VEHICLE DYNAMICS

FACHHOCHSCHULE REGENSBURG UNIVERSITY OF APPLIED SCIENCES HOCHSCHULE FUR TECHNIK WIRTSCHAFT SOZIALES

LECTURE NOTES Prof. Dr. Georg Rill NOVEMBER, 2002

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

Contents

Contents 1 Introduction 1.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 1.1.2 1.1.3 1.1.4 1.1.5 1.2 1.2.1 1.2.2 1.2.3 1.3 1.3.1 1.3.2 1.3.3 1.3.4 1.3.5 1.4 1.4.1 1.4.2 1.4.3 1.4.4 Vehicle Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multi Purpose Suspension Systems . . . . . . . . . . . . . . . . . . . Specic Suspension Systems . . . . . . . . . . . . . . . . . . . . . . . Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I 1 1 1 2 2 3 3 4 4 4 5 5 5 6 6 7 7 8 8 9 9

Wheel/Axle Suspension Systems . . . . . . . . . . . . . . . . . . . . . . . . .

Steering Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rack and Pinion Steering . . . . . . . . . . . . . . . . . . . . . . . . Lever Arm Steering System . . . . . . . . . . . . . . . . . . . . . . . Drag Link Steering System . . . . . . . . . . . . . . . . . . . . . . . Bus Steer System . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coordinate Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . Forces and Torques in the Tire Contact Area . . . . . . . . . . . . . . Dynamic Rolling Radius . . . . . . . . . . . . . . . . . . . . . . . . . 1.4.4.1 1.4.4.2 Denitions according to DIN 70 000

Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Toe and Camber Angle . . . . . . . . . . . . . . . . . . . . . . . . . 11 . . . . . . . . . . . . . 11 Calculation . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.4.5

Steering Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.4.5.1 1.4.5.2 1.4.5.3 Kingpin . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Caster and Kingpin Angle . . . . . . . . . . . . . . . . . . . 13 Caster and Kingpin Oset . . . . . . . . . . . . . . . . . . . 14 15

2 Tire 2.1 2.1.1 2.1.2 2.1.3 2.2 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7 2.2.8 2.2.9

Contact Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Contact Point . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Local Track Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Contact Point Velocity . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Wheel Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Longitudinal Force and Longitudinal Slip . . . . . . . . . . . . . . . . 19 Lateral Slip, Lateral Force and Self Aligning Torque . . . . . . . . . . 22 Generalized Tire Characteristics . . . . . . . . . . . . . . . . . . . . . 24 Wheel Load Inuence . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Self Aligning Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Camber Inuence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Bore Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Typical Tire Characteristics . . . . . . . . . . . . . . . . . . . . . . . 31 33

Tire Forces and Torques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3 Longitudinal Dynamics 3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.1.6 3.1.7 3.1.8 3.2 3.2.1 3.2.2

Accelerating and Braking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Simple Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Maximum Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . 34 Drive Torque at Single Axle . . . . . . . . . . . . . . . . . . . . . . . 34 Braking at Single Axle . . . . . . . . . . . . . . . . . . . . . . . . . . 35 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Optimal Distribution of Drive and Brake Forces . . . . . . . . . . . . . 37 Dierent Distributions of Brake Forces . . . . . . . . . . . . . . . . . 38 Anti-Lock-Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Plane Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Drive and Brake Pitch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

II

3.2.3 3.2.4 3.2.5 3.2.6 3.2.7

Linearization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 Equations of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Driving and Braking . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 Brake Pitch Pole . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 47

4 Lateral Dynamics 4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.2 4.2.1 4.2.2 4.2.3

Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Overturning Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Roll Support and Camber Compensation . . . . . . . . . . . . . . . . 50 Roll Center and Roll Axis . . . . . . . . . . . . . . . . . . . . . . . . 51 Roll Angle and Wheel Loads . . . . . . . . . . . . . . . . . . . . . . . 51 Kinematic Tire Model . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Ackermann Geometry . . . . . . . . . . . . . . . . . . . . . . . . . . 53 Vehicle Model with Trailer . . . . . . . . . . . . . . . . . . . . . . . . 54 4.2.3.1 4.2.3.2 4.2.3.3 4.2.3.4 4.2.3.5 Position . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Vehicle Movements . . . . . . . . . . . . . . . . . . . . . . 56 Entering a Curve . . . . . . . . . . . . . . . . . . . . . . . 57 Trailer Movements . . . . . . . . . . . . . . . . . . . . . . . 58 Course Calculations . . . . . . . . . . . . . . . . . . . . . . 59

Kinematic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.3

Simple Handling Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Lateral Slips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 Equations of Motion . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4.3.5.1 4.3.5.2 4.3.5.3 4.3.6 4.3.6.1 4.3.6.2 4.3.6.3 Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . 63 Low Speed Approximation . . . . . . . . . . . . . . . . . . . 64 High Speed Approximation . . . . . . . . . . . . . . . . . . 64 Side Slip Angle and Yaw Velocity . . . . . . . . . . . . . . . 65 Steering Tendency . . . . . . . . . . . . . . . . . . . . . . . 67 Slip Angles . . . . . . . . . . . . . . . . . . . . . . . . . . . 67

Steady State Solution . . . . . . . . . . . . . . . . . . . . . . . . . . 65

III

4.3.7

Inuence of Wheel Load on Cornering Stiness . . . . . . . . . . . . . 68 4.3.7.1 4.3.7.2 4.3.7.3 Linear Wheel Load Inuence . . . . . . . . . . . . . . . . . 68 Digressive Wheel Load Inuence . . . . . . . . . . . . . . . 69 Steering Tendency depending on Lateral Acceleration . . . . 70 71

5 Vertical Dynamics 5.1 5.2

Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Basic Tuning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 5.2.1 5.2.2 5.2.3 5.2.4 5.2.5 Simple Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Track . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Spring Preload . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 Free Vibrations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 Random Road Prole . . . . . . . . . . . . . . . . . . . . . . . . . . 77 Vehicle Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Quality Criteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Optimal Parameter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 5.3.4.1 5.3.4.2 5.3.4.3 Linear Characteristics . . . . . . . . . . . . . . . . . . . . . 79 Nonlinear Characteristics . . . . . . . . . . . . . . . . . . . 80 Limited Spring Travel . . . . . . . . . . . . . . . . . . . . . 81

5.3

Nonlinear Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 5.3.1 5.3.2 5.3.3 5.3.4

5.4

Dynamic Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 5.4.1 System Response in the Frequency Domain . . . . . . . . . . . . . . . 83 5.4.1.1 5.4.1.2 5.4.2 5.4.2.1 5.4.2.2 First Harmonic Oscillation . . . . . . . . . . . . . . . . . . . 83 Sweep-Sine Excitation . . . . . . . . . . . . . . . . . . . . . 84 Principle and Model . . . . . . . . . . . . . . . . . . . . . . 85 Dynamic Force Characteristics . . . . . . . . . . . . . . . . 87

Hydro-Mount . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

5.5

Dierent Inuences on Comfort and Safety . . . . . . . . . . . . . . . . . . . 88 5.5.1 5.5.2 Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Simulation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89

IV

6 Driving Behavior of Single Vehicles 6.1 6.1.1 6.1.2 6.1.3

91

Standard Driving Maneuvers . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 91 Step Steer Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Driving Straight Ahead . . . . . . . . . . . . . . . . . . . . . . . . . 93 6.1.3.1 6.1.3.2 Random Road Prole . . . . . . . . . . . . . . . . . . . . . 93 Steering Activity . . . . . . . . . . . . . . . . . . . . . . . . 95

6.2

Coach with dierent Loading Conditions . . . . . . . . . . . . . . . . . . . . 96 6.2.1 6.2.2 6.2.3 6.2.4 Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Roll Steer Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 97 Step Steer Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

6.3

Dierent Rear Axle Concepts for a Passenger Car . . . . . . . . . . . . . . . . 98

1 Introduction
1.1
1.1.1

Terminology
Vehicle Dynamics

The Expression Vehicle Dynamics encompasses the interaction of driver, vehicle load and environment Vehicle dynamics mainly deals with the improvement of active safety and driving comfort as well as the reduction of road destruction. In vehicle dynamics computer calculations test rig measurements and eld tests are employed. The interactions between the single systems and the problems with computer calculations and/or measurements shall be discussed in the following.

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

1.1.2

Driver

By various means of interference the driver can interfere with the vehicle: steering wheel lateral dynamics gas pedal brake pedal driver vehicle longitudinal dynamics clutch gear shift The vehicle provides the driver with some information: longitudinal, lateral, vertical vibrations: motor, aerodynamics, tyres vehicle sound: driver instruments: velocity, external temperature, ... The environment also inuences the driver: climate environment trac density driver track A drivers 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. In all tests, the drivers security must have absolute priority. Driving simulators provide an excellent means of analyzing the behavior of drivers even in limit situations without danger. For some years it has been tried to analyze the interaction between driver and vehicle with complex driver models.

1.1.3

Vehicle

The following vehicles are listed in the ISO 3833 directive: Motorcycles, Passenger Cars, Busses, Trucks

FH Regensburg, University of Applied Sciences Agricultural Tractors, Passenger Cars with Trailer Truck Trailer / Semitrailer, Road Trains.

Prof. Dr.-Ing. G. Rill

For computer calculations these vehicles have to be depicted in mathematically describable substitute systems. The generation of the equations of motions and 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.

1.1.4

Load

Trucks are conceived for taking up load. Thus their driving behavior changes. Load mass, inertia, center of gravity dynamic behaviour (liquid load)

In computer calculations problems occur with the determination of the inertias and the modelling of liquid loads. Even the loading and unloading process of experimental vehicles takes some eort. When making experiments with tank trucks, ammable liquids have to be substituted with water. The results thus achieved cannot be simply transferred to real loads.

1.1.5

Environment

The Environment inuences primarily the vehicle: Environment Road: bumps, coecient of friction Air: resistance, cross wind vehicle

but also inuences the driver Environment climate visibility driver

Through the interactions between vehicle and road, roads can quickly be destroyed. The greatest problem in eld test and laboratory experiments is the virtual impossibility of reproducing environmental inuences. The main problems in computer simulation are the description of random road bumps and the interaction of tires and road as well as the calculation of aerodynamic forces and torques.

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

1.2
1.2.1

Wheel/Axle Suspension Systems

General Remarks

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

1.2.2

Multi Purpose Suspension Systems

The Double Wishbone Suspension, the McPherson Suspension and the Multi-Link Suspension are multi purpose wheel suspension systems, Fig. 1.1.
E

E zR 2 yR D R N1 N3 G O2 xR Q F U1 U2 yB O1 1 P zB xB M S

G zR F yR R A B S Q U

O P xR D C yB zB xB
yR

zR

G xR Y W P U A

D Z R S B V X Q

Figure 1.1: Double Wishbone, McPherson and Multi-Link Suspension They are used as steered front or non steered rear axle suspension systems. These suspension 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. At pickups, trucks and busses often rigid axles are used. The rigid axles are guided either by
zA
zA Z2 Y2 Z1 Y1

X2 xA X1 yA

xA yA

Figure 1.2: Rigid Axles

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

leaf springs or by rigid links, Fig. 1.2. Rigid axles tend to tramp on rough road. Leaf spring guided rigid 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 rigid axle suspension systems additional links in longitudinal and lateral direction are used. Rigid axles suspended by air springs need at least four links for guidance. In addition to a good drive comfort air springs allow level control.

1.2.3

Specic Suspension Systems

The Semi-Trailing Arm, the SLA and the Twist Beam axle suspension are suitable only for non steered axles, Fig. 1.3.
zR yR

zA yA xA

xR

Figure 1.3: 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 SLA 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.

1.3
1.3.1

Steering Systems
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.

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

In order to achieve a good manuvrability a maximum steer angle of approx. 30 must be provided at the front wheels of passenger cars. Depending on the wheel base busses and trucks need maximum steer angles up to 55 at the front wheels. Recently some companies have started investigations on steer by wire techniques.

1.3.2

Rack and Pinion Steering

Rack and pinion is the most common steering system on passenger cars, Fig. 1.4. The rack may be located either in front of or behind the axle. The rotations of the steering wheel L are
uZ L rack pinion steer box

wheel and wheel body

nk drag li

1 L

Figure 1.4: Rack and Pinion Steering rstly transformed by the steering box to the rack travel uZ = uZ (L ) and then via the drag links transmitted to the wheel rotations 1 = 1 (uZ ), 2 = 2 (uZ ). Hence the overall steering ratio depends on the ratio of the steer box and on the kinematics of the steer linkage.

1.3.3

Lever Arm Steering System

Using a lever arm steering system Fig. 1.5, large steer angles at the wheels are possible. This

G
ste e r le ver 1

steer box

Q1 L

r stee
P2

leve

r2
Q2

drag link 1 P1

drag link 2
2

wheel and wheel body

Figure 1.5: Lever Arm Steering System

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

steering system is used on trucks with large wheel bases and independent wheel suspension at the front axle. Here the steering box can be placed outside of the axle center. The rotations of the steering wheel L are rstly transformed by the steering box to the rotation of the steer levers G = G (L ). The drag links transmit this rotation to the wheel 1 = 1 (G ), 2 = 2 (G ). Hence, again the overall steering ratio depends on the ratio of the steer box and on the kinematics of the steer linkage.

1.3.4

Drag Link Steering System

At rigid axles the drag link steering system is used, Fig. 1.6.
ver steer le
H O steer box

H wheel and wheel body

(90o rotated)

steer linkage

I L 1 K drag link 2

Figure 1.6: Drag Link Steering System The rotations of the steering wheel L are transformed by the steering box to the rotation of the steer lever arm H = H (L ) and further on to the rotation of the left wheel, 1 = 1 (H ). The drag link transmits the rotation of the left wheel to the right wheel, 2 = 2 (1 ).

1.3.5

Bus Steer System

In busses the driver sits more than 2 m in front of the front axle. Here, sophisticated steer systems are needed, Fig. 1.7. The rotations of the steering wheel L are transformed by the steering box to the rotation of the steer lever arm H = H (L ). Via the steer link the left lever arm is moved, H = H (G ). This motion is transferred by a coupling link to the right lever arm. Via the drag links the left and right wheel are rotated, 1 = 1 (H ) and 2 = 2 (H ).

Vehicle Dynamics
steer lever

FH Regensburg, University of Applied Sciences

H
steer box

steer link

I Q L
wheel and wheel body

left lever arm J

drag link

H P

K
coupl. link

Figure 1.7: Bus Steer System

1.4
1.4.1

Denitions
Coordinate Systems

In vehicle dynamics several dierent coordinate systems are used, Fig 1.8.

z0 x0 zF xF yF y0

eU

eN

e yR eS

Figure 1.8: Coordinate Systems The inertial system with the axes x0 , y0 , z0 is xed to the track. Within the vehicle xed system the xF -axis is pointed forward, the yF -axis left and the zF -axis upward. The position of the wheel is given by the unit vector eyR in direction of the wheel rotation axis.

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The unit vectors in the directions of circumferential and lateral forces eU and eS as well as the track normal eN follow from the contact geometry.

1.4.2

Forces and Torques in the Tire Contact Area

In any point of contact between tire and track normal and friction forces are delivered. According to the tires prole design the contact area forms a not necessarily coherent area. The eect of the contact forces can be fully described by a vector of force and torque in reference to a point in the contact patch. The vectors are described in a track-xed coordinate system. The z -axis is normal to the track, the x-axis is perpendicular to the z -axis and perpendicular to the wheel turning axis eyR . The demand for a right-handed coordinate system then also xes the y -axis.

Fx Fy Fz Mx My Mz

longitudinal or circumferential force lateral force vertical force or wheel load

tilting torque rolling resistance torque self aligning and bore torque

Fy

Mx Fx Fz

Mz

My

Figure 1.9: Contact Forces and Torques The components of the contact force are named according to the direction of the axes, Fig. 1.9. Non symmetric distributions of force in the contact patch cause torques around the x and y axes. The tilting torque Mx occurs when the tire is cambered. My also contains the rolling resistance of the tire. In particular the torque around the z -axis is relevant in vehicle dynamics. It consists of two parts, Mz = M B + M S . (1.1) Rotation of the tire around the z -axis causes the bore torque MB . The self aligning torque MS respects the fact that in general the resulting lateral force is not applied in the contact point.

