VEHICLE DYNAMICS

FACHHOCHSCHULE REGENSBURG
UNIVERSITY OF APPLIED SCIENCES
HOCHSCHULE FÜR
TECHNIK
WIRTSCHAFT
SOZIALES

LECTURE NOTES
Prof. Dr. Georg Rill

© October 2005

download: http://homepages.fh-regensburg.de/%7Erig39165/
Contents

Contents I

1 Introduction 1
1.1 Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.1 Vehicle Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.2 Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.2.3 Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2.4 Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.2.5 Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.1 Reference frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3.2 Toe and camber angle . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.3 Design Position of Wheel Rotation Axis . . . . . . . . . . . . . . . 6
1.3.4 Steering Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.4.1 Kingpin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
1.3.4.2 Caster and Kingpin Angle . . . . . . . . . . . . . . . . . . 8
1.3.4.3 Caster, Steering Oset and Disturbing Force Lever . . . . 9

2 Road 10
2.1 Modeling Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
2.2 Deterministic Proles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.1 Bumps and Potholes . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.2.2 Sine Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3 Random Proles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3.1 Statistical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.3.2 Classication of Random Road Proles . . . . . . . . . . . . . . . . 15
2.3.3 Realizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3.3.1 Sinusoidal Approximation . . . . . . . . . . . . . . . . . . 16
2.3.3.2 Shaping Filter . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.3.3 Two-Dimensional Model . . . . . . . . . . . . . . . . . . . 18

I
3 Tire 19
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.1 Tire Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.2 Tire Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.3 Tire Forces and Torques . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.4 Measuring Tire Forces and Torques . . . . . . . . . . . . . . . . . . 21
3.2 Contact Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.1 Basic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
3.2.2 Local Track Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.2.3 Tire Deection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.4 Length of Contact Patch . . . . . . . . . . . . . . . . . . . . . . . . 28
3.2.5 Static Contact Point . . . . . . . . . . . . . . . . . . . . . . . . . . 29
3.2.6 Contact Point Velocity . . . . . . . . . . . . . . . . . . . . . . . . . 31
3.2.7 Dynamic Rolling Radius . . . . . . . . . . . . . . . . . . . . . . . . 32
3.3 Forces and Torques caused by Pressure Distribution . . . . . . . . . . . . . 34
3.3.1 Wheel Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
3.3.2 Tipping Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
3.3.3 Rolling Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
3.4 Friction Forces and Torques . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.4.1 Longitudinal Force and Longitudinal Slip . . . . . . . . . . . . . . . 37
3.4.2 Lateral Slip, Lateral Force and Self Aligning Torque . . . . . . . . . 40
3.4.3 Wheel Load Inuence . . . . . . . . . . . . . . . . . . . . . . . . . . 41
3.4.4 Two-Dimensional Tire Characteristics . . . . . . . . . . . . . . . . . 43
3.4.5 Dierent Friction Coecients . . . . . . . . . . . . . . . . . . . . . 45
3.4.6 Self Aligning Torque . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.4.7 Camber Inuence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4.8 Bore Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.4.9 Typical Tire Characteristics . . . . . . . . . . . . . . . . . . . . . . 52

4 Suspension System 54
4.1 Purpose and Components . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
4.2 Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2.1 Multi Purpose Systems . . . . . . . . . . . . . . . . . . . . . . . . . 55
4.2.2 Specic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3 Steering Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3.1 Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.3.2 Rack and Pinion Steering . . . . . . . . . . . . . . . . . . . . . . . 57
4.3.3 Lever Arm Steering System . . . . . . . . . . . . . . . . . . . . . . 57
4.3.4 Drag Link Steering System . . . . . . . . . . . . . . . . . . . . . . . 58
4.3.5 Bus Steer System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
4.4 Standard Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4.1 Springs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.4.2 Damper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
4.4.3 Rubber Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

II
 4.5 Dynamic Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
4.5.1 Testing and Evaluating Procedures . . . . . . . . . . . . . . . . . . 63
4.5.2 Simple Spring Damper Combination . . . . . . . . . . . . . . . . . 66
4.5.3 General Dynamic Force Model . . . . . . . . . . . . . . . . . . . . . 68
4.5.3.1 Hydro-Mount . . . . . . . . . . . . . . . . . . . . . . . . . 69

5 Vertical Dynamics 72
5.1 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.2 Modelling Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.2.1 Full Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
5.2.2 Twodimensional Models . . . . . . . . . . . . . . . . . . . . . . . . 73
5.2.3 Simple Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
5.3 Basic Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
5.3.1 Natural Frequency and Damping Rate . . . . . . . . . . . . . . . . 76
5.3.2 Spring Rates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
5.3.2.1 Minimum Spring Rates . . . . . . . . . . . . . . . . . . . . 78
5.3.2.2 Nonlinear Springs . . . . . . . . . . . . . . . . . . . . . . . 79
5.3.3 Inuence of Damping . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.3.4 Optimal Damping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
5.3.4.1 Avoiding Overshoots . . . . . . . . . . . . . . . . . . . . . 81
5.3.4.2 Fast Approach to Steady State . . . . . . . . . . . . . . . 82
5.4 Sky Hook Damper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.4.1 Modelling Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
5.4.2 Eigenfrequencies and Damping Ratios . . . . . . . . . . . . . . . . . 88
5.4.3 Technical Realization . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.5 Nonlinear Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.5.1 Quarter Car Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.5.2 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92

6 Longitudinal Dynamics 94
6.1 Dynamic Wheel Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.1.1 Simple Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.1.2 Inuence of Grade . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.1.3 Aerodynamic Forces . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.2 Maximum Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.2.1 Tilting Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.2.2 Friction Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.3 Driving and Braking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.3.1 Single Axle Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6.3.2 Braking at Single Axle . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.3.3 Optimal Distribution of Drive and Brake Forces . . . . . . . . . . . 100
6.3.4 Dierent Distributions of Brake Forces . . . . . . . . . . . . . . . . 102
6.3.5 Anti-Lock-Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6.4 Drive and Brake Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

III
 6.4.1 Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
6.4.2 Equations of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.4.3 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
6.4.4 Driving and Braking . . . . . . . . . . . . . . . . . . . . . . . . . . 107
6.4.5 Brake Pitch Pole . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108

7 Lateral Dynamics 109

7.1 Kinematic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.1.1 Kinematic Tire Model . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.1.2 Ackermann Geometry . . . . . . . . . . . . . . . . . . . . . . . . . 109
7.1.3 Space Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
7.1.4 Vehicle Model with Trailer . . . . . . . . . . . . . . . . . . . . . . . 112
7.1.4.1 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . 112
7.1.4.2 Vehicle Motion . . . . . . . . . . . . . . . . . . . . . . . . 113
7.1.4.3 Entering a Curve . . . . . . . . . . . . . . . . . . . . . . . 114
7.1.4.4 Trailer Motions . . . . . . . . . . . . . . . . . . . . . . . . 115
7.1.4.5 Course Calculations . . . . . . . . . . . . . . . . . . . . . 116
7.2 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.2.1 Cornering Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7.2.2 Overturning Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
7.2.3 Roll Support and Camber Compensation . . . . . . . . . . . . . . . 121
7.2.4 Roll Center and Roll Axis . . . . . . . . . . . . . . . . . . . . . . . 124
7.2.5 Wheel Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
7.3 Simple Handling Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.3.1 Modeling Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.3.2 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.3.3 Tire Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7.3.4 Lateral Slips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
7.3.5 Equations of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.3.6 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.3.6.1 Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . 128
7.3.6.2 Low Speed Approximation . . . . . . . . . . . . . . . . . . 129
7.3.6.3 High Speed Approximation . . . . . . . . . . . . . . . . . 129
7.3.7 Steady State Solution . . . . . . . . . . . . . . . . . . . . . . . . . . 130
7.3.7.1 Side Slip Angle and Yaw Velocity . . . . . . . . . . . . . . 130
7.3.7.2 Steering Tendency . . . . . . . . . . . . . . . . . . . . . . 132
7.3.7.3 Slip Angles . . . . . . . . . . . . . . . . . . . . . . . . . . 133
7.3.8 Inuence of Wheel Load on Cornering Stiness . . . . . . . . . . . 133

8 Driving Behavior of Single Vehicles 136

8.1 Standard Driving Maneuvers . . . . . . . . . . . . . . . . . . . . . . . . . . 136
8.1.1 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 136
8.1.2 Step Steer Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
8.1.3 Driving Straight Ahead . . . . . . . . . . . . . . . . . . . . . . . . . 138

IV
 8.1.3.1 Random Road Prole . . . . . . . . . . . . . . . . . . . . 138
8.1.3.2 Steering Activity . . . . . . . . . . . . . . . . . . . . . . . 140
8.2 Coach with dierent Loading Conditions . . . . . . . . . . . . . . . . . . . 140
8.2.1 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
8.2.2 Roll Steering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
8.2.3 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 141
8.2.4 Step Steer Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
8.3 Dierent Rear Axle Concepts for a Passenger Car . . . . . . . . . . . . . . 143

V
1 Introduction

1.1 Literature
ˆ ATZ: Automobiltechnische Zeitschrift

ˆ Fachbuchgruppe Fahrwerktechnik:
Jörnsen Reimpell, Hrsg. Vogel Buchverlag Würzburg.

ˆ Dynamik der Kraftfahrzeuge: M. Mitschke, Bde. A,B,C; Springer-Verlag.

ˆ Mitschke, M.; Wallentowitz, H.: Dynamik der Kraftfahrzeuge. 4. Auage.

Springer-Verlag Berlin Heidelberg 2004.

ˆ Popp, K.; Schiehlen, W.: Fahrzeugdynamik. Teubner Stuttgart 1993.

ˆ Simulation von Kraftfahrzeugen: G. Rill, Vieweg-Verlag 1994.

ˆ Radführungen der Straÿenfahrzeuge: W. Matschinsky, Springer-Verlag.

ˆ Blundell, M.; Harty, D.: The Multibody System Approach to Vehicle Dynamics.
Elsevier Butterworth-Heinemann Publications, 2004.

ˆ Fundamentals of Vehicle Dynamics: Th., D. Gillespie, SAE, Inc.

ˆ ISO-Standards: (International Organisation for Standardization.)

z.B.: ISO 4138 Steady State Circular Test Procedure.

ˆ Kraftfahrtechnisches Handbuch: Robert Bosch GmbH (Hrsg.), 23. Au.,

Vieweg-Verlag.

ˆ Proceedings:

VDI-Tagungen: z.B.: Berechnung im Automobilbau.

SAE-Congress: (Society of Automotive Engineers).

FISITA: ( Féd. Internat. des Sociétés d'Ingénieurs de Techniques de l'Automobile).
IAVSD: (International Assosiation for Vehicle System Dynamics).

1
Vehicle Dynamics FH Regensburg, University of Applied Sciences

1.2 Terminology

1.2.1 Vehicle Dynamics

The expression `Vehicle Dynamics' encompasses the interaction of:

ˆ driver

ˆ vehicle

ˆ load

ˆ environment

Vehicle dynamics mainly deals with:

ˆ the improvement of active safety and driving comfort

ˆ the reduction of road destruction

In vehicle dynamics are employed:

ˆ computer calculations

ˆ test rig measurements

ˆ eld tests

In the following the interactions between the single systems and the problems with com-
puter calculations and/or measurements shall be discussed.

1.2.2 Driver

By various means the driver can interfere with the vehicle:

 
 steering wheel lateral dynamics 
  
accelerator pedal

 

 
 
driver brake pedal −→ vehicle

longitudinal dynamics
clutch

 


 
 

gear shift
  

The vehicle provides the driver with these information:

 
 vibrations: longitudinal, lateral, vertical 
vehicle sounds: motor, aerodynamics, tires −→ driver
instruments: velocity, external temperature, ...
 

2
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

The environment also inuences the driver:

 
 climate 
environment trac density −→ driver
track
 

The driver's reaction is very complex. To achieve objective results, an `ideal' driver is
used in computer simulations, and in driving experiments automated drivers (e.g. steering
machines) are employed.

Transferring results to normal drivers is often dicult, if eld tests are made with test
drivers. Field tests with normal drivers have to be evaluated statistically. Of course, the
driver's security must have absolute priority in all tests.

Driving simulators provide an excellent means of analyzing the behavior of drivers even
in limit situations without danger.

It has been tried to analyze the interaction between driver and vehicle with complex driver
models for some years.

1.2.3 Vehicle

The following vehicles are listed in the ISO 3833 directive:

ˆ motorcycles

ˆ passenger cars

ˆ busses

ˆ trucks

ˆ agricultural tractors

ˆ passenger cars with trailer

ˆ truck trailer / semitrailer

ˆ road trains

For computer calculations these vehicles have to be depicted in mathematically describable

substitute systems. The generation of the equations of motion, the numeric solution, as
well as the acquisition of data require great expenses. In times of PCs and workstations
computing costs hardly matter anymore.

At an early stage of development, often only prototypes are available for eld and/or
laboratory tests. Results can be falsied by safety devices, e.g. jockey wheels on trucks.

3
Vehicle Dynamics FH Regensburg, University of Applied Sciences

1.2.4 Load

Trucks are conceived for taking up load. Thus, their driving behavior changes.

mass, inertia, center of gravity
Load
dynamic behaviour (liquid load)

In computer calculations problems occur at the determination of the inertias and the
modeling of liquid loads.

Even the loading and unloading process of experimental vehicles takes some eort. When
carrying out experiments with tank trucks, ammable liquids have to be substituted with
water. Thus, the results achieved cannot be simply transferred to real loads.

1.2.5 Environment

The environment inuences primarily the vehicle:

 
road: irregularities, coecient of friction
Environment −→ vehicle
air: resistance, cross wind

but also aects the driver:

 
climate
environment −→ driver
visibility

Through the interactions between vehicle and road, roads can quickly be destroyed.

The greatest diculty with eld tests and laboratory experiments is the virtual impossi-
bility of reproducing environmental inuences.

The main problems with computer simulation are the description of random road irreg-
ularities and the interaction of tires and road as well as the calculation of aerodynamic
forces and torques.

1.3 Definitions

1.3.1 Reference frames

A reference frame xed to the vehicle and a ground-xed reference frame are used to
describe the overall motions of the vehicle, Figure 1.1. The ground-xed reference frame
with the axis x0 , y0 , z0 serves as an inertial reference frame. Within the vehicle-xed
reference frame the xF -axis points forward, the yF -axis to the left, and the zF -axis upward.

The wheel rotates around an axis which is xed to the wheel carrier. The reference frame
C is xed to the wheel carrier. In design position its axes xC , yC and zC are parallel to
the corresponding axis of vehicle-xed reference frame F.

4
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

zF
yF
xF

z0

zC y0
x0
yC

en xC eyR

Figure 1.1: Frames used in vehicle dynamics

The momentary position of the wheel is xed by the wheel center and the orientation of
the wheel rim center plane which is dened by the unit vector eyR into the direction of
the wheel rotation axis.

Finally, the normal vector en describes the inclination of the local track plane.

1.3.2 Toe and camber angle

front
δ δ

xF

vehicle
yF center
left right
wheel plane wheel
rear

Figure 1.2: Positive toe-in angle

According to the DIN 70 000 directive the angle δ between the vehicle center plane in
longitudinal direction and the intersection line of the tire center plane with the track
plane is named toe or toe-in angle. It will be positive, if the front part of the wheel is
oriented towards the vehicle center plane, Figure 1.2. Toe-in reduces the tendency of the
wheels to shimmy.

The camber angle γ is the angle between the wheel center plane and the local track normal
en . It will be positive, if the upper part of the wheel is inclined outwards, Figure 1.3. A
cambered wheel causes a non symmetric tire wear.

5
Vehicle Dynamics FH Regensburg, University of Applied Sciences

en en
γ γ
left right
zF wheel
wheel

yF

road

Figure 1.3: Positive camber angle

1.3.3 Design Position of Wheel Rotation Axis

The unit vector eyR

describes the wheel rotation axis. Its orientation with respect to the
∗
wheel carrier xed reference frame can be dened by the angles δ0 and γ0 or δ0 and γ0 ,
Fig. 1.4. In design position the corresponding axes of the frames C and F are parallel.

zC = zF

xC = xF

δ0 yC = yF
γ0*
γ0

eyR

Figure 1.4: Design position of wheel rotation axis

Then, for the left wheel we get

 
tan δ0
1
eyR,F = eyR,C = q  1  (1.1)
tan2 δ0 + 1 + tan2 γ0∗ − tan γ0∗

or  
sin δ0 cos γ0
eyR,F = eyR,C =  cos δ0 cos γ0  , (1.2)
− sin γ0
where δ0 yF -axis and the projection line of the wheel rotation
is the angle between the
axis into the xF - yF -plane, γ0∗ describes the angle between the yF -axis and the
the angle
0
projection line of the wheel rotation axis into the yF - zF -plane, whereas γ0 is the angle
between the wheel rotation axis eyR and its projection into the xF - yF -plane. Kinematics

6
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

and compliance test machines usually measure the angle γ0∗ . That is why, the automotive
industry mostly uses this angle instead of γ0 .
On a at and horizontal road where the track normal en points into the direction of
the vertical axes zC = zF the angles δ0 γ0
correspond with the toe angle δ and the
and
∗
camber angle γ0 . To specify the dierence between γ0 and γ0 the ratio between the third
and second component of the unit vector eyR is considered. The Equations 1.1 and 1.2
deliver
− tan γ0∗ − sin γ0 tan γ0
= or tan γ0∗ = . (1.3)
1 cos δ0 cos γ0 cos δ0
Hence, for small angles δ0  1 the dierence between the angles γ0 and γ0∗ is hardly
noticeable.

1.3.4 Steering Geometry

1.3.4.1 Kingpin

At the steered front axle, the McPherson-damper strut axis, the double wishbone axis,
and the multi-link wheel suspension or the enhanced double wishbone axis are mostly
used in passenger cars, Fig. 1.5 and Fig. 1.6.

zR

B
yR
M xR

kingpin axis A-B

Figure 1.5: Double wishbone wheel suspension

The wheel body rotates around the kingpin at steering motions. At the double wishbone
axis the ball joints A B , which determine the kingpin, are both xed to the wheel
and
body. Whereas the ball joint A is still xed to the wheel body at the standard McPherson
wheel suspension, the ball joint B is now xed to the vehicle body. At a multi-link axle the
kingpin is no longer dened by real joints. Here, as well as with the enhanced McPherson
wheel suspension, the kingpin changes its position relative to the wheel body at wheel
travel and steering motions.

7
Vehicle Dynamics FH Regensburg, University of Applied Sciences

B
zR zR

yR yR
xR M xR
M
A

kingpin axis A-B rotation axis

Figure 1.6: McPherson and multi-link wheel suspensions

1.3.4.2 Caster and Kingpin Angle

The unit vector eS describes the direction of the kingpin axis. Within the vehicle xed
reference frame F it can be xed by two angles. The caster angle ν denotes the angle
between the zF -axis and the projection line of eS into the xF -, zF -plane. In a similar
way the projection of eS into the yF -, zF -plane delivers the kingpin inclination angle σ ,
Fig. 1.7.

zF
zF eS σ

yF
xF

Figure 1.7: Kingpin and caster angle

At many axles the kingpin and caster angle can no longer be determined directly. Here, the
current rotation axis at steering motions, which can be taken from kinematic calculations
will deliver a virtual kingpin. The current values of the caster angle ν and the kingpin
inclination angle σ can be calculated from the components of the unit vector eS in the
direction of the kingpin, described in the vehicle xed reference frame

(1) (2)
−eS,F −eS,F
tan ν = (3)
and tan σ = (3)
, (1.4)
eS,F eS,F
(1) (2) (3)
where eS,F , eS,F , eS,F are the components of the unit vector eS,F expressed in the vehicle
xed reference frame F.

8
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

1.3.4.3 Caster, Steering Offset and Disturbing Force Lever

The contact point P, the local track normal en and the unit vectors ex and ey which
point into the direction of the longitudinal and lateral tire force result from the contact
geometry. The axle kinematics denes the kingpin line. In general, the point S where an
extension oft the kingpin line meets the road surface does not coincide with the contact
point P, Fig. 1.8. As both points are located on the local track plane, for the left wheel
the vector from S to P can be written as

rSP = −c ex + s ey , (1.5)

where c names the caster and s is the steering oset. Caster and steering oset will be
positive, if S is located in front of and inwards of P.

kingpin line

C d

en
ey ex
P
S
s c

Figure 1.8: Caster and steering oset

The distance d between the wheel center C and the king pin line represents the disturbing
force lever. It is an important quantity in evaluating the overall steering behavior.

9
2 Road

2.1 Modeling Aspects

Sophisticated road models provide the road height zR and the local friction coecient µL
at each point x, y , Fig. 2.1.

Road profile

z0 Segments
z(x,y)

y0 x0

µ(x,y)
Friction

Obstacle

Center Line L(s)

Figure 2.1: Sophisticated road model

The tire model is then responsible to calculate the local road inclination. By separating
the horizontal course description from the vertical layout and the surface properties of
the roadway almost arbitrary road layouts are possible.

Besides single obstacles or track grooves the irregularities of a road are of stochastic
nature. A vehicle driving over a random road prole mainly performs hub, pitch and roll
motions. The local inclination of the road prole also induces longitudinal and lateral
motions as well as yaw motions. On normal roads the latter motions have less inuence
on ride comfort and ride safety. To limit the eort of the stochastic description usually
simpler road models are used.

If the vehicle drives along a given path its momentary position can be described by the
path variable s = s(t). Hence, a fully two-dimensional road model can be reduced to a
parallel track model, Fig. 2.2.

10
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

z z2
z1
x
y

zR(x,y) s

z1(s)

Figure 2.2: Parallel track road model

Now, the road heights on the left and right track are provided by two one-dimensional
functions z1 = z1 (s) and z2 = z2 (s). Within the parallel track model no information
about the local lateral road inclination is available. If this information is not provided by
additional functions the impact of a local lateral road inclination to vehicle motions is not
taken into account.

For basic studies the irregularities at the left and the right track can considered to be
approximately the same, z1 (s) ≈ z2 (s). Then, a single track road model with zR (s) =
z1 (x) = z2 (x) can be used. Now, the roll excitation of the vehicle is neglected too.

2.2 Deterministic Profiles

2.2.1 Bumps and Potholes

Bumps and Potholes on the road are single obstacles of nearly arbitrary shape. Already
with simple rectangular cleats the dynamic reaction of a vehicle or a single tire to a
sudden impact can be investigated. If the shape of the obstacle is approximated by a
smooth function, like a cosine wave, then, discontinuities will be avoided. Usually the
obstacles are described in local reference frames, Fig. 2.3.

z
z
H H
B B
y
x y x
L L

Figure 2.3: Rectangular cleat and cosine-shaped bump

Then, the rectangular cleat is simply dened by

(
H if 0<x<L and − 12 B < y < 12 B
z(x, y) = (2.1)
0 else

11
Vehicle Dynamics FH Regensburg, University of Applied Sciences

and the cosine-shaped bump is given by

x 
  
1

2
H 1 − cos 2π if 0<x<L and − 21 B < y < 12 B
z(x, y) = L (2.2)
 0 else

where H, B andL denote height, width and length of the obstacle. Potholes are obtained
if negative values for the height (H < 0) are used.

2.2.2 Sine Waves

Using the parallel track road model, a periodic excitation can be realized by

z1 (s) = A sin (Ω s) , z2 (s) = A sin (Ω s − Ψ) , (2.3)

where s is the path variable, A denotes the amplitude, Ω the wave number, and the angle
Ψ describes a phase lag between the left and the right track. The special cases Ψ = 0 and
Ψ = π represent the in-phase excitation with z1 = z2 and the out of phase excitation with
z1 = −z2 .
If the vehicle runs with constant velocity ds/dt = v0 , the momentary position of the
vehicle is given by s = v0 t, where the initial position s = 0 at t = 0 was assumed. By
introducing the wavelength
2π
L = (2.4)
Ω
the term Ωs can be written as

2π 2π v0
Ωs = s= v0 t = 2π t = ω t . (2.5)
L L L
Hence, in the time domain the excitation frequency is given by f = ω/(2π) = v0 /L.
For most of the vehicles the rigid body vibrations are in between 0.5 Hz to 15 Hz . This
range is covered by waves which satisfy the conditions v0 /L ≥ 0.5 Hz and v0 /L ≤ 15 Hz .
For a given wavelength, lets say L = 4 m, the rigid body vibration of a vehicle are excited
min
if the velocity of the vehicle will be varied from v0 = 0.5 Hz ∗ 4 m = 2 m/s = 7.2 km/h
max
to v0 = 15 Hz ∗ 4 m = 60 m/s = 216 km/h. Hence, to achieve an excitation in the
whole frequency range with moderate vehicle velocities proles with dierent varying
wavelengths are needed.

2.3 Random Profiles

2.3.1 Statistical Properties

Road proles t the category of stationary Gaussian random processes. Hence, the irregu-
larities of a road can be described either by the prole itself zR = zR (s) or by its statistical
properties, Fig. 2.4.

12
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

zR
0.15 Gaussian
Realization density
0.10 function
[m] +σ
0.05
m Histogram
0
s
-0.05
−σ
-0.10

-0.15
-200 -150 -100 -50 0 50 100 [m] 150 200

Figure 2.4: Road prole and statistical properties

By choosing an appropriate reference frame, a vanishing mean value

ZX/2
1
m = E {zR (s)} = lim zR (s) ds = 0 (2.6)
X→∞ X
−X/2

can be achieved, where E {} denotes the expectation operator. Then, the Gaussian density
function which corresponds with the histogram is given by

zR2
1 −
p(zR ) = √ e 2σ 2 , (2.7)
σ 2π
where the deviation or the eective value σ is obtained from the variance of the process
zR = zR (s)
ZX/2
1
σ 2 = E zR2 (s) zR (s)2 ds .

= lim (2.8)
X→∞ X
−X/2

Alteration of σ eects the shape of the density function. In particular, the points of
inexion occur at ±σ . The probability of a value |z| < ζ is given by

Z+ζ z 2
1 −
P (±ζ) = √ e 2σ 2 dz . (2.9)
σ 2π
−ζ

In particular, one gets the values: P (±σ) = 0.683, P (±2σ) = 0.955, and P (±3σ) = 0.997.
Hence, the probability of a value |z| ≥ 3σ is 0.3%.
In extension to the variance of a random process the auto-correlation function is dened
by
ZX/2
1
R(ξ) = E {zR (s) zR (s+ξ)} = lim zR (s) zR (s+ξ) ds . (2.10)
X→∞ X
−X/2

13
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The auto-correlation function is symmetric, R(ξ) = R(−ξ), and it plays an important

part in the stochastic analysis. In any normal random process, as ξ increases the link
between zR (s) and zR (s+ξ) diminishes. For large values of ξ the two values are practically
unrelated. Hence, R(ξ → ∞) will tend to 0. In fact, R(ξ) is always less R(0), which
2
coincides with the variance σ of the process. If a periodic term is present in the process
it will show up in R(ξ).

Usually, road proles are characterized in the frequency domain. Here, the auto-correlation
function R(ξ) is replaced by the power spectral density (psd) S(Ω). In general, R(ξ) and
S(Ω) are related to each other by the Fourier transformation

Z∞ Z∞
1 −iΩξ 1
S(Ω) = R(ξ) e dξ and R(ξ) = S(Ω) eiΩξ dΩ , (2.11)
2π 2π
−∞ −∞

where i is the imaginary unit, and Ω in rad/m denotes the wave number. To avoid negative
wave numbers, usually a one-sided psd is dened. With

Φ(Ω) = 2 S(Ω) , if Ω≥0 and Φ(Ω) = 0 , if Ω<0, (2.12)

the relationship e±iΩξ = cos(Ωξ) ± i sin(Ωξ), and the symmetry property R(ξ) = R(−ξ)
Eq. (2.11) results in

Z∞ Z∞
2
Φ(Ω) = R(ξ) cos (Ωξ) dξ and R(ξ) = Φ(Ω) cos (Ωξ) dΩ . (2.13)
π
0 0

Now, the variance is obtained from

Z∞
2
σ = R(ξ = 0) = Φ(Ω) dΩ . (2.14)

In reality the psd Φ(Ω) will be given in a nite interval Ω1 ≤ Ω ≤ ΩN , Fig. 2.5. Then,

Φ N ∆Ω
∆Ω
Φ(Ωi)

Ω
Ω1 Ωi ΩN

Figure 2.5: Power spectral density in a nite interval

Eq. (2.14) can be approximated by a sum, which for N equal intervals will result in

N
X ΩN − Ω1
σ2 ≈ Φ(Ωi ) 4Ω with 4Ω = . (2.15)
i=1
N

14
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

2.3.2 Classification of Random Road Profiles

Road elevation proles can be measured point by point or by high-speed prolometers.

The power spectral densities of roads show a characteristic drop in magnitude with the
wave number, Fig. 2.6a. This simply reects the fact that the irregularities of the road may
amount to several meters over the length of hundreds of meters, whereas those measured
over the length of one meter are normally only some centimeter in amplitude.

Random road proles can be approximated by a psd in the form of

 −w
Ω
Φ (Ω) = Φ (Ω0 ) , (2.16)
Ω0

where, Ω = 2π/L in rad/m denotes the wave number and Φ0 = Φ (Ω0 ) in m2 /(rad/m)
describes the value of the psd at a the reference wave number Ω0 = 1 rad/m. The drop
in magnitude is modeled by the waviness w.

a) Measurements (country road) b) Range of road classes (ISO 8608)

10-3
Power spectral density Φ [m2/(rad/m)]

Φ0=256∗10−6
10-4

10-5
Class E
10-6 Φ0=1∗10−6

10-7

10-8
Class A

10-9 -2
10 10-1 100 101 102 10-2 10-1 100 101 102
Wave number Ω [rad/m] Wave number Ω [rad/m]

Figure 2.6: Road power spectral densities: a) Measurements, b) Classication

According to the international directive ISO 8608 typical road proles can be grouped
into classes from A to E. By setting the waviness to w = 2 each class is simply dened by
−6
its reference value Φ0 . Class A with Φ0 = 1 ∗ 10 m2 /(rad/m) characterizes very smooth
−6
highways, whereas Class E with Φ0 = 256 ∗ 10 m2 /(rad/m) represents rather rough
roads, Fig. 2.6b.

15
Vehicle Dynamics FH Regensburg, University of Applied Sciences

2.3.3 Realizations
2.3.3.1 Sinusoidal Approximation

A random prole of a single track can be approximated by a superposition of N →∞

sine waves
N
X
zR (s) = Ai sin (Ωi s − Ψi ) , (2.17)
i=1
where each sine wave is determined by its amplitude Ai and its wave number Ωi . By
dierent sets of uniformly distributed phase angles Ψi , i = 1(1)N in the range between
0 and 2π dierent proles can be generated which are similar in the general appearance
but dierent in details.

