Documente Academic
Documente Profesional
Documente Cultură
LECTURE NOTES
Prof. Dr. Georg Rill
October 2005
download: http://homepages.fh-regensburg.de/%7Erig39165/
Contents
Contents 1 Introduction
1.1 1.2 Literature 1.2.1 1.2.2 1.2.3 1.2.4 1.2.5 1.3 1.3.1 1.3.2 1.3.3 1.3.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vehicle Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vehicle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
I 1
1 2 2 2 3 4 4 4 4 5 6 7 7 8 9
Denitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference frames Toe and camber angle Steering Geometry 1.3.4.1 1.3.4.2 1.3.4.3
. . . . . . . . . . . . . . . . . . . . . . . . . . .
Kingpin . . . . . . . . . . . . . . . . . . . . . . . . . . . . Caster and Kingpin Angle . . . . . . . . . . . . . . . . . . Caster, Steering Oset and Disturbing Force Lever . . . .
Road
2.1 2.2 Modeling Aspects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deterministic Proles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 2.2.2 2.3 2.3.1 2.3.2 2.3.3 Bumps and Potholes Sine Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
10 11 11 12 12 12 15 16 16 17 18
Random Proles
Statistical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . Classication of Random Road Proles . . . . . . . . . . . . . . . . Realizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3.1 2.3.3.2 2.3.3.3 Sinusoidal Approximation Shaping Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Two-Dimensional Model . . . . . . . . . . . . . . . . . . .
Tire
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 3.1.2 3.1.3 3.1.4 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.2.7 3.3 3.3.1 3.3.2 3.3.3 3.4 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5 3.4.6 3.4.7 3.4.8 3.4.9 Tire Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tire Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tire Forces and Torques
19
19 19 19 20 21 23 23 25 26 28 29 31 32 34 34 35 36 37 37 40 41 43 45 46 48 50 52
Measuring Tire Forces and Torques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basic Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Local Track Plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tire Deection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Length of Contact Patch . . . . . . . . . . . . . . . . . . . . . . . . Static Contact Point Contact Point Velocity . . . . . . . . . . . . . . . . . . . . . . . . . Dynamic Rolling Radius
Contact Geometry
Forces and Torques caused by Pressure Distribution . . . . . . . . . . . . . Wheel Load . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tipping Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rolling Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Longitudinal Force and Longitudinal Slip . . . . . . . . . . . . . . . Lateral Slip, Lateral Force and Self Aligning Torque . . . . . . . . . Wheel Load Inuence . . . . . . . . . . . . . . . . . . . . . . . . . . Two-Dimensional Tire Characteristics . . . . . . . . . . . . . . . . . Dierent Friction Coecients Self Aligning Torque . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Suspension System
4.1 4.2 Purpose and Components 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 4.4 4.4.1 4.4.2 4.4.3 . . . . . . . . . . . . . . . . . . . . . . . . . . . Some Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Multi Purpose Systems . . . . . . . . . . . . . . . . . . . . . . . . . Specic Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Requirements
54
54 55 55 56 56 56 57 57 58 58 59 59 60 62
Steering Systems
Drag Link Steering System . . . . . . . . . . . . . . . . . . . . . . . Bus Steer System . . . . . . . . . . . . . . . . . . . . . . . . . . . . Springs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Damper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rubber Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . .
II
4.5
Dynamic Force Elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5.1 4.5.2 4.5.3 Testing and Evaluating Procedures Simple Spring Damper Combination 4.5.3.1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
63 63 66 68 69
Vertical Dynamics
5.1 5.2 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Modelling Aspects 5.2.1 5.2.2 5.2.3 5.3 5.3.1 5.3.2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Full Vehicle Model
72
72 72 72 73 75 76 76 78 78 79 81 81 81 82 87 87 88 90 90 90 92
Twodimensional Models
Simple Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Natural Frequency and Damping Rate Spring Rates 5.3.2.1 5.3.2.2
Basic Tuning
5.3.3 5.3.4
Inuence of Damping . . . . . . . . . . . . . . . . . . . . . . . . . . Optimal Damping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.4.1 5.3.4.2 Avoiding Overshoots . . . . . . . . . . . . . . . . . . . . . Fast Approach to Steady State . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.4
Modelling Aspects
Eigenfrequencies and Damping Ratios . . . . . . . . . . . . . . . . . Technical Realization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Quarter Car Model . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.5
Longitudinal Dynamics
6.1 Dynamic Wheel Loads 6.1.1 6.1.2 6.1.3 6.2 6.2.1 6.2.2 6.3 6.3.1 6.3.2 6.3.3 6.3.4 6.3.5 6.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Simple Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . Inuence of Grade Aerodynamic Forces
94
94 94 95 96 97 97 97 98 98 99
Single Axle Drive . . . . . . . . . . . . . . . . . . . . . . . . . . . . Braking at Single Axle . . . . . . . . . . . . . . . . . . . . . . . . . Dierent Distributions of Brake Forces Anti-Lock-Systems
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103
III
Vehicle Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Equations of Motion Driving and Braking Brake Pitch Pole . . . . . . . . . . . . . . . . . . . . . . . . . . 105 . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Equilibrium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Lateral Dynamics
7.1 Kinematic Approach 7.1.1 7.1.2 7.1.3 7.1.4 Kinematic Tire Model Ackermann Geometry
109
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 . . . . . . . . . . . . . . . . . . . . . . . . . 109 . . . . . . . . . . . . . . . . . . . . . . . . . 109
Space Requirement . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 Vehicle Model with Trailer . . . . . . . . . . . . . . . . . . . . . . . 112 7.1.4.1 7.1.4.2 7.1.4.3 7.1.4.4 7.1.4.5 Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . 112 . . . . . . . . . . . . . . . . . . . . . . . . 113 Vehicle Motion
7.2
Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 7.2.1 7.2.2 7.2.3 7.2.4 7.2.5 Cornering Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Overturning Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 . . . . . . . . . . . . . . . . . . . . . . . 124 Roll Support and Camber Compensation . . . . . . . . . . . . . . . 121 Roll Center and Roll Axis Wheel Loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
7.3
Simple Handling Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 7.3.1 7.3.2 7.3.3 7.3.4 7.3.5 7.3.6 Modeling Concept Kinematics Tire Forces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Lateral Slips . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 Equations of Motion 7.3.6.1 7.3.6.2 7.3.6.3 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Eigenvalues . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Low Speed Approximation . . . . . . . . . . . . . . . . . . 129 High Speed Approximation . . . . . . . . . . . . . . . . . 129
7.3.7
Steady State Solution . . . . . . . . . . . . . . . . . . . . . . . . . . 130 7.3.7.1 7.3.7.2 7.3.7.3 Side Slip Angle and Yaw Velocity . . . . . . . . . . . . . . 130 Steering Tendency Slip Angles . . . . . . . . . . . . . . . . . . . . . . 132 . . . . . . . . . . . 133 . . . . . . . . . . . . . . . . . . . . . . . . . . 133
7.3.8
136
Standard Driving Maneuvers . . . . . . . . . . . . . . . . . . . . . . . . . . 136 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 136 Step Steer Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 Driving Straight Ahead . . . . . . . . . . . . . . . . . . . . . . . . . 138
IV
. . . . . . . . . . . . . . . . . . . . 138 . . . . . . . . . . . . . . . . . . . 140
Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 . . . . . . . . . . . . . . 143 Steady State Cornering . . . . . . . . . . . . . . . . . . . . . . . . . 141 Step Steer Input
1 Introduction
1.1 Literature
ATZ: Automobiltechnische Zeitschrift Fachbuchgruppe Fahrwerktechnik:
Robert
Bosch
GmbH
(Hrsg.),
23.
Au.,
Vieweg-Verlag.
Proceedings:
( Fd. Internat. des Socits d'Ingnieurs de Techniques de l'Automobile). (International Assosiation for Vehicle System Dynamics).
Vehicle Dynamics
1.2 Terminology
1.2.1 Vehicle Dynamics
The expression `Vehicle Dynamics' encompasses the interaction of:
the improvement of active safety and driving comfort the reduction of road destruction
In the following the interactions between the single systems and the problems with computer calculations and/or measurements shall be discussed.
1.2.2 Driver
By various means the driver can interfere with the vehicle:
driver
lateral dynamics
longitudinal dynamics
vehicle
vehicle
longitudinal, lateral, vertical motor, aerodynamics, tires velocity, external temperature, ...
driver
driver
environment
The driver's reaction is very complex. To achieve objective results, an `ideal' driver is used in computer simulations, and in driving experiments automated drivers (e.g. steering machines) are employed. Transferring results to normal drivers is often dicult, if eld tests are made with test drivers. Field tests with normal drivers have to be evaluated statistically. Of course, the driver's security must have absolute priority in all tests. Driving simulators provide an excellent means of analyzing the behavior of drivers even in limit situations without danger. It has been tried to analyze the interaction between driver and vehicle with complex driver models for some years.
1.2.3 Vehicle
The following vehicles are listed in the ISO 3833 directive:
motorcycles passenger cars busses trucks agricultural tractors passenger cars with trailer truck trailer / semitrailer road trains
For computer calculations these vehicles have to be depicted in mathematically describable substitute systems. The generation of the equations of motion, the numeric solution, as well as the acquisition of data require great expenses. In times of PCs and workstations computing costs hardly matter anymore. At an early stage of development, often only prototypes are available for eld and/or laboratory tests. Results can be falsied by safety devices, e.g. jockey wheels on trucks.
Vehicle Dynamics
1.2.4 Load
Trucks are conceived for taking up load. Thus, their driving behavior changes. mass, inertia, center of gravity dynamic behaviour (liquid load)
Load
In computer calculations problems occur at the determination of the inertias and the modeling of liquid loads. Even the loading and unloading process of experimental vehicles takes some eort. When carrying out experiments with tank trucks, ammable liquids have to be substituted with water. Thus, the results achieved cannot be simply transferred to real loads.
1.2.5 Environment
The environment inuences primarily the vehicle: road: air: irregularities, coecient of friction resistance, cross wind
Environment
vehicle
environment
driver
Through the interactions between vehicle and road, roads can quickly be destroyed. The greatest diculty with eld tests and laboratory experiments is the virtual impossibility of reproducing environmental inuences. The main problems with computer simulation are the description of random road irregularities and the interaction of tires and road as well as the calculation of aerodynamic forces and torques.
1.3 Denitions
1.3.1 Reference frames
A reference frame xed to the vehicle and a ground-xed reference frame are used to describe the overall motions of the vehicle, Figure 1.1. The ground-xed reference frame
x0 , y0 , z0 serves as an inertial reference frame. Within the vehicle-xed reference frame the xF -axis points forward, the yF -axis to the left, and the zF -axis upward.
with the axis The wheel rotates around an axis which is xed to the wheel carrier. The reference frame
xC , yC
and
zC
are parallel to
F.
zF yF xF z0 zC yC en xC eyR y0
x0
The momentary position of the wheel is xed by the wheel center and the orientation of the wheel rim center plane which is dened by the unit vector the wheel rotation axis. Finally, the normal vector
eyR
en
left wheel
yF
right wheel
longitudinal direction and the intersection line of the tire center plane with the track plane is named toe or toe-in angle. It will be positive, if the front part of the wheel is oriented towards the vehicle center plane, Figure 1.2. Toe-in reduces the tendency of the wheels to shimmy. The camber angle
is the angle between the wheel center plane and the local track normal
en .
It will be positive, if the upper part of the wheel is inclined outwards, Figure 1.3. A
Vehicle Dynamics
en
eyR
and
are parallel.
zC = zF
xC = xF 0 0
* 0
yC = yF
eyR
Figure 1.4: Design position of wheel rotation axis Then, for the left wheel we get
eyR,F = eyR,C =
or
(1.1)
eyR,F = eyR,C
where
(1.2)
yF -axis and the projection line of the wheel rotation 0 describes the angle between the yF -axis and the 0 projection line of the wheel rotation axis into the yF - zF -plane, whereas 0 is the angle between the wheel rotation axis eyR and its projection into the xF - yF -plane. Kinematics
is the angle between the axis into the
xF - yF -plane,
the angle
0 .
and compliance test machines usually measure the angle industry mostly uses this angle instead of
0 . en
points into the direction of
On a at and horizontal road where the track normal the vertical axes
correspond with the toe angle and the camber angle 0 . To specify the dierence between 0 and 0 the ratio between the third the angles and and second component of the unit vector deliver
zC = zF
eyR
or
tan 0 =
tan 0 . cos 0 0
and
(1.3)
is hardly
zR B M A xR
yR
B , which determine the kingpin, are both xed to the wheel A is still xed to the wheel body at the standard McPherson wheel suspension, the ball joint B is now xed to the vehicle body. At a multi-link axle the
and kingpin is no longer dened by real joints. Here, as well as with the enhanced McPherson wheel suspension, the kingpin changes its position relative to the wheel body at wheel travel and steering motions.
Vehicle Dynamics
B zR zR
yR xR M A
yR M xR
rotation axis
eS describes the direction of the kingpin axis. Within the vehicle xed reference frame F it can be xed by two angles. The caster angle denotes the angle between the zF -axis and the projection line of eS into the xF -, zF -plane. In a similar way the projection of eS into the yF -, zF -plane delivers the kingpin inclination angle ,
zF
Fig. 1.7.
zF
eS
yF xF
Figure 1.7: Kingpin and caster angle At many axles the kingpin and caster angle can no longer be determined directly. Here, the current rotation axis at steering motions, which can be taken from kinematic calculations will deliver a virtual kingpin. The current values of the caster angle inclination angle
eS
in the
tan =
where
eS,F eS,F
(3)
(1)
and
tan =
eS,F eS,F
(3)
(2)
(1.4)
(1)
(2)
(3)
eS,F
F.
P,
en
ex
and
ey
which
point into the direction of the longitudinal and lateral tire force result from the contact geometry. The axle kinematics denes the kingpin line. In general, the point point
where an
extension oft the kingpin line meets the road surface does not coincide with the contact
P,
Fig. 1.8. As both points are located on the local track plane, for the left wheel
to
can be written as
rSP = c ex + s ey ,
where
positive, if
P.
kingpin line
C d ey en P ex S s c
2 Road
2.1 Modeling Aspects
Sophisticated road models provide the road height at each point
zR
x, y ,
Fig. 2.1.
Road profile
z0 y0
z(x,y) x0
Segments
(x,y)
Friction
Obstacle
Figure 2.1: Sophisticated road model The tire model is then responsible to calculate the local road inclination. By separating the horizontal course description from the vertical layout and the surface properties of the roadway almost arbitrary road layouts are possible. Besides single obstacles or track grooves the irregularities of a road are of stochastic nature. A vehicle driving over a random road prole mainly performs hub, pitch and roll motions. The local inclination of the road prole also induces longitudinal and lateral motions as well as yaw motions. On normal roads the latter motions have less inuence on ride comfort and ride safety. To limit the eort of the stochastic description usually simpler road models are used. If the vehicle drives along a given path its momentary position can be described by the path variable
s = s(t).
10
z x y
z2
z1
zR(x,y) z1(s)
Now, the road heights on the left and right track are provided by two one-dimensional functions
z1 = z1 (s)
and
z2 = z2 (s).
about the local lateral road inclination is available. If this information is not provided by additional functions the impact of a local lateral road inclination to vehicle motions is not taken into account. For basic studies the irregularities at the left and the right track can considered to be approximately the same,
z1 (s) z2 (s).
zR (s) =
z1 (x) = z2 (x)
can be used. Now, the roll excitation of the vehicle is neglected too.
z H B x L y H B x
y L
Figure 2.3: Rectangular cleat and cosine-shaped bump Then, the rectangular cleat is simply dened by
z(x, y) =
H 0
if else
0<x<L
and
1B < y < 1B 2 2
(2.1)
11
Vehicle Dynamics
z(x, y) =
where
1 2
H 1 cos 2 0
x L
if else
0<x<L
and
1 2B < y < 1B 2
(2.2)
H, B
if negative
L denote height, width and length of the obstacle. Potholes are obtained values for the height (H < 0) are used.
and
z1 (s) = A sin ( s) ,
where
z2 (s) = A sin ( s ) ,
(2.3)
s is the path variable, A denotes the amplitude, the wave number, and the angle describes a phase lag between the left and the right track. The special cases = 0 and = represent the in-phase excitation with z1 = z2 and the out of phase excitation with z1 = z2 .
If the vehicle runs with constant velocity vehicle is given by
s = v0 t,
where the
the term
can be written as
s =
2 2 v0 s= v0 t = 2 t = t . L L L f = /(2) = v0 /L.
(2.5)
For most of the vehicles the rigid body vibrations are in between range is covered by waves which satisfy the conditions For a given wavelength, lets say
L = 4 m, the rigid body vibration of a vehicle are excited min if the velocity of the vehicle will be varied from v0 = 0.5 Hz 4 m = 2 m/s = 7.2 km/h max = 15 Hz 4 m = 60 m/s = 216 km/h. Hence, to achieve an excitation in the to v0
whole frequency range with moderate vehicle velocities proles with dierent varying wavelengths are needed.
12
zR 0.15 0.10 [m] 0.05 0 -0.05 -0.10 -0.15 -200 -150 -100 -50 0 Realization Gaussian density function m s Histogram
50
100
[m] 150
200
X/2
zR (s) ds = 0
(2.6)
p(zR ) =
(2.7)
zR = zR (s)
X/2 2 2 = E zR (s)
1 = lim X X
X/2
zR (s)2 ds .
(2.8)
Alteration of
inexion occur at
|z| <
is given by
P () =
z2 2 e 2 dz .
(2.9)
and
P (3) = 0.997.
X/2
zR (s) zR (s+) ds .
(2.10)
13
Vehicle Dynamics
R() = R(),
zR (s) and zR (s+) diminishes. For large values of the two values are practically unrelated. Hence, R( ) will tend to 0. In fact, R() is always less R(0), which 2 coincides with the variance of the process. If a periodic term is present in the process it will show up in R().
Usually, road proles are characterized in the frequency domain. Here, the auto-correlation function
R()
S().
In general,
R()
and
S()
1 S() = 2
where
R() e
and
1 R() = 2
S() ei d ,
(2.11)
i is the imaginary unit, and in rad/m denotes the wave number. To avoid negative () = 2 S() ,
if
and
() = 0 ,
if
<0,
(2.12)
the relationship
ei = cos() i sin(),
R() = R()
and
2 () =
0
R() cos () d
R() =
0
() cos () d .