1.4.3

Dynamic Rolling Radius

At an angular rotation of , assuming the tread particles stick to the track, the deected tire moves on a distance of x, Fig. 1.10. With r0 as unloaded and rS = r0 r as loaded or static tire radius r0 sin = x (1.2)

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

deflected tire

rigid wheel

r0 r S

rD

vt

x
Figure 1.10: Dynamic Rolling Radius

and r0 cos hold. If the movement of a tire is compared to the rolling of a rigid wheel, its radius rD then has to be chosen so, that at an angular rotation of the tire moves the distance x = rD From (1.2) and (1.4) one gets r0 sin , where the trivial solution rD = r0 follows from at 0. rD = (1.5) . (1.4) = rS . (1.3)

At small, yet nite angular rotations the sine-function can be approximated by the rst terms of its Taylor-Expansion. Then, (1.5) reads as rD = r0
1 6

= r0

1 2 6

(1.6)

With the according approximation for the cosine-function rS = cos r0 = 1 1 2 2 or 2 = 2 1 rS r0 . (1.7)

follows from (1.3). Inserted into (1.6) r D = r0 remains. 1 1 3 1 rS r0 = 2 1 r0 + rS 3 3 (1.8)

10

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The radius rD depends on the wheel load Fz because of rS = rS (Fz ) and thus is named dynamic tire radius. With the rst approximation (1.8) it can be calculated from the undeformed radius r0 and the steady state radius rS . At a rotation with the angular velocity , the tread particles are transported through the contact area with the average velocity v t = rD (1.9)

1.4.4
1.4.4.1

Toe and Camber Angle

Denitions according to DIN 70 000

The angle between the vehicle center plane in longitudinal direction and the intersection line of the tire center plane with the track plane is named toe angle. It is positive, if the front part of the wheel is oriented towards the vehicle center plane. The camber angle is the angle between the wheel center plane and the track normal. It is positive, if the upper part of the wheel is inclined outwards. 1.4.4.2 Calculation

The calculation can be done via the unit vector eyR in the direction of the wheel turning axis. For the calculation of the toe angle the unit vector eyR is described in the vehicle xed coordinate system F , Fig. 1.11 eyR,F = eyR,F eyR,F eyR,F
(1) (2) (3) T

(1.10)

where the axes xF and zF span the vehicle center plane. The xF -axis points forward and the zF -axis points upward. The toe angle V can then be calculated from tan V = eyR,F eyR,F
(2) (1)

(1.11)

The camber angle follows from the scalar product between the unit vectors in the direction of the wheel turning axis and in the direction of the track normal sin = eT yR en . (1.12)

11

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

eyR zF yF e (2) yR,F V e (3) yR,F e (1) yR,F xF

Figure 1.11: Toe Angle

1.4.5
1.4.5.1

Steering Geometry
Kingpin

At the steered front axle the McPherson-damper strut axis, the double wishbone axis and multilink wheel suspension or dissolved double wishbone axis are frequently employed in passenger cars, Fig. 1.12 and Fig. 1.13.

zR B M A xR

yR

kingpin axis A-B

Figure 1.12: Double Wishbone Wheel Suspension The wheel body rotates around the kingpin at steering movements. At the double wishbone axis, the ball joints A and B , which determine the kingpin, are xed to the wheel body. The ball joint point A is also xed to the wheel body at the classic McPherson wheel suspension, but the point B is xed to the vehicle body.

12

FH Regensburg, University of Applied Sciences

B zR zR

Prof. Dr.-Ing. G. Rill

yR xR M A

yR M xR

kingpin axis A-B

rotation axis

Figure 1.13: McPherson and Multi-Link Wheel Suspensions At a multi-link axle, the kingpin is no longer dened by real link points. Here, as well as with the McPherson wheel suspension, the kingpin changes its position against the wheel body at wheel travel. 1.4.5.2 Caster and Kingpin Angle

The current direction of the kingpin can be dened by two angles within the vehicle xed coordinate system, Fig. 1.14. If the kingpin is projected into the yF -, zF -plane, the kingpin inclination angle can be read as the angle between the zF -axis and the projection of the kingpin. The projection of the kingpin into the xF -, zF -plane delivers the caster angle with the angle between the zF -axis and the projection of the kingpin. With many axles the kingpin and caster angle can no longer be determined directly. The current rotation axis at steering movements, that can be taken from kinematic calculations here delivers the kingpin. The current values of the caster angle and the kingpin inclination angle can be calculated from the components of the unit vector in the direction of the kingpin, described in the vehicle xed coordinate system tan = with eS,F = eS,F eS,F eS,F
(1) (2) (3)

eS,F eS,F
(3)

(1)

and

tan =

eS,F eS,F
(3)

(2)

(1.13)

(1.14)

13

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

zF

zF

eS

yF xF

Figure 1.14: Kingpin and Caster Angle 1.4.5.3 Caster and Kingpin Offset

In general, the point S where the kingpin runs through the track plane does not coincide with the contact point P , Fig. 1.15.

ey

ex S nK

rS

Figure 1.15: Caster and Kingpin Oset If the kingpin penetrates the track plane before the contact point, the kinematic kingpin oset is positive, nK > 0. The caster oset is positive, rS > 0, if the contact point P lies outwards of S .

14

2 Tire
2.1
2.1.1

Contact Geometry
Contact Point

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 turning axis, Fig. 2.1.

tire rim centre plane

ezR M e yR ex P0 ey M rS
b a

e yR en P0 P*

en

x0

z0 0

road: z = z ( x , y )

y0

local road plane

Figure 2.1: Contact Geometry The irregularities of the track are described by an arbitrary function of two spatial coordinates z = z (x, y ). (2.1)

At an uneven track the contact point P can not be calculated directly. One can rstly get an estimated value with the vector rM P = r0 ezB , (2.2) where r0 is the undeformed tire radius and ezB is the unity vector in the z direction of the body xed reference system.

15

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

The position of P with respect to the xed system x0 , y0 , z0 is determined by r 0P = r 0 M + r M P , (2.3)

where the vector r0M states the position of the rim center M . Usually the point P lies not on the track. The corresponding track point P0 follows from r0P ,0 (1) r0P0 ,0 = (2.4) r0P ,0 (2) . z (r0P ,0 (1), r0P ,0 (2)) In the point P0 now the track normal is calculated. Then the unit vectors in the tires circumferential direction and lateral direction can be calculated ex = eyR en , | eyR en | and ey = en ex . (2.5)

Calculating ex demands a normalization, for the unit vector in the direction of the wheel turning axis eyR is not always perpendicular to the track. The tire camber angle = arcsin eT yR en describes the inclination of the wheel turning axis against the track normal. The vector from the rim center M to the track point P0 is now split into three parts rM P0 = rS ezR + a ex + b ey , (2.7) (2.6)

where rS names the loaded or static tire radius and a, b are displacements in circumferential and lateral direction. The unit vector ezR = ex eyR . | ex eyR | (2.8)

is perpendicular to ex and eyR . Because the unit vectors ex and ey are perpendicular to en , the scalar multiplication of (2.7) with en results in
T eT n rM P0 = rS en ezR

or rS =

eT n rM P0 . eT n ezR

(2.9)

Now also the tire deection can be calculated r = r 0 rS , with r0 marking the undeected tire radius. The point P given by the vector rM P = rS ezR (2.11) (2.10)

16

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

lies within the rim center plane. The transition from P 0 to P takes place according to (2.7) by terms a ex and b ey , standing perpendicular to the track normal. The track normal however was calculated in the point P 0 . Therefore with an uneven track P no longer lies on the track. With the newly estimated value P = P now the equations (2.4) to (2.11) can be recurred 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 even. At an ordinary passenger-car tire, the contact patch has at normal load about the size of approximately 20 20 cm. There is obviously little sense in calculating a ctitious contact point to fractions of millimeters, when later the real track is approximated in the range of centimeters by a plane. If the track in the contact patch is replaced by a plane, no further iterative improvement is necessary at the hereby used initial value.

2.1.2

Local Track Plane

A plane is given by three points. With the tire width b, the undeformed tire radius r0 and the length of the contact area LN at given wheel load, estimated values for three track points can be given in analogy to (2.3) rM L = rM R = rM F =
LN 2 b 2 b 2

eyR r0 ezB , eyR r0 ezB , r0 ezB . (2.12)

exB

The points lie left, resp. right and to the front of a point below the rim center. The unit vectors exB and ezB point in the longitudinal and vertical direction of the vehicle. The wheel turning axis is given by eyR . According to (2.4) the corresponding points on the track L, R and F can be calculated. The vectors rRF = r0F r0R and rRL = r0L r0R rRF rRL . | rRF rRL | (2.13) lie within the track plane. The unit vector calculated by en = (2.14)

is perpendicular to the plane dened by the points L, R, and F and gives an average track normal over the contact area. Discontinuities which occur at step- or ramp-sized obstacles are smoothed that way. Of course it would be obvious to replace LN in (2.12) by the actual length L of the contact area and the unit vector ezB by the unit vector ezR which points upwards in the wheel center plane. The values however, can only be calculated from the current track normal. Here also an iterative solution would be possible. Despite higher computing eort the model quality cannot be improved by this, because approximations in the contact calculation and in the tire model limit the exactness of the tire model.

17

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

2.1.3

Contact Point Velocity

The absolute velocity of the contact point one gets from the derivation of the position vector v0P,0 = r 0P,0 = r 0M,0 + r M P,0 . (2.15)

Here r 0M,0 = v0M,0 is the absolute velocity of the wheel center and rM P,0 the vector from the wheel center M to the contact point P , expressed in the inertial frame 0. With (2.11) one gets d r M P,0 = (rS ezR,0 ) = r S ezR,0 rS e zR,0 . (2.16) dt Due to r0 = const. r S = r (2.17) follows from (2.10). The unit vector ezR is xed to the wheel body. Its time derivative is then given by e zR,0 = 0RK,0 ezR,0 (2.18)

where 0RK is the angular velocity of the wheel body RK relative to the inertial frame 0. With rM P,0 = rS ezR,0 and the relations (2.17) and (2.18), (2.16) reads as r M P,0 = r ezR,0 + 0RK,0 rM P,0 . (2.19)

The contact point velocity is then given by v0P,0 = v0M,0 + r ezR,0 + 0RK,0 rM P,0 (2.20)

where the velocity components from the wheel rotation have not yet been taken into account. Because the point P lies on the track, v0P,0 must not contain a component normal to the track eT (2.21) n v 0P = 0 . The tire deformation velocity is dened by this demand r = eT n (v0M + 0RK rM P ) . eT n ezR (2.22)

Then, one gets for the velocity components in longitudinal and lateral direction
T vx = eT x v0P = ex (v0M + 0RK rM P )

(2.23)

and
T vy = eT y v0P = ey (v0M +

r ezR + 0RK rM P ) ,

(2.24)

where the term which can be cancelled in v0P by the orthogonality relation ezR ex has already been omitted in (2.23).

18

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

2.2
2.2.1

Tire Forces and Torques

Wheel Load
z= (2.25)

The vertical tire force Fz can be calculated as a function of the normal tire deection eT r and the deection velocity z = eT r n n Fz = Fz ( z, z ) .

Because the tire can only deliver pressure forces to the road, the restriction Fz 0 holds. In a rst approximation Fz is separated into a static and a dynamic part Fz = FzS + FzD . The static part is described as a nonlinear function of the normal tire deection FzS = a1 z + a2 ( z )2 . (2.27) (2.26)

The constants a1 and a2 are calculated from the radial stiness at nominal payload cN z = and the radial stiness at double payload
N c2 = z

d FzS d z

(2.28)
S =F N Fz z

d FzS d z

.
S =2F N Fz z

(2.29)

The dynamic part is approximated by FzD = dR z , (2.30)

where dR is a constant describing the radial tire damping.

2.2.2

Longitudinal Force and Longitudinal Slip

To get some insight into the mechanism generating tire forces in longitudinal direction we consider a tire on a at test rig. The rim is rotating with the angular speed and the at track runs with speed vx . The distance between the rim center an the at track is controlled to the loaded tire radius corresponding to the wheel load Fz , Fig. 2.2. A tread particle enters at time t = 0 the contact area. If we assume adhesion between the particle and the track then the top of the particle runs with the track speed vx and the bottom with the average transport velocity vt = rD . Depending on the speed dierence v = rD v the tread particle is deected in longitudinal direction u = (rD vx ) t . (2.31)

19

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

rD vx

rD vx L

u max

Figure 2.2: Tire on Flat Track Test Rig The time a particle spends in the contact area can be calculated by T = L , rD || (2.32)

where L denotes the contact length, and T > 0 is assured by ||. The maximum deection occurs when the tread particle leaves at t = T the contact area umax = (rD vx ) T = (rD vx ) L . rD || (2.33)

The deected tread particle applies a force to the tire. In a rst approximation we get
t Fx = ct xu,

(2.34)

where ct x is the stiness of one tread particle in longitudinal direction. On normal wheel loads more than one tread particle is in contact with the track, Fig. 2.3a. The number p of the tread particles follows from p = L . s+a (2.35)

where s is the length of one particle and a denotes the distance between the particles. Particles entering the contact area are undeected on exit the have the maximum deection. According to (2.34) this results in a linear force distribution versus the contact length, Fig. 2.3b. For p particles the resulting force in longitudinal direction is given by Fx = With (2.35) and (2.33) this results in Fx = 1 L t L cx (rD vx ) . 2 s+a rD || (2.37) 1 t p c umax . 2 x (2.36)

20

FH Regensburg, University of Applied Sciences

a) L b) L

Prof. Dr.-Ing. G. Rill

c) r0

t cx * u

t cu * u max

L/2

Figure 2.3: a) Particles, b) Force Distribution, c) Tire Deformation A rst approximation of the contact length L is given by
2 (L/2)2 = r0 (r0

r)2 ,

where r0 is the undeected tire radius, and r denotes the tire deection, Fig. 2.3c. With r r0 one gets L2 8 r 0 r . (2.39) The tire deection can be approximated by r = Fz /cR . (2.40)

where Fz is the wheel load, and cR denotes the radial tire stiness. Now, (2.36) can be written as r0 c t rD v x x Fx = 4 Fz . (2.41) s + a cR rD || The non-dimensional relation between the sliding velocity of the tread particles in longitudinal S direction vx = vx rD and the average transport velocity rD || is the longitudinal slip sx = (vx rD ) . rD || (2.42)

In this rst approximation the longitudinal force Fx is proportional to the wheel load Fz and the longitudinal slip sx Fx = k Fz sx , (2.43) where the constant k collects the tire properties r0 , s, a, ct x and cR . The relation (2.43) holds only as long as all particles stick to the track. At average slip values the particles at the end of the contact area start sliding, and at high slip values only the parts at the beginning of the contact area stick to the road, Fig. . 2.4. The resulting nonlinear function of the longitudinal force Fx versus the longitudinal slip sx can 0 be dened by the parameters initial inclination (longitudinal stiness) dFx , location sM x and M G magnitude of the maximum Fx , start of full sliding sG and the sliding force F , Fig. 2.5. x x

(2.38)

21

Vehicle Dynamics
small slip values Fx = k * Fz* s x L Fx < = FH adhesion
t t

FH Regensburg, University of Applied Sciences

moderate slip values Fx = Fz * f ( s x ) L Fx = FH adhesion sliding
t t t

large slip values Fx = FG L Fx = FG sliding

Figure 2.4: Longitudinal Force Distribution for dierent Slip Values

Fx
M Fx G Fx

adhesion
0 dFx

sliding

sM x

sG x

sx

Figure 2.5: Typical Longitudinal Force Characteristics

2.2.3

Lateral Slip, Lateral Force and Self Aligning Torque

Similar to the longitudinal slip sx , given by Eq. (2.42), the lateral slip can be dened by
G vy sy = , rD ||

(2.44)

where the sliding velocity in lateral direction is given by

G vy = vy

(2.45)

and the lateral component of the contact point velocity vy follows from Eq. (2.24). As long as the tread particles stick to the road (small amounts of slip), an almost linear distribution of the forces along the contact area length L appears. At moderate slip values the particles at the end of the contact area start sliding, and at high slip values only the parts at the beginning of the contact area stick to the road, Fig. 2.6. The distribution of the lateral forces over the contact area length also denes the acting point of the resulting lateral force. At small slip values the working point lies behind the center of the contact area (contact point P). With rising slip values, it moves forward, sometimes even before the center of the contact area. At extreme slip values, when practically all particles are sliding, the resulting force is applied at the center of the contact area.

22

FH Regensburg, University of Applied Sciences

small slip values Fy = k * Fz * s y adhesion moderate slip values Fy = Fz * f ( s y ) adhesion

Prof. Dr.-Ing. G. Rill

large slip values Fy = FG

Fy

sliding

Fy L
full sliding

Fy n

Figure 2.6: Lateral Force Distribution over Contact Area The resulting lateral force Fy with the dynamic tire oset or pneumatic trail n as a lever generates the self aligning torque MS = n Fy . (2.46) 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 Fig. 2.7. Characteristic parameters for the lateral

sliding

n/L (n/L)0

adhesion adhesion/sliding

Fy Fy
M G adhesion

adhesion/ sliding

full sliding

s0 y
MS
adhesion

sG y

sy

Fy dF0 y

adhesion/sliding full sliding

sM y

sG y

sy

s0 y

sG y

sy

Figure 2.7: Typical Plot of Lateral Force, Tire Oset and Self Aligning Torque
0 force graph are initial inclination (cornering stiness) dFy , location sM y and magnitude of the M G G 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 area L. The initial G value (n/L)0 as well as the slip values s0 y and sy characterize the graph suciently.

