'

imax

(3.4)
in Figure 3.5. As
the

The lateral friction

coefficients at each

load

are plotted

with

cornering coefficients, the relationship between lateral friction is approximately linear for this tire. The lateral friction

coefficient and vertical

load

coefficient can

be

expressed as

fly=B5
Values
of

C5Fz
This
expression

(3.5)
is
used

the constants

B5 and C5 are provided in Table 3.1.

during simulation to predict the lateral friction coefficient for an arbitrary vertical load.
With the cornering
the
experimental coefficient and

lateral friction
angle a can

coefficient

known

at each

load for

data,

the

normalized

slip

be

calculated at each

data point from

the

expression

_=Cctan(q)

(3.6)

Similarly,

the

normalized

lateral force F

at each

data point is

F.=-F'

(3.7)

HyZ

17

Chapter 3

Tire Behavior

When the
data point, the
then curve fit.

normalized

lateral force is

plotted against

the normalized slip angle at each


normalized

results

lie

on a single curve as shown

in Figure 3.6. The


to fit this
here.26

data

are

While

various

functions

could

be

used

data,

a popular

function for
a

fitting tire data known as the "magic


combination of trigonometric
of various shapes such as

formula"

is

used

The

magic

formula is

functions

and

has the ability to accurately fit tire data curves


and

lateral force, longitudinal force,

aligning

moment.

The

normalized

lateral force is fit to the function

Fy
where

Dx sin(0)

(3.8)

6
and

Cx atan(fl^)

(3.9)

. _

y/

(l-El)a+

,_

E, ' atan(B,a) v ' ;

(3.10)
provide the

The

parameters

B}, C,, Dn
data. The

and

Ej

must

be determined to

best fit to the


script

normalized experimental

curve

fitting is implemented in the MATLAB

MagicFit.m. This
and uses

script reads

the normalized lateral force versus slip angle data from a file

the MATLAB Optimization Toolbox function

leastsq to do a non-linear least

squares

fit. The
each

leastsq function calls the function MagicError.m which computes the enors
and

between

data point

the curve fit function. The parameters

Bp C7, D,,

and

Et

are

found to
are

minimize

the sum of the squares of these enors. MagicFit.m and MagicError.m


and

listed in Appendix A. 1

Appendix A.2
and the

respectively.

Values for the

curve

fit
with

parameters are given

in Table 3. 1,

function is

plotted

in Figure 3.6 along

the

normalized

data. It
a

can

be

seen

from the

plot

that a good fit to the data has been obtained.


of normalized angle

With
now

function for the

normalized

lateral force in terms

slip

available, the tire lateral

force

can

be

calculated

for any

combination of vertical

load

18

Chapter 3

Tire Behavior

and

slip

angle.

First, the cornering coefficient and lateral friction coefficient are calculated
load using Eq. (3.3)
and

from the

vertical

Eq. (3.5). Second, the


and the

normalized

slip

angle

is

calculated

from the slip angle, the cornering coefficient,


normalized

lateral friction

coefficient

using Eq. (3.6). Next, the


angle

lateral force is

calculated

from the

normalized

slip

using Eq. (3.10), Eq. (3.9),


normalized

and

Eq. (3.8). The tire lateral force


and

can

then be found

from the

lateral force, the lateral friction coefficient,

the vertical load as

Fy=FyHyFz
This
procedure

(3.11)
which

is implemented in the MATLAB function NLTire.m


vertical

is listed

in Appendix A. 3. The function takes the tire


outputs the

load

and

slip

angle as

inputs

and

lateral force. Plots


of

of

lateral force

versus

slip
and

angle

from this function for


are shown

vertical

loads

2793 N, 4190 N, 5587 N, 6984 N,


with

8380 N

in Figure 3.7

as solid

lines along
The

the experimental data

points.

non-linear tire model

implemented in this
versus

section

accurately
of

reproduces the

experimentally determined lateral force


study.

slip

angle

relationship
produced

the tire used in this

This

model

is

capable of

predicting the lateral force


for inclusion in

by the tire at high slip


lateral dynamics

angles.

Thus the tire

model

is

suitable

a model of vehicle

where

high tire slip

angles are obtained.

While this tire

model

only determines lateral force

due to slip angle, it force


and

can

be

extended

to predict aligning moment due to


and

slip angle, lateral


slip.

aligning
should

moment

due to camber,

longitudinal force due to longitudinal

It

be

noted

that due to sign conventions in the SAE tire axis system and in

the SAE vehicle coordinate system, a positive tire

lateral force is

produced

by a negative

slip

angle.

The description

of

the models in this chapter assumed that a positive slip angle


convenience.

produced a positive

lateral force for

However,

when

the tire models are


ensure

integrated into the

vehicle model appropriate care must

be taken to

compatibility

with

19

Chapter 3

Tire Behavior

the sign convention required by the vehicle model. In the linear tire model the cornering
stiffness must

be

negative to meet sign convention requirements.

20

Chapter 3

Tire Behavior

Table 3.2: Experimental Tire Data

SUp Angle (deg)


0
1 2 3 4 5

Lateral Force @ Vertical Load

(N)
6984
-334

2793
-89

4190
-156

5587
-245

8380
-378

737

1001 2024 2847 3403

1223 2558 3603 4359

1357

1458
3025 4404

1388 1935
2358 2647

2869
4115

5026 5649
6005

5449 6183 6672


6950 7110 7166 7180 7153 7111 7050

3803 4026 4175


4241 4258 4246 4222 4168

4849
5182

6
7

2802 2910 2965 2969


2950 2930 2890

5350
5470

6230 6330 6370 6352 6322 6291 6235 6160 6090

8 9
10 11 12 13 14

5490
5450

5400 5350
5282

2840 2780 2750

4099
4037

5200 5121

6978 6900

15

3977

10

Slip Angle (deg)


Figure 3.3: Experimental Tire Data

21

Chapter 3

Tire Behavior

0.35

0.30

TO

0)
TJ

0.25

s^-

c o

0.20

*= CD o

O
O) c
k-

0.15

0)
^

0.10

0.05

2000

4000

6000

8000

10000

Vertical Load

(N)

Figure 3.4: Tire

Cornering Coefficient

1.2

1.0

g>

0.8

'o it=
a) o

O
c

0.6

o
o

it
to

0.4

a
a>

0.2

0.0 2000 4000

6000

8000

10000

Vertical Load

(N)

Figure 3.5: Tire Lateral Friction Coefficient

22

Chapter 3

Tire Behavior

1.2

1.0
CD

|
CD

0-8

5
CD N

0.6

15

0.4
o

2793 N
4190N

A
0.2 X
ill

5587 N

6984 N 8380 N
i II

0.0

II

....

0.0

0.5

1.0

1.5

2.0 Normalized

Slip

2.5 3.0 Angle

3.5

4.0

4.5

Figure 3.6: Tire Normalized Lateral Force

10

12

14

16

Slip Angle (N)


Figure 3. 7: Reconstructed Tire Lateral Force

23

Two Degree-of-Freedom Vehicle Model

4.1 Introduction
Under
The driver
normal

driving conditions the driver and vehicle form a closed-loop system.


inputs to
produce

observes the motion of the vehicle and provides control

the

desired lateral

motion.

However, this work is

concerned

primarily

with

predicting the open-loop

response of a road vehicle to control and

disturbance inputs.
used

The

simplest model which can

realistically be

to

examine

the lateral

response

of a road vehicle

is the two degree-of-freedom (DOF) has been


model used

"bicycle"

model.

As

noted

in
lateral

Chapter 2, this
response.

model

extensively in the literature to study


simplifies

road vehicle

Although this lateral

greatly

the

vehicle

system, much

can

be learned

about vehicle

response

through its use. The model demonstrates the effects of major tire properties, inertia properties, mass center
of practical significance

design

and operational parameters such as

location,

wheelbase, and forward speed. Conclusions

regarding
model.17

road vehicle

lateral directional
chapter

control and

stability

can

be drawn using this

simple

In this

the two degree-of-freedom

vehicle model

is described in detail. The


relationships

equations of motion are

derived from basic

principles of

dynamics. Next,

for

the tire slip angles are derived from the

vehicle

kinematics. From here the

model

is

developed in two forms, linear In


the

and non-linear.

linear form

of

the model,

additional assumptions are made which

simplify the linear

kinematic

relationships and

tire

mechanics.

This

simplification allows powerful

systems analysis

techniques to be used to gain significant

insight into the lateral dynamics

of road vehicles.

Transfer functions for the

response of

the vehicle to steering control,


measures of
steady-

aerodynamic side

force,

and road side slope are

developed. Several

state and

transient

response are

derived. Next, the

frequency response of the vehicle is


is
simulated

examined

using bode

plots.

Finally,

vehicle response

for

variety

of

steering

24

Chapter 4

Two Degree-of-Freedom Vehicle Model

inputs

and

for the disturbance inputs

by integrating the differential equations of motion with

respect to time.

In the

non-linear

form of the model, full


as

non-linear

kinematics
accounts

and a non-linear

tire

model are used.

The tire model,

described in Section 3.4,


versus

for the non-linear,


of the model

vertical

load-dependent lateral force


steering inputs is

slip

angle relationship.

Simulation

response to

performed and

the results are compared to the linear model

simulation.

4.2 Description

of

Model
used

The two degree-of-freedom model


vehicle

in this

chapter

is

shown

in Figure 4.1. The


v and yaw

is
r

modeled as a single

lumped

mass

rigid body
u

and

has lateral velocity

velocity

degrees

of

freedom. The forward velocity


the vehicle is represented

is

assumed

to be constant. The pair

of tires at each end of

by a single tire at the centerline of the car.


Ia,
with

The

vehicle

has

a wheel

base L,
a

a mass

m, and a yaw mass moment of inertia


axle and a

its

mass center

located
of

distance

front the front

distance b from the

rear axle.

Rotation

the front tire about the vertical axis relative to the

body is permitted and


The

is

measured

by the front steer angle 8, with clockwise rotation considered positive.


is the only
control control

front

steer angle

input considered. In this


as

work position control

is

assumed.

Position

is defined by SAE

"that

mode of vehicle control wherein

inputs

or restraints are placed upon

the steering system in the form of displacements at

some control point

in the steering

system

(front wheels, Pitman arm, steering wheel),


contrast

independent

of

the force

required."29

This is in

to force control where inputs are in

the form of forces or

moments and are

independent

of

displacement.
x-y-z

The
motion of

standard

SAE

vehicle-fixed coordinate system

is

used

to describe the

the vehicle. The


z-axis

x-axis

is

positive

in the forward
origin

direction, the y-axis is positive


coordinate system

to the right, and the

is

positive

down. The

for the

is

at

the

25

Chapter 4

Two Degree-of-Freedom Vehicle Model

Figure 4. 1 : Vehicle Model


vehicle mass

center, and the coordinate system translates and rotates with the vehicle.
permitted

Motion is only

in the x-y

plane.

4.2. 1 Assumptions
Several simplifying
model:

assumptions are made to

facilitate the development of the

Constant vehicle

parameters

Constant vehicle forward velocity


Motion in x-y
mass)
plane

only (ignore vertical, rolling,

and

pitching

motions of

sprung

Vehicle is rigid

26

Chapter 4

Two Degree-of-Freedom Vehicle Model

Vehicle is
Road

symmetrical about z-x plane

surface

is

smooth

Ignore

effects of

drive line (ignore in longitudinal

Ignore longitudinal gravity


Ignore
all

effects

road slope

direction)

longitudinal forces (tire

driving/braking forces, tire roiling resistance,

aerodynamic

drag)
kinematics
and

Ignore

suspension system

dynamics

Ignore steering Position

system

kinematics

and

dynamics

control

for steering input due to due to


roll of

Ignore tire Ignore tire

steer

sprung

mass

steer

chassis compliance

Ignore tire slip Ignore lateral


Tire

angles

resulting from lateral tire

scrub

and

longitudinal load transfer (vertical tire forces independent of time


and

remain

constant)

properties are

forward velocity
and

Ignore tire lateral forces due to camber, conicity, Ignore tire aligning Ignore
effect of

ply

steer.

moments

longitudinal tire slip

on

tire lateral force

Ignore tire dynamics (no Ignore tire deflections

delay in lateral force generation)

4.2.2 Vehicle Parameters


The
model

nominal values of

the vehicle parameters used

for the two degree-of-freedom


of metric units

in this study

are given

in Table 4.1. The International System

(SI) is

27

Chapter 4

Two Degree-of-Freedom Vehicle Model

used

for

all calculations.

The base

units are meter

(m), kilogram (kg),

and second

(s).

Force is
on

measured

in the derived

unit newton

(N). For convenience, the fraction

of weight

the front axle/is used to define the


mass center

position of the mass center

along the

wheelbase.

The

location parameters

a and

are then calculated

from/ and L.

These

parameters are representative of a production automobile.

Table 4.1: Vehicle Parameters

Parameter Vehicle
Yaw
mass

Symbol
m

Value
1775

kg
kg-m2

moment of

inertia
fraction

'
f
L
to front axle
a

1960 0.52 2.372

Front

axle weight

Wheelbase
Distance from
mass center
center

m m m m

1.139
1.233 1.25

Distance from mass


Distance from front

to

rear axle

b
force
c

axle to aerodynamic side

4.2.3

Free-Body
There
are

Diagram

three types of external forces acting on a vehicle which are considered in


aerodynamic side

this model: tire

lateral forces,

force,

and gravitational side

force. The tire


a

lateral forces i^and Fyr

occur

due to tire slip


c

angles.

The

aerodynamic side

force F a is

disturbance input acting


vehicle when

at a

distance

behind the front


The

axle.

This type

of

force
a

acts on a

it

encounters a crosswind.

gravitational side

force F is

disturbance
road.

input acting

at

the

vehicle mass center and

resulting from

a side slope

in the

All forces
are

are considered

to be the

positive when

acting in the

positive y-direction.

These forces

shown

acting

on

vehicle

in Figure 4.2.

28

Chapter 4

Two Degree-of-Freedom Vehicle Model

Figure 4.2:

Free-Body Diagram
Motion
vehicle are

4.3 Derivation
The

of

Equations

of

equations of motion

for the two degree-of-freedom for rigid body

derived using

basic

principles of Newtonian mechanics

motion relative to

translating and
acting

systems.31

rotating
on a

coordinate

The basic

equations

relating the forces

and moments

rigid

body to the acceleration of the body are


If
=g

Xmg=hg
where

(4.1)

and

Mg are external forces and moments about the mass center (in vector form)
body,
and

acting
vector

on

the

and

HG are the linear and angular momenta of the body (also in

form) measured relative to an inertia! reference frame. Since in this model only

29

Chapter 4

Two Degree-of-Freedom Vehicle Model

motion

in the x-y Eq.

plane

is

considered and all

longitudinal

(x-direction) forces

are

being

ignored,

(4.1) become
ma

(4.2)

Since the x-y-z


mass

coordinate system

is fixed to the

vehicle with

its

origin at

the vehicle
of

center, the translational velocity

of

the vehicle mass center and

rotational

velocity

the vehicle are identical to those of the x-y-z system. From Figure 4. 1, the velocity the origin of the x-y-z system is

V0 of

V0=wi + vj
and the angular

(4.3)

velocity Q. is Q
=

rk

(4.4)
changing
with

Since the

x-y-z system

is rotating, the

unit vectors are

time. Thus the

acceleration

a0

of

the origin expressed in

an

inertial

reference

frame

coincident with

the

x-y-z system

is

dV.
a
=

+ ilxV

dt
=

(4.5)
(v + ur)j
to the inertial frame is

(u- vr)i +

Similarly, the angular acceleration of the x-y-z

system relative

+ Q.XQ.

dt
=

(4.6)

rk

Thus the

acceleration values of

interest

are

=v

+ ur

(4.7)
Qz=r
These
values

apply both to the

vehicle-fixed

x-y-z coordinate system and to the vehicle

mass center.

30

Chapter 4

Two Degree-of-Freedom Vehicle Model

The

external

forces acting

on

the vehicle are shown in their

positive sense

in

Figure 4.2. A

positive steer angle results

in

positive

tire lateral forces. From the

free-body

diagram, it is

seen that

Fv

=Fyf cost
=

Fyr + Fya+Fy yg bFyr


-

(4.8)
ya

X Mz
Substitution
of

aF^
Eq.

cos 8

(c

a)Fy

Eq.

(4.7)

and

(4.8) in Eq. (4.2)

yields

the equations of motion

for the two degree-of-freedom vehicle:

F^ cos 8 + Fyr + Fya +Fyg=m(v + ur)


aF^ cos 8
Here
u
-

(4.9)

bFyr

(c

a)Fya is

Ia r
The
state variables are

is the
v and

vehicle

forward velocity
r.

and

a constant.

the

lateral velocity

the yaw velocity

F^ and Fyr are the front and rear tire lateral forces.
and gravitational side

Fya and F
while

are

the aerodynamic side force

force disturbance inputs,

the control input is the front steer angle

8.

4.4 Derivation

of

Tire

Slip

Angles
produced

As discussed in Chapter 3, the lateral force among


the
other

by a tire

depends upon,

things, the

vertical

load

on

the tire and the slip angle of the tire. In this model
and rear

vertical

load remains constant, but the front


yaw

tire slip

angles

vary

as

functions

of

the lateral velocity v and the

velocity r,

which are

the system state variables. Thus it is

necessary to

develop expressions

for the front

and rear

tire slip angles,

af and ar, in terms

of v and r.

Figure 4.3 is
and

kinematic diagram

of

the vehicle

showing the tire velocity

vectors

tire slip angles. Each slip angle is shown in its positive sense as the angle between the
vector.

tire and the tire velocity tire to its velocity

positive

slip

angle

implies

a clockwise rotation

from the

vector.

However, for a positive

steer

input

as shown

in the figure, the tire

31

Chapter 4

Two Degree-of-Freedom Vehicle Model

Figure 4.3: Kinematic Diagram velocity


vector

is actually

a counter-clockwise rotation

from the tire,

so

slip

angles are

negative.

In summary,

a positive steer

input

results

in

negative tire

slip

angles.

The first step in


the translational and

determining the tire slip angles is finding the tire velocities.

Since

rotational velocities of

the vehicle-fixed x-y-z coordinate system are


convenient

already known from Eq.


motion

(4.3)

and

Eq. (4.4), it is The

to

use

the principle

of relative

to derive the tire


are

velocities.31

velocities

Vf and Vr of the front and rear tires

respectively

Vf=V0+QxRf
V=V+fixR

(4.10)

where

R, and Rr are position vectors from the vehicle mass center to the front and rear tires:

32

Chapter 4

Two Degree-of-Freedom Vehicle Model

Rf
R,
Thus the tire
velocities are

=a\

(4.11)
=-b\

V. f

u\

ar)\ + (v V + '

Vr
In general, if the velocity

u\ +

(v

(4.12)
-

br)j
ith tire
are

and steer angle of the

known,

the slip angle is

fvA
a,
=

atan

-8,.

(4.13)

\Y*j
where

Vx and Vy are the

x-

and y-components of

the velocity of the tire. Since steer of the

rear

tire is not permitted in this model, the tire slip angles are, in their general non-linear

form,
^

a/=atan(

(v + ar\

j-S
(4'14)

ar

atan

fv-bA
V
u

4.5 Linear Model


Thus far the
general non-linear equations of motion and

tire slip

angle relationships

have been derived for the two degree-of-freedom vehicle. If additional


made

assumptions are

to linearize the model, analysis techniques for linear systems may be used to gain

more

insight into

road vehicle

lateral dynamics. In this section, tire slip

angles and

lateral

forces

are assumed

to be linear functions. Other research has shown that these assumptions

are valid

for

vehicle

lateral

accelerations

up to

about

0.35 g,

which corresponds to the

linear

range of the

tire lateral force versus slip angle

relationship.13

Most

automobile

driving is

done

within

this range, so results from the linear model are applicable over a wide range of

driving

situations.

33

Chapter 4

Two Degree-of-Freedom Vehicle Model

It is
models, to

common

in modeling

of vehicle

lateral dynamics, particularly

with

linear

use

the vehicle sideslip

angle

p instead of the lateral velocity v to describe the


sideslip
angle

lateral

motion of the vehicle.

The

vehicle

is the

angle

between the

vehicle-

fixed x-axis
Eq.

and

the vehicle velocity

vector

Vn

as shown

in Figure 4.3. Similar to

(4.14)

for tire slip angles, the

vehicle

sideslip

angle

is

p
A
positive vehicle

= atari-

(;)
may be
positive or

(4.15)
from the x-axis to the velocity
negative,

sideslip

angle

implies

clockwise rotation

vector.

For

a given steer

angle, the sideslip

angle

depending
lateral

upon

the forward
when

speed.+

In this thesis the

vehicle

sideslip

angle

is

used

in

place of

velocity

finding transfer functions,

steady-state response measures, transient

response

measures, and

frequency response. However,


lateral velocity

since simulation of

the non-linear
of the

model

is

more straightforward with

as a state

variable, simulation

linear model is

also performed with

lateral velocity

as a state variable to

facilitate parallel

development

of

the two simulation models.


of motion are

Once the linearized equations


variables

written, transfer functions for the state

in terms

of

the control and disturbance inputs are

developed. From these transfer

functions,
response

measures of steady-state and

transient response are derived and the

frequency

is

examined.

Simulation

of

the model is performed for various steering and

disturbance inputs.

4.5.1 Additional Assumptions


The
addition

following assumptions are made for the linear two degree-of-freedom model in

to those listed in Section 4.2.1:

See Section 4.5.18 for

more

information.

34

Chapter 4

Two Degree-of-Freedom Vehicle Model

Linear tire lateral force versus slip


Small
steer

angle

relationship

angle, tire slip angles,

vehicle

sideslip angle,

and road side slope angle.

4.5.2 Vehicle
With the

Sideslip Angle
small angle

assumption, Eq.

(4.15) for the vehicle sideslip

angle

becomes

Pu

(4.16)

However, if the vehicle sideslip angle is

small then

cos(P)

i^l
(4.17)

or u~V

where

Vis the

magnitude of the vehicle

velocity V0. Now Eq. (4.16) becomes

P~"
4.5.3 Tire

(4-18)

Slip

Angles
assumption, the tire slip
v + ar

With the

small angle

angles

become

af
a

-8

u v
=

br
u

(4.19)

Furthermore, if vehicle sideslip angle is used in place


angles can

of

lateral velocity, then the tire slip

be

expressed as

af K f =P

r-8

b
r=P--r

(4.20)

35

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.5.4 External Forces


From the
seen

and

Moments

free-body diagram of the two degree-of-freedom model in Figure 4.2 it is


on

that the four external forces acting

the

vehicle are

the front tire lateral force

F^ the

rear tire

lateral force Fyr, the

aerodynamic side

force Fya,
are

and the gravitational side

force Fyg.