(2.13)
= R( = 0) =
0
In reality the psd
() d . 1 N ,
(2.14)
()
N (i) 1 i N
Figure 2.5: Power spectral density in a nite interval Eq. (2.14) can be approximated by a sum, which for
2
i=1
(i )
with
N 1 . N
(2.15)
14
() = (0 )
where,
(2.16)
= 2/L
in
rad/m
describes the value of the psd at a the reference wave in magnitude is modeled by the waviness
w.
b) Range of road classes (ISO 8608) 0=256106
Class A 10-1 100 101 Wave number [rad/m] 102 10-2 10-1 100 101 Wave number [rad/m] 102
Figure 2.6: Road power spectral densities: a) Measurements, b) Classication According to the international directive ISO 8608 typical road proles can be grouped into classes from A to E. By setting the waviness to w = 2 each class is simply dened by 6 its reference value 0 . Class A with 0 = 1 10 m2 /(rad/m) characterizes very smooth 6 highways, whereas Class E with 0 = 256 10 m2 /(rad/m) represents rather rough roads, Fig. 2.6b.
15
Vehicle Dynamics
2.3.3 Realizations
2.3.3.1 Sinusoidal Approximation
A random prole of a single track can be approximated by a superposition of sine waves
N
(2.17)
zR (s) =
i=1
where each sine wave is determined by its amplitude dierent sets of uniformly distributed phase angles
i .
By
and
dierent proles can be generated which are similar in the general appearance
but dierent in details. The variance of the sinusoidal representation is then given by
X/2
1 2 = lim X X
X/2
For
Ai sin (i s i )
i=1 j=1
Aj sin (j s j )
ds .
(2.18)
i=j
and for
i=j
i=j
can
be solved immediately
Jii =
A2 sin2 (i si ) ds = i
(2.19)
sin x sin y =
the integrals for
(2.20)
i=j
Jij = =
ij = ij
1.
1 = lim X X
2 N
1 + lim X X
1 = 2
A2 . i
i=1
(2.22)
i=1
A2 i i 2i
16
On the other hand, the variance of a sinusoidal approximation to a random road prole is given by Eq. (2.15). So, a road prole given psd
zR = zR (s) , N
()
Ai =
and the wave numbers
2 (i )
i = 1(1)N ,
equal intervals
(2.23)
[m]
0.10 0.05 0
Figure 2.7: Realization of a country road A realization of the country road with a psd of
0 = 10 106 m2 /(rad/m) is shown in Fig. 2.7. According to Eq. (2.17) the prole z = z(s) was generated by N = 200 sine waves in the frequency range from 1 = 0.0628 rad/m to N = 62.83 rad/m. The amplitudes Ai , i = 1(1)N were calculated by Eq. (2.23) and the MATLAB function rand was used to produce uniformly distributed random phase angles in the range between 0 and 2 .
zR
reads as (2.24)
where
is a constant, and
w .
R = H() W H T () =
where lter. By setting
1 W 1 W = 2 , + i i + 2
(2.25)
H()
country road can be approximated very well, Fig. 2.8. The shape lter approach is also suitable for modeling parallel tracks. Here, the crosscorrelation between the irregularities of the left and right track have to be taken into account too.
17
Vehicle Dynamics
10-3
Measurements Shaping filter
102
zR = zR (x, y)
proximation is very laborious. Because a shaping lter is a dynamic system, the resulting road prole realizations are not reproducible. By adding band-limited white noise processes and taking the momentary position
x, y
z
x
1 0 -1 4
0 -2 -4
10
15
20
25
30
35
40
45
50
Figure 2.9: Two-dimensional road prole By assuming the same statistical properties in longitudinal and lateral direction twodimensional proles, like the one in Fig. 2.9, can be obtained.
18
3 Tire
3.1 Introduction
3.1.1 Tire Development
Some important mile stones in the development of tires are shown in Table 3.1. 1839 1845 1888 1893 1895 Charles Goodyear: vulcanization Robert William Thompson: rst pneumatic tire (several thin inated tubes inside a leather cover) John Boyd Dunlop: patent for bicycle (pneumatic) tires The Dunlop Pneumatic and Tyre Co. GmbH, Hanau, Germany Andr and Edouard Michelin: pneumatic tires for Peugeot Paris-Bordeaux-Paris (720 Miles): 1899 1904 1908 1922 1943 1946 50 tire deations, 22 complete inner tube changes
Continental: long-lived tires (approx. 500 Kilometer) Carbon added: black tires. Frank Seiberling: grooved tires with improved road traction Dunlop: steel cord thread in the tire bead Continental: patent for tubeless tires Radial Tire Table 3.1: Milestones in tire development
Of course the tire development did not stop in 1946, but modern tires are still based on this achievements.
8.5 kg
19
Vehicle Dynamics
Reinforcements: steel, rayon, nylon Rubber: natural/synthetic Compounds: carbon, silica, chalk, ... Softener: oil, resin Vulcanization: sulfur, zinc oxide, ... Miscellaneous
www.felge.de
180 mm
Figure 3.1: Tire footprint of a passenger car at normal loading condition: Continental 205/55 R16 90 H,
2.5
bar,
Fz = 4700 N
The eect of the contact forces can be fully described by a resulting force vector applied at a specic point of the contact patch and a torque vector. The vectors are described in a track-xed reference frame. The to the
z -axis
140 mm
is normal to the track, the
x-axis
is perpendicular
z -axis
eyR .
y -axis.
The components of the contact force vector are named according to the direction of the axes, Figure 3.2. A non symmetric distribution of the forces in the contact patch causes torques around the
and
Tx .
The torque
Ty
includes
the rolling resistance of the tire. In particular, the torque around the in vehicle dynamics. It consists of two parts,
z -axis
is important
Tz = TB + TS .
(3.1)
20
Fx Fy Fz Tx Ty Tz
eyR Fy Tx Fx Fz Tz Ty
tilting torque rolling resistance torque self aligning and bore torque
z -axis
MB .
MS
takes into account that ,in general, the resulting lateral force is not acting in
Test trailer
compensation wheel
test wheel
Figure 3.3: Layout of a tire test trailer surfaces like asphalt or concrete and dierent environmental conditions like dry, wet or icy are possible. Measurements with test trailers are quite cumbersome and in general they are restricted to passenger car tires. Indoor measurements of tire forces and torques can be performed on drums or on a at bed, Figure 3.4.
21
Vehicle Dynamics
tire
tire
perfect contact
rotation drum
tire
On drum test rigs the tire is placed either inside or outside of the drum. In both cases the shape of the contact area between tire and drum is not correct. That is why, one can not rely on the measured self aligning torque. Due its simple and robust design, wide applications including measurements of truck tires are possible. The at bed tire test rig is more sophisticated. Here, the contact patch is as at as on the road. But, the safety walk coating which is attached to the steel bed does not generate the same friction conditions as on a real road surface.
Radial 205/50 R15, FN= 3500 N, dry asphalt
4000 3000
2000 1000 0 -1000 -2000 -3000 -4000 -40 -30 -20 -10 0 10
Driving
Braking
20
30
40
Tire forces and torques are measured in quasi-static operating conditions. Hence, the measurements for increasing and decreasing the sliding conditions usually result in dierent graphs, Figure 3.5. In general, the mean values are taken as steady state results.
22
x0 -, y0 - z0 -system
is given by the
eyR
tire
M e yR en P0 P* x0 road: z = z ( x , y ) y0 0 z0
e yR
rMP ex b P0 a ey P
en
wheel carrier
x0
z0
The irregularities of the track can be described by an arbitrary function of two spatial coordinates
z = z(x, y).
At an uneven track the contact point
(3.2)
rM P = r0 ezB ,
where
(3.3)
r0
ezB
z -direction of the x0 , y0 ,
z0
is determined by
r0P ,0 = r0M,0 + rM P ,0
x = y , z
(3.4)
23
Vehicle Dynamics
r0M
M.
P0
follows from
r0P0 ,0 =
x y z (x , y )
,
(3.5)
where Eq. (3.2) was used to calculate the appropriate road height. In the point track normal
P0
the
en
ex =
where
eyR en | eyR en |
and
ey = en ex ,
(3.6)
ex
demands a normalization, as
eyR denotes the unit vector into the direction of the wheel rotation axis. Calculating eyR not always being perpendicular to the track. The tire = arcsin eT en yR
(3.7)
camber angle
describes the inclination of the wheel rotation axis against the track normal. The vector from the rim center
P0
rM P0 = rS ezR + a ex + b ey ,
where
rS
a, b
ferential and lateral direction, and the radial direction is given by the unit vector
ezR = ex eyR
which is perpendicular to in
(3.9)
ex
and
eyR .
en
results (3.10)
eT rM P0 = rS eT ezR + a eT ex + b eT ey . n n n n
As the unit vectors
ex
and
ey
are perpendicular to
en
eT rM P0 = rS eT ezR . n n
Hence, the static tire radius is given by
rS
The contact point
eT rM P0 = nT . en ezR
(3.12)
rM P = rS ezR
lies within the rim center plane. The transition from the point takes place according to Eq. (3.8) by the terms
(3.13)
P0
a ex
and
b ey
24
normal
en .
P0 .
With an uneven
Tire models which can be simulated within acceptable time assume that the contact patch is suciently at. At an ordinary passenger car tire, the contact area has at normal load approximately the size of
to fractions of millimeters, when later on the real track will be approximated by a plane in the range of centimeters. If the track in the contact area is replaced by a local plane, no further iterative improvements will be necessary for the contact point calculation.
eyr
e = x
(3.14)
r0Q ,0 = r0M,0 + rM Q ,0 i i
x i = yi , zi
i = 1(1)4 .
(3.15)
In order to sample the contact patch as good as possible the tire width tire radius
r0
rM Q = 1 rM Q = 3 rM Q = 4
x r0 ex
rM Q = x r0 ex 2
in the front, in the rear, in the left, and in the right of the contact patch. According to Eq. (3.5) the corresponding points on the track are given by
r0Qi ,0
x i = yi , z (xi , yi )
i = 1(1)4 .
(3.17)
The calculation of the track normal is straight forward now, Figure 3.7. The vectors
25
Vehicle Dynamics
eyR
* Q1
Q1 rQ2Q1
* Q4 Q4
en
rMP* P* P
* Q3 Q3
rQ3Q4 Q2
* Q2
and
(3.18)
ex , ey
r0P0 ,0 =
serves as rst improvement of point
in the rim center plane is obtained by Eqs. (3.12) and (3.13). On rough roads
the point
not always is located on the track. But, together with the local track normal
it represents the local road plane very well. As in reality, sharp bends and discontinuities which will occur at step- or ramp-sized obstacles are smoothed by this approach.
=0
A0 =
where
zb,
(3.20)
z = r0 r S .
Here, the width of the tire simply equals the width of the contact zone,
(3.21)
wC = b.
On a cambered tire the deected zone of the tire cross section depends on the contact situation. The magnitude of the tire ank radii
rSL = rs +
b tan 2
and
rSR = rs
b tan 2
(3.22)
26
=0 / eyR
=0 eyR eyR
rS z
en P
r0
rSL
rS P
b wC
rSR en r0 rS
rSR en P
b* wC
r0
wC = b
full contact
Figure 3.8: Tire deection
partial contact
determines the shape of the deected zone. The tire will be in full contact to the road if
rSL r0
and
rSR r0
A =
(3.23)
z = r0 r S .
Hence, at full contact the tire camber angle But, due to
(3.24)
wC =
b cos
(3.25)
the width of the contact area increases with the tire camber angle. The deected zone will change to a triangular shape if one of the ank radii exceeds the undeected tire radius. Assuming obtained by
rSL > r0
and
rSR < r0
A =
(3.26)
b =
Now, Eq. (3.26) reads as
(3.27)
(3.28)
27
Vehicle Dynamics
A0 = A
results in
b 1 r0 rS + 2 tan z = 2 b tan
where Eq. (3.22) was used to express the ank radius tire width
.
by the static tire radius
(3.29)
rSR
rS ,
the
wC =
(3.30)
where the Eqs. (3.27) and (3.22) where used to simplify the expression. If are replaced by
tan
and
sin
| tan |
and
| sin |
then, the Eqs. (3.29) and (3.30) will hold for positive
zF
and
zB z =
zF +
zB = r0 rS
(3.31)
will hold.
Belt Fz Rim
zF
undeformed belt
rS r0
zB
L/2
r0
zB
zF
zB
1 2
z.
(3.32)
Approximating the belt deection by truncating a circle with the radius of the undeformed tire results in
L 2
+ (r0
2 zB )2 = r0 .
(3.33)
28
zB 8 r0 zB .
In normal driving situations the belt deections are small, can be simplied and nally results in
r0 .
L2 = 2 r0 4
zB
or
L =
(3.34)
Inspecting the passenger car tire footprint in Figure 3.1 leads to a contact patch length of
L 140mm. For this tire the radial stiness and the inated radius are specied with cR = 265 000 N/m and r0 = 316.9 mm. The overall tire deection can be estimated by z = Fz /cR . At the load of Fz = 4700N the deection amounts to z = 4700N / 265 000N/m = 0.0177 m. Then, by approximating the belt deformation by the half of the tire deection, the length of the contact patch will become L = 8 0.3169 m 0.0177/2 m = 0.1498 m = 150 mm which corresponds quite well with the length of the tire footprint.
FZ
to the
Q,
Figure 3.10.
rS en ey r0-rSL A
y wC
Q Fz A r0-rSR
Figure 3.10: Lateral deviation of contact point at full contact The center of the trapezoidal area determines the lateral deviation area into a rectangular and a triangular section we will obtain
yQ .
By splitting the
yQ =
y A +y A . A ey
(3.35)
The minus sign takes into account that for positive camber angles the acting point will move to the right whereas the unit vector dening the lateral direction points to the left. The area of the whole cross section results from
A =
(3.36)
29
Vehicle Dynamics
wC
A =
geometric contact point,
z wC .
(3.37)
As the center of the rectangular section is located on the center line which runs through the
y =0
1 1 1 wC wC = wC . 2 3 6
(3.38)
(3.39)
where Eq. (3.22) was used to simplify the expression. Now, Eq. (3.35) can be written as
yQ =
wC
1 2
(3.40)
If the cambered tire has only a partial contact to the road then, according to the deection area a triangular pressure distribution will be assumed, Figure 3.11.
b/2
en ey P
y wC
Q Fz
Figure 3.11: Lateral deviation of contact point at partial contact Now, the location of the static contact point
is given by
yQ =
where the width of the contact area
1 b wC 3 2 cos
(3.41)
wC
b/(2 cos )
describes the distance from the geometric contact point angles. The static contact point
contact patch. The plus sign holds for positive and the minus sign for negative camber
r0Q = r0P + yQ ey
represents the contact patch much better than the geometric contact point
(3.42)
P.
30
r0Q = r0M + rM Q ,
where
rM Q describes the position of Q relative to the wheel center M . The absolute velocity of the contact v0Q,0 = r0Q,0 = r0M,0 + rM Q,0 ,
where
r0M,0 = v0M,0
rM Q
takes part on all those motions of the wheel carrier which do not contain elements of the wheel rotation and it In addition, it contains the tire deection Hence, its time derivative can be calculated from
z en,0 , en
(3.45)
0R
is the angular velocity of the wheel rim without any component in the direction
describes
z en,0 .
(3.46)
v0Q,0
eT v0P,0 = 0 n,0
As
or
z eT en,0 = 0 . n,0
(3.47)
en,0
is a unit vector,
eT en,0 = 1 n,0
(3.48)
Finally, the components of the contact point velocity in longitudinal and lateral direction are obtained from
(3.49)
(3.50)
eT en,0 = 0 and eT en,0 = 0 were used to simplify the expressions. x,0 y,0
31
Vehicle Dynamics
deflected tire
rigid wheel
r0 r S
rD
vt
x
Figure 3.12: Dynamic rolling radius
, assuming the tread particles stick to the track, the deected x, Figure 3.12. r
as loaded or static tire radius
r0
as unloaded and
r S = r0
r0 sin
and
= x = rS
(3.51)
r0 cos
hold.
(3.52)
If the motion of a tire is compared to the rolling of a rigid wheel, then, its radius have to be chosen so that at an angular rotation of
rD
will
r0 sin
= x = rD
rD =
For
r0 sin . rD = r0 .
(3.54)
At small, yet nite angular rotations the sine-function can be approximated by the rst terms of its Taylor-Expansion. Then, Eq. (3.54) reads as
r D = r0
1 6
= r0
1 2 6
(3.55)
rS = cos r0
= 1
1 2 2
or
2 = 2
rS r0
(3.56)
32
r D = r0
Due to
1 3
1 rD
rS r0
2 1 r0 + rS . 3 3
(3.57) it is
rS = rS (Fz )
called dynamic tire radius. If the tire rotates with the angular
v t = rD
(3.58)
will denote the average velocity at which the tread particles are transported through the contact patch.
10 [mm] 0
-10 -20 0 2 4 6
r -r
D
Fz [kN]
Figure 3.13: Dynamic tire radius In extension to Eq. (3.57), the dynamic tire radius is approximated in the tire model TMeasy by
rD = r0 + (1 )
r0
FzS c0
(3.59)
rS
where the static tire radius rS = r0 r has been approximated by using the linearized S tire deformation r = Fz /c0 . The parameter is modeled as a function of the wheel load Fz
= N + ( 2N N )
where
Fz 1 FzN Fz = FzN
(3.60)
N and 2N Fz = 2FzN .
[N/m], [N/m],
coefficient for dynamic tire radius fz=fz0 [-], coefficient for dynamic tire radius fz=2*fz0 [-],
the approximation of measured tire data can be done very well, Figure 3.13.
33
Vehicle Dynamics
Fz
z Fz = Fz ( z, z) .
(3.61)
Because the tire can only apply pressure forces to the road the normal force is restricted to
Fz 0.
In a rst approximation
Fz
Fz = FzS + FzD .
The static part is described as a nonlinear function of the normal tire deection
FzS = a1
The constants payload
z + a2 ( z)2 .
(3.63)
a1
and
a2
cN =
d FzS d z
and
S N Fz =Fz
c2N =
d FzS d z
.
S N Fz =2Fz
(3.64)
d FzS = a1 + 2 a2 z . d z
From Eq. (3.63) one gets
(3.65)
z =
a1
(3.66)
Because the tire deection is always positive, the minus sign in front of the square root has no physical meaning, and can be omitted therefore. Hence, Eq. (3.65) can be written as
d FzS = a1 + 2 a2 d z
a1 +
a2 + 4a2 FzS . 1
(3.67)
cN
or or
c2 N c2 2N
(3.68)
c2N =
nally leading to
a1 =
2 c2 c2 N 2N
and
a2 =
c2 c2 2N N . N 4 Fz
(3.69)
34
Results for a passenger car and a truck tire are shown in Figure 3.14. The parabolic approximation in Eq. (3.63) ts very well to the measurements. The radial tire stiness of the passenger car tire at the payload of
190 000N/m.