23

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

2.2.4

Generalized 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 parameters 0 0 M M M initial inclination dFx , dFy , location sM x , sy and magnitude of the maximum Fx , Fy as G G G well as sliding limit sG x , sy and sliding force Fx , Fy , Fig. 2.8.
Fx

Fx

Fx

0 dF x

Fy sx sM x sG x Fy
dF 0 FM FG

Fy Fy

F(s)
Fx

sy sG y sM y

dF y

sG sy sM s

sx

Figure 2.8: Generalized Tire Characteristics 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 physical signicance, the parameters can subsequently be improved by means of comparisons between the simulation and vehicle testing results as far as they are available. During general driving situations, e.g. acceleration or deceleration in curves, longitudinal sx and lateral slip sy appear simultaneously. The combination of the more or less diering longitudinal and lateral tire forces requires a normalization process. One way to perform the normalization is described in the following. The longitudinal slip sx and the lateral slip sy can vectorially be added to a generalized slip s = sx s x
2

sy s y

(2.47)

where the slips sx and sy were normalized in order to perform their similar weighting in s. For normalizing, the normation factors s x and s y are calculated from the location of the maxima M M M 0 0 sM , s the maximum values F , F and the initial inclinations dFx , dFx . x y x y

24

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

Similar to the graphs of the longitudinal and lateral forces the graph of the generalized tire force is dened by the characteristic parameters dF 0 , sM , F M , sG and F G . The parameters are calculated from the corresponding values of the longitudinal and lateral force
2 0s 0 sin (dFx x cos ) + dFy s y 2

dF 0 = sM = FM = sG = FG =

sM x cos s x

sM y sin s y
2

, (2.48)
2

M cos )2 + F M sin (Fx y

, ,

sG x cos s x

sG y sin s y
2

G cos )2 + F G sin (Fx y

where the slip normalization have also to be considered at the initial inclination. The angular functions sx /s sy /s x y cos = and sin = (2.49) s s grant a smooth transition from the characteristic curve of longitudinal to the curve of lateral forces in the range of = 0 to = 90 . The function F = F (s) is now described in intervals by a broken rational function, a cubic polynomial and a constant F G s , = M , 0 s sM ; sM dF 0 M s s 1 + + F0 M 2 F F (s) = (2.50) s sM M M G 2 M G F ( F F ) (3 2 ) , = , s < s s ; sG sM FG , s > sG . When dening the curve parameters, one just has to make sure that the condition dF 0 2 F sM is fullled, because otherwise the function has a turning point in the interval 0 < s sM .
M

Longitudinal and lateral force now follow from the according projections in longitudinal and lateral direction Fx = F cos and Fy = F sin . (2.51)

25

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

2.2.5

Wheel Load Inuence

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 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 digressive inuence of the wheel load on the characteristic curves of longitudinal and lateral forces. Longitudinal Force Fx Fz = 3.2 kN
0 dFx = 90 kN

Lateral Force Fy Fz = 3.2 kN

0 dFy = 70 kN

Fz = 6.4 kN
0 dFx = 160 kN

Fz = 6.4 kN
0 dFy = 100 kN

sM x = 0.090 sG x = 0.400

sM x = 0.110 sG x = 0.500

sM y = 0.180 sG y = 0.600

sM y = 0.200 sG y = 0.800

M M M M Fx = 3.30 kN Fx = 6.50 kN Fy = 3.10 kN Fy = 5.40 kN

G G Fx = 3.20 kN Fx = 6.00 kN

G G Fy = 3.10 kN Fy = 5.30 kN

Table 2.1: Characteristic Tire Data with Digressive Wheel Load Inuence

In order to respect this fact in a tire model, the characteristic data for two nominal wheel loads FzN and 2 FzN are given in Tab. 2.1.
0 0 M M From this data the initial inclinations dFx , dFy , the maximal forces Fx , Fx and the sliding G M forces Fx , Fy for arbitrary wheel loads Fz are calculated by quadratic functions. For the maximum longitudinal force it reads as M Fx (Fz ) =

Fz Fz M M (FzN ) 1 F M (2FzN ) N . 2 Fx (FzN ) 1 F M (2FzN ) Fx 2 x 2 x N Fz Fz

(2.52)

M G G The location of the maxima sM x , sy , and the slip values, sx , sy , at which full sliding appears, are dened as linear functions of the wheel load Fz . For the location of the maximum longitudinal force this results in M N M N M N sM x (Fz ) = sx (Fz ) + sx (2Fz ) sx (Fz )

Fz 1 FzN

(2.53)

With the numeric values from Tab. 2.1 a slight shift of the maxima towards higher slip values is also modelled, Fig. 2.9.

26

FH Regensburg, University of Applied Sciences

Fx = Fx (s ): Parameter F
x z

Prof. Dr.-Ing. G. Rill

Fy = Fy (s y): Parameter Fz

8000 6000 4000 2000 0 -2000 -4000 -6000 -8000 -0.5 0 0.5

8000 6000 4000 2000

Fz

Fz

0 -2000 -4000 -6000 -8000 -0.5 0 0.5

Figure 2.9: Wheel Load Inuence to Tire Forces

Fx = Fx(sx): Parameter sy 4000 3000 2000 1000 0 1000 2000 3000 4000 0.5 4000 3000 Fy = Fy(sy): Parameter sx

sy

2000

s
1000 0 1000 2000 3000 4000 0.5

0.5

0.5

sy = 0.0, 0.0375, 0.075, 0.1125, 0.15

sx = 0.0, 0.0375, 0.075, 0.1125, 0.15

Figure 2.10: Tire Forces vs. Longitudinal and Lateral Slip: Fz = 3.2 kN The bilateral inuence of longitudinal sx and lateral slip sy on the longitudinal Fx and lateral force Fy is depicted in Fig. 2.10. With the 20 parameters, which are, according to Tab. 2.1, necessary for the denition of the characteristic curves of longitudinal and lateral force, the tire model can be easily tted to measured characteristics. Because for description of the characteristic curves of longitudinal and lateral force only characteristic curve parameters are used, a desired tire behavior can also be constructed in a convenient manner.

2.2.6

Self Aligning Torque

According to Eq. (2.46) the self aligning torque can be calculated via the dynamic tire oset.

27

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

The approximation as a function of the lateral slip is done by a line and a cubic polynomial (n/L)0 (1 |sy |/s0 |sy | s0 Q) Q 2 0 E |sy | sQ sQ |sy | n E (2.54) = (n/L)0 s0 Q < |sy | sQ 0 E 0 L s s s Q Q Q 0 |sy | > sE Q The cubic polynomial reaches the sliding limit sE Q with a horizontal tangent and is continued with the value zero. 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 wheel load are given. The calculation of the parameters of arbitrary wheel loads is done similar to Eq. (2.53) by linear inter- or extrapolation.
Mz = Mz (s y ): Parameter F z

150

Tire Oset Parameters Fz = 3.2 kN Fz = 6.4 kN

100

Fz
50 0 -50 -100 -150 -0.5 0.5

(n/L)0 = 0.150 (n/L)0 = 0.130 s0 y = 0.200 sE y = 0.500 s0 y = 0.230 sE y = 0.450

Figure 2.11: Self Aligning Torque: Fz = 0, 2, 4, 6, 8 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. Fig. 2.6. The working point of the resulting force (dynamic tire oset) is then given by n(Fz 0, sy = 0) = 1 L. 6 (2.55)

The value n = 1 L can only serve as reference point, for the uneven distribution of pressure in 6 longitudinal direction of the contact area results in a change of the deexion prole and the dynamic tire oset. The self aligning torque in Fig. 2.11 has been calculated with the tire parameters from Tab. 2.1, the tire stiness cR = 180 kN/m and the undeected tire radius r0 = 0.293 m. The digressive inuence of the wheel load on the lateral force can be seen here as well.

28

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

With the parameters for the description of the tire oset it has been assumed that at double payload Fz = 2 FzN the related tire oset reaches the value of (n/L)0 = 0.13 at sy = 0. Because for Fz = 0 the value 1/6 0.17 can be assumed, a linear interpolation provides the value (n/L)0 = 0.15 for Fz = FzN . The slip value s0 y , 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. With rising wheel load it moves to higher values. The values for sE y are estimated too.

2.2.7

Camber Inuence

If the wheel rotation axis is inclined against the road a lateral force appears, dependent on the inclination angle. At a non-vanishing camber angle, = 0 the tread particles possess a lateral velocity when entering the contact area, which is dependent on wheel rotation speed and the camber angle , Fig. 2.12. At the center of the contact area (contact point) this

rim centre plane

4000

Fy = Fy (s y ): Parameter

3000 2000

eyR en ex
velocity deflection profile

1000 0 -1000 -2000 -3000 -4000 -0.5

ey

0.5

Figure 2.12: Cambered Tire Fy ( ) at Fz = 3.2 kN and = 0 , 2 , 4 , 6 , 8 component vanishes and at the end of the contact area it is of the same value but opposing the component at the beginning of the contact area. At normal friction and even distribution of pressure in the longitudinal direction of the contact area one gets a parabolic deection prole, which is equal to the average deection L sin 1 y = L (2.56) 2 R || 6 s s denes a camber-dependent lateral slip. A solely lateral tire movement without camber results in a linear deexion prole with the average deexion vy 1 L. (2.57) y vy = R || 2 sy

29

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

1 a comparison of Eq. (2.56) to Eq. (2.57) shows, that with s y = 3 s the lateral camber slip s can be converted to the equivalent lateral slip s y. In normal driving operation, the camber angle and thus the lateral camber slip are limited to small values. So the lateral camber force can be calculated over the initial inclination of the characteristic curve of lateral forces 0 Fy dFy sy .

(2.58)

0 If the global inclination dFy Fy /sy is used instead of the initial inclination dFy , one gets the camber inuence on the lateral force as shown in Fig. 2.12. The camber angle inuences the distribution of pressure in the lateral direction of the contact area, and changes the shape of the contact area from rectangular to trapezoidal. It is thus extremely dicult if not impossible to quantify the camber inuence with the aid of simple models. Therefore a plain approximation has been used, which still describes the camber inuence rather exactly.

2.2.8

Bore Torque
n = eT n 0R = 0 . (2.59)

If the wheel rotation 0R has a component in direction of the track normal en a very complicated deection prole of the tread particles in the contact area occurs. By a simple approach the resulting bore torque can be calculated by the parameter of the longitudinal force characteristics. Fig. 2.13 shows the contact area at zero camber ( = 0) and small slip values (sx 0, sy 0). The contact area is separated into small stripes of width dy . The longitudinal slip in a stripe at position y is then given by (n y ) sx (y ) = . (2.60) rD || For small slip values the nonlinear tire force characteristics can be linearized. The longitudinal force in the stripe can then be approximated by d Fx d s x Fx (y ) = y. (2.61) d sx d y With (2.60) one gets d Fx n Fx (y ) = y. (2.62) d sx rD || The forces Fx (y ) generate a bore torque in the contact point P
+B 2 +B 2

MB =

1 B
B 2

y Fx (y ) dy =

1 B
B 2

d Fx n y dy d sx rD ||

(2.63)

1 2 d Fx n B 12 d sx rD ||

1 d Fx B n B , 12 d sx rD | |

30

FH Regensburg, University of Applied Sciences

dy x UG L Q n P contact area B P y n dy x

Prof. Dr.-Ing. G. Rill

U(y) y

contact area -UG L

Figure 2.13: Bore Torque generated by Longitudinal Forces where sB =

n ||

(2.64)

can be considered as bore slip. Via dFx /dsx the bore torque takes into account the actual friction and slip conditions. The bore torque calculated by (2.63) is only a rst approximation. At large bore slips the longitudinal forces in the stripes are limited by the sliding values. Hence, the bore torque is limited by
+B 2 max MB = 2

1 B
0

G y Fx dy =

1 G B Fx , 4

(2.65)

G where Fx denotes the longitudinal sliding force.

2.2.9

Typical Tire Characteristics

The tire model TMeasy1 approximates the characteristic curves Fx = Fx (sx ), Fy = Fy () and Mz = Mz () quite well even for dierent wheel loads Fz , Fig. 2.14. TMeasy is able to handle the dierent tire types in a suitable manner. The soft truck tire of type Radial 315/80 R22.5 at p=8.5 bar (right column) and the large dierences between longitudinal and lateral force characteristics at the passenger car tire of type Radial 205/50 R15, 6J at p=2.0 bar (left column) are represented without any considerable tting problems. The one-dimensional characteristics are automatically converted to a two-dimensional combination characteristics which are shown in Fig. 2.15.

Hirschberg, W; Rill, G. Weinfurter, H.: User-Appropriate Tyre-Modelling for Vehicle Dynamics in Standard and Limit Situations. Vehicle System Dynamics 2002, Vol. 38, No. 2, pp. 103-125. Lisse: Swets & Zeitlinger.

31

Vehicle Dynamics
6 4 F [kN]

FH Regensburg, University of Applied Sciences

40 20 F [kN] 0 10 kN 20 kN 30 kN 40 kN 50 kN -20 0 s [%] x 20 40

2 0 1.8 kN 3.2 kN 4.6 kN 5.4 kN -20 0 s [%] x 20 40

-2 -4 -6 -40 6 4 Fy [kN]

-20 -40 -40

40 20 F [kN] 0 10 kN 20 kN 30 kN 40 kN

2 0 1.8 kN 3.2 kN 4.6 kN 6.0 kN

-2 -4 -6 150 100 50 Mz [Nm]

1500 1000 500 M [Nm] 0 -500 18.4 kN 36.8 kN 55.2 kN -20 -10
z

0 -50 1.8 kN 3.2 kN 4.6 kN 6.0 kN -20 -10

-100 -150 0 10

-1000 -1500 0

-20 -40

[o]

20

[o]

10

20

Figure 2.14: Tire Characteristics at Dierent Wheel Loads: Meas., TMeasy

30 3 2 F [kN] F [kN] -2 0 F [kN]
x

20 10
y

1 0

-1 -2 -3 -4 2 4

-10 -20 -30 -20 0 F [kN]

x

20

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

|| = 1, 2, 4, 6, 10, 14

Figure 2.15: Two-dimensional Tire Characteristics at Fz = 3.2 kN / Fz = 35 kN

32

3 Longitudinal Dynamics
3.1
3.1.1

Accelerating and Braking

Simple Model
v

Fx2 Fz2
a2

mg a1

Fx1 Fz1

Figure 3.1: Simple Model The forces in the wheel contact points are combined into one vertical and one circumferential force per axle. Aerodynamic forces (drag, positive and negative lift) are neglected. The road runs horizontally and be ideally even. Then no vertical acceleration and no pitch acceleration around the lateral vehicle axle occur: 0 = Fz1 + Fz2 m g and 0 = Fz1 a1 Fz2 a2 + (Fx1 + Fx2 ) h . The linear momentum in longitudinal direction results in mv = Fx1 + Fx2 , (3.3) (3.2) (3.1)

where v indicates the vehicles acceleration. This are only three equations for the four unknown forces Fx1 , Fx2 , Fz1 , Fz2 .

33

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

If we insert (3.3) in (3.2) we can eliminate two unknowns by one stroke 0 = Fz1 a1 Fz2 a2 + m v h. The equations (3.1) and (3.4) can now be resolved for the axle loads Fz1 = m g and Fz2 = m g h v a2 a1 + a2 a1 + a2 g a1 h v + a1 + a2 a1 + a2 g . (3.5) (3.4)

(3.6)

The weight mg is distributed among the axles according to position of the center of gravity. When accelerating v > 0, the front axle is relieved, as is the rear axle when decelerating v < 0.

3.1.2

Maximum Acceleration

Ordinary vehicles can only deliver pressure forces to the road. According to equation (3.6), the conditions Fz1 0 and Fz2 0 lead to the tilting conditions a1 v a2 . h g h (3.7)

The maximum acceleration is also limited by the friction conditions |Fx1 | Fz1 and |Fx2 | Fz2 (3.8)

where the same friction coecient has been assumed at front and rear axle. In the limit case |Fx1 | = Fz1 and |Fx2 | = Fz2 (3.9) the maximally achievable acceleration resp. deceleration follows from (3.3) and (3.1) |v max | = g . (3.10)

According to the vehicle dimensions and the friction values the maximal acceleration or deceleration is restricted either by (3.7) or by (3.10).

3.1.3

Drive Torque at Single Axle

With the rear axle driven in limit situations Fx1 = 0 and Fx2 = Fz2 . holds. With this, one gets from (3.3) mv 0+ = Fz2 , (3.12) (3.11)

34

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

where the subscript 0+ indicates that the front axle is neither driven nor braked and the rear axle is driven. Using 3.6 one gets mv 0+ = m g a1 h v 0+ + a1 + a2 a1 + a2 g . (3.13)

Hence, the maximum acceleration for a rear wheel driven vehicle is given by v 0+ = g a1 . h a1 + a2 1 a1 + a2 (3.14)

With front wheel drive one gets with Fx1 = Fz1 the maximum acceleration v +0 = g h 1+ a1 + a2 a2 , a1 + a2 (3.16) and Fx2 = 0 (3.15)

where the subscript +0 indicates now a driven front axle and a rear axle which is neither driven nor braked.

3.1.4

Braking at Single Axle

With an unbraked rear axle in the limit case it holds Fx1 = Fz1 With that one gets from (3.3) mv 0 = Fz1 , (3.18) where the subscript 0 indicates a braked front axle and a rear axle which is neither driven nor braked. With 3.5 one gets mv 0 = m g or v 0 = g a2 h v 0 a1 + a2 a1 + a2 g h 1 a1 + a2 a2 . a1 + a2 (3.19) and Fx2 = 0 . (3.17)

(3.20)

If only the rear axle is braked, using Fx1 = 0 and Fx2 = Fz2 , (3.21)

35

Vehicle Dynamics one gets now the maximal deceleration v 0 = g where the subscript rear axle.