With the linear tire

assumption that tire

lateral forces
employed.

linear functions

of

tire slip angle, the

model of

Section 3.3

can

be

From Eq.

(3.1)

the tire lateral forces are

Fyf *
Fyr
where

Cfaf ' '


^rr
effective

(4.21)

_ ~

C/and Crare the front


stiffnesses of both

and rear

tire cornering stiffnesses and are the

cornering

tires

on an axle.

Thus, for example, Cfis

twice the cornering

stiffness of a single

front tire. As

result,

F^ and Fyr are the sums of the tire lateral forces


angles are negative

of

both tires

on an axle.

Since the tire slip

for

a positive steer

angle, the

cornering
required

stiffnesses must also

be

negative

in

order

to produce the positive lateral

forces

by the

sign convention.

For further

explanation of

this tire model see Section 3.3. though the

Values for tire cornering


application of

stiffnesses

for the

vehicle studied are obtained

Eq.

(3.3)

and are given

in Table 4.2.

Table 4.2: Linear Tire Model Parameters

Parameter Front tire cornering Rear tire cornering


stiffness

Symbol
(two tires)

Value
-2461

C~f
Cr

N/deg N/deg

stiffness

(two tires)

-23

11

The

aerodynamic side

force

Fya is in general a function of the relative air speed


area.32

squared, the side force coefficient,

and a reference

However, for simplicity the

side

force itself is

used as

the disturbance input to the system. The aerodynamic side force is


on

positive when

acting

the

vehicle

in the

positive y-direction.

36

Chapter 4

Two Degree-of-Freedom Vehicle Model

Figure 4.4: Gravitational Side Force The


shown

gravitational side

force

Fyg is a function of the side slope in the road and is


force is
positive when

in Figure 4.4. The

gravitational side

acting

on

the vehicle in

the positive y-direction. Thus the gravitational side force can be expressed as

Fyg
where

=mg sinQ
and

(4.22)
road side

g is the

acceleration

due to gravity
on

9 is the

slope, which is positive for

a road which

is sloping down

the right side of the vehicle as shown in the figure. If the

assumption of a small road side slope angle

is used, then Eq.


=

(4.22)

simplifies to

F
With the
can now gravitational side

mgQ

(4.23)
of

force

expressed

in terms

the road side slope, the side slope 8

be

considered

to be the disturbance input instead of the force itself.

4.5.5 Equations
With Eq. (4.21),

of

Motion
the tire slip
angles

substitution of

from Eq. (4.20), the tire lateral forces from

and the gravitational side

force from Eq.

(4.23) into Eq. (4.9),

the linearized

equations of motion

become

37

Chapter 4

Two Degree-of-Freedom Vehicle Model

(Cr+Cr)p + ^
C*

r-r-Cf5 + Fya+mgQ
hi
r
-

mV$ + mVr

rt

-U

(aCf
The

bCr )p

f-

aCf8

(c

a)Fya =1J
been
to the

assumption that the steer angle

is

small

has

also

applied

reduce

equations to the above

form. This

assumption

is generally

valid

for maneuvers

at moderate

to high speeds. For very low speed maneuvers, such as parking, large
often required.

steer angles are

To simplify

manipulation of

the equations of motion, the


of

external

force

and moment

terms of the left sides of Eq.

(4.24) can be rewritten in terms

stability derivatives. This

technique has been used extensively


such as

by early researchers in automobile lateral dynamics


and

Leonard Segal, David Whitcomb,


equations of

William

Milliken.6,7,13

In

addition

to

simplifying the
which can give

motion, the derivatives themselves have


road vehicle

physical

meaning

further insight into

lateral dynamics.

The stability derivatives


moments

are

the rates of change of the external forces or external


to

acting

on

the

vehicle with respect

p,

r, or 8. There are three stability derivatives


with yaw moment.

associated with

lateral force

and

three associated
are

The

equations of

motion

in stability derivative form

Fpp + Yrr + Ys8 + Fya

+ mgQ

mV$ + mVr

(4'25)

Nfi
where

Nrr + N6S-(c-a)Fja=I,tr

the stability derivatives are defined as follows:

38

Chapter 4

Two Degree-of-Freedom Vehicle Model

V cf+cr
=
-

Damping

in

Sideslip Coupling
(4.26)

Yr

aCf
V
=

bCr

Lateral Force/Yaw

Y5

~Cf

Control Force
Directional

NP

aCf-bCr a2Cf+b2Cr
V

Stability

Nr-

Yaw

Damping

Ns-=

-aCf

Control Moment
model under consideration the

In the two degree-of-freedom


all constants.

stability derivatives

are

As such, the

equations of motion can

be

manipulated

in stability derivative

from without loss

of generality.

By noting that the tire cornering stiffnesses Cf and Cr are


definition,

always negative

by

the signs of the stability derivatives can be obtained. The


and yaw

damping-in-sideslip

derivative Yj, derivative Ys

damping derivative Nr are always negative. The control force


derivative Ns
are always positive.

and control moment

The lateral force/yaw

coupling derivative
negative

Yr and directional stability derivative A/j, are both either positive or


If

depending on the relative magnitudes of aCf and bCr.


derivatives
are positive and

aCf is

greater than the

bCr,

then the

the vehicle understeers. If

aCf is less than bCr,

then the

derivatives

are negative and

the vehicle is oversteer. If the terms are equal, the

derivatives

are zero and

the vehicle is neutral steer.


more

Understeer,

oversteer,

and neutral steer are

discussed in

detail in Section 4.5.14.

4.5.6 Transfer Functions


Now that the
equations of motion are available

in

simple, compact

form,
and

transfer

functions

can

easily be found relating the

outputs

p and r to the inputs 5, 6,

From
can

these inputs and outputs six transfer


used to examine

functions

can

be formed. The transfer functions


response,

be

many

aspects of system response such as steady-state

frequency

39

Chapter 4

Two Degree-of-Freedom Vehicle Model

response, and poles


expressions

and zeros.

The derivation

of

the transfer functions and all analytical

for measures

of system response

is done using Mathematica. The Mathematica

session

for the two degree-of-freedom vehicle is included in Appendix B. To find the transfer functions the
equations of motion are

first

written

in the Laplace

domain assuming that the initial

conditions are zero:

s--

1
mV
s--

( Y \
mV

1
mV
a c

> (g\

mV

Na

N.

m+

As);
equations of

Fjs)

W
yOj

(4.27)

From these Laplace-domain

motion, the transfer functions for vehicle

sideslip

angle are

found to be

75

NrYs+N8(mV-Yr)
i*v

!-

mV

N~

Y\ + !W+N9(mV-Yr)
LmV
s+
-

(4.28)

\^+'mVj 1

(c-a)(mV-Y)-Nr ^ T-i-

P
ya

mV

(')

I^mV
I

s2-

Y\{N^+N*(mV-Yr)
mV

(4.29)

I,mV

8S
Ya \
s2-

8Nr

(4.30)
^NrY^N^mV-Yr)

mV

IzjnV
are

Similarly,

the transfer

functions for yaw velocity

N
I
s-

8j|

Va-^P
I,mV
"i

s2-

N.
i

N^ + N^mV-Y,)
IumV

(4.31)

v'

mV

40

Chapter 4

Two Degree-of-Freedom Vehicle Model

a-c

s+

+ yp(c-fl) iVp p
-

J-(S)

I*

^mV
,

(4.32)

\*zz

ntV

ImV zz

gN L(s)
r

IV
=

is-l
2

BK)

(4 33) IjnV
of

(Nr

-H

Y,]5-1 NrY,+N,{mV-Yr)
mVj
the two degree-of-freedom

U
The
vehicle can

above transfer

functions for the lateral dynamics

be

used

to examine steady-state behavior. For each of the three types of inputs,


yaw

the steady-state step input response gains in vehicle sideslip angle,

velocity, tire slip

angles,

path

curvature, and lateral acceleration are found. The steer angle required to
turn radius is calculated. In addition, measures of steady-state
vehicle

produce a given

behavior

such as understeer

gradient, stability

factor,

neutral steer

point, static margin,

tangent speed, critical speed, and characteristic speed are defined and expressed in terms of the stability derivatives.

The found for

steady-state

step

response of vehicle

sideslip

angle and yaw

velocity

can

be

each of

the three inputs

by applying the Final Value Theorem to the transfer


however,
the stability of the

functions.33

Before the Final Value Theorem can be applied, be


verified.

system must

The

system

is

stable

if none

of

the poles have positive real parts.

Expressions for the

system poles are

derived in Section 4.5.24.

4.5.7 Vehicle

Sideslip Angle
of the

Gain
vehicle

Application

Final Value Theorem to the


aerodynamic side

sideslip

angle transfer

functions for
Eq. (4.29),

steer

angle,

force,
the

and road side slope

(Eq. (4.28),

and

Eq.

(4.30)

respectively)

gives

following steady-state response gains:

41

Chapter 4

Two Degree-of-Freedom Vehicle Model

P
8

NrY&+N&(mV-Yr) NrYA+N9(mV-Yr) (c-a)(mV-Yr)-Nr NrY+NJmV-Yr)


mgNr

(4.34)

P
ya

(4.35)

(4.36)

NrYf N9{mV-Yr)
Results for the sideslip
are given

angle response gains and

for the

response gains

that follow

in Table 4.3. The


and

vehicle and tire parameters used

for the

calculations are given

in

Table 4. 1

Table 4.2,

and

the vehicle forward speed is 100 km/hr.

4.5.8 Yaw

Velocity

Gain

Similarly,

the yaw velocity steady-state response gains are

"&-*&
NrY9+N,(mV-Yr)
rp(c-a) + JVp
ya

(4.37)

NrYf+N,{mV-Yr)
mgN?

(4.38)

NrYf+N9{mV-Yr)

(4.39)

4.5.9 Front Tire


Once
steady-state

Slip

Angle Gain
sideslip
angle and yaw

steady-state values of vehicle

velocity

are

known the

tire slip angles can be found using Eq. (4.20). The front tire

slip

angle gains

are

oc<
=

a(N9Ys-NJA)-V(mV-Yrpt+Nh)-VNr(YR +

Yh)

(4.40)

v(ivryp+ivp(mv-yr))

42

Chapter 4

Two Degree-of-Freedom Vehicle Model

a,
ya

V(c

a)(mV

-Yr)

a(c-

a)7p

aiVp

VNr

(4.41)

v{NrY^N^mV-Yr))
a,
e

mg{aN[,-VNr)

(4.42)

v(NrY^N9{mV-Yr))

4.5.10 Rear Tire

Slip

Angle Gain
are

Similarly, the steady-state rear tire slip angle gains

(4.43)

v(ivryp+ivp(mv-yr))
V(c
_ ~ -

a)(mV

-Yr)-b(c-

a)Y&

WVp VNr
-

Fya
O^
e

(4.44)

v(ivrrp+;vp(mv-yr))
-mg^+VN^
~v(NrYB +

(4-45)

NB{mv-Yr))

4.5.11 Path Curvature Gain


Another response
measure of

interest is the

curvature of the path that the vehicle

follows

when subject

to one of the inputs. The path curvature 1/R

is the

reciprocal of the

path radius and can

be found

by dividing the yaw velocity by the vehicle velocity.


r

1/R

V The
path curvature gains

(4.46)

for

each of the

inputs

are

1/R

N Y

-N

(4.47)

v(ivryp+;vp(mv-yr))
yp(c-a) + iVp
(4.48)

1/R
ya

v(NrY{i+N{i(mV-Yr))

43

Chapter 4

Two Degree-of-Freedom Vehicle Model

1/R

mgN,

(4.49)

v(Arryp+ivp(mV-yr))
Step Response
Gain
performance.

4.5.12 Lateral Acceleration Steady-State


Lateral
Lateral
acceleration

is

an

important

measure of vehicle

cornering

acceleration

is typically

expressed

in

units of

"g". The linear

vehicle model

being

considered

here is

valid

for lateral

accelerations

up to approximately 0.35

g.

Beyond that

level,

non-linearities, particularly in tire lateral force mechanics, become

significant.

Most
able

passenger car

driving is done below this limit,


of

although a

typical

passenger car

may be

to attain maximum lateral accelerations

approximately 0.7-0.8 g

with standard street

tires. Race cars without the aid of aerodynamic downforce

reach over

1.2 g

with special

tires designed specifically for racing, while race cars with aerodynamic downforce have

been known to
expressed

exceed

4 g lateral

acceleration.

The

steady-state

lateral

acceleration

Ay

in

"g"

units of

is
rV

Ay
The lateral
acceleration gains

(4.50)
8

for each

of

the inputs are

V(^YS-N&YP)
8

(4.51)

g(ivryp+ivp(mv-yr)) v(Y[i(c-a) + Nfi)

A,
ya

(4.52)

g(ivryp+ivp(mv-yr))
=

mVJVp

NY+NJmV-Yr)

(4.53)

44

Chapter 4

Two Degree-of-Freedom Vehicle Model

Table 4.3: Steady-State Response Gains


V=

100 km/hr

Response
Vehicle sideslip
Yaw velocity
angle

Steer Angle

(6) (P)
/

Aerodynamic Side Force (FJ 2.82xlO"4o/N

Road Side Slope (9)


0.0615/

-1.52

(r)
angle

0.197

rad/s
1

rad/s

/N

2.15xl0'4rad/s/

Front tire slip


Rear tire slip Path

(af)

-2.05

2.66xlO"4/N
1

0.0620 0.0609

angle

(ar)

-2.02

3.00xl0'4o/N
-2.54xlO-7l/m/N

curvature

(1/R) (Ay)

7.10xl0-3l/m/

7.76xl0-6l/m/

Lateral

acceleration

0.559

g/

6.11xl0-4g/
-2.00xl05g/N

4.5.13 Steady-State Steer Angle


The
R
can

steady-state steer angle required

for the

vehicle

to turn at a constant path radius

be found

by solving the yaw velocity gain for steer angle input expression of
steer angle

Eq.

(4.37) for the

8. After

some

manipulation, the
mV2Na

steer angle

is

=- +

(4.54)

/?(/vpy5-/v8yp)

Substitution

of

the stability derivative definitions into this expression yields

f
"

b
C

nYL
L R

(4.55)
steer angle can also

Examination
expressed

of

the

kinematics
of

of a

turning

vehicle

indicates that the

be

in terms

the tire slip angles

as17

8
From Eq.
angle

ar +ar

(4.56)
velocity
approaches zero the steer

(4.55) it can be

seen

that as the vehicle

becomes
L
Ac ker

(4.57)

45

Chapter 4

Two Degree-of-Freedom Vehicle Model

This

steer angle

is known

as the

Ackerman
at

steer angle and

is the

steer angle required

to

negotiate a turn of path radius

low

vehicle velocity.

At low

speeds

the lateral

acceleration, tire slip angles,


governed

and

tire lateral forces are

negligible and

turning behavior is

solely

by geometry considerations.
in this study the Ackerman
steer angle

For the 2.72 . The

vehicle used

for

50

m radius

turn is

steer angle required

to negotiate the turn at 100 km/hr is 2.82 .

4.5.14 Understeer Gradient


A
common measure of vehicle

turning behavior is

the

understeer

gradient, or

understeer

coefficient, K^. The

understeer

gradient, assuming a constant radius


acceleration.1

turn, is
of the

basically the rate of change of steer angle with lateral


understeer gradient can

The definition

be

seen

in the

steady-state steer angle expression

for

a constant

radius turn

from Eq. (4.54):


L
V2

Sss=- + Kus

(4.58)

gR

Thus the

understeer

gradient, expressed in

units of

radians, is

K
"*

mgNa ^N Y ivp-i5
-

N Y

(4.59)

JV6-p

or

substituting the stability derivative definitions


a

mg

(4.60)

The
vehicle

understeer gradient

is

a measure of the amount of understeer or oversteer a

has. If to

maintain a constant path radius

the steer angle must increase as vehicle


understeer gradient

forward velocity,
positive and

and

hence lateral acceleration, increases, the

is

the

vehicle

is

said

to be

understeer.

If the

steer angle must

decrease, the
does
not

understeer gradient

is

negative and

the

vehicle

is

oversteer.

If the

steer angle

46

Chapter 4

Two Degree-of-Freedom Vehicle Model

change the understeer gradient

is

zero and

the vehicle is

neutral steer.

The

understeer of a

vehicle can also

be

obtained

from relationship between the front


greater

and rear

tire slip

angles.

If If

the magnitude

of

the front tire slip angle is

than the rear, the

vehicle

is

understeer.

the rear tire slip angle is greater, the


vehicle

vehicle

is

oversteer.

If the slip

angles are

equal, the

is

neutral steer.

For this

simple model the understeer gradient

is

function

of tire

cornering influence the

stiffnesses and the weight

distribution. Other

vehicle characteristics which

understeer gradient

include lateral load transfer distribution,

roll

camber, roll steer, lateral


system

force

compliance

steer, tire aligning moments, tire longitudinal

forces, steering

compliance, and differential type.

The

understeer gradient of a vehicle can

be

measured experimentally.

The two
test.17

most

common methods of

testing

are

the constant radius test and the constant speed


of

In the

constant radius test the

forward velocity
steer angle

the vehicle is varied as the car is driven on a to maintain the constant radius. The

constant path radius.

The

is
as

varied

understeer gradient

is then calculated

the rate of change of steer angle with lateral

acceleration:

K.=
y

d8

(4.61)

In the
path radius

constant speed

test the vehicle forward velocity is held constant while the

is

varied and

the steer angle required to attain the radius is measured. The


constant speed

understeer gradient

from the

test is

d8

gL

K~
Most
designed into
conditions.

M,-^ between 1
and

(462)
10 deg. Understeer is

passenger cars

have

understeer gradients

passenger cars

to assure directional stability over a wide range of operating

The

vehicle used

for this

example

has

an understeer gradient of

0.0626 deg,

47

Chapter 4

Two Degree-of-Freedom Vehicle Model

indicating that the vehicle has a very small amount of understeer in the linear response
range.

4.5.15

Stability
A

Factor
cornering behavior is the stability
factor.6

second measure of vehicle

The

definition

of

the stability factor comes from the yaw velocity gain for steer angle input

expression of

Eq. (4.37). This

expression can

be

rewritten as

VIL l+
where
KV2

(4.63)
for K and substituting Eq. (4.37),

the K is the stability factor.

Solving Eq. (4.63)


mN

K=

(4.64)

l(n,y5-nsy?)
or

substituting the stability derivative definitions,

L2

(4.65)

If K is

positive

the

vehicle

is understeer,
4.69xl0"5

while

if K is

negative

it is

oversteer.

The

sample vehicle

has

stability factor

of

s2/m2.

Since the stability factor is positive,

the vehicle is

understeer.

4.5.16 Neutral Steer Point


The
force
can neutral steer point

is "the

point

along the

chassis at which an external

lateral
there

be

velocity."13

applied which produces no steady-state yaw

On

a real vehicle

is actually

a neutral steer

line, but for this

simple model which

does

not account

for body

roll and roll steer effects

there is only a neutral steer point. To find the neutral steer point a

transfer function can be

written

relating the

yaw

velocity to
axle.

fictitious lateral force Fm

applied

to the

vehicle at a

distance d behind the front

As in Section 4.5.8 the Final

48

Chapter 4

Two Degree-of-Freedom Vehicle Model

Value Theorem lateral force.

can

be

used

to find the

steady-state yaw

velocity

gain

for the fictitious

Y^d-a) + N^

(4.66)

NrY^N,{mV-Yr)
Setting this yaw
velocity
gain equal

to

zero and

solving for d,

= a-^

(4.67)

Yt
Substituting the stability derivative definitions, the neutral steer point is located at a distance
behind the front
axle of

-^-

(4.68)
is understeer, has if

Cf + Cr
If the
neutral steer point

is behind the
vehicle

mass

center, the

vehicle

while

it is in front
steer point

of

the mass center the


at a

is

oversteer.

The

sample vehicle

a neutral

located
m

distance

of

1.149

behind the front

axle.

Since the

mass center

is

located 1.139
the vehicle is

behind the front axle, the

neutral steer point

is behind the

mass center and

understeer.

4.5.17 Static Margin


Another way
of

describing understeer is with the static


mass center and

margin.

The

static margin

SM is the distance between the


wheelbase.13

the neutral steer point, normalized

by the

SM

--

^p

(4.69)

Lip
or, substituting the stability derivative

definitions,
aCf
=
7-i

SM

bCr

r-r

(4.70)

L(Cf+Cr)

49

Chapter 4

Two Degree-of-Freedom Vehicle Model

If the

static margin

is positive, the
has

vehicle

is

understeer.

If it is

negative

the

vehicle

is

oversteer.

The

sample vehicle

a static margin of

0.00428, indicating that the vehicle

is understeer.

4.5.18 Tangent Speed


At low front
that
axle.
speed

in

a steady-state

turn the rear axle travels on a smaller circle than the

As

speed

is increased the

radius of travel of the rear axle

increases

and surpasses

of

the front. Thus at high speed the rear axle travels on a larger circle than the front in a

steady-state turn.

The

speed at which the

front

and rear axles

travel on circles of the same


angle

radius

is

called the

tangent speed Vmn. At the tangent

speed the vehicle

sideslip

is

zero.

For

a right

turn, below the tangent

speed the vehicle

sideslip

angle

is positive,

while

above the

tangent speed it is negative. The tangent speed can be found

by multiplying the

steady-state

sideslip

angle gain

for

steer angle of

Eq.

(4.34) by the

steady-state steer angle

for

a constant radius rum of

Eq. (4.54), setting that

product equal to zero, and

solving for

the vehicle velocity V.

Vtan=

&r

rS

mNs

(4.71)

However,
of

since the yaw

velocity stability derivatives

Yr and Nr are both functions

velocity, it is necessary to substitute the stability derivative definitions and solve for V.
speed

When this is done the tangent

is

vm=4 V
The
sample vehicle used

bLCr
am

(4-72)

in this study has

tangent speed of 49.8 km/hr. Below

this speed the vehicle sideslip angle is positive for a right turn. Above this speed it is
negative.

