Vehicle Dynamics

Rochester Institute of Technology
RIT Scholar Works

Theses Thesis/Dissertation Collections
8-1-1996
Modeling of road vehicle lateral dynamics

Joseph Kiefer
Follow this and additional works at: http://scholarworks.rit.edu/theses
Recommended Citation
Kiefer, Joseph, "Modeling of road vehicle lateral dynamics" (1996). Thesis. Rochester Institute of Technology. Accessed from
This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusion
in Theses by an authorized administrator of RIT Scholar Works. For more information, please contact [email protected].
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
Professor - - - - - - - - -
Dr. Alan Nye
Professor _
Dr. Michael Hennessey
Professor _
Dr. Charles Haines
Thesis Advisor
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 of a road vehicle is studied through the development of a
mathematical model. 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 model tire lateral forces are represented by an approximate linear
relationship. Transfer functions are written for the vehicle system for various control and
disturbance inputs. Steady-state and transient response characteristics and frequency

response are studied, and numerical simulation of the model is performed.
In the non-linear model a 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.
m
Table of Contents
Disclosure Statement ii
Abstract iii
Table of Contents iv
List of Tables vi
List of Figures vii
List of Symbols x
Chapter 1 Introduction 1
Chapter 2 Literature Review 5
Chapter 3 Tire Behavior 10
3.1 Introduction 10
3.2 Lateral Force Mechanics 11
3.3 Linear Tire Model 14
3.4 Non-Linear Tire Model 14
Chapter 4 Two Degree-of-Freedom Vehicle Model 24
4. 1 Introduction 24
4.2 Description of Model 25

4.2. 1 Assumptions 26
4.2.2 Vehicle Parameters 27
4.2.3 Free-Body Diagram 28
4.3 Derivation of Equations of Motion 29

4.4 Derivation of Tire Slip Angles 31
4.5 Linear Model 33

4.5. 1 Additional Assumptions 34
4.5.2 Vehicle Sideslip Angle 35
4.5.3 Tire Slip Angles 35
4.5.4 External Forces and Moments 36
4.5.5 Equations of Motion 37
4.5.6 Transfer Functions 39
4.5.7 Vehicle Sideslip Angle Gain 41
4.5.8 Yaw Velocity Gain 42
4.5.9 Front Tire Slip Angle Gain 42
4.5.10 Rear Tire Slip Angle Gain 43
4.5.11 Path Curvature Gain 43
iv
Table of Contents
4.5. 12 Lateral Acceleration Steady-State Step Response Gain 44

4.5.13 Steady-State Steer Angle 45
4.5.14 Understeer Gradient 46
4.5.15 Stability Factor 48
4.5. 16 Neutral Steer Point 48
4.5.17 Static Margin 49
4.5.18 Tangent Speed 50
4.5.19 Critical Speed 51
4.5.20 Characteristic Speed 51
4.5.21 Characteristic Equation 52
4.5.22 Undamped Natural Frequency 53
4.5.23 Damping Ratio 53
4.5.24 System Poles 55
4.5.25 System Zeros 55
4.5.26 Frequency Response 58
4.5.27 Simulation 65
4.6 Non-Linear Model 89
4.6. 1 Model Equations 89
4.6.2 Simulation 90
Chapter 5 Conclusion 100
References 102
Appendix A Tire Model MATLAB Programs 105

A. 1 MagicFit.m 105
A.2 MagicError.m 106
A.3 NLTire.m 107
Appendix B Two DOF Model Mathematica Session 108
Appendix C Two DOF Model MATLAB Programs 123

C.l DOF2Control.m 123
C.2 DOF2Param.m 124
C.3 DOF2DependParam.m 126
C.4 SteerAngle.m 127
C5 DOF2LFreq.m 129
C.6 DOF2LSim.m 130
C7DOF2LDE.m 132
C.8 DOF2NLSim.m 133
C.9 DOF2NLDE.m 135
Appendix D Relevant Literature 136

List of Tables
Table 3.1: Non-Linear Tire Model Parameters 17
Table 3.2: Experimental Tire Data 21
Table 4.1: Vehicle Parameters 28
Table 4.2: Linear Tire Model Parameters 36
Table 4.3: Steady-State Response Gains 45
Table 4.4: System Zeros 56
vi
List of Figures
Figure 3. 1 : Tire Slip Angle 11
Figure 3.2: Tire Lateral Force Versus Slip Angle 12
Figure 3.3: Experimental Tire Data 21
Figure 3.4: Tire Cornering Coefficient 22
Figure 3.5: Tire Lateral Friction Coefficient 22
Figure 3.6: Tire Normalized Lateral Force 23
Figure 3. 7: Reconstructed Tire Lateral Force 23
Figure 4.1: Vehicle Model 26
Figure 4.2: Free-Body Diagram 29
Figure 4.3: Kinematic Diagram 32
Figure 4.4: Gravitational Side Force 37
Figure 4.5: Natural Frequency vs. Vehicle Velocity 54
Figure 4.6: Poles and Zeros 57
Figure 4. 7: Sideslip Angle / Steer Angle Frequency Response, V = 100 km/hr 61
Figure 4.8: Sideslip Angle /Aero Side Force Frequency Response, V = 100 km/hr 61
Figure 4.9: Sideslip Angle /Road Side Slope Frequency Response, V = 100 km/hr 62
Figure 4.10: Yaw Velocity / Steer Angle Frequency Response, V'= lOOkm/hr 62
Figure 4.11: Yaw Velocity /Aero Side Force Frequency Response, V=100 km/hr 63
Figure 4. 12: Yaw Velocity /Road Side Slope Frequency Response, V =

100 km/hr 63
Figure 4. 13: Sideslip Angle / Steer Angle Frequency Response, V = 30 km/hr 64
Figure 4. 14: Sideslip Angle / Steer Angle Frequency Response, V = 49.84 km/hr 64
Figure 4.15: Simulation Steer Angle Inputs 67
Figure 4.16: Linear Step Steer Lateral Velocity Response 71
vn
List of Figures
Figure 4.17: Linear Step Steer Yaw Velocity Response 71
Figure 4.18: Linear Step Steer Sideslip Angle Response 72
Figure 4. 19: Linear Step Steer Front Tire Slip Angle Response 72
Figure 4.20: Linear Step Steer Rear Tire Slip Angle Response 73
Figure 4.21 : Linear Step Steer Lateral Acceleration Response 73
Figure 4.22: Linear Ramp Step Steer Lateral Velocity Response 74
Figure 4.23: Linear Ramp Step Steer Yaw Velocity Response 74
Figure 4.24: Linear Ramp Step Steer Sideslip Angle Response 75
Figure 4.25: Linear Ramp Step Steer Front Tire Slip Angle Response 75
Figure 4.26: Linear Ramp Step Steer Rear Tire Slip Angle Response 76
Figure 4.27: Linear Ramp Step Steer Lateral Acceleration Response 76
Figure 4.28: Linear Ramp Square Steer Lateral Velocity Response 77
Figure 4.29: Linear Ramp Square Steer Yaw Velocity Response 77
Figure 4.30: Linear Ramp Square Steer Sideslip Angle Response 78
Figure 4.31: Linear Ramp Square Steer Front Tire Slip Angle Response 78
Figure 4.32: Linear Ramp Square Steer Rear Tire Slip Angle Response 79
Figure 4.33: Linear Ramp Square Steer Lateral Acceleration Response 79
Figure 4.34: Linear I Hz Sine Steer Lateral Velocity Response 80
Figure 4.35: Linear 1 Hz Sine Steer Yaw Velocity Response 80
Figure 4. 36: Linear I Hz Sine Steer Sideslip Angle Response 81
Figure 4.37: Linear I Hz Sine Steer Front Tire Slip Angle Response 81
Figure 4.38: Linear I Hz Sine Steer Rear Tire Slip Angle Response 82
Figure 4.39: Linear 1 Hz Sine Steer Lateral Acceleration Response 82
Figure 4.40: Linear Step Aero Side Force Lateral Velocity Response 83
Figure 4.41: Linear Step Aero Side Force Yaw Velocity Response 83
vui
List of Figures
Figure 4.42: Linear Step Aero Side Force Sideslip Angle Response 84
Figure 4.43: Linear Step Aero Side Force Front Tire Slip Angle Response 84
Figure 4.44: Linear Step Aero Side Force Rear Tire Slip Angle Response 85
Figure 4.45: Linear Step Aero Side Force Lateral Acceleration Response 85
Figure 4.46: Linear Step Road Side Slope Lateral Velocity Response 86
Figure 4.47: Linear Step Road Side Slope Yaw Velocity Response 86
Figure 4.48: Linear Step Road Side Slope Sideslip Angle Response 87
Figure 4.49: Linear Step Road Side Slope Front Tire Slip Angle Response 87
Figure 4.50: Linear Step Road Side Slope Rear Tire Slip Angle Response 88
Figure 4.51: Linear Step Road Side Slope Lateral Acceleration Response 88
Figure 4.52: Non-Linear Step Steer Lateral Velocity Response 94
Figure 4.53: Non-Linear Step Steer Yaw Velocity Response 94
Figure 4.54: Non-Linear Step Steer Sideslip Angle Response 95
Figure 4.55: Non-Linear Step Steer Front Tire Slip Angle Response 95
Figure 4.56: Non-Linear Step Steer Rear Tire Slip Angle Response 96
Figure 4.57: Non-Linear Step Steer Lateral Acceleration Response 96
Figure 4.58: Non-Linear Ramp Square Steer Lateral Velocity Response 97
Figure 4.59: Non-Linear Ramp Square Steer Yaw Velocity Response 97
Figure 4.60: Non-Linear Ramp Square Steer Sideslip Angle Response 98
Figure 4.61: Non-Linear Ramp Square Steer Front Tire Slip Angle Response 98
Figure 4.62: Non-Linear Ramp Square Steer Rear Tire Slip Angle Response 99
Figure 4.63: Non-Linear Ramp Square Steer Lateral Acceleration Response 99
IX
Symbolst
List of
l/R Path curvature (1/m)
a Distance from mass center to front axle (m)
a0 Acceleration of the origin of the vehicle-fixed coordinate system (m/s2)
ay Acceleration of vehicle mass center in y-direction (m/s )

"g"
Ay Acceleration of vehicle mass center in y-direction in units of (g)
b Distance from mass center to rear axle (m)
Bj Magic Formula curve fit parameter
B3 Tire cornering coefficient intercept (N/deg/N)
B5 Tire lateral friction coefficient intercept
c Distance from front axle to aerodynamic side force (m)
C} Magic Formula curve fit parameter
C3 Tire cornering coefficient slope (N/deg/N2)
C5 Tire lateral friction coefficient slope
Ca Tire cornering stiffness (N/rad)
Front tire cornering

Cf stiffness two tires (N/rad)
-
Cr Rear tire cornering stiffness -
two tires (N/rad)
Cc Tire cornering coefficient (N/deg)
d Distance from front axle to neutral steer point (m)
D} Magic Formula curve fit parameter
Ej Magic Formula curve fit parameter
/ Front axle weight fraction
F External force (N)
f italicized
By convention, scalar variables are and vectors are boldfaced.
List of Symbols
F^ Fictitious lateral force for finding neutral steer point (N)
Fy Tire lateral force (N)
Fy Normalized tire lateral force
Fya Aerodynamic side force disturbance (N)
F^ Front tire lateral force -
two tires (N)
Fyg Gravitational side force disturbance (N)
F Rear tire lateral force -

two tires (N)
Fz Tire vertical load (N)
g Acceleration due to gravity (m/s2)
G Linear momentum (kg-m/s)

H Angular momentum (about mass center) (kg-m2/s)
i Unit vector in x-direction of vehicle-fixed coordinate system
Izz Total vehicle yaw mass moment of inertia (kg-m2)
j Unit vector in y-direction of vehicle-fixed coordinate system
k Unit vector in z-direction of vehicle-fixed coordinate system
K Stability factor (s2/m2)
Kus Understeer gradient (rad)
L Wheelbase (m)
m Total vehicle mass (kg)
M External moment (about mass center) (N-m)
Nr Yaw damping derivative (N-m-s/rad)
/Vp Directional stability derivative (N-m/rad)
Ns Control moment derivative (N-m/rad)
r Yaw velocity (rad/s)
R Path radius (m)
XI
List of Symbols
Rf Position vector from vehicle mass center to front tire (m)
Rr Position vector from vehicle mass center to rear tire (m)

s Laplace-domain variable
SM Static margin
u Forward velocity (m/s)
v Lateral velocity (m/s)
V Magnitude of vehicle velocity (m/s)
Vchar Characteristic speed (m/s)
VcHt Critical speed (m/s)
Vf Velocity of front tire (m/s)
V0 Velocity of vehicle-fixed coordinate system (m/s)
Vr Velocity of rear tire (m/s)
V[an Tangent speed (m/s)
Vx Component of
velocity in x-direction (m/s)
Vy Component of velocity in y-direction (m/s)
Yr Lateral force/yaw coupling derivative (N-s/rad)
Tp Damping-in-sideslip derivative (N/rad)

Y6 Control force derivative (N/rad)
a Tire slip angle (rad)
a Tire normalized slip angle
af Front tire slip angle (rad)
ar Rear tire slip angle (rad)
(3 Vehicle sideslip angle (rad)
5 Front steer angle input (rad)
Âcker Ackerman steer angle (rad)
xn
List of Symbols
8/ Front steer angle (rad)
8r Rear steer angle (rad)
8M Steady-state steer angle for given path radius (rad)
yr Magic Formula intermediate variable
0 Road side slope (rad)
9 Magic Formula intermediate variable
Damping ratio
liy Tire lateral friction coefficient
oo Undamped natural frequency (rad/s)

Q. Angular velocity of vehicle-fixed coordinate system (rad/s)
Qz Angular velocity of vehicle-fixed coordinate system about z-axis (rad/s)
xin
1 Introduction
The first practical automobiles were built in 1886 by Karl Benz and Gottlieb
Daimler.1
The top speeds of these vehicles were only about fifteen miles per hour. With
much of the automotive industry's early engineering effort devoted to developing faster
vehicles, production car top speeds reached forty-five miles per hour by 1900 and eighty
miles per hour by 1915. This year, Craig Breedlove will attempt to be the first man to travel
faster than the speed of sound in a ground vehicle.
Since the top speeds of the first automobiles were relatively low, there was initially
little concern with the dynamic behavior of the vehicles. However, as cars quickly became
capable of achieving higher speeds, vehicle dynamics became an important concern for
automotive engineers. Of primary importance from a safety standpoint was the behavior of
vehicles in maneuvers such as turning and braking as top speeds increased. Also, since
early roads were of very poor quality by today's standards, isolation of the driver and
passengers from road disturbances became increasingly important.

The field of vehicle dynamics encompasses three basic modes of vehicle
performance. Vertical dynamics, or ride dynamics, basically refers to the vertical response
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
turning behavior. Achieving acceptable performance in each of these modes is necessary in
order for a vehicle to meet the requirements of the consumer, and the government, with
regards to comfort, controllability, and safety. 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 production. With such a long time from
initial concept to production, vehicle designs can 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 cycle. This allows
manufacturers to respond more quickly to changes in market demand. In addition, reducing
the length of the design cycle reduces the cost of developing a new vehicle.
One way that manufacturers can reduce design cycle length is by achieving the best
possible design before any prototypes are built. The development of 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
facilitated this effort. As computer speeds continually increase and engineering software
becomes more powerful, better vehicle designs can be obtained before prototypes are built,
resulting in fewer prototypes and reduced development time and cost.
Mathematical modeling of vehicle dynamics helps engineers reduce the time it takes
to achieve a design which will meet performance requirements for the consumer and for
government regulations. A proposed design can be studied to determine if it can meet goals
before any prototypes are built. The effects of design changes can be evaluated without
building costly prototypes. Development engineers can use mathematical models to assist
with the tuning of prototypes by identifying the changes which should be made to produce
desired ride and handling characteristics.
Computer simulation offers a controlled, repeatable environment where the effects
of individual parameters can be isolated without the influence of the variations in the
environment. Simulation can remove the performance of the test driver from the picture to
isolate the performance of the vehicle. In addition, simulation can be used to study
maneuvers that could result in costly damage to the vehicle or danger to the test driver. An
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
related technologies such anti-lock 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) or a combination
of modes. Depending upon its purpose, a model must include representations of
appropriate systems of the vehicle. Effects of the suspension system, steering system,
powertrain system, braking system, or 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 simple two degree-of-freedom vehicle model popularly known in the
literature as 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
such as tire properties, inertia properties, mass center location, wheelbase, and forward
speed.
Chapter 2 is a review of vehicle dynamics literature relating to lateral dynamics.
Since the majority of the forces acting on a vehicle are developed by the tires, an overview
of tire lateral force mechanics is provided in Chapter 3 along with a description of the two
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 presents the main focus of this research, the development and application
of the 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 and
non-linear. In the linear model transfer functions are written and used 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,

simulation of the non-linear model is performed and the results are compared with those of
the linear model. Conclusions are drawn regarding the range of applicability of the linear
model.
2 Literature Review
As mentioned in Chapter 1 as the top speeds of automobiles increased rapidly in the
early part of this century vehicle dynamics became an important consideration to engineers.
A large body of vehicle dynamics literature cunently exists covering all aspects of vertical,
longitudinal, and lateral performance. Fundamental to understanding lateral vehicle
dynamics is knowledge of the mechanism of tire lateral force generation, and much has
been written on this topic.
One of first papers concerning road vehicle lateral dynamics was written in 1908 by
Lanchester.2
William In this work Lanchester discussed the steering behavior of
automobiles. However, complete understanding of turning behavior was hampered in the
early years by a lack of understanding of tire mechanics. In 1925 George Broulhiet
Mechanism"
published a paper titled "The Suspension and the Automobile Steering which
described tire lateral force generation 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 were built which could measure the forces generated by a
tire under various conditions. These advancements paved the way for others to develop
detailed explanations and models of turning behavior.
One of the early pioneers in vehicle dynamics research was Maurice Olley. He was
responsible for the introduction of the independent front suspension in the United States for
Cadillac and described the operation of the system in the 1934 SAE paper "Independent
Whererfores."
Wheel Suspension: Its Whys and A report written in 1937 titled
Handling"
"Suspension and reviewed the research in lateral dynamics during the preceding
today.3
years and covered much of what is understood Olley was active in vehicle
dynamics from the early 1930's through his retirement in 1955. During the period of the
Chapter 2 Literature Review
"Olley' Notes"4'5
early 1960's he published a series of documents know today as s which
summarized his extensive knowledge of suspension systems and handling.
One of the most significant works concerning lateral vehicle dynamics was written
Segal.6
in 1956 by Leonard Segal, who worked at Cornell Aeronautical Laboratory,
applied to the road vehicle many of the analytical techniques which had been developed for
aircraft dynamics. In this work, entitled 'Theoretical Prediction and Experimental
Control,"
Substantiation of the Response of the Automobile to Steering Segal developed
equations of motion for a linear three degree-of-freedom (yaw, lateral, and roll) model of
vehicle turning behavior. Since digital computers were not available for his research, it was
necessary to have a 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
equations of motion and described the concepts of stability factor, neutral steer point, and
static margin. Segal backed his modeling efforts with experimental testing of a vehicle and
concluded that a linear model was sufficiently accurate for lateral motions of a reasonable
magnitude.
A second paper "Design Implications of a General Theory of Automobile Stability

Control"
and written by David Whitcomb and William Milliken, also of Cornell
Aeronautical Laboratory, was presented at the same time as Segal's paper as part of a five
Performance."7
paper series "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 body of established techniques for the analysis of second order
dynamic systems.
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
1993),8
Suspension and Handling (Bastow and Howard, Elementary Vehicle Dynamics
1972),9 1),10
(Cole, Tyres, Suspension and Handling (Dixon, 199 Vehicle Dynamics (Ellis,
1969),11 1989),12
Road Vehicle Dynamics (Ellis, Fundamentals of Vehicle Dynamics
1992),1 1995),13
(Gillespie, Race Car Vehicle Dynamics (Milliken and Milliken,
1969),14
Fundamentals of Vehicle Dynamics (Mola, The Automotive Chassis: Engineering
1996),15 1957),16
Principles (Reimpell and Stall, Mechanics of Vehicles (Taborek, and
n
Theory of Ground Vehicles (Wong, 1993). Most of these books utilize a two degree-of-
freedom vehicle model when explaining turning behavior.
Another significant contribution to the literature was made in 1976 by Bundorf and
Leffert.18
In this work the cornering compliance concept is described. With this technique
the contributions of various vehicle systems and characteristics to understeer are determined
and added 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
camber stiffnesses, tire aligning torque, and lateral load transfer on understeer without
developing the detailed vehicle models that would be necessary to simulate these effects
directly. Since the computing hardware and software needed to analyze sufficiently detailed
models was not readily available at the time, this concept was a significant advancement.
Since the work of Segal in 1956 many vehicle models have been developed which
expand on his model. The dynamics of other systems such as the steering system have
been integrated into the vehicle models. Lateral dynamics models have been expanded to
include longitudinal and vertical degrees of freedom. Non-linearities, particularly in tire
force generation, have been included in the models. Some examples in the literature can be
Allen,19 Heydinger,20 Xunmao.21

found in works by 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 and connected using joints. By modeling each suspension component
individually a very accurate kinematic representation of a complete vehicle can be obtained.