The Payload
Fz = 3 200 N can be specied with c0 = Fz = 35 000 N and the stiness c0 = 1 250 000N/m of a truck
10 8
100 80
Fz [kN]
Fz [kN]
6 4 2 0
60 40 20 0
10
20 30 z [mm]
40
50
20
40 60 z [mm]
80
Measurements, Approximation
FzD = dR
where
z,
(3.70)
dR
is a constant describing the radial tire damping, and the derivative of the tire
deformation
Fz
to the
Mx = Fz y
acting around a longitudinal axis in
P,
Figure 3.15.
Note: Figure 3.15 shows a negative tipping torque. Because a positive camber angle moves the contact point into the negative torque. As long as the cambered tire has full contact to the road the lateral displacement given by Eq. (3.40). Then, Eq. (3.71) reads as
Mx = Fz
b2 tan . 12 z cos
(3.72)
35
Vehicle Dynamics
en ey P Q
y
ey Fz
en P Fz Tx
If the wheel load is approximated by its linearized static part camber angles
Fz cN
and small
||
Mx = cN
where the term
1 b2 = c N b2 , 12 z 12
(3.73)
1 12
c N b2
en ey P
y
Q Fz
Figure 3.16: Cambered tire with partial contact The use of the tipping torque instead of shifting the contact point is limited to those cases where the tire has full or nearly full contact to the road. If the cambered tire has only partly contact to the road, the geometric contact point the contact area whereas the static contact point 3.16. In the following the static contact
represents the contact area more precisely than the geometric contact point
P.
Fz
y -axis.
If the tire rotates tread particles will be stued into the front of the contact area which causes a slight pressure increase, Figure 3.17. Now, the resulting vertical force is applied in front of the contact point and generates the rolling resistance torque
ty = Fz xR sign() ,
(3.74)
36
rotating xR en C
en ex non-rotating C ex
Fz
Fz
where radius
distance
that
ty
The
Fz
fR = v
of the vehicle
xR . r0
(3.75)
The dimensionless rolling resistance coecient slightly increases with the traveling velocity
fR = fR (v) .
Under normal operating conditions, coecient for typical passenger car
(3.76) resistance
20 km/h < v < 200 km/h, the rolling tires is in the range of 0.01 < fR < 0.02.
The rolling resistance hardly inuences the handling properties of a vehicle. But it plays a major part in fuel consumption.
and
vx .
bed is controlled to the loaded tire radius corresponding to the wheel load A tread particle enters at the time
Fz , Figure 3.18.
t = 0 the contact patch. If we assume adhesion between the particle and the track, then the top of the particle will run with the bed velocity vx and the bottom with the average transport velocity vt = rD . Depending on the velocity dierence v = rD vx the tread particle is deected in longitudinal direction u = (rD vx ) t .
(3.77)
T =
L , rD ||
(3.78)
37
Vehicle Dynamics
rD vx
rD vx L
u max
where
T >0
is assured by
||.
The maximum deection occurs when the tread particle leaves the contact patch at the time
L . rD ||
(3.79)
The deected tread particle applies a force to the tire. In a rst approximation we get
t Fx = ct u , x
where
(3.80)
ct x
On normal wheel loads more than one tread particle is in contact with the track, Figure 3.19a. The number
p =
where
L , s+a
(3.81)
a)
b)
t cx * u
t cu * u max
Figure 3.19:
Particles entering the contact patch are undeformed, whereas the ones leaving have the maximum deection. According to Eq. (3.80), this results in a linear force distribution
38
versus the contact length, Figure 3.19b. The resulting force in longitudinal direction for
particles is given by
Fx =
1 t p c umax . 2 x
(3.82)
Fx =
1 L t L cx (rD vx ) . 2 s+a rD ||
(3.83)
A rst approximation of the contact length L was calculated in Eq. (3.34). Approximating the belt deformation by zB 1 Fz /cR results in 2
L2 4 r 0
where
Fz , cR
(3.84)
cR
denotes the radial tire stiness, and nonlinearities and dynamic parts in the tire
Fx = 2
rD v x r0 c t x Fz . s + a cR rD ||
(3.85)
The nondimensional relation between the sliding velocity of the tread particles in lonS gitudinal direction vx = vx rD and the average transport velocity rD || form the longitudinal slip
sx =
The longitudinal force
(vx rD ) . rD || Fz
Fx
sx
Fx = k Fz sx ,
where the constant
r 0 , s , a, c t x
and
cR .
Eq. (3.87) holds only as long as all particles stick to the track. At moderate slip values the particles at the end of the contact patch start sliding, and at high slip values only the parts at the beginning of the contact patch still stick to the road, Figure 3.20.
Figure 3.20: Longitudinal force distribution for dierent slip values versus the longitudinal slip 0 sx can be dened by the parameters initial inclination (driving stiness) dFx , location M M S S sx and magnitude of the maximum Fx , start of full sliding sx and the sliding force Fx , Figure 3.21. The resulting nonlinear function of the longitudinal force
Fx
39
Vehicle Dynamics
Fx
M
adhesion
Fx S 0 Fx dFx
sliding
sM x
sS x
sx
sx ,
S vy sy = , rD ||
where the sliding velocity in lateral direction is given by
(3.88)
S vy = vy
and the lateral component of the contact point velocity
(3.89)
vy
As long as the tread particles stick to the road (small amounts of slip), an almost linear distribution of the forces along the length
slip values the particles at the end of the contact patch start sliding, and at high slip values only the parts at the beginning of the contact patch stick to the road, Figure 3.22.
Fy
sliding
Fy L
Fy n
The nonlinear characteristics of the lateral force versus the lateral slip can be described 0 M M by the initial inclination (cornering stiness) dFy , the location sy and the magnitude Fy
40
sliding
sS , y
S Fy
of the sliding
The distribution of the lateral forces over the contact patch length also denes the point of application of the resulting lateral force. At small slip values this point lies behind the center of the contact patch (contact point P). With increasing slip values it moves forward, sometimes even before the center of the contact patch. At extreme slip values, when practically all particles are sliding, the resulting force is applied at the center of the contact patch. The resulting lateral force
Fy
as a lever
TS = n Fy .
The lateral force
Fy
sy .
Typical plots of these quantities are shown in Figure 3.23. Characteristic parameters
n/L (n/L)0
adhesion adhesion/sliding full sliding full sliding
Fy Fy
M S adhesion
adhesion/ sliding
s0 y
MS
adhesion
sS y
sy
Fy dF0 y
sM y
sS y
sy
s0 y
sS y
sy
Figure 3.23: Typical plot of lateral force, tire oset and self aligning torque
M 0 of the lateral force graph are initial inclination (cornering stiness) dFy , location sy and S S M magnitude of the maximum Fy , begin of full sliding sy , and the sliding force Fy .
The dynamic tire oset has been normalized by the length of the contact patch L. The 0 S initial value (n/L)0 as well as the slip values sy and sy suciently characterize the graph.
41
Vehicle Dynamics
Longitudinal force
Fx
Lateral force
Fy
Fz = 3.0 kN
0 dFx = 70 kN
Fz = 6.0 kN
0 dFx = 220 kN
Fz = 3.0 kN
0 dFy = 72 kN
Fz = 6.0 kN
0 dFy = 130 kN
sM = 0.160 x sS = 0.500 x
sM = 0.120 x sS = 0.500 x
sM = 0.180 y sS = 0.500 y
sM = 0.200 y sS = 0.700 y
S S Fx = 2.65 kN Fx = 5.10 kN
S S Fy = 2.80 kN Fy = 5.30 kN
Table 3.3: Characteristic tire data with degressive wheel load inuence
friction force. With rising imperfection of the pressure distribution over the contact area, the ability to transmit forces of friction between tire and road lessens. In practice, this leads to a degressive inuence of the wheel load on the characteristic curves of longitudinal and lateral forces. In order to respect this fact in a tire model, the N N characteristic data for two nominal wheel loads Fz and 2 Fz are given in Table 3.3. M M 0 0 From this data the initial inclinations dFx , dFy , the maximal forces Fx , Fx and the M S sliding forces Fx , Fy for arbitrary wheel loads Fz are calculated by quadratic functions. For the maximum longitudinal force it reads as
M Fx (Fz ) =
(3.91)
Fx [kN]
Fy [kN]
-0.2 0 0.2 0.4
-2
-2
-4
-4
-6 -0.4
-6
-20
-10
10
20
sx [-]
[deg]
Figure 3.24: Longitudinal and lateral force characteristics: The location of the maxima
sM , sM , x y
42
Fz 1 FzN
(3.92)
With the numeric values from Tab. 3.3 a slight shift of the maxima with an increasing wheel load is also modeled, Figure 3.24.
Fx = Fx (sx )
force depending on the lateral slip Fy = Fy (sy ) can be dened by their characteristic 0 0 M M parameters initial inclination dFx , dFy , location sx , sy and magnitude of the maximum M M S S S S Fx , Fy as well as sliding limit sx , sy and sliding force Fx , Fy , Figure 3.25. During general driving situations, e.g. acceleration or deceleration in curves, longitudinal sx and lateral slip
sy
M
appear simultaneously.
Fx
Fx
Fx
0 dF x
Fy sx sM x sS x Fy
dF 0 FM FS
Fy Fy
F(s)
Fx
sy sS y
dF y sS sy sM s sM y
sx
sx
sy
2
s =
where the slips
sx sx
sy sy
, sx
and
(3.93)
sx
and
sy
sy .
43
Vehicle Dynamics
Similar to the graphs of the longitudinal and lateral forces the graph of the generalized tire 0 M M S S force is dened by the characteristic parameters dF , s , F , s and F . The parameters are calculated from the corresponding values of the longitudinal and lateral force
dF 0 = sM = FM = sS = FS =
sM x cos sx
sM y sin sy
2
,
(3.94)
,
2
sS x cos sx
sS y sin sy
2
where the slip normalization have also to be considered at the initial inclination. The angular functions
cos =
sx /x s s
and
sin =
sy /y s s
(3.95)
grant a smooth transition from the characteristic curve of longitudinal to the curve of lateral forces in the range of = 0 to = 90 .
F = F (s) is now described in intervals by a broken rational function, a cubic S polynomial, and by the sliding force F s sM dF 0 , = M , 0 s sM ; M s s 1 + + dF 0 M 2 F F (s) = (3.96) s sM F M (F M F S ) 2 (3 2 ) , = S , sM < s sS ; s sM FS , s > sS .
The function
0 When dening the curve parameters, one just has to make sure that the condition dF M 2 FM is fullled, because otherwise the function has a turning point in the interval 0 s s sM .
<
Now, the longitudinal and the lateral force follow from the according projections in longitudinal and lateral direction
Fx = F cos
and
Fy = F sin .
(3.97)
Hence, within TMeasy the one-dimensional characteristics are automatically converted to a two-dimensional combination characteristics, Figure 3.26.
44
30 20 10
1 0
-1 -2 -3 -4 2 4
20
|sx | = 1, 2, 4, 6, 10, 15 %;
|| = 1, 2, 4, 6, 10, 14 Fz = 3.2 kN
/
Fz = 35 kN
4000
Fy [N]
L /0
0.2 0.4 0.6 0.8 1.0
sy [-]
0.5
Fz = 3.2 kN
Figure 3.27: Lateral force characteristics for dierent friction coecients If only the coecient of friction is changed a simple but eective adaption of given model data is possible. A reduced or changed friction coecient mainly inuences the maximum force and the sliding force, whereas the initial inclination will remain unchanged. So, by setting
sM
L M L M s , FM F , 0 0
sS
L S L S s , FS F , 0 0
(3.98)
45
Vehicle Dynamics
the essential tire model parameter which are valid for the friction coecient justed to the new friction coecient Figure 3.27. If the road model provides not only the roughness information local friction coecient be simulated.
are ad-
L .
z = fR (x, y)
[z, L ] = fR (x, y)
then, braking on
-split
nN = n/L. It mainly depends on the lateral slip sy . The normalized tire oset starts at sy = 0 with an initial value (n/L)0 . It tends to zero, n/L 0 at large slip values, sy sS . Sometimes the dynamic tire oset overshoots to negative values before it reaches y S 0 zero again. This behavior can be modeled by introducing the parameter sy < sy , Figure
3.28.
n/L (n/L)0
n/L (n/L)0
s0 y
sS y
sy
s0 y
sy
In order to achieve a simple and smooth approximation of the normalized tire oset versus 0 the lateral slip, a linear and a cubic function are overlayed in the rst section sy sy
|sy | s0 y
and
s=
|sy | s0 y
(3.99)
s0 y w= S sy
(3.100)
0 weights the linear and the cubic function according to the values of the parameter sy S 0 S and sy . No overshoot occurs for sy = sy . Here, w = 1 and (1 w) = 0 will produce a
cubic transition from sy = s0 . y
n/L = (n/L)0
to
46
The characteristic curve parameters, which are used for the description of the dynamic tire oset, are at rst approximation not wheel load dependent. Similar to the description of the characteristic curves of longitudinal and lateral force, here also the parameters for single and double pay load are given. The calculation of the parameters of arbitrary wheel loads is done similar to Eq. (3.92) by linear inter- or extrapolation.
Fz = 3.0 kN
Fz = 6.0 kN
10
[deg]
(n/L)0
gets at rst approximation a triangular distribution of lateral forces over the contact area length cf. Figure 3.22. The working point of the resulting force (dynamic tire oset) is then given by
n(Fz 0, sy = 0) =
The value
1 L. 6
(3.101)
n = 1 L can only serve as reference point, for the uneven distribution of pressure 6
in longitudinal direction of the contact area results in a change of the deexion prole and the dynamic tire oset. The self aligning torque in Figure 3.29 has been calculated with the tire parameter from Table 3.3. The degressive inuence of the wheel load on the self aligning torque can be seen here as well. With the parameters for the description of the tire oset it has been assumed that at N the payload Fz = Fz the related tire oset reaches the value of (n/L)0 = 0.17 1/6 at sy = 0. The slip value s0 , at which the tire oset passes the x-axis, has been estimated. y Usually the value is somewhat higher than the position of the lateral force maximum. S With rising wheel load it moves to higher values. The values for sy are estimated too.
47
Vehicle Dynamics
n = sin .
en
rim centre plane
4000 3000
Fy = Fy (s y ): Parameter
eyR
2000 1000 0
ex ey
y()
-1000
rD ||
-2000 -3000
v()
-4000 -0.5
0
and
0.5
Fy ()
at
Fz = 3.2 kN
= 0 , 2 , 4 , 6 , 8
Now, the tread particles in the contact patch possess a lateral velocity which depends on their position
and is provided by
v () = n
L , = sin , 2 L/2
L/2 L/2 .
(3.103)
At the contact point it vanishes whereas at the end of the contact patch it takes on the same value as at the beginning, however, pointing into the opposite direction. Assuming that the tread particles stick to the track, the deection prole is dened by
y () = v () .
The time derivative can be transformed to a space derivative
(3.104)
y () =
where
d y () d d y () = rD || d dt d
(3.105)
rD ||
denotes the average transport velocity. Now, Eq. (3.104) can be written as
d y () rD || = sin d
or
d y () sin L = , d rD || 2 L/2
(3.106)
48
where
L/2
sy
which
s =
Then, Eq. (3.106) simplies to
sin L . rD || 2
(3.107)
d y () = s . d L/2
The shape of the lateral displacement prole is obtained by integration
(3.108)
1 L y = s 2 2
The boundary condition
L/2
+ C.
(3.109)
C.
One gets
y () = s
1 L 2 2
L/2
(3.111)
The lateral displacements of the tread particles caused by a camber slip are compared now with the ones caused by pure lateral slip, Figure 3.31. At a tire with pure lateral
Figure 3.31: Displacement proles of tread particles slip each tread particle in the contact patch possesses the same lateral velocity which results in transformed to the space derivative
yy was dyy /d . Hence, the deection prole is linear, and reads as yy = vy /(rD ||) = sy , where the denition in Eq. (3.88) was used to introduce the lateral slip sy . Then, the average deection of the tread particles under
where according to Eq. (3.105) the time derivative pure lateral slip is given by
dyy /d rD || = vy ,
yy = sy
L . 2
(3.112)
49
Vehicle Dynamics
The average deection of the tread particles under pure camber slip is obtained from
L/2
1 L 1 y = s 2 2 L
L/2
x L/2
1 L d = s . 3 2
(3.113)
s = y
the lateral camber slip
1 s 3 s . y
(3.114)
In normal driving conditions, the camber angle and thus, the lateral camber slip are limited to small values. So, the lateral camber force can be approximated by
0 Fy dFy s . y
If the global inclination
(3.115)
dFy = Fy /sy
0 dFy ,
one
gets the camber inuence on the lateral force as shown in Figure 3.30. The camber angle inuences the distribution of pressure in the lateral direction of the contact patch, and changes the shape of the contact patch from rectangular to trapezoidal. Thus, it is extremely dicult, if not impossible, to quantify the camber inuence with the aid of such simple models. But, it turns out that this approach is quit a good approximation.
0W = 0R + eyR
has a component in direction of the track normal
(3.116)
en
(3.117)
n = eT 0W = 0 . n
Then, a very complicated deection prole of the tread particles in the contact patch occurs. However, by a simple approach the resulting bore torque can be approximated quite good by the parameter of the generalized tire force characteristics. At rst, the complex shape of a tire's contact patch is approximated by a circle, Figure 3.32. By setting
R =
1 2
L B + 2 2
1 (L + B) 4 length L and
of the actual
contact patch. The integration over the whole circle area results in the bore torque
TB =
1 A
F r dA ,
A
(3.119)
50
d F r dr
ex
B
ey
circular approximation
denotes the force transmitted by the patch element dA, and 2 circle. With dA = r d dr and A = R Eq. (3.119) reads as where
R 2
TB
which immediately results in
1 = 2 R
0 0
F r rd dr
(3.120)
TB 1 = 2 R
2 F r r dr 2 = 2 R
F r2 dr .
0
(3.121)
For small slip values the force transmitted in the patch element can be approximated by
F = F (s) dF 0 s
where
dF 0
generalized tire force characteristics. Similar to Eqs. (3.86 and (3.88) we dene
s =
where
r n rD || rD
and
(3.123)
r n
denote the
dynamic tire radius and the angular velocity of the wheel. Now, Eq. (3.121) reads as
TB
2 = 2 R
0
dF 0
r n 2 r dr rD ||
(3.124)
51
Vehicle Dynamics
TB
2 n = 2 dF 0 R rD ||
0
2 1 R n n R4 0 r dr = 2 dF = R dF 0 R rD || 4 2 rD ||
3
(3.125)
where
sB =
R n rD || dF 0
the bore torque
(3.126)
can be considered as bore slip. Via the initial inclination the actual tire properties into account.