50

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.5.19 Critical Speed


Consideration
reveals that the yaw
of

Eq.

(4.37) for steady-state yaw velocity gain for steer angle input
gain could

velocity

become infinite if the denominator

were zero.

The

speed at which this

happens is
of

called

the critical speed. The critical speed can

be found by
of

setting the denominator stability

Eq.

(4.37) equal to zero and solving for

V. In terms

the

derivatives,

the critical speed is

NaYr-NY

Vcril
However,
the stability derivatives
substitute the

"

"

(4-73)

mNp

Yr and Nr are functions

of velocity.

Therefore, it is
critical speed

necessary to is then

stability derivative definitions

and solve

for V. The

CfCL2

v'-=-l^K^)
At this infinite
yaw

(4'74)

speed

the

vehicle

is

unstable.

small

steering input produces

theoretically

velocity

response.

Given that the tire cornering

stiffnesses are negative

by

definition, it can be
than

seen

from Eq. (4.74) that

a critical speed

only

exists

if bCr is is

greater

aCf In Section 4.5.5 it is


a vehicle

noted that when this condition exists the vehicle

oversteer.

Thus

has

a critical speed

only if it is

oversteer.

When

an oversteer vehicle reaches

its

critical speed

it becomes
The

directionally unstable. The more oversteer a vehicle has, the


critical speed

lower its

critical speed.

for

a neutral speed vehicle

is infinite,

and the

critical speed

does

not exist

for

an understeer vehicle.

Since the

sample vehicle

is understeer, it does

not

have

a critical speed.

4.5.20 Characteristic Speed


While
an understeer vehicle

has

no critical

speed, the characteristic speed is defined

in

a similar manner

to indicate the level of understeer present in the vehicle. The

51

Chapter 4

Two Degree-of-Freedom Vehicle Model

characteristic speed

is the

speed at which the steer angle required to produce

any

steady-

state rum radius

is twice the Ackerman

steer angle.

The characteristic

speed can

be found

by setting the steady-state steer angle of Eq. (4.54) equal to twice the Ackerman steer angle
of

Eq.

(4.57)
of

and

solving for the velocity V. Since the


which

solution

to this

equation

is

function

stability derivatives
must

depend on velocity, the stability derivative


resulting
equation solved

definitions
speed

be

substituted and the

for V. The

characteristic

is then
L CfC f
r

2
,

ychar=

'vmiaCf-bCA The

(4.75)

characteristic speed

is

seen

to have the

same

form

as

the

critical

speed, but with

the sign of the denominator reversed. A characteristic speed only exists if aC/is greater than

bCr. Since, from Section 4.5.5, this is the


understeer vehicles

condition

for

an understeer

vehicle, only
an

have

a characteristic speed.

Neutral

steer vehicles

have

infinite

characteristic speed and oversteer vehicles

have

no characteristic speed.

The

more

understeer a vehicle

has,

the lower its characteristic speed.

The
speed

sample vehicle

has

a characteristic speed of

525 km/hr. Since the

characteristic

is very high, the

vehicle

has

a small amount of understeer.

4.5.21 Characteristic Equation


The
measures of system response

derived in Section 4.5.7 through Section 4.5.20 The lateral transient


response of the two

are all measures of steady-state system response.

degree-of-freedom

road vehicle

is

now examined.

From the Laplace-domain


equation

equations of motion of

Eq. (4.27), the

characteristic

for the

system

is

(N
s1-*-

Y)s+ NrYf+Nf(mV-Yr)
H

V7*

mVJ

L,rnV

/\7

(4.76)

52

Chapter 4

Two Degree-of-Freedom Vehicle Model

From this

characteristic

equation, the

undamped natural

frequency, damping ratio,

and poles of the system can

be found.

4.5.22 Undamped Natural


From the

Frequency
Eq.

characteristic equation of

(4.76) the natural frequency

of

the system is

NY+NJmV-Y)
*
=
"

A V

r v ImV

(4-77)

Substituting the

stability derivative definitions yields

>,.=,

I^f +
ImV2

\L2CfCr
zz

aCf-bC '.""'

(4-78)

With the
natural

expression

in this form the

effects of various parameters on undamped

frequency,

and

consequently, system response time

can

be

seen.

From the first term


and the square

the
of

natural

frequency decreases with the yaw moment of inertia, the mass,


of

the forward velocity

the

vehicle.

In addition, it increases
wheelbase.

with

the product of tire the second


negative

cornering

stiffnesses and with

the square of the

The

numerator of

term is the directional stability iVp. This term is

positive

if the

vehicle

is understeer,
all else

if it is oversteer,

and zero

if it is
a

neutral steer.

Thus from this model,

being equal,

an understeer vehicle

has

higher

natural

frequency and lower response time than an

oversteer vehicle.

The
speed of

sample vehicle

has

an undamped natural

frequency of 1.01

Hz

at a

forward

100 km/hr. The

natural

frequency decreases as vehicle velocity increases as

shown

in Figure 4.5.

4.5.23

Damping
The

Ratio

damping ratio of the system can also be obtained from the characteristic
of

equation.

In terms

the stability derivatives the

damping ratio is

53

Chapter 4

Two Degree-of-Freedom Vehicle Model

50

100
V

150

200

(km/hr)
vs.

Figure 4.5: Natural Frequency

Vehicle

Velocity
(4.79)

IJi+mVNr

2^mV(iVryp+/Vp(mV-yr))
If the stability derivatives
are substituted

the

damping ratio becomes


(4.80)

Izz(Cf+Cr) + m(a2Cf+b2Cr)
2^Iam{l}CfCr +
mV2

(aCf

bCr ))

Depending on the values of the parameters, the vehicle may be underdamped,


critically damped,
or overdamped.

Neutral

steer vehicles

tend to be close to critically

damped, tending

with understeer vehicles

tending

toward underdamped and oversteer vehicles

toward

overdamped.13

As

with other

dynamic systems, the

damping ratio affects


disturbance

the response time and overshoot

of

the road vehicle to lateral control and

inputs. At
a

forward

speed of

100 km/hr the

sample vehicle

has

damping ratio of 0.990,

indicating that the vehicle is very slightly underdamped. Damping ratio decreases as vehicle
velocity increases. This
vehicle

is

overdamped

below

a speed of

approximately 63.7 km/hr.

54

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.5.24 System Poles


The Eq.
poles of the system can

be found

by

solving the

characteristic equation of

(4.76)

for

s.

Doing this yields the poles


mVNr +/ay

Pi2=

J(mVNr

Ijrf -Alumv\NTY^
H^mV

+N?{mV-

Yr))

(4.81)

The location

of

the poles in the

s-plane provides an

indication

of system response.

During the design stage the vehicle parameters may be manipulated to place the poles
that the desired system response is obtained. The pole locations also provide
of system stability.
an

such

indication

If there

are

any

poles

in the right half of the

5-plane

(i.e., they have

positive real

components), the

system

is

unstable.

At Since the

forward

speed of

100 km/hr the

sample vehicle

has

poles of

-6.301

0.918/.
since

poles are complex

conjugates, the

vehicle

is

underdamped.

Furthermore,

the real part


move

of

the poles is negative, the system is stable. As speed decreases the poles

together until

they

meet on

the

real axis at

-9.897

at a vehicle speed of

63.7 km/hr. At
poles

this speed the


move

vehicle

is critically damped. As
the

speed

increases from 100 km/hr the

farther

apart and approach

imaginary axis. However, even at a forward speed of

300 km/hr the


stable

poles remain

in the left half of the 5-plane,

indicating that the vehicle remains

for

reasonable speeds.

4.5.25 System Zeros


Further insight into the
system zeros.
nature of system response can

be

obtained

by examining the

The

system zeros can

be found for

each

input

and state variable combination

by setting the numerator of the corresponding transfer function equal to zero and solving
for
s.

The transfer functions


The
zero

are given

in Eq. (4.28) through Eq. (4.33).


to steer angle input is

for sideslip

angle response

55

Chapter 4

Two Degree-of-Freedom Vehicle Model

zp/s
The

NrY&+N&{mV-Yr)

(4.82)
force input is

zero

for sideslip

angle response

to

aerodynamic side

hipya
The
zero

'

Nr-(c-a)(mV-Yr)

(4.83)
input is

for sideslip

angle response to road side slope angle

Nr
p/e
The for
j

(4.84)
input is

zero

yaw

velocity

response to steer angle

Zr/6
The

(4.85)
mVNx

zero

for yaw velocity

response to aerodynamic side

force is

+
_;vp

yp(c-a)
-

zr IF
mV(c

(4.86)
does
not

a)

There is
appear

no zero

for

yaw

velocity

response to road side slope since s

in

numerator of

this transfer function. This is due to the fact that the gravitational

side

force

acts at the mass center and therefore results

in

no external yaw moment on

the

vehicle.

The
given

system zeros

for the

sample vehicle with a

forward

speed of

100 km/hr

are

in Table 4.4. Table 4.4: System Zeros V=100km/hr

Response
Vehicle sideslip Yaw velocity
angle

Steer Angle 21.53


-5.59

Aerodynamic
Side Force
-9.86

Road Side Slope


-7.06

-5.04

56

Chapter 4

Two Degree-of-Freedom Vehicle Model

The
poles.33

effect that zeros

have
a

on response

depends

upon

their location

relative

to the

The

closer a zero

is to

pole, the larger the

effect.

zero

that is in the right half of

the 5-plane

(i.e., is

positive

in sign) is

called a nonminimum-phase zero.

The

nonminimum-

phase zero can cause the response to

initially start in the opposite direction.


sideslip
angle response to steer angle

This

response

can

be

seen

in the lateral velocity

and

input which is

simulated

in Section 4.5.27. Plots

of the poles and zeros

for

each

input

and output

combination are provided

in Figure 4.6 for the

sample vehicle with a

forward

speed of

100

km/hr.
The
angle zero values of all of

the

zeros

increase

with vehicle speed.

The sideslip

angle

steer

is

negative at

low

speeds and positive at

high

speed.

This

zero changes sign

when

the vehicle speed reaches its tangent speed. All other

zeros remain negative

for

reasonable values of vehicle speed.

0.5

m CO

/\

a*

E
-0.5

?
-10

10

15

20

25

Re(s)
?Pole

Sideslip Angle / Steer Angle Zero


xSideslip Angle / Road Side Slope Zero
Yaw

ASideslip Angle /Aero Side Force Zero XYaw Velocity /Steer Angle Zero Figure 4.6: Poles

Velocity /Aero Side

Force Zero

and

Zeros

57

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.5.26

Frequency Response
It is
also

interesting to examine the frequency response of the vehicle. A driving

event where

frequency response may be of particular interest is a slalom test where the


spaced cones

vehicle

is driven through regularly

by means of a sinusoidal steering input.

The
and

frequency of the input required to negotiate the slalom depends upon the vehicle speed
the cone spacing. The performance
of

the

vehicle

in the

slalom

may be influenced

by

the magnitude of the input

frequency relative to the natural frequency of the vehicle.


also

Sinusoidal steering inputs may


change.

be

used

in emergency

maneuvers such as a

double lane

Examining the frequency response of the vehicle may provide an indication of its
in
such a maneuver.

performance

Since it is generally desirable to


slope,

minimize

the response of

a vehicle

to

disturbances
be
used

such as side winds and road side

frequency response

techniques

can

to examine the response of the vehicle to periodic disturbance

inputs. Phase lags in


obtain

response to

steering input require the driver to

adjust

his input to

the desired response, making the vehicle more difficult to drive. Smaller phase lags

tend to
a

improve

controllability.21

vehicle

The

frequency response of the sample vehicle with


using the bode plotting capability
of

forward velocity

of

100 km/hr is

examined

the

MATLAB Controls Toolbox. The


yaw

gain and phase responses of vehicle

sideslip

angle and

velocity to

steer

angle, aerodynamic side

force,

and road side slope are plotted

in

Figure 4.7 through Figure 4. 12. The MATLAB Appendix

script

DOF2LFreq.m,

which

is listed in

C.5, is

used

to facilitate plotting of the

frequency response.

The

script generates

gain and phase versus

input

frequency for the two degree-of-freedom model.


DOF2Control.m,
input
which sets program execution

DOF2LFreq.m

calls the scripts

parameters;

DOF2Param.m,

which sets vehicle and

magnitude

parameters;

and

58

Chapter 4

Two Degree-of-Freedom Vehicle Model

DOF2DependParam.m, which calculates vehicle parameters which depend on other


parameters.

These

scripts are

listed in Appendix C.l through Appendix C.3.

The
most of

frequency response of a road vehicle changes as forward velocity changes. For


magnitude of

the responses the

the

gain changes while

the

general shape of

the

curves remain

approximately

constant.

There is little

change

in the

phase plot

for

most of

the responses. The sideslip angle / steer angle


experiences significant change

response

is the only

response which

in the

shape of the gain and phase plots as

forward velocity

changes.

The sideslip

angle

steer angle

frequency response is influenced strongly by the


This is
a result

magnitude of the

forward

speed relative to the tangent speed of the vehicle.

of

the sideslip angle / steer angle zero changing sign at the tangent speed. The sideslip angle

steer angle

frequency response is plotted in Figure 4. 13


and

and

Figure 4. 14 for forward

speeds of

30 km/hr

49.84 km/hr
gain

respectively.

At 30 km/hr the
goes

is flat up to approximately 1 Hz
at

at

0.33

deg/deg and the phase


of a second order

from

at

0. 1 Hz to

100 Hz. The

phase response

is typical

system with a negative zero.

At 49.84 km/hr, the tangent speed, the

gain approaches zero

as

the

frequency approaches zero as expected from the definition of tangent speed.


there is a significant
peak

However,

in the

gain at

approximately 2 Hz,
90

which

is

the

undamped natural

frequency at 49.84 km/hr. The phase goes from


the undamped
natural

at

0. 1 Hz to

at

100 Hz, crossing


km/hr is There is
shown

zero at

frequency. The
from
180

frequency response at
at

100

in Figure 4.7. The lead


at

phase goes

at

0.1 Hz to

100 Hz.

180

phase

low

frequency because above the tangent speed a positive


sideslip
angle as shown

steady-state steer angle produces a negative

in Section 4.5.18.
The
phase

Also
goes

of

interest is the
0.1 Hz to
others

yaw

velocity / road
100 Hz. The
of a zero.

side slope phase response.

from

at

at

phase response of

this transfer function

differs from the

due to the lack

59

Chapter 4

Two Degree-of-Freedom Vehicle Model

At low frequencies the


values of the steady-state

gains

for each input

and output combination approach the

step input response

gains shown

in Table 4.3 for

forward

velocity V

100 km/hr.

60

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.6
O)

1.2
o>

-i^^^

-i--iT-i-i-i-r_

~i~

"

\-

~i

i~

i-

CD

0.8
CO

0.4
0.0 0.1
1 10

100

Frequency (Hz)
180
D)

CD

2.
CD
CO

90

----!---

_i

_-|_

|-

+^^J

--

--

co

'

-90

'

'

'

'

IT

0.1

10

100
V=

Frequency (Hz)
Figure 4.7:

Sideslip Angle / Steer Angle Frequency Response,

100 km/hr

3.0E-04

"& 2.0E-04
CD

"I
CD

1.0E-04

0.0E+00 0.1 10 100

Frequency (Hz)

CD

2_
CD
CO
-45

CO

-90

0.1

10

100

Frequency (Hz)
Figure 4.8:

Sideslip Angle /Aero Side Force Frequency Response,

100 km/hr

61

Chapter 4

Two Degree-of-Freedom Vehicle Model

CD

CD

2.
C co

0.1

10

100

Frequency (Hz)

CD
CD
CO
-

-45

CO

-90

0.1

10

100

Frequency (Hz)
Figure 4.9:

Sideslip Angle /Road Side Slope Frequency Response,

100 km/hr

0.20
CD

0.15
-52

co

0.10 0.05 0.00


0.1

CO

10

100

Frequency (Hz)

CD

2,
CD
CO
__ _ _ _ _,_ _ _ --,-

-45

-, -r

-i

-i "I -i

-^c

-i

i-i-|------|---|---i-T-i-r-TT

CO

-90

0.1

10

100

Frequency (Hz)
Figure 4.10: Yaw

Velocity / Steer Angle Frequency Response, V

100 km/hr

62

Chapter 4

Two Degree-of-Freedom Vehicle Model

8.0E-06

6.0E-06

---,---,-

rT^~

----n---i~~T-T-i-rTTr~___in

2. 4.0E-06
O 2.0E-06

0.0E+00
0.1 10 100

Frequency (Hz)
180
O)

CD

135 90

2.
CD
CO CO

45 0

\-

0.1

10

100

Frequency (Hz)
Figure 4.11: Yaw

Velocity /Aero Side Force Frequency Response,

V-100 km/hr

2.5E-04
.g

2.0E-04 1.5E-04 1.0E-04


r

CO

O 5.0E-05

0.0E+00
0.1

10

100

Frequency (Hz)
0

CD TO CD
CO

-45

-90

CO
-135

-180

0.1

10

100

Frequency (Hz)
Figure 4.12: Yaw

Velocity /Road Side Slope Frequency Response, V

100 km/hr

63

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.4
^m^

O)

CD T3
O)

0.3
0.2

-\

-\

\~

tt

-i_

~^^t

~i-n-i-rTTr

-i

-i

i~

CD

2,
C co

._____,__

_r

__!

-r

~r

-----|---r

-|

i^W"

ttt

-,--

-i

-j

|-

0.1
0.0 0.1 10
100

Frequency (Hz)

CD TO CD
CO
-45
i

t--i--|-|-t-]-it-----|----|--i ^ci

-i

-i -}

CO

-90

0.1

10

100

Frequency (Hz)
Figure 4.13:

Sideslip Angle /Steer Angle Frequency Response,

30 km/hr

0.3
,,

O)

CD

0.2
CD D
-H^

--

----I---

__l__l_l_

_l_l

,_

-f^l^I-

-I

-I

-I

-I

--t----f__l__l_H

"co

0.1

0.0 0.1
1

10

100

Frequency (Hz)
90 45

CD

2,
CD
CO

CO
-45

-90

0.1

10

100

Frequency (Hz)
Figure 4.14:

Sideslip Angle / Steer Angle Frequency Response,

49.84 km/hr

64

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.5.27 Simulation
Another tool that is very
simulation.
useful

in analyzing

vehicle

lateral dynamics is

numerical

Simulation is done by
be
used

integrating the differential equations of motion with


response of

respect to time and can

to predict the

the

vehicle

to arbitrary

control and

disturbance inputs. Non-linearities

are

generally

much easier

to handle
point

with numerical

simulation than with the analytical techniques used

up to this

in this

chapter.

To

maintain

consistency

with

the

non-linear model simulation which

follows in
simulation

Section 4.6, the lateral velocity instead


of

is

used as a state variable

for the linear model


were

the sideslip angle (3. Since the equations of motion


and

originally derived in

terms of yaw velocity and lateral velocity


and

then simplified to be in terms of yaw velocity to its original,


non-linear more

sideslip angle, the


The

model

has been

returned

general, form.

equations of motion

in their general,

form

are given

by Eq.

(4.9).

With the

small steer angle assumption used

for the linear

model

the equations become

Fyf+Fyr+Fya+Fyg=m(v + ur)
~

aFyf

bFyr

(c

a^Fya

IJ
and gravitational side

Expressions for the tire slip angles, tire lateral forces,


are

force

derived in Section 4.5.3 here for

and

Section 4.5.4 using the

small angle assumption and are

repeated

convenience.

v + ar

af

u
,

v-br
ar=

(4.19)

Fyf=Cfaf

Fy=mgQ

(4.23)

65

Chapter 4

Two Degree-of-Freedom Vehicle Model

Outputs from the


vehicle

simulation are

time histories of lateral velocity, yaw velocity,

sideslip angle, tire slip angles,

and

lateral

acceleration.

Inputs

can

be

step steer, step


road

ramp step steer, ramp


side slope.

square steer, sine steer,

step

aerodynamic side

force,

or

The

simulation

is implemented in the MATLAB


of

script

DOF2LSim.m,

which

is

listed in Appendix C.6. Integration


built-in MATLAB function ode23,
formulas.34

the differential

equations of motion

is done using the

which uses second and

third order Runge-Kutta


r over

The function

ode23 returns

the state variables v and

the time interval

specified

for the
each

simulation.

At

time step the

ode23

function

calls

the function DOF2LDE.m

which

calculates the state

derivatives

v and r

based upon the instantaneous

values of the state

variables v and r.

First,

the instantaneous steer angle is calculated


upon

by the function
time,
the type of

SteerAngle.m,

which

is listed in Appendix C.4, based

the current

input selected, the

magnitude of the

input,

and

the values of the input duration parameters.

Any arbitrary steer input, including steer inputs measured experimentally during vehicle
testing,
could

easily be implemented in this function. Use


the model.

of measured steer

input data

facilitates

experimental validation of

After the
values of

steer angle

is calculated, the tire slip

angles are calculated

from the

current

the state variables

v and r which are passed as parameters

into DOF2LDE.m. The

tire lateral forces are then calculated from the slip angles using Eq. (4.21).

Finally, the

state

derivatives

are calculated as

Fyf+Fyr+Fya+Fyg
m

ur

aFyf-bFyr-(c-a)Fya
=

(4-88>

66

Chapter 4

Two Degree-of-Freedom Vehicle Model

Eq.

(4.88)

are obtained

by

solving Eq.

(4.87) for

v and r.

listing of DOF2LDE.m is

provided

in Appendix C.7.
effect of

To illustrate the
at

forward velocity

on

response,

simulations are performed

low

speed

(30 km/hr), high

at the tangent speed

(49.84

km/hr),

at normal

highway speed (100


is

km/hr),

and at

speed

(150 km/hr). The for the

simulations are run until steady-state

reached.

Initial

conditions

simulations are zero.

Lateral velocity,

yaw

velocity,

sideslip angle, front tire slip angle,


for
each

rear

tire slip angle, and lateral

acceleration are plotted

input

studied.

The inputs

used

for the
1

simulation are a

step steer,

ramp step

steer with a

ramp time
of

of

0.2 sec,

ramp

square steer with a

ramp time

of

0.2

sec and a

dwell time

1.0 sec,
1

sine steer with a period of

1 sec,

10000 N step
square

aerodynamic side

force,
inputs

and a

step

road side slope.