The variance of the sinusoidal representation is then given by

ZX/2 X
N
! N
!
1 X
σ 2 = lim Ai sin (Ωi s − Ψi ) Aj sin (Ωj s − Ψj ) ds . (2.18)
X→∞ X
i=1 j=1
−X/2

For i=j and for i 6= j dierent types of integrals are obtained. The ones for i=j can
be solved immediately

A2i
Z  
1
A2i 2


Jii = sin (Ωi s−Ψi ) ds = Ωi s−Ψi − sin 2 (Ωi s−Ψi ) . (2.19)
2Ωi 2
Using the trigonometric relationship

1 1
sin x sin y = cos(x−y) − cos(x+y) (2.20)
2 2
the integrals for i 6= j can be solved too

Z
Jij = Ai sin (Ωi s−Ψi ) Aj sin (Ωj s−Ψj ) ds
Z Z
1 1 (2.21)
= A i Aj cos (Ωi−j s − Ψi−j ) ds − Ai Aj cos (Ωi+j s − Ψi+j ) ds
2 2
1 A i Aj 1 Ai Aj
=− sin (Ωi−j s − Ψi−j ) + sin (Ωi+j s − Ψi+j )
2 Ωi−j 2 Ωi+j
where the abbreviations Ωi±j = Ωi±Ωj and Ψi±j = Ψi±Ψj were used. The sine and cosine
terms in Eqs. (2.19) and (2.21) are limited to values of ±1. Hence, Eq. (2.18) simply
results in
N N N
2 1 X   X/2 1 X   X/2 1X 2
σ = lim Jii −X/2 + lim Jij −X/2 = A . (2.22)
X→∞ X
i=1
X→∞ X
i,j=1
2 i=1 i
| {z } | {z }
N
X A 2 0
i
Ωi
i=1
2Ω i

16
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

On the other hand, the variance of a sinusoidal approximation to a random road prole
is given by Eq. (2.15). So, a road prole zR = zR (s) described by Eq. (2.17) will have a
given psd Φ(Ω) if the amplitudes are generated according to
p
Ai = 2 Φ(Ωi ) 4Ω , i = 1(1)N , (2.23)

and the wave numbers Ωi are chosen to lie at N equal intervals 4Ω.

Road profile z=z(s)

0.10
[m]
0.05
0
-0.05
-0.10
0 10 20 30 40 50 60 70 80 [m] 90 100

Figure 2.7: Realization of a country road

A realization of the country road with a psd of Φ0 = 10 ∗ 10−6 m2 /(rad/m) is shown in

Fig. 2.7. According to Eq. (2.17) the prole z = z(s) was generated by N = 200 sine waves
in the frequency range from Ω1 = 0.0628 rad/m to ΩN = 62.83 rad/m. The amplitudes
Ai , i = 1(1)N were calculated by Eq. (2.23) and the MATLABr function rand was used
to produce uniformly distributed random phase angles in the range between 0 and 2π .

2.3.3.2 Shaping Filter

The white noise process produced by random number generators has a uniform spectral
density, and is therefore not suitable to describe real road proles. But, if the white noise
process is used as input to a shaping lter more appropriate spectral densities will be
obtained. A simple rst order shaping lter for the road prole zR reads as

d
zR (s) = −γ zR (s) + w(s) , (2.24)
ds
where γ is a constant, and w(s) is a white noise process with the spectral density Φw .
Then, the spectral density of the road prole is obtained from

1 1 ΦW
ΦR = H(Ω) ΦW H T (−Ω) = ΦW = 2 , (2.25)
γ + iΩ γ − iΩ γ + Ω2
where Ω is the wave number, and H(Ω) is the frequency response function of the shaping
lter.

By setting ΦW = 10 ∗ 10−6 m2 /(rad/m) and γ = 0.01 rad/m the measured psd of a typical
country road can be approximated very well, Fig. 2.8.

The shape lter approach is also suitable for modeling parallel tracks. Here, the cross-
correlation between the irregularities of the left and right track have to be taken into
account too.

17
Vehicle Dynamics FH Regensburg, University of Applied Sciences

10-3

Power spectral density Φ [m2/(rad/m)]

Measurements
Shaping filter
10-4

10-5

10-6

10-7

10-8

10-9 -2
10 10-1 100 101 102
Wave number Ω [rad/m]

Figure 2.8: Shaping lter as approximation to measured psd

2.3.3.3 Two-Dimensional Model

The generation of fully two-dimensional road proles zR = zR (x, y) via a sinusoidal ap-
proximation is very laborious. Because a shaping lter is a dynamic system, the result-
ing road prole realizations are not reproducible. By adding band-limited white noise
processes and taking the momentary position x, y as seed for the random number gener-
ator a reproducible road prole can be generated.

z
y
x 50
45
1 40
0 35 m
-1 30
4 25
2 20
0 15
-2 10
-4 5
0

Figure 2.9: Two-dimensional road prole

By assuming the same statistical properties in longitudinal and lateral direction two-
dimensional proles, like the one in Fig. 2.9, can be obtained.

18
3 Tire

3.1 Introduction

3.1.1 Tire Development

Some important mile stones in the development of tires are shown in Table 3.1.

1839 Charles Goodyear: vulcanization

1845 Robert William Thompson: rst pneumatic tire
(several thin inated tubes inside a leather cover)
1888 John Boyd Dunlop: patent for bicycle (pneumatic) tires
1893 The Dunlop Pneumatic and Tyre Co. GmbH, Hanau, Germany
1895 André and Edouard Michelin: pneumatic tires for Peugeot
50 tire deations,
Paris-Bordeaux-Paris (720 Miles):
22 complete inner tube changes
1899 Continental: long-lived tires (approx. 500 Kilometer)
1904 Carbon added: black tires.
1908 Frank Seiberling: grooved tires with improved road traction
1922 Dunlop: steel cord thread in the tire bead
1943 Continental: patent for tubeless tires
1946 Radial Tire

Table 3.1: Milestones in tire development

Of course the tire development did not stop in 1946, but modern tires are still based on
this achievements.

3.1.2 Tire Composites

Tires are very complex. They combine dozens of components that must be formed, as-
sembled and cured together. And their ultimate success depends on their ability to blend
all of the separate components into a cohesive product that satises the driver's needs. A
modern tire is a mixture of steel, fabric, and rubber. The main composites of a passenger
car tire with an overall mass of 8.5 kg are listed in Table 3.2.

19
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Reinforcements: steel, rayon, nylon 16%

Rubber: natural/synthetic 38%
Compounds: carbon, silica, chalk, ... 30%
Softener: oil, resin 10%
Vulcanization: sulfur, zinc oxide, ... 4%
Miscellaneous 2%

Table 3.2: Tire composites: 195/65 R 15 ContiEcoContact, data from www.felge.de

3.1.3 Tire Forces and Torques

In any point of contact between the tire and the road surface normal and friction forces
are transmitted. According to the tire's prole design the contact patch forms a not
necessarily coherent area, Figure 3.1.

180 mm

140 mm

Figure 3.1: Tire footprint of a passenger car at normal loading condition:

Continental 205/55 R16 90 H, 2.5 bar, Fz = 4700 N

The eect of the contact forces can be fully described by a resulting force vector applied
at a specic point of the contact patch and a torque vector. The vectors are described in a
track-xed reference frame. The z -axis is normal to the track, the x-axis is perpendicular
to the z -axis and perpendicular to the wheel rotation axis eyR . Then, the demand for a
right-handed reference frame also xes the y -axis.
The components of the contact force vector are named according to the direction of the
axes, Figure 3.2.

A non symmetric distribution of the forces in the contact patch causes torques around
the x and y axes. A cambered tire generates a tilting torque Tx . The torque Ty includes
the rolling resistance of the tire. In particular, the torque around the z -axis is important
in vehicle dynamics. It consists of two parts,

Tz = TB + TS . (3.1)

20
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Fx longitudinal force

Fy lateral force

Fz vertical force or wheel load

eyR

Tx tilting torque
Fy
Ty rolling resistance torque
Ty
Tx
Tz self aligning and bore torque Tz
Fx Fz

Figure 3.2: Contact forces and torques

The rotation of the tire around the z -axis causes the bore torque MB . The self aligning
torque MS takes into account that ,in general, the resulting lateral force is not acting in
the center of the contact patch.

3.1.4 Measuring Tire Forces and Torques

To measure tire forces and torques on the road a special test trailer is needed, Figure
3.4. Here, the measurements are performed under real operating conditions. Arbitrary

Test trailer

tire
exact contact
real road

compensation wheel

test wheel

Figure 3.3: Layout of a tire test trailer

surfaces like asphalt or concrete and dierent environmental conditions like dry, wet or
icy are possible. Measurements with test trailers are quite cumbersome and in general
they are restricted to passenger car tires.

Indoor measurements of tire forces and torques can be performed on drums or on a at
bed, Figure 3.4.

21
Vehicle Dynamics FH Regensburg, University of Applied Sciences

tire

too small safety walk

contact area tire
coating

safety walk coating perfect contact

rotation
drum

tire

too large contact area

Figure 3.4: Drum and at bed tire test rig

On drum test rigs the tire is placed either inside or outside of the drum. In both cases
the shape of the contact area between tire and drum is not correct. That is why, one
can not rely on the measured self aligning torque. Due its simple and robust design, wide
applications including measurements of truck tires are possible.

The at bed tire test rig is more sophisticated. Here, the contact patch is as at as on the
road. But, the safety walk coating which is attached to the steel bed does not generate
the same friction conditions as on a real road surface.

Radial 205/50 R15, FN= 3500 N, dry asphalt

4000

3000

2000
Longitud force Fx [N]

Driving

1000

-1000
Braking

-2000

-3000

-4000
-40 -30 -20 -10 0 10 20 30 40
Longitudinal slip [%]

Figure 3.5: Typical results of tire measurements

Tire forces and torques are measured in quasi-static operating conditions. Hence, the mea-
surements for increasing and decreasing the sliding conditions usually result in dierent
graphs, Figure 3.5. In general, the mean values are taken as steady state results.

22
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

3.2 Contact Geometry

3.2.1 Basic Approach

The current position of a wheel in relation to the xed x0 -, y0 - z0 -system is given by the
wheel center M and the unit vector eyR in the direction of the wheel rotation axis, Figure
3.6.

rim γ
tire centre
plane e zR

M
M e yR wheel
e yR rMP carrier
en
ex
en b
P0 a
P0 ey P
z0
x0
P* z0 local road plane
x0 y0
0
road: z = z ( x , y ) y0
0

Figure 3.6: Contact geometry

The irregularities of the track can be described by an arbitrary function of two spatial
coordinates
z = z(x, y). (3.2)

At an uneven track the contact point P can not be calculated directly. At rst, one can
get an estimated value with the vector

rM P ∗ = −r0 ezB , (3.3)

where r0 is the undeformed tire radius, and ezB is the unit vector in the z -direction of the
body xed reference frame.

The position of this rst guess P∗ with respect to the earth xed reference frame x0 , y0 ,
z0 is determined by
x∗
 

r0P ∗ ,0 = r0M,0 + rM P ∗ ,0 =  y∗  , (3.4)

z∗

23
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where the vector r0M describes the position of the rim center M. Usually, the point P∗
does not lie on the track. The corresponding track point P0 follows from

 
x∗
r0P0 ,0 =  y∗ , (3.5)
 
∗ ∗
z (x , y )

where Eq. (3.2) was used to calculate the appropriate road height. In the point P0 the
track normal en is calculated, now. Then the unit vectors in the tire's circumferential
direction and lateral direction can be determined. One gets

eyR ×en
ex = and ey = en ×ex , (3.6)
| eyR ×en |

where eyR denotes the unit vector into the direction of the wheel rotation axis. Calculating
ex demands a normalization, as eyR not always being perpendicular to the track. The tire
camber angle
γ = arcsin eTyR en


(3.7)

describes the inclination of the wheel rotation axis against the track normal.

The vector from the rim center M to the track point P0 is split into three parts now

rM P0 = −rS ezR + a ex + b ey , (3.8)

where rS denotes the loaded or static tire radius, a, b are distances measured in circum-
ferential and lateral direction, and the radial direction is given by the unit vector

ezR = ex ×eyR (3.9)

which is perpendicular to ex and eyR . A scalar multiplication of Eq. (3.8) with en results
in
eTn rM P0 = −rS eTn ezR + a eTn ex + b eTn ey . (3.10)

As the unit vectors ex and ey are perpendicular to en Eq. (3.10) simplies to

eTn rM P0 = −rS eTn ezR . (3.11)

Hence, the static tire radius is given by

eTn rM P0
rS = − T . (3.12)
en ezR
The contact point P given by the vector

rM P = −rS ezR (3.13)

lies within the rim center plane. The transition from the point P0 to the contact point P
takes place according to Eq. (3.8) by the terms a ex and b ey perpendicular to the track

24
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

normal en . The track normal, however, was calculated in the point P0 . With an uneven
track the point P no longer lies on the track and can therefor no longer considered as
contact point.

With the newly estimated value P ∗ = P now the Eqs. (3.5) to (3.13) can be repeated
until the dierence between P and P0 is suciently small.
Tire models which can be simulated within acceptable time assume that the contact patch
is suciently at. At an ordinary passenger car tire, the contact area has at normal load
approximately the size of 15×20cm. It makes no sense to calculate a ctitious contact point
to fractions of millimeters, when later on the real track will be approximated by a plane
in the range of centimeters. If the track in the contact area is replaced by a local plane,
no further iterative improvements will be necessary for the contact point calculation.

3.2.2 Local Track Plane

A plane is given by three points. In order to get a good approximation to the local track
unevenness four point will be used to determine the local track normal. Using the initial
guess in Eq. (3.3) and the wheel rotation axis eyr the circumferential direction can be
estimated by
eyR ×ezB
e∗x = . (3.14)
| eyR ×ezB |
Similar to Eq. (3.4) four points are generated now

 
x∗i
r0Q∗i ,0 = r0M,0 + rM Q∗i ,0 =  yi∗  , i = 1(1)4 . (3.15)
 

zi∗

In order to sample the contact patch as good as possible the tire width b and the unloaded
tire radius r0 are used to place the points via

rM Q∗1 = λx r0 ex∗ − r0 ezB ,

rM Q∗2 = −λx r0 ex∗ − r0 ezB ,
(3.16)
rM Q∗3 = λy b eyR − r0 ezB ,
rM Q∗4 = −λy b eyR − r0 ezB

in the front, in the rear, in the left, and in the right of the contact patch.

According to Eq. (3.5) the corresponding points on the track are given by

 
x∗i
r0Qi ,0 = yi∗ , i = 1(1)4 . (3.17)
 
∗ ∗
z (xi , yi )

The calculation of the track normal is straight forward now, Figure 3.7. The vectors

25
Vehicle Dynamics FH Regensburg, University of Applied Sciences

eyR M

Q1*
rMP*
Q1 en Q3*
Q3
rQ2Q1 P*

P
Q4*
rQ3Q4 Q2*
Q4
Q2

Figure 3.7: Local track plane

rQ2 Q1 = r0Q1 −r0Q2 and rQ4 Q3 = r0Q3 −r0Q4 dene the local track inclination in longitudinal
and lateral direction. Hence, the local track normal is dened by

rQ2 Q1 ×rQ4 Q3
en = . (3.18)
| rQ2 Q1 ×rQ4 Q3 |
The unit vectors ex , ey in longitudinal and lateral direction are calculated from Eq. (3.6).
The mean value of the track points

1
r0P0 ,0 = (r0Q1 ,0 + r0Q2 ,0 + r0Q3 ,0 + r0Q4 ,0 ) (3.19)
4
∗
serves as rst improvement of the contact point, P → P0 . Finally, the corresponding
point P in the rim center plane is obtained by Eqs. (3.12) and (3.13). On rough roads
the point P not always is located on the track. But, together with the local track normal
it represents the local road plane very well. As in reality, sharp bends and discontinuities
which will occur at step- or ramp-sized obstacles are smoothed by this approach.

3.2.3 Tire Deflection

For a vanishing camber angle γ =0 the deected zone has a rectangular shape, Figure
3.8. Its area is given by
A0 = 4z b , (3.20)

where b is the width of the tire, and the tire deection is obtained by

4z = r0 − rS . (3.21)

Here, the width of the tire simply equals the width of the contact zone, wC = b.
On a cambered tire the deected zone of the tire cross section depends on the contact
situation. The magnitude of the tire ank radii

b b
rSL = rs + tan γ and rSR = rs − tan γ (3.22)
2 2

26
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

γ=0 γ =/ 0

eyR eyR
eyR γ γ

rSL rS rSR
r0 rSR
rS en en r0 rS r0
en
P P
∆z P
b b*
wC = b
wC wC

full contact partial contact

Figure 3.8: Tire deection

determines the shape of the deected zone.

The tire will be in full contact to the road if rSL ≤ r0 and rSR ≤ r0 hold. Then, the
deected zone has a trapezoidal shape with an area of

1
Aγ = (r0 −rSR + r0 −rSL ) b = (r0 − rS ) b . (3.23)
2
Equalizing the cross sections A0 = Aγ results in

4z = r0 − rS . (3.24)

Hence, at full contact the tire camber angle γ has no inuence to the vertical tire force.
But, due to
b
wC = (3.25)
cos γ
the width of the contact area increases with the tire camber angle.

The deected zone will change to a triangular shape if one of the ank radii exceeds the
undeected tire radius. Assuming rSL > r0 and rSR < r0 the area of the deected zone is
obtained by
1
Aγ = (r0 −rSR ) b∗ , (3.26)
2
where the width of the deected zone follows from

r0 −rSR
b∗ = . (3.27)
tan γ
Now, Eq. (3.26) reads as
1 (r0 −rSR )2
Aγ = . (3.28)
2 tan γ

27
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Equalizing the cross sections A0 = Aγ results in

2
1 r0 − rS + 2b tan γ
4z = . (3.29)
2 b tan γ
where Eq. (3.22) was used to express the ank radius rSR by the static tire radius rS , the
tire width b and the camber angle γ. Now, the width of the contact area is given by

b∗ r0 − rSR r0 − rS + 2b tan γ
wC = = = , (3.30)
cos γ tan γ cos γ sin γ
where the Eqs. (3.27) and (3.22) where used to simplify the expression. If tan γ and sin γ
are replaced by | tan γ | and | sin γ | then, the Eqs. (3.29) and (3.30) will hold for positive
and negative camber angles.

3.2.4 Length of Contact Patch

To approximate the length of the contact patch the tire deformation is split into two
parts, Figure 3.9. By 4zF and 4zB the average tire ank and the belt deformation are
measured. Hence, for a tire with full contact to the road

4z = 4zF + 4zB = r0 − rS (3.31)

will hold.

undeformed
Belt belt

Fz ∆zF

Rim
rS r0 r0

L L/2
∆zB ∆zB

Figure 3.9: Length of contact patch

Assuming both deections being equal will lead to

1
4zF ≈ 4zB ≈ 4z . (3.32)
2
Approximating the belt deection by truncating a circle with the radius of the undeformed
tire results in  2
L
+ (r0 − 4zB )2 = r02 . (3.33)
2

28
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

In normal driving situations the belt deections are small, 4zB  r0 . Hence, Eq. (3.33)
can be simplied and nally results in

L2 p
= 2 r0 4zB or L = 8 r0 4zB . (3.34)
4
Inspecting the passenger car tire footprint in Figure 3.1 leads to a contact patch length of
L ≈ 140mm. For this tire the radial stiness and the inated radius are specied with cR =
265 000 N/m and r0 = 316.9 mm. The overall tire deection can be estimated by 4z =
Fz /cR . At the load of Fz = 4700N the deection amounts to 4z = 4700N / 265 000N/m =
0.0177 m. Then, by approximating the belt deformation
p by the half of the tire deection,
the length of the contact patch will become L = 8 0.3169 m 0.0177/2 m = 0.1498 m =
150 mm which corresponds quite well with the length of the tire footprint.

3.2.5 Static Contact Point

Assuming that the pressure distribution on a cambered tire with full road contact cor-
responds with the trapezoidal shape of the deected tire area, the acting point of the
resulting vertical tire force FZ will be shifted from the geometric contact point P to the
static contact point Q, Figure 3.10.

γ
rS
en

ey P Q
r0-rSL A Fz r0-rSR
A
y
wC

Figure 3.10: Lateral deviation of contact point at full contact

The center of the trapezoidal area determines the lateral deviation yQ . By splitting the
area into a rectangular and a triangular section we will obtain

y  A  + y 4 A4
yQ = − . (3.35)
A
The minus sign takes into account that for positive camber angles the acting point will
move to the right whereas the unit vector ey dening the lateral direction points to the
left. The area of the whole cross section results from

1
A = (r0 −rSL + r0 −rSR ) wC , (3.36)
2

29
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where the width of the contact area wC is given by Eq. (3.25). Using the Eqs. (3.22) and
(3.24) the expression can be simplied to

A = 4z wC . (3.37)

As the center of the rectangular section is located on the center line which runs through the
geometric contact point, y = 0 will hold. The distance from the center of the triangular
section to the center line is given by

1 1 1
y4 = wC − wC = wC . (3.38)
2 3 6
Finally, the area of the triangular section is dened by

1 1 1
A4 = (r0 −rSR − (r0 −rSL )) wC = (rSL − rSR )) wC = (b tan γ) wC , (3.39)
2 2 2
where Eq. (3.22) was used to simplify the expression. Now, Eq. (3.35) can be written as

1 1
6
wC 2
b tan γ wC b tan γ b2 tan γ
yQ = − = − wC = − . (3.40)
4z wC 12 4z 12 4z cos γ
If the cambered tire has only a partial contact to the road then, according to the deection
area a triangular pressure distribution will be assumed, Figure 3.11.

γ
b/2

en

ey P Q
Fz
y

wC

Figure 3.11: Lateral deviation of contact point at partial contact

Now, the location of the static contact point Q is given by

 
1 b
yQ = ± wC − , (3.41)
3 2 cos γ
where the width of the contact area wC is determined by Eq. (3.30) and the term b/(2 cos γ)
describes the distance from the geometric contact point P to the outer corner of the
contact patch. The plus sign holds for positive and the minus sign for negative camber
angles.

The static contact point Q described by the vector

r0Q = r0P + yQ ey (3.42)

represents the contact patch much better than the geometric contact point P.

30
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

3.2.6 Contact Point Velocity

To calculate the tire forces and torques which are generated by friction the contact point
velocity will be needed. The static contact point Q given by Eq. (3.42) can be expressed
as follows
r0Q = r0M + rM Q , (3.43)

where M denotes the wheel center and hence, the vector rM Q describes the position of
static contact point Q relative to the wheel center M . The absolute velocity of the contact
point will be obtained from

v0Q,0 = ṙ0Q,0 = ṙ0M,0 + ṙM Q,0 , (3.44)

where ṙ0M,0 = v0M,0 denotes the absolute velocity of the wheel center. The vector rM Q
takes part on all those motions of the wheel carrier which do not contain elements of the
wheel rotation and it In addition, it contains the tire deection 4z normal to the road.
Hence, its time derivative can be calculated from

∗
ṙM Q,0 = ω0R,0 ×rM Q,0 + 4ż en,0 , (3.45)

∗
where ω0R is the angular velocity of the wheel rim without any component in the direction
of the wheel rotation axis, 4ż denotes the change of the tire deection, and en describes
the road normal. Now, Eq. (3.44) reads as

∗
v0Q,0 = v0M,0 + ω0R,0 ×rM Q,0 + 4ż en,0 . (3.46)

As the point Q lies on the track, v0Q,0 must not contain any component normal to the
track

∗
eTn,0 v0P,0 = 0 eTn,0 v0M,0 + ω0R,0 ×rM Q,0 + 4ż eTn,0 en,0 = 0 .


or (3.47)

As en,0 is a unit vector, eTn,0 en,0 = 1 will hold, and then, the time derivative of the tire
deformation is simply given by

∗
4ż = − eTn,0 v0M,0 + ω0R,0


×rM Q,0 . (3.48)

Finally, the components of the contact point velocity in longitudinal and lateral direction
are obtained from

∗
vx = eTx,0 v0Q,0 = eTx,0 v0M,0 + ω0R,0


×rM Q,0 (3.49)

and
∗
vy = eTy,0 v0P,0 = eTy,0 v0M,0 + ω0R,0


×rM Q,0 , (3.50)

where the relationships eTx,0 en,0 = 0 and eTy,0 en,0 = 0 were used to simplify the expressions.

31
Vehicle Dynamics FH Regensburg, University of Applied Sciences

deflected tire rigid wheel

Ω Ω

r0 r rD
S
∆ϕ vt ∆ϕ

x x

Figure 3.12: Dynamic rolling radius

3.2.7 Dynamic Rolling Radius

At an angular rotation of 4ϕ, assuming the tread particles stick to the track, the deected
tire moves on a distance of x, Figure 3.12.
With r0 as unloaded and rS = r0 − 4r as loaded or static tire radius

r0 sin 4ϕ = x (3.51)

and
r0 cos 4ϕ = rS (3.52)

hold.

If the motion of a tire is compared to the rolling of a rigid wheel, then, its radius rD will
have to be chosen so that at an angular rotation of 4ϕ the tire moves the distance

r0 sin 4ϕ = x = rD 4ϕ . (3.53)

Hence, the dynamic tire radius is given by

r0 sin 4ϕ
rD = . (3.54)
4ϕ
For 4ϕ → 0 one obtains the trivial solution rD = r0 .
At small, yet nite angular rotations the sine-function can be approximated by the rst
terms of its Taylor-Expansion. Then, Eq. (3.54) reads as

4ϕ − 16 4ϕ3
 
1
rD = r0 = r0 1 − 4ϕ2 . (3.55)
4ϕ 6
With the according approximation for the cosine-function
 
rS 1 rS
= cos 4ϕ = 1 − 4ϕ2 or
2
4ϕ = 2 1− (3.56)
r0 2 r0

32
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

one nally gets   

1 rS 2 1
r D = r0 1− 1− = r0 + rS . (3.57)
3 r0 3 3
Due to rS = rS (Fz ) the ctive radius rD depends on the wheel load Fz . Therefore, it is
called dynamic tire radius. If the tire rotates with the angular velocity Ω, then

v t = rD Ω (3.58)

will denote the average velocity at which the tread particles are transported through the
contact patch.

10
[mm]

0
◦ Measurements
-10
r -r
D 0
− TMeasy tire model

-20
0 2 4 6 8
Fz [kN]

Figure 3.13: Dynamic tire radius

In extension to Eq. (3.57), the dynamic tire radius is approximated in the tire model
TMeasy by
FzS
 
rD = λ r0 + (1 − λ) r0 − (3.59)
c
| {z 0 }
≈ rS
where the static tire radius rS = r0 − 4r has been approximated by using the linearized
S
tire deformation 4r = Fz /c0 . The parameter λ is modeled as a function of the wheel
load Fz  
Fz
λ = λN + ( λ2N − λN ) −1 , (3.60)
FzN
where λN and λ2N denote the values for the pay load Fz = FzN and the doubled pay load
Fz = 2FzN .
With the TMeasy parameters for a passenger car tire

vertical tire stiffness at fz=fz0 [N/m], 190000.

vertical tire stiffness at fz=2*fz0 [N/m], 206000.

coefficient for dynamic tire radius fz=fz0 [-], 0.375

coefficient for dynamic tire radius fz=2*fz0 [-], 0.750

the approximation of measured tire data can be done very well, Figure 3.13.

33
Vehicle Dynamics FH Regensburg, University of Applied Sciences

3.3 Forces and Torques caused by Pressure Distribution

3.3.1 Wheel Load

The vertical tire force Fz can be calculated as a function of the normal tire deection 4z
and the deection velocity 4ż

Fz = Fz (4z, 4ż) . (3.61)

Because the tire can only apply pressure forces to the road the normal force is restricted
to Fz ≥ 0. In a rst approximation Fz is separated into a static and a dynamic part

Fz = FzS + FzD . (3.62)

The static part is described as a nonlinear function of the normal tire deection

FzS = a1 4z + a2 (4z)2 . (3.63)

The constants a1 and a2 may be calculated from the radial stiness at nominal and double
payload
d FzS d FzS
cN = and c2N = . (3.64)
d 4z FzS =FzN d 4z FzS =2FzN

The derivative of Eq. (3.63) results in

d FzS
= a1 + 2 a2 4z . (3.65)
d 4z
From Eq. (3.63) one gets
p
−a1 ± a21 + 4a2 FzS
4z = . (3.66)
2a2
Because the tire deection is always positive, the minus sign in front of the square root
has no physical meaning, and can be omitted therefore. Hence, Eq. (3.65) can be written
as p !
d FzS −a1 + a21 + 4a2 FzS
q
= a1 + 2 a2 = a21 + 4a2 FzS . (3.67)
d 4z 2a2
Combining Eqs. (3.64) and (3.67 results in

q
cN = a21 + 4a2 FzN or c2N = a21 + 4a2 FzN ,
q (3.68)

c2N = a21 + 4a2 2FzN or c22N = a21 + 8a2 FzN

nally leading to
c22N − c2N
q
a1 = 2 c2N − c22N and a2 = . (3.69)
4 FzN

34
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Results for a passenger car and a truck tire are shown in Figure 3.14. The parabolic
approximation in Eq. (3.63) ts very well to the measurements. The radial tire stiness
of the passenger car tire at the payload of Fz = 3 200 N can be specied with c0 =
190 000N/m. The Payload Fz = 35 000 N and the stiness c0 = 1 250 000N/m of a truck
tire are signicantly larger.

Passenger Car Tire: 205/50 R15 Truck Tire: X31580 R22.5

10 100

8 80

60
Fz [kN]

Fz [kN]
4 40

2 20

0 0
0 10 20 30 40 50 0 20 40 60 80
∆z [mm] ∆z [mm]

Figure 3.14: Tire radial stiness: ◦ Measurements,  Approximation

The dynamic part is roughly approximated by

FzD = dR 4ż , (3.70)

where dR is a constant describing the radial tire damping, and the derivative of the tire
deformation 4ż is given by Eq. (3.48).

3.3.2 Tipping Torque

The lateral shift of the vertical tire force Fz from the geometric contact point P to the
static contact point Q is equivalent to a force applied in P and the tipping torque

Mx = Fz y (3.71)

acting around a longitudinal axis in P, Figure 3.15.