TB
takes
The bore torque calculated by Eq. (3.125) is only a rst approximation. At large bore slips S the generalized tire force F is limited to the sliding force F . Then, Eq. (3.121) changes to
R max TB
2 = 2 R
0
F S r2 dr = FS
2 S R3 2 = FS R . F R2 3 3
max TB
(3.127)
depends on the
tire properties and the actual friction value. Now, the bore torque is given by
TB =
1 R n R dF 0 2 rD ||
with
|TB |
2 S F R 3
(3.128)
and
truck tire
-2 -4 -6 -40
sx [%]
Figure 3.33: Longitudinal force:
sx [%]
20
40
Meas.,
TMeasy
The tire model TMeasy which is based on this approach, can be used for passenger car tires as well as for truck tires. It approximates the characteristic curves
Fx = Fx (sx ),
52
40 20
passenger car
6 4 Fy [kN] F [kN] 2 0 1.8 kN 3.2 kN 4.6 kN 6.0 kN
truck
0 10 kN 20 kN 30 kN 40 kN
Mz
-100 -150 0 10
-1000 -1500 0
[o]
20
[o]
10
20
Meas.,
TMeasy
Fy = Fy ()
and 3.34.
and
Mz = Mz ()
Fz ,
Figures 3.33
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.
53
4 Suspension System
4.1 Purpose and Components
The automotive industry uses dierent kinds of wheel/axle suspension systems. Important criteria are costs, space requirements, kinematic properties, and compliance attributes. The main purposes of a vehicle suspension system are
carry the car and its weight, maintain correct wheel alignment, control the vehicles direction of travel, keep the tires in contact with the road, reduce the eect of shock forces.
force elements: coil spring, torsion bar, air spring, leaf spring, anti-roll bar, damper, bushings, hydro-mounts,
tires.
Tires are air springs that support the total weight of the vehicle. The air spring action of the tire is very important to the ride quality and safe handling of the vehicle.
54
Figure 4.1: 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.
leaf springs
links
Figure 4.2: Solid axles guided by leaf springs and links At pickups, trucks, and busses solid axles are used often. They are guided either by leaf springs or by rigid links, Fig. 4.2. Solid axles tend to tramp on rough roads. Leaf-spring-guided solid axle suspension systems are very robust. Dry friction between the leafs leads to locking eects in the suspension. Although the leaf springs provide axle guidance on some solid axle suspension systems, additional links in longitudinal and lateral direction are used. Thus, the typical wind-up eect on braking can be avoided. Solid axles suspended by air springs need at least four links for guidance. In addition to a good driving comfort air springs allow level control too.
55
Vehicle Dynamics
The semi-trailing arm is a simple and cheap design which requires only few space. It is mostly used for driven rear axles. The short-long-arm axle design allows a nearly independent layout of longitudinal and lateral axle motions. It is similar to the central control arm axle suspension, where the trailing arm is completely rigid and hence, only two lateral links are needed. The twist beam axle suspension exhibits either a trailing arm or a semi-trailing arm characteristic. It is used for non driven rear axles only. The twist beam axle provides enough space for spare tire and fuel tank.
56
30
must
be provided at the front wheels of passenger cars. Depending on the wheel base, busses and trucks need maximum steering angles up to 55 at the front wheels. Recently some companies have started investigations on `steer by wire' techniques.
uR S
nk drag li
rack
the drag links transmitted to the wheel rotations the steering linkage.
overall steering ratio depends on the ratio of the steering box and on the kinematics of
l ring stee
ever
drag link 1
drag link 2
1
wheel and wheel body
Figure 4.5: Lever arm steering system
Using a lever arm steering system Fig. 4.5, large steering angles at the wheels are possible. This steering system is used on trucks with large wheel bases and independent wheel
57
Vehicle Dynamics
suspension at the front axle. Here, the steering box can be placed outside of the axle center. Firstly, the rotations of the steering wheel rotation of the steer levers
1 = 1 (L ), 2 = 2 (L ).
g steerin r leve
O
steer box
(90o rotated)
steering link
1 drag link
Figure 4.6: Drag link steering system wheel
S are transformed by the steering box to the rotation of the steering lever arm L = L (S ) and further on to the rotation of the left wheel, 1 = 1 (L ). The drag link transmits the rotation of the left wheel to the right wheel, 2 = 2 (1 ). The steering ratio
is dened by the ratio of the steering box and the kinematics of the steering link. Here, the ratio
2 = 2 (1 )
angles at the front wheels are needed to achieve a good manoeuvrability. That is why, more sophisticated steering systems are needed, Fig. 4.7. The rotations of the steering
S are transformed by the steering box to the rotation of the steering lever arm L = L (S ). The left lever arm is moved via the steering link A = A (L ). This motion
wheel is transferred by a coupling link to the right lever arm. Finally, the left and right wheels are rotated via the drag links,
1 = 1 (A )
and
2 = 2 (A ).
58
steer in
g leve r
L
steering box steering link
drag link
coupl. link
1
wheel and wheel body
Coil spring
FS
u u FS
Leaf spring Air spring
u FS
Torsion bar
FS
Figure 4.8: Vehicle suspension springs
Coil springs, torsion bars, and leaf springs absorb additional load by compressing. Thus, the ride height depends on the loading condition. Air springs are rubber cylinders lled with compressed air. They are becoming more popular on passenger cars, light trucks, and heavy trucks because here the correct vehicle ride height can be maintained regardless of the loading condition by adjusting the air pressure.
59
Vehicle Dynamics
L LF L c
FS
u
FS
0 FS
L0
FS
Figure 4.9: Linear coil spring and general spring characteristics
A linear coil spring may be characterized by its free length Fig. 4.9. The force acting on the spring is then given by
LF
c,
FS = c LF L ,
where
(4.1)
L denotes the actual length of the spring. Mounted in a vehicle suspension the spring L0 < L F .
Now, Eq. (4.1) can be written as
has to support the corresponding chassis weight. Hence, the spring will be compressed to the conguration length
FS = c LF (L0 u)
where
0 = c LF L0 + c u = FS + c u ,
(4.2)
0 FS
FS
u FS = FS (u) .
(4.3)
Now, arbitrary spring characteristics can be approximated by elementary functions, like polynomials, or by tables which are then inter- and extrapolated by linear functions or cubic splines. The complex behavior of leaf springs and air springs can only be approximated by simple nonlinear spring characteristics,
FS = FS (u).
4.4.2 Damper
Dampers are basically oil pumps, Fig. 4.10. As the suspension travels up and down, the hydraulic uid is forced by a piston through tiny holes, called orices. This slows down the suspension movement.
60
FD v
Piston
Rebound Ch.
FD
Piston orifice
Today twin-tube and mono-tube dampers are used in vehicle suspension systems. Dynamic damper models compute the damper force via the uid pressure applied to each side of the piston. The change in uid pressures in the compression and rebound chambers are calculated by applying the conservation of mass. In standard vehicle dynamics applications simple characteristics
FD = FD (v)
are used to describe the damper force
(4.4)
FD
v.
To
obtain this characteristics the damper is excited with a sinusoidal displacement signal
u = u0 sin 2f t. By varying the frequency in several steps from f = f0 to f = fE dierent force displacement curves FD = FD (u) are obtained, Fig. 4.11. By taking the peak values of the damper force at the displacement u = u0 which corresponds with the velocity v = 2f u0 the characteristics FD = FD (v) is generated now. Here, the rebound cycle is
associated with negative damper velocities.
FD = FD(u)
1000
FD = FD(v)
FD [N]
0 f0
Compression
Rebound
fE 0 0.02 0.04 0.06 -1.6 -1.2 -0.8 -0.4 0 0.4 0.8 1.2 1.6
u [m]
v [m/s]
Figure 4.11: Damper characteristics generated from measurements Typical passenger car or truck dampers will have more resistance during its rebound cycle then its compression cycle.
61
Vehicle Dynamics
Topmount
Stop
Subframe mounts
Figure 4.12: Rubber elements in vehicle suspension During suspension travel, the control arm bushings provide a pivot point for the control arm. They also maintain the exact wheel alignment by xing the lateral and vertical location of the control arm pivot points. During suspension travel the rubber portion of the bushing must twist to allow control arm movement. Thus, an additional resistance to suspension movement is generated. Bump and rebound stops limit the suspension travel. The compliance of the topmount avoids the transfer of large shock forces to the chassis. The subframe mounts isolate the suspension system from the chassis and allow elasto-kinematic steering eects of the whole axle. It turns out, that those elastic elements can hardly be described by simple spring and damper characteristics, models are needed.
properties change with the frequency of the motion. Here, more sophisticated dynamic
62
xe (t) = A sin(2 f t) ,
with dierent frequencies
(4.5)
f0 f fE
and amplitudes
Amin A Amax .
Starting at
t = 0,
where
the system will usually be in a steady state condition after several periods
t nT ,
T = 1/f
and
n = 2, 3, . . .
n T t (n + 1)T , F (t)
sin(2 f t) + cos(2 f t) .
measured/ calculated
The coecients
and
(n+1)T
1 2
nT
dt
and
M inimum .
(4.7)
(n+1)T
= =
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 nT t (n + 1)T , T ; 2 cos2 dt = ,
(4.9)
where the integral limits and arguments of sine and cosine no longer have been written. Because it is integrated exactly over one period Eq. (4.9) for the integrals in
sin cos dt = 0 ;
sin2 dt =
T 2
(4.10)
63
Vehicle Dynamics
2 T
F sin dt ,
2 T
F cos dt
(4.11)
remains. However, these are exactly the rst two coecients of a Fourier-"-Approximation.
The rst order harmonic approximation in Eq. (4.6) can now be written as
F (t) = F sin (2 f t + )
where amplitude
(4.12)
are given by
F =
2 + 2
and
tan =
(4.13)
A simple force element consisting of a linear spring with the stiness with the constant
F = A
c2 + (2f d)2
and
tan =
and
d d 2f A = 2f . cA c
(4.15)
c=0
d=0
is characterized by
tan = 0 or = 0, whereas a pure damper response with c = 0 and F = 2f dA and tan or = 90 . Hence, the phase angle which
The dynamic stiness, dened by
F = A c and d = 0 results in
is also called
the dissipation angle can be used to evaluate the damping properties of the force element.
cdyn =
is used to evaluate the stiness of the element.
F A
(4.16)
In practice the frequency response of a system is not determined punctually, but continuously. For this, the system is excited by a sweep-sine. In analogy to the simple sine-function
xe (t) = A sin(2 f t) ,
where the period
(4.17)
T = 1/f
xe (t) = A 2 f cos(2 f t) =
2 A cos(2 f t) . T
(4.18)
(4.19)
64
(4.20)
h(t)
h(t) =
where
(4.21)
p>0
and
q>0
h(t) =
The initial condition
(4.22)
h(t = 0) = 0
C =
1 ln p . q
(4.23)
With Eqs. (4.23) and (4.22) Eq. (4.19) results in a sine-like function
xe (t) = A sin
p 2 ln , q pqt
(4.24)
which is characterized by linear fading periods. The important zero values for determining the period duration lie at
1 p ln = 0, 1, 2, q p q tn
and
or
p = en q , p q tn
mit
n = 0, 1, 2,
(4.25)
tn =
p (1 en q ) , n = 0, 1, 2, . q
(4.26)
The time dierence between two zero values yields the period
(4.27)
(n = 0)
and last
(n = N )
T0 = TN
p (1 eq ) q . p = (1 eq ) eN q = T0 eN q q f0
1 N
(4.28)
With the frequency range to investigate, given by the initial the parameters
fE ,
q/p
1 fE q = ln , N f0
q fE = f0 1 p f0
(4.29)
65
Vehicle Dynamics
with
xing the number of frequency intervals. The passing of the whole frequency range
tN +1 = N = 500
(4.30)
Hence, to test or simulate a force element in the frequency range from with intervals will only take
0.1Hz to f = 100Hz
728 s
12min.
Figure 4.13: Spring and damper in series The displacements of the force element and the spring itself are described by Then, the the forces acting in the spring and damper are given by
and
s.
FS = c s
The force balance spring
and
FD = d (u s) .
(4.31)
FD = FS displacement s
d (u s) = c s
or
d d s = s + u , c c d
and the spring stiness
(4.32)
acts as time
T = d/c. Hence, this force element will responds dynamically to any excitation.
u(t) = u0 sin t
respectively
u = u0 cos t
(4.33)
can be calculated easily. The steady state response will be of the same type as the excitation. Inserting
(4.34)
(4.35)
66
d u0 b = u0 a c 2 2 + (c/d)2
and
d d u0 a = u0 b + u0 c c c . 2 d + (c/d)2 c cos t d
(4.36)
a =
and
b =
(4.37)
FS = c s = c u0
which can be transformed to
+ (c/d)2
2
sin t +
(4.38)
FS = FS sin (t + )
where the force magnitude
(4.39)
FS
FS =
c u0 + (c/d)2
2
c u0 2 + (c/d)2
= arctan
c/d .
(4.40)
cdyn
400
4
300
[N/mm]
200 100 0 100
2 1
[o]
50
3 4
0 0 20 40
60
80
f [Hz]
Figure 4.14: Frequency response of a spring damper combination With increasing frequency the spring damper combination changes from a pure damper performance, cdyn 0 and 90 to a pure spring behavior, cdyn c and 0. The frequency range, where the element provides stiness and damping is controlled by the value for the damping constant
d.
67
Vehicle Dynamics
parallel force elements, Fig. 4.15. The static load is carried by a single spring with
the stiness
c0
F0 = F0 (u).
u d1 c0 s1 c1 c2 s2 cN sN FM F1 d2 FM F2 dN FM FN
Figure 4.15: Dynamic force model Within each force element the spring acts in serial to parallel combination of a damper and a dry friction element. Now, even hysteresis eects and the stress history of the force element can be taken into account. The forces acting in the spring and damper of force element
FSi = ci si
were
and
FDi = di (si u) ,
and
si
As long as the absolute value of the spring force FSi is lower than the maximum friction M force FF the damper friction combination will not move at all
u si = 0
In all other cases the force balance
for
M |FSi | FF .
(4.42)
M FSi = FDi FF
holds. Using Eq. 4.41 the force balance results in
(4.43)
di (si u) = FSi
which can be combined with Eq. 4.42 to
M FF
(4.44)
F + F M Si F di si = di u F F M
Si F
M FSi < FF
for
(4.45)
68
FSi = ci si .
In extension to this linear approach nonlinear springs and dampers may be used. To derive all the parameters an extensive set of static and dynamic measurements is needed.
4.5.3.1 Hydro-Mount
For the elastic suspension of engines in vehicles very often specially developed hydromounts are used. The dynamic nonlinear behavior of these components guarantees a good acoustic decoupling but simultaneously provides sucient damping.
Figure 4.16: Hydro-mount Fig. 4.16 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 ratio of the chamber cross section to the ring channel cross section is very large. Thus the uid is moved through the ring channel at very high speed. This results in remarkable inertia and resistance forces (damping forces). The force eect of a hydro-mount is combined from the elasticity of the main spring and the volume change in chamber 1. With
uF
MF ,
(4.46)
FH = cT xe + FF (xe uF ) cT .
holds, where the force eect of the main spring has been approximated by a linear spring with the constant
69
Vehicle Dynamics
With
MF R
as the actual mass in the ring channel and the cross sections
A K , AR
of
MF =
AK AR
MF R .
(4.47)
The uid in chamber 1 is not being compressed, unless the membrane can evade no longer. With the uid stiness
FF (xe uF ) =
|xe uf |
sF
(4.48)
FF = 0
is not realistic and leads to problems, even with the numeric solution.
The motions of the uid mass cause friction losses in the ring channel, which are as a rst approximation proportional to the speed,
FD = dF uF .
Then, the equation of motion for the uid mass reads as
(4.49)
MF uF = FF FD .
(4.50)
The membrane clearance makes Eq. (4.50) nonlinear and only solvable by numerical integration. The nonlinearity also aects the overall force in the hydro-mount, Eq. (4.46). The dynamic stiness and the dissipation angle of a hydro-mount are displayed in Fig. 4.17 versus the frequency. The simulation is based on the following system parameters
mF cT dF cF sF
= = = = =
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 can be achieved between noise isolation and vibration damping.
70
10
10
Figure 4.17: Dynamic stiness [N/mm] and dissipation angle [deg] for a hydro-mount
71
5 Vertical Dynamics
5.1 Goals
The aim of vertical dynamics is the tuning of body suspension and damping to guarantee good ride comfort, resp. a minimal stress of the load at sucient safety. The stress of the load can be judged fairly well by maximal or integral values of the body accelerations. The wheel load
Fz
Fx
Fy
by the coecient
Fz
on
Fx
and
Fy
Fx
and
Fy
at wheel load variations. Maximal driving safety can therefore be achieved with minimal variations of the wheel load. Small variations of the wheel load also reduce the stress on the track. The comfort of a vehicle is subjectively judged by the driver. In literature dierent 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.
72
Ford
Z Z
X X
Y Y
Thilo Seibert Ext. 37598 Vehicle Dynamics, Ford Research Center Aachen
Much simpler models can be used, however, for basic studies on ride comfort and ride safety. A two-dimensional vehicle model, for instance, suits perfectly with a single track road model, Fig. 5.2. Neglecting longitudinal accelerations, the vehicle chassis only per-
zA2
Figure 5.2: Vehicle model for basic comfort and safety analysis forms hub and pitch motions. Here, the chassis is considered as one rigid body. Then, mass and inertia properties can be represented by three point masses which are located in the chassis center of gravity and on top of the front and the rear axle. The lumped mass model has 4 degrees of freedom. The hub and pitch motion of the chassis are represented by the vertical motions of the chassis in the front
zC1
zA1
and
zA2
provides road irregularities in the space domain, where front and rear axle are given by
the vehicle and measured at the chassis center of gravity. Then, the irregularities at the
zR (s + a1 )
zR (s a2 ) gravity C .
and
a1
and
a2
73
Vehicle Dynamics
M1 + M + M2 = M
(5.1)
and they have to provide the same inertia around an axis located in the chassis center and pointing into the lateral direction
a2 M 1 + a2 M 2 = . 2 1
The correct location of the center of gravity is assured by
(5.2)
a1 M 1 = a2 M 2 .