The ramp step steer, ramp


steer

steer,

and sine steer

are shown

in Figure 4.15. The

input is

a positive steer

angle,

indicating a right rum.

Ramp Step
CD ;o

Steer Input
CD

Ramp Square Steer Input


S 1.0

1.0

<
CD CD
I

0.5

<
CD CD

0.5

CO

0.0
1

55 0.0
Time

El
1 2

(s)

Time

(s)

Sine Steer Input


O)

CD

1.0

2,
<D
D) c

0.0

<
CD CD

OT

-1-0

Time

(s)

Figure 4.15: Simulation Steer Angle Inputs

67

Chapter 4

Two Degree-of-Freedom Vehicle Model

The aerodynamic
slope

side

force input is

applied

in the

positive

y-direction, and the road side

is positive,

which means that the road slopes

down to the right.

Simulation
Figure 4.21. The
velocity
and

results

for the step

steer

input

are plotted

in Figure 4. 16 through
The
steady-state

response times

increase

with

forward

speed.

lateral

sideslip

angle are positive

below the tangent speed,


agrees with the

zero at

the tangent speed, the tangent speed

and negative above the tangent speed.

This

definition

of

presented

in Section 4.5.18. Above the tangent

speed the

lateral velocity

and

sideslip

angle

also

initially begin to increase from zero becoming positive and then decrease to negative
This is
a result of

values.

the system zero


explained

being positive when the forward speed is greater


a system with a positive zero

than the tangent speed. As

in Section 4.5.25

is

nonminimum-phase system and

typically exhibits the type

of

step

response shown

here,

initially in the opposite direction to the steady-state value. The front tire slip angles show
response similar

to the sideslip angle, but with initial values of

due to the

step lateral

steer.

The lateral

acceleration

has

a non-zero

initial

value

due to the rate

of change of

velocity

when

the step steer occurs. The steady-state values of yaw velocity, sideslip angle,
rear

front tire slip angle,

tire slip angle, and lateral acceleration agree

with

the steady-state
angles and yaw

response gains presented

in Table 4.3. In addition, the

steady-state

sideslip

velocities agree with

the values approached at low

frequency in the frequency response


Figure 4.14.
in Figure 4.22 through Figure 4.27.

plots of

Figure 4.7, Figure 4.10, Figure 4.13, step

and

Ramp
The ramp step
expected.

steer simulation results are plotted

steer response

is

similar

to the step steer response and


of

lags it slightly

as

The

steady-state values are

identical to those

the step response. The lateral

velocity

and

sideslip

angle still exhibit

the non-minimum phase system response above the

tangent speed, but the

magnitude of

the initial response is

less than it is for the step


acceleration are

steer.

Unlike

with

the step

input,

the front tire slip angle and

lateral

initially zero

68

Chapter 4

Two Degree-of-Freedom Vehicle Model

for the ramp step input. At 30 km/hr


angle response at

and

49.84 km/hr there

are peaks

in the front tire slip

0.2

sec which

is

when

the ramping of the steer input is completed.

The

responses to the

ramp

square steer

input

are plotted

in Figure 4.28 through


response

Figure 4.33. The ramp

square steer response

is identical to the ramp step


zero.

up

until

the time that the steer input is ramped back down to


reach steady-state

At the lower

speeds the responses

before the ramp down. However,


has been
reached.

at

150 km/hr the ramp down

occurs

before

steady-state

As

with

the ramp step response the front tire slip


as the

angles experience overshoot at

30 km/hr

and

49.84 km/hr

ramp up is

completed.

The input had


state yaw a

sine steer results are shown

in Figure 4.34 through Figure 4.39. The

sine steer

frequency of 1
and

Hz. From

visual

inspection

of

the plots it is seen that the

steady-

velocity

sideslip

angle gains and phases agree with

those obtained for 1 Hz


and

from the
The

frequency response

in Figure 4.7, Figure 4.10, Figure 4.13,


with

Figure 4.14.

response amplitude

increases As

forward velocity in

all cases except the

lateral

velocity

and

sideslip

angle.

expected

from the definition

of

tangent speed, at 49.84


at

km/hr the lateral velocity Results from the

and

sideslip

angle amplitudes are

less than those

30 km/hr. in

aerodynamic side

force step input

simulation are provided

Figure 4.40 through Figure 4.45. The


velocity.

magnitude of

the responses increases with forward


neutral steer

Since the

center of aerodynamic pressure

is located behind the

point,

a positive aerodynamic side

force

produces a negative yaw velocity.

The

steady-state yaw

velocity, sideslip angle, front tire slip angle, rear tire slip angle, and lateral acceleration
responses at

100 km/hr

agree with the steady-state gains

in Table 4.3. In addition, the


agree with

steady-state yaw

velocity
and

and

sideslip
as

angle at

100 km/hr

the

frequency response
The sideslip

gains of Figure

4.8

Figure 4. 1 1
rear

the input

frequency

approaches zero.

angles, front tire slip angles,


slopes at

tire slip angles, and lateral acceleration curves

have higher

lower

speeds

indicating that the response is faster at lower speeds.

69

Chapter 4

Two Degree-of-Freedom Vehicle Model

Road

side slope

step input results

are plotted

in Figure 4.46 through Figure 4.51.


increase
angle with

The lateral velocity,


velocity,
while the

yaw

velocity, and lateral

acceleration responses

forward

sideslip angle, front tire slip angle,


yaw

and rear

tire slip

decrease.

Again,
and

the steady-state

velocity, sideslip angle, front tire slip angle, rear tire slip angle,
obtained with

lateral

acceleration at

100 km/hr

the

simulation agree with

the steady-state

gains of

Table 4.3,

and the steady-state yaw

velocity

and

sideslip

angle agree with

frequency response gains of Figure 4.9 and Figure 4.12 as the input frequency approaches
zero.

The linear two degree-of-freedom model is

useful

for characterizing

and

predicting

the response of the automobile to control and disturbance inputs. Although this model
simplifies the vehicle
much can

greatly

system,

be learned

about road vehicle

lateral

dynamics through its


response can

study.

The

effects of

changing

vehicle and

tire parameters on system

quickly be
can

studied.

Powerful linear
applied

analysis techniques

based

on system

transfer

functions

be readily
results

to the vehicle model to gain significant

insight into

system

behavior. The

from the linear model

are

generally

valid

for lateral

accelerations

up to 0.35 g,

which constitutes most of normal passenger car

driving.

Beyond this level

a non-linear tire model

is

required

to accurately simulate tire behavior at

high slip

angles.

70

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5 V 0.0 V
= =

30 km/hr 49.84 km/hr

-0.5
_

I
-1.0

V=

100 km/hr

o CD
co

CD
CO J
-2.0

-2.5

V= 150 km/hr

-3.0

0.5

1 Time

1.5

(s)

Figure 4.16: Linear Step Steer Lateral


0.30
'

Velocity Response

>
0.25

150 km/hr

0.20
f0

/i

8
CD

-15

/^

100 km/hr

>

o.io

'/ /^
ifs^~

49.84 km/hr

0.05

30 km/hr

0.00

J
0.5
1

1.5

Time

(s)

Figure 4.17: Linear Step Steer Yaw

Velocity Response

71

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5
,
,

30 km/hr
49.84 km/hr

0.0
V
-0.5
=

C7> CD T7

-1.0

CD TO C

-1.5

V
-2.0

100 km/hr

<
Q.
CO

CD
Tl

CO

-2.5

-3.0

-3.5

V
1

150 km/hr

-4.0

0.5

1.5

Time

(s)

Figure 4.18: Linear Step Steer Sideslip Angle Response


0.0 V
-0.5
=

30 km/hr 49.84 km/hr

V
-1.0
^-s.

O)

CD

2.
CD TO

-1.S

<
Q.

-2.0

V=
-?.5

100 km/hr

CO
a)

H
H-*

r o

-3.0

LL
-3.5

-4.0

V= 150 km/hr
-4.5

0.5

1.5

Time

(s)

Figure 4.19: Linear Step Steer Front Tire

Slip Angle Response

72

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.0

V
-0.5

30 km/hr
49.84 km/hr

V
-1.0
^^

TO 0)

2,
CD TO C

-l.b

<
Q.

-2.0

V
-?.S

100 km/hr

CO
CD
1-

m CD

-3.0

DC
-3.5

-4.0

V
-4.5

150 km/hr

0.5

1.5

Time

(s)

Figure 4.20: Linear Step Steer Rear Tire


1.4

Slip Angle Response

V
1.2

150 km/hr

1.0
c

o ffl
OJ CD O O

0.

<

0.

V= 100 km/hr

s
CD
CO
-i

0.4

0.2

V V

49.84

km/hr"

"

30 km/hr

0.0
0.5

1.5

Time

(s)

Figure 4.21: Linear Step Steer Lateral Acceleration Response

73

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5 V 0.0
V
= =

30 km/hr 49.84 km/hr

"-5
_

""V

E,
&
o o
-1.0

V= 100 km/hr

CD

i
CD
I

"1-5

co J
-2.0

-2.5

V
..

150 km/hr

-3.0

0.5

1.5

Time

(s)

Figure 4.22: Linear Ramp


0.30
,

Step Steer Lateral Velocity Response

s^

V= 150

km/hr

0.25

0.20
T3

/f
l/^
V= 100 km/hr

8
CD

0.15

>

0.10

If
r

49.84 km/hr

0.05

30 km/hr

0.00

!
0.5

1.5 Time

(s)

Figure 4.23: Linear Ramp

Step Steer Yaw Velocity Response

74

Chapter 4

Two Degree-of-Freedom Vehicle Model

TO CD

TO

<
Q.
W

CD

g
CO

-3.0

0.5

1.5

Time

(s)

Figure 4.24: Linear Ramp


0.0

Step Steer Sideslip Angle Response

V
-0.5

30 km/hr 49.84 km/hr

V
-1.0
^-H^

TO CD

2,
CD TO C

-1.S

X
-2.0

^^^^

<
a.

V= 100 km/hr
-?.S

CO
CD
i

1-

C O

-3.0

-3.5

-4.0

V
1
-4.5

150 km/hr
'

0.5

1.5

Time

(s)

Figure 4.25: Linear Ramp

Step Steer Front Tire Slip Angle Response

75

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.0
V
-0.5
=

30 km/hr 49.84 km/hr

V
-1.0
^-^

TO

a)

2,
CD
TO

-l.b

<
a.

-2.0

V
-2.5

100 km/hr

CO
a>

H
-3.0

CD

oc
-3.5

-4.0

.;

>s
V= 150 km/hr

-4.5

0.5

1.5

Time

(s)

Figure 4.26: Linear Ramp


1.4

Step Steer Rear Tire Slip Angle Response

V= 150 km/hr

1.2

1.0
c

]3
CD O U

0. 8

<

0.

V= 100

km/hr

B
"5 0.4

0.2

V V
^^^

"

49.84 km/hr 30 km/hr

0.0

0.5

1 Time

1.5

(s)

Figure 4.27: Linear Ramp

Step Steer Lateral Acceleration Response

76

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5
>

V V

30 km/hr
49.84 km/hr

0.0

"^V

i
-1.0

\ ^Ss*^^

]/

\
\

V= 100 km/hr

o CD

I"'-5
CD
CO

J
-2.0

\V

150 km/hr

-2.5

-3.0

0.5

1.5

2.5

Time

(s)

Figure 4.28: Linear Ramp Square Steer Lateral


0.30
.

Velocity Response

>^V

150 km/hr

0.25

0.20

7
l/?
V
=

\
100 km/hr

\
A \
\

0.15

o o

2
co
>-

0.10

\/
^*

49.84 km/hr

V \
\

0.05

Wf

"^Nk

/]:*"'

0.00

.1

>^

.>

-0.05

0.5

1.5 Time (s)

2.5

Figure 4.29: Linear Ramp Square Steer Yaw

Velocity Response

77

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5
'

'

0.0

^Z-~~^__ "^l

30 km/hr
49.84 km/hr

N^

'

-0.5

W.
TO CD
-1.0

!
100 km/hr

\ ^v^V
-1.5

CD TO
c

<
Q.
CO
-2.0

V
'

\
\

rn

-2.5

-3.0

\. V
-

150 km/hr

-3.5

-4.0

1
0 0.5
1

1.5

2.5

Time

(s)

Figure 4.30: Linear Ramp Square Steer Sideslip Angle Response


0.5 0.0
-0.5

oU

Km/nr

\^

x^~~
r*"~

'

'

/
;
V
=

V
\\.

49.84 km/hr

TO
-1.0

"to TO
<
9.

-1-5

\ ^SV
V

100 km/hr

-2.0

CO

-2.5

\
P
-3.0 -3.5

;
:

150 km/hr

\,

I
1.5 Time (s) 2

-4.0

-4.5

0.5

2.5

Figure 4.31: Linear Ramp Square Steer Front Tire

Slip Angle Response

78

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5
0.0

Ys^:
-0.5

' ^S

/^

49.84 km/hr

TO
-1.0

/
/

T3

TO C

-1.5

<
a.
-2.0

\
-2.5

:^vV=

100 km/hr

CO

2
H
i

CO
-3.0

rr
-3.5

v-

ibUKm/nr

~/~

-4.0

l -4.5

....

i 1

0.5

1.5

2.5

Time

(s)

Figure 4.32: Linear Ramp Square Steer Rear Tire


1.2

Slip Angle Response

js^~\.
1.0

/
~

150 km/hr

0.8

/
0.6

\ \
=

j
o

Sv
0.4

100 km/hr

<

2
co

0.2

J/

49.84 km/nr

; \^;

X^

0.0

-0.2

0.5

1.5 Time (s)

2.5

Figure 4.33: Linear Ramp Square Steer Lateral Acceleration Response

79

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.0

y\
0.5 V
=

30 km/hn

r~\

\
>^-

0.0

-jf-<5

\\
-0.5

)V

49.84 km/hr

V \

>

2
CO

;V=

100 km/hr

-1.0

>^V
-1.5

150 km/hr

0.5

1.5

2.5

Time

(s)

Figure 4.34: Linear 1 Hz Sine Steer Lateral

Velocity Response

0.25 0.20
0.15
I"

v^\/

150 km/hr 100 km/hr:

\
/nA

1/^sN
//SV

0.10

'

\
==
-

49.84 km/hr

w
30 km/hr
-

\\ r\\\
Is-^
^dr^^sAW

T
u

T^x \\

fss\

r\\\
i^r\\\\

0.05
o

0.00

-0.05

-0.10

-0.15

i -0.20

0.5

1.5 Time (s)

2.5

Figure 4.35: Linear 1 Hz Sine Steer Yaw

Velocity Response

80

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.5

f\.

1.0

1
0.5
TO

'

30 km/hr

fl

VV

0.0
TO
C

<
a-0.5

\\
-1.0

49.84 km/hr

CO

g
CO

W/
\
i

V= 100 km/hr

W
\ /
i i

-1.5

V= 150
-2.0

km/hr
i

0.5

1.5

2.5

Time

(s)

Figure 4.36: Linear I Hz Sine Steer Sideslip Angle Response

1.0

0.5

V v

49.84

km/hr^y/

\V

sT*l

30

TO

km/h|>7/v\

\\
" "

yy/h\

0.0
TO C

<
-0.5

-Vs

'/

CO

~^\\

'/ /

<
LL

-1.0

Yv

100 km/hr

-1.5

V
j
-2.0

150 km/hr
1

0.5

1.5

2.5

Time

(s)

Figure 4.37: Linear 1 Hz Sine Steer Front Tire

Slip Angle Response

81

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.5
1
__

1.0

0.5
TO

V 0.0
TO C

30 km/hr^^-f

\ \\
\\

<
-0.5

==

49.84 km/nr

CO

2 F
CO

-1.0

//
\
\

\
\

DC

100 km/hr

-1.5

-2.0

V= 150

km/hr

-2.5

0.5

1.5 Time (s)

2.5

Figure 4.38: Linear 1 Hz Sine Steer Rear Tire

Slip Angle Response

0.5

f\
0.4

\v

150 km/hr

0.3

0.2

'

1
o

o.i

"^^\v

f^
= =

//

\V= 100 km/hr

L
'/its.'

49,84 km/hr

"

~/Tr<~

\-\

o.o

30

km/hr^^VVy;
\
1

//
/

co

"0.1

N\l^-

-0.2

VI
-0.3

.1

-0.4

0.5

1.5 Time (s)

2.5

Figure 4.39: Linear 1 Hz Sine Steer Lateral Acceleration Response

82

Chapter 4

Two Degree-of-Freedom Vehicle Model

3.0 V 2.5
=

150 km/hr
,

',
-52

2.0

E
o
-2

CD

1 5 J
"

"

100 km/hr
i
i i

>

2
*

1.0

V
0.5

49.84 km/hr

:
V
*
=

30 km/hr

0.0 0

0.2

0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.40: Linear Step Aero Side Force Lateral


0.00

Velocity Response

-0.02

JO

-0.04

T3 co

'o
o

-0.06

>

>

-0.08

-0.10

-0.12

0.2

0.4

0.6

0.8

1.2

1.4

1.6

Figure 4.41: Linear Step Aero Side Force Yaw

Velocity Response

83

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.0

V
r

150 km/hr

3.5
i

3.0
TO

100 km/hr

2.5
TO C

Jr

V V

49.84, km/hr
30 km/hr

< Jo

2.0

1.5
CO

1.0

i i

0.5
i

0.0

1
0.6

0.2

0.4

0.8 Time (s)

1.2

1.4

1.6

Figure 4.42: Linear Step Aero Side Force


3.5

Sideslip Angle Response

V= 150 km/hr

3.0

V
CD
<-D

100 km/hr

49.84 km/hr
|

1?
<

2.0
V 1.5
=

30 km/hr

f
p

1.0

0.5

0.0

V
0

i
1.4

0.2

0.4

0.6

0.8 Time (s)

1.2

1.6

Figure 4.43: Linear Step Aero Side Force Front Tire

Slip Angle Response

84

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.0
i

,V

150 km/hr
i

3.5
i

!V
^

100 km/hr

3.0

TO

2.5
TO C

!V
iV

49.84 km/hr

<
J2-

30 km/hr

2.0

--

-/-/-

co

2 F
CO

1.5

DC

1.0

0.5
i
i

0.0
0
0.2

0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.44: Linear Step Aero Side Force Rear Tire

Slip Angle Response

0.6

0.4

0.2
o

\
o o

V
0.0

30 krrl/hr

<

V
-o-2

49.84 km/hr

CO

V= 100 km/hr

-0.4

V= 150 km/hr

1
-0.6

0.2

0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.45: Linear Step Aero Side Force Lateral Acceleration Response

85

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.05

;
V
150 km/hr

0.04

jo
~

0.03
V= 100 km/hr
i

+-*

'a
o

>

0.02

/j

V
0.01

49.84 km/hr

l~r
V
=

30 km/hr

0.00 0

0.2

0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.46: Linear Step Road Side Slope Lateral


3.5E-04

Velocity Response

3.0E-04
V= 150 km/hr

2.5E-04
T3
~

V= 100 km/hr

2.0E-04

+-*

"o
o

S
co
>-

1.5E-04

V
1
.0E-04

49.84\ km/hr

V
5.0E-05

30 km/hr
1

0.0E+00

i 0
0.2 0.4

i_.
1

i
1.4 1.6

0.6

0.8 Time (s)

1.2

Figure 4.47: Linear Step Road Side Slope Yaw

Velocity Response

86

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.07

V 0.06

30 km/hr
j/0^

;
t
.

;
ii
.

;
I

100

km/hr-

/ Ai
0.05
TO
-

150 km/hr

49.84

knrvhr^-'

0.04
TO c

<

0.03
CO

T3

CO

0.02

0.01

0.00

i
0.2

0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.48: Linear Step Road Side Slope


0.07

Sideslip Angle Response

V= 150 km/hr

0.00 0 0.2 0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.49: Linear Step Road Side Slope Front Tire

Slip Angle Response

87

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.07

0.06

V
TO

100 km/hr

150 km/hr

0.05

TO C

0.04

<
Q.

(f)
(1>

0.03

H
\

CC

CD

0.02

0.01

0.00
0
0.2 0.4

0.6

0.8 Time (s)

1.2

1.4

1.6

Figure 4.50: Linear Step Road Side Slope Rear Tire


0.018
0.016

Slip Angle Response

0.014

S 0.012
c

0.010

|
2

0.008

% CO

0.006

-"

0.004

\\v
h
~V

30 km/hr v : \ == 49.84 km/hr

"

"

vVcV
0.002
^W

100lcm/hr
-i

!
V
'
=

'"^*H^^

150 km/hr

V^

'

^^^.
0.4

0.000 0 0.2 0.6 0.8 Time (s) 1 1.2


1.4

1.6

Figure 4.51: Linear Step Road Side Slope Lateral Acceleration Response

88

Chapter 4

Two Degree-of-Freedom Vehicle Model

4.6 Non-Linear Model


As previously
noted

the linear vehicle model is valid for lateral


a result of tire

accelerations

up to
respect

approximately 0.35

g.

This is primarily

lateral force being linear with


accelerations.

to slip angle at small slip angles, and hence small lateral

Beyond 0.35 g

when

higher slip accurately

angles are

being attained a non-linear tire model is usually necessary to


lateral forces.
models of tire

predict tire

As discussed in Chapter 3 many


chosen

behavior exist. The tire

model

for this

work

is

called

tire data nondimensionalization and was

originated

by Hugo

Radt. This tire In this

model

is discussed in detail in Section 3.4.


the equations

section

describing the non-linear two degree-of-freedom


of

vehicle model are presented.

Simulation

the model is

performed

for

selected

steering

inputs

and

the results are compared with the simulation of the linear model.

4.6.1 Model Equations


The
equations of motion

for the

non-linear

two degree-of-freedom vehicle are

derived in Section 4.3

and are repeated

here.
+
-

F^ cos 8 + F
aF^ cos 5
Expressions for the tire slip
-

F
-

m(v +

ur)

bFyr

(c

a)Fya

I^r

(4.9)

angles are

derived in Section 4.4.

tf=^f)-i atan 8 a ,
-

(4'14>
a,
=atan

v-M
u

j
as

The tire lateral force is


Section 3.4
can

given

by the following expressions


From these

described in
lateral force F

and repeated

here for

convenience.

equations the tire

be

calculated

based

upon

the tire

vertical

load

Fz and the tire slip angle a.