Note: Figure 3.15 shows a negative tipping torque. Because a positive camber angle moves
the contact point into the negative y -direction and hence, will generate a negative tipping
torque.

As long as the cambered tire has full contact to the road the lateral displacement y is
given by Eq. (3.40). Then, Eq. (3.71) reads as

b2 tan γ
Mx = − Fz . (3.72)
12 4z cos γ

35
Vehicle Dynamics FH Regensburg, University of Applied Sciences

en
∼ en
γ

ey P Q ey P
Fz Fz
y Tx

Figure 3.15: Tipping torque at full contact

If the wheel load is approximated by its linearized static part Fz ≈ cN 4z and small
camber angles |γ|  1 are assumed, then, Eq. (3.72) simplies to

b2 1
Mx = − cN 4z γ = − c N b2 γ , (3.73)
12 4z 12
1
where the term
12
c N b2 can be regarded as the tipping stiness of the tire.

en
ey Q
P y Fz

Figure 3.16: Cambered tire with partial contact

The use of the tipping torque instead of shifting the contact point is limited to those cases
where the tire has full or nearly full contact to the road. If the cambered tire has only
partly contact to the road, the geometric contact point P may even be located outside
the contact area whereas the static contact point Q is still a real contact point, Figure
3.16. In the following the static contact Q will be used as the contact point, because it
represents the contact area more precisely than the geometric contact point P.

3.3.3 Rolling Resistance

If a non-rotating tire has contact to a at ground the pressure distribution in the contact
patch will be symmetric from the front to the rear, Figure 3.17. The resulting vertical
force Fz is applied in the center C of the contact patch and hence, will not generate a
torque around the y -axis.
If the tire rotates tread particles will be stued into the front of the contact area which
causes a slight pressure increase, Figure 3.17. Now, the resulting vertical force is applied
in front of the contact point and generates the rolling resistance torque

ty = −Fz xR sign(Ω) , (3.74)

36
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

en rotating en
xR
ex non-rotating C ex C

Fz Fz

Figure 3.17: Pressure distribution at a non-rotation and rotation tire

where sign(Ω) assures that ty always acts against the wheel angular velocity Ω. The
distance xR from C to the working point of Fz usually is related to the unloaded tire
radius r0
xR
fR = . (3.75)
r0
The dimensionless rolling resistance coecient slightly increases with the traveling velocity
v of the vehicle
fR = fR (v) . (3.76)

Under normal operating conditions, 20 km/h < v < 200 km/h, the rolling resistance
coecient for typical passenger car tires is in the range of 0.01 < fR < 0.02.

The rolling resistance hardly inuences the handling properties of a vehicle. But it plays
a major part in fuel consumption.

3.4 Friction Forces and Torques

3.4.1 Longitudinal Force and Longitudinal Slip

To get a certain insight into the mechanism generating tire forces in longitudinal direction,
we consider a tire on a at bed test rig. The rim rotates with the angular velocity Ω and
the at bed runs with the velocity vx . The distance between the rim center and the at
bed is controlled to the loaded tire radius corresponding to the wheel load Fz , Figure 3.18.
A tread particle enters at the timet = 0 the contact patch. If we assume adhesion between
the particle and the track, then the top of the particle will run with the bed velocity vx
and the bottom with the average transport velocity vt = rD Ω. Depending on the velocity
dierence 4v = rD Ω − vx the tread particle is deected in longitudinal direction

u = (rD Ω − vx ) t . (3.77)

The time a particle spends in the contact patch can be calculated by

L
T = , (3.78)
rD |Ω|

37
Vehicle Dynamics FH Regensburg, University of Applied Sciences

rD Ω

vx
Ω

rD
u
vx
L
u max

Figure 3.18: Tire on at bed test rig

where L denotes the contact length, and T >0 is assured by |Ω|.

The maximum deection occurs when the tread particle leaves the contact patch at the
time t=T
L
umax = (rD Ω − vx ) T = (rD Ω − vx ) . (3.79)
rD |Ω|
The deected tread particle applies a force to the tire. In a rst approximation we get

Fxt = ctx u , (3.80)

where ctx represents the stiness of one tread particle in longitudinal direction.

On normal wheel loads more than one tread particle is in contact with the track, Figure
3.19a. The number p of the tread particles can be estimated by

L
p = , (3.81)
s+a
where s is the length of one particle and a denotes the distance between the particles.

a) b)
L L

s a cxt * u
cut * u max

Figure 3.19: a) Particles, b) Force distribution,

Particles entering the contact patch are undeformed, whereas the ones leaving have the
maximum deection. According to Eq. (3.80), this results in a linear force distribution

38
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

versus the contact length, Figure 3.19b. The resulting force in longitudinal direction for
p particles is given by
1 t
Fx = p c umax . (3.82)
2 x
Using the Eqs. (3.81) and (3.79) this results in

1 L t L
Fx = cx (rD Ω − vx ) . (3.83)
2 s+a rD |Ω|
A rst approximation of the contact length L was calculated in Eq. (3.34). Approximating
1
the belt deformation by 4zB ≈ F /c results in
2 z R
Fz
L2 ≈ 4 r 0 , (3.84)
cR
where cR denotes the radial tire stiness, and nonlinearities and dynamic parts in the tire
deformation were neglected. Now, Eq. (3.82) can be written as

r0 ctx rD Ω − v x
Fx = 2 Fz . (3.85)
s + a cR rD |Ω|
The nondimensional relation between the sliding velocity of the tread particles in lon-
S
gitudinal direction vx = vx − rD Ω and the average transport velocity rD |Ω| form the
longitudinal slip
−(vx − rD Ω)
sx = . (3.86)
rD |Ω|
The longitudinal force Fx is proportional to the wheel load Fz and the longitudinal slip
sx in this rst approximation
Fx = k Fz sx , (3.87)

where the constant k summarizes the tire properties r0 , s, a, ctx and cR .

Eq. (3.87) holds only as long as all particles stick to the track. At moderate slip values
the particles at the end of the contact patch start sliding, and at high slip values only the
parts at the beginning of the contact patch still stick to the road, Figure 3.20.

small slip values moderate slip values high slip values

Fx = k * Fz* s x Fx = Fz * f ( s x ) Fx = FG
L L L

t t t
Fx <= FH t t Fx = FG
Fx = FH
adhesion sliding
adhesion sliding

Figure 3.20: Longitudinal force distribution for dierent slip values

The resulting nonlinear function of the longitudinal force Fx

versus the longitudinal slip
sx can be dened by the parameters initial inclination (driving stiness) dFx0 , location
sM M S
x and magnitude of the maximum Fx , start of full sliding sx and the sliding force Fx ,
S

Figure 3.21.

39
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Fx
adhesion
M sliding
Fx
S
Fx dFx0

sM
x sSx sx

Figure 3.21: Typical longitudinal force characteristics

3.4.2 Lateral Slip, Lateral Force and Self Aligning Torque

Similar to the longitudinal slip sx , given by Eq. (3.86), the lateral slip can be dened by

−vyS
sy = , (3.88)
rD |Ω|
where the sliding velocity in lateral direction is given by

vyS = vy (3.89)

and the lateral component of the contact point velocity vy follows from Eq. (3.50).

As long as the tread particles stick to the road (small amounts of slip), an almost linear
distribution of the forces along the length L of the contact patch appears. At moderate
slip values the particles at the end of the contact patch start sliding, and at high slip
values only the parts at the beginning of the contact patch stick to the road, Figure 3.22.

small slip values moderate slip values large slip values

Fy = k * Fz * s y Fy = Fz * f ( s y ) Fy = FG
adhesion
adhesion

Fy

sliding

Fy
L
L

L
Fy n

sliding

Figure 3.22: Lateral force distribution over contact patch

The nonlinear characteristics of the lateral force versus the lateral slip can be described
0 M M
by the initial inclination (cornering stiness) dFy , the location sy and the magnitude Fy

40
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

of the maximum, the beginning of full sliding sSy , and the magnitude FyS of the sliding
force.

The distribution of the lateral forces over the contact patch length also denes the point
of application of the resulting lateral force. At small slip values this point lies behind
the center of the contact patch (contact point P). With increasing slip values it moves
forward, sometimes even before the center of the contact patch. At extreme slip values,
when practically all particles are sliding, the resulting force is applied at the center of the
contact patch.

The resulting lateral force Fy with the dynamic tire oset or pneumatic trail n as a lever
generates the self aligning torque

TS = −n Fy . (3.90)

The lateral force Fy as well as the dynamic tire oset are functions of the lateral slip
sy . Typical plots of these quantities are shown in Figure 3.23. Characteristic parameters

n/L
adhesion
(n/L)0
adhesion/sliding
Fy full sliding
adhesion/
adhesion
M sliding s0y sSy sy
Fy full sliding
S
Fy dF0y MS adhesion

adhesion/sliding

full sliding

sM
y sSy sy s0y sSy sy

Figure 3.23: Typical plot of lateral force, tire oset and self aligning torque

0 M
of the lateral force graph are initial inclination (cornering stiness) dFy , location sy and
M S S
magnitude of the maximum Fy , begin of full sliding sy , and the sliding force Fy .

The dynamic tire oset has been normalized by the length of the contact patch L. The
0 S
initial value (n/L)0 as well as the slip values sy and sy suciently characterize the graph.

3.4.3 Wheel Load Influence

The resistance of a real tire against deformations has the eect that with increasing
wheel load the distribution of pressure over the contact area becomes more and more
uneven. The tread particles are deected just as they are transported through the contact
area. The pressure peak in the front of the contact area cannot be used, for these tread
particles are far away from the adhesion limit because of their small deection. In the rear
of the contact area the pressure drop leads to a reduction of the maximally transmittable

41
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Longitudinal force Fx Lateral force Fy

Fz = 3.0 kN Fz = 6.0 kN Fz = 3.0 kN Fz = 6.0 kN
dFx0 = 70 kN dFx0 = 220 kN dFy0 = 72 kN dFy0 = 130 kN
sM
x = 0.160 sM
x = 0.120 sM
y = 0.180 sM
y = 0.200

FxM = 2.90 kN FxM = 5.60 kN FyM = 2.85 kN FyM = 5.40 kN

sSx = 0.500 sSx = 0.500 sSy = 0.500 sSy = 0.700
FxS = 2.65 kN FxS = 5.10 kN FyS = 2.80 kN FyS = 5.30 kN

Table 3.3: Characteristic tire data with degressive wheel load inuence

friction force. With rising imperfection of the pressure distribution over the contact area,
the ability to transmit forces of friction between tire and road lessens.

In practice, this leads to a degressive inuence of the wheel load on the characteristic
curves of longitudinal and lateral forces. In order to respect this fact in a tire model, the
N N
characteristic data for two nominal wheel loads Fz and 2 Fz are given in Table 3.3.
0 0 M M
From this data the initial inclinations dFx , dFy , the maximal forces Fx , Fx and the
S M
sliding forces Fx , Fy for arbitrary wheel loads Fz are calculated by quadratic functions.
For the maximum longitudinal force it reads as
 
Fz N Fz
 
1 M 1 M
FxM (Fz ) M N N M N
= N 2 Fx (Fz )− 2 Fx (2Fz ) − Fx (Fz )− 2 Fx (2Fz ) N . (3.91)
Fz Fz

6 6

4 4
Fy [kN]
Fx [kN]

2 2

0 0

-2 -2

-4 -4

-6 -6
-0.4 -0.2 0 0.2 0.4 -20 -10 0 10 20
sx [-] α [deg]

Figure 3.24: Longitudinal and lateral force characteristics: Fz = 1.8, 3.2, 4.6, 5.4, 6.0 kN

The location of the maxima sM M

x , sy , and the slip values, sSx , sSy , at which full sliding
appears, are dened as linear functions of the wheel load Fz . For the location of the

42
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

maximum longitudinal force this results in

  F 
z
sM
x (Fz ) = sM N
x (Fz ) + sM N
x (2Fz ) − sM N
x (Fz ) −1 . (3.92)
FzN

With the numeric values from Tab. 3.3 a slight shift of the maxima with an increasing
wheel load is also modeled, Figure 3.24.

3.4.4 Two-Dimensional Tire Characteristics

The longitudinal force as a function of the longitudinal slip Fx = Fx (sx ) and the lateral
force depending on the lateral slip Fy = Fy (sy ) can be dened by their characteristic
0 0 M M
parameters initial inclination dFx , dFy , location sx , sy and magnitude of the maximum
Fx , Fy as well as sliding limit sx , sy and sliding force FxS , FyS , Figure 3.25. During
M M S S

general driving situations, e.g. acceleration or deceleration in curves, longitudinal sx and

lateral slip sy appear simultaneously.

Fx S
M
Fx Fx

S
0
Fy M
dF x Fy Fy
sx
dF 0
sSx FS F(s)
FM
sM Fy
x
Fx
sy
0
dF y
sSy
sM
y
sS
s ϕ
sy sM
sx

Figure 3.25: Generalized tire characteristics

The longitudinal slip sx and the lateral slip sy can vectorally be added to a generalized
slip
s 2  2
sx sy
s = + , (3.93)
ŝx ŝy
where the slips sx and sy were normalized by appropriate weighting factors ŝx and ŝy .

43
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Similar to the graphs of the longitudinal and lateral forces the graph of the generalized tire
0 M M S S
force is dened by the characteristic parameters dF , s , F , s and F . The parameters
are calculated from the corresponding values of the longitudinal and lateral force

q
2
dF 0
= (dFx0 ŝx cos ϕ)2 + dFy0 ŝy sin ϕ ,
s 2  M 2
sM x
sy
sM = cos ϕ + sin ϕ ,
ŝx ŝy
q
2 (3.94)
FM = (FxM cos ϕ)2 + FyM sin ϕ ,
s 2  S 2
sSx sy
sS = cos ϕ + sin ϕ ,
ŝx ŝy
q
2
FS = (FxS cos ϕ)2 + FyS sin ϕ ,

where the slip normalization have also to be considered at the initial inclination. The
angular functions
sx /ŝx sy /ŝy
cos ϕ = and sin ϕ = (3.95)
s s
grant a smooth transition from the characteristic curve of longitudinal to the curve of
◦
lateral forces in the range of ϕ = 0 to ϕ = 90 .

The function F = F (s) is now described in intervals by a broken rational function, a cubic
S
polynomial, and by the sliding force F

σ s


 sM dF 0  M
, σ= M , 0 ≤ s ≤ sM ;
s s


1 + σ σ + dF 0 M − 2



 F
F (s) = s − sM (3.96)
M M S 2 M S

 F − (F − F ) σ (3 − 2 σ) , σ = , s < s ≤ s ;



 sS − sM
FS , s > sS .



0
When dening the curve parameters, one just has to make sure that the condition dF ≥
M
2 FsM is fullled, because otherwise the function has a turning point in the interval 0 <
s ≤ sM .
Now, the longitudinal and the lateral force follow from the according projections in lon-
gitudinal and lateral direction

Fx = F cos ϕ and Fy = F sin ϕ . (3.97)

Hence, within TMeasy the one-dimensional characteristics are automatically converted to

a two-dimensional combination characteristics, Figure 3.26.

44
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

30
3
20
2
1 10
F [kN]

F [kN]
0 0
y

y
-1 -10
-2
-20
-3
-30
-4 -2 0 2 4 -20 0 20
F [kN] F [kN]
x x

|sx | = 1, 2, 4, 6, 10, 15 %; |α| = 1, 2, 4, 6, 10, 14◦

Figure 3.26: Two-dimensional tire characteristics at Fz = 3.2 kN / Fz = 35 kN

3.4.5 Different Friction Coefficients

The tire characteristics are valid for one specic tire road combination only. Hence, dif-
ferent tire road combinations will demand for dierent model parameter.

4000 µF
µL /µ0
3000
Fy [N]

0.2
2000
0.4
1000
0 0.6

-1000 0.8

-2000 1.0

-3000
-4000 Fz = 3.2 kN
-0.5 0 0.5
sy [-]

Figure 3.27: Lateral force characteristics for dierent friction coecients

If only the coecient of friction is changed a simple but eective adaption of given model
data is possible. A reduced or changed friction coecient mainly inuences the maximum
force and the sliding force, whereas the initial inclination will remain unchanged. So, by
setting

µL M µL M µL S µL S
sM → s , FM → F , sS → s , FS → F , (3.98)
µ0 µ0 µ0 µ0

45
Vehicle Dynamics FH Regensburg, University of Applied Sciences

the essential tire model parameter which are valid for the friction coecient µ0 are ad-
justed to the new friction coecient µL . The result of this simple approach is shown in
Figure 3.27.

If the road model provides not only the roughness information z = fR (x, y) but also the
local friction coecient [z, µL ] = fR (x, y) then, braking on µ-split maneuvers can easily
be simulated.

3.4.6 Self Aligning Torque

According to Eq. (3.90) the self aligning torque can be calculated via the dynamic tire
oset. The dynamic tire oset n can be normalized by the length L of the contact area,
nN = n/L. It mainly depends on the lateral slip sy . The normalized tire oset starts
at sy = 0 with an initial value (n/L)0 . It tends to zero, n/L → 0 at large slip values,
sy ≥ sSy . Sometimes the dynamic tire oset overshoots to negative values before it reaches
0 S
zero again. This behavior can be modeled by introducing the parameter sy < sy , Figure
3.28.

n/L n/L
(n/L)0 (n/L)0

s0y sSy sy s0y sy

Figure 3.28: Normalized tire oset with and without overshoot

In order to achieve a simple and smooth approximation of the normalized tire oset versus
0
the lateral slip, a linear and a cubic function are overlayed in the rst section sy ≤ sy


|sy |
(1−w) (1−s) + w 1 − (3−2s) s2 |sy | ≤ s0y and s = 0
 





 sy
n n  0  S
|sy | − sy sy − |sy |
2
= − (1−w) s0y < |sy | ≤ sSy (3.99)
L L 0 s 0 s S − s0
y y y




0 |sy | > sSy



where the factor

s0y
w= S (3.100)
sy
0
weights the linear and the cubic function according to the values of the parameter sy
S 0 S
and sy . No overshoot occurs for sy = sy . Here, w = 1 and (1 − w) = 0 will produce a
cubic transition from n/L = (n/L)0 to n/L = 0 with vanishing inclinations at sy = 0 and
sy = s0y .

46
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

The characteristic curve parameters, which are used for the description of the dynamic
tire oset, are at rst approximation not wheel load dependent. Similar to the description
of the characteristic curves of longitudinal and lateral force, here also the parameters for
single and double pay load are given. The calculation of the parameters of arbitrary wheel
loads is done similar to Eq. (3.92) by linear inter- or extrapolation.

200

150

Tz [Nm]
100
Tire oset parameter Fz
50
Fz = 3.0 kN Fz = 6.0 kN
0
(n/L)0 = 0.170 (n/L)0 = 0.190
-50

s0y = 0.200 s0y = 0.220 -100

sE
y = 0.420 sE
y = 0.400
-150

-200
-30 -20 -10 0 10 20 30
α [deg]

Figure 3.29: Self aligning torque: Fz = 1.5, 3.0, 4.5, 6.0, 7.5 kN

The value of (n/L)0 can be estimated very well. At small values of lateral slip sy ≈ 0 one
gets at rst approximation a triangular distribution of lateral forces over the contact area
length cf. Figure 3.22. The working point of the resulting force (dynamic tire oset) is
then given by
1
n(Fz → 0, sy = 0) = L. (3.101)
6
The value n = 16 L can only serve as reference point, for the uneven distribution of pressure
in longitudinal direction of the contact area results in a change of the deexion prole
and the dynamic tire oset.

The self aligning torque in Figure 3.29 has been calculated with the tire parameter from
Table 3.3. The degressive inuence of the wheel load on the self aligning torque can be
seen here as well.

With the parameters for the description of the tire oset it has been assumed that at
N
the payload Fz = Fz the related tire oset reaches the value of (n/L)0 = 0.17 ≈ 1/6 at
sy = 0. The slip value s0y , at which the tire oset passes the x-axis, has been estimated.
Usually the value is somewhat higher than the position of the lateral force maximum.
S
With rising wheel load it moves to higher values. The values for sy are estimated too.

47
Vehicle Dynamics FH Regensburg, University of Applied Sciences

3.4.7 Camber Influence

At a cambered tire, Figure 3.30, the angular velocity of the wheel Ω has a component
normal to the road
Ωn = Ω sin γ . (3.102)

en γ

rim Fy = Fy (s y ): Parameter γ
4000
centre
plane
γ
3000
2000
eyR Ωn
Ω 1000
0
-1000
ex rD |Ω| -2000
-3000
vγ(ξ)
yγ(ξ)
ξ -4000
ey -0.5 0 0.5

Figure 3.30: Cambered tire Fy (γ) at Fz = 3.2 kN and γ = 0◦ , 2◦ , 4◦ , 6◦ , 8◦

Now, the tread particles in the contact patch possess a lateral velocity which depends on
their position ξ and is provided by

L ξ
vγ (ξ) = −Ωn , = −Ω sin γ ξ , −L/2 ≤ ξ ≤ L/2 . (3.103)
2 L/2
At the contact point it vanishes whereas at the end of the contact patch it takes on the
same value as at the beginning, however, pointing into the opposite direction. Assuming
that the tread particles stick to the track, the deection prole is dened by

ẏγ (ξ) = vγ (ξ) . (3.104)

The time derivative can be transformed to a space derivative

d yγ (ξ) d ξ d yγ (ξ)
ẏγ (ξ) = = rD |Ω| (3.105)
dξ dt dξ
where rD |Ω| denotes the average transport velocity. Now, Eq. (3.104) can be written as

d yγ (ξ) d yγ (ξ) Ω sin γ L ξ

rD |Ω| = −Ω sin γ ξ or = − , (3.106)
dξ dξ rD |Ω| 2 L/2

48
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

where L/2 was used to achieve dimensionless terms. Similar to the lateral slip sy which
is dened by Eq. (3.88) we can introduce a camber slip now

−Ω sin γ L
sγ = . (3.107)
rD |Ω| 2

Then, Eq. (3.106) simplies to

d yγ (ξ) ξ
= sγ . (3.108)
dξ L/2

The shape of the lateral displacement prole is obtained by integration

 2
1 L ξ
y γ = sγ + C. (3.109)
2 2 L/2
1


The boundary condition y ξ = L = 0 can be used to determine the integration constant
2
C . One gets
1 L
C = −sγ . (3.110)
2 2
Then, Eq. (3.109) reads as

"  2 #
1 L ξ
yγ (ξ) = −sγ 1− . (3.111)
2 2 L/2

The lateral displacements of the tread particles caused by a camber slip are compared
now with the ones caused by pure lateral slip, Figure 3.31. At a tire with pure lateral

a) camber slip b) lateral slip

y y yy(ξ)
yγ(ξ) _
_ yy
yγ

-L/2 0 L/2 ξ -L/2 0 L/2 ξ

Figure 3.31: Displacement proles of tread particles

slip each tread particle in the contact patch possesses the same lateral velocity which
results in dyy /dξ rD |Ω| = vy , ẏy was
where according to Eq. (3.105) the time derivative
transformed to the space derivative dyy /dξ . Hence, the deection prole is linear, and
reads as yy = vy /(rD |Ω|) ξ = −sy ξ , where the denition in Eq. (3.88) was used to
introduce the lateral slip sy . Then, the average deection of the tread particles under
pure lateral slip is given by
L
ȳy = −sy . (3.112)
2

49
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The average deection of the tread particles under pure camber slip is obtained from

ZL/2 "  2 #
1 L 1 x 1 L
ȳγ = −sγ 1− dξ = − sγ . (3.113)
2 2 L L/2 3 2
−L/2

A comparison of Eq. (3.112) with Eq. (3.113) shows, that by using

1
sγy = sγ (3.114)
3
the lateral camber slip sγ can be converted to an equivalent lateral slip sγy .
In normal driving conditions, the camber angle and thus, the lateral camber slip are
limited to small values. So, the lateral camber force can be approximated by

Fyγ ≈ dFy0 sγy . (3.115)

If the  global inclination dFy = Fy /sy is used instead of the initial inclination dFy0 , one
gets the camber inuence on the lateral force as shown in Figure 3.30.

The camber angle inuences the distribution of pressure in the lateral direction of the
contact patch, and changes the shape of the contact patch from rectangular to trape-
zoidal. Thus, it is extremely dicult, if not impossible, to quantify the camber inuence
with the aid of such simple models. But, it turns out that this approach is quit a good
approximation.

3.4.8 Bore Torque

In particular during steering motions the angular velocity of the wheel

∗
ω0W = ω0R + Ω eyR (3.116)

has a component in direction of the track normal en

ωn = eTn ω0W 6= 0 . (3.117)

Then, a very complicated deection prole of the tread particles in the contact patch
occurs. However, by a simple approach the resulting bore torque can be approximated
quite good by the parameter of the generalized tire force characteristics.

At rst, the complex shape of a tire's contact patch is approximated by a circle, Figure
3.32. By setting  
1 L B 1
R = + = (L + B) (3.118)
2 2 2 4
the radius of the circle can be adjusted to the length L and the width B of the actual
contact patch. The integration over the whole circle area results in the bore torque
Z
1
TB = F r dA , (3.119)
A A

50
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

ex
B
ϕ
dϕ
dr
F
r
L
ey ωn R

circular
normal shape of contact patch approximation

Figure 3.32: Bore torque approximation

where F
denotes the force transmitted by the patch element dA, and A is the area of the
2
circle. With dA = r dϕ dr and A = R π Eq. (3.119) reads as

ZR Z2π
1
TB = 2 F r rdϕ dr (3.120)
R π
0 0

which immediately results in

 R 2π
Z ZR
1  2
TB = 2 F r r dr = 2 F r2 dr . (3.121)
R π R
0 0 0

For small slip values the force transmitted in the patch element can be approximated by

F = F (s) ≈ dF 0 s (3.122)

where s denotes the slip of the patch element, and dF 0 is the initial inclination of the
generalized tire force characteristics. Similar to Eqs. (3.86 and (3.88) we dene

−r ωn
s = (3.123)
rD |Ω|
where r ωn describes the sliding velocity in the patch element, and rD and Ω denote the
dynamic tire radius and the angular velocity of the wheel.

Now, Eq. (3.121) reads as

ZR
2 −r ωn 2
TB = 2 dF 0 r dr (3.124)
R rD |Ω|
0

51
Vehicle Dynamics FH Regensburg, University of Applied Sciences

which nally results in

ZR
2 ωn 2 ωn R4 1 − R ωn
TB = − 2 dF 0 3
r dr = − 2 dF 0
= R dF 0 (3.125)
R rD |Ω| R rD |Ω| 4 2 rD |Ω|
0

where
− R ωn
sB = (3.126)
rD |Ω|
can be considered as bore slip. Via the initial inclination dF 0 the bore torque TB takes
the actual tire properties into account.

The bore torque calculated by Eq. (3.125) is only a rst approximation. At large bore slips
S
the generalized tire force F is limited to the sliding force F . Then, Eq. (3.121) changes
to
ZR
2 2 S R3 2 S
TBmax = 2 F S r2 dr = F = F R. (3.127)
R R2 3 3
0

Due to the generalized sliding force FS the maximum bore torque TBmax depends on the
tire properties and the actual friction value. Now, the bore torque is given by

1 − R ωn 2 S
TB = R dF 0 with |TB | ≤ F R (3.128)
2 rD |Ω| 3
where according to Eq. (3.118) the circle radius R can be replaced by the length L and
the width B of the contact patch.

3.4.9 Typical Tire Characteristics

passenger car tire truck tire

6
40
4
20
2
F [kN]

F [kN]

0 0 10 kN
x
x

1.8 kN 20 kN
-2 3.2 kN -20 30 kN
4.6 kN 40 kN
-4 5.4 kN 50 kN
-40
-6
-40 -20 0 20 40 -40 -20 0 20 40
sx [%] sx [%]

Figure 3.33: Longitudinal force: ◦ Meas., − TMeasy

The tire model TMeasy which is based on this approach, can be used for passenger car
tires as well as for truck tires. It approximates the characteristic curves Fx = Fx (sx ),

52
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

passenger car truck

6
40
4
20
2

F [kN]
Fy [kN]

0 0

y
1.8 kN 10 kN
-2 3.2 kN -20 20 kN
4.6 kN 30 kN
-4
6.0 kN 40 kN
-40
-6
150 1500

100 1000

50 500

[Nm]
[Nm]

0 0
z
Mz

M
-50 1.8 kN -500
3.2 kN 18.4 kN
-100 -1000 36.8 kN
4.6 kN
6.0 kN 55.2 kN
-150 -1500
-20 -10 0 10 20 -20 -10 0 10 20
α [o] α [o]

Figure 3.34: Lateral force and self aligning torque: ◦ Meas., − TMeasy

Fy = Fy (α) and Mz = Mz (α) quite well  even for dierent wheel loads Fz , Figures 3.33
and 3.34.

When experimental tire values are missing, the model parameters can be pragmatically
estimated by adjustment of the data of similar tire types. Furthermore, due to their phys-
ical signicance, the parameters can subsequently be improved by means of comparisons
between the simulation and vehicle testing results as far as they are available.

53
4 Suspension System

4.1 Purpose and Components

The automotive industry uses dierent kinds of wheel/axle suspension systems. Important
criteria are costs, space requirements, kinematic properties, and compliance attributes.

The main purposes of a vehicle suspension system are

ˆ carry the car and its weight,

ˆ maintain correct wheel alignment,

ˆ control the vehicle’s direction of travel,

ˆ keep the tires in contact with the road,

ˆ reduce the eect of shock forces.

Vehicle suspension systems consist of

ˆ guiding elements:

control arms, links,

struts,

leaf springs,

ˆ force elements:

coil spring, torsion bar, air spring, leaf spring,

anti-roll bar,

damper,

bushings, hydro-mounts,

ˆ tires.

Tires are air springs that support the total weight of the vehicle. The air spring action of
the tire is very important to the ride quality and safe handling of the vehicle.

54
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

4.2 Some Examples

4.2.1 Multi Purpose Systems

The double wishbone suspension, the McPherson suspension and the multi-link suspension
are multi purpose wheel suspension systems, Fig. 4.1.

Figure 4.1: Double wishbone, McPherson and multi-link suspension

They are used as steered front or non steered rear axle suspension systems. These sus-
pension systems are also suitable for driven axles.

In a McPherson suspension the spring is mounted with an inclination to the strut axis.
Thus, bending torques at the strut, which cause high friction forces, can be reduced.

leaf springs

links

Figure 4.2: Solid axles guided by leaf springs and links

At pickups, trucks, and busses solid axles are used often. They are guided either by leaf
springs or by rigid links, Fig. 4.2. Solid axles tend to tramp on rough roads.