Now, Eqs. (5.2) and (5.3) yield the main masses
(5.3)
M1 =
and the coupling mass
a1 (a1 +a2 ) M = M
and
M2 =
, a2 (a1 +a2 )
(5.4)
M a1 a2
(5.5)
follows from Eq. (5.1). If the mass and the inertia properties of a real vehicle happen to result in
= M a1 a2 then, = 0, and the vehicle can be represented by two uncoupled the coupling mass vanishes M
two mass systems describing the vertical motion of the axle and the hub motion of the chassis mass on top of each axle. mid size car full size car sports utility vehicle
vehicles properties front axle mass rear axle mass center of gravity chassis mass chassis inertia lumped mass model
commercial vehicle
heavy truck
[kg m2 ] M1 M [kg] M2
74
Depending on the actual mass and inertia properties the vertical dynamics of a vehicle can be investigated by two simple decoupled mass models describing the vibrations of the front and rear axle and the corresponding chassis masses. By using half of the chassis and half of the axle mass we nally end up in quarter car models. The data in Table 5.1 show that for a wide range of passenger cars the coupling mass is smaller than the corresponding chassis masses,
M < M1
and
M < M2 .
mass model or the quarter car model represent a quite good approximation to the lumped mass model. For commercial vehicles and trucks, where the coupling mass has the same magnitude as the corresponding chassis masses, the quarter car model serves for basic studies only.
mi
Mi , i = 1, 2.
can be investigated independently. The quarter car model is now further simplied to two single mass models, Fig. 5.3.
M cS `` ` `
` c `
6C z
`` cS `` dS ``
dS
6R z
m
`` `` c `` T c
6 zW
6R z
Figure 5.3: Simple vertical vehicle models The chassis model neglects the tire deection and the inertia forces of the wheel. For the high frequent wheel motions the chassis can be considered as xed to the inertia frame. The equations of motion for the models read as
M zC + dS zC + cS zC = dS zR + cS zR
and
(5.6)
m zW + dS zW + (cS + cT ) zW = cT zR ,
where
(5.7)
zC
and
zW
label the vertical motions of the corresponding chassis mass and the
wheel mass with respect to the steady state position. The constants
cS , dS
describe the
D FT = cT (zR zW )
(5.8)
75
Vehicle Dynamics
where
cT
zR
In this simple approach the damping eects in the tire are not taken into account.
zR = 0
and
zR = 0.
2 z + 2 0 z + 0 z = 0 ,
(5.9)
where 0 represents the undamped natural frequency, and is a dimensionless parameter called viscous damping ratio. For the chassis and the wheel model the new parameter are given by
Chassis:
z zC , z zW ,
C = W =
d S , 2 cS M dS 2 (cS +cT )m ,
2 2 0 0C =
cS ; M cS +cT . m
(5.10)
Wheel:
2 2 0 0W =
z(t) = z0 et ,
where
(5.11)
z0
and
2 (2 + 2 0 + 0 ) z0 et = 0 .
Non-trivial solutions
(5.12)
z0 = 0
are possible, if
2 2 + 2 0 + 0 = 0
will hold. The roots of the characteristic equation (5.13) depend on the value of
(5.13)
<1 : 1 :
1,2 = 0 pm i 0 1,2 = 0
1 2 , 2 1
(5.14)
Figure 5.4 shows the root locus of the eigenvalues for dierent values of the viscous damping rate
76
=0 =0.2
Im()/0
1.0
0.5
=1.25
-2 -1.5
=1
-1
=1.25 =1.5
-0.5 0 -0.5
Re()/0
-1.0
and
For
are both real and negative. Hence, Eq. (5.11) will produce a complex,
< 1 holds, the eigenvalues 1,2 will become conjugate of 1 . Now, the solution can be written as 1 2 t ,
z(t) = A e0 t sin 0
where
(5.15)
A and are constants which have to be adjusted to given initial conditions z(0) = z0 and z(0) = z0 . The real part Re (1,2 ) = 0 is negative and determines the decay of 1 2 part denes the actual frequency of the solution. The imaginary Im (1,2 ) = 0
the vibration. The actual frequency
= 0
tends to zero,
1 2
(5.16)
0,
1.
In a more general way the relative damping may be judged by the ratio
D =
Re(1,2 ) . | 1,2 |
(5.17)
D =
holds, because the absolute value of the complex eigenvalues is given by
(5.18)
| 1,2 | =
Re(1,2 )2 + Im(1,2 )2 =
( 0 )2 + 0
1 2
= 0 ,
(5.19)
D =
For
+ 0 = . 0
(5.20)
1 the eigenvalues become real and negative. Then, Eq. (5.17) will always produce D = 1. In this case the viscous damping rate is more
77
Vehicle Dynamics
u0 =
Mg . cS u.
(5.21)
At a conventional suspension without niveau regulation a load variation leads to changed spring deections deection follows from
M M+
u0 u0 + u =
Mg . cS
(5.22)
cS
Mg . u
(5.23)
In the standard design of a passenger car the engine is located in the front and the trunk in the rear part of the vehicle. Hence, most of the load is supported by the rear axle suspension. For an example we assume that wheel by
150 kg
spring deection caused by the load should not exceed half of the maximum value. Then, according to Eq. (5.23) a lower bound of the spring rate at the front axle can be estimated by
(5.24)
(5.25)
which is nearly two times the minimum value of the spring rate at the front axle. In order
In Tab. 5.1 the lumped mass chassis model of a full size passenger car is described by
M1 = M2 = 600 kg
and
M = 200.
decoupled two mass models we have to neglect the coupling mass or, in order to achieve
78
(5.26)
According to Eq. 5.10 the undamped natural eigen frequency of the simple chassis model 2 is then given by 0C = cS /M . Hence, for a spring rate of cS = 20000 N/m the undamped natural frequency of the unloaded car amounts to
f0C =
(5.27)
which is a typical value for most of all passenger cars. Due to the small amount of load the undamped natural frequency for the loaded car does not change very much,
f0C =
(5.28)
The corresponding cassis mass over the rear axle is given here by
(5.29)
0 f0C =
and the loaded car
(5.30)
L f0C =
are larger and dier more.
(5.31)
u0
0 FS = FS + c0 u
where value
1+k
u u
, u = 0,
and
(5.32)
0 FS
cS
terizes the intensity of the nonlinearity. The linear characteristic provides u the lin 0 FS ( u) = FS + cS u. To achieve the same value with the nonlinear spring
k>0 at u =
charac-
0 FS + c0
must hold, where
0 u (1 + k) = FS + cS
or
c0 (1 + k) = cS
(5.33)
cS
dFS = c0 du
1 + 3k
u u
(5.34)
79
Vehicle Dynamics
FS
dFS du
8000
u=u
FS [N]
6000
M g cS
0 FS
63 kN/m 44 kN/m
dFS du
u=0
4000 29 kN/m
20 kN/m
2000
u
0.05
u [m]
0.1
u=
is given by (5.35)
cL = c0 (1 + 3 k) .
The intensity of the nonlinearity
c0 = 20000 N/m
and rear axle will be the same for the unloaded vehicle. With
cS = 34300 N/m
Eq. (5.33)
k =
(5.36)
The solid line in Fig. 5.5 shows the resulting nonlinear spring characteristics which is characterized by the spring rates
cL = c0 (1 + 3k) = 20 000 (1 +
for the unloaded and the loaded vehicle. Again, the undamped
0 f0C =
or
L f0C =
(5.37)
for the unloaded and the loaded vehicle dier quite a lot. The unloaded and the loaded vehicle have the same undamped natural frequencies if
c0 cL = M M+ M
or
cL M+ M = c0 M
(5.38)
will hold. Combing this relationship with Eq. (5.35) one obtains
1 + 3k =
M M+ M
or
k =
1 3
M+ M 1 M
1 M . 3 M
(5.39)
80
Hence, for the quarter car model with sity of the nonlinear spring amounts to and
cS = 34300 N/m will produce the dotted line in Fig. 5.5. The spring rates c0 = cS /(1 + k) = 34 300 N/m / (1 + 0.1667) = 29 400 N/m for the unloaded and cL = c0 (1 + 3k) = 29 400 N/m (1 + 3 0.1667) = 44 100 N/m for the loaded vehicle follow from Eqs. (5.34) and (5.35). Now, the undamped natural frequency for the unloaded 0 0 f0C = c0 /M = 1.46 Hz and the loaded vehicle f0C = cL /(M + M ) = 1.46 Hz are in deed the same.
zC (t)
To achieve the same damping rates for the chassis and the wheel model dierent values for the damping parameter
dS
were needed.
With increased damping the overshoot eect in the time history of the chassis displacement and the wheel load becomes smaller and smaller till it vanishes completely at and
C = 1
W = 1.
=1
=1
=1
the eigenvalues of the single mass oscillator change from complex to real.
Thus, producing a non oscillating solution without any sine and cosine terms. According to Eq. (5.10)
C = 1 = 2
and
W = 1
dopt S
C =1 Comfort
cS M ,
and
dopt S
W =1 Safety
= 2
(cS +cT )m .
(5.40)
dopt S
C =1 Comfort
= 5292
N m/s
and
dopt S
W =1 Safety
= 6928
N m/s
(5.41)
will avoid an overshoot eect in the time history of the chassis displacement in the time history of the wheel load
zC (t) or in the
FT (t).
optimal comfort and optimal ride safety will be dierent. Hence, a simple linear damper can either avoid overshoots in the chassis motions or in the wheel loads.
81
Vehicle Dynamics
chassis model
350 kg 20000 N/m dS C [ - ] 0.25 0.50 0.75 1.00 1.25 dS [Ns/m] 1323 2646 3969 5292 6614 W [ - ] 0.25 0.50 0.75 1.00 1.25
wheel model
dS [Ns/m] 1732 3464 5196 6928 8660 dS 50 kg 220000 N/m 20000 N/m
displacement [mm]
200 150 100 50 0 -50 -100 0 0.5 1 20 15 10 5 0 -5 -10
displacement [mm]
t [s]
0.05
0.1
t [s] 0.15
acceleration [g]
C
-0.5
2000 1000
-1 0 0.5 1
t [s]
1.5
0.05
0.1
t [s]
0.15
zC (t)
which surely is not a desirable conguration, because then, it takes a very long
z(t = 0) = z0
z(t = 0) = 0.
82
z0 zS z(t) tE t
errors allows to judge the attenuation. If the overall quadratic error becomes a minimum
t=tE 2
=
t=0
z(t)2 dt M in ,
(5.42)
the system approaches the steady state position as fast as possible. In theory holds, for practical applications a nite
tE
tE
To judge ride comfort and ride safety the hub motion of the chassis zC , its acceleration D zC and the variations of the dynamic wheel load FT can be used. In the absence of road D irregularities zR = 0 the dynamic wheel load from Eq. (5.8) simplies to FT = cT zW . Hence, the demands
t=tE 2 C
=
t=0
g1 zC
g2 zC
dt M in
(5.43)
and
t=tE 2 S
=
t=0
cT zW
dt M in g1
and
(5.44)
will guarantee optimal ride comfort and optimal ride safety. By the factors acceleration and the hub motion can be weighted dierently.
g2
the
The equation of motion for the chassis model can be resolved for the acceleration
2 zC = 0C zC + 2C zC ,
where, the system parameter
(5.45)
M , dS
and
cS
C =
C 0C = dS /(2M )
t=tE 2 C
0C = cS /M .
Then, the
=
t=0 t=tE
2 2 g1 0C zC + 2C zC
2 2 + g2 zC dt
zC zC
=
t=0
2 g1
2 2 (0C ) 2 g1 2 0C
2 g2
2 g1 2 g1
2 0C
2C
2
zC zC
M in ,
(5.46)
2C
(2C )
83
Vehicle Dynamics
where
xT = zC zC C
t=tE 2 S
is the state vector of the chassis model. In a similar way Eq. (5.44)
can be transformed to
t=tE
=
t=0
c2 T
2 zW
dt =
t=0
zW zW
c2 0 T 0 0
zW zW
M in ,
(5.47)
where
xT = zW zW W
The problems given in Eqs. (5.46) and (5.47) are called disturbance-reaction problems. They can be written in a more general form
t=tE
xT (t) Q x(t) dt M in
t=0
where
(5.48)
x(t)
Q = QT
single mass oscillators described by Eq. (5.9) the state equation reads as
z z x
For tE
0 1 2 0 2 A
z z x
(5.49)
the time response of the system exposed to the initial disturbance x(t = 0) = x0 vanishes x(t ) = 0, and the integral in Eq.(5.48) can be solved by
t=tE
xT (t) Q x(t) dt = xT R x0 , 0
t=0
where the symmetric matrix
(5.50)
R = RT
AT R + R A + Q = 0 .
For the single mass oscillator the Ljapunov equation
2 0 0 1 2
0 1 2 0 2
(5.52)
2 2 0 R12 0 R12 + Q11 = 0 2 0 R22 + R11 2 R12 + Q12 = 0 R12 2 R22 + R12 2 R22 + Q22 = 0
which easily can be solved for the elements of
(5.53)
R11 =
1 + 2 0 4
Q11 Q12 +
2 0 Q22 , 4
R12 =
84
Q11 Q12 +
x0 = [ z0 0 ]T
1 + 2 0 4
2 0 Q22 . 4
(5.55)
Now, the integral in Eq. (5.46) evaluating the ride comfort is solved by
2 C
2 = z0C
C 1 + 2 0C 4C 0C 4C C
2 g1 +
2 2 g1 0C
2 + g2
2 2 g1 0C 2 C +
2 0C 2 g (2C )2 4C 1
2 2 = z0C 0C
g2 2 0C
g2 2 0C
(5.56)
C 0C C 0C .
g1 = 1
and
g2 = 0
zC
is weighted
2 C zC
which will become a minimum for
2 2 = z0C 0C
or
0C 4C
As mentioned before,
(5.57)
0C 0 cS 0
C .
surely is not a desirable conguration. A low undamped natural frequency achieved by a soft suspension spring or a large chassis mass
C 0C 0 is
However, a
M .
large chassis mass is uneconomic and the suspension stiness is limited by the the loading conditions. Hence, weighting the chassis accelerations only does not lead to a specic result for the system parameter. The combination of
2 C zC
which will become a minimum for
1 + C 4C
(5.58)
0C
2 z0C = 0C
2 C |zC
d C
1 +1 2 4C
= 0.
(5.59)
contradicts the demand for rapidly van1 ishing accelerations. The viscous damping ratio C = solves Eq. (5.59) and minimizes 2 the merit function in Eq. (5.58). But again, this value does not correspond with C which minimizes the merit function in Eq. (5.57). Hence, practical results can be achieved only if the chassis displacements and the chassis accelerations will be evaluated simultaneously. To do so, appropriate weighting factors have to be chosen. In the equation of motion for the chassis (5.6) the terms
0C
M zC
and
cS zC
g1 = M
and
g2 = cS
and
or
g1 = 1
g2 =
cS 2 = 0C M
(5.60)
85
Vehicle Dynamics
2 C
2 2 = z0C 0C
0C + C 0C 2C 0C 0.
(5.61)
will satisfy
2 C |zC
d C
2 2 = z0C 0C
0C + 0C 2 2C 1 2 2
= 0.
(5.62)
C =
or a damping parameter of
(5.63)
1 opt C = 2 2 dS Comfort
2 cS M ,
(5.64)
will provide optimal comfort by minimizing the merit function in Eq. (5.61). For the passenger car with parameter will amount to
M = 350 kg dopt S
C = 1 2
and
Comfort
= 3742
(5.65)
which is
70%
2 S
2 z0W 0W
W +
1 4W
c2 T
(5.66)
where the model parameter m, cS , dS and cT where replaced by the undamped natural 2 frequency 0W = (cS + cT )/m and by the damping ratio W = W 0W = dS /(2m). 2 A soft tire cT 0 make the safety criteria Eq. (5.66) small S 0 and thus, reduces the dynamic wheel load variations. However, the tire spring stiness can not be reduced to arbitrary low values, because this would cause too large tire deformations. Small wheel masses
m 0
cS
will increase
0W
and thus,
reduce the safety criteria Eq. (5.66). The use of light metal rims improves, because of wheel weight reduction, the ride safety of a car. Hard body suspensions contradict a good driving comfort. With xed values for if
cT
and
0W
2 2 z0W S = W 0W
1+
1 2 4W 1 2
c2 = 0 T
(5.67)
W =
(5.68)
86
dopt S
Safety
(cS + cT ) m
(5.69)
will guarantee optimal ride safety by minimizing the merit function in Eq. (5.66). For the passenger car with
M = 350 kg
W = 1 2 Safety
and
dopt S
which is
= 3464
(5.70)
50%
sky zC zC
dB dW
cS zW zR cT
dS m zW zR
cS
FD
m cT
a) Standard Damper
Figure 5.8: Quarter car model with standard and sky hook damper To take this situation into account the simple quarter car models of section 5.2.3 must be combined to a more enhanced model, Fig. 5.8a. Assuming a linear characteristics the suspension damper force is given by
FD = dS (zW zC ) ,
where
(5.71)
87
Vehicle Dynamics
The sky hook damping concept starts with two independent dampers for the body and the wheel/axle mass, Fig. 5.8b. A practical realization in form of a controllable damper will then provide the damping force
FD = dW zW dC zC ,
where instead of the single damping constant are available. The equations of motion for the quarter car model are given by
(5.72)
dS
dW
and
dC
M zC = FS + FD M g , m zW = FT FS FD m g , M, m ments, and g
where are the sprung and unsprung mass, is the constant of gravity.
(5.73)
zC , zW
0 FS = FS + cS (zW zC ) ,
where
(5.74)
0 FS = mC g
cS
0 FT = FT + cT (zR zW ) ,
where
(5.75)
0 FT = (M + m) g
cS
zR
describes
FT 0
zC
0 0
1 0 dC M
dC m
zC
(5.76)
zW z C zW x
= 0 cS M
cS M cS cS +cT m m
1 zW + 0 zR , 0 dW z M C u dW cT m zW m x B
0 0 M g , mg were compensated by the preloads FS , FT , the term B u excitation, x denotes the state vector, and A is the state matrix. In this
linear approach the tire lift o is no longer taken into consideration. The eigenvalues
of the state matrix A will characterize the eigen dynamics of the quarter
car model. In case of complex eigenvalues the damped natural eigenfrequencies are given
88
Frequencies [Hz]
12 10
Damping ratio = D
Wheel
8 6 4 2 0 0 1000 2000 3000
dS 0.5
0.4 0.2
Chassis
4000 5000
220000 N/m
3880 3220
0 0 1000 2000 3000 4000 5000
dS [N/(m/s)]
Figure 5.9: Quarter car model with standard damper
dS [N/(m/s)]
= Im(), and according to Eq. (vdyn-eq: = D = Re()/ ||. evaluates the damping ratio.
and
relative damping
dC = dS
dW = dS
standard damper described by Eq. (5.71). Fig. 5.9 shows the eigenfrequencies and the damping ratios
= D
Optimal ride comfort with a damping ratio of could be achieved with the damping parameter
1 C = 2 2 0.7 for the chassis motion parameter dS = 3880 N/(m/s), and the damping
f = /(2) dS .
dS = 3220 N/(m/s)
W = 0.5
3742 N/(m/s)
3464 N/(m/s)
Eqs. (5.65) and (5.70) with the single mass models. Hence, the very simple single mass models can be used for a rst damper layout. Usually, as it is here, optimal ride comfort and optimal ride safety cannot achieved both by a standard linear damper. The sky-hook damper, modeled by Eq. (5.72), provides with rameter. Their inuence to the eigenfrequencies Fig. 5.10. The the sky-hook damping parameter damping ratio
dW
and
dS
is shown in
dC , dW dW .
mainly depends on
dC ,
mainly depends on
wheel motion can be adjusted to nearly each design goal. Here, a sky-hook damper with
dC = 3900 N/(m/s) and dW = 3200 N/(m/s) would generate the damping ratios dC = 0.7 and dW = 0.5 hence, combining ride comfort and ride safety within one damper layout.