89

Chapter 4

Two Degree-of-Freedom Vehicle Model

CC=B3
Hy=B5

C3FZ C5Fz

(3.3) (3.5)

_=Qtan(a)

(3

6)

y/

(l-E1)a+-

x_

FatanfB.a) ^-^i

^x

(3.10)

C, atan(fl^)

(3.9)

Fy=DlSin(tJ)
Fy=FuFz
4.6.2 Simulation
Simulation in the MATLAB
similar of

(3.8) (3.11)

the non-linear two degree-of-freedom vehicle model is implemented

script

DOF2NLSim.m. This

script

is listed in Appendix C.8


the linear model. As
and

and

is very

to DOF2LSim.m

which performs simulation of

with

the linear
are

simulation, the scripts


called at

DOF2Control.m, DOF2Param.m,
set

DOF2DependParam.m

the

beginning of DOF2NLSim.m to
ode23

simulation, vehicle, and tire parameters.


to integrate the

The built-in MATLAB function


motion which are contained

is

used again

differential

equations of

in the function DOF2NLDE.m. This function


upon

calculates the

state

derivatives

v and r

based

the instantaneous values of the state variables

v and r

and

the current steer angle. DOFTNLDE.m is listed in Appendix C.9. The state

derivatives

are calculated as

laF^ cos(5) 2bFyr


-

(c

(4'89)
a)Fya

90

Chapter 4

Two Degree-of-Freedom Vehicle Model

These

expressions are obtained

by solving Eq. (4.9) for

v and r

The

most significant

difference between the

non-linear and

linear model

simulations

is in the

calculation of tire

lateral forces. The tire lateral forces


which

are calculated

by the

MATLAB function NLTire.m


vertical

is listed in Appendix A.3. This

script

takes the tire


the

load

and

slip

angle as arguments and returns the tire

lateral force. Note that

with

non-linear

tire model the lateral forces

F^ and F,r are for only one tire, while for


are

simplification

in the linear model they


get the

for two tires. Thus here they

are multiplied

by the

factor of two in Eq. (4.89) to


time step

lateral forces for two tires. NLTire.m is

called at each

by DOF2NLDE.m, which also calls the function SteerAngle.m to calculate the


steer angle.

instantaneous For
performed

comparison with

the linear model,


and

simulation of the non-linear model

is

for the step

steer

input

the ramp square

steer

input. As

with

the linear

model, simulations are performed for forward

velocities of

30 km/hr, 48.94 km/hr, 100

km/hr,
used

and

150 km/hr. The step

steer and

ramp

square steer

inputs

are

identical to those
non-linear

for the linear model,

having a magnitude of 1.

Tire

parameters

for the

tire

model are given

in Table 3.1. These

parameters are a result of

the curve

fitting of empirical

tire data done in Section 3.4. The

values of

the linear tire cornering stiffnesses used in

throughout Section 4.5 are derived from these parameters, so the linear tire model and non

linear tire
the linear

model agree at small

slip

angles.

Vehicle

parameters are

identical to those

used

in

simulation.

Results from the


on

simulations are provided

in Figure 4.52 through

Figure 4.63. Included


comparison.

these plots as dashed lines are the linear simulation results for

Lateral velocity,
angle, and
presented

yaw

velocity, sideslip angle,

front tire slip angle, linear

rear

tire slip

lateral

acceleration results

for both the

non-linear and

simulations are

in Figure 4.52 through Figure 4.57 for the step

steer

input. The linear

and

non-

91

Chapter 4

Two Degree-of-Freedom Vehicle Model

linear lateral
simulation

acceleration results agree within

1%

over

the

complete

duration

of

the

for forward

velocities of

30 km/hr

and

49.84 km/hr. These


respectively.

speeds correspond

to steady-state lateral accelerations of 0.05 g and 0.14 g


produces a

At 100 km/hr,

which

0.55 g

steady-state

lateral acceleration, the linear model lateral

acceleration

results exceed those of the non-linear

by 4.3% during the transient and


slip
angles reach

1.0%

once

steady-

state

is

reached.

At this

speed the tire

slightly

more

than

2. At these slip line. Thus

angles

the tire lateral force versus slip angle curve is


and vehicle the

still

very nearly

a straight

for this tire

linear tire
g.

model

is reasonably

accurate and useful

for lateral

accelerations

in

excess of

0.5
of

However,

at

150 km/hr the linear

model

lateral
the
non

accelerations exceed

those

the non-linear model

by over 27%.
0.95 g

At this

speed

linear model
predicts

predicts a steady-state

lateral have

acceleration of

while

the linear model

1.20

g.

The tire slip

angles

exceeded

where

the lateral force versus slip

angle curve

is approaching its
angles of

peak.

The linear tire

approximation

is

not

sufficiently

accurate at

slip

this magnitude.

At high sideslip angles,


that the

speeds the

linear model

predicts

that the magnitudes of lateral velocities,

and

tire slip angles are below those that the non-linear model predicts and

yaw velocities and

lateral

accelerations are above

those of the non-linear model. The

linear
linear

model also predicts

faster response than the in


all of

non-linear model.

At 150 km/hr the

non

model predicts overshoot

the

quantities

examined,

while

the linear model

predicts no overshoot.

Non-linear

and

linear

simulation results

for the ramp

square steer

input

are plotted

in Figure 4.58 through Figure 4.63. The differences between the


models

non-linear and

linear

for this input

are similar

to those of the step steer input. The two models agree very
and

well

for forward

velocities of

30 km/hr

49.84 km/hr. As

with

the step steer

input,

at

the higher speeds the linear


angle magnitudes

model predicts peak

lateral velocity, sideslip angle,

and tire

slip

below those

of

the

non-linear model and predicts peak yaw velocities and

92

Chapter 4

Two Degree-of-Freedom Vehicle Model

lateral

accelerations above those of the non-linear model.

Differences in peak lateral

acceleration reach

25%. Again,
at

the

linear model

predicts

faster response

than the

non-linear

model.

In particular,

150 km/hr the

response of the non-linear model

lags the linear


zero.

model

by approximately 0.5

seconds after the steer

input is

ramped

back down to

Here differences between the linear

and non-linear

lateral

accelerations reach

nearly 100%. low slip

Comparison
angles and

of

the linear and non-linear model simulations

shows

that at

lateral

accelerations the

linear

vehicle and

tire models

can produce results

comparable

to the non-linear model. Even for the 100 km/hr case


acceleration reaches

where

slip

angles exceed

and the

lateral

0.55 g the linear

model produces results

that are

acceptable

for most engineering purposes. When tire slip


a non-linear

angles and

lateral

accelerations

become high it is necessary to have

tire

model

to obtain

accurate results.

However,

since most

driving is done at low

slip

angles and

lateral accelerations, the linear

model and

the linear analysis techniques presented in Section 4.5.6 through Section 4.5.26

can

be

used

both to study

vehicle

behavior

and

to design vehicles to have desirable


of conditions.

performance characteristics over a wide

variety

operating

93

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5

V
0.0 V
-0.5

30 km/hr
49.84 km/hr

'

-1.0

CO

[
Linear Norl-Linear

100 km/hr

E
-1.5

o o

-2.0

>

2
To

-2.5

\s
\
*s

Linear

-3.0

V= 150 km/hr
-3.5
.

Non-Linear
-4.0

>

-4.5

12

3 Time

(s)

Figure 4.52: Non-Linear Step Steer Lateral

Velocity Response

0.30

/
0.25

\ Linear
.

\/

1 <^n km/hr

ij
;

If

^^^

! Non-Linear

0.20

co

;
0.15

V= 100

km/hr

>

V
>

49.84 km/hr

-10

0.05

0.00

Time

(s)

Figure 4.53: Non-Linear Step Steer Yaw

Velocity Response

94

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.0

V
1

=
*

30 km/hr

0.0 V
-1.0
-

49.84 km/hr

TO

\
-

^^^

' ^
_

V
,

100 km/hr

-2.0

TO C

'

<
-3.0

V.j
i
Linear

CO

CO
-4.0

V= 150 km/hr
t

-5.0
'

^-'

; Non-Linear

-6.0

12

Time

(s)

Figure 4.54: Non-Linear Step Steer Sideslip Angle Response


1.0 V
1
,

=
'

0.0

30 km/hr

V
-1.0

49.84 km/hr

TO o
-2.0

'

TO C

V
-3.0

100 km/hr

<
Q.

CO
-4.0

r-

4_d

*"*
*_

Linear

u-

-5.0

-,

150 km/hr

---

-6.0

^
1 1

Non-Linear

-7.0

12

3 Time

(s)

Figure 4.55: Non-Linear Step Steer Front Tire

Slip Angle Response

95

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.0

0.0
v"^

V
i
!

30 km/hr

'.
-.

-1-0

=
i

49.84 km/hr

TO

CD TO
C

-2.0

'

**"'

"~"

"

~~

<
j?-3.0

100 km/hr

co

2 p
CO

-4.0

V^*-^
x.
1

Linear

-5.0

v.

-v
,
"~

=
i

i50km/nr

_^

Non-Linear

-6.0

-7.0

12

Time

(s)

Figure 4.56: Non-Linear Step Steer Rear Tire


1.4

Slip Angle Response

Linear
1.2
/
V= 150
,
'

/ 1.0
C
-

km/hr

/
o
nS

;
0.8
. . .

Non-Linear

J-j/-

//
o

<

0.6

-If
--i__,

'

.
'

S
0.4 0.2

V= 100 km/hr

49.84 km/hr 30 km/hr

"

'

i^
0.0 12 3

Time

(s)

Figure 4.57: Non-Linear Step Steer Lateral Acceleration Response

96

Chapter 4

Two Degree-of-Freedom Vehicle Model

0.5
;V
=

30 km/hr

0.0

"^vV
y.

49.84 km/hr
.

-0.5

^^

/
"co"

-1.0

'*~^r

100 km/hr

f
/

-1.5

o o

7
\n
--

/ : / ; /
:

>
m
co

-2.0

-2.5

; Linear \_V,H /
,

-3.0

;
-3.5

\^

150 km/hr

Non-Linear
1.5

i.

-4.0

0.5

2 Time

2.5

3.5

(s)

Figure 4.58: Non-Linear Ramp Square Steer Lateral

Velocity Response

0.30

0.25

0.20

co

0.15
o

g
co

0.10

0.05

0.00

-0.05

0.5

1.5

2 Time

2.5

3.5

(s)

Figure 4.59: Non-Linear Ramp Square Steer Yaw

Velocity Response

97

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.0

V 0.0

30 km/hr

^VvV

<S

^
'
"""

49.84 km/hr

^'"*^

'-

-1.0

TO
'
-

-2.0

'

TO

<
-3.0

CO

V \\

/
Linear

g
CO
-4.0

Nv
-5.0
' 1

/
Non-Linear

v= 150 km/hr

1
-6.0
^__^j

0.5

1.5

2 Time

2.5

3.5

(s)

Figure 4.60: Non-Linear Ramp Square Steer Sideslip Angle Response


1.0

0.0

\/

3D

jh<V-.

:_^_

to

y V
-i.o
-2.0

'/
V
=

^r^~\ ^
>"

49.84 km/hr
=

.^

\Xv
c

100 km/hr

/^

<
Q.

CO
-3.0

:\

j
Y_ Linear

: 2

/ / 1 ;
:

-4.0

V \

/ s^/

/ /

/ ;
:
150 km/hr

Nw
-5.0

r^\j=

Non-Linear
i
-6.0

i 1

i
1.5

i
2 Time

i
2.5

1l

0.5

3.5

(s)

Figure 4.61: Non-Linear Ramp Square Steer Front Tire

Slip Angle Response

98

Chapter 4

Two Degree-of-Freedom Vehicle Model

1.0

0.0

^-_-

30 km/hr

^^sj_'

"^y
=

TO

1.0

\
-

49.84 km/hr
=

\\v
?
<
Q.
-2.0

100 km/hr

ffr
/

/-/

\^<^\

X
.,

CO
-3.0

\ \

^
_,

*/
:

/
/
/

:
,/

:/

1
CO
-4-0

\\

Linear

/
/
V

f
/
=

\N'

150 km/hr

-5.0

;
-6.0

Non-Lin ear
1 1

0.5

1.5

2.5

3.5

Time

(s)

Figure 4.62: Non-Linear Ramp Square Steer Rear Tire


1.2

Slip Angle Response

1.0

0.8

3
c

S
o o

0.6

<

0.4

2
15
0.2

0.0

-0.2

0.5

1.5

2 Time

2.5

3.5

(s)

Figure 4.63: Non-Linear Ramp Square Steer Lateral Acceleration Response

99

5
In
the

Conclusion
top
speeds of automobiles

early

part of

this century as the


consideration

increased

vehicle

dynamics became
higher
and

an

important

for engineers. Manufacturers had to


areas of

meet

higher

standards of

performance, particularly in the

safety

and comfort.

Mathematical modeling

of vehicle

dynamics has become

an excellent

way for

engineers

to

study

vehicle

behavior and to
There is

reduce the time and cost to

develop vehicles which meet

performance goals.

a great

deal

of literature on

the topic of vehicle dynamics.

Lateral

vehicle

dynamics in

particular

has been

topic of great interest due to its


of

relationship

with safety.

Two

areas of

focus in the literature concerning the modeling

lateral dynamics have been the two degree-of-freedom vehicle

model and models of tire

behavior. Since tires play

an

extremely important role in the lateral dynamics


representation of tire mechanics

of road

vehicles, sufficiently accurate


models.

is

essential

for vehicle

In Chapter 3

an overview of

tire lateral force mechanics was given. Two

representations of tire

lateral forces

were used.

In the linear tire

model

the lateral force was

considered

to be a linear function tire data

of

the tire slip angle. The non-linear tire model utilized a to


predict

method called

nondimensionalization

lateral force. In this

method

experimentally

measured

tire lateral force versus slip angle curves for several vertical loads

are normalized and curve

fit. Tire lateral force


angle.

can

then be predicted as a non-linear

function

of

both

vertical

load

and

slip

In Chapter 4 the
were

equations of motion

for

two degree-of-freedom vehicle model

derived from basic

principles of

Newtonian

mechanics.

The

model was

then

developed in two
tire
model.

forms, linear and non-linear. The linear vehicle model utilized the linear
were written

Transfer functions

relating both

yaw

velocity

and

sideslip

angle to

the inputs of steering,

aerodynamic side

force,

and road side slope angle.

Expressions for

100

Chapter 5

Conclusion

steady-state

step input

response gains were

derived from the transfer functions. Several

other measures of steady-state

stability

were

derived

including the understeer gradient and


natural

tangent speed. Expressions for transient response characteristics such as

frequency,

damping ratio, and poles and zeros were developed. Numerical simulation of the response
of

the model to step steer, ramp-step steer, ramp-square steer, sine steer, step

aerodynamic

side

force,

and

step

road side slope

inputs

was performed.

It

was seen

that the

steady-state

and

transient response characteristics of the vehicle were very dependent upon its forward

speed.

In particular,

when

the

forward

speed was above the tangent speed of the vehicle,

the zero associated with sideslip angle response to


of

steer

input became

positive.

The

effect

this on the vehicle

was seen

clearly in the input

frequency response and in the simulation.


the linear model
predicted

For

some combinations of speed and

magnitude

lateral

accelerations

higher than

were

actually

possible

due to the

assumption of linear

tire
and

behavior. In

all cases

tested the steady-state response gains,

frequency response,

simulation results were

in

agreement.

The

non-linear vehicle model used a

the

non-linear

tire

model

for predicting tire

lateral forces

during

simulation.

This

model was seen

to predict reasonable responses at

high slip
the

angles and

lateral

accelerations.

Comparison with the linear

model showed

that for

vehicle studied

the linear model was reasonably accurate for most engineering purposes
2

up to slip modeling

angles of

and

lateral

accelerations of 0.5g.

It

was seen

that for accurate

of vehicle response at

high slip

angles and

lateral

accelerations a non-linear

representation of

the tires was

necessary.

101

References

Gillespie,
1992.

Thomas D. Fundamentals of Vehicle Dynamics.

Warrendale, PA: SAE,


Automobile."

Lanchester, F. William. "Some Reflections Peculiar to the Design of an Proceedings of the Institution of Mechanical Engineers, Vol. 2, 1908, p. 187-257.

Olley, Maurice. "Suspension Center, 1937.


Olley, Maurice. "Notes
1961.
on

Handling."

and

Detroit, MI: Chevrolet Engineering

4.

Suspensions."

Detroit, MI: Chevrolet Engineering Center,

Olley, Maurice. "Suspensions Notes


1966.

n."

Detroit, MI: Chevrolet Engineering Center,

Segal, Leonard. "Theoretical Prediction and Experimental Substantiation of the Response of the Automobile to Steering Proceedings of the Automobile Division of the Institution of Mechanical Engineers, 1956-1957.
Control."

Whitcomb, David W.

William F. Milliken. "Design Implications of a General Proceedings of the Automobile Division and of the Institution of Mechanical Engineers, 1956-1957.
and

Theory of Automobile Stability


8

Control."

Bastow, D.
1993.

and

G Howard. Car Suspension

and

Handling. Warrendale, PA: SAE,

9.

Cole, D.E. Elementary Vehicle Dynamics. Ann Arbor, MI: University of Michigan,
1972.

10. Dixon, John C. Tyres, Suspension

and

Handling, Cambridge, England: Cambridge

University Press,
11.

1991.

Ellis, John R. Vehicle Dynamics. London: Business Books, 1969.

12. Ellis, John R. Road Vehicle Dynamics, Akron, OH: J.R. Ellis, 1989. 13. Milliken, William F. PA: SAE, 1995.
and

Doug L.

Milliken. Race Car Vehicle Dynamics.

Warrendale,

14. Mola, Simone. Fundamentals of Vehicle Dynamics, Detroit, MI: General Motors Institute, 1969. 15. Reimpell, Jornsen and Helmut Stall. The Automotive Chassis: Warrendale, PA: SAE, 1996.
16.

Engineering Principles.

Taborek, Jaroslav J. Mechanics of Vehicles. Cleveland, OH: Penton, 1957.

102

References

17. Wong, Jo Yung. 1993.


18.

Theory

of Ground Vehicles

New York: John

Wiley

& Sons, Inc.,

Bundorf, R.T.
of

and R.L. Leffert. 'The Vehicle Directional Control

Properties."

Cornering Compliance Concept for Description


SAE Paper No. 760713, Oct. 1976.
and

19. Allen, R. Wade, Theodore J. Rosenthal, Transient Analysis of Ground Vehicle 20. Heydinger,

Handling."

Henry T.

Szostak. "Steady State and SAE Paper No. 870495, 1987.


of

Dynamics."

Gary J.
Ph.D.

"Improved Simulation

and

Validation

Dissertation, Ohio State University, Columbus, Ohio,


Control."

Road Vehicle Handling 1990.

21

Xia, Xunmao. "A Nonlinear Analysis of Closed Loop Driver/Vehicle Performance with Four Wheel Ph.D. Dissertation, Department of Mechanical Steering Engineering, Clemson University, Clemson, SC, Dec. 1990.
Program."

22. Trom, J.D., J.L. Lopex, and M.J. Vanderploeg. "Modeling of a Mid-Size Passenger Car Using a Multibody Dynamics Transactions of the ASME, Journal of Mechanisms, Transmissions, and Automation in Design, Vol. 109, Dec. 1987. 23. Kortum, W. and W. Schiehlen. "General Purpose Vehicle System Dynamics Software Based on Multibody Vehicle System Dynamics, No. 14, 1985, p. 229263.
Formalisms."

24. Clarke, S.K. (Ed.). Mechanics of Pneumatic Tires, DOT HS-805952, US Government Printing Office, Washington, DC, 1981.
25. Gim, Gwanghun and Parviz E. Nikravesh. "An Analytical Model of Pneumatic Tyres for Vehicle Dynamic Simulations. Part 1: Pure International Journal of Vehicle Design, Vol. 11, No. 6, 1990.
Slips."

26. Bakker, Egbert, Lars Nyborg, and Hans B. Pacejka. "Tyre Modelling for Use in SAE Paper No. 870421, 1987. Vehicle Dynamics
Studies."

27. Radt, Hugo S. and D.A. Glemming. "Normalization of Tire Force Tire Science and Technology, Vol. 21, No. 2, Apr.-June 1993.

and

Moment

Data."

28. Allen, R. Wade, Raymond E. Magdaleno, Theodore J. Rosenthal, David H. Klyde, and Jeffrey R. Hogue. 'Tire Modeling Requirements for Vehicle Dynamics SAE Paper No. 950312, Feb. 1995.
Simulation."

29.

Society of Automotive Engineers.


1976.

"Vehicle Dynamics

Terminology."

SAE J670e,

30. Radt, Hugo S. "An Efficient Method for Treating Race Tire Force-Moment SAE Paper No. 942536, Dec. 1994.
31
.

Data."

Meriam, James L.
York: John

and

Wiley

L. Glenn Kraige. & Sons, 1992.

Engineering Mechanics: Dynamics.

New

103

References

32. Katz, Joseph. Race Car Aerodynamics: Designing for Speed. Cambridge, Massachusetts: Robert Bentley, Inc., 1995. 33. Franklin, Gene F., J. David Powell, and Abbas Emami-Naeini. Feedback Control of Dynamic Systems. New York: Addison-Wesley Publishing Company, Inc., 1994. 34. MATLAB Reference Guide. The MathWorks, Inc., 1994.