Leaf-spring-guided solid axle suspension systems are very robust. Dry friction between
the leafs leads to locking eects in the suspension. Although the leaf springs provide
axle guidance on some solid axle suspension systems, additional links in longitudinal and
lateral direction are used. Thus, the typical wind-up eect on braking can be avoided.

Solid axles suspended by air springs need at least four links for guidance. In addition to
a good driving comfort air springs allow level control too.

55
Vehicle Dynamics FH Regensburg, University of Applied Sciences

4.2.2 Specific Systems

The semi-trailing arm, the short-long-arm axle (SLA), and the twist beam axle suspension
are suitable only for non-steered axles, Fig. 4.3.

Figure 4.3: Specic wheel/axles suspension systems

The semi-trailing arm is a simple and cheap design which requires only few space. It is
mostly used for driven rear axles.

The short-long-arm axle design allows a nearly independent layout of longitudinal and
lateral axle motions. It is similar to the central control arm axle suspension, where the
trailing arm is completely rigid and hence, only two lateral links are needed.

The twist beam axle suspension exhibits either a trailing arm or a semi-trailing arm
characteristic. It is used for non driven rear axles only. The twist beam axle provides
enough space for spare tire and fuel tank.

4.3 Steering Systems

4.3.1 Requirements

The steering system must guarantee easy and safe steering of the vehicle. The entirety
of the mechanical transmission devices must be able to cope with all loads and stresses
occurring in operation.

56
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

In order to achieve a good maneuverability a maximum steering angle of approx. 30◦ must
be provided at the front wheels of passenger cars. Depending on the wheel base, busses
◦
and trucks need maximum steering angles up to 55 at the front wheels.

Recently some companies have started investigations on `steer by wire' techniques.

4.3.2 Rack and Pinion Steering

Rack-and-pinion is the most common steering system of passenger cars, Fig. 4.4. The rack
may be located either in front of or behind the axle. Firstly, the rotations of the steering

uR δS
pinion
nk
drag li
rack steering
wheel
δ1 box δ2
and
wheel
body

Figure 4.4: Rack and pinion steering

wheel δS are transformed by the steering box to the rack travel uR = uR (δS ) and then via
the drag links transmitted to the wheel rotations δ1 = δ1 (uR ), δ2 = δ2 (uR ). Hence, the
overall steering ratio depends on the ratio of the steering box and on the kinematics of
the steering linkage.

4.3.3 Lever Arm Steering System

steering box
ste
e ring 2
lev lever
er 1 ring
stee
δL drag link 2
drag link 1

δ1 δ2
wheel and
wheel body

Figure 4.5: Lever arm steering system

Using a lever arm steering system Fig. 4.5, large steering angles at the wheels are possible.
This steering system is used on trucks with large wheel bases and independent wheel

57
Vehicle Dynamics FH Regensburg, University of Applied Sciences

suspension at the front axle. Here, the steering box can be placed outside of the axle
center.

Firstly, the rotations of the steering wheel δS are transformed by the steering box to the
rotation of the steer levers δL = δL (δS ). The drag links transmit this rotation to the wheel
δ1 = δ1 (δL ), δ2 = δ2 (δL ). Hence, the overall steering ratio again depends on the ratio of
the steering box and on the kinematics of the steering linkage.

4.3.4 Drag Link Steering System

At solid axles the drag link steering system is used, Fig. 4.6. The rotations of the steering

g
steerin
leve r steer box
δL (90o rotated)
O

wheel
and steering link
wheel
body

δ1 δ2

drag link

Figure 4.6: Drag link steering system

wheel δS are transformed by the steering box to the rotation of the steering lever arm
δL = δL (δS ) and further on to the rotation of the left wheel, δ1 = δ1 (δL ). The drag link
transmits the rotation of the left wheel to the right wheel, δ2 = δ2 (δ1 ). The steering ratio
is dened by the ratio of the steering box and the kinematics of the steering link. Here,
the ratio δ2 = δ2 (δ1 ) given by the kinematics of the drag link can be changed separately.

4.3.5 Bus Steer System

In busses the driver sits more than 2 m in front of the front axle. In addition, large steering
angles at the front wheels are needed to achieve a good manoeuvrability. That is why,
more sophisticated steering systems are needed, Fig. 4.7. The rotations of the steering
wheel δS are transformed by the steering box to the rotation of the steering lever arm
δL = δL (δS ). The left lever arm is moved via the steering link δA = δA (δL ). This motion
is transferred by a coupling link to the right lever arm. Finally, the left and right wheels
are rotated via the drag links, δ1 = δ1 (δA ) and δ2 = δ2 (δA ).

58
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

steer
ing leve
r
δL

steering box
steering link

left
lever
arm

drag link δA coupl.

link

δ1 δ2
wheel and
wheel body

Figure 4.7: Typical bus steering system

4.4 Standard Force Elements

4.4.1 Springs

Springs support the weight of the vehicle. In vehicle suspensions coil springs, air springs,
torsion bars, and leaf springs are used, Fig. 4.8.

Coil spring FS
u

u FS Air spring
u
u Torsion bar
Leaf spring
FS
FS

Figure 4.8: Vehicle suspension springs

Coil springs, torsion bars, and leaf springs absorb additional load by compressing. Thus,
the ride height depends on the loading condition. Air springs are rubber cylinders lled
with compressed air. They are becoming more popular on passenger cars, light trucks, and
heavy trucks because here the correct vehicle ride height can be maintained regardless of
the loading condition by adjusting the air pressure.

59
Vehicle Dynamics FH Regensburg, University of Applied Sciences

∆L
FS
FS
u
LF
L FS0 c

c L0
u
FS

Figure 4.9: Linear coil spring and general spring characteristics

A linear coil spring may be characterized by its free length LF and the spring stiness c,
Fig. 4.9. The force acting on the spring is then given by

FS = c LF − L , (4.1)

where L denotes the actual length of the spring. Mounted in a vehicle suspension the spring
has to support the corresponding chassis weight. Hence, the spring will be compressed to
the conguration length L0 < L F . Now, Eq. (4.1) can be written as

= c LF − L0 + c u = FS0 + c u ,



FS = c LF − (L0 − u) (4.2)

where FS0 is the spring preload and u describes the spring displacement measured from
the spring's conguration length.

In general the spring force FS can be dened by a nonlinear function of the spring dis-
placement u
FS = FS (u) . (4.3)

Now, arbitrary spring characteristics can be approximated by elementary functions, like

polynomials, or by tables which are then inter- and extrapolated by linear functions or
cubic splines.

The complex behavior of leaf springs and air springs can only be approximated by simple
nonlinear spring characteristics, FS = FS (u). For detailed investigations sophisticated or
even dynamic spring models have to be used.

4.4.2 Damper

Dampers are basically oil pumps, Fig. 4.10. As the suspension travels up and down, the
hydraulic uid is forced by a piston through tiny holes, called orices. This slows down
the suspension movement.

60
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Remote Gas Chamber Remote Oil Ch.

Rebound Ch. Remote orifice

FD Piston FD
Compression
Piston orifice Chamber
v

Figure 4.10: Principle of a mono-tube damper

Today twin-tube and mono-tube dampers are used in vehicle suspension systems. Dynamic
damper models compute the damper force via the uid pressure applied to each side of
the piston. The change in uid pressures in the compression and rebound chambers are
calculated by applying the conservation of mass.

In standard vehicle dynamics applications simple characteristics

FD = FD (v) (4.4)

are used to describe the damper force FD as a function of the damper velocity v. To
obtain this characteristics the damper is excited with a sinusoidal displacement signal
u = u0 sin 2πf t. By varying the frequency in several steps from f = f0 to f = fE dierent
force displacement curves FD = FD (u) are obtained, Fig. 4.11. By taking the peak values
of the damper force at the displacement u = u0 which corresponds with the velocity
v = ±2πf u0 the characteristics FD = FD (v) is generated now. Here, the rebound cycle is
associated with negative damper velocities.

FD = FD(u) FD = FD(v)
1000
FD [N]

0
f0 Compression
-1000

-2000

Rebound
-3000 fE

-4000
-0.06 -0.04 -0.02 0 0.02 0.04 0.06 -1.6 -1.2 -0.8 -0.4 0 0.4 0.8 1.2 1.6
u [m] v [m/s]

Figure 4.11: Damper characteristics generated from measurements

Typical passenger car or truck dampers will have more resistance during its rebound cycle
then its compression cycle.

61
Vehicle Dynamics FH Regensburg, University of Applied Sciences

4.4.3 Rubber Elements

Force elements made of natural rubber or urethane compounds are used in many locations
on the vehicle suspension system, Fig. 4.12. Those elements require no lubrication, isolate
minor vibration, reduce transmitted road shock, operate noise free, oer high load carrying
capabilities, and are very durable.

Topmount

Stop

Subframe mounts
Control arm
bushings

Figure 4.12: Rubber elements in vehicle suspension

During suspension travel, the control arm bushings provide a pivot point for the control
arm. They also maintain the exact wheel alignment by xing the lateral and vertical
location of the control arm pivot points. During suspension travel the rubber portion of
the bushing must twist to allow control arm movement. Thus, an additional resistance to
suspension movement is generated.

Bump and rebound stops limit the suspension travel. The compliance of the topmount
avoids the transfer of large shock forces to the chassis. The subframe mounts isolate the
suspension system from the chassis and allow elasto-kinematic steering eects of the whole
axle.

It turns out, that those elastic elements can hardly be described by simple spring and
damper characteristics, FS = FS (u) and FD = FD (v), because their stiness and damping
properties change with the frequency of the motion. Here, more sophisticated dynamic
models are needed.

62
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

4.5 Dynamic Force Elements

4.5.1 Testing and Evaluating Procedures

The eect of dynamic force elements is usually evaluated in the frequency domain. For
this, on test rigs or in simulation the force element is excited by sine waves

xe (t) = A sin(2π f t) , (4.5)

with dierent frequencies f0 ≤ f ≤ fE and amplitudes Amin ≤ A ≤ Amax . Starting at

t = 0, the system will usually be in a steady state condition after several periods t ≥ nT ,
where T = 1/f and n = 2, 3, . . . have to be chosen appropriately. Due to the nonlinear
system behavior the system response is periodic, F (t + T ) = F (T ), where T = 1/f , yet
not harmonic. That is why, the measured or calculated force F will be approximated
within one period n T ≤ t ≤ (n + 1)T , by harmonic functions as good as possible

F (t) ≈ α sin(2π f t) + β cos(2π f t) . (4.6)

|{z} | {z }
measured/ rst harmonic approximation
calculated
The coecients α and β can be calculated from the demand for a minimal overall error

(n+1)T
Z
1  2
α sin(2π f t)+β cos(2π f t) − F (t) dt −→ M inimum . (4.7)
2
nT

The dierentiation of Eq. (4.7) with respect to α and β yields two linear equations as
necessary conditions

(n+1)T
Z  
α sin(2π f t)+β cos(2π f t) − F (t) sin(2π f t) dt = 0
nT
(4.8)
(n+1)T
Z  
α sin(2π f t)+β cos(2π f t) − F (t) cos(2π f t) dt = 0
nT

with the solutions

R R R R
F sin dt cos2 dt − F cos dt sin cos dt
α =
sin2 dt cos2 dt − 2 sin cos dt
R R R
, (4.9)
F cos dt sin2 dt − F sin dt sin cos dt
R R R R
β =
sin2 dt cos2 dt − 2 sin cos dt
R R R

where the integral limits and arguments of sine and cosine no longer have been written.
Because it is integrated exactly over one period nT ≤ t ≤ (n + 1)T , for the integrals in
Eq. (4.9)
T T
sin2 dt =
R R R
sin cos dt = 0 ; ; cos2 dt = (4.10)
2 2

63
Vehicle Dynamics FH Regensburg, University of Applied Sciences

hold, and as solution

Z Z
2 2
α = F sin dt , β = F cos dt (4.11)
T T
remains. However, these are exactly the rst two coecients of a Fourier-"-Approximation.

The rst order harmonic approximation in Eq. (4.6) can now be written as

F (t) = F̂ sin (2π f t + Ψ) (4.12)

where amplitude F̂ and phase angle Ψ are given by

p β
F̂ = α2 + β 2 and tan Ψ = . (4.13)
α

A simple force element consisting of a linear spring with the stiness c and a linear damper
with the constant d in parallel would respond with

F (t) = c xe + d ẋe = c A sin 2πf t + d 2πf A cos 2πf t . (4.14)

Here, amplitude and phase angle are given by

d 2πf A d
q
F̂ = A c2 + (2πf d)2 and tan Ψ = = 2πf . (4.15)
cA c

c 6= 0 and d = 0 is characterized by F̂ = A c and

Hence, the response of a pure spring,
tan Ψ = 0 or Ψ = 0, whereas a pure damper response with c = 0 and d 6= 0 results in
F̂ = 2πf dA and tan Ψ → ∞ or Ψ = 90◦ . Hence, the phase angle Ψ which is also called
the dissipation angle can be used to evaluate the damping properties of the force element.
The dynamic stiness, dened by
F̂
cdyn = (4.16)
A
is used to evaluate the stiness of the element.

In practice the frequency response of a system is not determined punctually, but continu-
ously. For this, the system is excited by a sweep-sine. In analogy to the simple sine-function

xe (t) = A sin(2π f t) , (4.17)

where the period T = 1/f appears as pre-factor at dierentiation

2π
ẋe (t) = A 2π f cos(2π f t) = A cos(2π f t) . (4.18)
T
A generalized sine-function can be constructed, now. Starting with

xe (t) = A sin(2π h(t)) , (4.19)

64
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

the time derivative results in

ẋe (t) = A 2π ḣ(t) cos(2π h(t)) . (4.20)

In the following we demand that the function h(t) generates periods fading linearly in
time, i.e:
1 1
ḣ(t) = = , (4.21)
T (t) p−qt
where p>0 and q>0 are constants yet to determine. Eq. (4.21) yields

1
h(t) = − ln(p − q t) + C . (4.22)
q
The initial condition h(t = 0) = 0 xes the integration constant

1
C = ln p . (4.23)
q
With Eqs. (4.23) and (4.22) Eq. (4.19) results in a sine-like function

 2π p 
xe (t) = A sin ln , (4.24)
q p−qt
which is characterized by linear fading periods.

The important zero values for determining the period duration lie at

1 p p
ln = 0, 1, 2, or = en q , mit n = 0, 1, 2, (4.25)
q p − q tn p − q tn
and
p
tn = (1 − e−n q ) , n = 0, 1, 2, . (4.26)
q
The time dierence between two zero values yields the period

p
Tn = tn+1 − tn = (1−e−(n+1) q − 1+e−n q )
q , n = 0, 1, 2, . (4.27)
p −n q −q
Tn = e (1 − e )
q
For the rst (n = 0) and last (n = N ) period one nds

p
T0 = (1 − e−q )
q . (4.28)
p
TN = (1 − e−q ) e−N q = T0 e−N q
q
With the frequency range to investigate, given by the initial f0 and nal frequency fE ,
the parameters q and the ratio q/p can be calculated from Eq. (4.28)

1
1 fE q n hf iN o
E
q = ln , = f0 1 − , (4.29)
N f0 p f0

65
Vehicle Dynamics FH Regensburg, University of Applied Sciences

with N xing the number of frequency intervals. The passing of the whole frequency range
then takes the time
1 − e−(N +1) q
tN +1 = . (4.30)
q/p
Hence, to test or simulate a force element in the frequency range from 0.1Hz to f = 100Hz
with N = 500 intervals will only take 728 s or 12min.

4.5.2 Simple Spring Damper Combination

Fig. 4.13 shows a simple dynamic force element where a linear spring with the stiness c
and a linear damper with the damping constant d are arranged in series.

c d

s u

Figure 4.13: Spring and damper in series

The displacements of the force element and the spring itself are described by u and s.
Then, the the forces acting in the spring and damper are given by

FS = c s and FD = d (u̇ − ṡ) . (4.31)

The force balance FD = FS results in a linear rst order dierential equation for the
spring displacement s

d d
d (u̇ − ṡ) = c s or ṡ = −s + u̇ , (4.32)
c c
where the ratio between the damping coecient d and the spring stiness c acts as time
constant, T = d/c. Hence, this force element will responds dynamically to any excitation.
The steady state response to a harmonic excitation

u(t) = u0 sin Ωt respectively u̇ = u0 Ω cos Ωt (4.33)

can be calculated easily. The steady state response will be of the same type as the exci-
tation. Inserting
s∞ (t) = u0 (a sin Ωt + b cos Ωt) (4.34)

into Eq. (4.32) results in

d d
u0 (aΩ cos Ωt − bΩ sin Ωt) = − u0 (a sin Ωt + b cos Ωt) + u0 Ω cos Ωt . (4.35)
c | {z } | {z } c | {z }
ṡ∞ s∞ u̇

66
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Collecting all sine and cosine terms we obtain two equations

d d d
− u0 bΩ = −u0 a and u0 a Ω = −u0 b + u0 Ω (4.36)
c c c
which can be solved for the two unknown parameter

Ω2 c Ω
a = and b = . (4.37)
Ω2 + (c/d)2 d Ω + (c/d)2
2

Hence, the steady state force response reads as

Ω  c 
FS = c s∞ = c u0 2 Ω sin Ωt + cos Ωt (4.38)
Ω + (c/d)2 d
which can be transformed to

FS = F̂S sin (Ωt + Ψ) (4.39)

where the force magnitude F̂S and the phase angle Ψ are given by

c u0 Ω p c u0 Ω c/d
F̂S = 2 2 Ω2 + (c/d)2 = p and Ψ = arctan . (4.40)
Ω + (c/d) Ω + (c/d)2
2 Ω

The dynamic stiness cdyn = F̂S /u0 and the phase angle Ψ are plotted in Fig. 4.14 for
dierent damping values.

400
cdyn 4
300
[N/mm] 3 2 c = 400 N/mm
200
1 c
100

0 d
100

Ψ [o] d1 = 1000 N/(m/s)

d2 = 2000 N/(m/s)
50
2 1 d1 = 3000 N/(m/s)
3
4 d2 = 4000 N/(m/s)
0
0 20 40 60 80 100
f [Hz]

Figure 4.14: Frequency response of a spring damper combination

With increasing frequency the spring damper combination changes from a pure damper
◦
performance, cdyn → 0 and Ψ ≈ 90 to a pure spring behavior, cdyn ≈ c and Ψ → 0. The
frequency range, where the element provides stiness and damping is controlled by the
value for the damping constant d.

67
Vehicle Dynamics FH Regensburg, University of Applied Sciences

4.5.3 General Dynamic Force Model

To approximate the complex dynamic behavior of bushings and elastic mounts dierent
spring damper models can be combined. A general dynamic force model is constructed
by N parallel force elements, Fig. 4.15. The static load is carried by a single spring with
the stiness c0 or an arbitrary nonlinear force characteristics F0 = F0 (u).

d1 FM
F1 d2 FM
F2 dN FM
FN
c0
s1 s2 sN
c1 c2 cN

Figure 4.15: Dynamic force model

Within each force element the spring acts in serial to parallel combination of a damper
and a dry friction element. Now, even hysteresis eects and the stress history of the force
element can be taken into account.

The forces acting in the spring and damper of force element i are given by

FSi = −ci si and FDi = di (ṡi − u̇) , (4.41)

were u and si describe the overall element and the spring displacement.

As long as the absolute value of the spring force FSi is lower than the maximum friction
M
force FF the damper friction combination will not move at all

u̇ − ṡi = 0 for |FSi | ≤ FFM . (4.42)

In all other cases the force balance

FSi = FDi ± FFM (4.43)

holds. Using Eq. 4.41 the force balance results in

di (ṡi − u̇) = FSi ∓ FFM (4.44)

which can be combined with Eq. 4.42 to


 F + FFM FSi < −FFM
 Si


di ṡi = di u̇ for −FFM ≤ FSi ≤ +FFM (4.45)


F − F M

+FFM < FSi
Si F

68
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

where according to Eq. 4.41 the spring force is given by FSi = −ci si .
In extension to this linear approach nonlinear springs and dampers may be used. To derive
all the parameters an extensive set of static and dynamic measurements is needed.

4.5.3.1 Hydro-Mount

For the elastic suspension of engines in vehicles very often specially developed hydro-
mounts are used. The dynamic nonlinear behavior of these components guarantees a
good acoustic decoupling but simultaneously provides sucient damping.

xe
main spring

chamber 1
cT cF
__ c__
T
membrane 2 uF 2

ring channel MF
dF
__ dF
__
2 2
chamber 2

Figure 4.16: Hydro-mount

Fig. 4.16 shows the principle and mathematical model of a hydro-mount. At small de-
formations the change of volume in chamber 1 is compensated by displacements of the
membrane. When the membrane reaches the stop, the liquid in chamber 1 is pressed
through a ring channel into chamber 2. The ratio of the chamber cross section to the ring
channel cross section is very large. Thus the uid is moved through the ring channel at
very high speed. This results in remarkable inertia and resistance forces (damping forces).

The force eect of a hydro-mount is combined from the elasticity of the main spring and
the volume change in chamber 1.

With uF labeling the displacement of the generalized uid mass MF ,

FH = cT xe + FF (xe − uF ) (4.46)

holds, where the force eect of the main spring has been approximated by a linear spring
with the constant cT .

69
Vehicle Dynamics FH Regensburg, University of Applied Sciences

With MF R as the actual mass in the ring channel and the cross sections A K , AR of
chamber and ring channel the generalized uid mass is given by

 A 2
K
MF = MF R . (4.47)
AR
The uid in chamber 1 is not being compressed, unless the membrane can evade no longer.
With the uid stiness cF and the membrane clearance sF , one gets
  

 c F (x e − u F ) + s F (xe − uF ) < −sF

FF (xe − uF ) = 0  for |xe − uf | ≤ sF (4.48)
 
 c (x − u ) − s

(xe − uf ) > +sF
F e F F

The hard transition from clearance FF = 0 and uid compression resp. chamber deforma-
tion with FF 6= 0 is not realistic and leads to problems, even with the numeric solution.
Therefore, the function (4.48) is smoothed by a parabola in the range |xe − uf | ≤ 2 ∗ sF .
The motions of the uid mass cause friction losses in the ring channel, which are as a rst
approximation proportional to the speed,

FD = dF u̇F . (4.49)

Then, the equation of motion for the uid mass reads as

MF üF = − FF − FD . (4.50)

The membrane clearance makes Eq. (4.50) nonlinear and only solvable by numerical in-
tegration. The nonlinearity also aects the overall force in the hydro-mount, Eq. (4.46).

The dynamic stiness and the dissipation angle of a hydro-mount are displayed in Fig. 4.17
versus the frequency.

The simulation is based on the following system parameters

mF = 25 kg generalized uid mass

cT = 125 000 N/m stiness of main spring

dF = 750 N/(m/s) damping constant

cF = 100 000 N/m uid stiness

sF = 0.0002 mm clearance in membrane bearing

By the nonlinear and dynamic behavior a very good compromise can be achieved between
noise isolation and vibration damping.

70
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Dynamic Stiffness [N/m] at Excitation Amplitudes A = 2.5/0.5/0.1 mm

400

300

200

100

Dissipation Angle [deg] at Excitation Amplitudes A = 2.5/0.5/0.1 mm

60

50

40

30

20

10

0
0 1
10 Excitation Frequency [Hz] 10

Figure 4.17: Dynamic stiness [N/mm] and dissipation angle [deg] for a hydro-mount

71
5 Vertical Dynamics

5.1 Goals
The aim of vertical dynamics is the tuning of body suspension and damping to guarantee
good ride comfort, resp. a minimal stress of the load at sucient safety.

The stress of the load can be judged fairly well by maximal or integral values of the body
accelerations.

The wheel load Fz is linked to the longitudinal Fx and lateral force Fy by the coecient
of friction. The digressive inuence of Fz on Fx and Fy as well as non-stationary processes
at the increase of Fx and Fy in the average lead to lower longitudinal and lateral forces
at wheel load variations.

Maximal driving safety can therefore be achieved with minimal variations of the wheel
load. Small variations of the wheel load also reduce the stress on the track.

The comfort of a vehicle is subjectively judged by the driver. In literature dierent ap-
proaches of describing the human sense of vibrations by dierent metrics can be found.

Transferred to vehicle vertical dynamics, the driver primarily registers the amplitudes and
accelerations of the body vibrations. These values are thus used as objective criteria in
practice.

5.2 Modelling Aspects

5.2.1 Full Vehicle Model

For detailed investigations of ride comfort and ride safety sophisticated road and vehicle
models are needed. The three-dimensional vehicle model, shown in Fig. 5.1, includes
an elastically suspended engine, and dynamic seat models. The elasto-kinematics of the
wheel suspension was described fully nonlinear. In addition, dynamic force elements for
the damper elements and the hydro-mounts are used. Such sophisticated models not only
provide simulation results which are in good conformity to measurements but also make
it possible to investigate the vehicle dynamic attitude in an early design stage.

72
 Ford

FH Regensburg, University of Applied Sciences

Time = 0.000000
© Prof. Dr.-Ing. G. Rill

ZZ

Y
Y
X
X
Figure 5.1: Full vehicle model

Thilo Seibert Ext. 37598 /export/ford/dffa089/u/tseiber1/vedyna/work/results/mview.mvw 07/02/98

Vehicle Dynamics, Ford Research Center Aachen AA/FFA
5.2.2 Twodimensional Models
Much simpler models can be used, however, for basic studies on ride comfort and ride
safety. A two-dimensional vehicle model, for instance, suits perfectly with a single track
road model, Fig. 5.2. Neglecting longitudinal accelerations, the vehicle chassis only per-

a2
a1 zC2
C M2

zC1 zB
zA2
M1 M*
m2
M, Θ hub
zR(s-a2)
zA1 C
pitch
m1 yB
zR(s+a1) xB
zR(s)
s

Figure 5.2: Vehicle model for basic comfort and safety analysis

forms hub and pitch motions. Here, the chassis is considered as one rigid body. Then,
mass and inertia properties can be represented by three point masses which are located in
the chassis center of gravity and on top of the front and the rear axle. The lumped mass
model has 4 degrees of freedom. The hub and pitch motion of the chassis are represented
by the vertical motions of the chassis in the front zC1 and in the rear zC2 . The coordinates
zA1 and zA2 describe the vertical motions of the front and rear axle. The function zR (s)
provides road irregularities in the space domain, where s denotes the distance covered by
the vehicle and measured at the chassis center of gravity. Then, the irregularities at the
front and rear axle are given by zR (s + a1 ) and zR (s − a2 ) respectively, where a1 and a2
locate the position of the chassis center of gravity C .

73
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The point masses must add up to the chassis mass

M1 + M ∗ + M2 = M (5.1)

and they have to provide the same inertia around an axis located in the chassis center C
and pointing into the lateral direction

a21 M1 + a22 M2 = Θ . (5.2)

The correct location of the center of gravity is assured by

a1 M 1 = a2 M 2 . (5.3)

Now, Eqs. (5.2) and (5.3) yield the main masses

Θ Θ
M1 = and M2 = , (5.4)
a1 (a1 +a2 ) a2 (a1 +a2 )

and the coupling mass  

Θ
M∗ = M 1− (5.5)
M a1 a2
follows from Eq. (5.1).

Θ = M a1 a2 then,
If the mass and the inertia properties of a real vehicle happen to result in
∗
the coupling mass vanishes M = 0, and the vehicle can be represented by two uncoupled
two mass systems describing the vertical motion of the axle and the hub motion of the
chassis mass on top of each axle.

mid full sports

vehicles commercial heavy
size size utility
properties vehicle truck
car car vehicle
front axle
m1 [kg] 80 100 125 120 600
mass
rear axle
m2 [kg] 80 100 125 180 1100
mass
center
a1 [m] 1.10 1.40 1.45 1.90 2.90
of
gravity a2 [m] 1.40 1.40 1.38 1.40 1.90
chassis
M [kg] 1100 1400 1950 3200 14300
mass
chassis
Θ [kg m2 ] 1500 2350 3750 5800 50000
inertia

lumped M1 545 600 914 925 3592

mass M ∗ [kg] 126 200 76 1020 5225
model M2 429 600 960 1255 5483

Table 5.1: Mass and inertia properties of dierent vehicles

74
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Depending on the actual mass and inertia properties the vertical dynamics of a vehicle
can be investigated by two simple decoupled mass models describing the vibrations of the
front and rear axle and the corresponding chassis masses. By using half of the chassis and
half of the axle mass we nally end up in quarter car models.

The data in Table 5.1 show that for a wide range of passenger cars the coupling mass is
smaller than the corresponding chassis masses, M ∗ < M1 and M ∗ < M2 . Here, the two
mass model or the quarter car model represent a quite good approximation to the lumped
mass model. For commercial vehicles and trucks, where the coupling mass has the same
magnitude as the corresponding chassis masses, the quarter car model serves for basic
studies only.

5.2.3 Simple Models

At most vehicles, c.f. Table 5.1, the axle mass is much smaller than the corresponding
chassis mass, mi  Mi , i = 1, 2. Hence, for a rst basic study axle and chassis motions
can be investigated independently. The quarter car model is now further simplied to two
single mass models, Fig. 5.3.

``
zC cS `` dS
M 6 ` `

` m zW
6
cS ``` dS
` `
``
`` c
zR
6 `` T
c zR
6
c

Figure 5.3: Simple vertical vehicle models

The chassis model neglects the tire deection and the inertia forces of the wheel. For the
high frequent wheel motions the chassis can be considered as xed to the inertia frame.

The equations of motion for the models read as

M z̈C + dS żC + cS zC = dS żR + cS zR (5.6)

and
m z̈W + dS żW + (cS + cT ) zW = cT zR , (5.7)

where zC and zW label the vertical motions of the corresponding chassis mass and the
wheel mass with respect to the steady state position. The constants cS , dS describe the
suspension stiness and damping. The dynamic wheel load is calculated by

FTD = cT (zR − zW ) (5.8)

75
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where cT is the vertical or radial stiness of the tire and zR denotes the road irregularities.
In this simple approach the damping eects in the tire are not taken into account.