89
Vehicle Dynamics
Frequencies [Hz]
12 10 8 6 4 2 0 0 1000 2000 3000 4000 5000 1 0.8 0.7
Damping ratios C, W
dC [N/(m/s)]
350 kg
dC
dC 0.6 dW 0.5
0.4 0.2 0
W
0 1000 2000 3000 4000 5000
dW [N/(m/s)]
dW [N/(m/s)]
FD = dW zW dC zC + =
dW zW dC zC (zW zC ) = d (zW zC ) S zW zC d S
(5.77)
the sky-hook damper can be realized by a standard damper in the form of Eq. (5.71). The new damping parameter dS now nonlinearly depends on the absolute vertical velocities of the chassis and the wheel dS = dS (zC , zW ). As, a standard damper operates in a dissipative mode only the damping parameter will be restricted to positive values, dS > 0. Hence, the passive realization of a sky-hook damper will only match with some properties of the ideal damper law in Eq. (5.72). But, compared with the standard damper it still can provide a better ride comfort combined with an increased ride safety.
M zC = FS + FD M g m zW = FT FS FD m g ,
where and
(5.78)
g = 9.81 m/s2 labels the constant of gravity, M , m are the masses of the chassis the wheel, FS , FD , FT describe the spring, the damper, and the vertical tire force,
90
zC
FD
nonlinear spring
nonlinear damper
FS FS u
m cT
FD
zW zR
Figure 5.11: Quarter car model with nonlinear spring and damper characteristics
zC
zW
FS =
0 FS
c0 u c0 u
1 + kr 1 + kc
u ur u uc
u<0
2
(5.79)
u0
where
0 FS = M g
u = zW zC
describes the spring travel. Here,
(5.80)
u < 0 marks tension (rebound), and u 0 compression. Two sets of kr , ur and kc , uc dene the spring nonlinearity during rebound and compression. For kr = 0 and kc = 0 a linear spring characteristics is obtained.
A degressive damper characteristics can be modeled by
FD (v) =
d0 v 1 pr v d0 v 1 + pc v v = 0,
v <0,
(5.81)
v 0,
and the damper velocity is dened by (5.82)
where
d0
v = zW zC . v<0 v 0.
The sign convention of the damper velocity was chosen consistent to the spring travel. Hence, rebound is characterized by and compression by The parameter
pr
91
Vehicle Dynamics
and
pc
make it possible to model the damper nonlinearity dierently in the rebound and
pr = 0
and
pc = 0.
The nonlinear spring design in Section 5.3.2 holds for the compression mode. Hence, using the same data we obtain:
kc = k = 0.1667.
mode, Fig. 5.12a.
By setting
c0 = 29 400 N/m, uc = u = umax /2 = 0.10/2 = 0.05 and ur = uc and kr = 0 a simple linear is used in the rebound
a) Spring
7000 5000
b) Damper
d0 = 4200 N/(m/s) pr = 0.4 1/(m/s) pc = 1.2 1/(m/s)
cS = 34300 N/m 6000 c0 = 29400 N/m ur = 0.05 m 5000 kr = 0 uc = 0.05 m 4000 kc = 0.1667
3000 2000 1000 0 -0.1
rebound u<0 compression u>0
FS [N/m]
FD [N/m]
2500
-2500
rebound v<0 compression v>0
-0.05
0.05 u [m]
0.1
-5000 -1
-0.5
0.5 v [m/s]
Figure 5.12: Spring and damper characteristics: - - - linear, nonlinear According to Section 5.3.4 damping coecients optimizing the ride comfort and the ride safety can be calculated from Eqs. (5.65) and (5.69). For equivalent linear spring rate,
M = 350 kg , m = 50 kg
and
2 cS M =
(cS + cT ) m =
1 The mean value d0 = 4200N/(m/s) may serve as compromise. With pr = 0.4(m/s) and pc = 1.2 (m/s)1 the nonlinearity becomes more intensive in compression than rebound,
Fig. 5.12b.
5.5.2 Results
The quarter car model is driven with constant velocity over a single obstacle. Here, a cosine shaped bump with a height of The results are plotted in Fig. 5.13. Compared to the linear model the nonlinear spring and damper characteristics result in 2 2 signicantly reduced peak values for the chassis acceleration (6.0m/s instead of 7.1m/s )
H = 0.08 m
and a length of
L = 2.0 m
was used.
92
0.5
1.5
0.5
1.5
0.5
1.5
time [s]
time [s]
time [s]
over a single obstacle
v = 20 km h
instead of
6660 N ).
t 0.25 s
can be avoided. While crossing the bump large damper velocities occur. Here, the degressive damper characteristics provides less damping compared to the linear damper which increases the suspension travel.
0.5
1.5
0.5
1.5
time [s]
time [s]
time [s]
Figure 5.14: Results for low damping compared to nonlinear model A linear damper with a lower damping coecient,
the peaks in the chassis acceleration and in the wheel load, but then the attenuation of the disturbances will take more time. Fig. 5.14. Which surely is not optimal.
93
6 Longitudinal Dynamics
6.1 Dynamic Wheel Loads
6.1.1 Simple Vehicle Model
The vehicle is considered as one rigid body which moves along an ideally even and horizontal road. At each axle the forces in the wheel contact points are combined in one normal and one longitudinal force.
S h
mg a2
Fz1
Fx1
a1
Fx2 Fz2
If aerodynamic forces (drag, positive and negative lift) are neglected at rst, the equations of motions in the
x-, z -plane
will read as
(6.1) (6.2)
(6.3)
a1 + a2
is the
These are only three equations for the four unknown forces
But, if we
insert Eq. (6.1) in Eq. (6.3), we can eliminate two unknowns at a stroke
0 = Fz1 a1 Fz2 a2 + m v h .
(6.4)
94
The equations Eqs. (6.2) and (6.4) can be resolved for the axle loads now
Fz1 = m g
h a2 mv , a1 + a2 a1 + a2 h a1 + mv . a1 + a2 a1 + a2
st Fz2 = m g
(6.5)
Fz2 = m g
The static parts
(6.6)
st Fz1 = m g
a2 , a1 + a2
a1 a1 + a2
(6.7)
describe the weight distribution according to the horizontal position of the center of gravity. The height of the center of gravity only inuences the dynamic part of the axle loads,
dyn Fz1 = m g
When accelerating
h v , a1 + a2 g
dyn Fz2 = +m g
h v . a1 + a2 g
(6.8)
v > 0,
v < 0.
v x Fx1 Fz1 a1 a2
mg
Fx2 Fz2
For a vehicle on a grade, Fig.6.2, the equations of motion Eq. (6.1) to Eq. (6.3) can easily be extended to
m v = Fx1 + Fx2 m g sin , 0 = Fz1 + Fz2 m g cos , 0 = Fz1 a1 Fz2 a2 + (Fx1 + Fx2 ) h ,
(6.9)
95
Vehicle Dynamics
where
denotes the grade angle. Now, the axle loads are given by
Fz1 = m g cos
h a2 h tan mv , a1 + a2 a1 + a2 a1 + h tan h + mv , a1 + a2 a1 + a2
(6.10)
Fz2 = m g cos
(6.11)
where the dynamic parts remain unchanged, whereas now the static parts also depend on the grade angle and the height of the center of gravity.
FD2
Fz2
Figure 6.3: Vehicle with aerodynamic forces If we assume a positive driving speed,
v > 0,
m v = Fx1 + Fx2 FAR , 0 = Fz1 FD1 + Fz2 FD2 m g , 0 = (Fz1 FD1 ) a1 (Fz2 FD2 ) a2 + (Fx1 + Fx2 ) h ,
where (6.12)
FAR
and
FD1 , FD2
describe the air resistance and the down forces. For the dynamic
Fz1 = FD1 + m g
a2 h (m v + FAR ) , a1 + a2 a1 + a2 a1 h + (m v + FAR ) . a1 + a2 a1 + a2
(6.13)
Fz2 = FD2 + m g
The down forces
(6.14)
FD1 , FD2
FAR
96
Fz1 0
and
Fz2 0
v a2 cos sin g h
and
v a1 cos sin . g h
(6.15)
a1 cos h
v + sin g
a2 cos . h
and decelerations (v the center of gravity.
(6.16)
Hence, the maximum achievable accelerations (v limited by the grade angle and the the tilting condition Eq. (6.16) results in
> 0) position a1 , a2 , h of
a2 a1 tan h h
(6.17)
which describes the climbing and downhill capacity of a vehicle. The presence of aerodynamic forces complicates the tilting condition. Aerodynamic forces become important only at high speeds. Here, the vehicle acceleration is normally limited by the engine power.
|Fx1 | Fz1
where the same friction coecient case
and
|Fx2 | Fz2
(6.18)
Fx1 = Fz1
Fx2 = Fz2
(6.19)
(6.20)
v g
That means climbing (v a friction coecient
=
max
(6.21)
or downhill stopping (v
requires at least
97
Vehicle Dynamics
According to the vehicle dimensions and the friction values the maximal acceleration or deceleration is restricted either by Eq. (6.16) or by Eq. (6.21). If we take aerodynamic forces into account, the maximum acceleration and deceleration on a horizontal road will be limited by
1 +
FD1 FD2 + mg mg
FAR mg
v g
1 +
FD1 FD2 + mg mg
FAR . mg
(6.22)
In particular the aerodynamic forces enhance the braking performance of the vehicle.
Fx1 = 0
and
Fx2 = Fz2
m vR WD = m g
a1 h vR WD + a1 + a2 a1 + a2 g
(6.23)
where the subscript R WD indicates the rear wheel drive. Hence, the maximum acceleration for a rear wheel driven vehicle is given by
vR WD = g
h 1 a1 + a2
a1 . a1 + a2
(6.24)
By setting
Fx1 = Fz1
and
Fx2 = 0, vF WD = g
a2 , h a1 + a2 1+ a1 + a2
(6.25)
where the subscript F WD denotes front wheel drive. Depending on the parameter a2 and h the accelerations may be limited by the tilting condition v a2 . g h
, a1 ,
The maximum accelerations of a single axle driven vehicle are plotted in Fig. 6.4. For rear
a2 /(a1+a2 ) which describes the static axle load distribution is in the range of 0.4 a2 /(a1+a2 ) 0.5. For = 1 and h = 0.55 this results in maximum accelerations in between 0.77 v/g 0.64. Front wheel driven passenger cars usually cover the range 0.55 a2 /(a1 +a2 ) 0.60 which produces accelerations in the range of 0.45 v/g 0.49. Hence, rear wheel driven vehicles can accelerate much
wheel driven passenger cars, the parameter faster than front wheel driven vehicles.
98
FWD
RWD
FWD
v/g
RWD
0.2
0.4
0.6
a2 / (a1+a2)
0.8
Fx1 = Fz1
and
Fx2 = 0
m vF WB = m g
h vF WB a2 a1 + a2 a1 + a2 g
(6.26)
where the subscript F WB indicates front wheel braking. Then, the maximum deceleration is given by
vF WB = g
If only the rear axle is braked (Fx1 deceleration
h 1 a1 + a2
a2 . a1 + a2
(6.27)
= 0, Fx2 = Fz2 ), h 1+ a1 + a2
vR WB = g
a1 , a1 + a2
(6.28)
a2 ,
where the subscript R WB denotes a braked rear axle. Depending on the parameters v and h, the decelerations may be limited by the tilting condition a1 . g h
, a1 ,
The maximum decelerations of a single axle braked vehicle are plotted in Fig. 6.5. For passenger cars the load distribution parameter to
0.6.
a2 /(a1 +a2 ) usually covers the range of 0.4 v/g = 0.51 to v/g = 0.77
to
will be achieved. This is a quite large value compared to the deceleration range of a braked rear axle which is in the range of
v/g = 0.49
v/g = 0.33.
99
Vehicle Dynamics
v/g
FWB
RWB
0 0.2 0.4 0.6
a2 / (a1+a2)
0.8
Fx1 Fx2 v = + . g mg mg
the longitudinal forces
(6.29)
Fx1
and
Fx2 .
and
limited by wheel load and friction. The optimal combination of the same skid resistance.
Fx1
Fx2
Fx1 = Fz1
With Eq. (6.5) and Eq. (6.6) one obtains
and
Fx2 = Fz2 .
(6.30)
Fx1 = mg
and
a2 v h g a1 v + h g
h a1 + a2 h . a1 + a2
(6.31)
Fx2 = mg
(6.32)
With Eq. (6.31) and Eq. (6.32) one gets from Eq. (6.29)
v = , g
where it has been assumed that
(6.33)
Fx1
and
Fx2
Fx1 v = mg g
a2 v h g
h a1 + a2
(6.34)
100
h . a1 + a2
and
Fx2 v = mg g
a1 v + h g
(6.35)
that grant the same skid resistance at both axles can be calculated now.
B2/mg
braking
-a1/h
Fx1/mg
dFx2 dFx1 0
-1
-2
B1/mg
a =1.15
driving
a =1.35
1
1 2
h=0.55 =1.20
tilting limits
2
a2/h
Figure 6.6: Optimal distribution of driving and braking forces Fig. 6.6 shows the curve of optimal drive and brake forces for typical passenger car values. At the tilting limits
at the lifting axle. The initial gradient only depends on the steady state distribution of the wheel loads. From Eqs. (6.34) and (6.35) it follows
Fx2/mg
Fx1 mg = v d g
a2 v 2 h g
h a1 + a2
(6.36)
and
Fx2 mg = v d g
a1 v +2 h g
h . a1 + a2
(6.37)
101
Vehicle Dynamics
For
v/g = 0
d Fx2 d Fx1
=
0
a1 . a2
(6.38)
Fx1/mg Fx2/mg
Fx1/mg Fx2/mg
Fx1/mg
limitation
Fx2/mg
constant distribution
reduction
Figure 6.7: Dierent distributions of brake forces When braking, the stability of a vehicle depends on the potential of generating a lateral force at the rear axle. Thus, a greater skid (locking) resistance is realized at the rear axle than at the front axle. Therefore, the brake force distribution are all below the optimal curve in the physically relevant area. This restricts the achievable deceleration, specially at low friction values. Because the optimal curve depends on the center of gravity of the vehicle an additional safety margin have to be installed when designing real brake force distributions. The distribution of brake forces is often tted to the axle loads. There, the inuence of the height of the center of gravity, which may also vary much on trucks, is not taken into account and has to be compensated by a safety margin from the optimal curve. Only the control of brake pressure in anti-lock-systems provides an optimal distribution of brake forces independently from loading conditions.
6.3.5 Anti-Lock-Systems
On hard braking maneuvers large longitudinal slip values occur. Then, the stability and/or steerability is no longer given because nearly no lateral forces can be generated. By controlling the brake torque or brake pressure respectively, the longitudinal slip can be restricted to values that allow considerable lateral forces. Here, the angular wheel acceleration
of the wheel are derived from the measured angular speeds of the wheel by dierentiation. The rolling condition is fullled with a longitudinal slip of
rD = x
102
holds, where
rD
of the vehicle. According to Eq. (6.21), the maximum acceleration/deceleration of a vehicle depends on the friction coecient,
|| = g . x ||
a simple
1 || . x rD
(6.40)
Because no reliable possibility to determine the local friction coecient between tire and road has been found until today, useful information can only be gained from Eq. (6.40) at optimal conditions on dry road. Therefore, the longitudinal slip is used as a second control variable. In order to calculate longitudinal slips, a reference speed is estimated from all measured wheel speeds which is used for the calculation of slip at all wheels, then. This method is too imprecise at low speeds. Therefore, no control is applied below a limit velocity. Problems also arise when all wheels lock simultaneously for example which may happen on icy roads. The control of the brake torque is done via the brake pressure which can be
increased,
held,
or
decreased
made slower than the increment. To prevent a strong yaw reaction, the
at the rear axle. Here, the break pressure at both wheels is controlled by the wheel running on lower friction. Thus, at least the brake forces at the rear axle cause no yaw torque. However, the maximum achievable deceleration is reduced by this.
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:
(6.41)
R1
and
R2
The forces between wheel body and vehicle body are labeled drive torques the wheel loads
FF 1
and
FF 2 .
At the wheels
MA1 , MA2 and brake torques MB1 , MB2 , longitudinal forces Fx1 , Fx2 and Fz1 , Fz2 apply. The brake torques are directly supported by the wheel
bodies, whereas the drive torques are transmitted by the drive shafts to the vehicle body.
103
Vehicle Dynamics
zA MA1 z1 MA1 R1 Fz1 a1 MB1 Fx1 R a2 R2 Fz2 MA2 MB2 Fx2 MB1 hR MA2 z2 MB2 FF2 FF1 xA A
The forces and torques that apply to the single bodies are listed in the last column of the tables 6.1 and 6.2. The velocity of the vehicle body and its angular velocity are given by
v0A,0
xA 0 = 0 + 0 ; 0 zA
0A,0
0 = A . 0
(6.42)
At small rotational motions of the body one gets for the velocities of the wheel bodies and wheels
v0RK1 ,0 = v0R1 ,0
xA 0 hR A + 0 = 0 + 0 + A 0 zA a1 xA 0 hR A + 0 = 0 + 0 + 0 zA +a2 A
x1 z1
z1
; .