104

Appendix A

Tire Model MATLAB Programs

A.l MagicFit.m
%MagicFit % % % % % Created 4/21/96 % J. Kiefer
% Initialization
clear all

Curve

Fitting

of

Tire Data to Magic Formula

Finds

parameters

for Magic Formula


slip
angle

curve
read

fit

of

tire lateral force

or

aligning

moment vs.

data

from file TireSlip.dat

ele;

% Load Data from File load TireSlip.dat


t y
=

TireSlip(:,l) TireSlip(:,2)

; ;

% Find Curve Fit Parameters


xO x
=

[.7407 1.35 1.00

-0.5];
,

leastsq (

'MagicError'

xO,

[]

[], t,

y)

% Construct Fit Function


tl
psi
=

linspace(0/max(t)
=

,10)

(l-x(4))*tl
=

x(4)/x(l)*atan(x(l)*tl);

theta

x(2)*atan(x(l)*psi);

x(3)*sin(theta) ;

% Plot Data
plot(tl, title
(['

and

Fit Function

F,
',

t, y,
D=-

'o')
(B='

Tire Data Magic Formula Fit


num2str(x(3))

num2str(x(l)

',

C='

num2str(x(2)

',

E='

num2str

(x(4) )

')'])

xlabel

'

'

ylabeK'y')
grid

105

Appendix

ATire

Model MATLAB Programs

A. 2 MagicError.m
function e %MagicError
%
=

MagicError(x,

t,

y)

Error in Magic Formula Curve Fit

%e

MagicError(x, t,
Calculates
x and

y)

vector of errors

of

Magic Formula

curve

fit

given parameters

data

(t,

y)

% %
Inputs:

Vector
% x(l) x(2) x(3) x(4)

of

curve

fit

parameters

B c
D E

% %

% %
% % % % % %
t y

Vector Vector

of of

independent data
dependent data
between data
and

Outputs:
e

Vector

of errors

fit function

Created 4/21/96
J.

Kiefer

psi

(l-x(4))*t
=

x(4)/x(l)*atan(x(l)*t);

theta
F
=

x(2)*atan(x(l)*psi);

x(3)*sin(theta);

F;

106

Appendix ATire Model MATLAB Programs

A. 3 NLTire.m
function Fy %NLTire %
=

NLTire(Fz,

alpha)

Nbn Linear Tire Model Lateral Force

%Fy
%

NLTire(Fz,

alpha)

%
% % % % % % % % % % % %

Calculates tire lateral force from inputs


angle.

of

tire

vertical

load

and

slip

Based

on

Radt's tire data

nondimensionalization model

and

the

Magic Formula D0F2NLDE.m.

model.

Force is for

one

tire.

Called

by

the function

Inputs:
alpha

Tire slip Tire

angle

(rad)

Fz Outputs:

vertical

load (N)

Fy
Created 2/18/96
J. Kiefer

Tire lateral force

(N)

global

Bl CI Dl El B3 C3 B5

C5;

% Normalization Parameters
Cc
mu
=

B3 B5

+ +

C3*Fz; C5*Fz;

% N/deg/N % N/N

Cornering
Friction

coefficient

coefficient

% Normalized Slip Angle


alphaN
=

Cc.*tan (alpha)

./mu*180/pi;

% Normalized Lateral Force


psiFN
=

(l-El)*alphaN
=

El/Bl*atan(Bl*alphaN)
;

thetaFN

Cl*atan(Bl*psiFN)
;

FyN

Dl*sin( thetaFN)

% Lateral Force

Fy

-FyN.*mu.*Fz;

107

Appendix B

Two DOF Model Mathematica Session

Stability Derivatives
SDRules
=

{Yb
Yd

->

Cf
-Cf,

Cr,
Nb
Nd
->

Yr
a
-a

->

(a
-

Cf

b
Nr

Cr)/V#
->

->

Cf

Cr,

(aA2

Cf

bA2

Cr)/V,

->

Cf}

Cf

b Cr
,

{Yb

->

Cf

Cr,

Yr

->

Yd

->

-Cf,

Nb

->

Cf

b Cr,

2
a

2
Cf
+

Cr
,

Nr

->

Nd

->

-(a

Cf)}

Transformed Equations
A
=

of Motion

{{s-Yb/(m

V),

l-Yr/(m

V)},

{-Nb/Izz,
MatrixForm Yb
s
-

s-Nr/Izz}};

[A]
Yr

1
m

Nb
-(-

Nr

-(

)
Izz

Izz
Bl
=

{Yd/(m

V),

Nd/Izz};

MatrixForm

[Bl]

Yd

Nd

Izz
B2
=

{l/(m

V),

(a-c)/Izz};

MatrixForm

[B2]

Izz

108

Appendix B

Two DOF Model Mathematica Session

B3 {g/V, 0}; MatrixForm[B3]


=

B4

{l/(m

V),

(a-d)/Izz>;

MatrixForm

[B4]

Izz

Transfer Function Denominator


Ds
=

Collect [Det

[A]

,s]

Nb
+

2
s
+

Nr

Yb
+

Nr
s

Yb

Nb Yr

(-(
Izz

)
m

)
V

Izz

Izz

Izz

Transfer Function Numerators


Nbd
=

Collect [Det [Transpose


Nd Nr Yd
s

[Rep lace Part [Transpose [A],Bl,l]]],s]


Nd Yr

Yd

Izz

Izz

Izz

Nba

Collect [Det [Transpose [ReplacePart

Transpose
a c

[A],B2,l]]],s]
Nr
s a

Yr

Yr

Izz Nbt
=

Izz

Izz

Izz

Izz

Collect [Det [Transpose [ReplacePart

Transpose [A],B3,l]]],s]

g Nr
-(

)
Izz

109

Appendix B

Two DOF Model Mathematica Session

Nrd

Collect [Det [Transpose [ReplacePart [


,

Transpose [A] Nd
s

Bl, 2 ] ] ]
Nb Yd

s]

Nd Yb
+

Izz Nra
=

Izz

Izz

Collect [Det [Transpose [ReplacePart [

Transpose
a
c

[A],B2,2]]],s]
Nb
a

Yb

Yb

Izz Nrt
=

Izz

Izz

Izz

Izz

Collect [Det [Transpose [ReplacePart [


,

Transpose [A]
g Nb

B3

2] ] ]

s]

Izz Nrn
=

Collect [Det [Transpose [ReplacePart [


,B4,2]

Transpose [A]
ad

] ]

s]

Nb

Yb

d Yb

Izz

Izz

Izz

Izz

Izz

Transfer Functions

Sideslip Angle
Gbd
=

Nbd/Ds Nd
Nr

Yd

Yd

Nd Yr

Izz

Izz

Izz

Nb
+ s

2
+

Nr

Yb
+ s

Nr

Yb

Nb

Yr

(-(
Izz

)
m

)
V Izz
m

Izz Gba
=

Izz Nba/Ds
a c

Nr

Yr

Yr

Izz

Izz

Izz

Izz

Izz

Nb
+ s

2
+

Nr

Yb
+

Nr
s

Yb

Nb Yr

(-(
Izz

)
m

)
V Izz
m

Izz

Izz

110

Appendix B

Two DOF Model Mathematica Session

Gbt

Nbt/Ds g Nr g
s

(
Izz V

Nb
+

2
s
+

Nr

Yb
+

Nr
s

Yb

Nb Yr

(-(
Izz

)
m

)
V
Izz
m

Izz

Izz

Yaw
Grd

Velocity
=

Nrd/Ds
Nd
s

Nd Yb
+

Nb Yd

Izz

Izz

Izz

Nb
+

2
s
+

Nr

Yb
+

Nr
s

Yb

Nb Yr

(-(
Izz

)
m

)
V
Izz
m

Izz

Izz Nra/Ds

Gra

Nb

Yb

Yb

Izz

Izz

Izz

Izz

Izz

Nb
+

2
s
+

Nr

Yb
+ s

Nr

Yb

Nb Yr

(-(
Izz

)
m

)
V Izz
m

Izz

Izz

Grt

Nrt/Ds g Nb

Nb
Izz V

2
+ s +

Nr

Yb
+ s

Nr

Yb

Nb Yr

(
Izz

(-(
Izz

)
m

)
V Izz
m

)
V

Izz

Grn

Nrn/Ds

ad

Nb

Yb

d Yb

Izz

Izz

Izz

Izz

Izz

Nb
+ s

2
+

Nr

Yb
+

Nr
s

Yb

Nb Yr

(-(
Izz

)
m

)
V
Izz
m

Izz

Izz

111

Appendix B

Two DOF Model Mathematica Session

Steady

State Step-Input Response Gains

Sideslip Angle
Sbd
=

Simplify [Limit [Gbd,


Nd V)
-

s->0]

-(m

Nr

Yd

Nd Yr

Nb V
=

Nr Yb

Nb Yr
s->0]

Sba

Simplify [Limit [Gba,

-Nr

-amV+cmV+aYr-cYr

Nb V

Nr

Yb

Nb Yr s->0]]

Sbt

Simplify [Limit [Gbt,


g
m

Nr

_(

)
m

Nb V

Nr

Yb

Nb Yr

Yaw
Srd

Velocity
=

Simplify [Limit [Grd,


Yb)
+
+

s->0]]

-(Nd

Nb Yd

Nb V
=

Nr Yb

Nb Yr
s->0]

Sra

Simplify [Limit [Gra,


-

Nb

Yb

Yb

Nb V
=

Nr

Yb

Nb Yr
s->0]

Srt

Simplify [Limit [Grt,


g
m

Nb

Nb V
=

Nr Yb

Nb Yr
s->0]

Srn

Simplify [Limit [Grn,


-

Nb

Yb

d Yb

Nb V

Nr Yb

Nb

Yr

Front Tire
Safd
=

Slip

Angle
+ a/V

Simplify [Sbd
2

Srd

1]

2
+

(m Nb V

Nd V

Nd Yb

Nr V Yb

Nb Yd

Nr V Yd

Nb V Yr

Nd V

Yr)

(V

(-(m Nb V)

Nr

Yb

Nb Yr)

112

Appendix B

Two DOF Model Mathematica Session

Safa

Simplify [Sba

a/V

Sra]
2 2
+a Yb-acYb-aVYr+

2
(-

(a Nb)

+NrV+amV

-cmV

V Yr)

(V

(-(m Nb V)
+ a/V

Nr

Yb

Nb Yr)

Saft

Simplify [Sbt
g
m

Srt]

(a Nb

Nr

V)
Nb Yr)

(m Nb V

Nr Yb

Rear Tire
Sard

Slip Angle
Simplify [Sbd
2
-

b/V

Srd]

Nd V

b Nd Yb

b Nb Yd

Nr V Yd

Nd V Yr

V Sara
=

(-(m Nb V)
-

Nr

Yb

Nb Yr)

Simplify [Sba

b/V

Sra]
2

2 (bNb
+

NrV

amV

-cmV

-abYb

bcYb-aVYr

V Yr)

(V

(-(m Nb V)
-

Nr

Yb

Nb

Yr) )

Sart

Simplify [Sbt
g
m

b/V

Srt]

(b Nb

Nr

V)
+

(-(m Nb V)

Nr

Yb

Nb Yr)

Path Curvature
Scd
=

Simplify[l/V

Srd]

(Nd Yb)

Nb Yd

V Sea

(m Nb V
=

Nr Yb

Nb Yr)

Simplify [1/V
Nb
-

Sra]
Yb

Yb

V Set

(m Nb V
=

Nr

Yb

Nb Yr)

Simplify[l/V

Srt]

Nb

(m Nb V

Nr Yb

Nb

Yr)

113

Appendix

BTwo

DOF Model Mathematica Session

Lateral Acceleration
SAd
=

Simplify [V/g
(-(Nd Yb)
+

Srd]

Nb Yd)

g
SAa

(m Nb V
=

Nr Yb

Nb Yr)

Simplify [V/g
(Nb
-

Sra]
c

Yb

Yb)
Nb Yr)

g SAt

(m Nb V
=

Nr Yb

Simplify [V/g
m

Srt]

Nb V

Nb V

Nr Yb

Nb Yr

Steer Angle Response to Path Radius


deltaR
=

delta
2

/.

Solve[Scd

==

1/R

delta, delta] [ [1, 1] ]

(m Nb V

Nr V Yb

Nb V Yr

_(
-(Nd

)
R Yb)
+

Nb

R Yd

Terml

Coefficient [Expand

[deltaR]

,V,

2]

VA2

2
m

Nb V

-(Nd

R Yb)
=

Nb

R Yd

TermlS

Simplify [Terml
2

/.

SDRules]

(a Cf

b Cr)

Cf

Cr
=

b Cf Cr R

TermlSa

Numerator

[TermlS]
a->L-b]

Simplify [Denominator [

TermlS]
2 (a Cf
-

/.

b Cr)

Cf

Cr

114

Appendix B

Two DOF Model Mathematica Session

Terms 2 3

ExpandNumerator

[Simplify [Coefficient

[Expand

deltaR], V]
-(Nr

V]]

Yb)
-

Nb V Yr

Nd

R Yb

Nb

R Yd

Terms23S

Simplify [Terms23

/.

SDRules

/.

a->L-b]

deltaRl

Terml

Terms23S

2
L
-

m
+
-

Nb V

(Nd R Yb)
=

Nb

R Yd Terms23S

deltaR2

TermlSa

2
L
-

(a Cf
+

b Cr)

Cf

Cr

Understeer Gradient
Kus
=

Coefficient

[Simplify [deltaRl

g]

,VA2]

Nb

-(Nd

Yb)

Nb Yd

Kusl

Simplify [Kus
-

/.

SDRules]

(a Cf

b Cr)

(a

b)

Cf

Cr

Stability
Kl
=

Factor

Simplify [K
2

/.

Solve [Srd

==

V/(L

(1+K

VA2)),

K][[l]]]

Nb V

L Nd Yb

Nr V Yb

L Nb Yd

Nb V Yr

2
L V

(-(Nd Yb)

Nb Yd)

115

Appendix B

Two DOF Model Mathematica Session

K2

Simplify [Numerator [Kl]


VA2]
VA2

Coefficient [Numerator [Kl]


+

/.

SDRules
,

/. VA2

a->L-b]

Coef f icient [

Numerator

[Kl]

VA2]

Denominator [Kl]

Nb

L K3

(-(Nd Yb)

Nb Yd) /.
m

Simplify [K2
(a Cf
-

SDRules]

b Cr)

Cf
=

Cr

b Cf Cr L

K4

Numerator [K3]
a
->

/Simplify[ (Denominator [K3]

/,

b)]
m

(a Cf

b Cr) 2

Cf

Cr

Neutral Steer Point


dl
=

Simplify [d
Nb

/.

Solve [Numerator

[Srn]

==

0,

d] [ [1] ] ]

--

Yb

d2

Simplify [dl
+

/,

SDRules]

(a

b)
+

Cr

Cf

Cr

d3

Simplify [d2
L

/.

>

b]

Cr

Cf

Cr

Static Margin
SM
=

(dl
Nb

a)/L

-(

)
L Yb

SMI

Simplify [SM
Cf)
L
+ +

/.

SDRules]

-(a

b Cr

Cf

Cr

116

Appendix B

Two DOF Model Mathematica Session

Tangent Speed
Vtan
=

/.
-

Solve[Sbd

deltaR

==

0,V][[2,1]]

Nr Yd
-(

Nd Yr

)
m

Nd Sqrt

Vtanl

[Simplify [Expand[
/.

(V

/.

Solve [V

==

Simplify [

Vtan

SDRules]

,V]

[ [2,1] ]) A2] ] ]

(a

b)
m

Cr

Sqrt[-(
a

)]

Vtan2

Sqrt

[Simplify [Numerator [VtanlA2]


A

/.

->

L-b] /

Denominator [Vtanl b Cr
L

2 ] ]

Sqrt[-(
a m

)]

Critical Speed
Vcrit
=

/.

Solve [Denominator

[Srd]

==

0,V][[1]]

Nr Yb
-(

Nb Yr

)
m

Nb Sqrt Solve [Simplify [ (V /. SDRules,V] [[2,1]])A2]]

Vcritl

[V

==

Vcrit

2 (a
Sqrt
+

b)
+

Cf

Cr

[
-

]
(a Cf m)
=

b Cr

Vcrit2

Sqrt

[Simplify [Numerator [VcritlA2]

/.

->

L-b]/

Denominator [Vcritl A2] ]

2
Cf Cr L

Sqrt

[
-

]
(a Cf m)
+

b Cr

Characteristic Speed
Vchar
=

/.
+

Solve[deltaR

==

L/R, V] [ [2, 1] ]
(2 Nd Yb
-

(-(Nr Yb)

Nb Yr

Sqrt [-4

LmNb

Nb Yd)

2
(Nr Yb
-

Nb Yr)

])

(2

Nb)

117

Appendix B

Two DOF Model Mathematica Session

Vcharl

Sqrt

[Simplify [ (V

/.

Solve [V

==

Vchar

/-

SDRules, V] [[2,1]])A2]]
(a
+

b)

Cf

Cr

(a

L)

Sqrt[
-(a

]
Cf m)
+

b Cr

Vchar2

Sqrt

[Simplify [Numerator [Vcharl A2]

/.

->

L-b]/

Denominator [Vcharl A2] ]


2
Cf Cr L

Sqrt[-(
-(a

)]
Cf
m)
+

b Cr

Yaw Radius
kz
=

of

Gyration

Sqrt[Izz/m]
Izz

Sqrt[
m

Geometry to Inertia Ratio


GIR
=

LA2/kzA2

2
L
m

Izz

Total
TCF

Cornering Factor
=

Cf Cr

Cr/mA2

Cf

2
m

Characteristic Equation
Ds
==

0 2
+

Nb
s

Nr Yb
+ + s

Nr

Yb

Nb Yr

(-(
Izz

)
m

)
V
Izz
m

==

Izz

Izz

a2

Coef f icient

[Ds,

sA2]

118

Appendix B

Two DOF Model Mathematica Session

al

Coef f icient [Ds, s]


Yb

Nr
-(

)
Izz
m

V
sA2
-

aO

Ds

a2

al

Nb
+

Nr

Yb
_

Nb Yr

Izz

Izz

Izz

Undamped Natural Frequency


wn
=

Simplify [ Sqrt [aO ] ]


in

Nb V

Nr Yb

Nb Yr

Sqrt[
Izz
wnl
=

]
m

Simplify [wn
2
a

/.

SDRules]
2 2
Cf Cr
+

2
-

Cf

Cr

b Cf Cr

Cf

b Cr

Sqrt[
2
Izz
m

]
V

Damping
zeta

Ratio

Simplify [al
-

Izz
+

V/(2

Sqrt[wnA2

(Izz

V)A2])]

(m Nr V

Izz

Yb)
Nb Yr)

Sqrt [Izz
=

mV

(hi Nb V

Nr Yb

zetal

Simplify [zeta
2

/.

SDRules]
2

-(Cf

Izz

Cr

Izz

Cf

Cr m)

/
2

2 (2
Sqrt [Izz
m

2
Cf Cr
+

(a

Cf Cr

b Cf Cr

Cf

2 b Cr
m

) ] )

119

Appendix B

Two DOF Model Mathematica Session

Poles
poles
=

Solve [Ds==0, s] ;
/poles

si

[[1,1]]
2

(m Nr V

Izz

Yb

Sqrt[(-(m Nr V)
+

Izz

Yb)

4
sla
=

Izz

(m Nb V
/.

Nr

Yb

Nb Yr)

] )

(2

Izz

V)

Simplify [si

SDRules]
2
Cf
+

( (Cf

Cr)

Izz

(a

Cr)
2

2
Cf
m
-

2
Cr m)
-

Sqrt[(-(Cf

Izz)
2

Cr

Izz

2
Cf Cr
+

2
Cf Cr
+

Izz

(a

b Cf Cr

Cf

b Cr
s2
=

) ] )

(2

Izz

V)

/.

poles

[[2,1]]
2

(m Nr V

Izz

Yb

Sqrt[(-(m Nr V)

Izz

Yb)

4
s2a
=

Izz

(m Nb V
/. 2

Nr

Yb

Nb Yr)

] )

(2

Izz

V)

Simplify [s2

SDRules]
2
Cf
+

( (Cf

Cr)

Izz

(a

Cr)
2

2
Cf
m
-

2
Cr m)

Sqrt[(-(Cf

Izz)
2

Cr

Izz

2
Cf Cr
+

2
Cf

Izz

(a

b Cf Cr

Cr

Cf

b Cr

) ] )

(2

Izz

V)

Zeros
Zbd
=

/.

Solve [Nbd
-

==

0,

s][[l,l]]

(m Nd V)

Nr

Yd

Nd Yr

_(

}
Izz Yd

120

Appendix B

Two DOF Model Mathematica Session

Zbdl

Simplify [Zbd
2

/.

SDRules]

2
Cr+amV

abCr+b

Izz

V
==

Zba

/.

Solve [Nba

0,

s][[l,l]]

-Nr

-amV+cmV+aYr-cYr

_(

)
Izz

Zbal

Simplify [Zba
2

/.

SDRules]
2 2
-cmV

acCf+abCr+b

Cr-bcCr+amV

Izz

Zbt

/.

Solve [Nbt

==

0,

s][[l,l]]

Nr

Izz Zbtl

Simplify [Zbt
2

/.

SDRules]

2
a

Cf

Cr

Izz V Zrd
=

/.

Solve [Nrd
+

==

0,

s][[l,l]]

(Nd Yb)

Nb Yd

_(

)
m

Nd V
/-

Zrdl

Simplify [Zrd

SDRules]

(a

b)
m

Cr

Zra

/.
a

Solve [Nra Yb
+ c

==

0,

s][[l,l]]

Nb
-(

Yb

)
(a
-

c)

Zral

Simplify [Zra
+

/.
-

SDRules]
c

-(c

Cf)

Cr

b Cr

Cr

121

Appendix B

Two DOF Model Mathematica Session

Zrt

Solve [Nrt

==

0,

s]

{{}}

122

Appendix C

Two DOF Model MATLAB Programs

C.l DOF2Control.m
%DOF2Control
Controls

2 DOF Model Execution Control 2 DOF

execution of

model.

Sets

control
.

input type (step, step


simulation parameters.

ramp,

ramp step / ramp down,

or

sine

steer)

Sets

% Created 1/11/96 % J. Kiefer

% Control Input Type


step
ramp
=

1; 2;
=

rampsquare sine
=

3;

% Step steer % Ramp step steer % Ramp square steer % Sine


steer

4;
=

input

1;

% Select

which control

input to

use

% Simulation Parameters
tO
tr
=

td
ts tf

tol

0.0; 0.2; 1.0; 1.0; 4.0; le-5;

% % % % % %

Initial time for

steer

input

s s s
s

Ramp

time

EWell time Period for sine Final time for


steer

simulation

Simulation accuracy

(default

le-3)

123

Appendix

CTwo

DOF Model MATLAB Programs

C.2 DOF2Param.m
%DOF2Param
2 DOF Model Independent Parameters
and

Simulation Control

Sets independendent vehicle,

tire,

control,

and

disturbance

parameters

for 2 DOF

model.