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
vehicle dynamics simulations becomes more practical. The biggest disadvantage with the
use of these codes is the large amount of information that is required to construct the
models. The dimensions, mass properties, and in some cases stiffnesses of each relevant
component must be known to build an accurate model. Commercial multibody codes used
for vehicle dynamics simulation include ADAMS, DADS, and Mechanica Motion.
Paramount to the development of successful vehicle dynamics models 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
geometry and material properties of the tire. Most of the major advancements in tire models
have occurred within the last fifteen years as digital computers have become readily
available. A comprehensive analysis of tire mechanics was performed under a government
1981.24
contract by Clarke in In 1990 a detailed theoretical tire model was developed by
Nikravesh.25
Gim and However, most of the popular tire models in existence
today are
based primarily upon empirical data. These models involve curve fitting of experimentally
measured tire data. One of the most popular empirical tire models known as the "Magic
Formula" 1987.26
was published by Bakker, Nyborg, and Pacejka in Other useful tire
Radt27 Allen.28
models include those by and The development of accurate tire models has
been critical to the success of vehicle dynamics modeling.
There exists a large body a literature regarding vehicle dynamics. The last forty
years in particular have seen many significant developments on the topic. Models of
vehicles and tires have been developed to the point where very accurate simulations of
lateral dynamic response can be performed. The advent of the digital computer has greatly
enhanced the ability of engineers to develop and utilize these models for practical gains. A
list of relevant sources from the vehicle dynamics literature reviewed during this research is
provided in Appendix D.
3 Tire Behavior
3.1 Introduction
With the exception of gravitational and aerodynamic forces, all of the forces acting
on a road vehicle are applied to the vehicle through its tires. In supporting the vehicle the
ground applies vertical forces to the tires. When the vehicle changes speed or 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
dynamics of road vehicles it is necessary to have a suitable representation of tire behavior.
Two tire models are used in this thesis: a simple linear model and a more accurate, more
widely applicable non-linear model.
The requirements of a tire model vary depending upon the aspects of vehicle
performance which are being modeled and the accuracy required. In general, there are 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 longitudinal, lateral, and
vertical motions of the vehicle are being studied, all six of these components must be
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
horizontal plane. Thus only forces in the lateral direction and moments about the vertical
axis of the vehicle need to be considered. The moment acting on the tire itself about its
vertical axis is called the tire aligning moment. The effect of the
aligning moments of the
tires on the overall dynamics of the vehicle is generally small compared to the effect of the
lateral forces of the tires. In the vehicle model which is presented here the aligning
moments of the tires are neglected. Thus the only aspect of tire behavior which is modeled
is lateral force generation.
10
Chapter 3 Tire Behavior
3.2 Lateral Force Mechanics
The mechanics of the lateral force generation of a tire is a complex process. A
complete discussion of this process is beyond the scope of this thesis. Many thorough
literature.1'1317
discussions of the mechanics of force generation exist in the The lateral
force
Fy generated by a pneumatic tire depends upon many variables including road surface
conditions, tire carcass construction, tread design, rubber compound, size, pressure,
temperature, speed, vertical load, longitudinal slip, inclination angle, and slip angle. For a
given tire on load

unchanging, dry road surface conditions, vertical 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 is represented by the symbol a and is defined by S AE as "the
X'
angle between the axis and the direction of travel of the center of tire
contact."29
This
X'
Figure 3. 1: Tire Slip Angle
11
definition The this system is at the center of

system.29
references the S AE tire axis origin of
X'
the tire contact patch. The axis is the intersection of the plane of the wheel and the plane
Z'
of the ground and is positive in the forward direction. The axis is perpendicular to the
Y'
plane of the road and is positive in the downward direction. The axis is in the plane of
the road and oriented to form a right-hand Cartesian coordinate system. The tire slip angle,
lateral force, and tire axis system are shown in Figure 3.1. A positive slip angle and lateral
force are shown. Simply stated, the slip angle is the angle between the direction the wheel
is pointing and the direction it is traveling at a given instant in time.
The lateral force produced by a tire is a non-linear function of, among other
variables, vertical load and slip angle. A typical lateral force versus slip angle curve for a
single vertical 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
design, and tire pressure. There is little sliding occurring between the tire and ground
within the contact patch. Lateral force is developed as a result of deformation of the tire.
\.<h.j...\....^r.
<D
v>r
o
2
0)
CO
/ i i
4 6 8 10 12 14 16
Slip Anlge (deg)
Figure 3.2: Tire Lateral Force Versus Slip Angle
12
The initial slope of lateral force versus slip angle curve is the cornering stiffness Ca of the
tire. The cornering stiffness is often used as a linear approximation to the relationship
between lateral force and slip angle (see Section 3.3). The cornering stiffness can be
normalized by dividing by the vertical load. This quantity is the cornering coefficient Cc of
the tire. The cornering stiffness and the cornering coefficient both vary with the vertical
load on the tire. In general, the cornering stiffness increases with vertical load, while the
cornering coefficient 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 vertical
force is the tire lateral friction coefficient \iy. The lateral friction coefficient usually decreases
as the vertical load on the tire increases. Beyond the

slip angle at which the peak lateral
force occurs, the lateral force begins to decrease. At high slip angles a larger portion of the
contact patch is sliding than at low slip angles. Here the lateral force produced depends
largely upon the tire rubber compound, the road surface, and the interface between them.
The curve shown in Figure 3.2 represents steady-state tire lateral force
characteristics. Because of the elasticity and damping inherent in a pneumatic tire it is
actually a dynamic system within itself. When a change in slip angle occurs, the change in
lateral force lags behind. Although the effects of tire dynamics be

can modeled by including
an additional differential equation in the vehicle model for each tire, the effects are generally
Hz.6
small below input frequencies of 3 Tire dynamics are
typically modeled when
simulating emergency crash avoidance maneuvers such as a sudden lane change. Tire
dynamics are neglected in the models of this thesis.
13
3.3 Linear Tire Model
As mentioned above the initial slope of the lateral force versus slip angle curve for a
single vertical load is the cornering stiffness Ca of the tire at that load. Under certain
conditions this characteristic can be 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 = Caa (3.1)
where
c.-Z
a (3.1)
da a=0
When combined with other assumptions regarding the vehicle, linearization of the
lateral force versus slip angle relationship permits modeling of the vehicle as a linear
system. Since there is a wide variety of powerful, well-developed analysis techniques for
linear systems, much can be learned about vehicle lateral dynamics from the study of a
linear model. The range of applicability of the linear tire model is examined by comparison
of simulations of 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, several approaches to modeling tire behavior can be
found in the literature. Some models are purely empirical, based upon curve fitting of
experimentally measured tire data. Other models are primarily theoretical, with some
14
parameters determined experimentally, such as the stiffness of the tire. Each type of model
has advantages and disadvantages. The type of tire model used in this thesis is the former,
based entirely on empirical data. This type of tire model is used because of its limited
complexity and its suitability to the tire data which is available to the researcher.
The tire model chosen for this study is called tire data nondimensionalization and
Radt.13,27,30
was originated by Hugo While this technique is able to predict tire aligning
moment, longitudinal force, and lateral force for combined lateral slip, camber, and
longitudinal slip, only the lateral force due to lateral slip is of interest in this thesis. The
effects of camber on lateral force are being ignored, and the vehicle model assumes a
constant forward velocity, so it is not necessary to consider longitudinal tire force. In this
study tire aligning moments are considered to have a negligible effect on the overall
dynamics of the vehicle.
There are two main steps in using the tire data nondimensionalization technique.
The first step is preprocessing experimental tire data to determine the parameters for the tire
model. The second step is using the model to calculate the tire lateral force for a given
vertical load and slip angle. In a vehicle dynamics simulation, the first step would typically
be done before ranning the simulation. The second 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 used for this study is based on experimental data provided by the
manufacturer for a production passenger vehicle tire. The tire data is shown in tabular form
at the end of this section in Table 3.2 and is plotted in Figure 3.3. Lateral force versus slip
angle curves are available for vertical loads of 2793 N, 4190 N, 5587 N, 6984 N, and
8380 N. The slip angle varies from 0 to 15. At each of the vertical loads the lateral force
at
0
slip angle is not zero as might be expected. This is due to conicity and/or ply steer in
the tire. Conicity arises from asymmetries in tire construction, while ply steer results from
errors in the angles of the belt cords in the tire. Both conicity and ply steer depend upon
15
quality control in the manufacturing process and can be random in nature, varying from tire
to tire. Since these effects are not important for the vehicle models under consideration
here, these effects have been eliminated from the experimental data by shifting each of the
lateral force curves to the left until they intersect the origin of the plot. This zeroed data is
used 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. Since the cornering stiffness is the initial slope of the
lateral force versus slip angle curve, and since the cornering coefficient is the cornering
stiffness divided by the vertical load, the cornering coefficient can be approximated at each
load by dividing the lateral force at 1
slip angle by the vertical load. Thus, from the
experimental data the cornering coefficient at a single load is
Cc=^==%1
(3-2)
The cornering coefficients at each load are plotted in Figure 3.4. As can be seen
from the figure, the relationship between cornering coefficient and vertical load is
approximately linear. For this tire the cornering coefficient as a function of vertical load can
be represented as
CC=B, +
C3FZ (3.3)
Values of the constants B3 and C3 are listed below in Table 3.1. This expression can be
used to predict the cornering coefficient for an

arbitrary load during a simulation.
Next the lateral friction coefficient
\yy at each load must be found. This is done by
dividing the maximum lateral force for a given vertical load by the vertical load itself. Thus
for a single vertical load, the lateral friction coefficient is
16
Table 3.1: Non-Linear Tire Model Parameters
Parameter Symbol Value

Tire cornering coefficient intercept B3 0.333
Tire cornering coefficient slope C3
Tire lateral friction coefficient intercept B5 1.173
Tire lateral friction coefficient slope c5
Magic Formula curve fit parameter B, 0.5835
Magic Formula curve fit parameter c, 1.7166
Magic Formula curve fit parameter D, 1.0005
Magic Formula curve fit parameter E, 0.2517
' imax
..
Py-
(3.4)
f
The lateral friction coefficients at each load are plotted in Figure 3.5. As with the
cornering coefficients, the relationship between lateral friction coefficient and vertical load
is approximately linear for this tire. The lateral friction coefficient can be expressed as
fly=B5 + C5Fz (3.5)
Values of the constants B5 and C5 are provided in Table 3.1. This expression is used
during simulation to predict the lateral friction coefficient for an arbitrary vertical load.
With the cornering coefficient and lateral friction coefficient known at each load for
the experimental data, the normalized slip angle a can 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
When the normalized lateral force is plotted against the normalized slip angle at each
data point, the results lie on a single curve as shown in Figure 3.6. The normalized data are
then curve fit. While various functions could be used to fit this data, a popular function for
formula" here.26
fitting tire data known as the "magic is used The magic formula is a
combination of trigonometric functions and has the ability to accurately fit tire data curves
of various shapes such as lateral force, longitudinal force, and aligning moment. The
normalized lateral force is fit to the function
Fy =
Dx sin(0) (3.8)
where
6 =
Cx atan(fl^) (3.9)
and
_
.
,_ E,' atan(B,a)
v ' ;
y/ =
(l-El)a+ (3.10)
The parameters B}, C,, Dn and Ej must be determined to provide the best fit to the
normalized experimental data. The curve fitting is implemented in the MATLAB script
MagicFit.m. This script reads the normalized lateral force versus slip angle data from a file
and uses the MATLAB Optimization Toolbox function leastsq to do a non-linear least
squares fit. The leastsq function calls the function MagicError.m which computes the enors
between each data point and the curve fit function. The parameters Bp C7, D,, and
Et are
found to minimize the sum of the squares of these enors. MagicFit.m and MagicError.m
are listed in Appendix A. 1 and Appendix A.2 respectively. Values for the curve fit
parameters are given in Table 3. 1, and the function is plotted in Figure 3.6 along with the
normalized data. It can be seen from the plot that a good fit to the data has been obtained.
With a function for the normalized lateral force in terms of normalized slip angle
now available, the tire lateral force can be calculated for any combination of vertical load
18
and slip angle. First, the cornering coefficient and lateral friction coefficient are calculated
from the vertical load using Eq. (3.3) and Eq. (3.5). Second, the normalized slip angle is
calculated from the slip angle, the cornering coefficient, and the lateral friction coefficient
using Eq. (3.6). Next, the normalized lateral force is calculated from the normalized slip
angle using Eq. (3.10), Eq. (3.9), and Eq. (3.8). The tire lateral force can then be found
from the normalized lateral force, the lateral friction coefficient, and the vertical load as
Fy=FyHyFz (3.11)
This procedure is implemented in the MATLAB function NLTire.m which is listed
in Appendix A. 3. The function takes the tire vertical load and slip angle as inputs and
outputs the lateral force. Plots of lateral force versus slip angle from this function for
vertical loads of 2793 N, 4190 N, 5587 N, 6984 N, and 8380 N are shown in Figure 3.7
as solid lines along with the experimental data points.
The non-linear tire model implemented in this section accurately reproduces the
experimentally determined lateral force versus slip angle relationship of the tire used in this
study. This model is capable of predicting the lateral force produced by the tire at high slip
angles. Thus the tire model is suitable for inclusion in a model of vehicle lateral dynamics
where high tire slip angles are obtained. While this tire model only determines lateral force
due to slip angle, it can be extended to predict aligning moment due to

slip angle, lateral
force and aligning moment due to camber, and longitudinal force due to longitudinal slip.
It should 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
produced a positive lateral force for convenience. However, when the tire models are
integrated into the vehicle model appropriate care must be taken to ensure compatibility with
19
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
Table 3.2: Experimental Tire Data
SUp Angle Lateral Force @ Vertical Load (N)

(deg) 2793 4190 5587 6984 8380
0 -89 -156 -245 -334 -378
1 737 1001 1223 1357 1458

2 1388 2024 2558 2869 3025
3 1935 2847 3603 4115 4404
4 2358 3403 4359 5026 5449
5 2647 3803 4849 5649 6183
6 2802 4026 5182 6005 6672
7 2910 4175 5350 6230 6950
8 2965 4241 5470 6330 7110
9 2969 4258 5490 6370 7166
10 2950 4246 5450 6352 7180
11 2930 4222 5400 6322 7153
12 2890 4168 5350 6291 7111
13 2840 4099 5282 6235 7050
14 2780 4037 5200 6160 6978
15 2750 3977 5121 6090 6900
6 8 10
Slip Angle (deg)

Figure 3.3: Experimental Tire Data
21
0.35
0.30
TO
0) 0.25
TJ
z
s^-
c
0.20
o
*=
CD
o
O
0.15
O)
c
k-
0)
^
0.10
o
O
0.05
2000 4000 6000 8000 10000

Vertical Load (N)
Figure 3.4: Tire Cornering Coefficient
1.2
1.0
c 0.8
g>
'o
it=
a)
o
O 0.6
c
o
o
it
a
0.4
a>
to
0.2
0.0
2000 4000 6000 8000 10000
Vertical Load (N)
Figure 3.5: Tire Lateral Friction Coefficient
22
1.2
1.0
CD
| 0-8
CD
5 0.6
CD
N
15
? 2793 N
0.4
o 4190N
A 5587 N
0.2 6984 N
X 8380 N
i II
ill ....
0.0 II
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Normalized Slip Angle
Figure 3.6: Tire Normalized Lateral Force
2 4 6 8 10 12 14 16
Slip Angle (N)
Figure 3. 7: Reconstructed Tire Lateral Force
23
4 Two Degree-of-Freedom Vehicle Model
4.1 Introduction
Under normal
driving conditions the driver and vehicle form a closed-loop system.
The driver observes the motion of the vehicle and provides control inputs to produce the
desired motion. However, this work is concerned primarily with predicting the open-loop
lateral response of a road vehicle to control and disturbance inputs.
The simplest model which can realistically be used to examine the lateral response
"bicycle"
of a road vehicle is the two degree-of-freedom (DOF) model. As noted in
Chapter 2, this model has been used extensively in the literature to study road vehicle lateral
response. Although this model greatly simplifies the vehicle system, much can be learned
about vehicle lateral response through its use. The model demonstrates the effects of major
design and operational parameters such as tire properties, inertia properties, mass center
location, wheelbase, and forward speed. Conclusions of practical significance regarding
lateral directional be drawn using this

model.17
road vehicle control and stability can simple
In this chapter the two degree-of-freedom vehicle model is described in detail. The
equations of motion are derived from basic principles of dynamics. Next, relationships for
the tire slip angles are derived from the vehicle kinematics. From here the model is
developed in two forms, linear and non-linear.
In the linear form of the model, additional assumptions are made which
simplify the
kinematic relationships and tire mechanics. This simplification allows powerful linear
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,
aerodynamic side force, and road side slope are developed. Several measures of steady-
state and transient response are derived. Next, the frequency response of the vehicle is
examined using bode plots. Finally, vehicle response is simulated for a 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 non-linear kinematics and a non-linear tire
model are used. The tire model, as described in Section 3.4, accounts for the non-linear,
vertical load-dependent lateral force versus slip angle relationship. Simulation of the model
response to steering inputs is performed and the results are compared to the linear model
simulation.
4.2 Description of Model
The two degree-of-freedom model used in this chapter is shown in Figure 4.1. The
vehicle is modeled as a single lumped mass rigid body and has lateral velocity v and yaw
velocity r degrees of freedom. The forward velocity u is assumed to be constant. The pair
of tires at each end of the vehicle is represented by a single tire at the centerline of the car.
The vehicle has a wheel base L, a mass m, and a yaw mass moment of inertia Ia, with its
mass center located a distance a front the front axle and a distance b from the rear axle.
Rotation of the front tire about the vertical axis relative to the body is permitted and
is measured by the front steer angle 8, with clockwise rotation considered positive. The
front steer angle is the only control input considered. In this work position control is
assumed. Position control is defined by SAE as "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),
the force This is in to force control where inputs are in

required."29
independent of contrast
the form of forces or moments and are independent of displacement.
The standard SAE vehicle-fixed coordinate system x-y-z is used to describe the
motion of the vehicle. The x-axis is positive in the forward direction, the y-axis is positive
to the right, and the z-axis is positive down. The origin for the coordinate system is at the
25
Figure 4. 1 : Vehicle Model
vehicle mass center, and the coordinate system translates and rotates with the vehicle.
Motion is only permitted in the x-y plane.
4.2. 1 Assumptions
Several simplifying assumptions are made to facilitate the development of the
model:
Constant vehicle parameters
Constant vehicle forward velocity
Motion in x-y plane only (ignore vertical, rolling, and

pitching motions of sprung
mass)
Vehicle is rigid
26
Vehicle is symmetrical about z-x plane
Road surface is smooth
Ignore effects of drive line
Ignore longitudinal gravity effects (ignore road slope in longitudinal direction)
Ignore all longitudinal forces (tire driving/braking forces, tire roiling resistance,
aerodynamic drag)
Ignore suspension system kinematics and dynamics
Ignore steering system kinematics and dynamics
Position control for steering input
Ignore tire steer due to roll of sprung mass
Ignore tire steer due to chassis compliance
Ignore tire slip angles resulting from lateral tire scrub
Ignore lateral and longitudinal load transfer (vertical tire forces remain constant)
Tire properties are independent of time and forward velocity
Ignore tire lateral forces due to camber, conicity, and ply steer.
Ignore tire aligning moments
Ignore effect of longitudinal tire slip on tire lateral force
Ignore tire dynamics (no delay in lateral force generation)
Ignore tire deflections
4.2.2 Vehicle Parameters
The nominal values of the vehicle parameters used for the two degree-of-freedom
model in this study are given in Table 4.1. The International System of metric units (SI) is
27
used for all calculations. The base units are meter (m), kilogram (kg), and second (s).
Force is measured in the derived unit newton (N). For convenience, the fraction of weight
on the front axle/is used to define the position of the mass center along the wheelbase.
The mass center location parameters a and b are then calculated from/ and L. These
parameters are representative of a production automobile.
Table 4.1: Vehicle Parameters

Vehicle mass m 1775 kg
Yaw moment of inertia ' 1960 kg-m2
Front axle weight fraction f 0.52

Wheelbase L 2.372 m
Distance from mass center to front axle a 1.139 m
Distance from mass center to rear axle b 1.233 m
Distance from front axle to aerodynamic side force c 1.25 m
4.2.3 Free-Body Diagram
There are three types of external forces acting on a vehicle which are considered in
this model: tire lateral forces, aerodynamic side force, and gravitational side force. The tire
lateral forces iând Fyr occur due to tire slip angles. The aerodynamic side force F a is a
disturbance input acting at a distance c behind the front axle. This type of force acts on a
vehicle when it encounters a crosswind. The gravitational side force F is a disturbance
input acting at the vehicle mass center and resulting from a side slope in the road. All forces
are considered to be positive when acting in the positive y-direction. These forces are
shown acting on the vehicle in Figure 4.2.
28
Figure 4.2: Free-Body Diagram
4.3 Derivation of Equations of Motion
The equations of motion for the two degree-of-freedom vehicle are derived using
basic principles of Newtonian mechanics for rigid body motion relative to translating and
The basic relating the forces

systems.31
rotating coordinate equations and moments

acting
on a rigid body to the acceleration of the body are
If =g
(4.1)
Xmg=hg
where F and
Mg are external forces and moments about the mass center (in vector form)
acting on the body, and G and
HG are the linear and angular momenta of the body (also in
vector form) measured relative to an inertia! reference frame. Since in this model only
29
motion in the x-y plane is considered and all longitudinal (x-direction) forces are being
ignored, Eq. (4.1) become
ma
(4.2)
Since the x-y-z coordinate system is fixed to the vehicle with its origin at the vehicle
mass center, the translational velocity of the vehicle mass center and rotational velocity of
the vehicle are identical to those of the x-y-z system. From Figure 4. 1, the velocity V0 of
the origin of the x-y-z system is
V0=wi + vj (4.3)
and the angular velocity Q. is
Q = rk (4.4)
Since the x-y-z system is rotating, the unit vectors are changing with 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)
= (u- vr)i + (v + ur)j
Similarly, the angular acceleration of the x-y-z system relative to the inertial frame is
+ Q.XQ.
dt (4.6)
=
rk
Thus the acceleration values of interest are
a =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
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+Fyyg
(4.8)
cos 8 (c
X Mz bFyr - -
= -
aF^ a)Fyya
Substitution of Eq. (4.7) and Eq. (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)

(4.9)
aF^ cos 8 bFyr (c Ia r
- - -
=
a)Fya
Here u is the vehicle forward velocity and is a constant. The state variables are the
lateral velocity v and the yaw velocity r.