5.3 Basic Tuning

5.3.1 Natural Frequency and Damping Rate

At an ideally even track the right side of the equations of motion (5.6), (5.7) vanishes
because of zR = 0 and żR = 0. The remaining homogeneous second order dierential
equations can be written in a more general form as

z̈ + 2 ζ ω0 ż + ω02 z = 0 , (5.9)

where ω0 represents the undamped natural frequency, and ζ is a dimensionless parameter

called viscous damping ratio. For the chassis and the wheel model the new parameter are
given by

d cS
Chassis: z → zC , ζ → ζC = √ S , ω02 → ω0C
2
= ;
2 cS M M
dS cS +cT
Wheel: z → zW , ζ → ζW = p , ω02 → ω0W
2
= .
2 (cS +cT )m m
(5.10)

The solution of Eq. (5.9) is of the type

z(t) = z0 eλt , (5.11)

where z0 and λ are constants. Inserting Eq. (5.11) into Eq. (5.9) results in

(λ2 + 2 ζ ω0 λ + ω02 ) z0 eλt = 0 . (5.12)

Non-trivial solutions z0 6= 0 are possible, if

λ2 + 2 ζ ω0 λ + ω02 = 0 (5.13)

will hold. The roots of the characteristic equation (5.13) depend on the value of ζ
p
ζ<1 : λ1,2 = −ζ ω0 pm i ω0 1−ζ 2 ,
 p  (5.14)
ζ≥1 : λ1,2 = −ω0 ζ ∓ ζ −1 .2

Figure 5.4 shows the root locus of the eigenvalues for dierent values of the viscous
damping rate ζ.

76
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Im(λ)/ω0
ζ=0
ζ=0.5 1.0
ζ=0.2
ζ=0.7

ζ=0.9 0.5

ζ=1.5 ζ=1.25 ζ=1 ζ=1.25 ζ=1.5

-1 0
-3 -2.5 -2 -1.5 -0.5 Re(λ)/ω0
ζ=0.9 -0.5
ζ=0.7
ζ=0.2
ζ=0.5 -1.0
ζ=0

Figure 5.4: Eigenvalues λ1 and λ2 for dierent values of ζ

For ζ ≥ 1 the eigenvalues λ1,2 are both real and negative. Hence, Eq. (5.11) will produce a
exponentially decaying solution. If ζ < 1 holds, the eigenvalues λ1,2 will become complex,
where λ2 is the complex conjugate of λ1 . Now, the solution can be written as
 p 
−ζω0 t 2
z(t) = A e sin ω0 1−ζ t − Ψ , (5.15)

where A and Ψ are constants which have to be adjusted to given initial conditions z(0) = z0
and ż(0) = ż0 . The real part Re (λ1,2 ) = −ζω0 is negative and determines the decay of
p
the solution. The imaginary Im (λ1,2 ) = ω0 1−ζ 2 part denes the actual frequency of
the vibration. The actual frequency
p
ω = ω0 1−ζ 2 (5.16)

tends to zero, ω → 0, if the viscous damping ratio will approach the value one, ζ → 1.
In a more general way the relative damping may be judged by the ratio

−Re(λ1,2 )
Dλ = . (5.17)
| λ1,2 |
For complex eigenvalues which characterize vibrations

Dλ = ζ (5.18)

holds, because the absolute value of the complex eigenvalues is given by

q r  p 2
| λ1,2 | = Re(λ1,2 )2 + Im(λ1,2 )2 = (−ζ ω0 )2 + ω0 1−ζ 2 = ω0 , (5.19)

and hence, Eq. (5.17) results in

+ζ ω0
Dλ = = ζ. (5.20)
ω0
For ζ ≥ 1 the eigenvalues become real and negative. Then, Eq. (5.17) will always produce
the relative damping value Dλ = 1. In this case the viscous damping rate ζ is more
sensitive.

77
Vehicle Dynamics FH Regensburg, University of Applied Sciences

5.3.2 Spring Rates

5.3.2.1 Minimum Spring Rates

The suspension spring is loaded with the corresponding vehicle weight. At linear spring
characteristics the steady state spring deection is calculated from

Mg
u0 = . (5.21)
cS
At a conventional suspension without niveau regulation a load variation M → M + 4M
leads to changed spring deections u0 → u0 + 4u. In analogy to (5.21) the additional
deection follows from
4M g
4u = . (5.22)
cS
If for the maximum load variation 4M the additional spring deection is limited to 4u
the suspension spring rate can be estimated by a lower bound

4M g
cS ≥ . (5.23)
4u

In the standard design of a passenger car the engine is located in the front and the trunk
in the rear part of the vehicle. Hence, most of the load is supported by the rear axle
suspension.

For an example we assume that 150 kg 500 kg are going to the

of the permissible load of
front axle. Then, each front wheel is loaded by 4MF = 150 kg/2 = 75 kg and each rear
wheel by 4MR = (500 − 150) kg/2 = 175 kg .
The maximum wheel travel is limited, u ≤ umax . At standard passenger cars it is in the
range of umax ≈ 0.8 m to umax ≈ 0.10 m. By setting 4u = umax /2 we demand that the
spring deection caused by the load should not exceed half of the maximum value. Then,
according to Eq. (5.23) a lower bound of the spring rate at the front axle can be estimated
by
cmin
S = ( 75 kg ∗ 9.81 m/s2 )/(0.08/2) m = 18400 N/m . (5.24)

The maximum load over one rear wheel was estimated here by 4MR = 175 kg . Assuming
that the suspension travel at the rear axle is slightly larger, umax ≈ 0.10 m the minimum
spring rate at the rear axle can be estimated by

cmin
S = ( 175 kg ∗ 9.81 m/s2 )/(0.10/2) m = 34300 N/m , (5.25)

which is nearly two times the minimum value of the spring rate at the front axle. In order
to reduce this dierence we will choose a spring rate of cS = 20 000 N/m at the front axle.
In Tab. 5.1 the lumped mass chassis model of a full size passenger car is described by
M1 = M2 = 600 kg and M ∗ = 200. To approximate the lumped mass model by two
decoupled two mass models we have to neglect the coupling mass or, in order to achieve

78
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

the same chassis mass, to distribute M∗ equally to the front and the rear. Then, the
corresponding cassis mass of a quarter car model is given here by

M = M1 + M ∗ /2 /2 = (600 kg + 200/2 kg)/2 = 350 kg .

(5.26)

According to Eq. 5.10 the undamped natural eigen frequency of the simple chassis model
2
is then given by ω0C = cS /M . Hence, for a spring rate of cS = 20000 N/m the undamped
natural frequency of the unloaded car amounts to
p
f0C = 20000 N/m ∗ 350 kg/(2 π) = 1.2 Hz , (5.27)

which is a typical value for most of all passenger cars. Due to the small amount of load
the undamped natural frequency for the loaded car does not change very much,
p
f0C = 20000 N/m ∗ (350 + 75) kg/(2 π) = 1.1 Hz . (5.28)

The corresponding cassis mass over the rear axle is given here by

M = M2 + M ∗ /2 /2 = (600 kg + 200/2 kg)/2 = 350 kg .

(5.29)

Now the undamped natural frequencies for the unloaded

0
p
f0C = 34300 N/m/350 kg/(2 π) = 1.6 Hz (5.30)

and the loaded car

L
p
f0C = 34300 N/m/(350 + 175) kg/(2 π) = 1.3 Hz (5.31)

are larger and dier more.

5.3.2.2 Nonlinear Springs

In order to reduce the spring rate at the rear axle and to avoid too large spring deections
when loaded nonlinear spring characteristics are used, Fig. 5.5. Adding soft bump stops
the overall spring force in the compression mode u≥0 can be modeled by the nonlinear
function
 2 !
u
FS = FS0 + c0 u 1+k , (5.32)
4u
where FS0 is the spring preload, cS describes the spring rate at u = 0, andk > 0 charac-
terizes the intensity of the nonlinearity. The linear characteristic provides at u = 4u the
value FSlin (4u) = FS0 + cS 4u. To achieve the same value with the nonlinear spring
FS0 + c0 4u (1 + k) = FS0 + cS 4u or c0 (1 + k) = cS (5.33)

must hold, where cS describes the spring rate of the corresponding linear characteristics.
The local spring rate is determined by the derivative

 2 !
dFS u
= c0 1 + 3k . (5.34)
du 4u

79
Vehicle Dynamics FH Regensburg, University of Applied Sciences

FS dFS 8000
du u=∆u

FS [N]

6000
∆M g 63 kN/m
44 kN/m
cS
dFS 4000
FS0 du u=0
29 kN/m
20 kN/m

2000
∆u u 0 0.05 u [m] 0.1

Figure 5.5: Principle and realizations of nonlinear spring characteristics

Hence, the spring rate for the loaded car at u = 4u is given by

cL = c0 (1 + 3 k) . (5.35)

The intensity of the nonlinearity k can be xed, for instance, by choosing an appropriate
spring rate for the unloaded vehicle. With c0 = 20000 N/m the spring rates on the front
and rear axle will be the same for the unloaded vehicle. With cS = 34300 N/m Eq. (5.33)
yields
cS 34300
k = −1 = − 1 = 0.715 . (5.36)
c0 20000
The solid line in Fig. 5.5 shows the resulting nonlinear spring characteristics which is
characterized by the spring rates c0 = 20 000 N/m and cL = c0 (1 + 3k) = 20 000 ∗ (1 +
3 ∗ 0.715) = 62 900 N/m for the unloaded and the loaded vehicle. Again, the undamped
natural frequencies

s s
0 20000 N/m 1 L 92000 N/m 1
f0C = = 1.20 Hz or f0C = = 1.74 Hz (5.37)
350 kg 2π (350+175) kg 2 π

for the unloaded and the loaded vehicle dier quite a lot.

The unloaded and the loaded vehicle have the same undamped natural frequencies if

c0 cL cL M + 4M
= or = (5.38)
M M + 4M c0 M

will hold. Combing this relationship with Eq. (5.35) one obtains

 
M 1 M + 4M 1 4M
1 + 3k = or k = −1 = . (5.39)
M + 4M 3 M 3 M

80
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Hence, for the quarter car model with M = 350 kg and 4M = 175 the inten-
sity of the nonlinear spring amounts to k = 1/3 ∗ 175/350 = 0.1667. This value
and cS = 34300 N/m will produce the dotted line in Fig. 5.5. The spring rates
c0 = cS /(1 + k) = 34 300 N/m / (1 + 0.1667) = 29 400 N/m for the unloaded and
cL = c0 (1 + 3k) = 29 400 N/m ∗ (1 + 3 ∗ 0.1667) = 44 100 N/m for the loaded vehicle fol-
low from Eqs. (5.34) and (5.35). Now, the undamped natural frequency for the unloaded
0
p 0
p
f0C = c0 /M = 1.46 Hz and the loaded vehicle f0C = cL /(M + 4M ) = 1.46 Hz are
in deed the same.

5.3.3 Influence of Damping

To investigate the inuence of the suspension damping to the chassis and wheel motion the
simple vehicle models are exposed to initial disturbances. Fig. 5.6 shows the time response
of the chassis zC (t)
and wheel displacement zW (t) as well as the chassis acceleration z̈C
0 D
and the wheel load FT = FT + FT for dierent damping rates ζC and ζW . The dynamic
0
wheel load follows from Eq. (5.8), and the static wheel load is given by FT = (M + m) g ,
where g labels the constant of gravity.

To achieve the same damping rates for the chassis and the wheel model dierent values
for the damping parameter dS were needed.

With increased damping the overshoot eect in the time history of the chassis displace-
ment and the wheel load becomes smaller and smaller till it vanishes completely at ζC = 1
and ζW = 1. The viscous damping rate ζ=1

5.3.4 Optimal Damping

5.3.4.1 Avoiding Overshoots

If avoiding overshoot eects is the design goal then, ζ =1 will be the optimal damping
ratio. For ζ=1 the eigenvalues of the single mass oscillator change from complex to real.
Thus, producing a non oscillating solution without any sine and cosine terms.

According to Eq. (5.10) ζC = 1 and ζW = 1 results in the optimal damping parameter

ζC =1 p ζW =1 p
dopt
S Comfort
= 2 cS M , and doptS Safety = 2 (cS +cT )m . (5.40)

So, the damping values

ζC =1 N ζW =1 N
dopt
S Comfort
= 5292 and dopt
S Safety
= 6928 (5.41)
m/s m/s

will avoid an overshoot eect in the time history of the chassis displacement zC (t) or in the
in the time history of the wheel load FT (t). Usually, as it is here, the damping values for
optimal comfort and optimal ride safety will be dierent. Hence, a simple linear damper
can either avoid overshoots in the chassis motions or in the wheel loads.

81
Vehicle Dynamics FH Regensburg, University of Applied Sciences

chassis model wheel model

ζC [ - ] dS [Ns/m] ζW [ - ] dS [Ns/m]
350 kg dS 20000 N/m
0.25 1323 0.25 1732
0.50 2646 0.50 3464
0.75 3969 50 kg
20000 N/m dS 0.75 5196
1.00 5292 1.00 6928
1.25 6614 1.25 8660 220000 N/m

displacement [mm] displacement [mm]

200 20

150 15
ζC ζW
100 10

50 5

0 0

-50 -5

-100 -10
0 0.5 1 t [s] 1.5 0 0.05 0.1 t [s] 0.15
acceleration [g] wheel load [N]
1 6000

5000
0.5
4000

0 3000
ζC 2000
-0.5 ζW
1000

-1 0
0 0.5 1 t [s] 1.5 0 0.05 0.1 t [s] 0.15

Figure 5.6: Time response of simple vehicle models to initial disturbances

The overshot in the time history of the chassis accelerations z̈C (t) will only vanish for
ζC → ∞ which surely is not a desirable conguration, because then, it takes a very long
time till the initial chassis displacement has fully disappeared.

5.3.4.2 Fast Approach to Steady State

Instead of avoiding overshoot eects we better demand that the time history of the system
response will approach the steady state value as fast as possible. Fig. 5.7 shows the typical
time response of a damped single-mass oscillator to the initial disturbance z(t = 0) = z0
and ż(t = 0) = 0.

82
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

z0

zS
tE t
z(t)

Figure 5.7: Evaluating a damped vibration

Counting the dierences of the system response z(t) from the steady state value zS = 0 as
errors allows to judge the attenuation. If the overall quadratic error becomes a minimum

t=t
Z E
2
 = z(t)2 dt → M in , (5.42)

t=0

the system approaches the steady state position as fast as possible. In theory tE → ∞
holds, for practical applications a nite tE have to be chosen appropriately.

To judge ride comfort and ride safety the hub motion of the chassis zC , its acceleration
z̈C and the variations of the dynamic wheel load FTD can be used. In the absence of road
D
irregularities zR = 0 the dynamic wheel load from Eq. (5.8) simplies to FT = −cT zW .
Hence, the demands

t=t
Z E h
2
2 i
2C = g1 z̈C + g2 zC dt → M in (5.43)

t=0

and
t=t
Z E

2
2S = −cT zW dt → M in (5.44)

t=0
will guarantee optimal ride comfort and optimal ride safety. By the factors g1 and g2 the
acceleration and the hub motion can be weighted dierently.

The equation of motion for the chassis model can be resolved for the acceleration

2


z̈C = − ω0C zC + 2δC żC , (5.45)

where, the system parameter M , dS and cS were substituted by the damping rate δC =
ζC ω0C = dS /(2M ) and by the undamped natural frequency ω0C = cS /M . Then, the
problem in Eq. (5.43) can be written as

t=t
Z E h
2 i
2C = g12 ω0C
2
zC + 2δC żC + g22 zC2 dt
t=0
t=t
   (5.46)
Z E 2 2
  g12 (ω0C ) + g22 g12 2
ω0C 2δC zC
= zC żC    → M in ,
2
g12 2
ω0C 2δC g12 (2δC ) żC
t=0

83
Vehicle Dynamics FH Regensburg, University of Applied Sciences

 
where xTC = zC żC is the state vector of the chassis model. In a similar way Eq. (5.44)
can be transformed to

t=t
Z E t=t
Z E " #" #
c2T 0 zW
2S c2T 2
 
= zW dt = zW żW → M in , (5.47)
0 0 żW
t=0 t=0
 
where xTW = zW żW denotes the state vector of the wheel model.

The problems given in Eqs. (5.46) and (5.47) are called disturbance-reaction problems.
They can be written in a more general form

t=t
Z E
xT (t) Q x(t) dt → M in (5.48)

t=0

where x(t) denotes the state vector and Q = QT is a symmetric weighting matrix. For
single mass oscillators described by Eq. (5.9) the state equation reads as
    
ż 0 1 z
= 2 . (5.49)
z̈ −ω0 −2δ ż
| {z } | {z } | {z }
ẋ A x

For tE → ∞ the time response of the system exposed to the initial disturbance x(t = 0) =
x0 vanishes x(t → ∞) = 0, and the integral in Eq.(5.48) can be solved by
t=t
Z E
xT (t) Q x(t) dt = xT0 R x0 , (5.50)

t=0

where the symmetric matrix R = RT is given by the Ljapunov equation

AT R + R A + Q = 0 . (5.51)

For the single mass oscillator the Ljapunov equation

       
0 −ω02 R11 R12 R11 R12 0 1 Q11 Q12
+ + . (5.52)
1 −2δ R12 R22 R12 R22 −ω02 −2δ Q12 Q22

results in 3 linear equations

−ω02 R12 − ω02 R12 + Q11 = 0

−ω02 R22 + R11 − 2δ R12 + Q12 = 0 (5.53)

R12 − 2δ R22 + R12 − 2δ R22 + Q22 = 0

which easily can be solved for the elements of R

ω02
 
δ 1 Q11 Q11 Q22
R11 = 2
+ Q11 − Q12 + Q22 , R12 = , R22 = + . (5.54)
ω0 4δ 4δ 2ω02 2
4δ ω0 4δ

84
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

For the initial disturbance x0 = [ z0 0 ]T Eq. (5.50) nally results in

t=t
Z E
ω02
  
T δ 1
x (t) Q x(t) dt = z02 R11 = z02 2
+ Q11 − Q12 + Q22 . (5.55)
ω0 4δ 4δ
t=0

Now, the integral in Eq. (5.46) evaluating the ride comfort is solved by

2
  
δC 1
2  ω0C 2
2C = 2
z0C + 2 2 2 2 2
g1 ω0C + g2 − g1 ω0C 2 δC + 2
g (2δC )
2
ω0C 4δC 4δC 1
" (5.56)
 2 !  2 #
2 2 ω0C g2 g2
= z0C ω0C g12 + 2
+ 2
ζC ω0C .
4ζC ω0C ω0C

where the abbreviation δC was nally replaced by ζC ω0C .

By setting g1 = 1 and g2 = 0 the time history of the chassis acceleration z̈C is weighted
only. Eq. (5.56) then simplies to

ω0C
2C z̈C
2
= z0C 2
ω0C (5.57)
4ζC
which will become a minimum for ω0C → 0 or ζC → ∞. As mentioned before, ζC → ∞
surely is not a desirable conguration. A low undamped natural frequency ω0C → 0 is
achieved by a soft suspension spring cS → 0 or a large chassis mass M → ∞. However, a
large chassis mass is uneconomic and the suspension stiness is limited by the the loading
conditions. Hence, weighting the chassis accelerations only does not lead to a specic
result for the system parameter.

The combination of g1 = 0 and g2 = 1 weights the time history of the chassis displacement
only. Then, Eq. (5.56) results in

z2
 
1
2C z = 0C + ζC (5.58)
C ω0C 4ζC
which will become a minimum for ω0C → ∞ or
d 2C |zC 2
 
z0C −1
= +1 = 0. (5.59)
d ζC ω0C 4ζC2
A high undamped natural frequency ω0C → ∞
contradicts the demand for rapidly van-
1
ishing accelerations. The viscous damping ratio ζC = solves Eq. (5.59) and minimizes
2
the merit function in Eq. (5.58). But again, this value does not correspond with ζC → ∞
which minimizes the merit function in Eq. (5.57).

Hence, practical results can be achieved only if the chassis displacements and the chassis
accelerations will be evaluated simultaneously. To do so, appropriate weighting factors
have to be chosen. In the equation of motion for the chassis (5.6) the terms M z̈C and
cS zC are added. Hence, g1 = M and g2 = cS or

cS 2
g1 = 1 and g2 = = ω0C (5.60)
M

85
Vehicle Dynamics FH Regensburg, University of Applied Sciences

provide system-tted weighting factors. Now, Eq. (5.56) reads as

 
ω0C
2C = 2
z0C 2
ω0C + ζC ω0C . (5.61)
2ζC
Again, a good ride comfort will be achieved by ω0C → 0. For nite undamped natural
frequencies Eq. (5.61) becomes a minimum, if the viscous damping rate ζC will satisfy

d 2C |zC
 
2 2 −ω0C
= z0C ω0C + ω0C = 0. (5.62)
d ζC 2ζC2
Hence, a viscous damping rate of
1√
ζC = 2 (5.63)
2
or a damping parameter of

1
√
opt ζC = 2 2
p
dS Comfort = 2 cS M , (5.64)

will provide optimal comfort by minimizing the merit function in Eq. (5.61).

For the passenger car with M = 350 kg and cS = 20 000 N/m the optimal damping
parameter will amount to
√
ζC = 12 2 N
dopt
S Comfort
= 3742 (5.65)
m/s
which is 70% of the value needed to avoid overshot eects in the chassis displacements.

The integral in Eq. (5.47) evaluating the ride safety is solved by

z2
 
1
2S = 0W ζW + c2T (5.66)
ω0W 4ζW
where the model parameter m, cS , dS and cT where replaced by the undamped natural
2
frequency ω0W = (cS + cT )/m and by the damping ratio δW = ζW ω0W = dS /(2m).
2
A soft tire cT → 0 make the safety criteria Eq. (5.66) small S → 0 and thus, reduces the
dynamic wheel load variations. However, the tire spring stiness can not be reduced to
arbitrary low values, because this would cause too large tire deformations. Small wheel
masses m → 0 and/or a hard body suspension cS → ∞ will increase ω0W and thus,
reduce the safety criteria Eq. (5.66). The use of light metal rims improves, because of
wheel weight reduction, the ride safety of a car. Hard body suspensions contradict a good
driving comfort.

With xed values for cT and ω0W the merit function in Eq. (5.66) will become a minimum
if
∂2S 2
 
z0W −1
= 1+ 2 c2T = 0 (5.67)
∂ζW ω0W 4ζW
will hold. Hence, a viscous damping rate of

1
ζW = (5.68)
2

86
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

or the damping parameter

p
dopt
S Safety
= (cS + cT ) m (5.69)

will guarantee optimal ride safety by minimizing the merit function in Eq. (5.66).

For the passenger car with M = 350 kg and cS = 20 000 N/m the optimal damping
parameter will now amount to

ζW = 12 N
dopt
S Safety
= 3464 (5.70)
m/s

which is 50% of the value needed to avoid overshot eects in the wheel loads.

5.4 Sky Hook Damper

5.4.1 Modelling Aspects

In standard vehicle suspension systems the damper is mounted between the wheel and
the body. Hence, the damper aects body and wheel/axle motions simultaneously.

sky

zC zC
M M dB
dW

cS dS cS FD

zW zW
m m

zR cT zR cT

a) Standard Damper b) Sky Hook Damper

Figure 5.8: Quarter car model with standard and sky hook damper

To take this situation into account the simple quarter car models of section 5.2.3 must be
combined to a more enhanced model, Fig. 5.8a.

Assuming a linear characteristics the suspension damper force is given by

FD = dS (żW − żC ) , (5.71)

where dS denotes the damping constant, and żC , żW are the time derivatives of the absolute
vertical body and wheel displacements.

87
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The sky hook damping concept starts with two independent dampers for the body and
the wheel/axle mass, Fig. 5.8b. A practical realization in form of a controllable damper
will then provide the damping force

FD = dW żW − dC żC , (5.72)

where instead of the single damping constant dS now two design parameter dW and dC
are available.

The equations of motion for the quarter car model are given by

M z̈C = FS + FD − M g ,
(5.73)
m z̈W = FT − FS − FD − m g ,

where M, m are the sprung and unsprung mass, zC , zW denote their vertical displace-
ments, and g is the constant of gravity.

The suspension spring force is modeled by

FS = FS0 + cS (zW − zC ) , (5.74)

where FS0 = mC g is the spring preload, and cS the spring stiness.

Finally, the vertical tire force is given by

FT = FT0 + cT (zR − zW ) , (5.75)

where FT0 = (M + m) g is the tire preload, cS the vertical tire stiness, and zR describes
the road roughness. The condition FT ≥ 0 takes the tire lift o into account.

5.4.2 Eigenfrequencies and Damping Ratios

Using the force denitions in Eqs. (5.72), (5.74) and (5.75) the equations of motion in
Eq. (5.73) can be transformed to the state equation

      
żC 0 0 1 0 zC 0
      h i
 żW
 =  0 0 0 1   zW  +  0  zR ,
     
 (5.76)
 z̈   − cS cS
− dMC dW  
ż   0 
 C   M M M  C    | {z }
z̈W cS cS +cT
− m dC dW
−m żW cT u
m m m
| {z } | {z } | {z } | {z }
ẋ A x B
where the weight forces M g , mg were compensated by the preloads FS0 , FT0 , the term B u
describes the excitation, x denotes the state vector, and A is the state matrix. In this
linear approach the tire lift o is no longer taken into consideration.

The eigenvalues λ of the state matrix A will characterize the eigen dynamics of the quarter
car model. In case of complex eigenvalues the damped natural eigenfrequencies are given

88
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Frequencies [Hz] Damping ratio ζ = Dλ

12 1

10 0.8
Wheel 350 kg
0.7
8
0.6
6 20000 N/m dS 0.5
0.4
4 50 kg

220000 N/m 0.2

2 Chassis 3880
3220
0 0
0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000
dS [N/(m/s)] dS [N/(m/s)]

Figure 5.9: Quarter car model with standard damper

by the imaginary parts, ω = Im(λ), and according to Eq. (vdyn-eq: relative damping
ratio lambda) ζ = Dλ = −Re(λ)/ |λ|. evaluates the damping ratio.
By setting dC = dS and dW = dS Eq. (5.76) represents a quarter car model with the
standard damper described by Eq. (5.71). Fig. 5.9 shows the eigenfrequencies f = ω/(2π)
and the damping ratios ζ = Dλ for dierent values of the damping parameter dS .
√
Optimal ride comfort with a damping ratio of ζC = 21 2 ≈ 0.7 for the chassis motion
could be achieved with the damping parameter dS = 3880 N/(m/s), and the damping
parameter dS = 3220 N/(m/s) would provide for the wheel motion a damping ratio of
ζW = 0.5 which correspond to minimal wheel load variations. This damping parameter
are very close to the values 3742 N/(m/s) and 3464 N/(m/s) which very calculated in
Eqs. (5.65) and (5.70) with the single mass models. Hence, the very simple single mass
models can be used for a rst damper layout. Usually, as it is here, optimal ride comfort
and optimal ride safety cannot achieved both by a standard linear damper.

The sky-hook damper, modeled by Eq. (5.72), provides with dW and dS two design pa-
rameter. Their inuence to the eigenfrequencies f and the damping ratios ζ is shown in
Fig. 5.10.

The the sky-hook damping parameter dC , dW have a nearly independent inuence on

the damping ratios. The chassis damping ratio ζC mainly depends on dC , and the wheel
damping ratio ζW mainly depends on dW . Hence, the damping of the chassis and the
wheel motion can be adjusted to nearly each design goal. Here, a sky-hook damper with
dC = 3900 N/(m/s) and dW = 3200 N/(m/s) would generate the damping ratios dC = 0.7
and dW = 0.5 hence, combining ride comfort and ride safety within one damper layout.

89
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Frequencies [Hz] Damping ratios ζC, ζW

12 1
dC [N/(m/s)] ζC
10 0.8 4500
350 kg
0.7 4000
8
dC 0.6 3500
dC 20000 N/m 3000
6 dW 0.5
2500
50 kg 0.4
4 2000
220000 N/m 1500
0.2
2 ζW
0 0
0 1000 2000 3000 4000 5000 0 1000 2000 3000 4000 5000
dW [N/(m/s)] dW [N/(m/s)]

Figure 5.10: Quarter car model with sky-hook damper

5.4.3 Technical Realization

By modifying the damper law in Eq. (5.72) to

dW żW − dC żC
FD = dW żW − dC żC + = (żW − żC ) = d∗S (żW − żC ) (5.77)
żW − żC
| {z }
d∗S
the sky-hook damper can be realized by a standard damper in the form of Eq. (5.71). The
∗
new damping parameter dS now nonlinearly depends on the absolute vertical velocities of
∗ ∗
the chassis and the wheel dS = dS (żC , żW ). As, a standard damper operates in a dissipative
∗
mode only the damping parameter will be restricted to positive values, dS > 0. Hence, the
passive realization of a sky-hook damper will only match with some properties of the ideal
damper law in Eq. (5.72). But, compared with the standard damper it still can provide a
better ride comfort combined with an increased ride safety.

5.5 Nonlinear Force Elements

5.5.1 Quarter Car Model

The principal inuence of nonlinear characteristics on driving comfort and safety can
already be studied on a quarter car model Fig. 5.11.

The equations of motion read as

M z̈C = FS + FD − M g
(5.78)
m z̈W = FT − FS − FD − m g ,

where g = 9.81 m/s2 labels the constant of gravity, M , m are the masses of the chassis
and the wheel, FS , FD , FT describe the spring, the damper, and the vertical tire force,

90
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

zC
nonlinear spring nonlinear damper
M
FS FD

FS FD
v
u zW
m

cT zR

Figure 5.11: Quarter car model with nonlinear spring and damper characteristics

and the vertical displacements of the chassis zC and the wheel zW are measured from the
equilibrium position.

In extension to Eq. (5.32) the spring characteristics is modeled by


 2 !

 u
 c0 u 1 + kr u<0


4ur

FS = FS0 +  2 ! (5.79)
u


 c0 u 1 + kc u≥0


4uc


where FS0 = M g is the spring preload, and

u = zW − zC (5.80)

describes the spring travel. Here, u < 0 marks tension (rebound), and u ≥ 0 compres-
sion. Two sets of kr , ur and kc , uc dene the spring nonlinearity during rebound and
compression. For kr = 0 and kc = 0 a linear spring characteristics is obtained.