% Created 1/7/96 % J. Kiefer % Initialization


clear

all;

clc;
global m
global

Izz L

c u

Cf Cr dO Fzf Fzr

Fyg

Fya tO tr td ts tf

input;

Bl CI Dl El B3 C3 B5

C5;

% Constants
g
=

9.81;

m/s~2

Acceleration due to gravity

% Vehicle Independent Parameters


m
=

Izz

f
L
u

1775; 1960; 0.52; 2.372; 100;


=

% % %

kg

Gross

vehicle mass

% kg-m^2

Yaw inertia
Fraction
of weight on

front

axle

Wheelbase
Vehicle forward
speed

% km/hr

% Control dO
==

and

Disturbance Inputs

1;
=

theta Fya
c
=

0;

% % %

deg deg
m

Steer input Side


slope

magnitude

0; 1.25;

% N

Aerodynamic

side

force
axle side

Distance from front


aerodynamic

to

force

% Linear Tire Model Parameters


Cf
Cr
==

-1230.5; -1155.5;

==

% N/deg % N/deg

Front cornering

stiffness

(one tire) (one tire)

Rear cornering stiffness

% Non Linear Tire Model Parameters


%
Bl CI
Dl
=

Normalized Lateral Force Magic Formula Parameters

0.5835;
1.7166 1.0005 0.2517
Coefficient Parameters

El

%
B3 C3
=

Cornering 0.333;
-1.352e-5;

%
B5
C5
=

Friction Coefficient Parameters

1.173;
-3.696e-5;

% Unit Conversions
u
=

u*1000/3600;

m/s

Vehicle forward speed

124

Appendix

CTwo

DOF Model MATLAB Programs

dO
Cf Cr

d0*pi/180;
Cf*180/pi*2; Cr*180/pi*2;

rad

Step

steer

input
(two tires) (two tires)

% N/rad % N/rad

Front tire cornering stiffness


Rear tire cornering stiffness

125

Appendix

CTwo

DOF Model MATLAB Programs

C.3 DOF2DependParam.m
%DOF2DependParam
% % % % Created 1/7/96 % J. Kiefer % Dependent Parameters
a
=

2 DOF Model Dependent Parameter Calculation

Calculates

values

of

dependent

parameters

for 2 DOF

model.

b
V

(l-f)*L; f*L;
u;
=

%
%

m m

Distance from front tire to C.G. Distance from Vehicle


speed rear

tire to C.G.

%
m*g*sin(theta*pi/180) ; m*g*f/2*cos(theta*pi/180); m*g*(l-f)/2*cos(theta*pi/180);

m/s

Fyg
Fzf
Fzr

% N % N % N

Side

slope

lateral force
normal

Front tire Rear tire

load (one tire) load (one tire)

normal

% Stability Derivatives
Yb
Yr
=

Cf

Cr;

% N/rad % N-s/rad % N/rad % N-m/rad % N-m-s/rad % N-m/rad

Damping-

in-sideslip
/
yaw

(a*Cf-b*Cr) /V;
-Cf

Lateral force

coupling

Yd

Control force
Directional stability Yaw

Nb
Nr Nd

a*Cf-b*Cr;

(a~2*Cf+b~2*Cr) /V;
-a*Cf ;

damping
moment

Control

126

Appendix

CTwo

DOF Model MATLAB Programs

C.4 SteerAngle.m
function delta %SteerAngle
%
=

SteerAngle (t,

input, t0,tr,td,ts,tf

,d0)

Calculate Steer Angle Based on Time

and

Control Input Selection

%delta % %
%

SteerAngle

(t, input, tO, tr,td,ts,tf,dO)


based
on

Determines

steer angle

current

time,

input selection,
ramp,

and
or

simulation parameters. sine.

Input

selection

may be step,

ramp square,
D0F3NLDE.m.

%
% % %

Called

by functions D0F2LDE.m, D0F2NLDE.m, D0F3LDE.m,

Inputs: t

% %
%

input

(s) Flag for input


Time

selection

1 2 3 4
to tr

step ramp
rampsquare

%
% % % % % % % % % % % % %
Created 1/7/96
J.

sine
simulation

Initial time for

(s)

Ramp
EWell

time

td ts
tf

(s) time (s)


sine
steer

Period for

(s) (s) (rad)

Final time for Steer input

simulation

dO

magnitude

Outputs:

delta

Steer

angle

(rad)

Kiefer

% Crash Through Result


delta
=

0;

% Step Steer if input == 1 delta


=

dO;
tO

if t

<

delta
end end

0;

% Ramp Step Steer if input == 2 delta


=

dO;
tO
+

if t

<

tr
=

delta
end

d0*(t-t0)/tr;

if t
end
end

<

tO

delta

0;

% Ramp Square Steer if input == 3

127

Appendix

CTwo

DOF Model MATLAB Programs

delta

0;
tO
+

if t
end

<

2*tr
=

td

delta

d0*(t0+td+2*tr-t)/tr;
+

if t

<

tO

tr
=

td

delta
end

dO;

if t

<

tO

tr
=

delta
end

d0*(t-t0)/tr;

if t
end end

<

tO

delta

0;

% Sine Steer

if input

==

4
=

delta

d0*sin(2*pi*(t-t0)/ts)
delta
=

if t
end end

<

tO

0;

128

Appendix

CTwo

DOF Model MATLAB Programs

C.5 DOF2LFreq.m
%DOF2LFreq
% For % % % %
Generates bode
plot a

Frequency Response
of

of

Linear 2 DOF Model

Single Set

Parameters

data for linear 2 DOF


speed,
and

model

for

outputs

of

sideslip

angle

and yaw

for inputs

of

steer angle

control,

aerodynamic

side

force disturbance,

and road side slope

disturbance.

% % Created 2/4/96
% J. Kiefer

D0F2Param;
D0F2Control ;

% Set independent % Set


execution

parameters parameters

control

D0F2DependParam;
% Transfer Function Denominator
D
=

% Calculate dependent

parameters

[1

-Nr/Izz-Yb/

(m*V) Nb/Izz+(Nr*Yb-Nb*Yr) / (Izz*m*V) ]

% Transfer Function Numerators Nbd


Nba
=

[Yd/(m*V) (Nd*Yr-Nr*Yd-Nd*m*V) / (Izz*m*V) ] ; [1/ (m*V) (c-a) /Izz+ ( (a-c) *Yr-Nr) / (Izz*m*V) ]
[g/V
-g*Nr/(Izz*V)] ;

Nbt
Nrd Nra Nrt

[Nd/Izz

(Nb*Yd-Nd*Yb) / (Izz*m*V) ] ; [(a-c) /Izz ( (c-a)*Yb+Nb) / (Izz*m*V) ] [g*Nb/ (Izz*V) ] ;

% Bode Plot Data


w

lcgspace(-l,2) *2*pi; bode (Nbd, D,w) ; [Mbd,Pbd,w] bode (Nba, D,w) ; [Mba,Pba,w] bode (Nbt, D,w) ; [Mbt,Pbt,w] bode (Nrd, D,w) ; [Mrd,Prd,w] bode (Nra, D,w) ; [Mra,Pra,w] bode (Nrt, D,w); [Mrt,Prt,w]
=
=

129

Appendix

CTwo

DOF Model MATLAB Programs

C.6 DOF2LSim.m
%D0F2LSim
%
% % % % %
Performs
simulation of

Simulation of Linear 2 DOF Model Response to Control and Disturbance

Inputs

linear 2 DOF
yaw

model

response

to

control

and

disturbance inputs.

Determines

speed,
and

front

and rear

tire slip angles,

front

lateral speed, sideslip angle, rear tire lateral forces, and


time. Reads data from

lateral

acceleration.

Plots these

responses versus

% %

D0F2Param, D0F2DependParam.

% Created 1/7/96

% J. Kiefer

D0F2Param; D0F2Control; D0F2DependParam;


% Perform
simulation

% Set independent % Set

parameters

execution control parameters parameters

% Calculate dependent

[t,x]
v r
=

ode23(,DOF2LDE',0,tf, [0

0]',tol);

x(:,l) ; x(:,2);

% Steer Angle
delta
=

for i

( length ( t) l:length(t)
zeros

1)

delta(i)
end

SteerAngle (t

(i) input, t0,tr,td,ts,tf, dO)


,

rad

Steer angle

% Vehicle beta
=

and

Tire

Slip Angles
% % %
rad

v/u;
=

Vehicle sideslip angle Front tires slip


angle

alphaF alphaR

(v+a*r) /u-delta; (v-b*r)/u;


and

rad
rad

Rear tires slip

angle

% External Forces
Fyf
Fyr
=

Moments

Cf*alphaF; Cr*alphaR;

% N % N

Front tires lateral force Rear tires lateral force

% State Derivatives
vdot rdot
=

(Fyf

Fyr
-

Fya
-

(a*Fyf

b*Fyr

u*r; Fyg) /m (c-a) *Fya) /Izz;


-

% Lateral Acceleration
ay
=

vdot

u*r;

m/s^

Lateral

acceleration

% Do Plots
subplot

(2, 2,1)

plot(t,v)
grid

title

'

Lateral Speed

'

xlabeK'Time

ylabel

(s) ') (' Speed (m/s) ') (2, 2, 2)

subplot

plot(t,r*180/pi)

130

Appendix

CTwo

DOF Model MATLAB Programs

grid

title ('Yaw Speed')


xlabelCTime

(s) ')
'

ylabel( 'Speed (deg/s)


subplot

(2, 2, 3)
,t,alphaR*180/pi,

plot(t,beta*180/pi,t,alphaF*180/pi,
grid

'-.

'

'

'

,t,delta*180/pi,

Sideslip Angle, Tire Slip Angles, (s) ) ylabel ( Slip Angle (deg) )
xlabeK 'Time
'
' '

title ( 'Vehicle

Steer Angle')

subplot

(2, 2, 4)

plot(t,ay/g)
grid

title ( Lateral Acceleration


xlabel( 'Time
ylabel

'

'

(s) ')
(g)
'

'Acceleration

131

Appendix

CTwo

DOF Model MATLAB Programs

C.7 DOF2LDE.m
function %DOF2NLDE
% %xdot
% % % % %
Determines derivatives
state vector. of
=

xdot

DOF2NLDE(t,x)
Non Linear Differential Equations for 2 DOF Model

D0F2NLDE(t,x)

lateral

speed and yaw speed given

time and
with ode23

Non linear tire

and non

linear slip

angles.

Used

for

simulation.

Inputs: t
x(l) x(2)

% %
%

Time

(s)
speed speed

Lateral Yaw

(m/s) (rad/s)
of

% %
% % % %

Outputs:
xdot(l)
xdot

Derivative Derivative

lateral

speed

(2)

of yaw speed

(m/s ^2) (rad/s''2)

Created 2/18/96
J.

Kiefer

global

Izz L

c u

dO Fzf Fzr Fyg Fya tO tr td ts tf input;

delta
alphaF

SteerAngle(t, input, tO, tr,td,ts,tf,dO);


=

atan(
atan(
=

(x(l)+a*x(2) ) /u) -delta; (x(l)-b*x(2) )/u)


;

alphaR

[Fyf, Mzf] [Fyr, Mzr]


xdot
=

NLTire(Fzf,

alphaF); alphaR);

NLTire(Fzr,
+

[-u*x(2)

(2*Fyf*cos (delta)

+2*Fyr+Fya+Fyg)/m

(2*a*Fyf*cos (delta)

-2*b*Fyr+

(a-c) *Fya) /Izz]

132

Appendix C

Two DOF Model MATLAB Programs

C.8 DOF2NLSim.m
%DOF2NLSim
% % % % % %
Performs
simulation of

Simulation

of

Non-Linear 2 DOF Model Response to Control

and

Disturbance

Inputs

non

linear 2 DOF
yaw

model

response

to

control

and

disturbance inputs. front


and rear

Determines

speed,
and

tire slip angles,

front

lateral speed, sideslip angle, rear tire lateral forces, and


versus

lateral

acceleration.

Plots these responses

time.

Reads data from

% %

D0F2Param, D0F2DependParam.

% Created 2/18/96

% J. Kiefer

D0F2Param; D0F2Control;
D0F2DependParam; % Perform
simulation

% Set independent % Set

parameters parameters

execution control

% Calculate dependent

parameters

[t,x]
v r
=
=

ode23(,DOF2NLDE',0,tf, [0

0]',tol);

x(:,l) ;

x(:,2) ;

% Steer Angle delta for i


=

zeros

(length(t)
=

1)

l:length(t)

delta(i)
end

SteerAngle

(t(i) input, tO,tr,td,ts,tf


,

,d0)

rad

Steer

angle

% Vehicle beta
=

and

Tire

Slip

Angles

atan(v/u) ;
=

% % %

rad
rad rad

Vehicle sideslip

angle

alphaF alphaR

atan(
atan(

(v+a*r) /u) -delta; (v-b*r) /u) ;


and

Front tires slip


Rear tires slip

angle

angle

% External Forces
Fyf Fyr
=

Moments

NLTire(Fzf, NLTire(Fzr,

alphaF); alphaR) ;

% N % N

Front tire lateral force Rear tire lateral

(one tire) (one tire)

force

% State Derivatives
vdot
=

(2*Fyf.*cos

rdot

(2*a*Fyf.*cos

u*r; (delta) + 2*Fyr + Fya + Fyg) /m 2*b*Fyr (c-a)*Fya)/Izz; (delta)


-

% Lateral Acceleration ay
=

vdot

u*r;

m/s^2

Lateral

acceleration

% Do Plots
subplot(2,2,l) plot(t,v)
grid

Speed') (s) ') ylabel ('Speed (m/s) ')


title
xlabeK'Time subplot

('

Lateral

(2, 2, 2)

plot(t,r*180/pi)

133

Appendix

CTwo

DOF Model MATLAB Programs

grid

title ('Yaw Speed')


xlabelCTime
ylabel
'

(s) ') ( Speed (deg/s)

'

subplot

(2, 2, 3)
'
' ,t,alphaR*180/pi,

plot(t,beta*180/pi,t,alphaF*180/pi,
grid

'-.

'

'

,t,delta*180/pi,

')

title

'Vehicle

xlabel( 'Time
ylabel
'

Sideslip Angle, Tire Slip Angles, Steer Angle') (s) ') ( Slip Angle (deg) )
'

subplot
plot

(2, 2, 4) (t, ay/g)


' '

grid

title ( Lateral Acceleration


xlabel( 'Time
ylabel

(s)

'

)
'

( 'Acceleration (g)

134

Appendix

CTwo

DOF Model MATLAB Programs

C.9 DOF2NLDE.m
function xdot %D0F2NLDE
%
=

D0F2NLDE(t,x)
Non Linear Differential Equations for 2 DOF Model

%xdot
% %
% % % % %

D0F2NLDE(t,x)
Determines derivatives
state vector. of

lateral

speed and yaw speed given time and

Non linear tire

and non

linear slip

angles.

Used

with ode23

for simulation.
Inputs:
t
x(l) x(2)

Time

(s)
speed

%
% %

Lateral speed
Yaw

(m/s) (rad/s)
of

Outputs:
xdot(l)
xdot

% %
%

Derivative

lateral

speed

(2)

Derivative

of yaw speed

(m/s^2) (rad/sA2)

% %

Created 2/18/96
J.

Kiefer

global m

Izz L

c u

dO Fzf Fzr Fyg Fya tO tr td ts tf input;

delta
alphaF

SteerAngle(t, input, tO, tr,td,ts,tf,dO);


=

atan((x(l)+a*x(2))/u)-delta;
atan(

alphaR

(x(l)-b*x(2) )/u)

Fyf Fyr

NLTire(Fzf, alphaF); NLTire(Fzr, alphaR);

xdot

[-u*x(2)

(2*Fyf*cos (delta) +2*Fyr+Fya+Fyg)/m


-2*b*Fyr+

(2*a*Fyf*cos (delta)

(a-c) *Fya) /Izz]

135

Appendix D

Relevant Literature

"Control

of

Vehicle
-

Dynamics."

Automotive Engineering,

May 1995, p.

87-93.

"Road Vehicles "Road Vehicles

Lateral Transient Response Test

Methods."

ISO 7401,

May

1988.

Steady State

Circular Test
and

Procedure."

ISO 4138, Aug. 1982.


Vocabulary."
-

"Road Vehicles Vehicle Dynamics Dec. 1991.


-

Road-Holding Ability

ISO 8855,

"Vehicle Dynamics

Terminology."

SAE J670e, Warrendale, PA: SAE, 1976.

1994 Motor Sports Engineering Conference Proceedings: Volume 1: Vehicle Design Issues. SAE Publication No. P-287, Dec. 1994.

Allen, R. Wade and Theodore J. Rosenthal. "A Computer Simulation Analysis


Critical Maneuvers for Assessing Ground Vehicle Dynamic No. 930760, Mar. 1993.
Stability."

of

Safety

SAE Paper

Allen, R. Wade

and

Simulation

Models."

Theodore J. Rosenthal. "Requirements for Vehicle Dynamics SAE Paper No. 940175, Feb. 1994.
and

Allen, R. Wade, Raymond E. Magdaleno, Theodore J. Rosenthal, David H. Klyde,

Jeffrey

Simulation."

R. Hogue. 'Tire Modeling Requirements for Vehicle Dynamics SAE Paper No. 950312, Feb. 1995.
and

Allen, R. Wade, Thomas T. Myers,

Considerations with Automatic 931979, Nov. 1993.


DriverA^ehicle
Interaction."

and

Theodore J. Rosenthal. "Vehicle Four Wheel Steering

Systems."

Stability
SAE Paper No.

Allen, R. Wade, Theodore J. Rosenthal,


of

and Jeffrey R. Hogue. "Modeling and Simulation SAE Paper No. 960177, Feb. 1996. and

Allen, R. Wade, Theodore J. Rosenthal,


Transient Analysis
of

Henry

Ground Vehicle

Handling."

T. Szostak. "Steady State and SAE Paper No. 870495, 1987.


and

Allen, R. Wade, Theodore J. Rosenthal, David H. Klyde, Keith J. Owens,


Szostak. "Validation Dynamics Stability
of
Analysis."

Henry T.

Ground Vehicle Computer Simulations Developed for SAE Paper No. 920054, Feb. 1992.
and

Allen, R. Wade, Henry T. Szostak, Theodore J. Rosenthal, David H. Klyde,


Owens. "Characteristics Influencing SAE Paper No. 910234,
Stability."

Keith J.

Ground Vehicle Lateral/Directional Dynamic Feb. 1991. A. Sitchin. "Vehicle Dynamic Handling Correlation, and Application Using

Antoun, R.J, P.B. Hackert, M.C. O'Leary,


--

and

ADAMS."

Model Development, Computer Simulation SAE Paper No. 860574, 1986.


of

Araki, Kazuo and Hideo Sakai. "Study

Tire Model Consisting of Theoretical and Experimental Equations for Vehicle Dynamics Analysis Part 2: Under the SAE Paper No. Condition of Various Velocity on the Asphaltic Road
-

Surface."

960996, Feb. 1996.


Ashley, Steven. "Spin Control for
Cars."

Mechanical Engineering, Vol.

117, No. 6, June

1995,

p.

66-68.

136

Appendix D

Relevant Literature

Bakker, Egbert, Lars Nyborg, and Hans B.


Dynamics
Studies."

Pacejka. 'Tyre Modelling for Use in Vehicle SAE Paper No. 870421, 1987.
Studies."

Bakker, Egbert,
Barak, Pinhas.

Hans B. Pacejka, and Lars Lidner. "A New Tire Model with an SAE Paper No. 890087, 1989. Application in Vehicle Dynamics "Magic Numbers in Design Paper No. 911921, 1991.
of

Suspensions for Passenger

Cars."

SAE

Barbieri, Nilson. "Suspensions Bastow, D.


and

Optimization."

SAE Paper No. 921491, 1992. Handling. Warrendale, PA: SAE, 1993.
and

G. Howard. Car Suspension


and

and

Bernard, James E. Bernard, James E.

Calculations."

C.L. Clover. 'Tire Modeling for Low-Speed SAE Paper No. 950311, Feb. 1995.

High-Speed

and

Dynamics."

C.L. Clover. "Validation of Computer Simulations SAE Paper No. 940231, Feb. 1994.

of

Vehicle

Bixel, Ronald A., Gary J. Heydinger, N.J. Durisek,


Measurement."

and Dennis A. Guenther. "New Developments in Vehicle Center of Gravity and Inertial Parameter Estimation and SAE Paper No. 950356, Feb. 1995.

Blank, Matthew

Donald Margolis. "The Effect of Normal Force Variation on the SAE Paper No. 960484, Feb. 1996. Lateral Dynamics of
and
Automobiles."

Bowman, J. Eric

and

Braking

System."

E.H. Law. "A Feasibility Study of an Automotive SAE Paper No. 930762, Mar. 1993.
and
Friction."

Slip Control

Breuer, Bert, Thomas Bachmann, Stefan Ernesti,


Instruments for On-Board Measurement 942470, Dec. 1994.

of

Jorg Stocker. "Methods and SAE Paper No. Tyre/Road

Bundorf, R.T.

and R.L. Leffert. 'The Vehicle Directional Control and

Properties."

Cornering Compliance Concept for Description of


SAE Paper No. 760713, Oct. 1976.
Italy."

Cambiaghi, Danilo
1994.

Marco Gadola.

"Computer- Aided

Development at the

University
and

of

Brescia,

Racing Car Design and SAE Paper No. 942507, Dec.

Captain, K.M., A.B. Boghani,


Vehicle
Simulation."

D.N. Wormley. "Analytical Tire Models for Dynamic Vehicle System Dynamics, Vol. 8, 1979, p. 1-32.

Car Suspension Systems 1991.

and

Vehicle Dynamics. SAE Publication No. SP-878, Sept.


Suspension Stiffness

Chen H Fred

and

Responses."

Dennis A. Guenther. "The Effects SAE Paper No. 911928, 1991.

of

on

Handling

Traction Chocholek, S.E. "The Development of a Differential for the Improvement of IMechE Paper No. C368/88, 1988.
Control."

Composite Suspension Chrstos, Jeffrey P. "A Simplified Method for the Measurement of SAE Paper No. 910232, 1991.
Parameters."

Clover, Chris L.

and

Distribution

on

James E. Bernard. "The Influence of Lateral Load Transfer SAE Paper No. 930763, Mar. 1993. Directional
Response."