FH Regensburg, University of Applied Sciences

h 1+ a1 + a2

a1 , a1 + a2

(3.22)

indicates a front axle which is neither driven nor braked and a braked

3.1.5

Example

Typical values for a passenger car are: a1 = 1.25 m; a2 = 1.25 m; h = 0.55 m With a friction coecient of = 1 the maximal accelerations calculated from (3.14) and (3.16) result in v 0+ 1 1.25 = = 0.64 0.55 g 1.25 + 1.25 11 1.25 + 1.25 and 1 1.25 v +0 = 0.41 = 0.55 1.25 + 1.25 g 1+1 1.25 + 1.25 The maximal decelerations follow from (3.20) and (3.22) v 0 = g and 1 11 0.55 1.25 + 1.25 1 1+1 1.25 = 0.64 1.25 + 1.25

0.55 1.25 + 1.25 Because a load distribution of 50/50 between the axles was assumed, the maximal accelerations an decelerations have the same absolute value. If only the front axle is braked, the maximal deceleration still is about 2/3 of the maximally possible of v /g = = 1. Braking only the rear axle is often not sucient, because here only about 40% of the maximally possible deceleration can be achieved. In vehicles with front drive the maximal acceleration is augmented by shifting the center of gravity to the front. On the other hand the maximal acceleration can also be augmented by shifting the center of gravity to the rear.

v 0 = g

1.25 = 0.41 1.25 + 1.25

36

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

3.1.6

Optimal Distribution of Drive and Brake Forces

The sum of the circumferential forces accelerates or decelerates the vehicle. In dimensionless style (3.3) reads v Fx1 Fx2 = + . (3.23) g mg mg A certain acceleration or deceleration can only be achieved by dierent combinations of the circumferential forces Fx1 and Fx2 . According to (3.9) the circumferential forces are limited by wheel load and friction. The optimal combination of Fx1 and Fx2 is achieved, when front and rear axle have the same skid resistance. Fx1 = Fz1 and Fx2 = Fz2 . (3.24) With (3.5) and (3.6) one gets Fx1 = mg and Fx2 = mg a2 v h g a1 v + h g h a1 + a2 h . a1 + a2 (3.25)

(3.26)

With (3.25) and (3.26) one gets from (3.23) v = , g where it has been assumed that Fx1 and Fx2 have the same sign. With (3.27 inserted in (3.25) and (3.26) one gets Fx1 v = mg g and Fx2 v = mg g a2 v h g a1 v + h g h a1 + a2 h . a1 + a2 (3.28) (3.27)

(3.29)

remain. Depending on the desired acceleration v > 0 or deceleration v < 0 the circumferential forces that grant the same skid resistance at both axles can now be calculated. Fig.3.2 shows the curve of optimal drive and brake forces for typical passenger car values. At the tilting limits v/g = a1 /h and v/g = +a2 /h no circumferential forces can be delivered at the lifting axle. The initial gradient only depends on the steady state distribution of wheel loads. From (3.28) and (3.29) it follows Fx1 d a2 v h mg = 2 (3.30) v h g a1 + a2 d g

37

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

braking
Fx1/mg
dFx2 dFx1 0

-a1/h

-1

-2

a =1.15

driving

a =1.35
1

1 2

h=0.55 =1.20

tilting limits
2

a2/h

Figure 3.2: Optimal Distribution of Drive and Brake Forces and d

Fx2 mg = v d g

Fx2/mg

a1 v +2 h g

h . a1 + a2

(3.31)

For v/g = 0 the initial gradient remains as d F x2 d F x1 =

0

a1 . a2

(3.32)

3.1.7

Different Distributions of Brake Forces

In practice it is tried to approximate the optimal distribution of brake forces by constant distribution, limitation or reduction of brake forces as good as possible. Fig. 3.3. When braking, the vehicles stability is dependent on the potential of lateral force (cornering stiness) at the rear axle. In practice, a greater skid (locking) resistance is thus realized at the rear axle than at the front axle. Because of this, the brake force balances in the physically relevant area are all below the optimal curve. This restricts the achievable deceleration, specially at low friction values.

38

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

Fx1/mg Fx2/mg

Fx1/mg Fx2/mg

Fx1/mg

limitation

Fx2/mg

constant distribution

reduction

Figure 3.3: Dierent Distributions of Brake Forces Because the optimal curve is dependent on the vehicles center of gravity additional safeties have to be installed when designing real distributions of brake forces. Often the distribution of brake forces is tted to the axle loads. There the inuence of the height of the center of gravity, which may also vary much on trucks, remains unrespected and has to be compensated by a safety distance from the optimal curve. Only the control of brake pressure in anti-lock-systems provides an optimal distribution of brake forces independent from loading conditions.

3.1.8

Anti-Lock-Systems

Lateral forces can only be scarcely transmitted, if high values of longitudinal slip occur when decelerating a vehicle. Stability and/or steerability is then no longer given. By controlling the brake torque, respectively brake pressure, the longitudinal slip can be restricted to values that allow considerable lateral forces. is used here as control variable. Angular wheel accelerations The angular wheel acceleration are derived from the measured angular wheel speeds by dierentiation. With a longitudinal slip of sL = 0 the rolling condition is fullled. Then = x rD (3.33)

holds, where rD labels the dynamic tyre radius and x is the vehicles acceleration. According to (3.10), the maximum acceleration/deceleration of a vehicle is dependent on the friction coecient, |x | = g . With a known friction coecient a simple control law can be realized for every wheel | 1 |x | | (3.34) rD . Because until today no reliable possibility to determine the local friction coecient between tyre and road has been found, useful information can only be gained from (3.34) 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 then used for the calculation of slip at all wheels. This method is too imprecise at low speeds. Below a limit velocity no control occurs therefore. Problems also occur when for example all wheels lock simultaneously which may happen on icy roads.

39

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

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 -split braking at the rear axle. The break pressure at both wheels is controlled the wheel running on lower friction. Thus the brake forces at the rear axle cause no yaw torque. The maximally achievable deceleration however is reduced by this.

3.2
3.2.1

Drive and Brake Pitch

Plane Vehicle Model

zA MA1 z1 MA1
R1

FF1

xA A MB1 hR MA2 z2 MA2 MB2 FF2

MB1 Fz1 a1 Fx1 R a2 R2 Fz2

MB2 Fx2

Figure 3.4: Plane Vehicle Model The vehicle model drawn in Fig. 3.4 consists of ve rigid bodies. The body has three degrees of freedom: Longitudinal motion xA , vertical motion zA and pitch A . 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 ) , 1 = 1 (z1 ) ; x2 = x2 (z2 ) , 2 = 2 (z2 ) . (3.35)

The rotation angles R1 and R2 describe the wheel rotations relative to the wheel bodies.

40

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The forces between wheel body and vehicle body are labelled FF 1 and FF 2 . At the wheels torques of drive MA1 , MA2 and brake MB 1 , MB 2 , circumferential forces Fx1 , Fx2 and the wheel loads Fz1 , Fz2 apply. The brake torques are supported directly by the wheel bodies, the drive torques are supported directly by the vehicle via the drive train. The forces and torques that apply to the single bodies are listed in the last column of the tables 3.1 and 3.2.

3.2.2

Position

Position vector and rotation matrix xA ; 0 r0A,0 = R + hR + zA

A0 A

cos A 0 sin A 0 1 0 = sin A 0 cos A

(3.36)

describe the position of the bodys center of gravity relative to the earth xed coordinate system 0. With R = const. and hR = const one can immediately get from this the velocity and angular velocity of the body. 0 x A 0 A . v0A,0 = 0 + 0 ; 0A,0 = (3.37) 0 z A 0 The position of the wheel bodies is given by r0RK1 ,0 = r0A,0 + A0A rARK1 ,A and A0RK1 = A0A AARK1 as well as r0RK2 ,0 = r0A,0 + A0A rARK2 ,A and A0RK2 = A0A AARK2 with AARK2 cos 2 0 sin 2 0 1 0 = sin 2 0 cos 2 (3.41) with rARK2 ,A a2 + x2 0 = hR + z2 (3.40) with AARK1 cos 1 0 sin 1 0 1 0 = sin 1 0 cos 1 (3.39) with rARK1 ,A a1 + x 1 0 = hR + z1 (3.38)

41

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

With well balanced wheels the position vectors rARK1 ,A and rARK2 ,A also tell the position of the wheel pivots. If toe and camber angle are neglected, the following holds for the wheels rotation matrices: cos R1 0 sin R1 , (3.42) 0 1 0 A0R1 = A0A AARK1 ARKR1 with ARRK1 = sin R1 0 cos R1 cos R2 0 sin R2 . (3.43) 0 1 0 A0R2 = A0A AARK2 ARKR2 with ARRK2 = sin R2 0 cos R2

3.2.3

Linearization

At small rotational motions of the body and small spring motions one gets for the speed of the wheel bodies and wheels x1 A z x A 0 hR 1 z1 + 0 ; 0 v0RK1 ,0 = v0R1 ,0 = 0 + 0 + (3.44) z A 0 z 1 a1 A A x A 0 hR + 0 = 0 + 0 + A 0 z A +a2
x2 z2

z 2

. (3.45)

v0RK2 ,0 = v0R2 ,0

0 z 2

The angular velocities of the wheel bodies and wheels are given by 0 0 0 0 0 A + 1 and 0R1 ,0 = A + 1 + R1 0RK1 ,0 = 0 0 0 0 0 (3.46) as well as 0 0 0 0 0 A + 2 and 0R2 ,0 = A + 2 + R2 0RK2 ,0 = 0 0 0 0 0 (3.47) With the generalized velocities z = A 1 2 x A z A R1 R2
T

(3.48)

the velocities and angular velocities (3.37), (3.44), (3.45), (3.46), (3.47) can be written as
7

v 0i =
j =1

v0i zj zj

and 0i =
j =1

0i zj zj

(3.49)

42

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

3.2.4

Equations of Motion

0i 0i The partial velocities v and partial angular velocities for the ve bodies i = 1(1)5 and zj zj for the 7 generalized speeds j = 1(1)7 are arranged in the tables 3.1 and 3.2. With the aid

bodies chassis mA wheel body front mRK 1 wheel front mR 1 wheel body rear mRK 2 wheel rear mR 2

x A 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0

z A 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1

partial velocities v0i /zj A z 1 R1 z 2 0 0 0 0 0 0 0 0 0 0 0 0 x1 hR 0 0 z1 0 0 0 0 a1 0 0 1 x1 hR 0 0 z1 0 0 0 0 a1 0 0 1 x2 hR 0 0 z2 0 0 0 0 a2 0 0 1 x2 hR 0 0 z2 0 0 0 0 a2 0 0 1

R2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

applied forces Fie 0 0 FF 1 + FF 2 mA g 0 0 FF 1 mRK 1 g Fx1 0 Fz1 mR1 g 0 0 FF 2 mRK 2 g Fx2 0 Fz2 mR2 g

Table 3.1: Partial Velocities and Applied Forces

bodies chassis A wheel body front RK 1 wheel front R1 wheel body rear RK 2 wheel rear R2

partial angular A x A z A 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0

velocities 0i /zj applied torques z 1 R1 z 2 R2 Mie 0 0 0 0 0 0 0 0 0 MA1MA2 a1 FF 1 + a2 FF 2 0 0 0 0 0 0 0 0 0 0 1 0 0 0 M B1 z1 0 0 0 0 0 0 0 0 0 0 1 1 0 0 M M A1 B 1 R Fx1 z1 0 0 0 0 0 0 0 0 0 0 2 0 0 0 M B2 z2 0 0 0 0 0 0 0 0 0 0 2 0 0 z2 1 MA2 MB 2 R Fx2 0 0 0 0 0

Table 3.2: Partial Angular Velocities and Applied Torques

43

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

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

M (i, j ) =
k=1

v0k zi
5

mk

v0k + zj
T e Fk

k=1 5

0k zi

k
T

0k ; zj
e Mk ;

i, j = 1(1)7 ;

(3.50)

Q(i) =
k=1

v0k zi

+
k=1

0k zi

i = 1(1)7 .

(3.51)

The equations of motion for the plane vehicle model are then given by Mz = Q. (3.52)

3.2.5

Equilibrium

With the abbreviations m1 = mRK 1 + mR1 ; and h = hR + R The components of the vector of generalized forces and torques read as Q(1) = Fx1 + Fx2 ; Q(2) = Fz1 + Fz2 mG g ; Q(3) = a1 Fz1 + a2 Fz2 h(Fx1 + Fx2 ) + a1 m1 g a2 m2 g ; Q(4) = Fz1 FF 1 +
x1 z1

m2 = mRK 2 + mR2 ;

mG = mA + m1 + m2

(3.53) (3.54)

(3.55)

Fx1 m1 g +

1 (MA1 z1

R Fx1 ) ;

(3.56)

Q(5) = MA1 MB 1 R Fx1 ; Q(6) = Fz2 FF 2 +

x2 z2

Fx2 m2 g +

2 (MA2 z2

R Fx2 ) ;

(3.57)

Q(7) = MA2 MB2 R Fx2 . Without drive and brake forces MA1 = 0 ; MA2 = 0 ; MB 1 = 0 ; MB 2 = 0 (3.58)

from (3.55), (3.56) and (3.57) one gets the steady state circumferential forces, the spring preloads and wheel loads
0 Fx 1 = 0; 0 FF 1 = b a+b 0 Fx 2 = 0;

mA g ;
b a+b

0 FF 2 =

a a+b

mA g ;
a a+b

(3.59) mA g .

Fz01

= m1 g +

mA g ;

Fz02

= m2 g +

44

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

3.2.6

Driving and Braking

Assuming that on accelerating or decelerating the vehicle x A = 0 the wheels neither slip nor lock, A + x1 z R R1 = x A hR ; z1 1 (3.60) x2 R R2 = x A hR A + z 2 .
z2

holds. In steady state the pitch motion of the body and the vertical motion of the wheels reach constant values
st A = A = const. ; st z1 = z1 = const. ; st z2 = z2 = const.

(3.61)

and (3.60) simplies to R R1 = x A ; R R2 = x A . (3.62) With(3.61), (3.62) and (3.54) the equation of motion (3.52) results in
a a mG x A = Fx 1 + Fx2 ;

0 = Fza1 + Fza2 ; hR (m1 + m2 ) x A + R1

x1 z1 x A R

+ R2

x A R

a a = a Fza1 + b Fza2 (hR + R)(Fx 1 + Fx2 ) ; (3.63) x1 z1 a Fx 1+ 1 (MA1 z1 a R Fx 1) ;

m1 x A +

1 z1

R1 R1

x A R x A R x A R x A R

a = Fza1 FF 1+

a = MA1 MB 1 R Fx 1; a = Fza2 FF 2+ x2 z2 a Fx 2+ 2 (MA2 z2 a R Fx 2) ;

(3.64)

x2 z2

m2 x A +

2 z2

R2 R2

a = MA2 MB2 R Fx 2;

(3.65)

where the steady state spring forces, circumferential forces and wheel loads have been separated into initial and acceleration-dependent terms
st 0 a Fxi = Fxi + Fxi ; st 0 a Fzi = Fzi + Fzi ; st 0 a FF i = FF i + FF i ;

i = 1, 2 .

(3.66)

a a With given torques of drive and brake the vehicle acceleration x A , the wheel forces Fx 1 , Fx2 , a a Fza1 , Fza2 and the spring forces FF 1 , FF 2 can be calculated from (3.63), (3.64) and (3.65)

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
a FF 1 a FF 2 Fza1 Fza2 a = cA1 z1 , a = cA2 z2 , a a a = cR1 (zA a A + z1 ), a a a = cR2 (zA + b A + z2 ) .

(3.67)

besides the vertical motions of the wheels.

45

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

a Especially the pitch of the vehicle A = 0, caused by drive or brake is, if too distinct, felt as annoying.

The drive and brake pitch and/or a rotation of the wheel bodies on jounce and rebound i = 1, 2 can be reduced by a suitable axle kinematics.

i , zi

3.2.7

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. 3.5.

pitch pole

x-, z- motion of the contact points during compression and rebound

Figure 3.5: Brake Pitch Pole Increasing the pitch pole height above the track level means a decrease in the brake pitch angle.

46

4 Lateral Dynamics
4.1
4.1.1

Steady State Cornering

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. 4.1.

2 1 m ay h2 mg

h1

F yL FzL s/2

F yR s/2

FzR

Figure 4.1: Overturning Hazard on Trucks The balance of torques at the already inclined vehicle delivers for small angles 1 (FzL FzR ) s = m ay (h1 + h2 ) + m g [(h1 + h2 )1 + h2 2 ] , 2 1, 2 1

(4.1)

47

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

where ay indicates the lateral acceleration and m is the sprung mass. On a left-hand tilt, the right tire raises
K FzR = 0

(4.2)

and the left tire carries all the vehicle weight

K FzL = mg .

(4.3)

Using (4.2) and (4.3) one gets from (4.1) s aK h2 y K 2 = 1 K . h1 + h2 h1 + h2 2 g The vehicle turns over, when the lateral acceleration ay rises above the limit aK y
K K Roll of axle and body reduce the overturning limit. The angles 1 and 2 can be calculated from the tire stiness cR and the bodys roll stiness.

(4.4)

On a straight-ahead drive, the vehicle weight is equally distributed to both tires

stat stat = FzR = FzL

1 mg . 2 Fz

(4.5)

With
K stat FzL = FzL +

(4.6)

and the relations (4.3), (4.5) one gets for the increase of the wheel load at the overturning limit 1 Fz = m g . (4.7) 2 The resulting tire deection then follows from Fz = cR where cR is the radial tire stiness. Because the right tire simultaneously rebounds for the same amount, for the roll angle of the axle 2 r mg K K 2 r = s 1 or 1 = = . (4.9) s s cR holds. In analogy to (4.1) the balance of torques at the body delivers cW 2 = m ay h2 + m g h2 (1 + 2 ) , where cW names the roll stiness of the body suspension. (4.10) r, (4.8)

48

FH Regensburg, University of Applied Sciences Accordingly, at the overturning limit ay = aK y

K 2 =

Prof. Dr.-Ing. G. Rill

aK mgh2 mgh2 y + K g cW mgh2 cW mgh2 1

(4.11)

holds. Not allowing the vehicle to overturn already at aK y = 0 demands a minimum of roll min stiness cW > cW = mgh2 . With (4.9) and (4.11) the overturning condition (4.4) reads as (h1 + h2 ) aK aK s 1 1 1 1 y y = (h1 + h2 ) h2 h2 , g 2 cR g cW 1 cW 1 cR cR c R = mg s have been used. Resolved for the normalized lateral acceleration s aK y 2 = g h1 + h2 + remains.
overturning limit a y /g 0.6 0.5 0.4 0.3 0.2 5 0.1 0 0 10 20 normalized roll stiffness c W * 0 0 10 20 normalized roll stiffness c W * 15 10 20 roll angle =K +K 1 2

(4.12)

where, for abbreviation purposes, the dimensionless stinesses and c W = cW m g h2 (4.13)

h2 cW 1

1 c R

(4.14)

Figure 4.2: Tilting Limit for a Truck at Steady State Cornering At heavy trucks, a twin tire axle can be loaded with m = 13 000 kg . The radial stiness of one tire is cR = 800 000 N/m and the track with can be set to s = 2 m. The values h1 = 0.8 m

49

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

and h2 = 1.0 m hold at maximal load. This values deliver the results shown in Fig. 4.2 Even at a rigid body suspension c W the vehicle turns over at a lateral acceleration of ay 0.5 g . The roll angle of the vehicle then 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 then turn over at a roll angle of 10 .