F^ and Fyr are the front and rear tire lateral forces.
Fya and F are the aerodynamic side force and gravitational side force disturbance inputs,
while the control input is the front steer angle 8.
4.4 Derivation of Tire Slip Angles
As discussed in Chapter 3, the lateral force produced by a tire depends upon,
among other things, the vertical load on the tire and the slip angle of the tire. In this model
the vertical load remains constant, but the front and rear tire slip angles vary as functions of
the lateral velocity v and the yaw velocity r, which are the system state variables. Thus it is
for the front tire slip angles,

necessary to develop expressions and rear
af and ar, in terms
of v and r.
Figure 4.3 is a kinematic diagram of the vehicle showing the tire velocity vectors
and tire slip angles. Each slip angle is shown in its positive sense as the angle between the
tire and the tire velocity vector. A positive slip angle implies a clockwise rotation from the
tire to its velocity vector. However, for a positive steer input as shown in the figure, the tire
31
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 determining the tire slip angles is finding the tire velocities. Since
the translational and rotational velocities of the vehicle-fixed x-y-z coordinate system are
already known from Eq. (4.3) and Eq. (4.4), it is convenient to use the principle of relative
The Vf and Vr of the front and rear tires

velocities.31
motion to derive the tire velocities
respectively are
Vf=V0+QxRf
(4.10)
V=V+fixR
where R, and Rr are position vectors from the vehicle mass center to the front and rear tires:
32
Rf =a\
(4.11)
R, =-b\
Thus the tire velocities are
V.f =
u\ + (v
V +
ar)\
'
(4.12)
Vr =
u\ + (v -
br)j
In general, if the velocity and steer angle of the ith tire are 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,
(v + ar\
a/=atan(
^
j-S
(4'14)
fv-bA
ar -
atan
V u J
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 assumptions are
made 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
tire lateral force versus slip angle

relationship.13
range of the Most automobile driving is

done within this range, so results from the linear model are applicable over a wide range of
driving situations.
33
It is common in modeling of vehicle lateral dynamics, particularly with linear
models, to use the vehicle sideslip angle p instead of the lateral velocity v to describe the
lateral motion of the vehicle. The vehicle sideslip angle is the angle between the vehicle-
fixed x-axis and the vehicle velocity vector Vn as shown in Figure 4.3. Similar to
Eq. (4.14) for tire slip angles, the vehicle sideslip angle is
p
(;)
= atari-
(4.15)
A positive vehicle
sideslip angle implies clockwise rotation from the x-axis to the velocity
vector. For a given steer angle, the sideslip angle may be positive or negative, depending
upon the forward speed.+
In this thesis the vehicle sideslip angle is used in place of lateral
velocity when finding transfer functions, steady-state response measures, transient
response measures, and frequency response. However, since simulation of the non-linear
model is more straightforward with lateral velocity as a state variable, simulation of the
linear model is also performed with lateral velocity as a state variable to facilitate parallel
development of the two simulation models.
Once the linearized equations of motion are written, transfer functions for the state
variables in terms of the control and disturbance inputs are developed. From these transfer
functions, measures of steady-state and transient response are derived and the
frequency
response is examined. Simulation of the model is performed for various
steering and
disturbance inputs.
4.5.1 Additional Assumptions
The following assumptions are made for the linear two degree-of-freedom model in
addition to those listed in Section 4.2.1:
See Section 4.5.18 for more information.
34
Linear tire lateral force versus slip angle relationship
Small steer angle, tire slip angles, vehicle sideslip angle, and road side slope angle.
4.5.2 Vehicle Sideslip Angle

With the small angle assumption, Eq. (4.15) for the vehicle sideslip angle becomes
P-
(4.16)
u
However, if the vehicle sideslip angle is small then
cos(P)
=
i^l
or (4.17)
u~V
where Vis the magnitude of the vehicle velocity V0. Now Eq. (4.16) becomes
(4-18)
P~"
4.5.3 Tire Slip Angles
With the small angle assumption, the tire slip angles become
v + ar
=
af u
-8
v br (4.19)
a =
Furthermore, if vehicle sideslip angle is used in place of lateral velocity, then the tire slip
angles can be expressed as
+
aff =P
K r-8
V
b (4.20)
r=P--r
35
4.5.4 External Forces and Moments
From the free-body diagram of the two degree-of-freedom model in Figure 4.2 it is
seen that the four external forces acting the front tire lateral force
on the vehicle are
F^ the
rear tire lateral force Fyr, the aerodynamic side force Fya, and the gravitational side force Fyg.
With the assumption that tire lateral forces are linear functions of tire slip angle, the
linear tire model of Section 3.3 can be employed. From Eq. (3.1) the tire lateral forces are
Fyf
*
=
Cfaf
' '
_
(4.21)
Fyr
~
^rr
where C/and Crare the front and rear tire cornering stiffnesses and are the effective
cornering stiffnesses of both tires on an axle. Thus, for example, Cfis twice the cornering
stiffness of a single front tire. As a result, F^ and Fyr are the sums of the tire lateral forces
of both tires on an axle. Since the tire slip angles are negative for a positive steer angle, the
cornering stiffnesses must also be negative in order to produce the positive lateral forces
required by the sign convention. For further explanation of this tire model see Section 3.3.
Values for tire cornering stiffnesses for the vehicle studied are obtained though the
application of Eq. (3.3) and are given in Table 4.2.
Table 4.2: Linear Tire Model Parameters

Front tire cornering (two tires)
stiffness
C~f -2461
N/deg
Rear tire cornering stiffness (two tires) Cr -23 11 N/deg
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
positive when acting on the vehicle in the positive y-direction.
36
Figure 4.4: Gravitational Side Force
The force
gravitational side
Fyg is a function of the side slope in the road and is
shown in Figure 4.4. The gravitational side force is positive when acting on the vehicle in
the positive y-direction. Thus the gravitational side force can be expressed as
Fyg =mg sinQ (4.22)
where g is the acceleration due to gravity and 9 is the road side slope, which is positive for
a road which is sloping down on 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 =
mgQ
(4.23)
With the gravitational side force expressed in terms of the road side slope, the side slope 8
can now be considered to be the disturbance input instead of the force itself.
4.5.5 Equations of Motion
With substitution of the tire slip angles from Eq. (4.20), the tire lateral forces from
Eq. (4.21), and the gravitational side force from Eq. (4.23) into Eq. (4.9), the linearized
equations of motion become
37
(Cr+Cr)p + ^ r-r-Cf5 + Fya+mgQ =

mV$ + mVr
C*
rt -U hi
f-
(aCf bCr )p + (c -
aCf8
-
-
r -
a)Fya =1J
The assumption that the steer angle is small has also been applied to reduce the
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 steer angles are
often required.
To simplify manipulation of the equations of motion, the external force and moment
terms of the left sides of Eq. (4.24) can be rewritten in terms of stability derivatives. This
technique has been used extensively by early researchers in automobile lateral dynamics
Milliken.6,7,13
such as Leonard Segal, David Whitcomb, and William In addition to
simplifying the equations of motion, the derivatives themselves have physical meaning
which can give further insight into road vehicle lateral dynamics.
The stability derivatives are the rates of change of the external forces or external
moments acting on the vehicle with respect to p, r, or 8. There are three stability derivatives
associated with lateral force and three associated with yaw moment. The equations of
motion in stability derivative form are
Fpp + Yrr + Ys8 + Fya + mgQ =

mV$ + mVr
(4'25)
+ Nrr + N6S-(c-a)Fja=I,tr
Nfi
where the stability derivatives are defined as follows:
38
Damping in Sideslip
V cf+cr
- -
=
aCf
-
bCr Lateral Force/Yaw

Yr Coupling
V
Y5 = Control Force
~Cf
(4.26)
Directional Stability
NP =
aCf-bCr
Nr-
a2Cf+b2Cr
Yaw Damping
V
Ns-= Control Moment
-aCf
In the two degree-of-freedom model under consideration the stability derivatives are
all constants. 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 always negative by
definition, the signs of the stability derivatives can be obtained. The
damping-in-sideslip
derivative Yj, and yaw damping derivative Nr are always negative. The control force
derivative Ys and control moment derivative Ns are always positive. The lateral force/yaw
coupling derivative Yr and directional stability derivative A/j, are both either positive or
negative depending on the relative magnitudes of aCf and bCr. If
aCf is greater than the bCr,
then the derivatives the vehicle understeers. If
aCf is less than bCr,
are positive and 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. Understeer, oversteer, and neutral steer are
discussed in more detail in Section 4.5.14.
4.5.6 Transfer Functions
Now that the equations of motion are available in a simple, compact form, transfer
functions can easily be found relating the outputs p and r to the inputs 5, 6, and F . From
these inputs and outputs six transfer functions can be formed. The transfer functions can be
used to examine many aspects of system response such as steady-state

response, frequency
39
response, and poles and zeros. The derivation of the transfer functions and all analytical
expressions 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:
i
( Y \ 1 >
s--
1 (g\
mV mV mV mV
m+ Fjs) + W (4.27)
Na N. a c
s--
As); yOj
From these Laplace-domain equations of motion, the transfer functions for vehicle
sideslip angle are found to be
75 j
NrYs+N8(mV-Yr)
i*v
!- mV
(4.28)
N~
Y\ + !W+N9(mV-Yr) LmV
\^+'mVj
1
s+ -
(c-a)(mV-Y)-Nr
^ T-i- -
P mV I^mV
(') =
(4.29)
ya
s2-
I
Y\{N^+N*(mV-Yr)
mV I,mV
8Nr
8S
Ya \
(4.30)
^NrY^N^mV-Yr)
s2-
mV ) IzjnV
Similarly, the transfer functions for yaw velocity are
N Va-^P
8j|
I I,mV
s-
(4.31)
s2-
N.
i
+ -
"i
N^ + N^mV-Y,)
v'
mV IumV
40
a-c yp(c-fl)
p
+
iVp
s+
-
-
^mV
J-(S) = I* , r (4.32)
ImV
K
\*zz
ntV zz
gN
r IV
L(s)
BK)
= is-l (4 33)
S
2 -
(Nr-H
Y,]5-1 NrY,+N,{mV-Yr)
U mVj IjnV
The above transfer functions for the lateral dynamics of the two degree-of-freedom
vehicle can be used to examine steady-state behavior. For each of the three types of inputs,
the steady-state step input response gains in vehicle sideslip angle, yaw velocity, tire slip
angles, path curvature, and lateral acceleration are found. The steer angle required to
produce a given turn radius is calculated. In addition, measures of steady-state vehicle
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 steady-state step response of vehicle sideslip angle and yaw velocity can be
found for each of the three inputs by applying the Final Value Theorem to the transfer
functions.33
Before the Final Value Theorem can be applied, however, the stability of the
system must be verified. 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 Gain
Application of the Final Value Theorem to the vehicle sideslip angle transfer
functions for steer angle, aerodynamic side force, and road side slope (Eq. (4.28),
Eq. (4.29), and Eq. (4.30) respectively) gives the following steady-state response gains:
41
P NrY&+N&(mV-Yr)
(4.34)
8 NrYA+N9(mV-Yr)
P (c-a)(mV-Yr)-Nr
(4.35)
ya NrY+NJmV-Yr)
mgNr
(4.36)
+
NrYf N9{mV-Yr)
Results for the sideslip angle response gains and for the response gains that follow
are given in Table 4.3. The vehicle and tire parameters used for the calculations are given in
Table 4. 1 and 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
"&-*&
(4.37)
NrY9+N,(mV-Yr)
rp(c-a) + JVp
(4.38)
ya NrYf+N,{mV-Yr)
r mgN?
(4.39)
e NrYf+N9{mV-Yr)
4.5.9 Front Tire Slip Angle Gain
Once steady-state values of vehicle sideslip angle and yaw

velocity are known the
steady-state 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
a, V(c -
a)(mV
-Yr)
+ a(c-
a)7p +
aiVp
-
VNr
(4.41)
ya
v{NrY^N^mV-Yr))
a, mg{aN[,-VNr)
(4.42)
e
v(NrY^N9{mV-Yr))
4.5.10 Rear Tire Slip Angle Gain
Similarly, the steady-state rear tire slip angle gains are
(4.43)
8
v(ivryp+ivp(mv-yr))
V(c a)(mV
WVp VNr
- -Yr)-b(c- - -
_ a)Y&
(4.44)
~
Fya v(ivrrp+;vp(mv-yr))
O^ -mg^+VN^
(4-45)
e
~v(NrYB +
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 =
(4.46)
V
The path curvature gains for each of the inputs are
1/R N Y -N Y
(4.47)
8
v(ivryp+;vp(mv-yr))
1/R yp(c-a) + iVp
(4.48)
ya
v(NrY{i+N{i(mV-Yr))
43
mgN,
1/R
(4.49)
6
v(Arryp+ivp(mV-yr))
4.5.12 Lateral Acceleration Steady-State Step Response Gain
Lateral acceleration is an important measure of vehicle cornering performance.
Lateral 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
passenger car driving is done below this limit, although a typical passenger car may be able
to attain maximum lateral accelerations of 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 4 g lateral The lateral

exceed acceleration. steady-state acceleration
Ay
"g"
expressed in units of is
rV
Ay =
(4.50)
8
The lateral acceleration gains for each of the inputs are
V(^YS-N&YP)
(4.51)
8
g(ivryp+ivp(mv-yr))
A, v(Y[i(c-a) + Nfi)
(4.52)
ya
g(ivryp+ivp(mv-yr))
mVJVp
=
(4.53)
NY+NJmV-Yr)
44
Table 4.3: Steady-State Response Gains

V= 100 km/hr
Response Steer Angle Aerodynamic Road Side

(6) Side Force (FJ Slope (9)
Vehicle sideslip angle (P) / 0.0615/
-1.52
2.82xlO"4o/N
Yaw velocity (r) 0.197 rad/s
1
2.15xl0'4rad/s/
rad/s /N
Front tire slip angle (af) 1
0.0620 1
-2.05
2.66xlO"4/N
Rear tire slip angle (ar) 1
0.0609 1
-2.02
3.00xl0'4o/N
Path curvature (1/R) 7.10xl0-3l/m/ 7.76xl0-6l/m/
-2.54xlO-7l/m/N
Lateral acceleration (Ay) 0.559 g/
6.11xl0-4g/
-2.00xl05g/N
4.5.13 Steady-State Steer Angle
The steady-state steer angle required for the vehicle to turn at a constant path radius
R can be found by solving the yaw velocity gain for steer angle input expression of
Eq. (4.37) for the steer angle 8. After some manipulation, the steer angle is
mV2Na
=- +
R
(4.54)
/?(/vpy5-/v8yp)
Substitution of the stability derivative definitions into this expression yields
f ,
a b
nYL
"
R C C L R
(4.55)
Examination of the kinematics of a turning vehicle indicates that the steer angle can also be
in terms the tire slip angles

as17
expressed of
8 =
ar +ar (4.56)
From Eq. (4.55) it can be seen that as the vehicle velocity approaches zero the steer
angle becomes
L
Ac ker
R
(4.57)
45
This steer angle is known as the Ackerman steer angle and is the steer angle required to
negotiate a turn of path radius R at low vehicle velocity. At low speeds the lateral
acceleration, tire slip angles, and tire lateral forces are negligible and turning behavior is
governed solely by geometry considerations.

For the vehicle used in this study the Ackerman steer angle for a 50 m radius turn is
2.72 . The 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 turn, is
basically the rate of change of steer angle with lateral The definition
acceleration.1
of the
understeer gradient can be seen in the steady-state steer angle expression for a constant
radius turn from Eq. (4.54):
V2
L
Sss=- + Kus
(4.58)
R gR
Thus the understeer gradient, expressed in units of radians, is
mgNa
^-
K "*
=
(4.59)
N Y
ivp-i5
-
N Y
JV6-p
or substituting the stability derivative definitions
a b mg
K =
(4.60)
C C L
The understeer gradient is a measure of the amount of understeer or oversteer a
vehicle has. If to maintain a constant path radius the steer angle must increase as vehicle
forward velocity, and hence lateral acceleration, increases, the understeer gradient is
positive and the vehicle is said to be understeer. If the steer angle must decrease, the
understeer gradient is negative and the vehicle is oversteer. If the steer angle does not
46
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 and rear tire slip angles. If
the magnitude of the front tire slip angle is greater than the rear, the vehicle is understeer. If
the rear tire slip angle is greater, the vehicle is oversteer. If the slip angles are equal, the
vehicle is neutral steer.
For this simple model the understeer gradient is a function of tire cornering
stiffnesses and the weight distribution. Other vehicle characteristics which influence the
understeer gradient include lateral load transfer distribution, roll camber, roll steer, lateral
force compliance steer, tire aligning moments, tire longitudinal forces, steering system
compliance, and differential type.
The understeer gradient of a vehicle can be measured experimentally. The two most
test.17
common methods of testing are the constant radius test and the constant speed In the
constant radius test the forward velocity of the vehicle is varied as the car is driven on a
constant path radius. The steer angle is varied to maintain the constant radius. The
understeer gradient is then calculated as the rate of change of steer angle with lateral
acceleration:
d8
K.=
(4.61)
y
In the constant speed test the vehicle forward velocity is held constant while the
path radius is varied and the steer angle required to attain the radius is measured. The
understeer gradient from the constant speed test is
d8 gL
K~ =
(462)
M,-^
Most passenger cars have understeer gradients between 1 and 10 deg. Understeer is
designed into passenger cars to assure directional stability over a wide range of operating
conditions. The vehicle used for this example has an understeer gradient of 0.0626 deg,
47
indicating that the vehicle has a very small amount of understeer in the linear response
range.
4.5.15 Stability Factor

factor.6
A second measure of vehicle
cornering behavior is the stability 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
KV2
(4.63)
l+
where the K is the stability factor. Solving Eq. (4.63) for K and substituting Eq. (4.37),
mN
K= .
5 (4.64)
l(n,y5-nsy?)
or substituting the stability derivative definitions,
L2
(4.65)
C C
If K is positive the vehicle is understeer, while if K is negative it is oversteer. The
4.69xl0"5
sample vehicle has a stability factor of s2/m2. Since the stability factor is positive,
the vehicle is understeer.
4.5.16 Neutral Steer Point
The neutral steer point is "the point along the chassis at which an external lateral
velocity."13
force can be applied which produces no steady-state yaw On a real vehicle there
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 a fictitious lateral force Fm
applied to the vehicle at a distance d behind the front axle. As in Section 4.5.8 the Final
48
Value Theorem can be used to find the steady-state yaw velocity gain for the fictitious
lateral force.
r Y^d-a) + N^
(4.66)
NrY^N,{mV-Yr)
Setting this yaw velocity gain equal to zero and solving for d,
d = a-^
(4.67)
Yt
Substituting the stability derivative definitions, the neutral steer point is located at a distance
behind the front axle of
d = -^-
(4.68)
Cf + Cr
If the neutral steer point is behind the mass center, the vehicle is understeer, while if
it is in front of the mass center the vehicle is oversteer. The sample vehicle has a neutral
steer point located at a distance of 1.149 m behind the front axle. Since the mass center is
located 1.139 m behind the front axle, the neutral steer point is behind the mass center and
the vehicle is understeer.
4.5.17 Static Margin
Another way of describing understeer is with the static margin. The static margin
SM is the distance between the mass center and the neutral steer point, normalized by the
wheelbase.13
SM =
^p
(4.69)
--
Lip
or, substituting the stability derivative definitions,
aCf
-
bCr
SM = 7-i
r-r (4.70)
L(Cf+Cr)
49
If the static margin is positive, the vehicle is understeer. If it is negative the vehicle
is oversteer. The sample vehicle has a static margin of 0.00428, indicating that the vehicle
is understeer.
4.5.18 Tangent Speed
At low speed in a steady-state turn the rear axle travels on a smaller circle than the
front axle. As speed is increased the radius of travel of the rear axle increases and surpasses
that 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
radius is called the tangent speed Vmn. At the tangent speed the vehicle angle is
sideslip
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.
&r rS
Vtan=
M (4.71)
mNs
However, since the yaw velocity stability derivatives Yr and Nr are both functions
of velocity, it is necessary to substitute the stability derivative definitions and solve for V.
When this is done the tangent speed is
v bLCr
vm=4V am
(4-72)
The sample vehicle used in this study has a 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
4.5.19 Critical Speed
Consideration of Eq. (4.37) for steady-state yaw velocity gain for steer angle input
reveals that the yaw gain could become infinite if the denominator were zero. The
velocity
speed at which this happens is called the critical speed. The critical speed can be found by
setting the denominator of Eq. (4.37) equal to zero and solving for V. In terms of the
stability derivatives, the critical speed is
NaYr-NY
" "
Vcril =
(4-73)
mNp
However, the stability derivatives Yr and Nr are functions of velocity. Therefore, it is
necessary to substitute the stability derivative definitions and solve for V. The critical speed
is then
CfCL2
(4'74)
v'-=-l^K^)
At this speed the vehicle is unstable. A small steering input produces a theoretically
infinite yaw velocity response. Given that the tire cornering stiffnesses are negative
by
definition, it can be seen from Eq. (4.74) that a critical speed
only exists if bCr is greater
than
aCf In Section 4.5.5 it is noted that when this condition exists the vehicle is oversteer.
Thus a vehicle has a critical speed only if it is oversteer. When an oversteer vehicle reaches
its critical speed it becomes directionally unstable. The more oversteer a vehicle has, the
lower its critical speed. The 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
steady-
characteristic speed is the speed at which the steer angle required to produce any
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) and solving for the velocity V. Since the solution to this equation is a
function of
stability derivatives which depend on velocity, the stability derivative
definitions must be substituted and the resulting equation solved for V. The characteristic
speed is then
2
CfC L
f r
ychar=
\
'vmiaCf-bCA
,
(4.75)
The 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 condition for an understeer vehicle, only
understeer vehicles have a characteristic speed. Neutral steer vehicles have an infinite
characteristic speed and oversteer vehicles have no characteristic speed. The more
understeer a vehicle has, the lower its characteristic speed.
The sample vehicle has a characteristic speed of 525 km/hr. Since the characteristic
speed 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
are all measures of steady-state system response. The lateral transient response of the two
degree-of-freedom road vehicle is now examined.
From the Laplace-domain equations of motion of Eq. (4.27), the characteristic
equation for the system is
(N Y)s+ NrYf+Nf(mV-Yr)
H
s1-
+ 0 (4.76)
/\7
-*- - -
=
V7* mVJ L,rnV
52
From this characteristic equation, the undamped natural frequency, damping ratio,
and poles of the system can be found.
4.5.22 Undamped Natural Frequency

From the characteristic equation of Eq. (4.76) the natural frequency of the system is
NY+NJmV-Y)
* = "
(4-77)
AV r v
ImV
Substituting the stability derivative definitions yields
\L2CfCr aCf-bC
'.""'
>,.=, I^f +
ImV2
I
(4-78)
zz
With the expression in this form the effects of various parameters on undamped
natural frequency, and consequently, system response time can be seen. From the first term
the natural frequency decreases with the yaw moment of inertia, the mass, and the square
of the forward velocity of the vehicle. In addition, it increases with the product of tire
cornering stiffnesses and with the square of the wheelbase. The numerator of the second
term is the directional stability iVp. This term is positive if the vehicle is understeer, negative
if it is oversteer, and zero if it is neutral steer. Thus from this model, all else being equal,
an understeer vehicle has a higher natural frequency and lower response time than an
oversteer vehicle.
The sample vehicle has an undamped natural frequency of 1.01 Hz at a forward
speed of 100 km/hr. The natural frequency decreases as vehicle velocity increases as
shown in Figure 4.5.
4.5.23 Damping Ratio
The damping ratio of the system can also be obtained from the characteristic
equation. In terms of the stability derivatives the damping ratio is
53
50 100 150 200