(6.44) (6.43)
0 z1
x2 z2
z2
v0RK2 ,0 = v0R2 ,0
0 z2
The angular velocities of the wheel bodies and wheels are obtained from
0RK1 ,0
0 0 = A + 1 0 0
and
0R1 ,0
0 0 0 = A + 1 + R1 0 0 0
(6.45)
104
as well as
0RK2 ,0
0 0 = A + 2 0 0
and
0R2 ,0
0 0 0 = A + 2 + R2 0 0 0
(6.46)
z =
can be written as
xA zA A 1 R1 2 R2
(6.47)
the velocities and angular velocities given by Eqs. (6.42), (6.43), (6.44), (6.45), and (6.46)
v0i =
j=1
v0i zj zj
7
and
0i =
j=1
0i zj zj
(6.48)
i = 1(1)5
j = 1(1)7
z1 0 0 0
x1 z1
mA
mRK1
wheel front
mR1
mRK2
wheel rear
mR2
xA 1 0 0 1 0 0 1 0 0 1 0 0 1 0 0
zA 0 0 1 0 0 1 0 0 1 0 0 1 0 0 1
A 0 0 0 hR 0 a1 hR 0 a1 hR 0 a2 hR 0 a2
0 1
x1 z1
0 1 0 0 0 0 0 0
v0i /zj R1 z2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 x2 0 z2 0 0 0 1 x2 0 z2 0 0 0 1
applied forces
R2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Fie 0 0 FF 1 +FF 2 mA g 0 0 FF 1 mRK1 g Fx1 0 Fz1 mR1 g 0 0 FF 2 mRK2 g Fx2 0 Fz2 mR2 g
Table 6.1: Partial velocities and applied forces With the aid of the partial velocities and partial angular velocities the elements of the mass matrix
Q can
be calculated.
M (i, j) =
k=1
v0k zi
v0k mk + zj
k=1
0k zi
0k ; zj
i, j = 1(1)7 ;
(6.49)
105
Vehicle Dynamics
RK1
wheel front
R1
RK2
wheel rear
R2
xA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
zA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
A 0 1 0 0 1 0 0 1 0 0 1 0 0 1 0
z1 0 0 0 0
1 z1
0 0
1 z1
0 0 0 0 0 0 0
R1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
0i /zj applied torques z2 R2 Mie 0 0 0 0 0 MA1MA2 a1 FF 1 +a2 FF 2 0 0 0 0 0 0 0 0 MB1 0 0 0 0 0 0 0 0 MA1 MB1 R Fx1 0 0 0 0 0 0 2 0 MB2 z2 0 0 0 0 0 0 2 1 MA2 MB2 R Fx2 z2 0 0 0
Q(i) =
k=1
v0k zi
T e Fk
+
k=1
0k zi
T e Mk ;
i = 1(1)7 .
(6.50)
Then, the equations of motion for the plane vehicle model are given by
Mz = Q.
(6.51)
6.4.3 Equilibrium
With the abbreviations
m1 = mRK1 + mR1 ;
and
m2 = mRK2 + mR2 ; h = hR + R
mG = mA + m1 + m2
(6.52)
(6.53)
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
(6.54)
Fx1 m1 g +
1 (MA1 z1
R Fx1 ) ;
(6.55)
106
2 (MA2 z2
Q(6) = Fz2 FF 2 +
x2 z2
Fx2 m2 g +
R Fx2 ) ;
(6.56)
MA1 = 0 ;
MA2 = 0 ;
MB1 = 0 ;
MB2 = 0
(6.57)
from Eqs. (6.54), (6.55) and (6.56) one gets the steady state longitudinal forces, the spring preloads, and the wheel loads
0 Fx2 = 0 ;
mA g ;
a2 a1 +a2
0 FF 2 =
a a+b
mA g ;
a1 a1 +a2
(6.58)
= m1 g +
mA g ;
0 Fz2
= m2 g +
mA g .
R R1 = xA hR A + R R2 = xA hR A +
x1 z1 x2 z2
z1 ; z2
(6.59)
hold. In steady state the pitch motion of the body and the vertical motion of the wheels reach constant values
st A = A = const. ,
and Eq. (6.59) simplies to
st z1 = z1 = const. ,
st z2 = z2 = const.
(6.60)
R R1 = xA ;
R R2 = xA .
(6.61)
With Eqs. (6.60), (6.61) and (6.53) the equation of motion (6.51) results in
hR (m1 +m2 ) xA + R1
x1 z1
xA R
+ R2 =
a Fz1
xA R
m1 xA +
1 z1
R1 xA R
a FF 1
x1 z1
a Fx1
1 (MA1 z1
a R Fx1 )
R1
x2 z2
xA R xA R xA R
(6.63)
m2 xA +
2 z2
R2 R2
(6.64)
107
Vehicle Dynamics
where the steady state spring forces, longitudinal forces, and wheel loads have been separated into initial and acceleration-dependent terms
st 0 a FF i = FF i + FF i ;
i = 1, 2 .
(6.65)
a With given torques of drive and brake the vehicle acceleration xA , the wheel forces Fx1 , a a a a a Fx2 , Fz1 , Fz2 and the spring forces FF 1 , FF 2 can be calculated from Eq. (6.62), Eq. (6.63) and Eq. (6.64)
Via the spring characteristics which have been assumed as linear the accelerationdependent forces also cause a vertical displacement and pitch motion of the body besides the vertical motions of the wheels,
a FF 1 a FF 2 a Fz1 a Fz2
Especially the pitch of the vehicle if too distinct.
(6.66)
By an axle kinematics with 'anti dive' and/or 'anti squat' properties, in longitudinal direction during the suspension travel.
brake pitch angle can be reduced by rotating the wheel body and moving the wheel center
x-, z -plane,
Fig. 6.9.
pitch pole
108
7 Lateral Dynamics
7.1 Kinematic Approach
7.1.1 Kinematic Tire Model
When a vehicle drives through a curve at low lateral acceleration, small lateral forces will be needed for course holding. Then, hardly lateral slip occurs at the wheels. In the ideal case at vanishing lateral slip the wheels only move in circumferential direction. The velocity component of the contact point in the lateral direction of the tire vanishes then
vy = eT v0P = 0 . y
vehicles moving in the low lateral acceleration range.
(7.1)
This constraint equation can be used as 'kinematic tire model' for course calculation of
M,
Fig. 7.1.
At slowly moving vehicles the lay out of the steering linkage is usually done according to the Ackermann geometry. Then, the following relations apply
a a and tan 2 = , (7.2) R R+s where s labels the track width and a denotes the wheel base. Eliminating the curve radius R, we get a a tan 1 tan 2 = or tan 2 = . (7.3) a a + s tan 1 +s tan 1 a A a The deviations 2 = 2 2 of the actual steering angle 2 from the Ackermann steering A angle 2 , which follows from Eq. (7.3), are used, especially on commercial vehicles, to tan 1 =
judge the quality of a steering system. At a rotation around the momentary pole every point of the vehicle. The angle
M,
axis of the vehicle is called side slip angle. The side slip angle at point
is given by (7.4)
tan P =
where
x R P to
or
tan P =
x tan 1 , a
109
Vehicle Dynamics
v P P a x 1
Rmax ,
whereas a point on the inner side of the vehicle at the location of the rear
Rmin .
a, b are the wheel base and the width of the vehicle, and f Rmin
the front of the vehicle to the front axle. Then, the space requirement can be specied as a function of the cornering radius
R = Rmax Rmin
R = Rmax Rmin =
The space requirement space of
(7.6)
2.5
Rmin = 5.0 m, a bus requires needs only 1.5 times the width.
110
ma
Rmin
7 6 5 R [m] 4 3 2 1 0 0 10
20 30 R min [m]
40
50
111
Vehicle Dynamics
r02,0
and the rotation matrix
x0 , y0 , z0 x = y R
(7.7)
A02
Here, the tire radius
(7.8)
a b y2
A1
x2
A2
y0
y 3
A3
x0
with
r21,2
a = 0 0
(7.9)
112
with
A21
(7.10)
describe the position and the orientation of the front axle, where wheel base and
labels the
(7.11)
r2K,2
and the rotation matrix
b = 0 0
and
rK3,2
c = 0 0
(7.12)
with
A23
(7.13)
dene the position and the orientation of the trailer axis, with between vehicle and trailer, and coupling point
b, c
3.
v02,2
vx2 = 0 . 0
(7.14)
v02,0 = r02,0
x = y . 0 0
(7.15)
(7.16)
113
Vehicle Dynamics
compared to Eq. (7.15) results in two rst order dierential equations for the position coordinates
and
(7.18)
v01,2
(7.19)
ex1,2
cos = sin 0
and
ey1,2
sin = cos 0
(7.20)
dene the longitudinal and lateral direction at the front axle. According to Eq. (7.1) the velocity component lateral to the wheel must vanish,
(7.21)
(7.22)
remains. From Eq. (7.21) a rst order dierential equation follows for the yaw angle
=
7.1.4.3 Entering a Curve
In analogy to Eq. (7.2) the steering angle or with
vx2 tan . a
(7.23)
k = 1/R
tan =
a = ak . R
(7.24)
= vx2 k .
With the curvature gradient
(7.25)
k = k(t) = kC
t , T
(7.26)
114
k = kC .
the entering of a curve is described as a continuous transition from a straight line with the curvature
k=0
The yaw angle of the vehicle can be calculated by simple integration now
(t) =
where at time
t=0
t=T
a vanishing yaw
vx2 kC t2 , T 2 angle, (t = 0) = 0,
t=T
position of the vehicle follows with Eq. (7.27) from the dierential equations Eq. (7.17)
x = vx2
t=0
cos
vx2 kC t2 T 2
dt ,
y = vx2
t=0
sin
vx2 kC t2 T 2
dt .
(7.28)
vx2 = const., Eq. (7.28) is the parameterized form of a clothoide. tan-function can be approximated t T
(7.29)
From Eq. (7.24) the necessary steering angle can be calculated, too. If only small steering angles are necessary for driving through the curve, the by its argument, and
= (t) a k = a kC
holds, i.e. the driving through a clothoide is manageable by a continuous steer motion.
(7.30)
r02,0 = v02,0
02,0
Eqs. (7.16) and (7.19). The position vector from the rear axle to the axle of the trailer is
b c cos c sin , = 0
(7.31)
r2K,2
and
rK3,3
are dened in Eq. (7.12). The time derivative of Eq. (7.31) results
r23,2
(7.32)
v03,2
x2 , y2 , z2 now vx2 0 b c cos c sin vx2 + c sin (+ ) c sin + c cos = b c cos (+ ) . = 0 + 0 0 0 0 0 v02,2 02,2 r23,2 r23,2
(7.33)
115
Vehicle Dynamics
The longitudinal and lateral direction at the trailer axle are dened by the unit vectors
ex3,2
cos = sin 0
and
ey3,2
(7.34)
(7.35)
(7.36)
remains. If Eq. (7.23) is inserted into Eq. (7.35) now, one will get a rst order dierential
vx2 a
a sin + c
b cos + 1 tan c
(7.37)
The dierential equations Eq. (7.17) and Eq. (7.23) describe position and orientation within the Eq. (7.37).
x0 , y0
plane. The position of the trailer relative to the vehicle follows from
20 [m] 10 0
-30
-20
-10
10
20
30 [m] 40
50
60
[o]
10
15
20
[s]
25
30
Figure 7.5: Entering a curve For a given set of vehicle parameters velocity,
a, b , c ,
= (t),
can be calculated by numerical integration of the dierential equations Eqs. (7.17), (7.23) and (7.37). If the steering angle is slowly increased at constant driving speed, the vehicle drives a gure which will be similar to a clothoide, Fig. 7.5.
116
B,
v0C,B
where
v cos = v sin 0
and
0F,B
0 = 0 , v
(7.38)
denotes the side slip angle of the vehicle measured at the center of gravity. The on an at horizontal road
=
where
v , R
(7.39)
a2
Fx2 Fy2
a1
C R yB
xB
Fx1
Fy1
In the body xed reference frame, linear and angular momentum result in
m m
v2 sin R v2 cos R
= Fx1 cos Fy1 sin + Fx2 , = Fx1 sin + Fy1 cos + Fy2 , 0 = a1 (Fx1 sin + Fy1 cos ) a2 Fy2 ,
(7.40)
(7.41) (7.42)
117
Vehicle Dynamics
where
longitudinal and vertical direction applied at the front and rear axle, and average steer angle at the front axle.
species the
The engine torque is distributed by the center dierential to the front and rear axle. Then, in steady state condition we obtain
Fx1 = k FD
where
and
Fx2 = (1 k) FD ,
FD
mv 2 = sin , R mv 2 = cos , R
(7.44)
FD
(7.45)
or
a2 tan = tan a1 + a2
(7.46)
holds. This fully corresponds with the Ackermann geometry. But, the Ackermann geometry applies only for small lateral accelerations. In real driving situations, the side slip angle of a vehicle at the center of gravity is always smaller than the Ackermann side slip a2 angle. Then, due to tan < tan a driving force FD > 0 is needed to overcome the a1 +a2 'cornering resistance' of the vehicle.
118
2 1 m ay h2 mg
h1
F yL FzL s/2
F yR s/2
FzR
The balance of torques at the height of the track plane applied at the already inclined vehicle results in
s = m ay (h1 + h2 ) + m g [(h1 + h2 )1 + h2 2 ] , 2 where ay describes the lateral acceleration, m is the sprung mass, and small the axle and the body were assumed, 1 1, 2 1. (FzL FzR )
On a left-hand tilt, the right tire raises
T FzR = 0 ,
whereas the left tire carries the complete vehicle weight
(7.48)
T FzL = m g .
Using Eqs. (7.48) and (7.49) one gets from Eq. (7.47)
(7.49)
s aT h2 y T T 2 = 1 2 . g h1 + h2 h1 + h2
The vehicle will turn over, when the lateral acceleration
(7.50)
T rises above the limit ay . Roll T T of axle and body reduce the overturning limit. The angles 1 and 2 can be calculated from the tire stiness cR and the roll stiness of the axle suspension.
ay
119
Vehicle Dynamics
If the vehicle drives straight ahead, the weight of the vehicle will be equally distributed to both sides
1 mg . 2 Fz
(7.51)
(7.52)
and Eqs. (7.49), (7.51), one obtains for the increase of the wheel load at the overturning limit
Fz =
1 mg . 2
(7.53)
Fz = cR
where
r,
(7.54)
cR
Because the right tire simultaneously rebounds with the same amount, for the roll angle of the axle
T r = s 1
or
T 1 =
2 s
mg s cR
(7.55)
holds. In analogy to Eq. (7.47) the balance of torques at the body applied at the roll center of the body yields
cW 2 = m ay h2 + m g h2 (1 + 2 ) ,
where limit
(7.56)
cW names the roll stiness of the body suspension. In particular, at the overturning ay = aT y aT mgh2 mgh2 y T T 2 = + 1 (7.57) g cW mgh2 cW mgh2
T applies. Not allowing the vehicle to overturn already at ay = 0 demands a minimum of min roll stiness cW > cW = mgh2 . With Eqs. (7.55) and (7.57) the overturning condition Eq. (7.50) reads as
(h1 + h2 )
aT aT s 1 1 1 1 y y h2 , = (h1 + h2 ) h2 g 2 cR g cW 1 cW 1 cR
(7.58)
cR c = m g R s
and
c = W
cW m g h2
(7.59)
aT y = g
s 2 h2 h1 + h2 + cW 1
1 c R
(7.60)
120
0.6 0.5
overturning limit ay
20
15 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 * 10
Figure 7.8: Tilting limit for a typical truck at steady state cornering
remains. At heavy trucks, a twin tire axle may be loaded with of one tire is
cR = 800 000 N/m, and the track h1 = 0.8 m and h2 = 1.0 m hold at maximal load. These values produce the results shown in Fig. 7.8. Even with a rigid body suspension cW , the vehicle turns over at a lateral acceleration of ay 0.5 g . Then, the roll angle of the vehicle solely results from the tire
deection. At a normalized roll stiness of
c = 5, W
ay 0.45 g
and
so reaches already 90% of the maximum. The vehicle will turn over at a roll angle of = 1 + 2 10 then.
ay ,
centrifugal forces
will be applied to the single masses. At the simple roll model in Fig. 7.9, these are the
m A ay
and
m R ay ,
where
mA
mR
m A ay
z1 = z2 .
At steady state cornering, the vehicle forces are balanced. With the principle of virtual work
W = 0 ,
the equilibrium position can be calculated. At the simple vehicle model in Fig. 7.9 the suspension forces
(7.61)
FF 1 , FF 2
cA
and
121
Vehicle Dynamics
b/2
FF1 z1 mR a y S1 Q1 F z1 1 y1 F y1
cQ , cR .
The work
W = V
via the
potential
V.
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 2 cS (z1 z2 )2 1 2 cQ (yA + h0 A + y1 + r0 1 )2 1 b 2 cR zA + 2 A + z1
where the abbreviation (7.62)
1 2
cQ (yA + h0 A + y2 + r0 2 )2
2
1 2
b cR zA 2 A + z2
hR = h0 r0
cS
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 =
the work
y z1 , z
1 =
1 z
and
y2 =
y z2 , z
2 =
2 z
(7.63)
y = [ yA , zA , A , z1 , z2 ]T .
Due to
(7.64)
W = W (y)
the principle of virtual work Eq. (7.61) leads to
(7.65)
W =
W y = 0 . y
(7.66)
122
Because of
y = 0,
Ky = b
results from Eq. (7.66). The matrix
(7.67)
are given by
Q y cQ z
2 cQ 0 K = 2 cQ h0 yQ c z Q Q y cQ z
and
0 2 cR 0 cR cR
2 cQ h0 0 c
Q b c +h0 y cQ 2 R z
y Q c z Q
cR
Q b c +h0 y cQ 2 R z
c + cS + cR A cS ay .
b 2 cR h0 y cQ z
cR Q y b 2 cR h0 z cQ cS cA + cS + cR
(7.68)
b =
(7.69)
y y Q = + r0 , z z z
c A
= cA + cQ
y z
c =
2 cQ h2 0
+ 2 cR
b 2
(7.70)
The system of linear equations Eq. (7.67) can be solved numerically, e.g. with MATLAB. Thus, the inuence of axle suspension and axle kinematics on the roll behavior of the vehicle can be investigated.
a)
b)
2
roll center
1 0
roll center
2 0
Figure 7.10: Roll behavior at cornering: a) without and b) with camber compensation If the wheels only move vertically to the body at jounce and rebound, at fast cornering the wheels will be no longer perpendicular to the track Fig. 7.10 a. The camber angles
1 > 0
and
2 > 0
which leads to a reduction of the maximally transmittable lateral forces. Thus, at more
123
Vehicle Dynamics
sportive vehicles axle kinematics are employed, where the wheels are rotated around the longitudinal axis at jounce and rebound, compensation can be achieved with
1 = 1 (z1 ) and 2 = 2 (z2 ). Hereby, a camber 1 0 and 2 0. Fig. 7.10 b. By the rotation of
the wheels around the longitudinal axis on jounce and rebound, the wheel contact points are moved outwards, i.e against the lateral force. By this, a 'roll support' is achieved that reduces the body roll.