4.1.2

Roll Support and Camber Compensation

When a vehicle drives through a curve with the lateral acceleration ay , centrifugal forces are delivered to the single masses. At the even roll model in Fig. 4.3 these are the forces mA ay and mR ay , where mA names the body mass and mR the wheel mass.
b/2 zA mA a y FF2 h0 mR a y r0 z2 S2 Q2 Fy2 y2 F y2 2 mR a y SA A yA FF1 z1 S1 Q1 F z1 1 y1 F y1 b/2

Figure 4.3: Plane Roll Model Through the centrifugal force mA ay applied to the body at the center of gravity, a roll torque is generated, that rolls the body with the angle A and leads to a counterwise deection of the tires z1 = z2 .
a) A b) A

2
roll center

1 0

roll center

2 0

Figure 4.4: Roll Behavior at Cornering: a) without and b) with Camber Compensation

50

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

If the wheels only move vertically to the body at bound and rebound, then, at fast cornering the wheels are no longer perpendicular to the track Fig. 4.4 a. The camber angles 1 > 0 and 2 > 0 result in an unfavorable pressure distribution in the contact area, which leads to a reduction of the maximally transmittable lateral forces. At more sportive vehicles thus axle kinematics are employed, where the wheels are rotated around the longitudinal axis at bound and rebound, 1 = 1 (z1 ) and 2 = 2 (z2 ). With this, a camber compensation can be achieved with 1 0 and 2 0. Fig. 4.4 b. By the rotation of the wheels around the longitudinal axis on jounce, 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.

4.1.3

Roll Center and Roll Axis

The roll center can be constructed from the lateral movement of the wheel contact points Q1 and Q2 , Fig. 4.4.

roll axis roll center rear

Figure 4.5: Roll Axis The line through the roll center at the front and rear axle is called roll axis, Fig. 4.5.

roll center front

4.1.4

Roll Angle and Wheel Loads

At steady state cornering, the vehicle is balanced. With the principle of virtual work W = 0 the equilibrium position can be calculated. At the plane vehicle model in Fig. 4.3 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 cQ , cR . The (4.15)

51

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

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 A + y1 )2 mR ay (yA + hR A + y2 )2
1 2 2 cA z1 1 2 2 cA z2

1 c (z1 z2 )2 2 S 1 c (yA + h0 A + y1 + r0 1 )2 2 Q
b 1 c zA + 2 A + z1 2 R 2 1 2

(4.16) cQ (yA + h0 A + y2 + r0 2 )2
2

1 2

b cR zA 2 A + 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 y1 = y z1 , z 1 = 1 z and y2 = y z2 , z 2 = 2 z (4.17)

the work W can be described as function of the position vector y = [ yA , zA , A , z1 , z2 ]T . Due to W = W (y ) principle of virtual work (4.15) leads to W = W y = 0 . y (4.20) (4.19) (4.18)

Because of y = 0 a system of linear equations in the form of Ky = b results from (4.20). The matrix K and the vector b are given by 2 cQ 0 K = 2 cQ h0 yQ c z Q Q y c z Q 0 2 cR 0 cR cR 2 cQ h0 0 c
Q b c + h0 y c 2 R z Q

(4.21)

y Q c z Q

cR
Q b c + h0 y c 2 R z Q

c A + cS + cR cS

b 2 cR h0 y c z Q

cR y Q b 2 cR h0 z cQ cS cA + cS + cR

Q y c z Q

(4.22)

52

FH Regensburg, University of Applied Sciences and

Prof. Dr.-Ing. G. Rill

b =

mA + 2 mR 0 (m1 + m2 ) hR mR y/z mR y/z

ay .

(4.23)

The following abbreviations have been used: y Q y = + r0 , z z z c A = cA + cQ y z

2

c =

2 cQ h2 0

+ 2 cR

b 2

(4.24)

The system of linear equations (4.21) 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.

4.2
4.2.1

Kinematic Approach
Kinematic Tire Model

When a vehicle drives through the curve at low lateral acceleration, low lateral forces are needed for course holding. At the wheels then hardly lateral slip occurs. In the ideal case, with vanishing lateral slip, the wheels only move in circumferential direction. The speed component of the contact point in the tires lateral direction then vanishes vy = eT y v 0P = 0 . (4.25)

This kinematic constraint equation can be used for course calculation of slowly moving vehicles.

4.2.2

Ackermann Geometry

Within the validity limits of the kinematic tire model the necessary steering angle of the front wheels can be constructed via given momentary turning center M , Fig. 4.6. At slowly moving vehicles the lay out of the steering linkage is usually done according to the Ackermann geometry. With given steering angle of a wheel, e.g. 2 , the wheel base a xes the momentarily driven curve radius tan 2 = a ; R or R = a . tan 2 (4.26)

For the second wheel however tan 1 =

a , R+b

(4.27)

53

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

2 M

Figure 4.6: Ackermann Steering Geometry at a two-axled Vehicle holds, with b describing the track width. With (4.26), (4.27) reads as tan 1 = a a +b tan 2 (4.28)

which can be written as tan 1 =

tan 2 . b 1 + tan 2 a

(4.29)

a A a The deviations 1 = 1 1 of the actual steering angle 1 from the Ackermann steering A angle 1 , which follows from (4.29), are used to judge a steering system.

At a rotation around the momentary pole M the direction of speed is xed for every point of the vehicle. The angle between the speed vector v and the vehicles longitudinal axis is called side slip angle.

4.2.3
4.2.3.1

Vehicle Model with Trailer

Position

Fig. 4.7 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

54

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

a b y2

A1

x2

A2

y0

y 3

A3

x0

Figure 4.7: Kinematic Model with Trailer axle center r0A2 ,0 and the rotation matrix A02

xF = yF R

(4.30)

cos sin 0 cos 0 . = sin 0 0 1

(4.31)

Here, the tire radius R is considered to be constant, and xF , yF as well as are generalized coordinates. The position vector r0A1 ,0 = r0A2 ,0 + A02 rA2 A1 ,2 and the rotation matrix A01 = A02 A21 mit A21 cos sin 0 cos 0 = sin 0 0 1 (4.33) mit rA2 A1 ,2 a = 0 0 (4.32)

55

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

describe the position and the orientation of the front axle, where a = const labels the wheel base and the steering angle. The position vector r0A3 ,0 = r0A2 ,0 + A02 rA2 K,2 + A23 rKA3 ,3 with rA2 K,2 and the rotation matrix A03 = A02 A23 mit A23 cos sin 0 cos 0 = sin 0 0 1 (4.36) b = 0 0 c = 0 0 (4.34)

and rKA3 ,2

(4.35)

dene the position and the orientation of the trailer axis, with labelling the bend angle between vehicle and trailer and b, c marking the distances from the rear axle A2 to the coupling point K and from the coupling point K to the trailer axis A3 . 4.2.3.2 Vehicle Movements

According to the kinematic tire model, cf. section 4.2.1, the velocity at the rear axle can only have a component in the vehicles longitudinal direction v x2 v0A2 ,2 = 0 . (4.37) 0 The time derivative of (4.30) results in x F F . = y 0

v0A2 ,0 = r 0A2 ,0

(4.38)

With the transformation of (4.37) into the system 0 v x2 cos vx2 v0A2 ,0 = A02 v0A2 ,2 = A02 0 = sin vx2 0 0

(4.39)

one gets by equalizing with (4.38) two rst order dierential equations for the position coordinates xF and yF x F = cos vx2 , (4.40) y F = sin vx2 .

56

FH Regensburg, University of Applied Sciences The velocity at the front axis follows from (4.32)

Prof. Dr.-Ing. G. Rill

v0A1 ,0 = r 0A1 ,0 = r 0A2 ,0 + 02,0 A02 rA2 A1 ,2 . Transformed into the vehicle xed system x2 , y2 , z2 v x2 0 a v x2 . v0A1 ,2 = 0 + 0 0 = a 0 0 0 v0A2 ,2 remains. The unit vectors ex1,2 cos = sin 0 and ey1,2 sin = cos 0 02,2 rA2 A1 ,2

(4.41)

(4.42)

(4.43)

dene the longitudinal and lateral direction at the front axle. According to (4.25) the velocity component lateral to the wheel must vanish, eT = 0. y 1,2 v0A1 ,2 = sin vx2 + cos a In longitudinal direction then eT = vx1 x1,2 v0A1 ,2 = cos vx2 + sin a remains. From (4.44) a rst order dierential equation follows for the yaw angle = v x2 tan . a (4.46) (4.45) (4.44)

4.2.3.3

Entering a Curve

In analogy to (4.26) the steering angle can be related to the current track radius R or with k = 1/R to the current track curvature tan = a = ak . R (4.47)

The dierential equation for the yaw angle then reads as = v x2 k . With the curvature gradient k = k (t) = kC t T (4.49) (4.48)

57

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

The entering of a curve is described as a continuous transition from a line with the curvature k = 0 into a circle with the curvature k = kC . The yaw angle of the vehicle can now be calculated by simple integration (t) = vx2 kC t2 , T 2 (4.50)

where at time t = 0 a vanishing yaw angle, (t = 0) = 0, has been assumed. The vehicles position then follows with (4.50) from the dierential equations (4.40)
t=T

x F = v x2
t=0

cos

vx2 kC t2 T 2

t=T

dt ,

yF = v x2
t=0

sin

vx2 kC t2 T 2

dt .

(4.51)

At constant vehicle speed vx2 = const. (4.51) is the parameterized form of a clothoide. From (4.47) 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 = (t) a k = a kC (4.52) T holds, i.e. the driving through a clothoide is manageable by continuous steer motion. 4.2.3.4 Trailer Movements

The velocity of the trailer axis can be received by dierentiation of the position vector (4.34) v0A3 ,0 = r 0A3 ,0 = r 0A2 ,0 + 02,0 A02 rA2 A3 ,2 + A02 r A2 A3 ,2 . With rA2 A3 ,2 = rA2 K,2 + A23 rKA3 ,3 and r A2 A3 ,2 b c cos c sin = 0 (4.53)

(4.54)

c cos c sin 0 = 0 c sin = c cos 0 0 23,2 A23 rKA3 ,3

(4.55)

it remains, if (4.53) is transformed into the vehicle xed frame x2 , y2 , z2 v x2 0 b c cos c sin vx2 + c sin ( + ) c sin + c cos = b c cos ( + ) . v0A3 ,2 = 0 + 0 0 0 0 0 v0A2 ,2 02,2 rA2 A3 ,2 r A2 A3 ,2 (4.56)

58

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The longitudinal and lateral direction at the trailer axis are dened by the unit vectors cos sin ex3,2 = sin and ey3,2 = cos . (4.57) 0 0 At the trailer axis the lateral velocity must also vanish + ) + cos b c cos ( + ) = 0 . (4.58) eT y 3,2 v0A3 ,2 = sin vx2 + c sin ( In longitudinal direction + ) + sin b c cos ( + ) = vx3 (4.59) eT x3,2 v0A3 ,2 = cos vx2 + c sin ( remains. When (4.46) is inserted into (4.58), one gets a dierential equation of rst order for the bend angle v x2 a b = sin + (4.60) cos + 1 tan . a c c The dierential equations (4.40) and (4.46) describe position and orientation within the x0 , y0 plane. The position of the trailer relative to the vehicle follows from (4.60). 4.2.3.5 Course Calculations

For a given set of vehicle parameters a, b, c, and predened time functions of the vehicle speed, vx2 = vx2 (t) and the steering angle, = (t) the course of vehicle and trailer can be calculated by numerical integration of the dierential equations (4.40), (4.46) and (4.60).
20 [m] 10 0 front axle rear axle trailer axle

-30

-20

-10

10 [m]

20

30

40

50

60

30 [Grad] 20 10 0 0

front axle steer angle

10

15 [s]

20

25

30

Figure 4.8: Entering a Curve If the steering angle is slowly increased at constant driving speed, then the vehicle drives gure which is similar to a clothoide, Fig. 4.8.

59

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

4.3
4.3.1

Simple Handling Model

Forces

The tire forces at the wheel of one axle are combined to one resulting force. Tire torques, the rolling resistance and aerodynamic forces and torques applied at the vehicle remain unrespected. Unlike with the kinematic tire model, now small lateral movements in the contact points are permitted. At small lateral slips, the lateral force can be approximated by a linear approach Fy = cS sy (4.61)

where cS is a constant depending on the wheel load Fz and the lateral slip sy is dened by (2.44). Because the vehicle is neither accelerated nor decelerated, the rolling condition is fullled at every wheel rD = eT (4.62) x v 0P . Here rD is the dynamic tire radius, v0P the contact point velocity and ex the unit vector in longitudinal direction. With the lateral tire speed vy = eT y v 0P and the rolling condition (4.62) the lateral slip can be calculated from sy = eT y v 0P , T | ex v0P | (4.64) (4.63)

with ey labelling the unit vector in the tires lateral direction. So the movements of the vehicle model, Fig. 4.9, within the x0 , y0 plane are only inuenced by the lateral forces Fy1 = cS 1 syA1 ; Fy2 = cS 2 syA2 . (4.65)

4.3.2

Kinematics

At the single-track vehicle model, as drawn in Fig. 4.9, the yaw motion of the vehicle is described by the angle . The vehicle velocity at the center of gravity is given as constant v = const. (4.66)

With the side slip angle the vehicle velocity can be expressed in the vehicle xed frame xF , yF , zF v cos vSF ,F = v sin . (4.67) 0

60

FH Regensburg, University of Applied Sciences

y0
yF a2 y2
v

Prof. Dr.-Ing. G. Rill

a1
y
1

A1 SF
xF

Fy1

A2 Fy2

x2

x0

Figure 4.9: Single Track Vehicle Model For the calculation of the lateral slips, the speed vectors and the unit vectors in longitudinal and lateral direction of the single axles are needed. One gets cos sin v cos ex1 ,F = sin , ey1 ,F = cos , v0A1 ,F = v sin + a1 (4.68) 0 0 0 and ex2 ,F 1 = 0 , 0 0 = 1 , 0 v cos . = v sin a2 0

ey2 ,F

v0A2 ,F

(4.69)

4.3.3

Lateral Slips

With (4.69), the lateral slip at the front axle follows from (4.64): s y A1 = + sin (v cos ) cos (v sin + a1 ) . | cos (v cos ) + sin (v sin + a1 )| (4.70)

The lateral slip at the rear axle is given by s y A2 = v sin a2 . | v cos | (4.71)

The yaw velocity of the vehicle , the side slip angle and the steering angle are considered to be small | a1 | | v | ; | a2 | |v | (4.72)

61

Vehicle Dynamics | |

FH Regensburg, University of Applied Sciences 1 and | | 1. (4.73)

Because the side slip angle always labels the smaller angle between speed vector and vehicle longitudinal axis, instead of v sin v the approximation v sin |v | has to be used. Respecting (4.72), (4.73) and (4.74), from (4.70) and (4.71) then follow syA1 = and syA2 = + a1 v + |v | |v | a2 . |v | (4.75) (4.74)

(4.76)

4.3.4

Equations of Motion

To derive the equations of motion, the velocities, angular velocities and the accelerations are needed. For small side slip angles 1, (4.67) can be approximated by v vSF ,F = |v | . 0 0 = 0 .

(4.77)

The angular velocity is given by 0F,F (4.78)

If the vehicle accelerations are also expressed in the vehicle xed frame xF , yF , zF , one nds at constant vehicle speed v = const and with neglecting small higher order terms 0 . aSF ,F = 0F,F vSF ,F + v SF ,F = v (4.79) + |v | 0 The angular acceleration is given by 0F,F where the substitution = (4.81) 0 = 0 (4.80)

62

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

was used. The linear momentum in the vehicles lateral direction reads as ) = Fy1 + Fy2 , m (v + |v |

(4.82)

where, due to the small steering angle, the term Fy1 cos has been approximated by Fy1 and m describes the vehicle mass. With (4.81) the angular momentum delivers = a1 Fy1 a2 Fy2 , (4.83) where names the inertia of vehicle around the vertical axis. With the linear description of the lateral forces (4.65) and the lateral slips (4.75), (4.76) one gets from (4.82) and (4.83) two coupled, but linear rst order dierential equations a1 v a2 cS 2 v = cS 1 + + + , (4.84) |v | |v | |v | m |v | m |v | |v | a1 v a2 cS 2 a1 c S 1 + |v | |v | which can be written in the form of a state equation v a1 cS 1 a2 cS 2 cS 1 + cS 2 m |v | |v | m |v ||v | = a1 c S 1 a2 c S 2 a2 c S 1 + a2 2 cS 2 1 x |v | = A + a2 |v | v cS 1 |v | m |v | + v a1 c S 1 |v | B , (4.85)

u .

(4.86) If a system can be, at least approximatively, described by a linear state equation, then, stability, steady state solutions, transient response, and optimal controlling can be calculated with classic methods of system dynamics.

4.3.5
4.3.5.1

Stability
Eigenvalues x = Ax (4.87) (4.88) (4.89) (4.90)

The homogeneous state equation describes the eigen-dynamics. If the approach xh (t) = x0 et is inserted into (4.87), then the homogeneous equation remains ( E A) x0 = 0 . Non-trivial solutions x0 = 0 one gets for det | E A| = 0 . The eigenvalues provide information about the stability of the system.