V (km/hr)
Figure 4.5: Natural Frequency vs. Vehicle Velocity
IJi+mVNr
(4.79)
2^mV(iVryp+/Vp(mV-yr))
If the stability derivatives are substituted the damping ratio becomes
Izz(Cf+Cr) + m(a2Cf+b2Cr)
(4.80)
(aCf bCr ))
mV2
2Îam{l}CfCr + -
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, with understeer vehicles tending toward underdamped and oversteer vehicles
As dynamic systems, the

overdamped.13
tending toward with other

damping ratio affects
the response time and overshoot of the road vehicle to lateral control and disturbance
inputs.
At a forward speed of 100 km/hr the sample vehicle has a 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
4.5.24 System Poles
The poles of the system can be found by solving the characteristic equation of
Eq. (4.76) for s. Doing this yields the poles
mVNr +/ay J(mVNr +

Ijrf -Alumv\NTY^
+N?{mV-
Yr))
Pi2=
(4.81)
H^mV
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 such
that the desired system response is obtained. The pole locations also provide an indication
of system stability. 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 a forward speed of 100 km/hr the sample vehicle has poles of -6.301 0.918/.
Since the poles are complex conjugates, the vehicle is underdamped. Furthermore, since
the real part of the poles is negative, the system is stable. As speed decreases the poles
move together until they meet on the real axis at -9.897 at a vehicle speed of 63.7 km/hr. At
this speed the vehicle is critically damped. As speed increases from 100 km/hr the poles
move farther apart and approach the imaginary axis. However, even at a forward speed of
300 km/hr the poles remain in the left half of the 5-plane, indicating that the vehicle remains
stable for reasonable speeds.
4.5.25 System Zeros
Further insight into the nature of system response can be obtained by examining the
system zeros. 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 are given in Eq. (4.28) through Eq. (4.33).
The zero for sideslip angle response to steer angle input is
55
NrY&+N&{mV-Yr)
zp/s
-
(4.82)
The zero for sideslip angle response to aerodynamic side force input is
Nr-(c-a)(mV-Yr)
hipya
-
'
(4.83)
The zero for sideslip angle response to road side slope angle input is
Nr
p/e (4.84)
j
The zero for yaw velocity response to steer angle input is
Zr/6
=
(4.85)
mVNx
The zero for yaw velocity response to aerodynamic side force is
_;vp
+ yp(c-a)
zr IF (4.86)
mV(c -
a)
There is no zero for yaw velocity response to road side slope since s does not
appear 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 system zeros for the sample vehicle with a forward speed of 100 km/hr are
given in Table 4.4.
Table 4.4: System Zeros
V=100km/hr
Response Steer Angle Aerodynamic Road

Side Force Side Slope
Vehicle sideslip angle 21.53 -9.86 -7.06
Yaw velocity -5.59 -5.04
56
The effect that zeros have on response depends upon their location relative to the
that is in the right half of

poles.33
The closer a zero is to a pole, the larger the effect. A zero
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. This response
can be seen in the lateral velocity and sideslip angle response to steer angle 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 values of all of the zeros increase with vehicle speed. The sideslip angle / steer
angle zero 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 /\ a*
CO
-0.5
-10 5 10 15 20 25
Re(s)
?Pole Sideslip Angle / Steer Angle Zero
ASideslip Angle /Aero Side Force Zero xSideslip Angle / Road Side Slope Zero
XYaw Velocity /Steer Angle Zero Yaw Velocity /Aero Side Force Zero
Figure 4.6: Poles and Zeros
57
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
vehicle is driven through regularly spaced cones by means of a sinusoidal steering input.
The frequency of the input required to negotiate the slalom depends upon the vehicle speed
and 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.
Sinusoidal steering inputs may also be used in emergency maneuvers such as a double lane
change.
Examining the frequency response of the vehicle may provide an indication of its
performance in such a maneuver. Since it is generally desirable to minimize the response of
a vehicle to disturbances such as side winds and road side slope, frequency response
techniques can be used to examine the response of the vehicle to periodic disturbance
inputs.
Phase lags in response to steering input require the driver to adjust his input to
obtain the desired response, making the vehicle more difficult to drive. Smaller phase lags
improve
controllability.21
tend to vehicle The frequency response of the sample vehicle with

a forward velocity of 100 km/hr is examined using the bode plotting capability of the
MATLAB Controls Toolbox. The gain and phase responses of vehicle angle and
sideslip
yaw velocity to steer angle, aerodynamic side force, and road side slope are plotted in
Figure 4.7 through Figure 4. 12. The MATLAB script DOF2LFreq.m, which is listed in
Appendix 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.
DOF2LFreq.m calls the scripts DOF2Control.m, which sets program execution
parameters;
DOF2Param.m, which sets vehicle and input magnitude parameters; and
58
DOF2DependParam.m, which calculates vehicle parameters which depend on other

parameters. These scripts are listed in Appendix C.l through Appendix C.3.
The frequency response of a road vehicle changes as forward velocity changes. For
most of the responses the magnitude of 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 response is the only response which
experiences significant change 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
magnitude of the forward speed relative to the tangent speed of the vehicle. This is a result
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 Figure 4. 14 for forward
speeds of 30 km/hr and 49.84 km/hr respectively.
At 30 km/hr the gain is flat up to approximately 1 Hz at 0.33 deg/deg and the phase
0
goes from at 0. 1 Hz to at 100 Hz. The phase response is typical of a second order
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.
However, there is a significant peak in the gain at approximately 2 Hz, which is the
90
undamped natural frequency at 49.84 km/hr. The phase goes from at 0. 1 Hz to at
100 Hz, crossing zero at the undamped natural frequency. The frequency response at 100
180
km/hr is shown in Figure 4.7. The phase goes from at 0.1 Hz to at 100 Hz.
180
There is a phase lead at low frequency because above the tangent speed a positive
steady-state steer angle produces a negative sideslip angle as shown in Section 4.5.18.
Also of interest is the yaw velocity / road side slope phase response. The phase
0
goes from at 0.1 Hz to at 100 Hz. The phase response of this transfer function
differs from the others due to the lack of a zero.
59
At low frequencies the gains for each input and output combination approach the
values of the steady-state

step input response gains shown in Table 4.3 for a forward
velocity V = 100 km/hr.
60
1.6
O)
1.2 v i
T ,
-i^^^ t 1 i t
- - _ _ _
r
- -
n
- -i--i-
T-i-i-i-r-
- _ _ ~
t
- - ~i~ "
r
- \-
i ~i
- i~ i-
o>
CD
0.8
CO
O 0.4
0.0
0.1 1 10 100
Frequency (Hz)
180
D)
CD
90
-
- ----!---
+ _
_i
_-|_ |- --
i * i t
i t l + l t i I
l l
+^^J l --
2.
CD
CO
co
'
' ' ' '
: ! : i i : : : j IT
-90
0.1 1 10 100
Frequency (Hz)
Figure 4.7: Sideslip Angle / Steer Angle Frequency Response, V= 100 km/hr
3.0E-04
"& 2.0E-04
CD
"I 1.0E-04
CD
0.0E+00
0.1 10 100
Frequency (Hz)
CD
2_
CD -45
CO
CO
-90
0.1 10 100
Frequency (Hz)
Figure 4.8: Sideslip Angle /Aero Side Force Frequency Response, V = 100 km/hr
61
CD
CD
2.
C
co
O
0.1 1 10 100
Frequency (Hz)
CD
-
CD -45
CO
CO
-90
0.1 1 10 100
Frequency (Hz)
Figure 4.9: Sideslip Angle /Road Side Slope Frequency Response, V = 100 km/hr
0.20
CD
0.15
-52
co 0.10
CO 0.05
0.00
0.1 10 100
Frequency (Hz)
CD
2,
CD __ _ _ _ _,_ _ _
T
--,-
-, -r -i -i "I -i -^c
- - -
r
-
p r i -i i i-i-|------|---|---i-T-i-r-TT
-45
CO
CO
-90
0.1 1 10 100
Frequency (Hz)
Figure 4.10: Yaw Velocity / Steer Angle Frequency Response, V = 100 km/hr
62
8.0E-06
| 6.0E-06
rT^~
- - ---,---
-,- -
j
-
i t i ----n---i~~T-T-i-rTTr~___in i i t
2. 4.0E-06
O 2.0E-06
0.0E+00
0.1 10 100
Frequency (Hz)
180
O) 135
CD
2.
CD 90
CO
CO
\-
45
0
0.1 1 10 100
Frequency (Hz)
Figure 4.11: Yaw Velocity /Aero Side Force Frequency Response, V-100 km/hr
2.5E-04
2.0E-04
.g
^
CO
1.5E-04
1.0E-04 r
O 5.0E-05
0.0E+00
0.1 10 100
Frequency (Hz)
-45
CD
TO
CD -90
CO
CO
-135
-180
0.1 10 100
Frequency (Hz)
Figure 4.12: Yaw Velocity /Road Side Slope Frequency Response, V = 100 km/hr
63
0.4
^m^
O)
CD
0.3
- - - - - - - - - - - \~ - - _ _ -i_ ~ - - - - - - - - - - - i~
1 r -\ -\ r t tt ~^^t ~i-n-i-rTTr 1 r -i -i r t t
T3
O)
CD
2, 0.2 ._____,__
_r __! i | -r ~r i -----|---r
-
-|
-
i
- i^W"
ttt
~ ~ ~ - -,-- -
r
-
-i
-
-j
- |-
r t t
C
co
O 0.1
0.0
0.1 10 100
Frequency (Hz)
CD
TO
CD -
-45 i t--i--|-|-t-]-it-----|----|--i ^ci r n -i -i -}

CO
CO
-90
0.1 10 100
Frequency (Hz)
Figure 4.13: Sideslip Angle /Steer Angle Frequency Response, V = 30 km/hr
0.3
,, .
O)
CD
0.2
-- ----I---
+ __l__l_l_
+ _l_l + _ _ _ ,_ _ _ I-
-f^l^- -
1 -
+ -I -I -I -I _ - _
--t----f__l__l_H
CD
D
-H^
c
"co 0.1
0.0
0.1 1 10 100
Frequency (Hz)
90
CD
45
2,
CD 0
CO
CO
-45
-90
0.1 10 100
Frequency (Hz)
Figure 4.14: Sideslip Angle / Steer Angle Frequency Response, V = 49.84 km/hr
64
4.5.27 Simulation
Another tool that is very useful in analyzing vehicle lateral dynamics is numerical
simulation. Simulation is done by integrating the differential equations of motion with

respect to time and can be used to predict the response of the vehicle to arbitrary control and
disturbance inputs. Non-linearities are generally much easier to handle with numerical
simulation than with the analytical techniques used up to this point in this chapter.
To maintain consistency with the non-linear model simulation which follows in
Section 4.6, the lateral velocity v is used as a state variable for the linear model simulation
instead of the sideslip angle (3. Since the equations of motion were originally derived in
terms of yaw velocity and lateral velocity and then simplified to be in terms of yaw velocity
and sideslip angle, the model has been returned to its original, more general, form.
The equations of motion in their general, non-linear 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)
bFyr ~
(c ~
a^Fya
=
IJ
aFyf
Expressions for the tire slip angles, tire lateral forces, and gravitational side force
are derived in Section 4.5.3 and Section 4.5.4 using the small angle assumption and are
repeated here for convenience.
v + ar
af
=
8
u
, (4.19)
v-br
ar=
Fyf=Cfaf
Fy=mgQ (4.23)
65
Outputs from the simulation are time histories of lateral velocity, yaw velocity,
vehicle sideslip angle, tire slip angles, and lateral acceleration. Inputs can be a step steer,
ramp step steer, ramp square steer, sine steer, step aerodynamic side force, or step road
side slope.
The simulation is implemented in the MATLAB script DOF2LSim.m, which is
listed in Appendix C.6. Integration of the differential equations of motion is done using the
built-in MATLAB function ode23, which uses second and third order Runge-Kutta
formulas.34
The function ode23 returns the state variables v and r over the time interval
specified for the simulation.
At each 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 by the function
SteerAngle.m, which is listed in Appendix C.4, based upon the current time, the type of
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 of measured steer input data
facilitates experimental validation of the model.
After the steer angle is calculated, the tire slip angles are calculated from the current
values of 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
v =
Fyf+Fyr+Fya+Fyg -
ur
m
(4-88>
.
aFyf-bFyr-(c-a)Fya
r =
66
Eq. (4.88) are obtained by solving Eq. (4.87) for v and r. A listing of DOF2LDE.m is
provided in Appendix C.7.
To illustrate the effect of forward velocity on response, simulations are performed
at low speed (30 km/hr), at the tangent speed (49.84 km/hr), at normal highway speed (100
km/hr), and at high speed (150 km/hr). The simulations are run until steady-state is
reached. Initial conditions for the simulations are zero. Lateral velocity, yaw velocity,
sideslip angle, front tire slip angle, rear tire slip angle, and lateral acceleration are plotted
for each input studied.
1 1
The inputs used for the simulation are a step steer, a ramp step steer with a
1
ramp time of 0.2 sec, a ramp square steer with a ramp time of 0.2 sec and a dwell time
1
of 1.0 sec, a sine steer with a period of 1 sec, a 10000 N step aerodynamic side force,
1
and a step road side slope. The ramp step steer, ramp square steer, and sine steer inputs
are shown in Figure 4.15. The steer input is a positive steer angle, indicating a right rum.
Ramp Step Steer Input Ramp Square Steer Input

CD CD
;o
1.0 S 1.0
<
CD
CD
I
CO
0.5
0.0
Time
1
<
CD
CD
0.5
55 0.0
El 1 2
(s) Time (s)
Sine Steer Input
O) 1.0
CD
2,
<D
D)
c 0.0
<
CD
CD
OT -1-0
0 1 2 3
Time (s)
Figure 4.15: Simulation Steer Angle Inputs
67
The aerodynamic side force input is applied in the positive y-direction, and the road side
slope is positive, which means that the road slopes down to the right.
Simulation results for the step steer input are plotted in Figure 4. 16 through
Figure 4.21. The response times increase with forward speed. The steady-state lateral
velocity and sideslip angle are positive below the tangent speed, zero at the tangent speed,
and negative above the tangent speed. This agrees with the definition of the tangent speed
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
values. This is a result of the system zero being positive when the forward speed is greater
than the tangent speed. As explained in Section 4.5.25 a system with a positive zero is a
nonminimum-phase system and typically exhibits the type of response shown here,
step
initially in the opposite direction to the steady-state value. The front tire slip angles show
1
response similar to the sideslip angle, but with initial values of due to the steer.
step
The lateral acceleration has a non-zero initial value due to the rate of change of lateral
velocity when the step steer occurs. The steady-state values of yaw velocity,
sideslip angle,
front tire slip angle, rear tire slip angle, and lateral acceleration agree with the steady-state
response gains presented in Table 4.3. In addition, the steady-state sideslip angles and yaw
velocities agree with the values approached at low frequency in the frequency response
plots of Figure 4.7, Figure 4.10, Figure 4.13, and Figure 4.14.
Ramp step steer simulation results are plotted in Figure 4.22 through Figure 4.27.
The ramp step steer response is similar to the step steer response and lags it slightly as
expected. The steady-state values are identical to those of 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 steer.
Unlike the step input, the front tire slip angle and lateral
with acceleration are initially zero
68
for the ramp step input. At 30 km/hr and 49.84 km/hr there are peaks in the front tire slip
angle response at 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
Figure 4.33. The ramp square steer response is identical to the ramp step response up until
the time that the steer input is ramped back down to zero. At the lower speeds the responses
reach steady-state before the ramp down. However, at 150 km/hr the ramp down occurs
before steady-state has been reached. As with the ramp step response the front tire slip
angles experience overshoot at 30 km/hr and 49.84 km/hr as the ramp up is completed.
The sine steer results are shown in Figure 4.34 through Figure 4.39. The sine steer
input had a frequency of 1 Hz. From visual inspection of the plots it is seen that the steady-
state yaw velocity and sideslip angle gains and phases agree with those obtained for 1 Hz
from the frequency response in Figure 4.7, Figure 4.10, Figure 4.13, and Figure 4.14.
The response amplitude increases with forward velocity in all cases except the lateral
velocity and sideslip angle. As expected from the definition of tangent speed, at 49.84
km/hr the lateral velocity and sideslip angle amplitudes are less than those at 30 km/hr.
Results from the aerodynamic side force step input simulation are provided in
Figure 4.40 through Figure 4.45. The magnitude of the responses increases with forward
velocity. Since the center of aerodynamic pressure is located behind the neutral steer 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
steady-state yaw velocity and sideslip angle at 100 km/hr agree with the frequency response
gains of Figure 4.8 and Figure 4. 1 1 as the input frequency approaches zero. The sideslip
angles, front tire slip angles, rear tire slip angles, and lateral acceleration curves have higher
slopes at lower speeds indicating that the response is faster at lower speeds.
69
Road side slope step input results are plotted in Figure 4.46 through Figure 4.51.
The lateral velocity, yaw velocity, and lateral acceleration responses increase with forward
tire slip decrease.

velocity, while the
sideslip angle, front tire slip angle, and rear angle
Again, the steady-state yaw velocity, sideslip angle, front tire slip angle, rear tire slip angle,
and lateral acceleration at 100 km/hr obtained with 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
greatly simplifies the vehicle system, much can be learned about road vehicle lateral
dynamics through its study. The effects of changing vehicle and tire parameters on system
response can quickly be studied. Powerful linear analysis techniques based on system
transfer functions can be readily applied to the vehicle model to gain significant insight into
system behavior. The results 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
0.5
V = 30 km/hr
0.0
V = 49.84 km/hr
-0.5
_
I V= 100 km/hr
-1.0
o
o
CD
co
CD
CO
J
-2.0
-2.5
V= 150 km/hr
-3.0
0.5 1 1.5
Time (s)
Figure 4.16: Linear Step Steer Lateral Velocity Response
0.30
'
V 150 km/hr
> =
0.25
0.20
/i
f0
8 -15
/^ V =
100 km/hr
CD
>
o.io '//^ V = 49.84 km/hr
ifs^~
0.05 V = 30 km/hr
0.00
J
0.5 1 1.5
Time (s)
Figure 4.17: Linear Step Steer Yaw Velocity Response
71
0.5 ,
,
V = 30 km/hr
0.0
V = 49.84 km/hr
-0.5
C7> -1.0
CD
T7
CD -1.5
TO
C V = 100 km/hr
<
Q. -2.0
CO
CD
Tl
-2.5
CO
-3.0
-3.5
V =
150 km/hr
1
-4.0
0.5 1.5
Time (s)
Figure 4.18: Linear Step Steer Sideslip Angle Response
0.0
V = 30 km/hr
-0.5
V = 49.84 km/hr
-1.0
^-s.
O)
CD
2. -1.S
CD
TO
-2.0
<
Q. V= 100 km/hr
CO
-?.5
a)
H
H-*
r -3.0
o
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
0.0
V = 30 km/hr
-0.5
V = 49.84 km/hr
-1.0
^^
TO
0)
2, -l.b
CD
TO
C
-2.0
<
Q. V = 100 km/hr
CO
-?.S
CD
1-
-3.0
m
CD
DC
-3.5
-4.0
V = 150 km/hr
-4.5
0.5 1.5
Time (s)
Figure 4.20: Linear Step Steer Rear Tire Slip Angle Response
1.4
V = 150 km/hr
1.2
1.0
c
o
ffl 0.
OJ
CD
O
O
< 0. V= 100 km/hr
s
CD
CO
-i
0.4 -
0.2 V = 49.84
km/hr" "
V = 30 km/hr
1
0.0
0.5 1 1.5
Time (s)
Figure 4.21: Linear Step Steer Lateral Acceleration Response
73
0.5
V = 30 km/hr
0.0
V = 49.84 km/hr
_
"-5 \ ""V
E, V= 100 km/hr
& -1.0
o
o
CD
i "1-5
CD
I
co
J
-2.0
-2.5
V = 150 km/hr
..
-3.0
0.5 1.5
Time (s)
Figure 4.22: Linear Ramp Step Steer Lateral Velocity Response
0.30
,
V= 150 km/hr
s^
0.25
0.20
/f
T3
V= 100 km/hr
0.15
l/^
8
If
CD
>
0.10
V =
49.84 km/hr
r
0.05 V = 30 km/hr
0.00
/ !
0.5 1.5
Time (s)
Figure 4.23: Linear Ramp Step Steer Yaw Velocity Response
74
TO
CD
TO
<
Q.
W
CD
g
CO
-3.0
0.5 1.5
Time (s)
Figure 4.24: Linear Ramp Step Steer Sideslip Angle Response
0.0
V = 30 km/hr
-0.5
V = 49.84 km/hr
-1.0
^-H^
TO
CD
2, -1.S
CD
TO
X ^
C ^^^^
-2.0
<
a. V= 100 km/hr
CO
-?.S
CD
i
1-
C -3.0
O
-3.5
-4.0
V = 150 km/hr
'
1 1
-4.5
0.5 1.5
Time (s)
Figure 4.25: Linear Ramp Step Steer Front Tire Slip Angle Response
75
0.0
V = 30 km/hr
-0.5
V = 49.84 km/hr
-1.0
^-^
TO
a)
2, -l.b
CD
TO
-2.0
<
a. V = 100 km/hr
CO -2.5
a>
H
-3.0
CD
oc
-3.5
.;
>s
-4.0
V= 150 km/hr
-4.5
0.5 1.5
Time (s)
Figure 4.26: Linear Ramp Step Steer Rear Tire Slip Angle Response
1.4
V= 150 km/hr
1.2 -
1.0
c
o
]3 0. 8 -
CD
O
U
< 0. V= 100 km/hr
B
"5
0.4
0.2 V = 49.84 km/hr