Cole, D.E. Elementary Vehicle Dynamics. Ann Arbor, MI: University

of

Michigan, 1972.

137

Appendix D

Relevant Literature

Crahan, Thomas C. "Modeling Steady-State Suspension Kinematics and Vehicle Dynamics Part I: Theory and of Road Racing Cars SAE Paper No. 942505,
Methodology."
-

Dec. 1994.

Crahan, Thomas C. "Modeling Steady-State Suspension Kinematics and Vehicle Dynamics of Road Racing Cars Part II: SAE Paper No. 942506, Dec. 1994.
Examples."
-

Crolla, D.A.

and

Vehicle

Model."

M.B.A. Abdel-Hady. "Semi-Active Suspension Control for SAE Paper No. 91 1904, Sept. 1991.
of the

Full

Day, Terry D. "An Overview


1995.

HVE Vehicle

Model."

SAE Paper No. 950308, Feb.

Dickison, J.G.

and

Vehicle

Development."

A.J. Yardley. "Development and Application of a Functional Model to SAE Paper No. 930835, Mar. 1993.

Dixon, John C. Tyres, Suspension and Handling, Cambridge, England: Cambridge University Press, 1991. Dreyer, Andreas
with and Heinz-Dieter Heitzer. "Control Strategies for Active Chassis Systems Respect to Road SAE Paper No. 910660, Feb. 1991.
Friction."

Egnaczak, Bernard C. "Supplement to: 'The Development of a Differential for the Improvement of Traction Auto Tech 89, Session 5 Traction Control, Nov. 14, 1989.
Control."

ElBeheiry, ElSayed M.

and

Suspensions Based

on a

Dean C. Karnopp. "Optimization of Active and Passive Full Car SAE Paper No. 951063, Feb. 1995.
Model."

Ellis, John R. Road Vehicle Dynamics, Akron, OH: J.R. Ellis, 1989.
Ellis, John R. Vehicle Dynamics. London: Business Books, 1969. Floyd, R. Scott
and E. Harry Law. "Simulation Aerodynamic Interactions of Race and and
Cars."

Analysis of Suspension and SAE Paper No. 942537, Dec. 1994.

Franklin, Gene F., J. David Powell,

Dynamic Systems. New York:

Addison-

Abbas Emami-Naeini. Feedback Control of Wesley Publishing Company, Inc., 1994.

Garrot, W. Riley, Douglas L. Wilson, and Richard A. Scott. "Digital Simulation for Automobile Simulation, Sept. 1981, p. 83-91.
Maneuvers."

Gillespie, T.D. Fundamentals of Vehicle Dynamics. Warrendale, PA: SAE, 1992.

Gim, Gwanghun
Gim, Gwanghun

and

Cornering

Namcheol Kang. "Requirements of a Tire Model for Practical SAE Paper No. 960179, Feb. 1996. Simulations of
Vehicles."

Parviz E. Nikravesh. "A Three-Dimensional Tire Model for SAE Paper No. 931913, Nov. 1993. State Simulations of
and
Vehicles."

Steady-

Gim, Gwanghun

and Parviz E. Nikravesh. "An Analytical Model of Pneumatic Tyres for International Journal of Vehicle Vehicle Dynamic Simulations. Part 1: Pure 1990. No. Vol. Design, 6, 11,
Slips."

Gim, Gwanghun

Parviz E. Nikravesh. "An Analytical Model Vehicle Dynamic Simulations. Part 2: Comprehensive of Vehicle Design, Vol. 12, No. 1, 1991.
and and

of

Slips."

Pneumatic Tyres for International Journal

Gruening, James

Computer

Simulation."

James E. Bernard. "Verification of Vehicle Parameters for Use in SAE Paper No. 960176, Feb. 1996.

138

Appendix D

Relevant Literature

Gruening, James,
Force
and

Keith A. Williams, Kurt Hoffmeister, and James E. Bernard. 'Tire Moment SAE Paper No. 960182, Feb. 1996.
Processor." Model."

Guntur, R.

S. Sankar. "A Friction Circle Concept for Dugoff s Tyre Friction International Journal of Vehicle Design, Vol. 1, No. 4, 1980.
and

Haney, Paul and Jeff Braun. Inside Racing Technology. Redwood City, CA: TV Motorsports, 1995.

Heydinger, Gary J. "Improved Simulation and Validation of Road Vehicle Handling Ph.D. Dissertation, Ohio State University, Columbus, Ohio, 1990.
Dynamics."

Heydinger, Gary J., W. Riley Garrot,

and

Lag

on

Simulated Transient Vehicle


and

Jeffrey P.

Response."

Chrstos. "The Importance of Tire SAE Paper No. 910235, 1991.

Heydinger, Gary J., Paul A. Grygier,


Applied to Vehicle

Handling

Dynamics."

Seewoo Lee. "Pulse Testing Techniques SAE Paper No. 930828, Mar. 1993.

Holmes, H. and D. Alexander. Formula Car Technology. Santa Ana, CA: Steve Smith Autosports, 1980. Hopkins, Patrick and L. Daniel Metz. "Oversteer/Understeer Characteristics SAE Paper No. 942485, Dec. 1994.
Differential."

of a

Locked

Howard, Geoffrey. Chassis & Suspension Engineering, London, England: Osprey Publishing Limited, 1987. Huang, Feng, J. Roger Chen,
for Vehicle Parameter
and
Estimation."

Lung-Wen Tsai. "The Use of Random Steer Test Data SAE Paper No. 930830, Mar. 1993.
Traction."

Huchtkoetter, Heinrich
in

Heinz Klein. "The Effect of Various Drive Vehicles on Handling and 960717, Feb. 1996.
and Front-Wheel

Limited-Slip Differentials
SAE Paper No.
Control."

Ikushima, Y.

K Sawase. "A Study on the Effects SAE Paper No. 950303, Feb. 1995.
and
Speed."

of

the Active Yaw Moment

Jung, Shinsub

and Dennis A. Guenther. "An Examination of the Maneuverability SAE Paper No. 910241, Feb. 1991. Wheel Steer Vehicle at Low

of an

All

Kaminaga, M., M. Murata,

and Y. Tateishi. "Factoring Nonlinear Kinematics into New SAE Paper No. Suspension Design: A CAE Approach to Vehicle Roll Feb. 1994. 940871,
Dynamics."
Systems."

Karnopp, Dean. "Active Damping in Road Vehicle Suspension Dynamics, Vol. 12, 1983, p. 291-316. Kasprzak, James L.
R. Scott Floyd. "Use SAE Paper No. 942504, Dec. 1994.
and of

Vehicle System
Dampers."

Simulation to Tune Race Car

Katz, Joseph. Race Car Aerodynamics. Cambridge, MA: Robert Bentley, Inc., 1995. Klein, Richard H., Gary L. Teper,
and

Dolly
Ko, Y.
and

Type Combination

Vehicles."

James D. Fait. "Lateral/Directional Stability SAE Paper No. 960184, Feb. 1996.
with an

of

Tow

T. Oh. "Motion Control of the Vehicle SAE Paper No. 940865, Feb. 1994.

Active Suspension

System."

Koibuchi, Ken, Masaki Yamamoto, Yoshiki Fukada,


Control in Limit

and

Cornering by

Active

Brake."

Shoji Inagaki. "Vehicle Stability SAE Paper No. 960487, Feb. 1996.

139

Appendix D

Relevant Literature

Korturn, W.

and W. Schiehlen. "General Purpose Vehicle System Dynamics Software Vehicle System Dynamics, No. 14, 1985, p. Based on Multibody 229-263.
Formalisms."

Kramer,

Kenneth D. and Dale E. Calkins. "Lateral Response SAE Paper No. 942523, Dec. 1994.
Optimization."

of

Formula SAE Race

Car."

La Joie, Joseph C. "Race Car Performance 1994.

SAE Paper No. 942492, Dec.


Technology."

Langer, William. "Vehicle Testing with Flat Surface Roadway No. 960731, Feb. 1996.
Vehicle."

SAE Paper

Lee, Allan Y. "Emulating the Lateral Dynamics of a Range of Vehicles Using SAE Paper No. 950304, Feb. 1995. Wheel-Steering

a Four-

Lee, Allan Y. "Performance of Four-Wheel-Steering Vehicles in Lane Change SAE Paper No. 950316, Feb. 1995. Lee, Seewoo, Jeffrey P. Chrstos,
Inputs."

Maneuvers."

and Dennis A. Guenther. "Modeling of Dynamic Characteristics of Tire Lateral and Longitudinal Force Responses to Dynamic SAE Paper No. 950314, Feb. 1995.
and

Lee, Seewoo, Gary J. Heydinger,

Input Techniques to the Study Dynamics in the Frequency

of

Domain."

Dennis A. Guenther. "The Application of Pulse Tire Lateral Force and Self- Aligning Moment SAE Paper No. 950317, Feb. 1995.
Results."

Lund, Yvonne I.

and James E. Bernard. "The Relationship Between the Complexity of SAE Paper No. 920052, Linear Models and the Utility of the Computer Feb. 1992.

Maalej, Aref Y. "Application of Suspension Derivative Formulation to Ground Vehicle Ph.D. Dissertation, The Ohio State University, Modeling and 1988. Columbus, OH,
Simulation."

Mabrouka, Hani, H. Fred Chen, Aref Y. Maalej,


Lateral Tire Flexibility 910239, Feb. 1991.
on

the

Steering

and Dennis A. Guenther. "Effect of SAE Paper No. Dynamic


Behavior."

Mashadi, Behrooz

and

David A. Crolla. "Vehicle

Around Non-Linear

Operating

Conditions."

Handling Analysis Using Linearization


SAE Paper No. 960482, Feb. 1996.
Program."

McConville, James B.

and John C. Angell. 'The Dynamic Simulation of a Moving Vehicle Subject to Transient Steering Inputs Using the ADAMS Computer ASME Paper No. 84-DET-2, 1984. and

Metz, L. Daniel

D.M. Alter. 'Transient

and

Steady State Performance Characteristics


Model."

of a Two- Wheel-Steer and

Four-Wheel-Steer Vehicle

SAE Paper No.

911926, 1991.
Metz, L. Daniel, Michael Dover, John Fisher, Victoria McCleary,
Configurations."

Errol Shavers. Vehicle "Comparison of Linear Roll Dynamics Properties for Various SAE Paper No. 920053, 1992.
and and

Metz, L. Daniel, Troy S. Torbeck, Kevin H. Forbes,


Maneuver Capability Without 942469, Dec. 1994.
and

L.

Gregory Metz.
Flat
Tire."

In the Presence

of a

"Evasive SAE Paper No.

140

Appendix D

Relevant Literature

Metz, L. Daniel. "Dynamics 930765, Mar. 1993.

of Four-Wheel

Steer Off-Highway

Vehicles."

SAE Paper No.

Milliken, William F.
SAE, 1995.

and

Doug L.

Milliken. Race Car Vehicle Dynamics. Warrendale, PA:


Method."

Milliken, William F.
1983,
p.

and

R.S. Rice. "Moment

IMechE Paper No. CI 13/83,

31-60.
and

Milliken, William F., Peter G. Wright,

Comprehensive Tool for Race Car 1994.

Development."

Douglas L. Milliken. "Moment Method A SAE Paper No. 942538, Dec.


-

Mimuro, Tetsushi, Masayoshi Ohsaki, Hiromichi Yasunaga,


Parameter Evaluation Method 901734, 1990. 1969.
of Lateral

and

Transient

Response."

Kohji Satoh. "Four SAE Paper No.

Mola, Simone. Fundamentals of Vehicle Dynamics, Detroit, MI: General Motors Institute, Moline, D., S. Floyd, S. Vaduri,
Active
Suspensions."

and E.H. Law. "Simulation and Evaluation SAE Paper No. 940864, Feb. 1994.

of Semi-

Mori, Yoshinori, Hironobu Matsushita, Takashi Yonekawa, Yoshihisa Nagahara,


Hiroshi Shimomura. "A Simulation System for Vehicle Dynamics Paper No. 910240, Feb. 1991.

and

Control."

SAE

Nalecz, Andrzej G. "Analysis


at

High

Speed."

of the Dynamic Response of a Four Wheel Steering Vehicles International Journal of Vehicle Design, Vol. 9, No. 2, 1988.

Nalecz, Andrzej G. "Development and Validation of Light Vehicle Dynamics Simulation SAE Paper No. 920056, Feb. 1992.
(LVDS)."

Nalecz, Andrzej G.
Wheel

Alan C. Bindemann. "Investigation into the Stability of Four International Journal of Vehicle Design, Vol. 9, No. 2, Steering 159-178. p. 1988,
and
Vehicles."

Naude, Alwyn F.

and Jasper L. Steyn. "Objective Evaluation Characteristics of a Vehicle in a Double Lane Change 930826, Mar. 1993.

of the

Manoeuvre."

Simulated Handling SAE Paper No.

Negrut, D.

and J.S. Freeman. "Dynamic Tire Modelling for Application with Vehicle SAE Paper No. 940223, Feb. 1994. Simulations Incorporating
Terrain."

Neto, Mauro Speranza, Fernando Riberio da Silva, and Jose Francisco Martinex. "Design Methodology in Vehicle Dynamics, Using the Procedures of Modeling, Simulation, SAE Paper No. 921480, 1992. and Analysis of System
Dynamics."

New Developments in Vehicle Dynamics, Simulation, Publication No. SP-1074, Feb. 1995.

and

Suspension Systems. SAE

Nikravesh, Parviz E.

Jong-Nyun Lee. "Optimal Four-Wheel Steering Strategy Using SAE Paper No. 931915, Nov. 1993. Nonlinear Analytical Vehicle
and
Models."
Handling."

Olley, Maurice. "Suspension


1937.

and

Detroit, MI: Chevrolet Engineering Center,

Olley, Maurice. "Notes


1961.

Suspensions."

on

Detroit, MI: Chevrolet Engineering Center,

141

Appendix D

Relevant Literature

Olley, Maurice.

"Suspensions Notes

n."

Detroit, MI: Chevrolet Engineering Center,

1966.

Palmeri, Paolo S., Alberto Moschetti,

and

Thema Full Active Suspension

System."

Luigi Gortan. "H-Infmity Control for Lancia SAE Paper No. 950583, Feb. 1995.

Petersen, Michael R.
with

and John M. Starkey. "Nonlinear Vehicle Performance Simulation Test Correlation and Sensitivity Analysis." SAE Paper No. 960521 Feb 1996.

Post, J.W.

and E.H. Law. "Modeling, Characterization and Simulation of Automobile Power Steering Systems for the Prediction of On-Center Handling." SAE Paper No. 960178, Feb. 1996.
Data."

Radt, Hugo S. "An Efficient Method for Treating Race Tire Force-Moment Paper No. 942536, Dec. 1994. Radt, Hugo S.
and

SAE

D.A. Gleniming. "Normalization of Tire Force and Moment Science and Technology, Vol. 21, No. 2, Apr.-June 1993, p. 91-119.
and
Derivatives."

Data."

Tire

Radt, Hugo S.

Donald J. Van Dis. "Vehicle Handling Responses SAE Paper No. 960483, Feb. 1996.
Cars."

Using Stability

Reichelt, Werner. "Correlation Analysis of Open/Closed Loop Data for Objective Assessment of Handling Characteristics of SAE Paper No. 910238, Feb.
1991.

Reimpell, Jornsen and Helmut Stoll. The Automotive Chassis: Engineering Principles. Warrendale, PA: SAE, 1996. Rice, R.S.
and

Utilizing the Moment


Sayers, Micheal W.
Dynamics
and
Models."

William F. Milliken. "Static Stability and Control of the Automobile SAE Paper No. 800847, June 1980.
Method."

C. Mink. "A Simulation Graphical User Interface for Vehicle SAE Paper No. 950169, Feb. 1995.
Trucks."

Sayers, Michael W.

and Stephen M. Riley. "Modeling Assumptions for Realistic Simulations of the Yaw and Roll Behavior of SAE Multibody Heavy Paper No. 960173, Feb. 1996. and

Schuring, Dieterich J., Wolfgang Pelz,


Tire

Cornering

and

Braking

Forces."

Marion G. Pottinger. "A Model for Combined SAE Paper No. 960180, Feb. 1996.
Concept."

Schuring, Dieterich J., Wolfgang Pelz, Implementation of the 'Magic 1993.


of

and

Formula'

Marion G. Pottinger. "An Automated SAE Paper No. 931909, Nov.

Segal, Leonard. "Theoretical Prediction and Experimental Substantiation of the Response


the Automobile to
the

Proceedings of the Automobile Division of Institution of Mechanical Engineers, No. 7, 1956-1957, p. 310-330.

Steering

Control."

Sharp, R.S.

D.A. Crolla. "Road Vehicle Suspension System Design Vehicle System Dynamics, Vol. 16, 1987, p. 167-192.
and
and
Moments."

Review."

Shimada, K.

Stabilizing Yaw
Smith, C. Prepare
to

Y. Shibahata. "Comparison of Three Active Chassis Control Methods for SAE Paper No. 940870, Feb. 1994.

Smith, C. Engineer to Win. Osceola, WI: Motorbooks International, 1984.


Win.

Fallbrook, CA: Aero Publishers, Inc.,

1975.

142

Appendix D

Relevant Literature

Smith, C. Tune

to Win.

Fallbrook, CA: Aero Publishers, Inc., 1978.


Automobiles."

Smith, Norman. 'Transient Control Response of Dynamics, Vol. 6, No. 2-3, Sept. 1977, p. 63-67.
Sohn, H.S., S.C. Lee, M.W. Suh,
Characteristics 1994.
on

Vehicle System

and Y.M. Song. "The Influences of Chassis Geometric Performances." Vehicle Dynamic SAE Paper No. 940872, Feb.

Song, Jun-gyu and Yong-San Yoon. "Design


Control Algorithm
of

of

Four-Wheel

Steer."

Two-Wheel Steer Vehicle Using Optimal SAE Paper No. 931914, 1993.

Staniforth, A. Competition Car Suspension. Newbury Park, CA: Haynes Publications Inc., 1991.
Sultan, Mohammad O., Gary J. Heydinger, Nicholas J. Durisek,
"A Study of Vehicle Class Segregation Paper No. 950307, Feb. 1995.
and

Using Linear Handling

Dennis A. Guenther. SAE


Models."

Taborek, Jaroslav J. Mechanics of Vehicles. Cleveland, OH: Penton, 1957. Thomas, D.W. "Vehicle Modeling
Oct. 1987.
and

Service Loads

Analysis."

SAE Paper No. 871940,

Trom, J.D., J.L. Lopex,


a

M.J. Vanderploeg. "Modeling of a Mid-Size Passenger Car Transactions of the ASME, Journal of Dynamics Using Multibody Automation in Design, Vol. 109, Dec. 1987. and Mechanisms, Transmissions,
and
Program."

Trom, J.D., M.J. Vanderploeg,


Vehicle Optimization 110.

and

Problems."

James E. Bernard. "Application of Inverse Models to Vehicle System Dynamics, Vol. 19, 1990, p. 97-

Turpin, D.R.

and

Simulation."

D.F. Evans. "High-Fidelity Road/Tire Interaction Models for Real Time SAE Paper No. 950170, Feb. 1995.
and Mechanics.

Van Valkenburgh, P. Race Car Engineering Valkenburgh, 1986.


van

Seal Beach, CA: Paul Van


and Harmut SAE Paper No.

Zanten, Anton Th., Rainer Erhardt, Albert Lutz, Wilfried Neuwald,


Bartels. "Simulation for the Development 960486, Feb. 1996.
of

the

Bosch-

VDC."

Vanderlploeg, M.J., J.D. Trom,

James E. Bernard. "Evaluation Path Follow Performance Using a Linear Inverse Vehicle 880644, 1988.
and

of

Model."

Four-Wheel Steer SAE Paper No.

Vedamuthu, S.

E.H. Law. "An Investigation Determining Automobile Handling


and

of

the Pulse Steer Method for

Qualities."

SAE Paper No.

930829, Mar. 1993.

Vehicle Dynamics and Electronic Controlled Suspensions. SAE Publication No. Feb. 1991. Vehicle Dynamics
and

SP-861,

Rollover Propensity Research. SAE Publication No.

SP-909, Feb.

1992.
Vehicle Dynamics
and

Simulation. SAE Publication No. SP-950, Mar. 1993.


SAE Publication No. SP-1031, Feb. 1994.

Vehicle Suspension System Advancements.

143

Appendix D

Relevant Literature

Whatmough,
Low

Speeds."

K.J. "Real-Time Wheel Brake and Tire Lateral Force Models Refined for SAE Paper No. 940178, Feb. 1994.
and

Whitcomb, David W.
of

William F. Milliken. "Design Implications


Control."
,

of a

General

Theory

Automobile Stability and Proceedings of the Automobile Division of the Institution of Mechanical Engineers Aug. 1956, p. 83-107. R.S.

Wilson, D.A.,

Sharp,

and

S.A. Hassan. "Application


Suspension."

Theory to the Design of Automobile


15, 1986,
p.

Linear Optimal Control Vehicle System Dynamics, Vol.


of

105-118.

Wong, Jo Yung. Theory of Ground Vehicles. New York: John Wiley & Sons, Inc., 1993. Wright, Peter. "Out at the
Four Wheel
Edge!"

Racecar, Vol. 5, No. 3, 1995,


of

p.

15-18.

Xia, Xunmao. "A Nonlinear Analysis

Closed Loop Driver/Vehicle Performance with Ph.D. Dissertation, Department of Mechanical Steering Clemson Engineering, University, Clemson, SC, Dec. 1990.
Control."

Xia, Xunmao Xia, Xunmao

and E.H. Law. "Nonlinear Analysis Performance with Four Wheel Steering
and

of

Control."

Closed Loop Driver/Automobile SAE Paper No. 920055, 1992.


on

Handling

Performance."

J.N. Willis. 'The Effects of Tire Cornering Stiffness SAE Paper No. 950313, Feb. 1995.

Vehicle Linear
Stability."

Yamamoto, Masaki. "Active Control Strategy for Improved Handling Paper No. 911902, Sept. 1991. Yasui, Yoshiyuki, Kenji Tozu, Noriaki Hattori,
of
Control."

and

SAE

and Masakazu Sugisawa. "Improvement Vehicle Directional Stability for Transient Steering Maneuvers Using Active SAE Paper No. 960485, Feb. 1996. Brake

144