63

Vehicle Dynamics 4.3.5.2 Low Speed Approximation

FH Regensburg, University of Applied Sciences

The state matrix c S 1 + c S 2 v a1 c S 1 a2 c S 2 m |v | |v | m |v ||v | = a2 c S 1 + a2 2 cS 2 0 1 |v | (4.91)

Av0

approximates at v 0 the eigen-dynamics of vehicles at low speeds. The matrix (4.91) has the eigenvalues 1 v 0 = cS 1 + cS 2 m |v | and 2v0 =
2 a2 1 c S 1 + a2 c S 2 . |v |

(4.92)

The eigenvalues are real and, independent from the driving direction, always negative. Thus, vehicles at low speed possess an asymptotically stable driving behavior! 4.3.5.3 High Speed Approximation

At highest driving velocities v , the state matrix can be approximated by v 0 |v | . Av = a c a c 1 S1 2 S2 0 Using (4.93) one receives from (4.90) the relation 2 v with the solutions 1,2v = v a1 c S 1 a2 c S 2 = 0 |v | v a1 c S 1 a2 c S 2 . |v |

(4.93)

(4.94)

(4.95)

When driving forward with v > 0, the root argument is positive, if a1 cS 1 a2 cS 2 > 0 or a1 cS 1 > a2 cS 2 holds. Then however, one eigenvalue is positive and the system is unstable. Two zero-eigenvalues 1 = 0 and 2 = 0 one gets for a1 cS 1 a2 cS 2 = 0 or a1 cS 1 = a2 cS 2 . (4.97) (4.96)

64

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The driving behavior is then indierent. Slight parameter variations however can lead to an unstable behavior. With a1 cS 1 a2 cS 2 < 0 or a1 cS 1 < a2 cS 2 (4.98) and v > 0 the root argument in (4.95) becomes negative. The eigenvalues are then imaginary, and disturbances lead to undamped vibrations. To avoid instability, high-speed vehicles have to satisfy the condition (4.98). The root argument in (4.95) changes at backward driving its sign. A vehicle showing stable driving behavior at forward driving becomes unstable at fast backward driving!

4.3.6
4.3.6.1

Steady State Solution

Side Slip Angle and Yaw Velocity

With a given steering angle = 0 , after a certain time, a stable system reaches steady state. With xst = const. or x st = 0, the state equation (4.86) is reduced to a linear system of equations A xst = B u . (4.99)

With the elements from the state matrix A and the vector B one gets from (4.99) two equations to determine the steady state side slip angle st and the steady state angular velocity st at a constant given steering angle = 0 |v | (cS 1 + cS 2 ) st + (m v |v | + a1 cS 1 a2 cS 2 ) st = v cS 1 0 ,
2 |v | (a1 cS 1 a2 cS 2 ) st + (a2 1 cS 1 + a2 cS 2 ) st = v a1 cS 1 0 ,

(4.100) (4.101)

where the rst equation has been multiplied by m |v | |v | and the second with |v |. The solution can be derived from v cS 1 0 st = v a1 cS 1 0 |v | (cS 1 + cS 2 ) |v | (a1 cS 1 a2 cS 2 ) and |v | (cS 1 + cS 2 ) st = v cS 1 0 (4.103) |v | (a1 cS 1 a2 cS 2 ) v a1 cS 1 0 |v | (cS 1 + cS 2 ) |v | (a1 cS 1 a2 cS 2 ) m v |v | + a1 cS 1 a2 cS 2
2 a2 1 c S 1 + a2 c S 2

m v |v | + a1 cS 1 a2 cS 2
2 a2 1 c S 1 + a2 c S 2

(4.102)

m v |v | + a1 cS 1 a2 cS 2
2 a2 1 c S 1 + a2 c S 2

65

Vehicle Dynamics The denominator results in

FH Regensburg, University of Applied Sciences

detD = |v | cS 1 cS 2 (a1 + a2 )2 m v |v | (a1 cS 1 a2 cS 2 ) For a non vanishing denominator detD = 0 steady state solutions exist a1 v cS 2 (a1 + a2 ) 0 , = |v | a + a m v |v | a1 cS 1 a2 cS 2 1 2 cS 1 cS 2 (a1 + a2 ) a2 m v |v | a1 cS 1 a2 cS 2 0 . a1 + a2 m v |v | cS 1 cS 2 (a1 + a2 ) v

(4.104)

st

(4.105)

st =

(4.106)

At forward driving vehicles v > 0 the steady state side slip angle, starts with the kinematic value a2 v v v 0 v 0 st = 0 and st = 0 (4.107) |v | a1 + a2 a1 + a2 and decreases with increasing speed. At speeds larger then vst=0 = the side slip angle changes the sign. Because the rear wheels are not steered, higher slip angles at the rear axle can only be reached by slanting the car.
2 [deg] 0 -2 -4 -6 -8 -10 0 a1*c S1/a2*c S2 = 0.66667 a1*cS1/a2*c S2 = 1 a1*cS1/a2*c S2 = 1.3333 10 20 v [m/s] 30 40 r [m] 150 100 50 0 a1*cS1/a2*cS2 = 0.66667 a1*cS1/a2*cS2 = 1 a1*cS1/a2*cS2 = 1.3333 0 10 20 v [m/s] 30 40 steady state side slip angle 200 radius of curvrature

a2 cS 2 (a1 + a2 ) a1 m

(4.108)

m=700 kg ; =1000 kg m2 ;

a1 =1.2 m; a2 =1.3 m;

cS 1 = 80 000 N m;

cS 2

110 770 N m = 73 846 N m 55 385 N m

Figure 4.10: Steady State Cornering

66

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

In Fig. 4.10 the side slip angle , and the driven curve radius R are plotted versus the driving speed v . The steering angle has been set to 0 = 1.4321 , in order to let the vehicle drive a circle with the radius R0 = 100 m at v 0. The actually driven circle radius R has been calculated via v . st = R

(4.109)

Some concepts for an additional steering of the rear axle were trying to keep the vehicles side slip angle to zero by an appropriate steering or controlling. Due to numerous problems production stage could not yet be reached. 4.3.6.2 Steering Tendency

After reaching the steady state solution, the vehicle moves in a circle. When inserting (4.109) into (4.106) and resolving for the steering angle, one gets 0 = a1 + a2 v 2 v a1 c S 1 a2 c S 2 m . R R |v | cS 1 cS 2 (a1 + a2 ) (4.110)

The rst term is the Ackermann steering angle, which follows from (4.26) with the wheel base a = a1 + a2 and the approximation for small steering angles tan 0 0 . The Ackermann-steering angle provides a good approximation for slowly moving vehicles, because at v 0 the second expression in (4.110) becomes neglectably small. At higher speeds, depending on the value of a1 cS 1 a2 cS 2 and the driving direction (forward: v > 0, backward: v < 0), the necessary steering angle diers from the Ackermann-steering angle. The dierence is proportional to the lateral acceleration ay = v2 . R (4.111)

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 cS 1 , cS 2 . The various steering tendencies are arranged in the table 4.1. 4.3.6.3 Slip Angles

st = 0, With the conditions for a steady state solution st = 0 and the relation (4.109), the equations of motion (4.82) and (4.83) can be dissolved for the lateral forces Fy1st = Fy2st = a2 v2 m , a1 + a2 R a1 v2 m a1 + a2 R

or

a1 Fy2st = . a2 Fy1st

(4.112)

67

Vehicle Dynamics

FH Regensburg, University of Applied Sciences a1 c S 1 <1 a2 c S 2 a1 c S 1 =1 a2 c S 2 a1 c S 1 >1 a2 c S 2

understeer neutral oversteer

A 0 > 0

or a1 cS 1 < a2 cS 2 or a1 cS 1 = a2 cS 2 or a1 cS 1 > a2 cS 2

or or or

A 0 = 0

A 0 < 0

Table 4.1: Steering Tendency of a Vehicle at Forward Driving

With the linear tire model (4.61) one gets Fy1st = cS 1 sst yA1 and Fy2st = cS 2 sst y A2 , (4.113)

st where sst yA1 and syA2 label the steady state lateral slips at the axles. From (4.112) and (4.113) now follows cS 2 sst sst a1 Fy2st a1 c S 1 yA2 y A2 = = = or . (4.114) st a2 Fy1st a2 c S 2 cS 1 syA1 sst y A1

So, the steering tendency can also be determined from the slip angle at the axles.

4.3.7
4.3.7.1

Inuence of Wheel Load on Cornering Stiffness

Linear Wheel Load Inuence

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 S 1 = cS Fz 1 and clin S 2 = cS Fz 2 . (4.115)

holds. The weight of the vehicle G = m g is distributed over the axles according to the position of the center of gravity a2 a1 Fz1 = G and .Fz2 = G (4.116) a1 + a2 a1 + a2 With (4.115) and (4.116) one gets a1 clin S 1 = a1 c S and a2 clin S 2 = a2 c S a2 G a1 + a2 (4.117)

a1 G. (4.118) a1 + a2 A vehicle with identical tires would thus be steering neutrally at a linear inuence of wheel load on the slip stiness, because of
lin a1 clin S 1 = a2 c S 2

(4.119)

68

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The fact that the lateral force is applied behind the center of the contact area at the caster oset distance, leads, because of a1 a1 |v n and a2 a2 + |v n to a stabilization of v | L1 v | L1 the driving behavior, independent from the driving direction. 4.3.7.2 Digressive Wheel Load Inuence At a real tire, a digressive inuence of wheel load on the tire forces is observed, Fig. 4.11.
6 5 4

3 2 1 0

Fz [kN]

Fz [N ] 0 1000 2000 3000 4000 5000 6000 7000 8000

Fy [N ] 0 758 1438 2043 2576 3039 3434 3762 4025

According to (4.83) 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 (4.120) holds. At a vehicle with the wheel base a = 2.45 m the axle loads Fz1 = 4000 N and Fz2 = 3000 N deliver 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 receives from the table in 4.11 Fy1 = 2576 N and Fy2 = 2043 N . With this, the condition (4.120) delivers 1.05 2576 1.45 2043 = 257.55 . The value is signicantly negative and thus stabilizing. Vehicles with a1 < a2 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 ), a stable driving behavior can generally only be achieved with dierent tires.

Fy [kN]

Figure 4.11: Lateral Force Fy over Wheel Load Fz at dierent Slip Angles

69

Vehicle Dynamics 4.3.7.3

FH Regensburg, University of Applied Sciences

Steering Tendency depending on Lateral Acceleration

At increasing lateral acceleration the vehicle is more and more supported by the outer wheels. At a suciently rigid vehicle body the wheel load dierences can dier, because of dierent kinematics (roll support) or dierent roll stinesses, Fig. 4.12.
+TT -TT

PR0+P PF0+P PF0-P

PR0-P PF0+PF

PR0+PR PF0-PF

PR0-PR

Figure 4.12: Wheel Load Dierences at exible and sti Frame Due to the digressive inuence of wheel load, the deliverable 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 decreases more than at the rear axle and the vehicle becomes more stable with increasing lateral force, i.e. more understeering.

70

5 Vertical Dynamics
5.1 Goals

The aim of vertical dynamics is the tuning of body suspension and damping to guarantee good driving 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 instationary 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 wheel load. Small variations of wheel load also reduce the stress on the track. The comfort of a vehicle is subjectively judged by the driver. In literature, dierent approaches 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
5.2.1

Basic Tuning
Simple Models

Fig. 5.1 shows simple quarter car models, that are suitable for basic investigations of body and axle vibrations. At normal vehicles the wheel mass m is in relation to the respective body mass M much smaller m M . The coupling of wheel and body movement can thus be neglected for basic investigations. In describing the vertical movements of the body, the wheel movements remain unrespected. If the wheel movements are in the foreground, then body movements can be neglected. The equations of motion for the models read as Mz B + dS z B + cS zB = dS z R + cS zR (5.1)

71

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

M cS `` `
`` c `

6 zB

`` ` cS ` dS ``

dS
6 zR

m
`` `` c `` T c

6 zW

6 zR

Figure 5.1: Simple Vehicle and Suspension Model and mz W + dS z W + (cS + cT ) zW = cT zR , (5.2) where zB and zW label the vertical movements of the body and the wheel mass out of the equilibrium position. The constants cS , dS describe the body suspension and damping, and cT the vertical stiness of the tire. The tire damping is hereby neglected against the body damping.

5.2.2

Track

The track is given as function in the space domain zR = zR (x) . (5.3)

In (5.1) also the time gradient of the track irregularities is necessary. From (5.3) rstly follows z R = d zR dx . dx dt (5.4)

At the simple model the speed, with which the track irregularities are probed equals the vehicle speed dx/dt = v . If the vehicle speed is given as time function v = v (t), the covered distance x can be calculated by simple integration.

5.2.3

Spring Preload

The suspension spring is loaded with the respective vehicle load. At linear spring characteristics the steady state spring deection is calculated from f0 = Mg . cS (5.5)

72

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

At a conventional suspension without niveau regulation a load variation M M + M leads to changed spring deections f0 f0 + f . In analogy to (5.5) the additional deection follows from Mg f = . (5.6) cS If for the maximum load variation M max the additional spring deection is limited to f max the suspension spring rate can be estimated by a lower bound cS M max g . f max (5.7)

5.2.4

Eigenvalues

At an ideally even track the right side of the equations of motion (5.1), (5.2) vanishes because of zR = 0 and z R = 0. The remaining homogeneous second order dierential equations can be written as 2 z + 2z + 0 z = 0. (5.8) The respective attenuation constants and the undamped natural circular frequency 0 for the models in Fig. 5.1 can be determined from a comparison of (5.8) with (5.1) and (5.2). The results are arranged in table 5.1. undamped attenuation Eigenfrequency constant B = dS 2M dS R = 2m
2 B = 0 2 W 0

Motions

Dierential Equation

Body Wheel

Mz B + dS z B + cS zB = 0 mz W + dS z W + (cS + cT ) zW = 0

cS M cS + cT = m

Table 5.1: Attenuation Constants and undamped natural Frequencies

With z = z0 et the equation

2 (2 + 2 + 0 ) z0 et = 0 .

(5.9) (5.10) (5.11)

follows from (5.8). For

2 2 + 2 + 0 = 0

also non-trivial solutions are possible. The characteristical equation (5.11) has got the solutions 1,2 =
2 2 0

(5.12)

73

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

2 For 2 0 the eigenvalues 1,2 are real and, because of 0 not positive, 1,2 0. Disturbances z (t = 0) = z0 with z (t = 0) = 0 then subside exponentially. 2 With 2 < 0 the eigenvalues become complex

1,2 = i The system now executes damped oscillations. The case

2 2 = 0 ,

2 0 2 .

(5.13)

bzw.

= 0

(5.14)

describes, in the sense of stability, an optimal system behavior. Wheel and body mass, as well as tire stiness are xed. The body spring rate can be calculated via load variations, cf. section 5.2.3. With the abbreviations from table 5.1 now damping parameters can be calculated from (5.14) which provide with (dS )opt1 = 2 M optimal body vibrations and with (dS )opt2 = 2 m optimal wheel vibrations. cS + cT = 2 m (cS + cT ) m (5.16) cS = 2 M cS M (5.15)

5.2.5

Free Vibrations

Fig. 5.2 shows the time response of a damped single-mass oscillator to an initial disturbance as results from the solution of the dierential equation (5.8). The system here has been started without initial speed z (t = 0) = 0 but with the initial disturbance z (t = 0) = z0 . If the attenuation constant is increased at rst the system approaches the steady state position zG = 0 faster and faster, but then, a slow asymptotic behavior occurs. Counting dierences from the steady state positions as errors (t) = z (t) zG , allows judging the quality of the vibration. The overall error is calculated by
t=tE 2 G

=
t=0

z (t)2 dt ,

(5.17)

where the time tE have to be chosen appropriately. If the overall error becomes a Minimum
2 G

M inimum

(5.18)

the system approaches the steady state position as fast as possible.

74

FH Regensburg, University of Applied Sciences

z0

Prof. Dr.-Ing. G. Rill

z(t) t

Figure 5.2: Damped Vibration To judge driving comfort and safety the deections zB and accelerations z B of the body and the dynamic wheel load variations are used. The system behavior is optimal if the parameters M , m, cS , dS , cT result from the demands for comfort
t=tE 2 GC

=
t=0

g1 zB

g2 z B

dt M inimum

(5.19)

and safety
t=tE 2 GS

=
t=0

cT zW

dt M inimum .

(5.20)

With the factors g1 and g2 deections and accelerations can be weighted dierently. In the equations of motion for the body (5.1) the terms M z B and cS zB are added. With g1 = M and g2 = cS one gets system-tted weighting factors. At the damped single-mass oscillator, the integrals in (5.19) can, for tE , still be solved analytically. One gets dS cS 2 2 cS 1 (5.21) + 2 GC = zB0 M 2 M dS and dS m 2 2 2 1 . (5.22) + GS = zW0 cT 2 cS + cT dS Small body suspension stinesses cS 0 or large body masses M make the comfort criteria (5.21) small 2 GC 0 and so guarantee a high driving comfort. A great body mass however is uneconomic. The body suspension stiness cannot be reduced arbitrary low values, because then load variations would lead to too great changes in static deection. At xed values for cS and M the damper can be designed in a way that minimizes the comfort criteria (5.21). From the necessary condition for a minimum 2 GC 2 cS 1 = zB 0 dS M 2 1 cS 2 2 M dS = 0 (5.23)

75

Vehicle Dynamics the optimal damper parameter

FH Regensburg, University of Applied Sciences

(dS )opt3 = that guarantees optimal comfort follows.