"
V =
30 km/hr
^^^
1 1 1
0.0
0.5 1 1.5
Time (s)
Figure 4.27: Linear Ramp Step Steer Lateral Acceleration Response
76
0.5
>
V = 30 km/hr
0.0
"^V V = 49.84 km/hr
\ ^Ss*^^ ]/
i \ V= 100 km/hr
-1.0
u
o \
CD
I"'-5
CD
CO
J : \V = 150 km/hr
-2.0
-2.5
-3.0
0.5 1.5 2.5

Time (s)
Figure 4.28: Linear Ramp Square Steer Lateral Velocity Response
0.30 .
>^V = 150 km/hr \

0.25
0.20
7 \ \
l/? V = 100 km/hr A \
~
0.15
o
o
co
0.10 1
\/ V = 49.84 km/hr \
V
\
>-
Wf
^*
"^Nk \ \
0.05 /]:*"'
>^ .>
.1
0.00
-0.05
0.5 1.5 2.5

Time (s)
Figure 4.29: Linear Ramp Square Steer Yaw Velocity Response
77
0.5
' '
0.0
^Z-~~^__ V = 30 km/hr
N^
"^l ' V = 49.84 km/hr
-0.5
W. !
TO -1.0
CD
\ ^v^V =
100 km/hr J
CD -1.5
TO
c
<
V
Q. -2.0
CO
'
\
rn -2.5
\
-3.0
\. V = 150 km/hr
y
-
-3.5
-4.0
1
0 0.5 1 1.5 2 2.5
Time (s)
Figure 4.30: Linear Ramp Square Steer Sideslip Angle Response
0.5
0.0 c v oU Km/nr
s
x^~~
r*"~
' '
^
\^
-0.5
/
V ; V = 49.84 km/hr
TO
-1.0
\\.
"to -1-5
TO
<
\ ^SV = 100 km/hr /
9. -2.0
CO
V
S -2.5
\
P -3.0
; \ V = 150 km/hr /
-3.5
: \, I
-4.0
-4.5
0 0.5 1 1.5 2 2.5

Time (s)
Figure 4.31: Linear Ramp Square Steer Front Tire Slip Angle Response
78
0.5
0.0
'
/^
Ys^: ^S
-0.5
V V = 49.84 km/hr j /
TO
-1.0
T3
TO -1.5
C
< \ :^vV= 100 km/hr / \
/
a. -2.0
CO
2 -2.5
H
i
CO
-3.0
rr
~/~
-3.5 \ v- ibUKm/nr
-4.0
l .... i i
-4.5
0 0.5 1 1.5 2 2.5

Time (s)
Figure 4.32: Linear Ramp Square Steer Rear Tire Slip Angle Response
1.2
js^~\.
1.0
/ V = 150 km/hr V
~ 0.8
C
o
/ \ \
j 0.6
o
o Sv = 100 km/hr \ i \
< 0.4
2
co
0.2
J/ V = 49.84 km/nr ; x
\^;
X^
0.0
i i 1
-0.2
0.5 1.5 2.5

Time (s)
Figure 4.33: Linear Ramp Square Steer Lateral Acceleration Response
79
1.0
y\
0.5
V = 30 km/hn
r~\ \
~
0.0 >^-
-jf-<5
o
\\ )V = 49.84 km/hr V
>
2 -0.5
\ ;V= 100 km/hr \
CO
-1.0
>^V = 150 km/hr
-1.5
0 0.5 1 1.5 2 2.5

Time (s)
Figure 4.34: Linear 1 Hz Sine Steer Lateral Velocity Response
0.25
I"
0.20 v^\/ = 150 km/hr \ \
0.15 -
1/^sN = 100 km/hr: /nA fss\
W 0.10 '
//SV ==
\
w
49.84 km/hr n \\
Is-^ T n
T^x \\
-
30 km/hr -
r\\\ u
r\\\
0.05 ^dr^^sAW i^r\\\\
o
o
0.00
-0.05
-
-0.10
-0.15
i i 1 i i
-0.20
0.5 1.5 2.5

Time (s)
Figure 4.35: Linear 1 Hz Sine Steer Yaw Velocity Response
80
1.5
i f\. y \
1.0
1 \ '
1
0.5
TO
V = 30 km/hr
fl VV /
0.0
TO
C
<
a-
\\ : V = 49.84 km/hr \ |
-0.5
CO
g
CO
-1.0
W/ \ V= 100 km/hr
: W \ /
-1.5
V= 150 km/hr
i i i i
-2.0
0 0.5 1 1.5 2 2.5

Time (s)
Figure 4.36: Linear I Hz Sine Steer Sideslip Angle Response
1.0
0.5 V = 49.84 km/hr^y/ \V !

sT*l
TO
v = 30
km/h|>7/v\ \\ yy/h\
0.0
TO
C
<
-Vs ^ " "
-0.5
'/ / ~^\\ '/ /
CO
p
< -1.0
Yv = 100 km/hr :
LL
-1.5
V = 150 km/hr
j 1 1 i
-2.0
0 0.5 1 1.5 2 2.5

Time (s)
Figure 4.37: Linear 1 Hz Sine Steer Front Tire Slip Angle Response
81
1.5
1
__
1.0
0.5 -
TO
V = 30 km/hr^^-f \ \\
0.0
TO
C
<
-0.5
\ V == 49.84 km/nr \\
\
CO
2
F -1.0
// !
CO
DC \ V = 100 km/hr \
-1.5
\ M
-2.0
V= 150 km/hr
-2.5
0.5 1.5 2.5

Time (s)
Figure 4.38: Linear 1 Hz Sine Steer Rear Tire Slip Angle Response
0.5
f\
0.4
J \v = 150 km/hr
0.3
// \V= 100 km/hr L
f^
3
0.2 '
Z
o
1 o.i "^^\v = 49,84 km/hr

"
A
'/its.' ~/Tr<~
\-\
3 o.o
V = 30 km/hr^^VVy; //
N\l^-
I
co
"0.1 h
\ 1 / /
-0.2
VI y
-0.3
.1 1 1
-0.4
0.5 1.5 2.5

Time (s)
Figure 4.39: Linear 1 Hz Sine Steer Lateral Acceleration Response
82
3.0
V = 150 km/hr
2.5
',
-52 2.0
E
o
"
-2
CD
1 5
"J
V = 100 km/hr
> i
2 i
* 1.0
V = 49.84 km/hr
0.5
:
V = 30 km/hr
*
1 1 1
0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.40: Linear Step Aero Side Force Lateral Velocity Response
0.00
-0.02
JO -0.04
T3
co
'o -0.06
o
>
> -0.08
-0.10
-0.12
0 0.2 0.4 0.6 _

0.8 1 1.2 1.4 1.6
Figure 4.41: Linear Step Aero Side Force Yaw Velocity Response
83
4.0
V = 150 km/hr
r
3.5
i
3.0 V = 100 km/hr
TO
2.5
Jr V =
49.84, km/hr
TO
C
2.0 V = 30 km/hr
<
Jo i
1.5 -
CO
1.0 i
0.5
i
0.0 t i 1 i
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time (s)
Figure 4.42: Linear Step Aero Side Force Sideslip Angle Response
3.5
V= 150 km/hr
3.0
V = 100 km/hr
<-D
CD
V =
49.84 km/hr
|
1? 2.0
< V =
30 km/hr
f 1.5
p 1.0 -
0.5
V i i 1 1 J I i
0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.43: Linear Step Aero Side Force Front Tire Slip Angle Response
84
4.0 i
,V
= 150 km/hr
3.5
i i
!V = 100 km/hr
^ 3.0
TO
!V = 49.84 km/hr
2.5
TO
C
< iV = 30 km/hr
J2- 2.0
-- - -/-/-
co
2
F 1.5
CO
DC
1.0 -
0.5
i i i \ 1
0.0
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.44: Linear Step Aero Side Force Rear Tire Slip Angle Response
0.6
0.4
0.2
o
\
V = 30 krrl/hr
0.0
o
o
< V = 49.84 km/hr
2
3
CO
-o-2
V= 100 km/hr
-0.4
V= 150 km/hr
1 1 1 1
-0.6
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time (s)
Figure 4.45: Linear Step Aero Side Force Lateral Acceleration Response
85
0.05
V = 150 km/hr
0.04 -
jo
~
0.03
+-*
V= 100 km/hr
'a
o
i
>
/j
2 0.02
.
V = 49.84 km/hr
0.01 l~r
V = 30 km/hr
i i i 1
0.00
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.46: Linear Step Road Side Slope Lateral Velocity Response
3.5E-04
1
3.0E-04
V= 150 km/hr
2.5E-04 -
V= 100 km/hr
T3
2
~
2.0E-04
+-*
"o
o
S 1.5E-04 -
co
>-
V = 49.84\ km/hr
1 .0E-04
V = 30 km/hr
1
5.0E-05
i i_. i i
0.0E+00
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.47: Linear Step Road Side Slope Yaw Velocity Response
86
0.07
V =
30 km/hr ; ; ;
j/0^ t . ii . i I
0.06
km/hr-
V = 100
V = 150 km/hr
/ Ai = 49.84
knrvhr^-'
0.05 -
TO
0.04 -
TO
c
<
0.03
CO
T3
CO
0.02
0.01
0.00 L i i i i
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Time (s)
Figure 4.48: Linear Step Road Side Slope Sideslip Angle Response
0.07
V= 150 km/hr
0.00
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.49: Linear Step Road Side Slope Front Tire Slip Angle Response
87
0.07
0.06 -
V =
100 km/hr V = 150 km/hr
TO 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 1 1.2 1.4 1.6
Time (s)
Figure 4.50: Linear Step Road Side Slope Rear Tire Slip Angle Response
0.018
0.016
0.014
S 0.012
c
o
2 0.010
| 0.008
2
%
CO
0.006
-"
W = 30 km/hr
" ~ "
\ v :
\\v == 49.84 km/hr
0.004
vVcV = 100lcm/hr !
0.002 h ~V ^W '"^*H^^ V = 150 km/hr
V^ '
^^^. -i
'
0.000
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Time (s)
Figure 4.51: Linear Step Road Side Slope Lateral Acceleration Response
88
4.6 Non-Linear Model
As previously noted the linear vehicle model is valid for lateral accelerations up to
approximately 0.35 g. This is primarily a result of tire lateral force being linear with respect
to slip angle at small slip angles, and hence small lateral accelerations. Beyond 0.35 g when
higher slip angles are being attained a non-linear tire model is usually necessary to
accurately predict tire lateral forces.
As discussed in Chapter 3 many models of tire behavior exist. The tire model
chosen for this work is called tire data nondimensionalization and was originated by Hugo
Radt. This tire model is discussed in detail in Section 3.4.
In this section the equations describing the non-linear two degree-of-freedom

vehicle model are presented. Simulation of 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 + F + F =
m(v + ur)
(4.9)
aF^ cos 5 bFyr (c I^r
- - -
=
a)Fya
Expressions for the tire slip angles are derived in Section 4.4.
atf=^f)-i
, atan-
8 | -
(4'14>
v-M
,
a, =atan
u j
The tire lateral force is given by the following expressions as described in
Section 3.4 and repeated here for convenience. From these equations the tire lateral force F
can be calculated based upon the tire vertical load Fz and the tire slip angle a.
89
CC=B3 +
C3FZ (3.3)
Hy=B5 +
C5Fz (3.5)
_=Qtan(a)
(3 6)
_ /
(l-E1)a+-
x_ FatanfB.a)
^-^- , ^x
y/ =
(3.10)
i
0 =
C, atan(fl^) (3.9)
Fy=DlSin(tJ) (3.8)
Fy=FuFz (3.11)
4.6.2 Simulation
Simulation of the non-linear two degree-of-freedom vehicle model is implemented
in the MATLAB script DOF2NLSim.m. This script is listed in Appendix C.8 and is very
similar to DOF2LSim.m which performs simulation of the linear model. As with the linear
simulation, the scripts DOF2Control.m, DOF2Param.m, and DOF2DependParam.m are
called at the beginning of DOF2NLSim.m to set simulation, vehicle, and tire parameters.
The built-in MATLAB function ode23 is used again to integrate the differential equations of
motion which are contained in the function DOF2NLDE.m. This function calculates the
state derivatives v and r based upon 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
m
(4'89)
(c
laF^ cos(5) 2bFyr
- - -
=
a)Fya
=
r
90
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 are calculated by the
MATLAB function NLTire.m which is listed in Appendix A.3. This script takes the tire
vertical load and slip angle as arguments and returns the tire lateral force. Note that with the
non-linear tire model the lateral forces

F^ and F,r are for only one tire, while for
simplification in the linear model they are for two tires. Thus here they are multiplied by the
factor of two in Eq. (4.89) to get the lateral forces for two tires. NLTire.m is called at each
time step by DOF2NLDE.m, which also calls the function SteerAngle.m to calculate the
instantaneous steer angle.
For comparison with the linear model, simulation of the non-linear model is
performed for the step steer input and 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, and 150 km/hr. The step steer and ramp square steer inputs are identical to those
used for the linear model, having a magnitude of 1. Tire parameters for the non-linear 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 model agree at small slip angles. Vehicle parameters are identical to those used in
the linear simulation. Results from the simulations are provided in Figure 4.52 through
Figure 4.63. Included on these plots as dashed lines are the linear simulation results for
comparison.
Lateral velocity, yaw velocity, sideslip angle, front tire slip angle, rear tire slip
angle, and lateral acceleration results for both the non-linear and linear simulations are
presented in Figure 4.52 through Figure 4.57 for the step steer input. The linear and non-
91
linear lateral acceleration results agree within 1% over the complete duration of the
simulation for forward velocities of 30 km/hr and 49.84 km/hr. These speeds correspond
to steady-state lateral accelerations of 0.05 g and 0.14 g respectively. At 100 km/hr, which
produces a 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 1.0% once steady-
state is reached. At this speed the tire slip angles reach slightly more than 2. At these slip
angles the tire lateral force versus slip angle curve is still very nearly a straight line. Thus
for this tire and vehicle the linear tire model is reasonably accurate and useful for lateral
accelerations in excess of 0.5 g. However, at 150 km/hr the linear model lateral
accelerations exceed those of the non-linear model by over 27%. At this speed the non
linear model predicts a steady-state lateral acceleration of 0.95 g while the linear model
6
predicts 1.20 g. The tire slip angles have exceeded where the lateral force versus slip
angle curve is approaching its peak. The linear tire approximation is not sufficiently
accurate at slip angles of this magnitude.
At high speeds the linear model predicts that the magnitudes of lateral velocities,
sideslip angles, and tire slip angles are below those that the non-linear model predicts and
that the yaw velocities and lateral accelerations are above those of the non-linear model. The
linear model also predicts faster response than the non-linear model. At 150 km/hr the non
linear model predicts overshoot in all of 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 non-linear and linear
models for this input are similar to those of the step steer input. The two models agree very
well for forward velocities of 30 km/hr and 49.84 km/hr. As with the step steer input, at
the higher speeds the linear model predicts peak lateral velocity, sideslip angle, and tire slip
angle magnitudes below those of the non-linear model and predicts peak yaw velocities and
92
lateral accelerations above those of the non-linear model. Differences in peak lateral
acceleration reach 25%. Again, the linear model predicts faster response than the non-linear
model. In particular, at 150 km/hr the response of the non-linear model lags the linear
model by approximately 0.5 seconds after the steer input is ramped back down to zero.
Here differences between the linear and non-linear lateral accelerations reach nearly 100%.
Comparison of the linear and non-linear model simulations shows that at low slip
angles and lateral accelerations the linear vehicle and tire models can produce results
comparable to the non-linear model. Even for the 100 km/hr case where slip angles exceed
2
and the lateral acceleration reaches 0.55 g the linear model produces results that are
acceptable for most engineering purposes. When tire slip angles and lateral accelerations
become high it is necessary to have a non-linear 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
performance characteristics over a wide variety of operating conditions.
93
0.5
V = 30 km/hr
0.0
V = 49.84 km/hr
-0.5
'
-1.0
[ V = 100 km/hr
CO
E Linear
-1.5
Norl-Linear
o
o
-2.0
>
\s
2 -2.5
\ *s
Linear
To
-3.0
V= 150 km/hr
-3.5 .
Non-Linear
-4.0
>
-4.5
12 3 4
Time (s)
Figure 4.52: Non-Linear Step Steer Lateral Velocity Response
0.30
/ \ Linear
.
\/ 1 <^n km/hr
0.25
If ! Non-Linear
^^^
ij
0.20
;
V= 100 km/hr
co ;
o 0.15
o
>
V =
49.84 km/hr
> -10 h
0.05
i i i
0.00
Time (s)
Figure 4.53: Non-Linear Step Steer Yaw Velocity Response
94
1.0
V = 30 km/hr
1 *
0.0
V = 49.84 km/hr
-
-1.0
TO \ ^^^ ^ _
' V = 100 km/hr
, ,
-
-2.0
'
TO
C
;
<
-3.0
V.j
CO
i Linear
CO
-4.0
V= 150 km/hr
t i
-5.0
'
; Non-Linear
^-'
-6.0
0 12 3 4
Time (s)
Figure 4.54: Non-Linear Step Steer Sideslip Angle Response
1.0
V = 30 km/hr
0.0 -
1 '
, p
V = 49.84 km/hr
-1.0
1 '
TO
-2.0
TO
C V =
100 km/hr
<
Q. -3.0
CO
r- -4.0
\ *"*
*-
Linear
4_d
_
c
o
- -
u- V
---
-5.0 -, = 150 km/hr
-6.0
^ , Non-Linear
1 1 1
-7.0
0 12 3 4
Time (s)
Figure 4.55: Non-Linear Step Steer Front Tire Slip Angle Response
95
1.0
0.0 V = 30 km/hr
i
v"^
! 1
'. V = 49.84 km/hr

-.
-1-0 i i
TO
-2.0 **"'
CD \ '
\
"~" " ~~ r
TO
C
V = 100 km/hr
<
j?- -3.0
co
2
p -4.0
V^*-^ . Linear
CO
c x.
-5.0 -
1
v. , -v = i50km/nr
i
,
"~
_^
-
Non-Linear
-6.0
-7.0
0 12 3 4
Time (s)
Figure 4.56: Non-Linear Step Steer Rear Tire Slip Angle Response
1.4
Linear
1.2
/ V= 150 km/hr
/ ,
1.0 -
'
/ ^
C
o ; Non-Linear
J-j/- L
0.8
. . .
nS
//
o
.
< 0.6 -If ' --i-
'
__,
2 V= 100 km/hr
S
3 0.4
0.2 V = 49.84 km/hr

" '
i^ V = 30 km/hr
i
0.0
12 3 4
Time (s)
Figure 4.57: Non-Linear Step Steer Lateral Acceleration Response
96
0.5
;V = 30 km/hr
0.0
"^vV = 49.84 km/hr
-0.5 ^^ . 4
"co"
y.
/ /
-1.0
V 100 km/hr
/ :
=
b
'*~^r
f :
-1.5
/ /
o
o 7 / ;
> -2.0
m
\n ; Linear ,
co -2.5
--
\_V,H /
-3.0
V = 150 km/hr
; \^ /
-3.5
Non-Linear
i i i. 1
-4.0
0 0.5 1 1.5 2 2.5 3 3.5

Time (s)
Figure 4.58: Non-Linear Ramp Square Steer Lateral Velocity Response
0.30
0.25
0.20
co
0.15
o
o
g 0.10
co
0.05
0.00
-0.05
0 0.5 1 1.5 2 2.5 3 3.5

Time (s)
Figure 4.59: Non-Linear Ramp Square Steer Yaw Velocity Response
97
1.0
V = 30 km/hr
^
0.0 <S
^VvV 49.84 km/hr
'
"""
r
=
^'"*^ '-
-1.0
TO
'
-2.0
L - - -
'
/
TO
< V /
-3.0
CO
\\ Linear / /
g
CO
-4.0
Nv /
v= 150 km/hr
-5.0 '
1
1 Non-Linear
^__^j
-6.0
0 0.5 1 1.5 2 2.5 3 3.5

Time (s)
Figure 4.60: Non-Linear Ramp Square Steer Sideslip Angle Response
1.0
0.0
\/ = 3D jh<V-. :_^_
'/ ^r^~\ ^
y
V V = 49.84 km/hr .^
>"
X
to -i.o
\Xv = 100 km/hr /^ / /

c -2.0
<
Q.
CO
-3.0
:\ j / / :
: \ / /
Y_ Linear 1 ; / ;
2 -4.0
.
V
\ /
s^/ :
/
/ : ;
Nw r^\j= 150 km/hr

-5.0
Non-Linear
i i i i i 1l
-6.0
0 0.5 1 1.5 2 2.5 3 3.5 4

Time (s)
Figure 4.61: Non-Linear Ramp Square Steer Front Tire Slip Angle Response
98
1.0
V = 30 km/hr ; ;
0.0
^-_-
- -
^^sj_'
"^y
\ V 49.84 km/hr /--
ffr
=
TO -
1.0 -
\\v = 100 km/hr /

/ /
? -2.0
\^<^\ */
<
Q.
X ^ : / : 7
CO .,
\\ _, /
/ :/
,/
-3.0
CO
\\ \N'
Linear /
1
f
-4-0
/
V = 150 km/hr
-5.0
/
; Non-Lin ear
1 1
-6.0
0 0.5 1 1.5 2 2.5 3 3.5 4

Time (s)
Figure 4.62: Non-Linear Ramp Square Steer Rear Tire Slip Angle Response
1.2
1.0 h
0.8
3
c
o
S 0.6
o
o
< 0.4 -
2
15
0.2
0.0
-0.2
0 0.5 1 1.5 2 2.5 3 3.5 4