A degressive damper characteristics can be modeled by


d0 v


v <0,


1 − pr v

FD (v) = (5.81)
d0 v


v ≥0,



1 + pc v
where d0 denotes the damping constant at v = 0, and the damper velocity is dened by

v = żW − żC . (5.82)

The sign convention of the damper velocity was chosen consistent to the spring travel.
Hence, rebound is characterized by v<0 and compression by v ≥ 0. The parameter pr

91
Vehicle Dynamics FH Regensburg, University of Applied Sciences

and pc make it possible to model the damper nonlinearity dierently in the rebound and
compression mode. A linear damper characteristics is obtained with pr = 0 and pc = 0.
The nonlinear spring design in Section 5.3.2 holds for the compression mode. Hence,
using the same data we obtain: c0 = 29 400 N/m, uc = 4u = umax /2 = 0.10/2 = 0.05 and
kc = k = 0.1667. By setting ur = uc and kr = 0 a simple linear is used in the rebound
mode, Fig. 5.12a.

a) Spring b) Damper
7000 5000
cS = 34300 N/m d0 = 4200 N/(m/s)
6000 c0 = 29400 N/m pr = 0.4 1/(m/s)

FD [N/m]
ur = 0.05 m pc = 1.2 1/(m/s)
FS [N/m]

2500
5000 kr = 0
uc = 0.05 m
4000 kc = 0.1667
0
3000

2000
-2500
1000 rebound compression rebound compression
u<0 u>0 v<0 v>0
0 -5000
-0.1 -0.05 0 0.05 0.1 -1 -0.5 0 0.5 1
u [m] v [m/s]

Figure 5.12: Spring and damper characteristics: - - - linear,  nonlinear

According to Section 5.3.4 damping coecients optimizing the ride comfort and the ride
safety can be calculated from Eqs. (5.65) and (5.69). For cS = 34 300 N/m which is the
equivalent linear spring rate, M = 350 kg , m = 50 kg and cT = 220 000 N/m we obtain

√ √
(dS )C
opt = 2 cS M = 2 34 300 350 = 4900 N/(m/s) ,
(5.83)

(dS )Sopt
p p
= (cS + cT ) m = (18 000 + 220 000) 50 = 3570 N/(m/s) .

−1
The mean value d0 = 4200N/(m/s) may serve as compromise. With pr = 0.4(m/s) and
pc = 1.2 (m/s)−1 the nonlinearity becomes more intensive in compression than rebound,
Fig. 5.12b.

5.5.2 Results

The quarter car model is driven with constant velocity over a single obstacle. Here, a
cosine shaped bump with a height of H = 0.08 m and a length of L = 2.0 m was used.
The results are plotted in Fig. 5.13.

Compared to the linear model the nonlinear spring and damper characteristics result in
2 2
signicantly reduced peak values for the chassis acceleration (6.0m/s instead of 7.1m/s )

92
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Chassis acceleration [m/s2] Wheel load [N] Suspension travel [m]

10 7000 0.04
6660
7.1 6160
6.0 6000
5 0.02
5000
0 4000 0

-5 3000 -0.02
2000
-10 linear -0.04
1000 nonlinear

-15 0 -0.06
0 0.5 1 1.5 0 0.5 1 1.5 0 0.5 1 1.5
time [s] time [s] time [s]

Figure 5.13: Quarter car model driving with v = 20 km h over a single obstacle

and for the wheel load (6160 N instead of 6660 N ). Even the tire lift o at t ≈ 0.25 s
can be avoided. While crossing the bump large damper velocities occur. Here, the degres-
sive damper characteristics provides less damping compared to the linear damper which
increases the suspension travel.

Chassis acceleration [m/s2] Wheel load [N] Suspension travel [m]

10 7000 0.04

6000 0.02
5
5000
0
0 4000
-0.02
-5 3000
-0.04
2000
linear,
-10 low damping -0.06
1000
nonlinear
-15 0 -0.08
0 0.5 1 1.5 0 0.5 1 1.5 0 0.5 1 1.5
time [s] time [s] time [s]

Figure 5.14: Results for low damping compared to nonlinear model

A linear damper with a lower damping coecient, d0 = 3000N/m for instance, also reduces
the peaks in the chassis acceleration and in the wheel load, but then the attenuation of
the disturbances will take more time. Fig. 5.14. Which surely is not optimal.

93
6 Longitudinal Dynamics

6.1 Dynamic Wheel Loads

6.1.1 Simple Vehicle Model

The vehicle is considered as one rigid body which moves along an ideally even and hor-
izontal road. At each axle the forces in the wheel contact points are combined in one
normal and one longitudinal force.

h mg

Fx1 Fx2
Fz1 a1 a2 Fz2

Figure 6.1: Simple vehicle model

If aerodynamic forces (drag, positive and negative lift) are neglected at rst, the equations
of motions in the x-, z -plane will read as

m v̇ = Fx1 + Fx2 , (6.1)

0 = Fz1 + Fz2 − m g , (6.2)

0 = Fz1 a1 − Fz2 a2 + (Fx1 + Fx2 ) h , (6.3)

where v̇ indicates the vehicle's acceleration, m is the mass of the vehicle, a1 + a2 is the
wheel base, and h is the height of the center of gravity.

These are only three equations for the four unknown forces Fx1 , Fx2 , Fz1 , Fz2 . But, if we
insert Eq. (6.1) in Eq. (6.3), we can eliminate two unknowns at a stroke

0 = Fz1 a1 − Fz2 a2 + m v̇ h . (6.4)

94
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

The equations Eqs. (6.2) and (6.4) can be resolved for the axle loads now

a2 h
Fz1 = m g − m v̇ , (6.5)
a1 + a2 a1 + a2

a1 h
Fz2 = m g + m v̇ . (6.6)
a1 + a2 a1 + a2
The static parts
st a2 st a1
Fz1 = mg , Fz2 = mg (6.7)
a1 + a2 a1 + a2
describe the weight distribution according to the horizontal position of the center of
gravity. The height of the center of gravity only inuences the dynamic part of the axle
loads,
dyn h v̇ dyn h v̇
Fz1 = −m g , Fz2 = +m g . (6.8)
a1 + a2 g a1 + a2 g
When accelerating v̇ > 0, the front axle is relieved as the rear axle is when decelerating
v̇ < 0.

6.1.2 Influence of Grade

v z

Fx1
mg
h
Fz1 a1
Fx2
a2 α
Fz2

Figure 6.2: Vehicle on grade

For a vehicle on a grade, Fig.6.2, the equations of motion Eq. (6.1) to Eq. (6.3) can easily
be extended to
m v̇ = Fx1 + Fx2 − m g sin α ,
0 = Fz1 + Fz2 − m g cos α , (6.9)

0 = Fz1 a1 − Fz2 a2 + (Fx1 + Fx2 ) h ,

95
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where α denotes the grade angle. Now, the axle loads are given by

a2 − h tan α h
Fz1 = m g cos α − m v̇ , (6.10)
a1 + a2 a1 + a2

a1 + h tan α h
Fz2 = m g cos α + m v̇ , (6.11)
a1 + a2 a1 + a2
where the dynamic parts remain unchanged, whereas now the static parts also depend on
the grade angle and the height of the center of gravity.

6.1.3 Aerodynamic Forces

The shape of most vehicles or specic wings mounted at the vehicle produce aerodynamic
forces and torques. The eect of these aerodynamic forces and torques can be represented
by a resistant force applied at the center of gravity and down forces acting at the front
and rear axle, Fig. 6.3.

FD1 FD2

FAR

h
mg
Fx1 Fx2
a1 a2
Fz1 Fz2

Figure 6.3: Vehicle with aerodynamic forces

If we assume a positive driving speed, v > 0, the equations of motion will read as

m v̇ = Fx1 + Fx2 − FAR ,

0 = Fz1 −FD1 + Fz2 −FD2 − m g , (6.12)

0 = (Fz1 −FD1 ) a1 − (Fz2 −FD2 ) a2 + (Fx1 + Fx2 ) h ,

where FAR and FD1 , FD2 describe the air resistance and the down forces. For the dynamic
axle loads we get

a2 h
Fz1 = FD1 + m g − (m v̇ + FAR ) , (6.13)
a1 + a2 a1 + a2

a1 h
Fz2 = FD2 + m g + (m v̇ + FAR ) . (6.14)
a1 + a2 a1 + a2
The down forces FD1 , FD2 increase the static axle loads, and the air resistance FAR
generates an additional dynamic term.

96
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

6.2 Maximum Acceleration

6.2.1 Tilting Limits

Ordinary automotive vehicles can only apply pressure forces to the road. If we take the
demands Fz1 ≥ 0 and Fz2 ≥ 0 into account, Eqs. (6.10) and (6.11) will result in

v̇ a2 v̇ a1
≤ cos α − sin α and ≥ − cos α − sin α . (6.15)
g h g h
These two conditions can be combined in one

a1 v̇ a2
− cos α ≤ + sin α ≤ cos α . (6.16)
h g h
Hence, the maximum achievable accelerations (v̇ > 0) and decelerations (v̇ < 0) are
limited by the grade angle α and the position a1 , a2 , h of the center of gravity. For v̇ → 0
the tilting condition Eq. (6.16) results in

a1 a2
− ≤ tan α ≤ (6.17)
h h
which describes the climbing and downhill capacity of a vehicle.

The presence of aerodynamic forces complicates the tilting condition. Aerodynamic forces
become important only at high speeds. Here, the vehicle acceleration is normally limited
by the engine power.

6.2.2 Friction Limits

The maximum acceleration is also restricted by the friction conditions

|Fx1 | ≤ µ Fz1 and |Fx2 | ≤ µ Fz2 (6.18)

where the same friction coecient µ has been assumed at front and rear axle. In the limit
case
Fx1 = ± µ Fz1 and Fx2 = ± µ Fz2 (6.19)

the linear momentum in Eq. (6.9) can be written as

m v̇max = ± µ (Fz1 + Fz2 ) − m g sin α . (6.20)

Using Eqs. (6.10) and (6.11) one obtains

 
v̇
= ± µ cos α − sin α . (6.21)
g max

That means climbing (v̇ > 0, α > 0) or downhill stopping (v̇ < 0, α < 0) requires at least
a friction coecient µ ≥ tan |α|.

97
Vehicle Dynamics FH Regensburg, University of Applied Sciences

According to the vehicle dimensions and the friction values the maximal acceleration or
deceleration is restricted either by Eq. (6.16) or by Eq. (6.21).

If we take aerodynamic forces into account, the maximum acceleration and deceleration
on a horizontal road will be limited by

   
FD1 FD2 FAR v̇ FD1 FD2 FAR
−µ 1 + + − ≤ ≤ µ 1 + + − . (6.22)
mg mg mg g mg mg mg

In particular the aerodynamic forces enhance the braking performance of the vehicle.

6.3 Driving and Braking

6.3.1 Single Axle Drive

With the rear axle driven in limit situations, Fx1 = 0 and Fx2 = µ Fz2 hold. Then, using
Eq. (6.6) the linear momentum Eq. (6.1) results in

 
a1 h v̇R WD
m v̇R WD = µ m g + , (6.23)
a1 + a2 a1 + a2 g

where the subscript R WD indicates the rear wheel drive. Hence, the maximum acceleration
for a rear wheel driven vehicle is given by

v̇R WD µ a1
= . (6.24)
g h a1 + a2
1−µ
a1 + a2

By setting Fx1 = µ Fz1 and Fx2 = 0, the maximum acceleration for a front wheel driven
vehicle can be calculated in a similar way. One gets

v̇F WD µ a2
= , (6.25)
g h a1 + a2
1+µ
a1 + a2
where the subscript F WD denotes front wheel drive. Depending on the parameter µ , a1 ,
a2 and h the accelerations may be limited by the tilting condition v̇g ≤ ah2 .
The maximum accelerations of a single axle driven vehicle are plotted in Fig. 6.4. For rear
wheel driven passenger cars, the parameter a2 /(a1+a2 ) which describes the static axle load
distribution is in the range of 0.4 ≤ a2 /(a1+a2 ) ≤ 0.5. For µ = 1 and h = 0.55 this results
in maximum accelerations in between 0.77 ≥ v̇/g ≥ 0.64. Front wheel driven passenger
cars usually cover the range 0.55 ≤ a2 /(a1 +a2 ) ≤ 0.60 which produces accelerations in
the range of 0.45 ≤ v̇/g ≥ 0.49. Hence, rear wheel driven vehicles can accelerate much
faster than front wheel driven vehicles.

98
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

range of load distribution

1

RWD

FWD
FWD
0.8

v/g
.
0.6

0.4

RWD
0.2

0
0 0.2 0.4 0.6 0.8 1
a2 / (a1+a2)

Figure 6.4: Single axle driven passenger car: µ = 1, h = 0.55 m, a1 +a2 = 2.5 m

6.3.2 Braking at Single Axle

If only the front axle is braked, in the limit case Fx1 = −µ Fz1 and Fx2 = 0 will hold. With
Eq. (6.5) one gets from Eq. (6.1)

 
a2 h v̇F WB
m v̇F WB = −µ m g − , (6.26)
a1 + a2 a1 + a2 g

where the subscript F WB indicates front wheel braking. Then, the maximum deceleration
is given by
v̇F WB µ a2
= − . (6.27)
g h a1 + a2
1−µ
a1 + a2
If only the rear axle is braked (Fx1 = 0, Fx2 = −µ Fz2 ), one will obtain the maximum
deceleration
v̇R WB µ a1
= − , (6.28)
g h a1 + a2
1+µ
a1 + a2
where the subscript R WB denotes a braked rear axle. Depending on the parameters µ , a1 ,
a2 ,
and h, the decelerations may be limited by the tilting condition
v̇
g
≥ − ah1 .
The maximum decelerations of a single axle braked vehicle are plotted in Fig. 6.5. For
passenger cars the load distribution parameter a2 /(a1 +a2 ) usually covers the range of 0.4
to 0.6. If only the front axle is braked, decelerations from v̇/g = −0.51 to v̇/g = −0.77
will be achieved. This is a quite large value compared to the deceleration range of a braked
rear axle which is in the range of v̇/g = −0.49 to v̇/g = −0.33. Therefore, the braking
system at the front axle has a redundant design.

99
Vehicle Dynamics FH Regensburg, University of Applied Sciences

0
range of
load
distribution
-0.2

v/g
FWB
.
-0.4

-0.6

-0.8
RWB
-1
0 0.2 0.4 0.6 0.8 1
a2 / (a1+a2)

Figure 6.5: Single axle braked passenger car: µ = 1, h = 0.55 m, a1 +a2 = 2.5 m

6.3.3 Optimal Distribution of Drive and Brake Forces

The sum of the longitudinal forces accelerates or decelerates the vehicle. In dimensionless
style Eq. (6.1) reads as
v̇ Fx1 Fx2
= + . (6.29)
g mg mg
A certain acceleration or deceleration can only be achieved by dierent combinations of
the longitudinal forces Fx1 and Fx2 . According to Eq. (6.19) the longitudinal forces are
limited by wheel load and friction.

The optimal combination of Fx1 and Fx2 will be achieved, when front and rear axle have
the same skid resistance.

Fx1 = ± ν µ Fz1 and Fx2 = ± ν µ Fz2 . (6.30)

With Eq. (6.5) and Eq. (6.6) one obtains

 
Fx1 a2 v̇ h
= ±ν µ − (6.31)
mg h g a1 + a2
and  
Fx2 a1 v̇ h
= ±ν µ + . (6.32)
mg h g a1 + a2
With Eq. (6.31) and Eq. (6.32) one gets from Eq. (6.29)

v̇
= ±ν µ , (6.33)
g
where it has been assumed that Fx1 and Fx2 have the same sign. Finally, if Eq. (6.33 is
inserted in Eqs. (6.31) and (6.32) one will obtain
 
Fx1 v̇ a2 v̇ h
= − (6.34)
mg g h g a1 + a2

100
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

and  
Fx2 v̇ a1 v̇ h
= + . (6.35)
mg g h g a1 + a2
Depending on the desired acceleration v̇ > 0 or deceleration v̇ < 0, the longitudinal forces
that grant the same skid resistance at both axles can be calculated now.

B2/mg
braking
Fx1/mg -a1/h

dFx2
0 -1 -2 B1/mg
dFx1 0

a =1.15
1
a =1.35
driving

2
1 h=0.55
µ=1.20

tilting limits

2
a2/h

Fx2/mg

Figure 6.6: Optimal distribution of driving and braking forces

Fig. 6.6 shows the curve of optimal drive and brake forces for typical passenger car values.
At the tilting limits v̇/g = −a1 /h and v̇/g = +a2 /h, no longitudinal forces can be applied
at the lifting axle. The initial gradient only depends on the steady state distribution of
the wheel loads. From Eqs. (6.34) and (6.35) it follows

Fx1
d  
mg a2 v̇ h
= −2 (6.36)
v̇ h g a1 + a2
d
g
and
Fx2
d  
mg a1 v̇ h
= +2 . (6.37)
v̇ h g a1 + a2
d
g

101
Vehicle Dynamics FH Regensburg, University of Applied Sciences

For v̇/g = 0 the initial gradient remains as

d Fx2 a1
= . (6.38)
d Fx1 0 a2

6.3.4 Different Distributions of Brake Forces

Practical applications aim at approximating the optimal distribution of brake forces by

constant distribution, limitation, or reduction of brake forces as good as possible. Fig. 6.7.

Fx1/mg Fx1/mg Fx1/mg

constant
Fx2/mg

Fx2/mg
Fx2/mg

limitation reduction
distribution

Figure 6.7: Dierent distributions of brake forces

When braking, the stability of a vehicle depends on the potential of generating a lateral
force at the rear axle. Thus, a greater skid (locking) resistance is realized at the rear axle
than at the front axle. Therefore, the brake force distribution are all below the optimal
curve in the physically relevant area. This restricts the achievable deceleration, specially
at low friction values.

Because the optimal curve depends on the center of gravity of the vehicle an additional
safety margin have to be installed when designing real brake force distributions. The
distribution of brake forces is often tted to the axle loads. There, the inuence of the
height of the center of gravity, which may also vary much on trucks, is not taken into
account and has to be compensated by a safety margin from the optimal curve. Only the
control of brake pressure in anti-lock-systems provides an optimal distribution of brake
forces independently from loading conditions.

6.3.5 Anti-Lock-Systems

On hard braking maneuvers large longitudinal slip values occur. Then, the stability and/or
steerability is no longer given because nearly no lateral forces can be generated. By control-
ling the brake torque or brake pressure respectively, the longitudinal slip can be restricted
to values that allow considerable lateral forces.

Here, the angular wheel acceleration Ω̇ is used as a control variable. Angular accelerations
of the wheel are derived from the measured angular speeds of the wheel by dierentiation.
The rolling condition is fullled with a longitudinal slip of sL = 0. Then

rD Ω̇ = ẍ (6.39)

102
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

holds, where rD labels the dynamic tire radius and ẍ names the longitudinal acceleration
of the vehicle. According to Eq. (6.21), the maximum acceleration/deceleration of a vehicle
depends on the friction coecient, |ẍ| = µ g . For a given friction coecient µ a simple
control law can be realized for each wheel

1
|Ω̇| ≤ |ẍ| . (6.40)
rD
Because no reliable possibility to determine the local friction coecient between tire and
road has been found until today, useful information can only be gained from Eq. (6.40)
at optimal conditions on dry road. Therefore, the longitudinal slip is used as a second
control variable.

In order to calculate longitudinal slips, a reference speed is estimated from all measured
wheel speeds which is used for the calculation of slip at all wheels, then. This method
is too imprecise at low speeds. Therefore, no control is applied below a limit velocity.
Problems also arise when all wheels lock simultaneously for example which may happen
on icy roads.

The control of the brake torque is done via the brake pressure which can be increased,
held, or decreased by a three-way valve. To prevent vibrations, the decrement is usually
made slower than the increment.

To prevent a strong yaw reaction, the select low principle is often used with µ-split braking
at the rear axle. Here, the break pressure at both wheels is controlled by the wheel running
on lower friction. Thus, at least the brake forces at the rear axle cause no yaw torque.
However, the maximum achievable deceleration is reduced by this.

6.4 Drive and Brake Pitch

6.4.1 Vehicle Model

The vehicle model in Fig. 6.8 consists of ve rigid bodies. The body has three degrees
of freedom: Longitudinal motion xA , vertical motion zA and pitch βA . The coordinates
z1 and z2 describe the vertical motions of wheel and axle bodies relative to the body.
The longitudinal and rotational motions of the wheel bodies relative to the body can be
described via suspension kinematics as functions of the vertical wheel motion:

x1 = x1 (z1 ) , β1 = β1 (z1 ) ;
(6.41)
x2 = x2 (z2 ) , β2 = β2 (z2 ) .

The rotation angles ϕR1 and ϕR2 describe the wheel rotations relative to the wheel bodies.

The forces between wheel body and vehicle body are labeled FF 1 and FF 2 . At the wheels
drive torques MA1 , MA2 and brake torques MB1 , MB2 , longitudinal forces Fx1 , Fx2 and
the wheel loads Fz1 , Fz2 apply. The brake torques are directly supported by the wheel
bodies, whereas the drive torques are transmitted by the drive shafts to the vehicle body.

103
Vehicle Dynamics FH Regensburg, University of Applied Sciences

zA

xA
MA1 FF1
βA
z1
MA1 MB1
hR FF2
MA2
ϕR1
MB1 z2

Fz1 Fx1 MA2 MB2

a1 R ϕR2
MB2

a2
Fz2 Fx2

Figure 6.8: Simple vehicle model

The forces and torques that apply to the single bodies are listed in the last column of the
tables 6.1 and 6.2.

The velocity of the vehicle body and its angular velocity are given by

     
ẋA 0 0
v0A,0 =  0  +  0 ; ω0A,0 =  β̇A  . (6.42)
0 żA 0

At small rotational motions of the body one gets for the velocities of the wheel bodies
and wheels

∂x1
       
ẋA 0 −hR β̇A ∂z1
ż1
v0RK1 ,0 = v0R1 ,0 =  0  +  0  +  0  +  0 ; (6.43)
0 żA −a1 β̇A ż1
∂x2
       
ẋA 0 −hR β̇A ∂z2
ż2
v0RK2 ,0 = v0R2 ,0 =  0  +  0  +  0  +  0 . (6.44)
0 żA +a2 β̇A ż2
The angular velocities of the wheel bodies and wheels are obtained from

         
0 0 0 0 0
ω0RK1 ,0 =  β̇A  +  β̇1  and ω0R1 ,0 =  β̇A  +  β̇1  +  ϕ̇R1 
0 0 0 0 0
(6.45)

104
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

as well as
         
0 0 0 0 0
ω0RK2 ,0 =  β̇A  +  β̇2  and ω0R2 ,0 =  β̇A  +  β̇2  +  ϕ̇R2 
0 0 0 0 0
(6.46)
Introducing a vector of generalized velocities

 T
z = ẋA żA β̇A β̇1 ϕ̇R1 β̇2 ϕ̇R2 , (6.47)

the velocities and angular velocities given by Eqs. (6.42), (6.43), (6.44), (6.45), and (6.46)
can be written as
7 7
X ∂v0i X ∂ω0i
v0i = zj and ω0i = zj (6.48)
j=1
∂zj j=1
∂zj

6.4.2 Equations of Motion

∂v0i ∂ω0i
The partial velocities
∂zj
and partial angular velocities
∂zj
for the ve bodies i = 1(1)5
and for the seven generalized speeds j = 1(1)7 are arranged in the tables 6.1 and 6.2.

partial velocities ∂v0i /∂zj applied forces

bodies ẋA żA β̇A ż1 ϕ̇R1 ż2 ϕ̇R2 Fie

1 0 0 0 0 0 0 0
chassis
mA 0 0 0 0 0 0 0 0
0 1 0 0 0 0 0 FF 1 +FF 2 −mA g
∂x1
wheel body
1 0 −hR ∂z1 0 0 0 0
front 0 0 0 0 0 0 0 0
mRK1 0 1 −a1 1 0 0 0 −FF 1 −mRK1 g
∂x1
wheel
1 0 −hR ∂z1 0 0 0 Fx1
front 0 0 0 0 0 0 0 0
mR1 0 1 −a1 1 0 0 0 Fz1 −mR1 g
∂x2
wheel body
1 0 −hR 0 0 ∂z2 0 0
rear 0 0 0 0 0 0 0 0
mRK2 0 1 a2 0 0 1 0 −FF 2 −mRK2 g
∂x2
wheel
1 0 −hR 0 0 ∂z2 0 Fx2
rear 0 0 0 0 0 0 0 0
mR2 0 1 a2 0 0 1 0 Fz2 −mR2 g

Table 6.1: Partial velocities and applied forces

With the aid of the partial velocities and partial angular velocities the elements of the
mass matrix M and the components of the vector of generalized forces and torques Q can
be calculated.

5  T 5  T
X ∂v0k ∂v0k X ∂ω0k ∂ω0k
M (i, j) = mk + Θk ; i, j = 1(1)7 ; (6.49)
k=1
∂zi ∂zj k=1
∂zi ∂zj

105
Vehicle Dynamics FH Regensburg, University of Applied Sciences

partial angular velocities ∂ω0i /∂zj applied torques

bodies ẋA żA β̇A ż1 ϕ̇R1 ż2 ϕ̇R2 Mie

0 0 0 0 0 0 0 0
chassis
ΘA 0 0 1 0 0 0 0 −MA1−MA2 −a1 FF 1 +a2 FF 2
0 0 0 0 0 0 0 0
wheel body
0 0 0 0 0 0 0 0
∂β1
front 0 0 1 ∂z1
0 0 0 MB1
ΘRK1 0 0 0 0 0 0 0 0
wheel
0 0 0 0 0 0 0 0
∂β1
front 0 0 1 ∂z1
1 0 0 MA1 −MB1 −R Fx1
ΘR1 0 0 0 0 0 0 0 0
wheel body
0 0 0 0 0 0 0 0
∂β2
rear 0 0 1 0 0 ∂z2
0 M B2
ΘRK2 0 0 0 0 0 0 0 0
wheel
0 0 0 0 0 0 0 0
∂β2
rear 0 0 1 0 0 ∂z2
1 M A2 −M B2 −R Fx2
ΘR2 0 0 0 0 0 0 0 0

Table 6.2: Partial angular velocities and applied torques

5  T 5  T
X ∂v0k X ∂ω0k
Q(i) = Fke + Mke ; i = 1(1)7 . (6.50)
k=1
∂zi k=1
∂zi
Then, the equations of motion for the plane vehicle model are given by

M ż = Q . (6.51)

6.4.3 Equilibrium

With the abbreviations

m1 = mRK1 + mR1 ; m2 = mRK2 + mR2 ; mG = mA + m1 + m2 (6.52)

and
h = hR + R (6.53)

The components of the vector of generalized forces and torques read as

Q(1) = Fx1 + Fx2 ;

Q(2) = Fz1 + Fz2 − mG g ; (6.54)

Q(3) = −a1 Fz1 + a2 Fz2 − h(Fx1 + Fx2 ) + a1 m1 g − a2 m2 g ;

∂x1 ∂β1
Q(4) = Fz1 − FF 1 + ∂z1
Fx1 − m1 g + ∂z1
(MA1 − R Fx1 ) ;
(6.55)
Q(5) = MA1 − MB1 − R Fx1 ;

106
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

∂x2 ∂β2
Q(6) = Fz2 − FF 2 + ∂z2
Fx2 − m2 g + ∂z2
(MA2 − R Fx2 ) ;
(6.56)
Q(7) = MA2 − MB2 − R Fx2 .
Without drive and brake forces

MA1 = 0 ; MA2 = 0 ; MB1 = 0 ; MB2 = 0 (6.57)

from Eqs. (6.54), (6.55) and (6.56) one gets the steady state longitudinal forces, the spring
preloads, and the wheel loads

0 0
Fx1 = 0; Fx2 = 0;
b a
FF0 1 = a+b
mA g ; FF0 2 = a+b
mA g ; (6.58)
0 a2 0 a1
Fz1 = m1 g + a1 +a2
mA g ; Fz2 = m2 g + a1 +a2
mA g .

6.4.4 Driving and Braking

Assuming that on accelerating or decelerating the vehicle the wheels neither slip nor lock,

∂x1
R ϕ̇R1 = ẋA − hR β̇A + ∂z1
ż1 ;
(6.59)
∂x2
R ϕ̇R2 = ẋA − hR β̇A + ∂z2
ż2

hold. In steady state the pitch motion of the body and the vertical motion of the wheels
reach constant values

βA = βAst = const. , z1 = z1st = const. , z2 = z2st = const. (6.60)

and Eq. (6.59) simplies to

R ϕ̇R1 = ẋA ; R ϕ̇R2 = ẋA . (6.61)

With Eqs. (6.60), (6.61) and (6.53) the equation of motion (6.51) results in

a a
mG ẍA = Fx1 + Fx2 ;
a a
0 = Fz1 + Fz2 ;
ẍA ẍA a a a a
−hR (m1 +m2 ) ẍA + ΘR1 R
+ ΘR2 R
= −a Fz1 + b Fz2 − (hR + R)(Fx1 + Fx2 );
(6.62)
∂β1 ∂β1
∂x1
∂z1
m1 ẍA + ∂z1
ΘR1 ẍRA = a
Fz1 − FFa 1 + ∂x1
∂z1
a
Fx1 + ∂z1
(MA1 − a
R Fx1 ) ;
(6.63)
ẍA a
ΘR1 R
= MA1 − MB1 − R Fx1 ;

∂x2 ∂β2 ẍA a ∂x2 ∂β2

∂z2
m2 ẍA + ∂z2
ΘR2 R
= Fz2 − FFa 2 + ∂z2
a
Fx2 + ∂z2
(MA2 a
− R Fx2 );
(6.64)
ẍA a
ΘR2 R
= MA2 − MB2 − R Fx2 ;

107
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where the steady state spring forces, longitudinal forces, and wheel loads have been sep-
arated into initial and acceleration-dependent terms

st 0 a
Fxi = Fxi + Fxi ; Fzist = Fzi0 + Fzia ; FFsti = FF0 i + FFa i ; i = 1, 2 . (6.65)

a
With given torques of drive and brake the vehicle acceleration ẍA , the wheel forces Fx1 ,
a a a a a
Fx2 , Fz1 , Fz2 and the spring forces FF 1 , FF 2 can be calculated from Eq. (6.62), Eq. (6.63)
and Eq. (6.64)

Via the spring characteristics which have been assumed as linear the acceleration-
dependent forces also cause a vertical displacement and pitch motion of the body besides
the vertical motions of the wheels,

FFa 1 = cA1 z1a ,

FFa 2 = cA2 z2a ,
(6.66)
a
Fz1 = −cR1 (zAa − a βAa + z1a ) ,
a
Fz2 = −cR2 (zAa + b βAa + z2a ) .

Especially the pitch of the vehicle βAa 6= 0, caused by drive or brake will be felt as annoying,
if too distinct.