2 cS M ,

(5.24)

Small tire spring stinesses cT 0 make the safety criteria (5.22) small 2 GS 0 and thus reduce dynamic wheel load variations. The tire spring stiness can however not be reduced to arbitrary low values, because this would cause too great tire deformation. Small wheel masses m 0 and/or a hard body suspension cS also reduce the safety criteria (5.22). The use of light metal rims increases, because of wheel weight reduction, the driving safety of a car. Hard body suspensions contradict driving comfort. With xed values for cS , cT and m here the damper can also be designed to minimize the safety criteria (5.22). From the necessary condition of a minimum 2 1 GS 2 = zW c2 0 T dS 2 the optimal damper parameter (dS )opt4 = follows, which guarantees optimal safety. (cS + cT ) m , (5.26) m 1 2 cS + cT dS = 0 (5.25)

5.3

Nonlinear Force Elements

subsectionQuarter Car Model The principal inuence of nonlinear characteristics on driving comfort and safety can already be displayed on a quarter car model Fig. 5.3.
progressive spring FF
xR

zB M

degressive damper FD
v

FR
x

m cT

zW zR

Figure 5.3: Quarter Car Model with nonlinear Characteristics

76

FH Regensburg, University of Applied Sciences The equations of motion are given by Mz B = F M g mz W = Fz F m g ,

Prof. Dr.-Ing. G. Rill

(5.27)

where g = 9.81 m/s2 labels the constant of gravity and M , m are the masses of body and wheel. The coordinates zB and zW are measured from the equilibrium position. Thus, the wheel load Fz is calculated from the tire deection zW zR via the tire stiness cT Fz = (M + m) g + cT (zR zW ) . (5.28)

The rst term in (5.28) describes the static part. The condition Fz 0 takes the wheel lift o into consideration. Body suspension and damping are described with nonlinear functions of the spring travel x = zW zB and the spring velocity v = z W z B , where x > 0 and v > 0 marks the spring and damper compression. The damper characteristics are modelled as digressive functions with the parameters pi 0, i = 1(1)4 1 p1 v 1 + p2 v v 0 (Druck) FD (v ) = . (5.31) 1 v < 0 (Zug) p3 v 1 p4 v A linear damper with the constant d is described by p1 = p3 = d and p2 = p4 = 0. For the spring characteristics the approach FF (x) = M g + FR x xR 1 p5 |x| 1 p5 xR (5.32) (5.30) (5.29)

is used, where M g marks the spring preload. With parameters within the range 0 p5 < 1, one gets dierently progressive characteristics. The special case p5 = 0 describes a linear spring with the constant c = FR /xR . All spring characteristics run through the operating point xR , FR . Thus, at a real vehicle, one gets the same roll angle, independent from the chosen progression at a certain lateral acceleration.

5.3.1

Random Road Prole

The vehicle moves with the constant speed vF = const. When starting at t = 0 at the point xF = 0, the current position of the car is given by xF (t) = vF t . (5.33)

77

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

road profil [m]

0.1 0.05 0 -0.05 -0.1 0 20

40

60 [m] 80

100

Figure 5.4: Track prole 1 The irregularities of the track can thus be written as time function zR = zR (xF (t)) The calculation of optimal characteristics, i.e. the determination of the parameters p1 to p5 , is done for three dierent tracks. Each track consists of a number of single obstacles, which lengths and heights are distributed randomly. Fig. 5.4 shows the rst track prole zS1 (x). Proles number two and three are generated from the rst by multiplication with the factors 3 and 5, zS2 (x) = 3 zS1 (x), zS3 (x) = 5 zS1 (x).

5.3.2

Vehicle Data

The values, arranged in table 5.2, describe the respective body mass of a fully loaded and an empty bus over the rear axle, the mass of the rear axle and the sum of tire stinesses at the twin tire rear axle. vehicle data M [kg] m [kg] FR [N] xR [m] cT [N/m] fully loaded 11 000 800 40 000 0.100 3 200 000 unloaded 6 000 800 22 500 0.100 3 200 000 Table 5.2: Vehicle Data

The vehicle possesses niveau-regulation. Therefore also the force FR at the reference deection xR has been tted to the load. The vehicle drives at the constant speed vF = 20 m/s. The ve parameters, pi , i = 1(1)5, which describe the nonlinear spring-damper characteristics, are calculated by minimizing quality functions.

78

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

5.3.3

Quality Criteria

In a rst quality function, driving comfort and safety are to be judged by body accelerations and wheel load variations GK 1 = 1 tE t0
tE t0

z B g

FzD FzS

(5.34)

comfort

safety

The body acceleration z B has been normalized to the constant of gravity g . The dynamic share of the normal force FzD = cT (zR zW ) follows from (5.28) with the static normal force FzS = (M + m) g . At real cars the spring travel is limited. The quality criteria is therefore extended accordingly GK 2 = 1 tE t0
tE t0

z B g

PD PS

x xR

(5.35)

comfort

safety

spring travel

where the spring travel x, dened by (5.29), has been related to the reference travel xr . According to the covered distance and chosen driving speed, the times used in (5.34) and (5.35) have been set to t0 = 0 s and tE = 8 s

5.3.4

Optimal Parameter

5.3.4.1 Linear Characteristics Judging the driving comfort and safety after the criteria GK 1 and restricting to linear characteristics, with p1 = p3 and p2 = p4 = p5 = 0, one gets the results arrayed in table5.3. The spring constants c = FR /xr for the fully loaded and the empty vehicle are dened by the optimal parameter road load p1 p2 p3 p4 p5 1 + 35766 0 35766 0 0 2 + 35763 0 35763 0 0 3 + 35762 0 35762 0 0 1 20298 0 20298 0 0 2 20300 0 20300 0 0 3 19974 0 19974 0 0 parts in quality criteria comfort safety 0.002886 0.002669 0.025972 0.024013 0.072143 0.066701 0.003321 0.003961 0.029889 0.035641 0.083040 0.098385

Table 5.3: Linear Spring and Damper Parameter optimized via GK 1 numerical values in table 5.2. One gets:cempty = 225 000N/m and cloaded = 400 000N/m.

79

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

As expected the results are almost independent from the track. The optimal value of the damping parameter d = p1 = p3 however is strongly dependent on the load state. The optimizing quasi ts the damper constant to the changed spring rate. The loaded vehicle is more comfortable and safer. 5.3.4.2 Nonlinear Characteristics

The results of the optimization with nonlinear characteristics are arrayed in the table 5.4. optimal parameter p2 p3 p4 0.000 20028 1.316 2.689 57892 1.175 3.048 311773 4.295 0.000 17255 0.337 0.808 27703 0.454 0.227 64345 0.714 parts in quality criteria comfort safety 0.000265 0.001104 0.009060 0.012764 0.040813 0.050069 0.000819 0.003414 0.012947 0.031285 0.060992 0.090250

road load p1 1 + 16182 2 + 52170 3 + 1875 1 13961 2 16081 3 9942

p5 0.9671 0.6983 0.0000 0.9203 0.6567 0.0000

Table 5.4: Nonlinear Spring and Damper Characteristics optimized via GK 1

The optimizing has been started with the linear parameters from table 5.3. Only at the extreme track irregularities of prole 3, linear spring characteristics, with p5 = 0, appear, Fig. 5.6. At moderate track irregularities, one gets strongly progressive springs. The dampers are digressive and dier in jounce and rebound. In comparison to the linear model a signicant improvement can be noted, especially in comfort. While driving over prole 2 with the loaded vehicle, the body accelerations are displayed in Fig. 5.5.
10 5 0 -5 -10 0 2 4 [s] 6 8 body accelerations [m/s2 ]

Figure 5.5: Body Accelerations optimized via GK 1

( linear, nonlinear)

80

FH Regensburg, University of Applied Sciences

spring force [kN]

Prof. Dr.-Ing. G. Rill

40 20 0 -20 -40 -0.1 -0.05

0 0.05 spring travel [m]

0.1

Figure 5.6: Optimal Spring Characteristics for fully loaded Vehicle; Criteria: GK 1 The extremely progressive spring characteristics, optimal at smooth tracks (prole 1), cannot be realize practically in that way. Due to the small spring stiness around the equilibrium position, small disturbances cause only small aligning forces. Therefore it would take long to reach the equilibrium position again. Additionally, friction forces in the body suspension would cause a large deviation of the equilibrium position. 5.3.4.3 Limited Spring Travel

Practically relevant results can only be achieved, if additionally the spring travels are judged. Firstly, linear characteristics are assumed again, table 5.5. optimal parameter road load p1 p2 p3 p4 p5 1 + 68727 0 68727 0 0 2 + 68666 0 68666 0 0 3 + 72882 0 72882 0 0 1 35332 0 35332 0 0 2 35656 0 35656 0 0 3 37480 0 37480 0 0 parts in quality criteria comfort safety s. travel 0.003854 0.003673 0.006339 0.034657 0.033025 0.057097 0.098961 0.094431 0.148757 0.004417 0.004701 0.006638 0.040049 0.042507 0.059162 0.112143 0.116722 0.155290

Table 5.5: Linear Spring and Damper Characteristics optimized via GK 2 The judging numbers for comfort and safety have worsened by limiting the spring travel in comparison to the values from table 5.3. In order to receive realistic spring characteristics, now the parameter p5 has been limited upwards to p5 0.6. Starting with the linear parameters from table 5.5, an optimization via criteria GK 2 delivers the results arranged in table 5.6. A vehicle with GK 2 -optimized characteristics manages the travel over uneven tracks with signicantly less spring travel than a vehicle with GK 1 -optimized characteristics, Fig. 5.7.

81

Vehicle Dynamics

FH Regensburg, University of Applied Sciences optimal parameter p2 p3 p4 12.89 102997 3.437 5.505 107498 1.234 4.844 152732 1.165 5.244 50353 2.082 0.601 37392 0.101 1.668 68917 0.643 parts in quality criteria comfort safety s. travel 0.001747 0.002044 0.005769 0.015877 0.018500 0.050073 0.064980 0.068329 0.116555 0.002380 0.003943 0.005597 0.024524 0.033156 0.059717 0.085001 0.102876 0.125042

road load p1 1 + 175530 2 + 204674 3 + 327864 1 66391 2 37246 3 89007

p5 0.4722 0.6000 0.5140 0.5841 0.5459 0.3614

Table 5.6: Nonlinear Spring and Damper Characteristics optimized via GK 2

spring travel [m] 0.1 0.05 0 -0.05 -0.1 0 2 4 [s] 6 8

Figure 5.7: Spring Travels on Prole 2

(- - - GK 1 , GK 2 )

The reduced spring travel however reduces comfort and safety. Still, in most cases, the according part of the quality criteria in table 5.6 lie even below the values of the linear model from table 5.3, where the spring travels have not been evaluated. By the use of nonlinear characteristics, the comfort and safety of a vehicle can so be improved, despite limitation of the spring travel. The optimal damper characteristics strongly depend on the roughness of the track, Fig. 5.8.
100 50 0 -50 -100 -1 -0.5
compression rebound

damper force [kN]

0 [m/s] 0.5

Figure 5.8: Optimal Damper Characteristics according to Table 5.6

82

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

Optimal comfort and safety are only guaranteed if the dampers are tted to the load as well as to the roughness of the track.

5.4
5.4.1
5.4.1.1

Dynamic Force Elements

System Response in the Frequency Domain
First Harmonic Oscillation

The eect of dynamic force elements is usually judged in the frequency domain. For this, on test rigs or in simulation, the force element is periodically excited with dierent frequencies f0 fi fE and amplitudes Amin Aj Amax xe (t) = Aj sin(2 fi t) . (5.36)

Starting at t = 0 at t = T0 with T0 = 1/f0 the system usually is in a steady state condition. Due to the nonlinear system behavior the system response is periodic, yet not harmonic. For the evaluation thus the answer, e.g. the measured or calculated force F , each within the intervals tSi t tSi + Ti , is approximated by harmonic functions as good as possible F (t)
measured or calculated

i sin(2 fi t) + i cos(2 fi t) .
f irst harmonic approximation

(5.37)

The coecients i and i can be calculated from the demand for a minimal overall error
tSi+Ti

1 2
tSi

i sin(2 fi t)+ i cos(2 fi t) F (t)

dt

M inimum .

(5.38)

The dierentiation of (5.38) with respect to i and i delivers two linear equations as necessary conditions
tSi+Ti 2

i sin(2 fi t)+ i cos(2 fi t) F (t)

tSi tSi+Ti 2

sin(2 fi t) dt = 0 (5.39)

i sin(2 fi t)+ i cos(2 fi t) F (t)

tSi

cos(2 fi t) dt = 0

with the solutions i = i = F sin dt cos2 dt F cos dt sin cos dt sin2 dt cos2 dt 2 sin cos dt F cos dt sin2 dt F sin dt sin cos dt sin2 dt cos2 dt 2 sin cos dt , (5.40)

83

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

where the integral limits and arguments of sine and cosine have no longer been written. Because it is integrated exactly over one period tSi t tSi + Ti , for the integrals in (5.40) sin cos dt = 0 ; holds, and as solution i = 2 Ti F sin dt , i = 2 Ti F cos dt . (5.42) sin2 dt = Ti ; 2 cos2 dt = Ti 2 (5.41)

remains. These however are exactly the rst two coecients of a FourierApproximation. In practice, the frequency response of a system is not determined punctual, but continuous. For this, the system is excited by a sweep-sine. 5.4.1.2 Sweep-Sine Excitation

In analogy to the simple sine-function xe (t) = A sin(2 f t) , where the period duration T = 1/f appears as pre-factor at dierentiation x e (t) = A 2 f cos(2 f t) = 2 A cos(2 f t) , T (5.44) (5.43)

now a generalized sine-function can be constructed. Starting with xe (t) = A sin(2 h(t)) the time derivative results in (t) cos(2 h(t)) . x e (t) = A 2 h Now we demand, that the function h(t) delivers a period, that fades linear in time, i.e: (t) = h 1 1 = , T (t) pqt (5.47) (5.46) (5.45)

where p > 0 and q > 0 are constants yet to determine. From (5.47) h(t) = 1 ln(p q t) + C q (5.48)

follows. The initial condition h(t = 0) = 0 xes the integration constant C = 1 ln p . q (5.49)

84

FH Regensburg, University of Applied Sciences Inserting (5.49) in (5.48), a sine-like function follows from (5.45) xe (t) = A sin delivering linear fading period durations. The important zero values for determining the period duration lie at 1 p ln = 0, 1, 2, q p q tn and tn = or 2 p ln , q pqt

Prof. Dr.-Ing. G. Rill

(5.50)

p = en q , mit n = 0, 1, 2, p q tn

(5.51)

p (1 en q ) , n = 0, 1, 2, . q p (1 e(n+1) q 1+ en q ) q

(5.52)

The time dierence between two zero points determines the period Tn = tn+1 tn = Tn =

p n q e (1 eq ) q

, n = 0, 1, 2, .

(5.53)

For the rst (n = 0) and last (n = N ) period one nds T0 = TN p (1 eq ) q . p q N q N q = (1 e ) e = T0 e q

(5.54)

With the frequency range to investigate given by the initial f0 and nal fE frequency, the parameters q and the relation q/p can be calculated from (5.54) 1 fE q = ln , N f0 q fE = f0 1 p f0
1 N

(5.55)

with N xing the number of frequency intervals. The passing of the whole frequency range then takes 1 e(N +1) q tN +1 = (5.56) q/p seconds.

5.4.2
5.4.2.1

Hydro-Mount
Principle and Model

For 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.

85

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

xe main spring chamber 1 membrane ring channel chamber 2 cT __ 2 cF uF MF dF __ 2 dF __ 2 c __ T 2

Figure 5.9: Hydro-Mount Fig. 5.9 shows the principle and mathematical model of a hydro-mount. At small deformations 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 relation of the chamber cross section to ring channel cross section is very large. Thus the uid is moved through the ring channel at very high speed. From this remarkable inertia and resistance forces (damping forces) result. 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 labelling the displacement of the generalized uid mass MF , FH = cT xe + FF (xe uF ) (5.57)

holds, where the force eect of the main spring has been approximated by a linear spring with the constant cT . With MF R as actual mass in the ring channel and the cross sections AK , AR of chamber and ring channel the generalized uid mass is given by MF = AK AR
2

MF R .

(5.58)

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 (xe uF ) + sF (xe uF ) < sF F 0 for |xe uf | sF FF (xe uF ) = (5.59) c (x u ) s (x u ) > +s
F e F F e f F

86

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

The hard transition from clearance FF = 0 and uid compression, resp. chamber deformation with FF = 0 is not realistic and leads to problems, even with the numeric solution. The function (5.59) is therefore smoothed by a parable in the range |xe uf | 2 sF . The motions of the uid mass cause friction losses in the ring channel, which are, at rst approximation, proportional to the speed, FD = dF u F . The equation of motion for the uid mass then reads as MF u F = FF FD . (5.61) (5.60)

The membrane clearing makes (5.61) nonlinear, and only solvable by numerical integration. The nonlinearity also aects the overall force (5.57) in the hydro-mount. 5.4.2.2 Dynamic Force Characteristics

The dynamic stiness and the dissipation angle of a hydro bearing are displayed in Fig. 5.10 over the frequency. The dissipation angle is a measurement for the damping.
400 300 200 100 0 60 50 40 30 20 10 0 10
0

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

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

Excitation Frequency [Hz]

10

Figure 5.10: Dynamic Stiness [N/mm] and Dissipation Angle [deg] for a Hydro-Mount The simulation is based on the following system parameters

87

Vehicle Dynamics mF cT dF cF sF = 25 kg = 125 000 N/m = 750 N/(m/s) = 100 000 N/m = 0.0002 mm

FH Regensburg, University of Applied Sciences generalized uid mass stiness of main spring damping constant uid stiness clearance in membrane bearing

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

5.5
5.5.1

Different Inuences on Comfort and Safety

Vehicle Model

Ford motor company uses the vehicle dynamics program VeDynA (Vehicle Dynamic Analysis) for comfort calculations. The theoretical basics of the program modelling, generating the equations of motion, and numeric solution have been published in the book G.Rill: Simulation von Kraftfahrzeugen, Vieweg 1994 Through program extensions, adaption to dierent operating systems, installation of interfaces to other programs and a menu-controlled in- and output, VeDynA has been subsequently developed to marketability by the company TESIS GmbH in Munich.
Ford At the tire model tmeasy(tire model easy to use), as integrated in VeDynA, the tire forces are calculated dynamically with respect to the tire deformation. For every tire a contact calculation is made. The local inclination of the track is determined from three track points. From the statistic characteristics of a track, spectral density and waviness, two-dimensional, irregular tracks calculated. Time =are 0.000000

Z Z

X X

Y Y

Figure 5.11: Car Model

/export/ford/dffa089/u/tseiber1/vedyna/work/results/mview.mvw 07/02/98 AA/FFA

Thilo Seibert Ext. 37598 Vehicle Dynamics, Ford Research Center Aachen

88

FH Regensburg, University of Applied Sciences The

Prof. Dr.-Ing. G. Rill

vehicle model is specially distinguished by the following details: nonlinear elastic kinematics of the wheel suspensions, friction-aected and elastically suspended dampers, fully elastic motor suspension by static and dynamic force elements (rubber elements and/or hydro-mounts, integrated passenger-seat models. Beyond this, interfaces to external tire- and force element models are provided. A specially developed integration procedure allows real-time simulation on a PC.