Time (s)
Figure 4.63: Non-Linear Ramp Square Steer Lateral Acceleration Response
99
5 Conclusion
In the early part of this century as the top speeds of automobiles increased vehicle
dynamics became an important consideration for engineers. Manufacturers had to meet
higher and higher standards of performance, particularly in the areas of safety and comfort.
Mathematical modeling of vehicle dynamics has become an excellent way for engineers to
study vehicle behavior and to reduce the time and cost to develop vehicles which meet
performance goals. There is a great deal of literature on the topic of vehicle dynamics.
Lateral vehicle dynamics in particular has been a topic of great interest due to its
relationship with safety. Two areas of focus in the literature concerning the modeling of
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 of road
vehicles, sufficiently accurate representation of tire mechanics is essential for vehicle
models.
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 of the tire slip angle. The non-linear tire model utilized a
method called tire data nondimensionalization to predict 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 can then be predicted as a non-linear function
of both vertical load and slip angle.
In Chapter 4 the equations of motion for a two degree-of-freedom vehicle model
were derived from basic principles of Newtonian mechanics. The model was then
developed in two forms, linear and non-linear. The linear vehicle model utilized the linear
tire model. Transfer functions were written 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
tangent speed. Expressions for transient response characteristics such as natural 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 steer input became positive. The effect
of this on the vehicle was seen clearly in the frequency response and in the simulation. For
some combinations of speed and input magnitude the linear model predicted lateral
accelerations higher than were actually possible due to the assumption of linear tire
behavior. In all cases tested the steady-state response gains, frequency response, and
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 angles and lateral accelerations. Comparison with the linear model showed that for
the vehicle studied the linear model was reasonably accurate for most engineering purposes
2
lateral accelerations of 0.5g. It was seen that for accurate
up to slip angles of and
modeling of vehicle response at high slip angles and lateral accelerations a non-linear
representation of the tires was necessary.
101
References
1 .
Gillespie, Thomas D. Fundamentals of Vehicle Dynamics. Warrendale, PA: SAE,
1992.
Automobile."
2 .
Lanchester, F. William. "Some Reflections Peculiar to the Design of an
Proceedings of the Institution of Mechanical Engineers, Vol. 2, 1908, p. 187-257.
Handling."
3 .
Olley, Maurice. "Suspension and Detroit, MI: Chevrolet Engineering
Center, 1937.
Suspensions."
4. Olley, Maurice. "Notes on Detroit, MI: Chevrolet Engineering Center,
1961.
n."
5 .
Olley, Maurice. "Suspensions Notes Detroit, MI: Chevrolet Engineering Center,
1966.
6 .
Segal, Leonard. "Theoretical Prediction and Experimental Substantiation of the
Control."
Response of the Automobile to Steering Proceedings of the Automobile
Division of the Institution of Mechanical Engineers, 1956-1957.
7 .
Whitcomb, David W. andWilliam F. Milliken. "Design Implications of a General
Control."
Theory of Automobile Stability and Proceedings of the Automobile Division

of the Institution of Mechanical Engineers, 1956-1957.
8 .
Bastow, D. and G Howard. Car Suspension and Handling. Warrendale, PA: SAE,
1993.
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, 1991.
11. 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. and Doug L. Milliken. Race Car Vehicle Dynamics. Warrendale,
PA: SAE, 1995.
14. Mola, Simone. Fundamentals of Vehicle Dynamics, Detroit, MI: General Motors
Institute, 1969.
15. Reimpell, Jornsen and Helmut Stall. The Automotive Chassis: Engineering Principles.
Warrendale, PA: SAE, 1996.
16. Taborek, Jaroslav J. Mechanics of Vehicles. Cleveland, OH: Penton, 1957.
102
References
17. Wong, Jo Yung. Theory of Ground Vehicles . New York: John Wiley & Sons, Inc.,
1993.
18. Bundorf, R.T. and R.L. Leffert. 'The Cornering Compliance Concept for Description
Properties."
of Vehicle Directional Control SAE Paper No. 760713, Oct. 1976.
19. Allen, R. Wade, Theodore J. Rosenthal, and Henry T. Szostak. "Steady State and
Handling."
Transient Analysis of Ground Vehicle SAE Paper No. 870495, 1987.
20. Heydinger, Gary J. "Improved Simulation and Validation of Road Vehicle Handling
Dynamics."
Ph.D. Dissertation, Ohio State University, Columbus, Ohio, 1990.
21 .
Xia, Xunmao. "A Nonlinear Analysis of Closed Loop Driver/Vehicle Performance
Control."
with Four Wheel Ph.D. Dissertation, Department of Mechanical
Steering
Engineering, Clemson University, Clemson, SC, Dec. 1990.
22. Trom, J.D., J.L. Lopex, and M.J. Vanderploeg. "Modeling of a Mid-Size Passenger
Program."
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
Formalisms."
Based on Multibody Vehicle System Dynamics, No. 14, 1985, p. 229-
263.
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
Slips."
for Vehicle Dynamic Simulations. Part 1: Pure International Journal of Vehicle
Design, Vol. 11, No. 6, 1990.
26. Bakker, Egbert, Lars Nyborg, and Hans B. Pacejka. "Tyre Modelling for Use in
Studies."
Vehicle Dynamics SAE Paper No. 870421, 1987.
Data."
27. Radt, Hugo S. and D.A. Glemming. "Normalization of Tire Force and Moment
Tire Science and Technology, Vol. 21, No. 2, Apr.-June 1993.
28. Allen, R. Wade, Raymond E. Magdaleno, Theodore J. Rosenthal, David H. Klyde,

and Jeffrey R. Hogue. 'Tire Modeling Requirements for Vehicle Dynamics
Simulation."
SAE Paper No. 950312, Feb. 1995.
Terminology."
29. Society of Automotive Engineers. "Vehicle Dynamics SAE J670e,
1976.
Data."
30. Radt, Hugo S. "An Efficient Method for Treating Race Tire Force-Moment
SAE Paper No. 942536, Dec. 1994.
31 .
Meriam, James L. andL. Glenn Kraige. Engineering Mechanics: Dynamics. New
York: John Wiley & Sons, 1992.
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 Curve Fitting of Tire Data to Magic Formula

%
% Finds parameters for Magic Formula curve fit of tire lateral force or
% aligning moment vs. slip angle data read from file TireSlip.dat
%
% Created 4/21/96
% J. Kiefer
% Initialization
clear all
ele;
% Load Data from File

load TireSlip.dat
t =
TireSlip(:,l) ;
y =
TireSlip(:,2) ;
% Find Curve Fit Parameters

xO =
[.7407 1.35 1.00 -0.5];
'MagicError'
x =
leastsq ( , xO, [] , [], t, y)
% Construct Fit Function

tl =
linspace(0/max(t) ,10) ;
psi = (l-x(4))*tl + x(4)/x(l)*atan(x(l)*tl);
theta =
x(2)*atan(x(l)*psi);
F =
x(3)*sin(theta) ;
% Plot Data and Fit Function
plot(tl, F, t, y, 'o')
(B=' C='
title
(['
Tire Data Magic Formula Fit num2str(x(l) ) ', num2str(x(2) ) . .
E='
',
D=-
num2str(x(3)) ', num2str (x(4) ) ')'])
' '
xlabel ( t )
ylabeK'y')
grid
105
Appendix ATire Model MATLAB Programs
A. 2 MagicError.m
function e =
MagicError(x, t, y)
%MagicError Error in Magic Formula Curve Fit
%
%e =
MagicError(x, t, y)
Calculates vector of errors of Magic Formula curve fit given parameters

x and data (t, y)
%
% Inputs:
Vector of curve fit parameters

% x(l) B
% x(2) c
% x(3) D
% x(4) E
% t Vector of independent data
% y Vector of dependent data
% Outputs:
% e Vector of errors between data and fit function
%
% Created 4/21/96
% J. Kiefer
psi =
(l-x(4))*t +
x(4)/x(l)*atan(x(l)*t);
theta =
x(2)*atan(x(l)*psi);
F =
x(3)*sin(theta);
e =
y
-
F;
106
Appendix ATire Model MATLAB Programs
A. 3 NLTire.m
function Fy =
NLTire(Fz, alpha)
%NLTire Nbn Linear Tire Model Lateral Force
%
%Fy =
NLTire(Fz, alpha)
%
% Calculates tire lateral force from inputs of tire vertical load and slip
% angle. Based on Radt's tire data nondimensionalization model and the
% Magic Formula model. Force is for one tire. Called by the function
% D0F2NLDE.m.
%
% Inputs:
% alpha Tire slip angle (rad)
% Fz Tire vertical load (N)
% Outputs:
% Fy Tire lateral force (N)
%
% Created 2/18/96
% J. Kiefer
global Bl CI Dl El B3 C3 B5 C5;
% Normalization Parameters
Cc =
B3 + C3*Fz; % N/deg/N Cornering coefficient
mu =
B5 + C5*Fz; % N/N Friction 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 -> Cf + Cr, Yr -> (a Cf -
b Cr)/V#
Yd -> -Cf, Nb -> a Cf -
b Cr, Nr -> (aA2 Cf +
bA2 Cr)/V, Nd -> -a Cf}
a Cf -
b Cr
{Yb -> Cf + Cr, Yr ->
, Yd ->
-Cf, Nb -> a Cf -
b Cr,
V
2 2
a Cf + b Cr
Nr ->
, Nd -> -(a Cf)}
V
Transformed Equations of Motion
A = {{s-Yb/(m V), l-Yr/(m V)},

{-Nb/Izz, s-Nr/Izz}};
MatrixForm [A]
Yb Yr
s
-
1
m V m V
Nb Nr
-(-
) -( ) + s
Izz Izz
Bl = {Yd/(m V), Nd/Izz};

MatrixForm [Bl]
Yd
m V
Nd
Izz
B2 = {l/(m V), (a-c)/Izz};

MatrixForm [B2]
m V
a -
c
Izz
108
B3 {g/V,
=
0};
MatrixForm[B3]
B4 = {l/(m V), (a-d)/Izz>;

MatrixForm [B4]
m V
a -
Izz
Transfer Function Denominator
Ds = Collect [Det [A] ,s]
Nb 2 Nr Yb Nr Yb Nb Yr
+ + (-(
s + s ) )
Izz Izz m V Izz m V Izz m V
Transfer Function Numerators
Nbd =
Collect [Det [Transpose [Rep lace Part [Transpose [A],Bl,l]]],s]
Nd Nr Yd s Yd Nd Yr
Izz Izz m V m V Izz m V
Nba = Collect [Det [Transpose [ReplacePart [

Transpose [A],B2,l]]],s]
a c Nr s a Yr c Yr
Izz Izz Izz m V m V Izz m V Izz m V
Nbt = Collect [Det [Transpose [ReplacePart [

Transpose [A],B3,l]]],s]
g Nr g s
-( ) +
Izz V V
109
Nrd = Collect [Det [Transpose [ReplacePart [

Transpose [A] , Bl, 2 ] ] ] , s]
Nd s Nd Yb Nb Yd
+
Izz Izz m V Izz m V
Nra =
Collect [Det [Transpose [ReplacePart [
Transpose [A],B2,2]]],s]
a c Nb a Yb c Yb
Izz Izz Izz m V Izz m V Izz m V
Nrt =
Collect [Det [Transpose [ReplacePart [
Transpose [A] , B3 , 2] ] ] , s]
g Nb
Izz V
Nrn = Collect [Det [Transpose [ReplacePart [

Transpose [A] ,B4,2] ] ] , s]
ad Nb a Yb d Yb
Transfer Functions
Sideslip Angle
Gbd = Nbd/Ds
Nd Nr Yd s Yd Nd Yr
Izz Izz m V m V Izz m V
+ s + + s (-( ) )
Gba = Nba/Ds
a c Nr s a Yr c Yr
Izz Izz Izz m V m V Izz m V Izz m V
+ s + + s (-( ) )
110
Gbt = Nbt/Ds
g Nr g s
( ) +
Izz V V
+ + (-(
s + s ) )
Yaw Velocity
Grd = Nrd/Ds
Nd s Nd Yb Nb Yd
+
Izz Izz m V Izz m V
+ s + + s (-( ) )
Gra = Nra/Ds
a c Nb a Yb c Yb
+ s + + s (-( ) )
Grt = Nrt/Ds
g Nb
Izz V ( + s + + s (-( ) ) )
Grn = Nrn/Ds
ad Nb a Yb d Yb
+ s + + s (-( ) )
111
Steady State Step-Input Response Gains
Sideslip Angle
Sbd =
Simplify [Limit [Gbd, s->0] ]
-(m Nd V) -
Nr Yd + Nd Yr
m Nb V + Nr Yb -
Nb Yr
Sba =
Simplify [Limit [Gba, s->0] ]
-Nr -amV+cmV+aYr-cYr
m Nb V + Nr Yb -
Nb Yr
Sbt =
Simplify [Limit [Gbt, s->0]]
g m Nr
_( )
m Nb V + Nr Yb -
Nb Yr
Yaw Velocity
Srd =
Simplify [Limit [Grd, s->0]]
-(Nd Yb) + Nb Yd
m Nb V + Nr Yb -
Nb Yr
Sra =
Simplify [Limit [Gra, s->0] ]
Nb -
a Yb + c Yb
m Nb V + Nr Yb -
Nb Yr
Srt =
Simplify [Limit [Grt, s->0] ]
g m Nb
m Nb V + Nr Yb -
Nb Yr
Srn =
Simplify [Limit [Grn, s->0] ]
Nb -
a Yb + d Yb
m Nb V + Nr Yb -
Nb Yr
Front Tire Slip Angle
Safd =
Simplify [Sbd + a/V Srd -
1]
2 2
(m Nb V + m Nd V + a Nd Yb + Nr V Yb -
a Nb Yd + Nr V Yd
Nb V Yr -
Nd V Yr) / (V (-(m Nb V) -
Nr Yb + Nb Yr) )
112
Safa =
Simplify [Sba + a/V Sra]
2 2 2
(- Yb-acYb-aVYr+
(a Nb) +NrV+amV -cmV +a
c V Yr) / (V (-(m Nb V) -
Nr Yb + Nb Yr) )
Saft =
Simplify [Sbt + a/V Srt]
g m (a Nb -
Nr V)
V (m Nb V + Nr Yb -
Nb Yr)
Rear Tire Slip Angle

Sard =
Simplify [Sbd -
b/V Srd]
2
m Nd V -
b Nd Yb + b Nb Yd + Nr V Yd -
Nd V Yr
V (-(m Nb V) -
Nr Yb + Nb Yr)
Sara =
Simplify [Sba -
b/V Sra]
2 2
(bNb + NrV + amV -cmV -abYb + bcYb-aVYr +
c V Yr) / (V (-(m Nb V) -
Nr Yb + Nb Yr) )
Sart =
Simplify [Sbt -
b/V Srt]
g m (b Nb + Nr V)
V (-(m Nb V) -
Nr Yb + Nb Yr)
Path Curvature
Scd = Simplify[l/V Srd]
(Nd Yb) + Nb Yd
V (m Nb V + Nr Yb -
Nb Yr)
Sea =
Simplify [1/V Sra]
Nb -
a Yb + c Yb
V (m Nb V + Nr Yb -
Nb Yr)
Set = Simplify[l/V Srt]
g m Nb
V (m Nb V + Nr Yb -
Nb Yr)
113
Appendix BTwo DOF Model Mathematica Session
Lateral Acceleration
SAd =
Simplify [V/g Srd]
V (-(Nd Yb) + Nb Yd)
g (m Nb V + Nr Yb -
Nb Yr)
SAa =
Simplify [V/g Sra]
V (Nb -
a Yb + c Yb)
g (m Nb V + Nr Yb -
Nb Yr)
SAt =
Simplify [V/g Srt]
m Nb V
m Nb V + Nr Yb -
Nb Yr
Steer Angle Response to Path Radius
deltaR = delta /. Solve[Scd == 1/R / delta, delta] [ [1, 1] ]
2
-
(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 /. SDRules]
2
(a Cf -
b Cr) m V
a Cf Cr R + b Cf Cr R
TermlSa = Numerator [TermlS] / Simplify [Denominator [

TermlS] /. a->L-b]
2
(a Cf -
b Cr) m V
Cf Cr L R
114
Terms 2 3 = ExpandNumerator [Simplify [Coefficient [Expand [

deltaR], V] 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
-
+
R -
(Nd R Yb) + Nb R Yd
deltaR2 = TermlSa + Terms23S
2
L (a Cf -
b Cr) m V
-
+
R Cf Cr L R
Understeer Gradient
Kus = Coefficient [Simplify [deltaRl R g] ,VA2]
g m Nb
-(Nd Yb) + Nb Yd
Kusl =
Simplify [Kus /. SDRules]
(a Cf -
b Cr) g m
(a + b) Cf Cr
Stability Factor
Kl =
Simplify [K /. Solve [Srd == V/(L (1+K VA2)), K][[l]]]
2
m Nb V + L Nd Yb + Nr V Yb -
L Nb Yd -
Nb V Yr
2
L V (-(Nd Yb) + Nb Yd)
115
K2 =
Simplify [Numerator [Kl] -
Coefficient [Numerator [Kl] ,
VA2] VA2 /. SDRules /. a->L-b] + Coef f icient [

Numerator [Kl] , VA2] VA2 / Denominator [Kl]
m Nb
L (-(Nd Yb) + Nb Yd)
K3 =
Simplify [K2 /. SDRules]
(a Cf -
b Cr) m
a Cf Cr L + b Cf Cr L
K4 = Numerator [K3] /Simplify[ (Denominator [K3] /,

a -> L -
b)]
(a Cf -
b Cr) m
2
Cf Cr L
Neutral Steer Point
dl =
Simplify [d /. Solve [Numerator [Srn] ==
0, d] [ [1] ] ]
Nb
a - --
Yb
d2 =
Simplify [dl /, SDRules]
(a + b) Cr
Cf + Cr
d3 =
Simplify [d2 /. a > L b]
Cr L
Cf + Cr
Static Margin
SM = (dl a)/L
Nb
-( )
L Yb
SMI =
Simplify [SM /. SDRules]
-(a Cf) + b Cr
Cf L + Cr L
116
Tangent Speed
Vtan = v /. Solve[Sbd deltaR ==

0,V][[2,1]]
Nr Yd -
Nd Yr
-( )
m Nd
Vtanl = Sqrt [Simplify [Expand[ (V /. Solve [V ==

Simplify [
Vtan /. SDRules] ,V] [ [2,1] ]) A2] ] ]
b (a + b) Cr
Sqrt[-( )]
a m
Vtan2 = Sqrt [Simplify [Numerator [VtanlA2] /. a -> L-b] /

A
Denominator [Vtanl 2 ] ]
b Cr L
Sqrt[-( )]
a m
Critical Speed
Vcrit = v /. Solve [Denominator [Srd] ==

0,V][[1]]
Nr Yb -
Nb Yr
-( )
m Nb
Vcritl = Sqrt [Simplify [ (V /. Solve [V == Vcrit /

SDRules,V] [[2,1]])A2]]
2
(a + b) Cf Cr
Sqrt [ ]
-
(a Cf m) + b Cr m
Vcrit2 = Sqrt [Simplify [Numerator [VcritlA2] /. a -> L-b]/

Denominator [Vcritl A2] ]
2
Cf Cr L
Sqrt [ ]
-
(a Cf m) + b Cr m
Characteristic Speed
Vchar = v /. Solve[deltaR == 2 L/R, V] [ [2, 1] ]
(-(Nr Yb) + Nb Yr + Sqrt [-4 LmNb (2 Nd Yb -
2 Nb Yd) +
2
(Nr Yb -
Nb Yr) ]) / (2 m Nb)
117
Vcharl = Sqrt [Simplify [ (V /. Solve [V ==

Vchar /-
SDRules, V] [[2,1]])A2]]
(a + b) Cf Cr (a + b -
2 L)
Sqrt[ ]
-(a Cf m) + b Cr m
Vchar2 = Sqrt [Simplify [Numerator [Vcharl A2] /. a -> L-b]/

Denominator [Vcharl A2] ]
2
Cf Cr L
Sqrt[-( )]
-(a Cf m) + b Cr m
Yaw Radius of Gyration
kz =
Sqrt[Izz/m]
Izz
Sqrt[ ]
m
Geometry to Inertia Ratio

GIR =
LA2/kzA2
2
L m
Izz
Total Cornering Factor

TCF = Cf Cr/mA2
Cf Cr
2
m
Characteristic Equation
Ds == 0
+ s + + s (-( ) ) == 0
a2 = Coef f icient [Ds, sA2]
118
al = Coef f icient [Ds, s]
Nr Yb
-( )
Izz m V
aO = Ds -
a2 sA2 -
al s
Nb Nr Yb Nb Yr
+ _
Izz Izz m V Izz m V
Undamped Natural Frequency
wn =
Simplify [ Sqrt [aO ] ]
in Nb V + Nr Yb -
Nb Yr
Sqrt[ ]
Izz m V
wnl =
Simplify [wn /. SDRules]
2 2 2 2
a Cf Cr + 2 a b Cf Cr + b Cf Cr + a Cf m V -
b Cr m V
Sqrt[ ]
2
Izz m V
Damping Ratio
zeta =
Simplify [al Izz m V/(2 Sqrt[wnA2 (Izz m V)A2])]
-
(m Nr V + Izz Yb)
2 Sqrt [Izz mV (hi Nb V + Nr Yb -
Nb Yr) ]
zetal =
Simplify [zeta /. SDRules]
2 2
-(Cf Izz + Cr Izz + a Cf m + b Cr m) /
2 2 2
(2 Sqrt [Izz m (a Cf Cr + 2 a b Cf Cr + b Cf Cr + a Cf m V -
2
b Cr m V ) ] )
119
Poles
poles = Solve [Ds==0, s] ;
si = s /- poles [[1,1]]
2
(m Nr V + Izz Yb -
Sqrt[(-(m Nr V) Izz Yb)

-
-
4 Izz m V (m Nb V + Nr Yb -
Nb Yr) ] ) / (2 Izz m V)
sla =
Simplify [si /. SDRules]
2 2
( (Cf + Cr) Izz + (a Cf + b Cr) m -
2 2 2
Sqrt[(-(Cf Izz) -
Cr Izz -
a Cf m -
b Cr m) -
2 2 2
4 Izz m (a Cf Cr + 2 a b Cf Cr + b Cf Cr + a Cf m V -
2
b Cr m V ) ] ) / (2 Izz m V)
s2 = s /. poles [[2,1]]
2
(m Nr V + Izz Yb + Sqrt[(-(m Nr V) -
Izz Yb)
4 Izz m V (m Nb V + Nr Yb -
Nb Yr) ] ) / (2 Izz m V)
s2a =
Simplify [s2 /. SDRules]
2 2
( (Cf + Cr) Izz + (a Cf + b Cr) m +
2 2 2
Sqrt[(-(Cf Izz) -
Cr Izz -
a Cf m -
b Cr m)
2 2 2
4 Izz m (a Cf Cr + 2 a b Cf Cr + b Cf Cr + a Cf m V -
2
b Cr m V ) ] ) / (2 Izz m V)
Zeros
Zbd = s /. Solve [Nbd ==