By an axle kinematics with 'anti dive' and/or 'anti squat' properties, the drive and/or
brake pitch angle can be reduced by rotating the wheel body and moving the wheel center
in longitudinal direction during the suspension travel.

6.4.5 Brake Pitch Pole

For real suspension systems the brake pitch pole can be calculated from the motions of
the wheel contact points in the x-, z -plane, Fig. 6.9.

pitch pole

x-, z- motion of the contact points

during compression and rebound

Figure 6.9: Brake pitch pole

Increasing the pitch pole height above the track level means a decrease in the brake pitch
angle. However, the pitch pole is not set above the height of the center of gravity in
practice, because the front of the vehicle would rise at braking then.

108
7 Lateral Dynamics

7.1 Kinematic Approach

7.1.1 Kinematic Tire Model

When a vehicle drives through a curve at low lateral acceleration, small lateral forces
will be needed for course holding. Then, hardly lateral slip occurs at the wheels. In the
ideal case at vanishing lateral slip the wheels only move in circumferential direction. The
velocity component of the contact point in the lateral direction of the tire vanishes then

vy = eTy v0P = 0 . (7.1)

This constraint equation can be used as 'kinematic tire model' for course calculation of
vehicles moving in the low lateral acceleration range.

7.1.2 Ackermann Geometry

Within the validity limits of the kinematic tire model the necessary steering angle of the
front wheels can be constructed via the given momentary pivot pole M, Fig. 7.1.

At slowly moving vehicles the lay out of the steering linkage is usually done according to
the Ackermann geometry. Then, the following relations apply
a a
tan δ1 = and tan δ2 = , (7.2)
R R+s
where s labels the track width and a denotes the wheel base. Eliminating the curve radius
R, we get
a a tan δ1
tan δ2 = a or tan δ2 = . (7.3)
+s a + s tan δ1
tan δ1
a A a
The deviations 4δ2 = δ2 − δ2 of the actual steering angle δ2 from the Ackermann steering
A
angle δ2 , which follows from Eq. (7.3), are used, especially on commercial vehicles, to
judge the quality of a steering system.

At a rotation around the momentary pole M, the direction of the velocity is xed for
every point of the vehicle. The angle β between the velocity vector v and the longitudinal
axis of the vehicle is called side slip angle. The side slip angle at point P is given by
x x
tan βP = or tan βP = tan δ1 , (7.4)
R a
where x denes the distance of P to the inner rear wheel.

109
Vehicle Dynamics FH Regensburg, University of Applied Sciences

δ1 δ2

v
βP

P
a

δ1
M βP δ2

R s

Figure 7.1: Ackermann steering geometry at a two-axled vehicle

7.1.3 Space Requirement

The Ackermann approach can also be used to calculate the space requirement of a vehicle
during cornering, Fig. 7.2. If the front wheels of a two-axled vehicle are steered according
to the Ackermann geometry, the outer point of the vehicle front will run on the maximum
radius Rmax , whereas a point on the inner side of the vehicle at the location of the rear
axle will run on the minimum radius Rmin . Hence, it holds

2
Rmax = (Rmin + b)2 + (a + f )2 , (7.5)

where a, b are the wheel base and the width of the vehicle, and f species the distance from
the front of the vehicle to the front axle. Then, the space requirement 4R = Rmax − Rmin
can be specied as a function of the cornering radius Rmin for a given vehicle dimension

q
4R = Rmax − Rmin = (Rmin + b)2 + (a + f )2 − Rmin . (7.6)

The space requirement 4R of a typical passenger car and a bus is plotted in Fig. 7.3
versus the minimum cornering radius. In narrow curves Rmin = 5.0 m, a bus requires a
space of 2.5 times the width, whereas a passenger car needs only 1.5 times the width.

110
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

R
ma
af x

Rmin
M

Figure 7.2: Space requirement

7
bus: a=6.25 m, b=2.50 m, f=2.25 m
6 car: a=2.50 m, b=1.60 m, f=1.00 m

4
∆ R [m]

0
0 10 20 30 40 50
R min [m]

Figure 7.3: Space requirement of a typical passenger car and bus

111
Vehicle Dynamics FH Regensburg, University of Applied Sciences

7.1.4 Vehicle Model with Trailer

7.1.4.1 Kinematics

Fig. 7.4 shows a simple lateral dynamics model for a two-axled vehicle with a single-axled
trailer. Vehicle and trailer move on a horizontal track. The position and the orientation
of the vehicle relative to the track xed frame x0 , y0 , z0 is dened by the position vector
to the rear axle center  
x
r02,0 =  y  (7.7)
 

R
and the rotation matrix  
cos γ − sin γ 0
A02 =  sin γ cos γ 0  . (7.8)
0 0 1
Here, the tire radius R is considered to be constant, and x, y as well as the yaw angle γ
are generalized coordinates.

1
x
1
y

a
A1
y2
b γ
x2
K A2
c

y0

κ
3
x
y3

A3 x0

Figure 7.4: Kinematic model with trailer

The position vector

 
a
r01,0 = r02,0 + A02 r21,2 with r21,2 = 0  (7.9)
0

112
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

and the rotation matrix

 
cos δ − sin δ 0
A01 = A02 A21 with A21 =  sin δ cos δ 0  (7.10)
0 0 1

describe the position and the orientation of the front axle, where a = const labels the
wheel base and δ the steering angle.

The position vector

r03,0 = r02,0 + A02 r2K,2 + A23 rK3,3 (7.11)

with    
−b −c
r2K,2 = 0  and rK3,2 = 0  (7.12)
0 0
and the rotation matrix
 
cos κ − sin κ 0
A03 = A02 A23 with A23 =  sin κ cos κ 0  (7.13)
0 0 1

dene the position and the orientation of the trailer axis, with κ labeling the bend angle
between vehicle and trailer, and b, c marking the distances from the rear axle 2 to the
coupling point K and from the coupling point K to the trailer axis 3.

7.1.4.2 Vehicle Motion

According to the kinematic tire model, cf. section 7.1.1, the velocity at the rear axle can
only have a component in the longitudinal direction of the tire which here corresponds
with the longitudinal direction of the vehicle

 
vx2
v02,2 =  0 . (7.14)
0

The time derivative of Eq. (7.7) results in

 
ẋ
v02,0 = ṙ02,0 =  ẏ  . (7.15)
0

The transformation of Eq. (7.14) into the system 0

  
vx2 cos γ vx2
v02,0 = A02 v02,2 = A02  0  =  sin γ vx2  (7.16)
0 0

113
Vehicle Dynamics FH Regensburg, University of Applied Sciences

compared to Eq. (7.15) results in two rst order dierential equations for the position
coordinates x and y
ẋ = vx2 cos γ , ẏ = vx2 sin γ . (7.17)

The velocity at the front axle follows from Eq. (7.9)

v01,0 = ṙ01,0 = ṙ02,0 + ω02,0 × A02 r21,2 . (7.18)

Transformed into the vehicle xed system x2 , y2 , z2 we obtain

       
vx2 0 a vx2
v01,2 =  0  +  0  ×  0  =  a γ̇  . (7.19)
0 γ̇ 0 0
| {z } | {z } | {z }
v02,2 ω02,2 r21,2

The unit vectors

   
cos δ − sin δ
ex1,2 =  sin δ  and ey1,2 =  cos δ  (7.20)
0 0

dene the longitudinal and lateral direction at the front axle. According to Eq. (7.1) the
velocity component lateral to the wheel must vanish,

eTy1,2 v01,2 = − sin δ vx2 + cos δ a γ̇ = 0 . (7.21)

Whereas in longitudinal direction the velocity

eTx1,2 v01,2 = cos δ vx2 + sin δ a γ̇ = vx1 (7.22)

remains. From Eq. (7.21) a rst order dierential equation follows for the yaw angle

vx2 (7.23)
γ̇ = tan δ .
a

7.1.4.3 Entering a Curve

In analogy to Eq. (7.2) the steering angle δ can be related to the current track radius R
or with k = 1/R to the current track curvature

a
tan δ = = ak . (7.24)
R
Then, the dierential equation for the yaw angle reads as

γ̇ = vx2 k . (7.25)

With the curvature gradient

t
k = k(t) = kC , (7.26)
T

114
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

the entering of a curve is described as a continuous transition from a straight line with
the curvature k=0 into a circle with the curvature k = kC .
The yaw angle of the vehicle can be calculated by simple integration now

vx2 kC t2
γ(t) = , (7.27)
T 2
where at time t=0 a vanishing yaw angle, γ(t = 0) = 0, has been assumed. Then, the
position of the vehicle follows with Eq. (7.27) from the dierential equations Eq. (7.17)

Zt=T Zt=T
vx2 kC t2 vx2 kC t2
   
x = vx2 cos dt , y = vx2 sin dt . (7.28)
T 2 T 2
t=0 t=0

At constant vehicle speed, vx2 = const., Eq. (7.28) is the parameterized form of a clothoide.
From Eq. (7.24) the necessary steering angle can be calculated, too. If only small steering
angles are necessary for driving through the curve, the tan-function can be approximated
by its argument, and
t
δ = δ(t) ≈ a k = a kC (7.29)
T
holds, i.e. the driving through a clothoide is manageable by a continuous steer motion.

7.1.4.4 Trailer Motions

The velocity of the trailer axis can be obtained by dierentiation of the position vector
Eq. (7.11)
v03,0 = ṙ03,0 = ṙ02,0 + ω02,0 × A02 r23,2 + A02 ṙ23,2 . (7.30)

The velocity ṙ02,0 = v02,0 and the angular velocity ω02,0 of the vehicle are dened in
Eqs. (7.16) and (7.19). The position vector from the rear axle to the axle of the trailer is
given by
 
−b − c cos κ
r23,2 = r2K,2 + A23 rK3,3 =  −c sin κ  , (7.31)
0
where r2K,2 and rK3,3 are dened in Eq. (7.12). The time derivative of Eq. (7.31) results
in     
0 −c cos κ c sin κ κ̇
ṙ23,2 =  0  ×  −c sin κ  =  −c cos κ κ̇  . (7.32)
κ̇ 0 0
| {z } | {z }
ω23,2 A23 rK3,3
Eq. (7.30) is transformed into the vehicle xed frame x2 , y2 , z2 now
         
vx2 0 −b − c cos κ c sin κ κ̇ vx2 + c sin κ (κ̇+ γ̇)
v03,2 =  0  +  0 × −c sin κ  +  −c cos κ κ̇  =  −b γ̇ − c cos κ (κ̇+ γ̇)  .
0 γ̇ 0 0 0
| {z } | {z } | {z } | {z }
v02,2 ω02,2 r23,2 ṙ23,2
(7.33)

115
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The longitudinal and lateral direction at the trailer axle are dened by the unit vectors
   
cos κ − sin κ
ex3,2 =  sin κ  and ey3,2 =  cos κ  . (7.34)
0 0
At the trailer axis the lateral velocity must also vanish

eTy3,2 v03,2 = − sin κ vx2 + c sin κ (κ̇+ γ̇) + cos κ −b γ̇ − c cos κ (κ̇+ γ̇) = 0 ,



(7.35)

whereas in longitudinal direction the velocity

eTx3,2 v03,2 = cos κ vx2 + c sin κ (κ̇+ γ̇) + sin κ −b γ̇ − c cos κ (κ̇+ γ̇) = vx3



(7.36)

remains. If Eq. (7.23) is inserted into Eq. (7.35) now, one will get a rst order dierential
equation for the bend angle
 
vx2
 
a b
κ̇ = − sin κ + cos κ + 1 tan δ . (7.37)
a c c
The dierential equations Eq. (7.17) and Eq. (7.23) describe position and orientation
within the x0 , y0 plane. The position of the trailer relative to the vehicle follows from
Eq. (7.37).

7.1.4.5 Course Calculations

20 front axle
rear axle
[m]
trailer axle
10

0
-30 -20 -10 0 10 20 30 [m] 40 50 60
30
[o] front axle steering angle δ
20

10

0
0 5 10 15 20 [s] 25 30

Figure 7.5: Entering a curve

For a given set of vehicle parameters a, b , c , and predened time functions of the vehicle
velocity, vx2 = vx2 (t) and the steering angle, δ = δ(t), the course of vehicle and trailer
can be calculated by numerical integration of the dierential equations Eqs. (7.17), (7.23)
and (7.37). If the steering angle is slowly increased at constant driving speed, the vehicle
drives a gure which will be similar to a clothoide, Fig. 7.5.

116
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

7.2 Steady State Cornering

7.2.1 Cornering Resistance

In a body xed reference frame B, Fig. 7.6, the velocity state of the vehicle can be
described by
   
v cos β 0
v0C,B =  v sin β  and ω0F,B = 0  , (7.38)
0 ω
where β denotes the side slip angle of the vehicle measured at the center of gravity. The
angular velocity of a vehicle cornering with constant velocity v on an at horizontal road
is given by
v
ω= , (7.39)
R
where R denotes the radius of curvature.

a2 Fx2
Fy2

a1 C
ω R
β yB
v
xB

Fx1 Fy1

Figure 7.6: Cornering resistance

In the body xed reference frame, linear and angular momentum result in

v2
 
m − sin β = Fx1 cos δ − Fy1 sin δ + Fx2 , (7.40)
R
 2 
v
m cos β = Fx1 sin δ + Fy1 cos δ + Fy2 , (7.41)
R
0 = a1 (Fx1 sin δ + Fy1 cos δ) − a2 Fy2 , (7.42)

117
Vehicle Dynamics FH Regensburg, University of Applied Sciences

where m denotes the mass of the vehicle, Fx1 , Fx2 , Fy1 , Fy2 are the resulting forces in
longitudinal and vertical direction applied at the front and rear axle, and δ species the
average steer angle at the front axle.

The engine torque is distributed by the center dierential to the front and rear axle. Then,
in steady state condition we obtain

Fx1 = k FD and Fx2 = (1 − k) FD , (7.43)

where FD is the driving force and by k dierent driving conditions can be modeled:

k=0 rear wheel drive Fx1 = 0, Fx2 = FD

Fx1 k
0<k<1 all wheel drive =
Fx2 1−k
k=1 front wheel drive Fx1 = FD , Fx2 = 0

If we insert Eq. (7.43) into Eq. (7.40) we will get

  mv 2
k cos δ + (1−k) FD − sin δ Fy1 = − sin β ,
R
mv 2 (7.44)
k sin δ FD + cos δ Fy1 + Fy2 = cos β ,
R
a1 k sin δ FD + a1 cos δ Fy1 − a2 Fy2 = 0 .

These equations can be resolved for the driving force

a2
cosβ sin δ − sin β cosδ
a1 + a2 mv 2
FD = . (7.45)
k + (1 − k) cos δ R
The driving force will vanish, if

a2 a2
cosβ sin δ = sin β cosδ or tan δ = tan β (7.46)
a1 + a2 a1 + a2
holds. This fully corresponds with the Ackermann geometry. But, the Ackermann geom-
etry applies only for small lateral accelerations. In real driving situations, the side slip
angle of a vehicle at the center of gravity is always smaller than the Ackermann side slip
a2
angle. Then, due to tan β < tan δ a driving force FD > 0 is needed to overcome the
a1 +a2
'cornering resistance' of the vehicle.

7.2.2 Overturning Limit

The overturning hazard of a vehicle is primarily determined by the track width and the
height of the center of gravity. With trucks however, also the tire deection and the body
roll have to be respected., Fig. 7.7.

118
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

α2 α1

m ay

h2
mg

h1

F yL F yR
FzR
FzL s/2
s/2

Figure 7.7: Overturning hazard on trucks

The balance of torques at the height of the track plane applied at the already inclined
vehicle results in
s
(FzL − FzR ) = m ay (h1 + h2 ) + m g [(h1 + h2 )α1 + h2 α2 ] , (7.47)
2
where ay describes the lateral acceleration, m is the sprung mass, and small roll angles of
the axle and the body were assumed, α1  1, α2  1.

On a left-hand tilt, the right tire raises

T
FzR = 0, (7.48)

whereas the left tire carries the complete vehicle weight

T
FzL = mg . (7.49)

Using Eqs. (7.48) and (7.49) one gets from Eq. (7.47)

s
aTy 2 h2
= − α1T − α2T . (7.50)
g h1 + h2 h1 + h2
T
The vehicle will turn over, when the lateral acceleration ay
rises above the limit ay . Roll
T T
of axle and body reduce the overturning limit. The angles α1 and α2 can be calculated
from the tire stiness cR and the roll stiness of the axle suspension.

119
Vehicle Dynamics FH Regensburg, University of Applied Sciences

If the vehicle drives straight ahead, the weight of the vehicle will be equally distributed
to both sides
stat stat 1
FzR = FzL = mg . (7.51)
2
With
T stat
FzL = FzL + 4Fz (7.52)

and Eqs. (7.49), (7.51), one obtains for the increase of the wheel load at the overturning
limit
1
4Fz = mg . (7.53)
2
Then, the resulting tire deection follows from

4Fz = cR 4r , (7.54)

where cR is the radial tire stiness.

Because the right tire simultaneously rebounds with the same amount, for the roll angle
of the axle
2 4r mg
2 4r = s α1T or α1T = = (7.55)
s s cR
holds. In analogy to Eq. (7.47) the balance of torques at the body applied at the roll
center of the body yields

cW ∗ α2 = m ay h2 + m g h2 (α1 + α2 ) , (7.56)

where cW names the roll stiness of the body suspension. In particular, at the overturning
limit ay = aTy
T
aTy mgh2 mgh2
α2 = + α1T (7.57)
g cW − mgh2 cW − mgh2
T
applies. Not allowing the vehicle to overturn already at ay = 0 demands a minimum of
min
roll stiness cW > cW = mgh2 . With Eqs. (7.55) and (7.57) the overturning condition
Eq. (7.50) reads as

aTy s 1 aTy 1 1 1
(h1 + h2 ) = − (h1 + h2 ) ∗ − h2 ∗
− h2 ∗ , (7.58)
g 2 cR g cW − 1 cW − 1 cR ∗
where, for abbreviation purposes, the dimensionless stinesses

cR cW
c∗R = m g and c∗W = (7.59)
m g h2
s
have been used. Resolved for the normalized lateral acceleration

s
aTy 2 1
= − (7.60)
g h2 c∗R
h1 + h2 + ∗
cW − 1

120
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

roll angle α=αT1+α2

T
overturning limit ay
0.6 20

0.5
15
0.4

0.3 10

0.2
5
0.1

0 0
0 10 20 0 10 20
*
normalized roll stiffness c W *
normalized roll stiffness c W

Figure 7.8: Tilting limit for a typical truck at steady state cornering

remains.

At heavy trucks, a twin tire axle may be loaded with m = 13 000 kg . The radial stiness
of one tire is cR = 800 000 N/m, and the track width can be set to s = 2 m. The values
h1 = 0.8 m and h2 = 1.0 m hold at maximal load. These values produce the results shown
∗
in Fig. 7.8. Even with a rigid body suspension cW → ∞, the vehicle turns over at a lateral
acceleration of ay ≈ 0.5 g . Then, the roll angle of the vehicle solely results from the tire
deection.

At a normalized roll stiness of c∗W = 5, the overturning limit lies at ay ≈ 0.45 g and
so reaches already 90% of the maximum. The vehicle will turn over at a roll angle of
α = α1 + α2 ≈ 10◦ then.

7.2.3 Roll Support and Camber Compensation

When a vehicle drives through a curve with the lateral acceleration ay , centrifugal forces
will be applied to the single masses. At the simple roll model in Fig. 7.9, these are the
forces m A ay and m R ay , where mA names the body mass and mR the wheel mass.

Through the centrifugal force m A ay applied to the body at the center of gravity, a torque
is generated, which rolls the body with the angle αA and leads to an opposite deection
of the tires z1 = −z2 .
At steady state cornering, the vehicle forces are balanced. With the principle of virtual
work
δW = 0 , (7.61)

the equilibrium position can be calculated.

At the simple vehicle model in Fig. 7.9 the suspension forces FF 1 , FF 2 and tire forces
Fy1 , Fz1 , Fy2 , Fz2 , are approximated by linear spring elements with the constants cA and

121
Vehicle Dynamics FH Regensburg, University of Applied Sciences

b/2 b/2

zA αA

mA a y SA yA

FF2 FF1
h0
z2 α2 z1 α1
mR a y S2 mR a y S1
r0 y2 y1
Q2 Q1
F y2 F y1
Fy2 F z1

Figure 7.9: Simple vehicle roll model

cQ , cR . The work W of these forces can be calculated directly or using W = −V via the
potential V. At small deections with linearized kinematics one gets

W = −mA ay yA
−mR ay (yA + hR αA + y1 )2 − mR ay (yA + hR αA + y2 )2
− 21 cA z12 − 1
2
cA z22
(7.62)
− 21 cS (z1 − z2 )2
− 21 cQ (yA + h0 αA + y1 + r0 α1 )2 − 12 cQ (yA + h0 αA + y2 + r0 α2 )2

2
2
− 21 cR zA + 2b αA + z1 − 21 cR zA − 2b αA + z2 ,
where the abbreviation hR = h0 − r0 has been used, and cS describes the spring constant
of the anti roll bar, converted to the vertical displacement of the wheel centers.

The kinematics of the wheel suspension are symmetrical. With the linear approaches

∂y ∂α ∂y ∂α
y1 = z1 , α1 = α1 and y2 = − z2 , α2 = − α2 (7.63)
∂z ∂z ∂z ∂z
the work W can be described as a function of the position vector

y = [ yA , zA , αA , z1 , z2 ]T . (7.64)

Due to
W = W (y) (7.65)

the principle of virtual work Eq. (7.61) leads to

∂W
δW = δy = 0 . (7.66)
∂y

122
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Because of δy 6= 0, a system of linear equations in the form of

Ky = b (7.67)

results from Eq. (7.66). The matrix K and the vector b are given by

 
∂y Q Q
2 cQ 0 2 cQ h0 ∂z Q
c − ∂y∂z cQ
 
 

 0 2 cR 0 cR cR 

 
Q Q
K =  2 cQ h0 b
c +h0 ∂y∂z cQ b ∂y (7.68)
 
 0 cα 2 R
− 2 cR −h0 ∂z cQ 

 
 ∂yQ c cR b Q
c +h0 ∂y∂z cQ c∗A + cS + cR −cS 
 ∂z Q 2 R 
 
Q Q
− ∂y∂z cQ cR − 2b cR −h0 ∂y∂z cQ −cS ∗
cA + cS + cR

and  
mA + 2 mR
0
 
 
 
b = −
 (m1 + m2 ) hR  ay .
 (7.69)

mR ∂y/∂z
 
 
−mR ∂y/∂z
The following abbreviations have been used:

2  2
∂y Q

∂y ∂α ∂y b
= + r0 , c∗A = cA + cQ , cα = 2 cQ h20 + 2 cR . (7.70)
∂z ∂z ∂z ∂z 2

The system of linear equations Eq. (7.67) can be solved numerically, e.g. with MATLAB.
Thus, the inuence of axle suspension and axle kinematics on the roll behavior of the
vehicle can be investigated.

a) αA b) αA

γ2 γ2 0
γ1 γ1 0 roll center
roll center

Figure 7.10: Roll behavior at cornering: a) without and b) with camber compensation

If the wheels only move vertically to the body at jounce and rebound, at fast cornering
the wheels will be no longer perpendicular to the track Fig. 7.10 a. The camber angles
γ1 > 0 and γ2 > 0 result in an unfavorable pressure distribution in the contact area,
which leads to a reduction of the maximally transmittable lateral forces. Thus, at more

123
Vehicle Dynamics FH Regensburg, University of Applied Sciences

sportive vehicles axle kinematics are employed, where the wheels are rotated around the
longitudinal axis at jounce and rebound, α1 = α1 (z1 ) and α2 = α2 (z2 ). Hereby, a camber
compensation can be achieved with γ1 ≈ 0 and γ2 ≈ 0. Fig. 7.10 b. By the rotation of
the wheels around the longitudinal axis on jounce and rebound, the wheel contact points
are moved outwards, i.e against the lateral force. By this, a 'roll support' is achieved that
reduces the body roll.

7.2.4 Roll Center and Roll Axis

roll axis
roll center front roll center rear

Figure 7.11: Roll axis

The 'roll center' can be constructed from the lateral motion of the wheel contact points
Q1 and Q2 , Fig. 7.10. The line through the roll center at the front and rear axle is called
'roll axis', Fig. 7.11.

7.2.5 Wheel Loads

-TT
+TT

PR0+∆P PR0-∆P PR0+∆PR

PR0-∆PR
PF0+∆P PF0-∆P PF0+∆PF PF0-∆PF

Figure 7.12: Wheel loads for a exible and a rigid chassis

The roll angle of a vehicle during cornering depends on the roll stiness of the axle and
on the position of the roll center. Dierent axle layouts at the front and rear axle may
result in dierent roll angles of the front and rear part of the chassis, Fig. 7.12.

124
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

On most passenger cars the chassis is rather sti. Hence, front and rear part of the chassis
are forced by an internal torque to an overall chassis roll angle. This torque aects the
wheel loads and generates dierent wheel load dierences at the front and rear axle. Due
to the degressive inuence of the wheel load to longitudinal and lateral tire forces the
steering tendency of a vehicle can be aected.

7.3 Simple Handling Model

7.3.1 Modeling Concept

x0
a2

a1
Fy2 y0

x2
y2
C
γ
xB
Fy1 β
v yB
y1
x1
δ

Figure 7.13: Simple handling model

The main vehicle motions take place in a horizontal plane dened by the earth-xed
frame 0, Fig. 7.13. The tire forces at the wheels of one axle are combined to one resulting
force. Tire torques, rolling resistance, and aerodynamic forces and torques, applied at the
vehicle, are not taken into consideration.

7.3.2 Kinematics

The vehicle velocity at the center of gravity can be expressed easily in the body xed
frame xB , yB , zB  
v cos β
vC,B =  v sin β  , (7.71)
0
where β denotes the side slip angle, and v is the magnitude of the velocity.

125
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The velocity vectors and the unit vectors in longitudinal and lateral direction of the axles
are needed for the computation of the lateral slips. One gets
     
cos δ − sin δ v cos β
ex1 ,B =  sin δ  , ey1 ,B =  cos δ  , v01,B =  v sin β + a1 γ̇  (7.72)
0 0 0

and     
1 0 v cos β
ex2 ,B = 0 , ey2 ,B = 1 , v02,B =  v sin β − a2 γ̇  , (7.73)
0 0 0
where a1 and a2 are the distances from the center of gravity to the front and rear axle,
and γ̇ denotes the yaw angular velocity of the vehicle.

7.3.3 Tire Forces

Unlike with the kinematic tire model, now small lateral motions in the contact points
are permitted. At small lateral slips, the lateral force can be approximated by a linear
approach
Fy = cS sy , (7.74)

where cS is a constant depending on the wheel load Fz , and the lateral slip sy is dened by
Eq. (3.88). Because the vehicle is neither accelerated nor decelerated, the rolling condition
is fullled at each wheel
rD Ω = eTx v0P . (7.75)

Here, rD is the dynamic tire radius, v0P the contact point velocity, and ex the unit vector
in longitudinal direction. With the lateral tire velocity

vy = eTy v0P (7.76)

and the rolling condition Eq. (7.75), the lateral slip can be calculated from

−eTy v0P
sy = T , (7.77)
| ex v0P |
with ey labeling the unit vector in the lateral direction direction of the tire. So, the lateral
forces are given by
Fy1 = cS1 sy1 ; Fy2 = cS2 sy2 . (7.78)

7.3.4 Lateral Slips

With Eq. (7.73), the lateral slip at the front axle follows from Eq. (7.77):

+ sin δ (v cos β) − cos δ (v sin β + a1 γ̇)

sy1 = . (7.79)
| cos δ (v cos β) + sin δ (v sin β + a1 γ̇) |

126
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

The lateral slip at the rear axle is given by

v sin β − a2 γ̇
sy2 = − . (7.80)
| v cos β |

The yaw velocity of the vehicle γ̇ , the side slip angle β and the steering angle δ are
considered to be small
| a1 γ̇ |  |v| ; | a2 γ̇ |  |v| (7.81)

|β |  1 and |δ|  1 . (7.82)

Because the side slip angle always labels the smaller angle between the velocity vector
and the vehicle longitudinal axis, instead of v sin β ≈ v β the approximation

v sin β ≈ |v| β (7.83)

has to be used. Now, Eqs. (7.79) and (7.80) result in

a1 v
sy1 = −β − γ̇ + δ (7.84)
|v| |v|

and
a2
sy2 = −β + γ̇ , (7.85)
|v|
where the consequences of Eqs. (7.81), (7.82), and (7.83) were already taken into consid-
eration.

7.3.5 Equations of Motion

The velocities, angular velocities, and the accelerations are needed to derive the equations
of motion, For small side slip angles β  1, Eq. (7.71) can be approximated by

 
v
vC,B =  |v| β  . (7.86)
0

The angular velocity is given by

 
0
ω0F,B =  0 . (7.87)
γ̇

If the vehicle accelerations are also expressed in the vehicle xed frame xF , yF , zF , one
will nd at constant vehicle speed v = const and with neglecting small higher-order terms
 
0
aC,B = ω0F,B × vC,B + v̇C,B =  v γ̇ + |v| β̇  . (7.88)
0

127
Vehicle Dynamics FH Regensburg, University of Applied Sciences

The angular acceleration is given by

 
0
ω̇0F,B =  0 , (7.89)
ω̇

where the substitution

γ̇ = ω (7.90)

was used. The linear momentum in the lateral direction of the vehicle reads as

m (v ω + |v| β̇) = Fy1 + Fy2 , (7.91)

where, due to the small steering angle, the term Fy1 cos δ has been approximated by Fy1 ,
and m describes the vehicle mass. With Eq. (7.90) the angular momentum yields

Θ ω̇ = a1 Fy1 − a2 Fy2 , (7.92)

where Θ names the inertia of vehicle around the vertical axis. With the linear description
of the lateral forces Eq. (7.78) and the lateral slips Eqs. (7.84), (7.85), one gets from
Eqs. (7.91) and (7.92) two coupled, but linear rst order dierential equations

a1 v a2
   
cS1 cS2 v
β̇ = −β − ω+ δ + −β + ω − ω (7.93)
m |v| |v| |v| m |v| |v| |v|
a1 v a2
   
a1 cS1 a2 cS2
ω̇ = −β − ω+ δ − −β + ω , (7.94)
Θ |v| |v| Θ |v|
which can be written in the form of a state equation
   
cS1 + cS2 a2 cS2 − a1 cS1 v   v cS1
 − −
   
  
β̇ m |v| m |v||v| |v|  β  |v| m |v|
 
=  +  δ . (7.95)
 
ω̇  ω

a21 cS1 + a22 cS2
  |{z}
 a2 cS2 − a1 cS1 | {z }  v a1 cS1  u
−
| {z }
ẋ x
Θ Θ |v| |v| Θ
| {z } | {z }
A B
If a system can be at least approximatively described by a linear state equation, stability,
steady state solutions, transient response, and optimal controlling can be calculated with
classic methods of system dynamics.