5.5.2

Simulation Results

The vehicle, a Ford Mondeo, occupied by two persons, drives with v = 80 km/h over a country road. The thereby occurring accelerations at the drivers seat rail and the wheel load variations are displayed in Fig. 5.12. The peak values of the accelerations and the maximal wheel load variations are arranged in the tables 5.7 and 5.8 for the standard car and several modications. acceleration standard friction seat model engine mounts comfort bushings x min x max y min y max z min z max -0.7192 -0.7133 +0.6543 +0.6100 -1.4199 -1.2873 +1.3991 +1.2529 -4.1864 +3.0623 -3.9986 2.7769 -0.7403 +0.6695 -1.4344 +1.3247 -4.1788 +3.1176 -0.5086 +0.5092 -0.7331 +0.8721 -3.6950 +2.8114 -0.7328 +0.6886 -1.5660 +1.2564 -4.2593 +3.1449

Table 5.7: Peak Acceleration Values

Fz front left front right rear left rear right

standard friction seat model engine mounts comfort bushings 2.3830 2.4208 2.1450 2.3355 2.4507 2.3856 2.2616 2.2726 2.4124 2.4436 2.1600 2.3730 2.3891 2.3891 2.1113 2.2997 Fz = Fzmax Fzmin 2.2394 2.4148 2.1018 2.1608

Table 5.8: Wheel Load Variations

It can be seen, that the damper friction, the passengers, the engine suspension and the compliance of the wheel suspensions, (here:represented by comfort bushings) inuence especially the accelerations and with this the driving comfort. At ne tuning thus all these inuences must be respected.

89

Vehicle Dynamics
road profil [m]

FH Regensburg, University of Applied Sciences

body longitudinal acceleration [m/s 2]

0.1 0.05 0 -0.05 -0.1

500

[m]

1000

-5

500

[m]

1000

body vertical acceleration [m/s 2 ]

body lateral acceleration [m/s 2 ]

-5

500

[m]

1000

-5

500

[m]

1000

6 5 4 3 2 1 0 0

wheel load front left [kN]

6 5 4 3 2 1

wheel load front right [kN]

500

[m]

1000

500

[m]

1000

6 5 4 3 2 1 0 0

wheel load rear left [kN]

6 5 4 3 2 1

wheel load rear right [kN]

500

[m]

1000

500

[m]

1000

Figure 5.12: Road Prole, Accelerations and Wheel Loads

90

6 Driving Behavior of Single Vehicles

6.1
6.1.1

Standard Driving Maneuvers

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 and, with this, because 2 of ay = v , the lateral acceleration, until reaching the limit. Typical results are displayed in R Fig. 6.1.
80 60 steer angle [deg] 40 20 0 4 3 2 1 0 4 2 0 -2 -4 6 5 wheel loads [kN] 0 0.2 0.4 0.6 0.8 lateral acceleration [g] roll angle [deg] 4 3 2 1 0 0 0.2 0.4 0.6 0.8 lateral acceleration [g]

Figure 6.1: Steady State Cornering: Rear-Wheel-Driven Car on R = 100 m

side slip angle [deg]

91

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

The vehicle is under-steering and thus stable. The inclination in the diagram steering angle over lateral velocity decides, according to (??) with (??), about the steering tendency and stability behavior. 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 a anti roll bar at the front axle allows to increase the wheel load dierence there and to decrease it at the rear axle accordingly. The digressive inuence of the wheel load on the tire properties, cornering stiness and maximally possible lateral force is thus stressed more strongly at the front axle and the vehicle becomes more under-steering and stable at increasing lateral acceleration, until, in the limit situation, it drifts out of the curve over the front axle. Problems occur at front driven vehicles, because, due to the 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.

6.1.2

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. In Fig. 6.2 a real steering angle gradient is displayed.
40 steering angle [deg] 30 20 10 0

0.2

0.4 0.6 time [s]

0.8

Figure 6.2: Step Steer Input Not the angle at the steering wheel is the decisive factor for the driving behavior, but the steer angle at the wheels, which can dier from the steering wheel angle because of elasticities,

92

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

friction inuences and a servo-support. At very fast steering movements also the dynamic raise of 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.
0.6 0.5 lateral acceleration [g] yaw velocity [deg/s] side slip angle [deg] 0 2 4 0.4 0.3 0.2 0.1 0 3 2.5 roll angle [deg] 2 1.5 1 0.5 0 12 10 8 6 4 2 0 1 0.5 0 -0.5 -1 -1.5 -2 0 2 4

[t]

Figure 6.3: Step Steer: Passenger Car at v = 100 km/h The vehicle behaves dynamically very well, Fig. 6.3. Almost no exceeds at roll angle and lateral acceleration. Small exceeds at yaw velocity and sideslip angle.

6.1.3
6.1.3.1

Driving Straight Ahead

Random Road Prole

The irregularities of a track are of stochastic nature. Fig. 6.4 shows a country road prole in dierent scalings. To limit the eort at 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 ) (6.1)

93

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

0.05 0.04 0.03 0.02 0.01 0 -0.01 -0.02 -0.03 -0.04 -0.05 0

10

20

30

40

50

60

70

80

90

100

Figure 6.4: Track Irregularities or one uses an isotropic track. At an isotropic track the statistic properties are directionindependent. Then a two-dimensional track with its stochastic properties can be described by a single random process z = z (x, y ) z = z (s) ; (6.2) A normally distributed, stationary and ergodic random process z = z (s) is completely characterized by the rst two expectation values, mean value
s

1 mz = lim s 2s
s

z (s) ds

(6.3)

and correlating function

s

1 Rzz ( ) = lim s 2s
s

z (s) z (s ) ds

(6.4)

. A vanishing mean value mz = 0 can always be achieved by an appropriate coordinate transformation. The correlation function is symmetric, Rzz ( ) = Rzz ( ) and 1 Rzz (0) = lim s 2s
s s

(6.5)

z (s)

ds

(6.6)

describes the squared average of zs . Stochastic track irregularities are mostly described by power spectral densities (abbreviated by psd). Correlating function and the one-sided power spectral density are linked by the Fouriertransformation Rzz ( ) =
0

Szz () cos( ) d

(6.7)

94

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

where denotes the space circular frequency. With (6.7) follows from (6.6)

Rzz (0) =
0

Szz () d .

(6.8)

The psd thus gives information, how the square average is compiled from the single frequency shares. The power spectral densities of real tracks can be approximated by the relation1 Szz () = S0 0
w

(6.9)

Where the reference frequency is xed to 0 = 1 m1 . The reference psd S0 = Szz (0 ) 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 reference-psd and waviness lie within the range 1 106 m3 S0 100 106 m3 6.1.3.2 Steering Activity and

A straightforward drive upon an uneven track makes continuous steering corrections necessary. The histograms of the steering angle at a driving speed of v = 90km/h are displayed in Fig. 6.5.
-6 3 -5 3

highway: S 0=1*10 1000

m ; w=2 1000

country road: S0=2*10

m ; w=2

500

500

-2

[deg] 2

-2

[deg] 2

Figure 6.5: Steering Activity on dierent Roads The track quality is reected in the amount of steering actions. The steering activity is often used to judge a vehicle in practice.
1

cf.: M. Mitschke: Dynamik der Kraftfahrzeuge (Band B), Springer-Verlag, Berlin 1984, S. 29.

95

Vehicle Dynamics

FH Regensburg, University of Applied Sciences

6.2
6.2.1

Coach with different Loading Conditions

Data

At trucks and coaches the dierence between empty and laden is sometimes very large. In the table 6.1 all relevant data of a travel coach in fully laden and empty condition are arrayed. inertias [kg m2 ] 12 500 0 0 0 155 000 0 0 0 155 000 15 400 0 250 0 200 550 0 250 0 202 160

vehicle empty

mass [kg ] 12 500

center of gravity [m] 3.800 | 0.000 | 1.500

fully laden

18 000

3.860 | 0.000 | 1.600

Table 6.1: Data for a Laden and Empty Coach

The coach has a wheel base of a = 6.25 m. The front axle with the track width sv = 2.046 m has a double wishbone single wheel suspension. The twin-tire rear axle with the track widths i so h = 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.

6.2.2

Roll Steer Behavior

suspension travel [cm] 10 5 0 -5 -10 -1 0 steer angle [deg] 1

Figure 6.6: 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 steer eect, Fig. 6.6.

96

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

6.2.3

Steady State Cornering

Fig. 6.7 shows the results of a steady state cornering on a 100 m-Radius. The fully occupied vehicle is slightly more understeering than the empty one. The higher wheel loads cause greater tire aligning torques and increase the digressive wheel load inuence on the increase of the lateral forces. Additionally roll steering at the rear axle occurs.
steer angle 250 200 [m] 150 100 50 0 0.1 0.2 0.3 0.4 lateral acceleration a y [g] wheel loads [kN] 100
LW

[deg] 200 150 100 50 0 -100

vehicle course

0 [m]

100

100

wheel loads [kN]

50

50

0.1 0.2 0.3 0.4 lateral acceleration a y [g]

0.1 0.2 0.3 0.4 lateral acceleration a y [g]

Figure 6.7: Steady State Cornering: Coach - - empty, fully occupied In the limit range both vehicles can not be kept on the given radius. Due to the high position 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 occurs already at ay 0.35 g .

6.2.4

Step Steer Input

The results of a step steer input at the driving speed of v = 80 km/h can be seen in Fig. 6.8. To achieve comparable acceleration values in steady state condition, the step steer input was done at the empty vehicle with = 90 Grad and at the fully occupied one with = 135 Grad. The steady state roll angle is at the fully occupied bus 50% larger than at the empty one. By the niveau-control the air spring stiness increases with the load. Because the damper eect remains unchange, the fully laden vehicle is not damped as well as the empty one. The results are higher exceeds in the lateral acceleration, the yaw speed and sideslip angle.

97

Vehicle Dynamics
lateral acceleration a y [g] 0.4 0.3 0.2 0.1 0 0 2 4 6

FH Regensburg, University of Applied Sciences

yaw velocity 10 8 6 4 2 8 0 0 2 4 6 8

Z [deg/s]

8 6 4 2 0 0

roll angle

[deg]
2 1 0 -1 -2

side slip angle

[deg]

[s] 6

[s] 6

Figure 6.8: Step Steer: - - Coach empty, Coach fully occupied

6.3

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. The semi-trailing axle realized in serial production represents, according to the roll support, Fig. 6.9, a compromise between the trailing arm and the single wishbone.
10 vertical motion [cm] 5 0 -5 -10 -5 0 lateral motion [cm] 5

Figure 6.9: Rear Axle Kinematics: Semi-Trailing Arm, - - Single Wishbone, Trailing Arm The inuences on the driving behavior at steady state cornering on a 100 m radius are shown in Fig. 6.10.

98

FH Regensburg, University of Applied Sciences

Prof. Dr.-Ing. G. Rill

steer angle 100

LW

[deg] 5 4 3 2 1 0

roll angle

[Grad]

50

0.2

0.4

0.6

0.8

0.2

0.4

0.6

0.8

6 4 2 0 0

wheel loads front [kN]

6 4 2

wheel loads rear [kN]

0.2 0.4 0.6 0.8 lateral acceleration a y [g]

0.2 0.4 0.6 0.8 lateral acceleration a y [g]

Figure 6.10: Steady State Cornering, Semi-Trailing Arm, - - Single Wishbone, Trailing Arm Substituting the semi-trailing arm at the standard car by a single wishbone, one gets, without adaption of the other system parameters, a vehicle, which oversteers in the limit range. The single wishbone causes, compared to the semi-trailing arm a notably higher roll support. This increases the wheel load dierence at the rear axle, Fig. 6.10. Because the wheel load dierence is simultaneously reduced at the front axle, the understeer tendency is reduced. In the limit range, this even leads to oversteer behavior. 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.

99

Index
Ackermann Geometry, 53 Ackermann Steering Angle, 53, 67 Anti Roll Bar, 52 Anti-Lock-Systems, 39 Axle Kinematics, 46 Double Wishbone, 12 McPherson, 12 Multi-Link, 12 Axle Load, 34 Axle Suspension Rigid Axle, 5 Twist Beam, 5 Bend Angle, 59 Brake Pitch Pole, 46 Braking, 45 Example, 36 Maximum Deceleration, 35, 36 Optimal Force Distribution, 37 Pitch Angle, 40 Unbraked Front Axle, 36 Unbraked Rear Axle, 35 Camber Angle, 11, 16 Camber Compensation, 50, 51 Caster Angle, 13 Caster Oset, 14 Comfort, 71 Contact Geometry, 15 Contact Point, 17 Contact Point Velocity, 18 Cornering Stiness, 67 Curve Radius, 53 Damper Characteristic, 77 Drag Link, 6, 7 Driver, 2 Driving, 45 Example, 36 Maximum Acceleration, 35 Optimal Force Distribution, 37 Pitch Angle, 40 Driving Comfort, 75 Driving Safety, 71 Dynamic Force Elements, 83 Eigenvalues, 63, 73 Environment, 3 First Harmonic Oscillation, 83 FourierApproximation, 84 Free Vibrations, 74 Frequency Domain, 83 Front Wheel Drive, 35 Generalized Fluid Mass, 86 Hydro-Mount, 85 Kingpin, 12 Kingpin Angle, 13 Kingpin Inclination, 13 Kingpin Oset, 14 Lateral Acceleration, 50, 67 Lateral Force, 60 Lateral Slip, 60, 61 Load, 3 Maximum Acceleration, 34 Maximum Deceleration, 34 Optimal Damper, 79 Optimal Damping, 74, 76 Optimal Parameter, 79 Optimal Spring, 79 Oversteer, 67 Overturning Limit, 47 Parallel Tracks, 93 Pinion, 6 Power Spectral Density, 94 Preload, 72 Quality Criteria, 79 Quarter Car Model, 76 Rack, 6 Random Road Prole, 77, 93 Rear Wheel Drive, 35 Road, 15 Road Normal, 16 Roll Axis, 51 Roll Center, 51 Roll Steer, 96 Roll Stiness, 48 Roll Support, 50, 51 Rolling Condition, 60

Vehicle Dynamics
Safety, 71 Side Slip Angle, 54 Spring Characteristic, 77 Spring Rate, 73 Stability, 63 State Equation, 63 Steady State Cornering, 47, 91, 97 Steer Box, 6, 7 Steer Lever, 7 Steering Activity, 95 Steering Angle, 57 Steering System Drag Link Steering, 7 Lever Arm, 6 Rack and Pinion, 6 Steering Tendency, 67 Step Steer Input, 92, 97 Suspension Model, 71 Suspension Spring Rate, 73 Sweep-Sine, 84 System Response, 83 Tilting Condition, 34 Tire Bore Slip, 31 Bore Torque, 9, 30, 31 Camber Angle, 16 Camber Inuence, 29 Camber Slip, 29 Characteristic Data, 26, 28 Characteristics, 31 Circumferential Direction, 16 Contact Area, 9 Contact Forces, 9 Contact Length, 21 Contact Point, 15 Contact Torques, 9 Deection, 16, 21 Deformation Velocity, 18 Dynamic Oset, 23 Dynamic Radius, 10, 11 Generalized Force, 24 Generalized Slip, 24 Lateral Direction, 16 Lateral Force, 9, 25 Lateral Force Distribution, 22 Lateral Slip, 22 Lateral Velocity, 18 Linear Model, 60 Loaded Radius, 11, 16 Longitudinal Force, 9, 20, 21, 25 Longitudinal Force Characteristics, 21 Longitudinal Force Distribution, 21 Longitudinal Slip, 21 Longitudinal Velocity, 18 Normal Force, 9

FH Regensburg, University of Applied Sciences

Pneumatic Trail, 23 Radial Damping, 19 Radial Direction, 16 Radial Stiness, 19, 49 Rolling Resistance, 9 Self Aligning Torque, 9, 23, 27 Sliding Velocity, 22 Static Radius, 11, 16 Tilting Torque, 9 Transport Velocity, 11 Tread Deection, 20 Tread Particles, 19 Undeformed Radius, 11 Vertical Force, 19 Wheel Load Inuence, 26 Tire Model Kinematic, 53 Linear, 68 TMeasy, 31 Toe Angle, 11 Track, 72 Track Curvature, 57 Track Normal, 17 Track Radius, 57 Track Width, 49, 54 Trailer, 54, 58 Turning Center, 53 Understeer, 67 Vehicle, 2 Vehicle Comfort, 71 Vehicle Data, 78 Vehicle Dynamics, 1 Vehicle Model, 33, 40, 50, 54, 60, 71, 76, 88 Virtual Work, 51 Waviness, 95 Wheel Base, 53 Wheel Load, 9 Wheel Suspension Central Control Arm, 5 Double Wishbone, 4 McPherson, 4 Multi-Link, 4 Semi-Trailing Arm, 5, 98 Single Wishbone, 98 SLA, 5 Trailing Arm, 98 Yaw Angle, 58 Yaw Velocity, 61

ii