0, s][[l,l]]
-
(m Nd V) -
Nr Yd + Nd Yr
_( }
Izz Yd
120
Zbdl =
Simplify [Zbd /. SDRules]
2 2
abCr+b Cr+amV
Izz V
Zba = s /. Solve [Nba ==

0, s][[l,l]]
-Nr -amV+cmV+aYr-cYr
_( )
Izz
Zbal =
Simplify [Zba /. SDRules]
2 2 2
acCf+abCr+b Cr-bcCr+amV -cmV
Izz V
Zbt = s /. Solve [Nbt ==

0, s][[l,l]]
Nr
Izz
Zbtl =
Simplify [Zbt /. SDRules]
2 2
a Cf + b Cr
Izz V
Zrd = s /. Solve [Nrd ==

0, s][[l,l]]
(Nd Yb) + Nb Yd
_( )
m Nd V
Zrdl =
Simplify [Zrd /- SDRules]
(a + b) Cr
a m V
Zra = s /. Solve [Nra ==

0, s][[l,l]]
Nb -
a Yb + c Yb
-( )
(a -
c) m V
Zral =
Simplify [Zra /. SDRules]
-(c Cf) + a Cr + b Cr -
c Cr
a m V -
c m V
121
Zrt =
Solve [Nrt ==
0, s]
{{}}
122
Appendix C Two DOF Model MATLAB Programs
C.l DOF2Control.m
%DOF2Control 2 DOF Model Execution Control
Controls execution of 2 DOF model. Sets control input type (step, step
% ramp, ramp step / ramp down, or sine steer) . Sets simulation parameters.
% Created 1/11/96
% J. Kiefer
% Control Input Type

step
=
1; % Step steer
ramp =
2; % Ramp step steer
rampsquare =
3; % Ramp square steer
sine =
4; % Sine steer
input =
1; % Select which control input to use
% Simulation Parameters
tO =
0.0; % s Initial time for steer input
tr =
0.2; % s Ramp time
td =
1.0; % s EWell time
ts =
1.0; % s Period for sine steer
tf =
4.0; % s Final time for simulation
tol =
le-5; % 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 Izz L a b c u Cf Cr dO Fzf Fzr Fyg Fya tO tr td ts tf input;
global Bl CI Dl El B3 C3 B5 C5;
% Constants
g =
9.81; % m/s~2 Acceleration due to gravity
% Vehicle Independent Parameters

m =
1775; % kg Gross vehicle mass
Izz =
1960; % kg-m^2 Yaw inertia
f =
0.52; % Fraction of weight on front axle
L =
2.372; % m Wheelbase
u =
100; % km/hr Vehicle forward speed
% Control and Disturbance Inputs

dO ==
1; % deg Steer input magnitude
theta =
0; % deg Side slope
Fya =
0; % N Aerodynamic side force
c =
1.25; % m Distance from front axle to
% aerodynamic side force
% Linear Tire Model Parameters

Cf ==
-1230.5; % N/deg Front cornering stiffness (one tire)
Cr ==
-1155.5; % N/deg Rear cornering stiffness (one tire)
% Non Linear Tire Model Parameters

% Normalized Lateral Force Magic Formula Parameters
Bl =
0.5835;
CI =
1.7166
Dl = 1.0005
El = 0.2517
% Cornering Coefficient Parameters
B3 =
0.333;
C3 =
-1.352e-5;
% Friction Coefficient Parameters

B5 =
1.173;
C5 =
-3.696e-5;
% Unit Conversions
u =
u*1000/3600; % m/s Vehicle forward speed
124
dO =
d0*pi/180; % rad Step steer input
Cf =
Cf*180/pi*2; % N/rad Front tire cornering stiffness (two tires)
Cr =
Cr*180/pi*2; % N/rad Rear tire cornering stiffness (two tires)
125
C.3 DOF2DependParam.m
%DOF2DependParam 2 DOF Model Dependent Parameter Calculation

%
% Calculates values of dependent parameters for 2 DOF model.
%
% Created 1/7/96
% J. Kiefer
% Dependent Parameters
a =
(l-f)*L; % m Distance from front tire to C.G.
b =
f*L; % m Distance from rear tire to C.G.
V =
u; % m/s Vehicle speed
Fyg =
m*g*sin(theta*pi/180) ; % N Side slope lateral force
Fzf =
m*g*f/2*cos(theta*pi/180); % N Front tire normal load (one tire)
Fzr =
m*g*(l-f)/2*cos(theta*pi/180); % N Rear tire normal load (one tire)
% Stability Derivatives
Yb = Cf + Cr; % N/rad Damping-
in-sideslip
Yr =
(a*Cf-b*Cr) /V; % N-s/rad Lateral force / yaw coupling
Yd =
-Cf ; % N/rad Control force
Nb =
a*Cf-b*Cr; % N-m/rad Directional stability
Nr =
(a~2*Cf+b~2*Cr) /V; % N-m-s/rad Yaw damping
Nd =
-a*Cf ; % N-m/rad Control moment
126
C.4 SteerAngle.m
function delta = SteerAngle (t, input, t0,tr,td,ts,tf ,d0)
%SteerAngle Calculate Steer Angle Based on Time and Control Input Selection
%
%delta = SteerAngle (t, input, tO, tr,td,ts,tf,dO)
%
% Determines steer angle based on current time, input selection, and
% simulation parameters. Input selection may be step, ramp, ramp square, or
% sine. Called by functions D0F2LDE.m, D0F2NLDE.m, D0F3LDE.m, D0F3NLDE.m.

%
% Inputs:
% t Time (s)
% input Flag for input selection
% 1 =
step
% 2 =
ramp
% 3 =
rampsquare
% 4 = sine
% to Initial time for simulation (s)

% tr Ramp time (s)
% td EWell time (s)
% ts Period for sine steer (s)
% tf Final time for simulation (s)
% dO Steer input magnitude (rad)
%
% Outputs:
% delta Steer angle (rad)
%
% Created 1/7/96
% J. Kiefer
% Crash Through Result

delta =
0;
% Step Steer
if input == 1
delta =
dO;
if t < tO
delta 0;
end
end
% Ramp Step Steer

if input == 2
delta =
dO;
if t < tO + tr
delta =
d0*(t-t0)/tr;
end
if t < tO
delta =
0;
end
end
% Ramp Square Steer

if input == 3
127
delta =
0;
if t < tO + 2*tr + td
delta =
d0*(t0+td+2*tr-t)/tr;
end
if t < tO + tr + td
delta =
dO;
end
if t < tO + tr
delta =
d0*(t-t0)/tr;
end
if t < tO
delta =
0;
end
end
% Sine Steer
if input == 4
delta =
d0*sin(2*pi*(t-t0)/ts) ;
if t < tO
delta =
0;
end
end
128
C.5 DOF2LFreq.m
%DOF2LFreq Frequency Response of Linear 2 DOF Model

% For a Single Set of Parameters
%
% Generates bode plot data for linear 2 DOF model for outputs of
% sideslip angle and yaw speed, and for inputs of steer angle control,
% aerodynamic side force disturbance, and road side slope disturbance.
%
% Created 2/4/96
% J. Kiefer
D0F2Param; % Set independent parameters
D0F2Control ; % Set execution control parameters
D0F2DependParam; % Calculate dependent parameters
% Transfer Function Denominator

D = [1 -Nr/Izz-Yb/ (m*V) Nb/Izz+(Nr*Yb-Nb*Yr) / (Izz*m*V) ] ;
% Transfer Function Numerators

Nbd =
[Yd/(m*V) (Nd*Yr-Nr*Yd-Nd*m*V) / (Izz*m*V) ] ;
Nba =
[1/ (m*V) (c-a) /Izz+ ( (a-c) *Yr-Nr) / (Izz*m*V) ] ;
Nbt = [g/V -g*Nr/(Izz*V)] ;
Nrd = [Nd/Izz (Nb*Yd-Nd*Yb) / (Izz*m*V) ] ;

Nra =
[(a-c) /Izz ( (c-a)*Yb+Nb) / (Izz*m*V) ] ;
Nrt =
[g*Nb/ (Izz*V) ] ;
% Bode Plot Data

w lcgspace(-l,2) *2*pi;
=
[Mbd,Pbd,w] bode (Nbd, D,w) ;

=
[Mba,Pba,w] bode (Nba, D,w) ;

=
[Mbt,Pbt,w] bode (Nbt, D,w) ;

=
[Mrd,Prd,w] bode (Nrd, D,w) ;

=
[Mra,Pra,w] bode (Nra, D,w) ;

=
[Mrt,Prt,w] bode (Nrt, D,w);

=
129
C.6 DOF2LSim.m
%D0F2LSim Simulation of Linear 2 DOF Model Response to Control and Disturbance

% Inputs
%
% Performs simulation of linear 2 DOF model response to control and
% disturbance inputs. Determines yaw speed, lateral speed, sideslip angle,

% front and rear tire slip angles, front and rear tire lateral forces, and
% lateral acceleration. Plots these responses versus time. Reads data from
% D0F2Param, D0F2DependParam.
%
% Created 1/7/96
% J. Kiefer
D0F2Control; % Set execution control parameters
% Perform simulation
[t,x] =
ode23(,DOF2LDE',0,tf, [0 0]',tol);
=
v x(:,l) ;
r =
x(:,2);
% Steer Angle
delta = zeros ( length ( t) , 1) ;
for i =
l:length(t)
delta(i) = SteerAngle (t (i) input, t0,tr,td,ts,tf, dO)
, ; % rad Steer angle
end
% Vehicle and Tire Slip Angles

beta =
v/u; % rad Vehicle sideslip angle
alphaF =
(v+a*r) /u-delta; % rad Front tires slip angle
alphaR =
(v-b*r)/u; % rad Rear tires slip angle
% External Forces and Moments
Fyf =
Cf*alphaF; % N Front tires lateral force
Fyr =
Cr*alphaR; % N Rear tires lateral force
% State Derivatives
(Fyf Fyr + Fya + Fyg) /m u*r;
-
vdot = +
rdot = (a*Fyf -
b*Fyr -
(c-a) *Fya) /Izz;
% Lateral Acceleration
= + % m/s^ Lateral acceleration
ay vdot u*r;
% Do Plots
subplot (2, 2,1)
plot(t,v)
grid
' '
title ( Lateral Speed )
xlabeK'Time (s) ')
('
ylabel Speed (m/s) ')
subplot (2, 2, 2)
plot(t,r*180/pi)
130
grid
title ('Yaw Speed')

xlabelCTime (s) ')
'
ylabel( 'Speed (deg/s) )
subplot (2, 2, 3)
' ' '
plot(t,beta*180/pi,t,alphaF*180/pi, ,t,alphaR*180/pi, '-. ,t,delta*180/pi, : )
grid
title ( 'VehicleSideslip Angle, Tire Slip Angles, Steer Angle')

'
xlabeK 'Time (s) )
' '
ylabel ( Slip Angle (deg) )
subplot (2, 2, 4)
plot(t,ay/g)
grid
' '
title ( Lateral Acceleration )
xlabel( 'Time (s) ')
'
ylabel ( 'Acceleration (g) )
131
C.7 DOF2LDE.m
function xdot =
DOF2NLDE(t,x)
%DOF2NLDE Non Linear Differential Equations for 2 DOF Model
%
%xdot =
D0F2NLDE(t,x)
%
% Determines derivatives of lateral speed and yaw speed given time and
% state vector. Non linear tire and non linear slip angles. Used with ode23
% for simulation.
% Inputs:
% t Time (s)
% x(l) Lateral speed (m/s)
% x(2) Yaw speed (rad/s)
% Outputs:
% xdot(l) Derivative of lateral speed (m/s ^2)
% xdot (2) Derivative of yaw speed (rad/s''2)
%
% Created 2/18/96
% J. Kiefer
global m Izz L a b c u dO Fzf Fzr Fyg Fya tO tr td ts tf input;
delta =
SteerAngle(t, input, tO, tr,td,ts,tf,dO);
alphaF =
atan( (x(l)+a*x(2) ) /u) -delta;
alphaR =
atan( (x(l)-b*x(2) )/u) ;
[Fyf, Mzf] =
NLTire(Fzf, alphaF);
[Fyr, Mzr] =
NLTire(Fzr, alphaR);
xdot =
[-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 Simulation of Non-Linear 2 DOF Model Response to Control and Disturbance

% Inputs
%
% Performs simulation of non linear 2 DOF model response to control and
% disturbance inputs. Determines yaw speed, lateral speed, sideslip angle,

% front and rear tire slip angles, front and rear tire lateral forces, and
% lateral acceleration. Plots these responses versus time. Reads data from
% D0F2Param, D0F2DependParam.
%
% Created 2/18/96
% J. Kiefer
D0F2Control; % Set execution control parameters
% Perform simulation
[t,x] =
ode23(,DOF2NLDE',0,tf, [0 0]',tol);
v =
x(:,l) ;
r =
x(:,2) ;
% Steer Angle
delta = zeros (length(t) , 1) ;
for i =
l:length(t)
delta(i) = SteerAngle (t(i) input, tO,tr,td,ts,tf
, ,d0) ; % rad Steer angle
end
% Vehicle and Tire Slip Angles

beta =
atan(v/u) ; % rad Vehicle sideslip angle
alphaF =
atan( (v+a*r) /u) -delta; % rad Front tires slip angle
alphaR =
atan( (v-b*r) /u) ; % rad Rear tires slip angle
% External Forces and Moments
Fyf =
NLTire(Fzf, alphaF); % N Front tire lateral force (one tire)
Fyr =
NLTire(Fzr, alphaR) ; % N Rear tire lateral force (one tire)
% State Derivatives
(delta) + 2*Fyr + Fya + Fyg) /m u*r;
-
vdot = (2*Fyf.*cos
rdot = (2*a*Fyf.*cos (delta) 2*b*Fyr (c-a)*Fya)/Izz;
- -
% Lateral Acceleration
+ % m/s^2 Lateral acceleration
ay =
vdot u*r;
% Do Plots
subplot(2,2,l)
plot(t,v)
grid
('
title Lateral
Speed')
xlabeK'Time (s) ')
ylabel ('Speed (m/s) ')
subplot (2, 2, 2)
plot(t,r*180/pi)
133
grid
title ('Yaw Speed')

xlabelCTime (s) ')
' '
ylabel ( Speed (deg/s) )
subplot (2, 2, 3)
' ' ' '
plot(t,beta*180/pi,t,alphaF*180/pi, ,t,alphaR*180/pi, '-. ,t,delta*180/pi, : ')
grid
title ( 'Vehicle Sideslip Angle, Tire Slip Angles, Steer Angle')

xlabel( 'Time (s) ')
' '
ylabel ( Slip Angle (deg) )
subplot (2, 2, 4)
plot (t, ay/g)
grid
' '
title ( Lateral Acceleration )
'
xlabel( 'Time (s) )
'
ylabel ( 'Acceleration (g) )
134
C.9 DOF2NLDE.m
function xdot =
D0F2NLDE(t,x)
%D0F2NLDE Non Linear Differential Equations for 2 DOF Model
%
%xdot =
D0F2NLDE(t,x)
%
% Determines derivatives of lateral speed and yaw speed given time and
% state vector. Non linear tire and non linear slip angles. Used with ode23
% for simulation.
%
% Inputs:
% t Time (s)
% x(l) Lateral speed (m/s)
% x(2) Yaw speed (rad/s)
% Outputs:
% xdot(l) Derivative lateral
of speed (m/s^2)
% xdot (2) Derivative of yaw speed (rad/sA2)
%
% Created 2/18/96
% J. Kiefer
global m Izz L a b c u dO Fzf Fzr Fyg Fya tO tr td ts tf input;
delta =
SteerAngle(t, input, tO, tr,td,ts,tf,dO);
alphaF =
atan((x(l)+a*x(2))/u)-delta;
alphaR =
atan( (x(l)-b*x(2) )/u) ;
Fyf =
NLTire(Fzf, alphaF);
Fyr =
NLTire(Fzr, alphaR);
xdot =
[-u*x(2) + (2*Fyf*cos (delta) +2*Fyr+Fya+Fyg)/m
(2*a*Fyf*cos (delta) -2*b*Fyr+
(a-c) *Fya) /Izz] ;
135
Appendix D Relevant Literature
Dynamics."
"Control of Vehicle Automotive Engineering, May 1995, p. 87-93.

Methods."
"Road Vehicles -
Lateral Transient Response Test ISO 7401, May 1988.
Procedure."
"Road Vehicles -
Steady State Circular Test ISO 4138, Aug. 1982.

Vocabulary."
"Road Vehicles Vehicle Dynamics
-
and Road-Holding Ability -

ISO 8855,
Dec. 1991.
Terminology."
"Vehicle Dynamics 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 of Safety
Stability."
Critical Maneuvers for Assessing Ground Vehicle Dynamic SAE Paper
No. 930760, Mar. 1993.
Allen, R. Wade and Theodore J. Rosenthal. "Requirements for Vehicle Dynamics

Models."
Simulation SAE Paper No. 940175, Feb. 1994.
Allen, R. Wade, Raymond E. Magdaleno, Theodore J. Rosenthal, David H. Klyde, and
Jeffrey R. Hogue. 'Tire Modeling Requirements for Vehicle Dynamics

Simulation."
Allen, R. Wade, Thomas T. Myers, and Theodore J. Rosenthal. "Vehicle Stability

Systems."
Considerations with Automatic and Four Wheel Steering SAE Paper No.
931979, Nov. 1993.
Allen, R. Wade, Theodore J. Rosenthal, and Jeffrey R. Hogue. "Modeling and Simulation
Interaction."
of DriverAêhicle SAE Paper No. 960177, Feb. 1996.
Allen, R. Wade, Theodore J. Rosenthal, and Henry T. Szostak. "Steady State and
Handling."
Transient Analysis of Ground Vehicle SAE Paper No. 870495, 1987.
Allen, R. Wade, Theodore J. Rosenthal, David H. Klyde, Keith J. Owens, and Henry T.
Szostak. "Validation of Ground Vehicle Computer Simulations Developed for
Analysis."
Dynamics Stability SAE Paper No. 920054, Feb. 1992.
Allen, R. Wade, Henry T. Szostak, Theodore J. Rosenthal, David H. Klyde, and Keith J.
Owens. "Characteristics Influencing Ground Vehicle Lateral/Directional Dynamic
Stability."
Antoun, R.J, P.B. Hackert, M.C. O'Leary, and A. Sitchin. "Vehicle Dynamic Handling
Computer Simulation Model Development,
--
Correlation, and Application Using

ADAMS."
SAE Paper No. 860574, 1986.
Araki, Kazuo and Hideo Sakai. "Study Tire Model Consisting of Theoretical and
of
Experimental Equations for Vehicle Dynamics Analysis Part 2: Under the

-
Surface."
Condition of Various Velocity on the Asphaltic Road SAE Paper No.

960996, Feb. 1996.
117, No. 6, June
Cars."
Ashley, Steven. "Spin Control for Mechanical Engineering, Vol.

1995, p. 66-68.
136
Bakker, Egbert, Lars Nyborg, and Hans B. Pacejka. 'Tyre Modelling for Use in Vehicle
Studies."
Dynamics SAE Paper No. 870421, 1987.
Bakker, Egbert,Hans B. Pacejka, and Lars Lidner. "A New Tire Model with an
Studies."
Application in Vehicle Dynamics SAE Paper No. 890087, 1989.
Cars."
Barak, Pinhas."Magic Numbers in Design of Suspensions for Passenger SAE

Paper No. 911921, 1991.
Optimization."
Barbieri, Nilson. "Suspensions SAE Paper No. 921491, 1992.
Bastow, D. and G. Howard. Car Suspension and Handling. Warrendale, PA: SAE, 1993.
Bernard, James E. and C.L. Clover. 'Tire Modeling for Low-Speed and High-Speed
Calculations."
Bernard, James E. and C.L. Clover. "Validation of Computer Simulations of Vehicle

Dynamics."
Bixel, Ronald A., Gary J. Heydinger, N.J. Durisek, and Dennis A. Guenther. "New
Developments in Vehicle Center of Gravity and Inertial Parameter Estimation and

Measurement."
Blank, Matthew Donald Margolis. "The Effect of Normal Force Variation on the
and
Automobiles."
Lateral Dynamics of SAE Paper No. 960484, Feb. 1996.
Bowman, J. Eric and E.H. Law. "A Feasibility Study of an Automotive

System."
Slip Control
Braking SAE Paper No. 930762, Mar. 1993.
Breuer, Bert, Thomas Bachmann, Stefan Ernesti, and Jorg Stocker. "Methods and
Friction."
Instruments for On-Board Measurement of Tyre/Road SAE Paper No.