7.3.6 Stability

7.3.6.1 Eigenvalues

The homogeneous state equation

ẋ = A x (7.96)

128
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

describes the eigen-dynamics. If the approach

xh (t) = x0 eλt (7.97)

is inserted into Eq. (7.96), the homogeneous equation will remain

(λ E − A) x0 = 0 . (7.98)

One gets non-trivial solutions x0 6= 0 for

det |λ E − A| = 0 . (7.99)

The eigenvalues λ provide information concerning the stability of the system.

7.3.6.2 Low Speed Approximation

The state matrix

 
 − cS1 + cS2 a2 cS2 − a1 cS1 v 
−
Av→0

=  m |v| m |v||v| |v| 
 (7.100)
 2 2

 a1 cS1 + a2 cS2 
0 −
Θ |v|

approximates the eigen-dynamics of vehicles at low speeds, v → 0. The matrix in

Eq. (7.100) has the eigenvalues

cS1 + cS2 a21 cS1 + a22 cS2

λ1v→0 = − and λ2v→0 = − . (7.101)
m |v| Θ |v|

The eigenvalues are real and always negative independent from the driving direction.
Thus, vehicles possess an asymptotically stable driving behavior at low speed!

7.3.6.3 High Speed Approximation

At high driving velocities, v → ∞, the state matrix can be approximated by

 
v
0 −
Av→∞

= 
 |v| 
. (7.102)
a2 cS2 − a1 cS1 
0
Θ
Using Eq. (7.102) one receives from Eq. (7.99) the relation

v a2 cS2 − a1 cS1
λ2v→∞ + = 0 (7.103)
|v| Θ

129
Vehicle Dynamics FH Regensburg, University of Applied Sciences

with the solutions r

v a2 cS2 − a1 cS1
λ1,2v→∞ = ± − . (7.104)
|v| Θ
When driving forward with v > 0, the root argument will be positive, if

a2 cS2 − a1 cS1 < 0 (7.105)

holds. Then however, one eigenvalue is positive, and the system is unstable. Two zero-
eigenvalues λ1 = 0 and λ2 = 0 are obtained for

a1 cS1 = a2 cS2 . (7.106)

The driving behavior is indierent then. Slight parameter variations, however, can lead
to an unstable behavior. With

a2 cS2 − a1 cS1 > 0 or a1 cS1 < a2 cS2 (7.107)

and v > 0 the root argument in Eq. (7.104) becomes negative. Then, the eigenvalues
are imaginary, and disturbances lead to undamped vibrations. To avoid instability, high-
speed vehicles have to satisfy the condition Eq. (7.107). The root argument in Eq. (7.104)
changes at backward driving its sign. Hence, a vehicle showing stable driving behavior at
forward driving becomes unstable at fast backward driving!

7.3.7 Steady State Solution

7.3.7.1 Side Slip Angle and Yaw Velocity

At a given steering angle δ = δ0 , a stable system reaches steady state after a certain time.
With xst = const. or ẋst = 0, the state equation Eq. (7.95) is reduced to a system of linear
equations
A xst = −B u . (7.108)

With the elements from the state matrix A and the vector B, one gets from Eq. (7.108)
two equations to determine the steady state side slip angle βst and the steady state angular
velocity ωst at a constant given steering angle δ = δ0
|v| (cS1 + cS2 ) βst + (m v |v| + a1 cS1 −a2 cS2 ) ωst = v cS1 δ0 , (7.109)

|v| (a1 cS1 − a2 cS2 ) βst + (a21 cS1 + a22 cS2 ) ωst = v a1 cS1 δ0 , (7.110)

where the rst equation has been multiplied by −m |v| |v| and the second with −Θ |v|.
The solution can be derived from

v cS1 δ0 m v |v| + a1 cS1 −a2 cS2

v a1 cS1 δ0 a21 cS1 + a22 cS2
βst = (7.111)
|v| (cS1 + cS2 ) m v |v| + a1 cS1 −a2 cS2
|v| (a1 cS1 − a2 cS2 ) a21 cS1 + a22 cS2

130
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

and

|v| (cS1 + cS2 ) v cS1 δ0

|v| (a1 cS1 − a2 cS2 ) v a1 cS1 δ0
ωst = (7.112)
|v| (cS1 + cS2 ) m v |v| + a1 cS1 −a2 cS2
|v| (a1 cS1 − a2 cS2 ) a21 cS1 + a22 cS2
The denominator results in

detD = |v| cS1 cS2 (a1 + a2 )2 + m v |v| (a2 cS2 − a1 cS1 ) . (7.113)

For a non vanishing denominator detD 6= 0, steady state solutions exist

a1
a2 − m v |v|
v cS2 (a1 + a2 )
βst = δ0 , (7.114)
|v| a + a + m v |v| a2 cS2 − a1 cS1
1 2
cS1 cS2 (a1 + a2 )
v
ωst = a2 cS2 − a1 cS1 δ0 . (7.115)
a1 + a2 + m v |v|
cS1 cS2 (a1 + a2 )
At forward driving vehicles v > 0, the steady state side slip angle starts with the kinematic
value
v→0 v a2 v→0 v
βst = δ0 and ωst = δ0 (7.116)
|v| a1 + a2 a1 + a2
and decreases with increasing speed. At speeds larger than

s
a2 cS2 (a1 + a2 )
vβst=0 = (7.117)
a1 m

the side slip angle changes the sign. Using the kinematic value of the yaw velocity
Eq. (7.115) can be written as

v 1
ωst = δ0 , (7.118)
a1 + a2
v 2
 
|v|
1 +
v vch

where s
cS1 cS2 (a1 + a2 )2
vch = (7.119)
m (a2 cS2 − a1 cS1 )
is called the 'characteristic' speed of the vehicle.

In Fig. 7.14 the side slip angle β , and the driven curve radius R are plotted versus the
driving speed v . The steering angle has been set to δ0 = 1.4321◦ , in order to let the vehicle

131
Vehicle Dynamics FH Regensburg, University of Applied Sciences

steady state side slip angle radius of curvrature

2 200

0
150
-2
β [deg]

r [m]
-4 100

-6 a1*c S1/a2*c S2 = 0.66667 a1*cS1/a2*cS2 = 0.66667

50
a1*cS1/a2*c S2 = 1 a1*cS1/a2*cS2 = 1
-8
a1*cS1/a2*c S2 = 1.3333 a1*cS1/a2*cS2 = 1.3333
-10 0
0 10 20 30 40 0 10 20 30 40
v [m/s] v [m/s]
110 770 N m
m=700 kg; a1 =1.2 m;
cS1 = 80 000 N m; cS2 = 73 846 N m
Θ=1000 kg m2 ; a2 =1.3 m;
55 385 N m
Figure 7.14: Steady state cornering

drive a circle with the radius R0 = 100 m at v → 0. The actually driven circle radius R
has been calculated via
v
ωst = . (7.120)
R
Some concepts for an additional steering of the rear axle were trying to keep the side
slip angle of the vehicle, measured at the center of the vehicle to zero by an appropriate
steering or controlling. Due to numerous problems, production stage could not yet be
reached.

7.3.7.2 Steering Tendency

After reaching the steady state solution, the vehicle moves on a circle. When inserting
Eq. (7.120) into Eq. (7.115) and resolving for the steering angle, one gets

a1 + a2 v 2 v a2 cS2 − a1 cS1
δ0 = + m . (7.121)
R R |v| cS1 cS2 (a1 + a2 )
The rst term is the Ackermann steering angle, which follows from Eq. (7.2) with the
wheel base a = a1 + a2 and the approximation for small steering angles tan δ0 ≈ δ0 .
The Ackermann-steering angle provides a good approximation for slowly moving vehicles,
because the second expression in Eq. (7.121) becomes very small at v → 0. Depending on
the value of a2 cS2 − a1 cS1 and the driving direction (forward: v > 0, backward: v < 0),
the necessary steering angle diers from the Ackermann-steering angle at higher speeds.
The dierence is proportional to the lateral acceleration

v2
ay = . (7.122)
R

132
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

At v > 0 the steering tendency of a vehicle is dened by the position of the center of
gravity a1 , a2 and the cornering stinesses at the axles cS1 , cS2 . The various steering
tendencies are arranged in the table 7.1.

a1 cS1
• understeering δ0 > δ0A or a1 cS1 < a2 cS2 or <1
a2 cS2
a1 cS1
• neutral δ0 = δ0A or a1 cS1 = a2 cS2 or =1
a2 cS2
a1 cS1
• oversteering δ0 < δ0A or a1 cS1 > a2 cS2 or >1
a2 cS2
Table 7.1: Steering tendency of a vehicle at forward driving

7.3.7.3 Slip Angles

With the conditions for a steady state solution β̇st = 0, ω̇st = 0 and the relation
Eq. (7.120), the equations of motion Eq. (7.91) and Eq. (7.92) can be dissolved for the
lateral forces

a2 v2
Fy1st = m , a1 Fy2st
a1 + a2 R or = . (7.123)
a2 Fy1st
a1 v2
Fy2st = m
a1 + a2 R
With the linear tire model in Eq. (7.74) one gets

st
Fy1 = cS1 sst
y1 and
st
Fy2 = cS2 sst
y2 , (7.124)

where sst
yA1 and sst
yA2 label the steady state lateral slips at the axles. Now, from Eqs. (7.123)
and (7.124) it follows

a1 st
Fy2 cS2 sst
y2 a1 cS1 sst
y2
= st = or = st . (7.125)
a2 Fy1 cS1 sst
y1 a2 cS2 sy1
That means, at a vehicle with understeering tendency (a1 cS1 < a2 cS2 ) during steady
state cornering the slip angles at the front axle are larger than the slip angles at the rear
st st
axle, sy1 > sy2 . So, the steering tendency can also be determined from the slip angle at
the axles.

7.3.8 Influence of Wheel Load on Cornering Stiffness

With identical tires at the front and rear axle, given a linear inuence of wheel load on
the raise of the lateral force over the lateral slip,

clin
S1 = cS Fz1 and clin
S2 = cS Fz2 . (7.126)

133
Vehicle Dynamics FH Regensburg, University of Applied Sciences

holds. The weight of the vehicle G = mg is distributed over the axles according to the
position of the center of gravity

a2 a1
Fz1 = G and .Fz2 = G (7.127)
a1 + a2 a1 + a2
With Eq. (7.126) and Eq. (7.127) one obtains

a2
a1 clin
S1 = a1 cS G (7.128)
a1 + a2
and
a1
a2 clin
S2 = a2 cS G. (7.129)
a1 + a2
Thus, a vehicle with identical tires would be steering neutrally at a linear inuence of the
wheel load on the cornering stiness, because of

a1 clin lin
S1 = a2 cS2 (7.130)

The lateral force is applied behind the center of the contact patch at the caster oset
v
distance. Hence, the lever arms of the lateral forces change to a1 → a1 − n and
|v| L1
v
a2 → a2 + |v| nL1 , which will stabilize the vehicle, independently from the driving direction.

5 Fz [N ] Fy [N ]
α 0 0
4 1000 758
2000 1438
Fy [kN]

3
3000 2043

2 4000 2576
5000 3039
1 6000 3434
7000 3762
0
0 1 2 3 4 5 6 7 8 8000 4025
Fz [kN]

Figure 7.15: Lateral force Fy over wheel load Fz at dierent slip angles

At a real tire, a degressive inuence of the wheel load on the tire forces is observed,
Fig. 7.15. According to Eq. (7.92) the rotation of the vehicle is stable, if the torque from
the lateral forces Fy1 and Fy2 is aligning, i.e.

a1 Fy1 − a2 Fy2 < 0 (7.131)

holds. At a vehicle with the wheel base a = 2.45 m the axle loads Fz1 = 4000 N and Fz2 =
3000 N yield the position of the center of gravity a1 = 1.05 m and a2 = 1.40 m. At equal
slip on front and rear axle one gets from the table in 7.15 Fy1 = 2576 N and Fy2 = 2043 N .

134
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

With this, the condition in Eq. (7.131) yields 1.05 ∗ 2576 − 1.45 ∗ 2043 = −257.55 . The
value is signicantly negative and thus stabilizing.

Vehicles with a1 < a 2 have a stable, i.e. understeering driving behavior. If the axle load at
the rear axle is larger than at the front axle (a1 > a2 ), generally a stable driving behavior
can only be achieved with dierent tires.

At increasing lateral acceleration the vehicle is more and more supported by the outer
wheels. The wheel load dierences can dier at a suciently rigid vehicle body, because of
dierent kinematics (roll support) or dierent roll stiness. Due to the degressive inuence
of wheel load, the lateral force at an axle decreases with increasing wheel load dierence.
If the wheel load is split more strongly at the front axle than at the rear axle, the lateral
force potential at the front axle will decrease more than at the rear axle and the vehicle
will become more stable with an increasing lateral force, i.e. more understeering.

135
8 Driving Behavior of Single Vehicles

8.1 Standard Driving Maneuvers

8.1.1 Steady State Cornering

The steering tendency of a real vehicle is determined by the driving maneuver called
steady state cornering. The maneuver is performed quasi-static. The driver tries to keep
the vehicle on a circle with the given radius R. He slowly increases the driving speed v
2
and, with this also the lateral acceleration due ay = vR until reaching the limit. Typical
results are displayed in Fig. 8.1.

80 4

60 2
side slip angle [deg]
steer angle [deg]

40 0

20 -2

0 -4
4 6

5
3
wheel loads [kN]
roll angle [deg]

2 3

2
1
1

0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
lateral acceleration [g] lateral acceleration [g]

Figure 8.1: Steady state cornering: rear-wheel-driven car on R = 100 m

The vehicle is under-steering and thus stable according to Eq. (7.121) with Eq. (7.122).
The inclination in the diagram steering angle versus lateral velocity decides about the
steering tendency and stability behavior.

136
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

The nonlinear inuence of the wheel load on the tire performance is here used to design
a vehicle that is weakly stable, but sensitive to steer input in the lower range of lateral
acceleration, and is very stable but less sensitive to steer input in limit conditions.

With the increase of the lateral acceleration the roll angle becomes larger. The overturning
torque is intercepted by according wheel load dierences between the outer and inner
wheels. With a suciently rigid frame the use of an anti roll bar at the front axle allows
to increase the wheel load dierence there and to decrease it at the rear axle accordingly.

Thus, the digressive inuence of the wheel load on the tire properties, cornering stiness
and maximum possible lateral force, is stressed more strongly at the front axle, and the
vehicle becomes more under-steering and stable at increasing lateral acceleration, until it
drifts out of the curve over the front axle in the limit situation.

Problems occur at front driven vehicles, because due to the demand for traction, the front
axle cannot be relieved at will.

Having a suciently large test site, the steady state cornering maneuver can also be
carried out at constant speed. There, the steering wheel is slowly turned until the vehicle
reaches the limit range. That way also weakly motorized vehicles can be tested at high
lateral accelerations.

8.1.2 Step Steer Input

The dynamic response of a vehicle is often tested with a step steer input. Methods for
the calculation and evaluation of an ideal response, as used in system theory or control
technics, can not be used with a real car, for a step input at the steering wheel is not
possible in practice. A real steering angle gradient is displayed in Fig. 8.2.

40
steering angle [deg]

30

20

10

0
0 0.2 0.4 0.6 0.8 1
time [s]

Figure 8.2: Step steer input

Not the angle at the steering wheel is the decisive factor for the driving behavior, but the
steering angle at the wheels, which can dier from the steering wheel angle because of
elasticities, friction inuences, and a servo-support. At very fast steering movements, also
the dynamics of the tire forces plays an important role.

In practice, a step steer input is usually only used to judge vehicles subjectively. Exceeds
in yaw velocity, roll angle, and especially sideslip angle are felt as annoying.

137
Vehicle Dynamics FH Regensburg, University of Applied Sciences

0.6 12

0.5 10

lateral acceleration [g]

0.4 8

yaw velocity [deg/s]

0.3 6

0.2 4

0.1 2

0 0
3 1

2.5 0.5

side slip angle [deg]

2 0
roll angle [deg]

1.5 -0.5

1 -1

0.5 -1.5

0 -2
0 2 4 0 2 [t] 4

Figure 8.3: Step steer: passenger car at v = 100 km/h

The vehicle under consideration behaves dynamically very well, Fig. 8.3. Almost no over-
shoots occur in the time history of the roll angle and the lateral acceleration. However,
small overshoots can be noticed at yaw the velocity and the sideslip angle.

8.1.3 Driving Straight Ahead

8.1.3.1 Random Road Profile

The irregularities of a track are of stochastic nature. Fig. 8.4 shows a country road prole in
dierent scalings. To limit the eort of the stochastic description of a track, one usually
employs simplifying models. Instead of a fully two-dimensional description either two
parallel tracks are evaluated

z = z(x, y) → z1 = z1 (s1 ) , and z2 = z2 (s2 ) (8.1)

or one uses an isotropic track. The statistic properties are direction-independent at an

isotropic track. Then, a two-dimensional track can be approximated by a single random
process
z = z(x, y) → z = z(s) ; (8.2)

138
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

0.05
0.04
0.03
0.02
0.01
0
-0.01
-0.02
-0.03
-0.04 5
4
-0.05 3
2
0 1
10 20 30 40 50 60 70 80 90 100 0

Figure 8.4: Track irregularities

A normally distributed, stationary and ergodic random process z = z(s) is completely

characterized by the rst two expectation values, the mean value

Zs
1
mz = lim z(s) ds (8.3)
s→∞ 2s
−s

and the correlation function

Zs
1
Rzz (δ) = lim z(s) z(s − δ) ds . (8.4)
s→∞ 2s
−s

A vanishing mean value mz = 0 can always be achieved by an appropriate coordinate

transformation. The correlation function is symmetric,

Rzz (δ) = Rzz (−δ) , (8.5)

and
Zs
1
2
Rzz (0) = lim z(s) ds (8.6)
s→∞ 2s
−s
describes the variance of zs .
Stochastic track irregularities are mostly described by power spectral densities (abbrevi-
ated by psd). Correlating function and the one-sided power spectral density are linked by
the Fourier-transformation
Z∞
Rzz (δ) = Szz (Ω) cos(Ωδ) dΩ (8.7)

where Ω denotes the space circular frequency. With Eq. (8.7) follows from Eq. (8.6)

Z∞
Rzz (0) = Szz (Ω) dΩ . (8.8)

139
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Thus, the psd gives information, how the variance is compiled from the single frequency
shares.

The power spectral densities of real tracks can be approximated by the relation

 −w
Ω
Szz (Ω) = S0 , (8.9)
Ω0

where the reference frequency is xed to Ω0 = 1 m−1 . The reference psd S0 = Szz (Ω0 )
acts as a measurement for unevennes and the waviness w indicates, whether the track has
notable irregularities in the short or long wave spectrum. At real tracks, the reference-
−6
psd S0 lies within the range from 1 ∗ 10 m3 to 100 ∗ 10−6 m3 and the waviness can be
approximated by w = 2.

8.1.3.2 Steering Activity

-6 3 -5 3
highway: S 0=1*10 m ; w=2 country road: S0=2*10 m ; w=2
1000 1000

500 500

0 0
-2 0 [deg] 2 -2 0 [deg] 2

Figure 8.5: Steering activity on dierent roads

A straightforward drive upon an uneven track makes continuous steering corrections nec-
essary. The histograms of the steering angle at a driving speed of v = 90 km/h are
displayed in Fig. 8.5. The track quality is reected in the amount of steering actions. The
steering activity is often used to judge a vehicle in practice.

8.2 Coach with different Loading Conditions

8.2.1 Data

The dierence between empty and laden is sometimes very large at trucks and coaches.
In the table 8.1 all relevant data of a travel coach in fully laden and empty condition are
listed.

140
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

vehicle mass [kg] center of gravity [m] inertias [kg m2 ]

12 500 0 0
empty 12 500 −3.800 | 0.000 | 1.500 0 155 000 0
0 0 155 000
15 400 0 250
fully laden 18 000 −3.860 | 0.000 | 1.600 0 200 550 0
250 0 202 160
Table 8.1: Data for a laden and empty coach

The coach has a wheel base of a = 6.25m. The front axle with the track width sv = 2.046m
has a double wishbone single wheel suspension. The twin-tire rear axle with the track
o i
widths sh = 2.152 m and sh = 1.492 m is guided by two longitudinal links and an a-arm.
The air-springs are tted to load variations via a niveau-control.

8.2.2 Roll Steering

10
suspension travel [cm]

-5

-10
-1 0 1
steer angle [deg]

Figure 8.6: Roll steer: - - front,  rear

While the kinematics at the front axle hardly cause steering movements at roll motions,
the kinematics at the rear axle are tuned in a way to cause a notable roll steering eect,
Fig. 8.6.

8.2.3 Steady State Cornering

Fig. 8.7 shows the results of a steady state cornering on a 100 m-Radius. The fully occu-
pied vehicle is slightly more understeering than the empty one. The higher wheel loads
cause greater tire aligning torques and increase the degressive wheel load inuence on the
increase of the lateral forces. Additionally roll steering at the rear axle occurs.

141
Vehicle Dynamics FH Regensburg, University of Applied Sciences

steer angle δ [deg] vehicle course

LW
250 200

200 150

[m]
150 100

100 50

50 0
0 0.1 0.2 0.3 0.4 -100 0 100
lateral acceleration a y [g] [m]

wheel loads [kN] wheel loads [kN]

100 100

50 50

0 0
0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4
lateral acceleration a y [g] lateral acceleration a y [g]

Figure 8.7: Steady state cornering: coach - - empty,  fully occupied

Both vehicles can not be kept on the given radius in the limit range. Due to the high posi-
tion of the center of gravity the maximal lateral acceleration is limited by the overturning
hazard. At the empty vehicle, the inner front wheel lift o at a lateral acceleration of
ay ≈ 0.4 g . If the vehicle is fully occupied, this eect will occur already at ay ≈ 0.35 g .

8.2.4 Step Steer Input

The results of a step steer input at the driving speed of v = 80 km/h can be seen in
Fig. 8.8. To achieve comparable acceleration values in steady state condition, the step
◦
steer input was done at the empty vehicle with δ = 90 and at the fully occupied one
◦
with δ = 135 . The steady state roll angle is 50% larger at the fully occupied bus than
at the empty one. By the niveau-control, the air spring stiness increases with the load.
Because the damper eect remains unchanged, the fully laden vehicle is not damped as
well as the empty one. This results in larger overshoots in the time histories of the lateral
acceleration, the yaw angular velocity, and the sideslip angle.

142
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

lateral acceleration a y [g] yaw velocity ω Z [deg/s]

0.4 10

8
0.3
6
0.2
4
0.1
2

0 0
0 2 4 6 8 0 2 4 6 8

roll angle α [deg] side slip angle β [deg]

8
2
6
1

4 0

-1
2
-2
0
0 2 4 [s] 6 8 0 2 4 [s] 6 8

Figure 8.8: Step steer input: - - coach empty,  coach fully occupied

8.3 Different Rear Axle Concepts for a Passenger Car

A medium-sized passenger car is equipped in standard design with a semi-trailing rear
axle. By accordingly changed data this axle can easily be transformed into a trailing arm
or a single wishbone axis. According to the roll support, the semi-trailing axle realized
in serial production represents a compromise between the trailing arm and the single
wishbone, Fig. 8.9, .

The inuences on the driving behavior at steady state cornering on a 100 m radius are
shown in Fig. 8.10.

Substituting the semi-trailing arm at the standard car by a single wishbone, one gets,
without adaption of the other system parameters a vehicle oversteering in the limit range.
Compared to the semi-trailing arm the single wishbone causes a notably higher roll sup-
port. This increases the wheel load dierence at the rear axle, Fig. 8.10. Because the wheel
load dierence is simultaneously reduced at the front axle, the understeering tendency is
reduced. In the limit range, this even leads to an oversteering behavior.

143
Vehicle Dynamics FH Regensburg, University of Applied Sciences

10

vertical motion [cm]

5

-5

-10
-5 0 5
lateral motion [cm]

Figure 8.9: Rear axle:  semi-trailing arm, - - single wishbone, ··· trailing arm

steer angle δ [deg] roll angle α [Grad]

LW
100 5
4
3
50
2
1
0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8

wheel loads front [kN] wheel loads rear [kN]

6 6

4 4

2 2

0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
lateral acceleration a y [g] lateral acceleration a y [g]

Figure 8.10: Steady state cornering,  semi-trailing arm, - - single wishbone, · · · trailing
arm

The vehicle with a trailing arm rear axle is, compared to the serial car, more understeering.
The lack of roll support at the rear axle also causes a larger roll angle.

144
Index

Ackermann geometry, 109 Damping rate, 76

Ackermann steering angle, 109, 132 Deviation, 13
Aerodynamic forces, 96 Disturbance-reaction problem, 84
Air resistance, 96 Disturbing force lever, 9
All wheel drive, 118 Down forces, 96
Anti dive, 108 Downhill capacity, 97
Anti roll bar, 122 Drag link, 57, 58
Anti squat, 108 Drive pitch angle, 103
Anti-lock-system, 102 Driver, 3
Auto-correlation, 13 Driving force distribution, 100
Axle kinematics, 108 Driving safety, 72
Double wishbone, 7 Dynamic axle load, 95
McPherson, 7 Dynamic force elements, 63
Multi-link, 7 Dynamic wheel loads, 94
Axle load, 95
Eective value, 13
Axle suspension
Eigenvalues, 128
Solid axle, 55
Environment, 4
Twist beam, 56

First harmonic oscillation, 63

Bend angle, 113, 116
Fourier-approximation, 64
Brake pitch angle, 103
Frequency domain, 63
Brake pitch pole, 108
Friction, 97
Braking force distribution, 100
Front wheel drive, 98, 118
Camber angle, 6, 24
Generalized uid mass, 70
Camber compensation, 121, 124
Grade, 95
Camber slip, 49
Caster, 8, 9
Hydro-mount, 69
Characteristic speed, 131
Climbing capacity, 97 Kingpin, 7
Comfort, 72 Kingpin Angle, 8
Contact point, 24
Cornering resistance, 117, 118 Lateral acceleration, 121, 132

Cornering stiness, 41, 133 Lateral force, 126

Curvature gradient, 114 Lateral slip, 126

i
Vehicle Dynamics FH Regensburg, University of Applied Sciences

Ljapunov equation, 84 State vector, 88

Load, 4 Steady state cornering, 117, 136, 141
Steering activity, 140
Maximum acceleration, 97, 98
Steering angle, 114
Maximum deceleration, 97, 99
Steering box, 57, 58
Mean value, 13
Steering lever, 58
Steering oset, 9
Natural frequency, 76
Steering system
Optimal damping, 81, 87 Drag link steering system, 58
Oversteering, 133 Lever arm, 57
Overturning limit, 118 Rack and pinion, 57
Steering tendency, 125, 132
Parallel track model, 10
Step steer input, 137, 143
Parallel tracks, 138
Suspension model, 75
Pinion, 57
Suspension spring rate, 78
Pivot pole, 109
System response, 63
Power spectral density, 139

Tilting condition, 97
Quarter car model, 87, 90
Tire

Rack, 57 Bore slip, 52

Random road prole, 138 Bore torque, 20, 50

Rear wheel drive, 98, 118 Camber angle, 24

Reference frames Camber inuence, 48

Ground xed, 5 Characteristics, 52

Inertial, 5 Circumferential direction, 24

Vehicle xed, 5 Composites, 19

Relative damping rate, 77 Contact forces, 20

Ride comfort, 83 Contact patch, 20

Ride safety, 83 Contact point, 23

Road, 10, 23 Contact point velocity, 31

Roll axis, 124 Contact torques, 20

Roll center, 124 Cornering stiness, 41

Roll steer, 141 Deection, 26

Roll stiness, 120 Deformation velocity, 31

Roll support, 121, 124 Development, 19

Rolling condition, 126 Dynamic oset, 41

Dynamic radius, 32, 33
Safety, 72 Friction coecient, 45
Side slip angle, 109 Lateral direction, 24
Sky hook damper, 87 Lateral force, 20
Space requirement, 110 Lateral force characteristics, 41
Spring rate, 78 Lateral force distribution, 40
Stability, 128 Lateral slip, 40
State equation, 128 Lateral velocity, 31
State matrix, 88

ii
FH Regensburg, University of Applied Sciences © Prof. Dr.-Ing. G. Rill

Lift o, 88 Vehicle model, 75, 90, 94, 103, 112, 121,
Linear model, 126 125
Loaded radius, 24, 32 Vertical dynamics, 72
Longitudinal force, 20, 38, 39 Virtual work, 121
Longitudinal force characteristics, 39
Waviness, 140
Longitudinal force distribution, 39
Wheel
Longitudinal slip, 39
Angular velocity, 50
Longitudinal velocity, 31
Wheel base, 109
Model, 52
Wheel load, 20
Normal force, 20
Wheel loads, 94
Pneumatic trail, 41
Wheel rotation axis, 5
Radial damping, 35
Wheel Suspension
Radial direction, 24
Semi-trailing arm, 143
Radial stiness, 34, 121
Single wishbone, 143
Rolling resistance, 20, 36, 37
Trailing arm, 143
Self aligning torque, 20, 41
Wheel suspension
Sliding velocity, 40
Central control arm, 56
Static radius, 24, 32, 33
Double wishbone, 55
Tilting torque, 20
McPherson, 55
Track normal, 24, 26
Multi-Link, 55
Transport velocity, 33
Semi-trailing arm, 56
Tread deection, 38
SLA, 56
Tread particles, 37
Unloaded radius, 32
Yaw angle, 112, 115
Vertical force, 34
Yaw velocity, 126
Wheel load inuence, 41
Tire Model
Kinematic, 109
Linear, 133
TMeasy, 52
Toe angle, 6
Toe-in angle, 6
Track, 23
Track curvature, 114
Track normal, 5
Track radius, 114
Track width, 109, 121
Trailer, 112, 115

Understeering, 133

Variance, 13
Vehicle, 3
Vehicle comfort, 72
Vehicle dynamics, 2

iii