942470, Dec. 1994.
Bundorf, R.T.and R.L. Leffert. 'The Cornering Compliance Concept for Description of
Properties."
Vehicle Directional Control SAE Paper No. 760713, Oct. 1976.
Cambiaghi, Danilo and Marco Gadola. "Computer- Aided

Italy."
Racing Car Design and
Development at the University of Brescia, SAE Paper No. 942507, Dec.
1994.
Captain, K.M., A.B. Boghani, and D.N. Wormley. "Analytical Tire Models for Dynamic
Vehicle System Dynamics, Vol. 8, 1979, p. 1-32.
Simulation."
Vehicle
Car Suspension Systems and Vehicle Dynamics. SAE Publication No. SP-878, Sept.
1991.
Chen H Fred and Dennis A. Guenther. "The Effects of Suspension Stiffness on Handling
SAE Paper No. 911928, 1991.
Responses."
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 James E. Bernard. "The Influence of Lateral Load Transfer
SAE Paper No. 930763, Mar. 1993.
Response."
Distribution on Directional
Cole, D.E. Elementary Vehicle Dynamics. Ann Arbor, MI: University of Michigan, 1972.
137
Crahan, Thomas C. "Modeling Steady-State Suspension Kinematics and Vehicle Dynamics

Methodology."
of Road Racing Cars Part I: Theory and
-
SAE Paper No. 942505,

Dec. 1994.
Crahan, Thomas C. "Modeling Steady-State Suspension Kinematics and Vehicle Dynamics

Examples."
of Road Racing Cars Part II: -
Crolla, D.A. and M.B.A. Abdel-Hady. "Semi-Active Suspension Control for a Full
Model."
Vehicle SAE Paper No. 91 1904, Sept. 1991.
Model."
Day, Terry D. "An Overview of the HVE Vehicle SAE Paper No. 950308, Feb.
1995.
Dickison, J.G. and A.J. Yardley. "Development and Application of a Functional Model to
Development."
Vehicle SAE Paper No. 930835, Mar. 1993.
Dixon, John C. Tyres, Suspension and Handling, Cambridge, England: Cambridge

University Press, 1991.
Dreyer, Andreas and Heinz-Dieter Heitzer. "Control Strategies for Active Chassis Systems
Friction."
with Respect to Road SAE Paper No. 910660, Feb. 1991.
Egnaczak, Bernard C. "Supplement to: 'The Development of a Differential for the

Control."
Improvement of Traction Auto Tech 89, Session 5 Traction Control,
Nov. 14, 1989.
ElBeheiry, ElSayed M. and Dean C. Karnopp. "Optimization of Active and Passive

Model."
Suspensions Based Full Car
on a SAE Paper No. 951063, Feb. 1995.
Ellis, John R. Road Vehicle Dynamics, Akron, OH: J.R. Ellis, 1989.
Ellis, John R. Vehicle Dynamics. London: Business Books, 1969.
and E. Harry Law. "Simulation
Floyd, R. Scott and Analysis of Suspension and
Cars."
Aerodynamic Interactions of Race SAE Paper No. 942537, Dec. 1994.
Franklin, Gene F., J. David Powell, and Abbas Emami-Naeini. Feedback Control of
Addison-
Dynamic Systems. New York: Wesley Publishing Company, Inc., 1994.
Garrot, W. Riley, Douglas L. Wilson, and Richard A. Scott. "Digital Simulation for
Maneuvers."
Automobile Simulation, Sept. 1981, p. 83-91.
Gillespie, T.D. Fundamentals of Vehicle Dynamics. Warrendale, PA: SAE, 1992.
Gim, Gwanghun and Namcheol Kang. "Requirements of a Tire Model for Practical
Vehicles."
Cornering Simulations of SAE Paper No. 960179, Feb. 1996.

Steady-
Gim, Gwanghun Parviz E. Nikravesh. "A Three-Dimensional Tire Model for
and
Vehicles."
State Simulations of SAE Paper No. 931913, Nov. 1993.
Gim, Gwanghun and Parviz E. Nikravesh. "An Analytical Model of Pneumatic Tyres for
Slips."
Vehicle Dynamic Simulations. Part 1: Pure International Journal of Vehicle

Design, Vol. 11, No. 6, 1990.
Gim, Gwanghun Parviz E. Nikravesh. "An Analytical Model

and of Pneumatic Tyres for
Slips."
Vehicle Dynamic Simulations. Part 2: Comprehensive International Journal

of Vehicle Design, Vol. 12, No. 1, 1991.
Gruening, James and James E. Bernard. "Verification of Vehicle Parameters for Use in
Simulation."
Computer SAE Paper No. 960176, Feb. 1996.
138
Gruening, James, Keith A. Williams, Kurt Hoffmeister, and James E. Bernard. 'Tire
Processor."
Force and Moment SAE Paper No. 960182, Feb. 1996.
Guntur, R. S. Sankar. "A Friction Circle Concept for Dugoff s Tyre Friction
and
Model."
International Journal of Vehicle Design, Vol. 1, No. 4, 1980.
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
Dynamics."
Ph.D. Dissertation, Ohio State University, Columbus, Ohio, 1990.
Heydinger, Gary J., W. Riley Garrot, and Jeffrey P. Chrstos. "The Importance of Tire
Response."
Lag on Simulated Transient Vehicle SAE Paper No. 910235, 1991.
Heydinger, Gary J., Paul A. Grygier, and Seewoo Lee. "Pulse Testing Techniques
Dynamics."
Applied to Vehicle Handling 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 of a Locked
Differential."
Howard, Geoffrey. Chassis & Suspension Engineering, London, England: Osprey

Publishing Limited, 1987.
Huang, Feng, J. Roger Chen, and Lung-Wen Tsai. "The Use of Random Steer Test Data
Estimation."
for Vehicle Parameter SAE Paper No. 930830, Mar. 1993.
Huchtkoetter, Heinrich Heinz Klein. "The Effect of Various

and
Limited-Slip Differentials
Traction."
in Front-Wheel
Drive Vehicles on Handling and SAE Paper No.
960717, Feb. 1996.
Control."
Ikushima, Y. K Sawase. "A Study on the Effects
and of the Active Yaw Moment
Jung, Shinsub and Dennis A. Guenther. "An Examination of the Maneuverability of an All
Speed."
Wheel Steer Vehicle at Low SAE Paper No. 910241, Feb. 1991.
Kaminaga, M., M. Murata, and Y. Tateishi. "Factoring Nonlinear Kinematics into New
Dynamics."
Suspension Design: A CAE Approach to Vehicle Roll SAE Paper No.
940871, Feb. 1994.
Systems."
Karnopp, Dean. "Active Damping in Road Vehicle Suspension Vehicle System

Dynamics, Vol. 12, 1983, p. 291-316.
Dampers."
Kasprzak, James L. R. Scott Floyd. "Use
and of Simulation to Tune Race Car
Katz, Joseph. Race Car Aerodynamics. Cambridge, MA: Robert Bentley, Inc., 1995.
Klein, Richard H., Gary L. Teper, and James D. Fait. "Lateral/Directional Stability of Tow
Vehicles."
Dolly Type Combination SAE Paper No. 960184, Feb. 1996.

System."
Ko, Y. T. Oh. "Motion Control of the Vehicle
and with an Active Suspension
Koibuchi, Ken, Masaki Yamamoto, Yoshiki Fukada, andShoji Inagaki. "Vehicle Stability
Brake."
Control in Limit Cornering by Active SAE Paper No. 960487, Feb. 1996.
139
Korturn, W.and W. Schiehlen. "General Purpose Vehicle System Dynamics Software

Formalisms."
Based on Multibody Vehicle System Dynamics, No. 14, 1985, p.

229-263.
Car."
Kramer, Kenneth D. and Dale E. Calkins. "Lateral Response of Formula SAE Race
Optimization."
La Joie, Joseph C. "Race Car Performance SAE Paper No. 942492, Dec.
1994.
Technology."
Langer, William. "Vehicle Testing with Flat Surface Roadway SAE Paper
No. 960731, Feb. 1996.
Lee, Allan Y. "Emulating the Lateral Dynamics of a Range of Vehicles Using a Four-
Vehicle."
Wheel-Steering SAE Paper No. 950304, Feb. 1995.

Maneuvers."
Lee, Allan Y. "Performance of Four-Wheel-Steering Vehicles in Lane Change

Lee, Seewoo, Jeffrey P. Chrstos, and Dennis A. Guenther. "Modeling of Dynamic

Characteristics of Tire Lateral and Longitudinal Force Responses to Dynamic
Inputs."
Lee, Seewoo, Gary J. Heydinger, and Dennis A. Guenther. "The Application of Pulse
Input Techniques to the Study Tire Lateral Force and Self- Aligning Moment
of
Domain."
Dynamics in the Frequency SAE Paper No. 950317, Feb. 1995.
Lund, Yvonne I.and James E. Bernard. "The Relationship Between the Complexity of
Results."
Linear Models and the Utility of the Computer SAE Paper No. 920052,
Feb. 1992.
Maalej, Aref Y. "Application of Suspension Derivative Formulation to Ground Vehicle

Simulation."
Modeling and Ph.D. Dissertation, The Ohio State University,

Columbus, OH, 1988.
Mabrouka, Hani, H. Fred Chen, Aref Y. Maalej, and Dennis A. Guenther. "Effect of
Behavior."
Lateral Tire Flexibility on the Steering Dynamic SAE Paper No.

910239, Feb. 1991.
Mashadi, Behrooz and David A. Crolla. "Vehicle
Conditions."
Handling Analysis Using Linearization
Around Non-Linear Operating SAE Paper No. 960482, Feb. 1996.
McConville, James B. and John C. Angell. 'The Dynamic Simulation of a Moving Vehicle
Program."
Subject to Transient Steering Inputs Using the ADAMS Computer

ASME Paper No. 84-DET-2, 1984.
Metz, L. Daniel and 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, Errol Shavers.
and
"Comparison of Linear Roll Dynamics Properties for Various Vehicle

Configurations."
SAE Paper No. 920053, 1992.
Metz, L. Daniel, Troy S. Torbeck, Kevin H. Forbes, and L. Gregory Metz. "Evasive
Tire."
Maneuver Capability Without and In the Presence of a Flat SAE Paper No.
942469, Dec. 1994.
140
Vehicles."
Metz, L. Daniel. "Dynamics of Four-Wheel Steer Off-Highway SAE Paper No.
930765, Mar. 1993.
Milliken, William F. and Doug L. Milliken. Race Car Vehicle Dynamics. Warrendale, PA:
SAE, 1995.
Method."
Milliken, William F. and R.S. Rice. "Moment IMechE Paper No. CI 13/83,
1983, p. 31-60.
Milliken, William F., Peter G. Wright, and Douglas L. Milliken. "Moment Method A -
Development."
Comprehensive Tool for Race Car SAE Paper No. 942538, Dec.
1994.
Mimuro, Tetsushi, Masayoshi Ohsaki, Hiromichi Yasunaga, and Kohji Satoh. "Four
Response."
Parameter Evaluation Method of Lateral Transient SAE Paper No.
901734, 1990.
Mola, Simone. Fundamentals of Vehicle Dynamics, Detroit, MI: General Motors Institute,
1969.
Moline, D., S. Floyd, S. Vaduri, and E.H. Law. "Simulation and Evaluation of Semi-
Suspensions."
Active SAE Paper No. 940864, Feb. 1994.
Mori, Yoshinori, Hironobu Matsushita, Takashi Yonekawa, Yoshihisa Nagahara, and

Control."
Hiroshi Shimomura. "A Simulation System for Vehicle Dynamics SAE
Paper No. 910240, Feb. 1991.
Nalecz, Andrzej G. "Analysis of the Dynamic Response of a Four Wheel Steering Vehicles
Speed."
at High International Journal of Vehicle Design, Vol. 9, No. 2, 1988.
Nalecz, Andrzej G. "Development and Validation of Light Vehicle Dynamics Simulation

(LVDS)."
Nalecz, Andrzej G. Alan C. Bindemann. "Investigation into the Stability of Four

and
Vehicles."
Wheel Steering International Journal of Vehicle Design, Vol. 9, No. 2,
1988, p. 159-178.
Naude, Alwyn F. and Jasper L. Steyn. "Objective Evaluation of the Simulated Handling
Manoeuvre."
Characteristics of a Vehicle in a Double Lane Change SAE Paper No.

930826, Mar. 1993.
Negrut, D.and J.S. Freeman. "Dynamic Tire Modelling for Application with Vehicle
Terrain."
Simulations Incorporating SAE Paper No. 940223, Feb. 1994.
Neto, Mauro Speranza, Fernando Riberio da Silva, and Jose Francisco Martinex. "Design
Methodology in Vehicle Dynamics, Using the Procedures of Modeling, Simulation,
Dynamics."
and Analysis of System SAE Paper No. 921480, 1992.
New Developments in Vehicle Dynamics, Simulation, and Suspension Systems. SAE

Publication No. SP-1074, Feb. 1995.
Nikravesh, Parviz E. Jong-Nyun Lee. "Optimal Four-Wheel Steering Strategy Using

and
Models."
Nonlinear Analytical Vehicle SAE Paper No. 931915, Nov. 1993.
Handling."
Olley, Maurice. "Suspension and Detroit, MI: Chevrolet Engineering Center,

1937.
Suspensions."
Olley, Maurice. "Notes on Detroit, MI: Chevrolet Engineering Center,

1961.
141
n."
Olley, Maurice. "Suspensions Notes Detroit, MI: Chevrolet Engineering Center, 1966.
Palmeri, Paolo S., Alberto Moschetti, and Luigi Gortan. "H-Infmity Control for Lancia
System."
Thema Full Active Suspension SAE Paper No. 950583, Feb. 1995.
Petersen, Michael R.and John M. Starkey. "Nonlinear Vehicle Performance Simulation

withTest 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.
Radt, Hugo S. "An Efficient Method for Treating Race Tire Force-Moment Data."
SAE
Paper No. 942536, Dec. 1994.
Radt, Hugo S. D.A. Gleniming. "Normalization of Tire Force and Moment

and Data."
Tire
Science and Technology, Vol. 21, No. 2, Apr.-June 1993, p. 91-119.
Radt, Hugo S. and Donald J. Van Dis. "Vehicle Handling Responses Using Stability
Derivatives."
Reichelt, Werner. "Correlation Analysis of Open/Closed Loop Data for Objective

Cars."
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 William F. Milliken. "Static Stability and Control of the Automobile
Method."
Utilizing the Moment SAE Paper No. 800847, June 1980.
Sayers, Micheal W. and C. Mink. "A Simulation Graphical User Interface for Vehicle
Models."
Dynamics SAE Paper No. 950169, Feb. 1995.
Sayers, Michael W.and Stephen M. Riley. "Modeling Assumptions for Realistic

Trucks."
Multibody Simulations of the Yaw and Roll Behavior of
Heavy SAE
Paper No. 960173, Feb. 1996.
Schuring, Dieterich J., Wolfgang Pelz, and Marion G. Pottinger. "A Model for Combined
Forces."
Tire Cornering and Braking SAE Paper No. 960180, Feb. 1996.
Schuring, Dieterich J., Wolfgang Pelz, and Marion G. Pottinger. "An Automated
Formula' Concept."
Implementation of the 'Magic SAE Paper No. 931909, Nov.
1993.
Segal, Leonard. "Theoretical Prediction and Experimental Substantiation of the Response

Control."
of the Automobile to Steering Proceedings of the Automobile Division of
the Institution of Mechanical Engineers, No. 7, 1956-1957, p. 310-330.
Review."
Sharp, R.S. D.A. Crolla. "Road Vehicle Suspension System Design
and A
Vehicle System Dynamics, Vol. 16, 1987, p. 167-192.
Shimada, K. and Y. Shibahata. "Comparison of Three Active Chassis Control Methods for
Moments."
Stabilizing Yaw SAE Paper No. 940870, Feb. 1994.
Smith, C. Engineer to Win. Osceola, WI: Motorbooks International, 1984.

Smith, C. Prepare to Win. Fallbrook, CA: Aero Publishers, Inc., 1975.
142
Smith, C. Tune to Win. Fallbrook, CA: Aero Publishers, Inc., 1978.

Automobiles."
Smith, Norman. 'Transient Control Response of Vehicle System
Dynamics, Vol. 6, No. 2-3, Sept. 1977, p. 63-67.
Sohn, H.S., S.C. Lee, M.W. Suh, and Y.M. Song. "The Influences of Chassis Geometric
Performances."
Characteristics on Vehicle Dynamic SAE Paper No. 940872, Feb.
1994.
Song, Jun-gyu and Yong-San Yoon. "Design of Two-Wheel Steer Vehicle Using Optimal
Steer."
Control Algorithm of Four-Wheel 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, and Dennis A. Guenther.
Models."
"A Study of Vehicle Class Segregation Using Linear Handling SAE
Paper No. 950307, Feb. 1995.
Taborek, Jaroslav J. Mechanics of Vehicles. Cleveland, OH: Penton, 1957.

Analysis."
Thomas, D.W. "Vehicle Modeling and Service Loads SAE Paper No. 871940,
Oct. 1987.
Trom, J.D., J.L. Lopex, M.J. Vanderploeg. "Modeling of a Mid-Size Passenger Car
and
Program."
Using Multibody
a Dynamics Transactions of the ASME, Journal of
Mechanisms, Transmissions, and Automation in Design, Vol. 109, Dec. 1987.
Trom, J.D., M.J. Vanderploeg, and James E. Bernard. "Application of Inverse Models to
Problems."
Vehicle Optimization Vehicle System Dynamics, Vol. 19, 1990, p. 97-
110.
Turpin, D.R. and D.F. Evans. "High-Fidelity Road/Tire Interaction Models for Real Time
Simulation."
Van Valkenburgh, P. Race Car Engineering and Mechanics. Seal Beach, CA: Paul Van
Valkenburgh, 1986.
van Zanten, Anton Th., Rainer Erhardt, Albert Lutz, Wilfried Neuwald, and Harmut
VDC."
Bosch- SAE Paper No.
Bartels. "Simulation for the Development of the
960486, Feb. 1996.

Vanderlploeg, M.J., J.D. Trom, James E. Bernard. "Evaluation
and of Four-Wheel Steer
Model."
Path Follow Performance Using a Linear Inverse Vehicle SAE Paper No.
880644, 1988.
Vedamuthu, S. E.H. Law. "An Investigation
and of the Pulse Steer Method for
Qualities."
Determining Automobile Handling SAE Paper No. 930829, Mar. 1993.

Vehicle Dynamics and Electronic Controlled Suspensions. SAE Publication No. SP-861,
Feb. 1991.
Vehicle Dynamics and Rollover Propensity Research. SAE Publication No. SP-909, Feb.
1992.
Vehicle Dynamics and Simulation. SAE Publication No. SP-950, Mar. 1993.
Vehicle Suspension System Advancements. SAE Publication No. SP-1031, Feb. 1994.
143
Whatmough, K.J. "Real-Time Wheel Brake and Tire Lateral Force Models Refined for
Speeds."
Low SAE Paper No. 940178, Feb. 1994.
Whitcomb, David W. and William F. Milliken. "Design Implications of a General Theory

Control."
of Automobile Stability and Proceedings of the Automobile Division of
the Institution of Mechanical Engineers Aug. 1956, p. 83-107.
,
Wilson, D.A., R.S. Sharp, and S.A. Hassan. "Application ofLinear Optimal Control
Suspension."
Theory to the Design of Automobile Vehicle System Dynamics, Vol.

15, 1986, p. 105-118.
Wong, Jo Yung. Theory of Ground Vehicles. New York: John Wiley & Sons, Inc., 1993.
Edge!"
Wright, Peter. "Out at the Racecar, Vol. 5, No. 3, 1995, p. 15-18.
Xia, Xunmao. "A Nonlinear Analysis Closed Loop Driver/Vehicle Performance with
of
Control."
Four Wheel Steering Ph.D. Dissertation, Department of Mechanical
Engineering, Clemson University, Clemson, SC, Dec. 1990.
Xia, Xunmao and E.H. Law. "Nonlinear Analysis of Closed Loop Driver/Automobile
Control."
Performance with Four Wheel Steering SAE Paper No. 920055, 1992.
Xia, Xunmao and J.N. Willis. 'The Effects of Tire Cornering Stiffness on Vehicle Linear
Performance."
Handling SAE Paper No. 950313, Feb. 1995.

Stability."
Yamamoto, Masaki. "Active Control Strategy for Improved Handling and SAE
Paper No. 911902, Sept. 1991.
Yasui, Yoshiyuki, Kenji Tozu, Noriaki Hattori, and Masakazu Sugisawa. "Improvement
ofVehicle Directional Stability for Transient Steering Maneuvers Using Active

Control."
Brake SAE Paper No. 960485, Feb. 1996.
144

Vehicle Dynamics

Uploaded by

Copyright:

Available Formats

Vehicle Dynamics

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Vehicle Dynamics

Uploaded by

Copyright:

Available Formats

Rochester Institute of Technology

RIT Scholar Works

Modeling of road vehicle lateral dynamics

Follow this and additional works at: http://scholarworks.rit.edu/theses

DEPARTMENT OF MECHANICAL ENGINEERING

Rochester Institute of Technology to reproduce my thesis entitled Modeling of Road

August 16, 1996

The lateral dynamics of a road vehicle is studied through the development of a

Newtonian mechanics. Both linear and non-linear models are developed.

disturbance inputs. Steady-state and transient response characteristics and frequency

data nondimensionalization is utilized. Simulation is performed and compared with the

List of Figures vii

Chapter 2 Literature Review 5

Chapter 3 Tire Behavior 10

3.2 Lateral Force Mechanics 11

3.3 Linear Tire Model 14

3.4 Non-Linear Tire Model 14

Chapter 4 Two Degree-of-Freedom Vehicle Model 24

4.2 Description of Model 25

4.3 Derivation of Equations of Motion 29

4.5 Linear Model 33

4.5. 12 Lateral Acceleration Steady-State Step Response Gain 44

Chapter 5 Conclusion 100

Appendix A Tire Model MATLAB Programs 105

Appendix B Two DOF Model Mathematica Session 108

Appendix C Two DOF Model MATLAB Programs 123

Appendix D Relevant Literature 136

Table 3.1: Non-Linear Tire Model Parameters 17

Table 3.2: Experimental Tire Data 21

Table 4.1: Vehicle Parameters 28

Table 4.2: Linear Tire Model Parameters 36

Table 4.3: Steady-State Response Gains 45

Table 4.4: System Zeros 56

Figure 3. 1 : Tire Slip Angle 11

Figure 3.2: Tire Lateral Force Versus Slip Angle 12

Figure 3.3: Experimental Tire Data 21

Figure 3.4: Tire Cornering Coefficient 22

Figure 3.5: Tire Lateral Friction Coefficient 22

Figure 3.6: Tire Normalized Lateral Force 23

Figure 3. 7: Reconstructed Tire Lateral Force 23

Figure 4.1: Vehicle Model 26

Figure 4.2: Free-Body Diagram 29

Figure 4.3: Kinematic Diagram 32

Figure 4.4: Gravitational Side Force 37

Figure 4.5: Natural Frequency vs. Vehicle Velocity 54

Figure 4.6: Poles and Zeros 57

Figure 4. 7: Sideslip Angle / Steer Angle Frequency Response, V = 100 km/hr 61

Figure 4. 12: Yaw Velocity /Road Side Slope Frequency Response, V =

Figure 4. 13: Sideslip Angle / Steer Angle Frequency Response, V = 30 km/hr 64

Figure 4.15: Simulation Steer Angle Inputs 67

Figure 4.16: Linear Step Steer Lateral Velocity Response 71

Figure 4.17: Linear Step Steer Yaw Velocity Response 71

Figure 4.18: Linear Step Steer Sideslip Angle Response 72

Figure 4.21 : Linear Step Steer Lateral Acceleration Response 73

Figure 4.22: Linear Ramp Step Steer Lateral Velocity Response 74

Figure 4.23: Linear Ramp Step Steer Yaw Velocity Response 74

Figure 4.24: Linear Ramp Step Steer Sideslip Angle Response 75

Figure 4.27: Linear Ramp Step Steer Lateral Acceleration Response 76