Figure 7.11: Roll axis The 'roll center' can be constructed from the lateral motion of the wheel contact points
Q1
and
Q2 ,
Fig. 7.10. The line through the roll center at the front and rear axle is called
PR0-P PF0+PF
PR0+PR PF0-PF
PR0-PR
Figure 7.12: Wheel loads for a exible and a rigid chassis The roll angle of a vehicle during cornering depends on the roll stiness of the axle and on the position of the roll center. Dierent axle layouts at the front and rear axle may result in dierent roll angles of the front and rear part of the chassis, Fig. 7.12.
124
On most passenger cars the chassis is rather sti. Hence, front and rear part of the chassis are forced by an internal torque to an overall chassis roll angle. This torque aects the wheel loads and generates dierent wheel load dierences at the front and rear axle. Due to the degressive inuence of the wheel load to longitudinal and lateral tire forces the steering tendency of a vehicle can be aected.
a2 a1 Fy2 x2 v C y2
y0
Fy1 x1
xB y1
yB
Figure 7.13: Simple handling model The main vehicle motions take place in a horizontal plane dened by the earth-xed frame
0,
Fig. 7.13. The tire forces at the wheels of one axle are combined to one resulting
force. Tire torques, rolling resistance, and aerodynamic forces and torques, applied at the vehicle, are not taken into consideration.
7.3.2 Kinematics
The vehicle velocity at the center of gravity can be expressed easily in the body xed frame
(7.71)
where
125
Vehicle Dynamics
The velocity vectors and the unit vectors in longitudinal and lateral direction of the axles are needed for the computation of the lateral slips. One gets
ex1 ,B
and
cos = sin , 0
ey1 ,B
sin = cos , 0 0 = 1 , 0
v01,B
v cos = v sin + a1 0
(7.72)
ex2 ,B
where and
1 = 0 , 0
ey2 ,B
v02,B
v cos = v sin a2 , 0
(7.73)
a1
and
a2
are the distances from the center of gravity to the front and rear axle,
Fy = cS sy ,
where
(7.74)
cS
is dened by
Eq. (3.88). Because the vehicle is neither accelerated nor decelerated, the rolling condition is fullled at each wheel
rD = eT v0P . x
Here,
(7.75)
rD
v0P
ex
vy = eT v0P y
and the rolling condition Eq. (7.75), the lateral slip can be calculated from
(7.76)
eT v0P y , sy = T | ex v0P |
with
(7.77)
ey
labeling the unit vector in the lateral direction direction of the tire. So, the lateral
(7.78)
sy1 =
(7.79)
126
sy2 =
The yaw velocity of the vehicle considered to be small
(7.80)
are
| a1 | | | 1
|v| ;
and
| a2 | ||
(7.81) (7.82)
Because the side slip angle always labels the smaller angle between the velocity vector and the vehicle longitudinal axis, instead of
v sin v
v sin |v|
has to be used. Now, Eqs. (7.79) and (7.80) result in
(7.83)
sy1 =
and
(7.84)
sy2 = +
(7.85)
where the consequences of Eqs. (7.81), (7.82), and (7.83) were already taken into consideration.
vC,B
The angular velocity is given by
can be approximated by
(7.86)
0F,B
0 = 0 .
(7.87)
If the vehicle accelerations are also expressed in the vehicle xed frame will nd at constant vehicle speed
xF , yF , zF ,
one
aC,B
v = const and with neglecting small higher-order terms 0 = 0F,B vC,B + vC,B = v + |v| . (7.88) 0
127
Vehicle Dynamics
0F,B
where the substitution
0 = 0 , =
(7.89)
(7.90)
was used. The linear momentum in the lateral direction of the vehicle reads as
(7.91)
Fy1 cos
Fy1 ,
describes the vehicle mass. With Eq. (7.90) the angular momentum yields
= a1 Fy1 a2 Fy2 ,
where
(7.92)
names the inertia of vehicle around the vertical axis. With the linear description
of the lateral forces Eq. (7.78) and the lateral slips Eqs. (7.84), (7.85), one gets from Eqs. (7.91) and (7.92) two coupled, but linear rst order dierential equations
cS2 m |v|
a2 |v|
a2 |v|
v |v| ,
(7.93)
a2 cS2
(7.94)
u .
(7.95)
If a system can be at least approximatively described by a linear state equation, stability, steady state solutions, transient response, and optimal controlling can be calculated with classic methods of system dynamics.
7.3.6 Stability
7.3.6.1 Eigenvalues
The homogeneous state equation
x = Ax
(7.96)
128
xh (t) = x0 et
is inserted into Eq. (7.96), the homogeneous equation will remain
(7.97)
( E A) x0 = 0 .
One gets non-trivial solutions
(7.98)
x0 = 0
for (7.99)
det | E A| = 0 .
The eigenvalues
(7.100)
approximates the eigen-dynamics of vehicles at low speeds, Eq. (7.100) has the eigenvalues
The matrix in
1v0 =
and
2v0 =
(7.101)
The eigenvalues are real and always negative independent from the driving direction. Thus, vehicles possess an asymptotically stable driving behavior at low speed!
v ,
Av 0 = a2 cS2 a1 cS1
v |v| . 0
(7.102)
Using Eq. (7.102) one receives from Eq. (7.99) the relation
2 v +
(7.103)
129
Vehicle Dynamics
1,2v =
When driving forward with
(7.104)
v > 0,
(7.105)
holds. Then however, one eigenvalue is positive, and the system is unstable. Two zero-
1 = 0
and
2 = 0
a1 cS1 = a2 cS2 .
to an unstable behavior. With
(7.106)
The driving behavior is indierent then. Slight parameter variations, however, can lead
or
(7.107)
v > 0
the root argument in Eq. (7.104) becomes negative. Then, the eigenvalues
are imaginary, and disturbances lead to undamped vibrations. To avoid instability, highspeed vehicles have to satisfy the condition Eq. (7.107). The root argument in Eq. (7.104) changes at backward driving its sign. Hence, a vehicle showing stable driving behavior at forward driving becomes unstable at fast backward driving!
xst = const.
or
= 0 , a stable system reaches steady state after a certain time. xst = 0, the state equation Eq. (7.95) is reduced to a system of linear A xst = B u .
(7.108)
equations
B,
st
= 0
|v| (cS1 + cS2 ) st + (m v |v| + a1 cS1 a2 cS2 ) st = v cS1 0 , |v| (a1 cS1 a2 cS2 ) st + (a2 cS1 + a2 cS2 ) st = v a1 cS1 0 , 1 2
where the rst equation has been multiplied by The solution can be derived from
m |v| |v|
|v|.
v cS1 0 st = v a1 cS1 0
m v |v| + a1 cS1 a2 cS2 a2 cS1 + a2 cS2 1 2 m v |v| + a1 cS1 a2 cS2 a2 cS1 + a2 cS2 1 2
(7.111)
130
v cS1 0
and
(7.113)
detD = 0,
st
a1 v cS2 (a1 + a2 ) = 0 , |v| a + a + m v |v| a2 cS2 a1 cS1 1 2 cS1 cS2 (a1 + a2 ) a2 m v |v| v a1 + a2 + m v |v| a2 cS2 a1 cS1 0 . cS1 cS2 (a1 + a2 ) v 0 a1 + a2
(7.114)
st =
(7.115)
v > 0, the steady state side slip angle starts with the kinematic v a2 0 |v| a1 + a2
and
v0 st =
v0 st =
(7.116)
vst=0 =
a2 cS2 (a1 + a2 ) a1 m
(7.117)
the side slip angle changes the sign. Using the kinematic value of the yaw velocity Eq. (7.115) can be written as
st =
v a1 + a2 1 +
1 |v| v v 2 vch
0 ,
(7.118)
where
vch =
(7.119)
is called the 'characteristic' speed of the vehicle. In Fig. 7.14 the side slip angle driving speed
, and the driven curve radius R are plotted versus the v . The steering angle has been set to 0 = 1.4321 , in order to let the vehicle
131
Vehicle Dynamics
2 0 -2 [deg] -4 -6 -8 -10 0
radius of curvrature
a1 =1.2 m; a2 =1.3 m;
cS1 = 80 000 N m;
cS2
R0 = 100 m
at
v 0. v . R
st =
(7.120)
Some concepts for an additional steering of the rear axle were trying to keep the side slip angle of the vehicle, measured at the center of the vehicle to zero by an appropriate steering or controlling. Due to numerous problems, production stage could not yet be reached.
0 =
(7.121)
The rst term is the Ackermann steering angle, which follows from Eq. (7.2) with the wheel base
a = a1 + a2
tan 0 0 .
The Ackermann-steering angle provides a good approximation for slowly moving vehicles, because the second expression in Eq. (7.121) becomes very small at the value of
a2 cS2 a1 cS1
the necessary steering angle diers from the Ackermann-steering angle at higher speeds. The dierence is proportional to the lateral acceleration
ay =
v2 . R
(7.122)
132
cS1 , cS2 .
At
gravity
v > 0 the a1 , a2
steering tendency of a vehicle is dened by the position of the center of and the cornering stinesses at the axles The various steering
understeering
A 0 > 0
or
or
neutral
A 0 = 0
or
or
oversteering
A 0 < 0
or
or
st = 0, st = 0
Eq. (7.120), the equations of motion Eq. (7.91) and Eq. (7.92) can be dissolved for the
Fy1st = Fy2st =
a2 v2 , m a1 + a2 R a1 v2 m a1 + a2 R
and
or
a1 Fy2st . = a2 Fy1st
(7.123)
(7.124)
sst yA1
and
sst yA2
label the steady state lateral slips at the axles. Now, from Eqs. (7.123)
or
(7.125)
during steady
state cornering the slip angles at the front axle are larger than the slip angles at the rear st st axle, sy1 > sy2 . So, the steering tendency can also be determined from the slip angle at the axles.
clin = cS Fz1 S1
and
clin = cS Fz2 . S2
(7.126)
133
Vehicle Dynamics
G = mg
Fz1 =
a2 G a1 + a2
and
.Fz2 =
a1 G a1 + a2
(7.127)
a1 clin = a1 cS S1
and
a2 G a1 + a2 a1 G. a1 + a2
(7.128)
a2 clin = a2 cS S2
(7.129)
Thus, a vehicle with identical tires would be steering neutrally at a linear inuence of the wheel load on the cornering stiness, because of
a1 clin = a2 clin S1 S2
(7.130)
The lateral force is applied behind the center of the contact patch at the caster oset v distance. Hence, the lever arms of the lateral forces change to a1 a1 n and |v| L1 v a2 a2 + |v| nL1 , which will stabilize the vehicle, independently from the driving direction.
6 5 4
Fz [N ]
0 1000 2000 3000 4000 5000 6000 7000
Fy [N ]
0 758 1438 2043 2576 3039 3434 3762 4025
Fy [kN]
3 2 1 0
Fz [kN]
Figure 7.15: Lateral force
8000
Fy
Fz
At a real tire, a degressive inuence of the wheel load on the tire forces is observed, Fig. 7.15. According to Eq. (7.92) the rotation of the vehicle is stable, if the torque from the lateral forces
Fy1
and
Fy2
is aligning, i.e.
(7.131)
a = 2.45 m the axle loads Fz1 = 4000 N and Fz2 = 3000 N yield the position of the center of gravity a1 = 1.05 m and a2 = 1.40 m. At equal slip on front and rear axle one gets from the table in 7.15 Fy1 = 2576 N and Fy2 = 2043 N .
134
The
a1 < a 2
the rear axle is larger than at the front axle (a1 can only be achieved with dierent tires.
At increasing lateral acceleration the vehicle is more and more supported by the outer wheels. The wheel load dierences can dier at a suciently rigid vehicle body, because of dierent kinematics (roll support) or dierent roll stiness. Due to the degressive inuence of wheel load, the lateral force at an axle decreases with increasing wheel load dierence. If the wheel load is split more strongly at the front axle than at the rear axle, the lateral force potential at the front axle will decrease more than at the rear axle and the vehicle will become more stable with an increasing lateral force, i.e. more understeering.
135
speed
Typical
wheel loads [kN] 0 0.2 0.4 0.6 0.8 lateral acceleration [g]
R = 100 m
The vehicle is under-steering and thus stable according to Eq. (7.121) with Eq. (7.122). The inclination in the diagram steering angle versus lateral velocity decides about the steering tendency and stability behavior.
136
The nonlinear inuence of the wheel load on the tire performance is here used to design a vehicle that is weakly stable, but sensitive to steer input in the lower range of lateral acceleration, and is very stable but less sensitive to steer input in limit conditions. With the increase of the lateral acceleration the roll angle becomes larger. The overturning torque is intercepted by according wheel load dierences between the outer and inner wheels. With a suciently rigid frame the use of an anti roll bar at the front axle allows to increase the wheel load dierence there and to decrease it at the rear axle accordingly. Thus, the digressive inuence of the wheel load on the tire properties, cornering stiness and maximum possible lateral force, is stressed more strongly at the front axle, and the vehicle becomes more under-steering and stable at increasing lateral acceleration, until it drifts out of the curve over the front axle in the limit situation. Problems occur at front driven vehicles, because due to the demand for traction, the front axle cannot be relieved at will. Having a suciently large test site, the steady state cornering maneuver can also be carried out at constant speed. There, the steering wheel is slowly turned until the vehicle reaches the limit range. That way also weakly motorized vehicles can be tested at high lateral accelerations.
0.2
0.8
Figure 8.2: Step steer input Not the angle at the steering wheel is the decisive factor for the driving behavior, but the steering angle at the wheels, which can dier from the steering wheel angle because of elasticities, friction inuences, and a servo-support. At very fast steering movements, also the dynamics of the tire forces plays an important role. In practice, a step steer input is usually only used to judge vehicles subjectively. Exceeds in yaw velocity, roll angle, and especially sideslip angle are felt as annoying.
137
Vehicle Dynamics
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
[t]
v = 100 km/h
The vehicle under consideration behaves dynamically very well, Fig. 8.3. Almost no overshoots occur in the time history of the roll angle and the lateral acceleration. However, small overshoots can be noticed at yaw the velocity and the sideslip angle.
z = z(x, y)
z1 = z1 (s1 ) ,
and
z2 = z2 (s2 )
(8.1)
or one uses an isotropic track. The statistic properties are direction-independent at an isotropic track. Then, a two-dimensional track can be approximated by a single random process
z = z(x, y)
z = z(s) ;
(8.2)
138
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
A normally distributed, stationary and ergodic random process characterized by the rst two expectation values, the mean value
z = z(s)
is completely
1 mz = lim s 2s
s
and the correlation function
z(s) ds
(8.3)
1 Rzz () = lim s 2s
s
A vanishing mean value
z(s) z(s ) ds .
(8.4)
mz = 0
Rzz () = Rzz () ,
and
(8.5)
z(s)
ds
(8.6)
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 Fourier-transformation
Rzz () =
0
where
Szz () cos() d
(8.7)
denotes the space circular frequency. With Eq. (8.7) follows from Eq. (8.6)
Rzz (0) =
0
Szz () d .
(8.8)
139
Vehicle Dynamics
Thus, the psd gives information, how the variance is compiled from the single frequency shares. The power spectral densities of real tracks can be approximated by the relation
Szz () = S0
where the reference frequency is xed to
(8.9)
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, the reference6 psd S0 lies within the range from 1 10 m3 to 100 106 m3 and the waviness can be approximated by
w = 2.
m ; w=2 1000
m ; w=2
500
500
-2
[deg] 2
-2
[deg] 2
Figure 8.5: Steering activity on dierent roads A straightforward drive upon an uneven track makes continuous steering corrections necessary. The histograms of the steering angle at a driving speed of steering activity is often used to judge a vehicle in practice.
v = 90 km/h
are
displayed in Fig. 8.5. The track quality is reected in the amount of steering actions. The
140
[m]
vehicle
mass
[kg]
center of gravity
inertias
[kg m2 ]
empty
12 500
fully laden
18 000
12 500 0 0 0 155 000 0 0 0 155 000 15 400 0 250 0 200 550 0 250 0 202 160
has a double wishbone single wheel suspension. The twin-tire rear axle with the track i o widths sh = 2.152 m and sh = 1.492 m is guided by two longitudinal links and an a-arm. The air-springs are tted to load variations via a niveau-control.
Figure 8.6: Roll steer: - - front, rear While the kinematics at the front axle hardly cause steering movements at roll motions, the kinematics at the rear axle are tuned in a way to cause a notable roll steering eect, Fig. 8.6.
100 m-Radius.
pied vehicle is slightly more understeering than the empty one. The higher wheel loads cause greater tire aligning torques and increase the degressive wheel load inuence on the increase of the lateral forces. Additionally roll steering at the rear axle occurs.
141
Vehicle Dynamics
LW
vehicle course
150 100 50 0 0.1 0.2 0.3 0.4 lateral acceleration a y [g] wheel loads [kN]
-100
0 [m]
100
100
100
50
50
Both vehicles can not be kept on the given radius in the limit range. 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
ay 0.35 g .
v = 80 km/h
can be seen in
Fig. 8.8. To achieve comparable acceleration values in steady state condition, the step steer input was done at the empty vehicle with = 90 and at the fully occupied one with = 135 . The steady state roll angle is 50% larger at the fully occupied bus than at the empty one. By the niveau-control, the air spring stiness increases with the load. Because the damper eect remains unchanged, the fully laden vehicle is not damped as well as the empty one. This results in larger overshoots in the time histories of the lateral acceleration, the yaw angular velocity, and the sideslip angle.
142
yaw velocity 8 6 4 2
Z [deg/s]
8 6 4 2 0 0
roll angle
[deg]
2 1 0 -1 -2
[deg]
[s] 6
[s] 6
- - coach empty,
100 m
radius are
143
Vehicle Dynamics
trailing arm
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
6 4 2
Figure 8.10: Steady state cornering, semi-trailing arm, - - single wishbone, arm
trailing
The vehicle with a trailing arm rear axle is, compared to the serial car, more understeering. The lack of roll support at the rear axle also causes a larger roll angle.
144
Index
Ackermann geometry, 109 Ackermann steering angle, 109, 132 Aerodynamic forces, 96 Air resistance, 96 All wheel drive, 118 Anti dive, 108 Anti roll bar, 122 Anti squat, 108 Anti-lock-system, 102 Auto-correlation, 13 Axle kinematics, 108 Double wishbone, 7 McPherson, 7 Multi-link, 7 Axle load, 95 Axle suspension Solid axle, 55 Twist beam, 56 Bend angle, 113, 116 Brake pitch angle, 103 Brake pitch pole, 108 Braking force distribution, 100 Camber angle, 6, 24 Camber compensation, 121, 124 Camber slip, 49 Caster, 8, 9 Characteristic speed, 131 Climbing capacity, 97 Comfort, 72 Contact point, 24 Cornering resistance, 117, 118 Cornering stiness, 41, 133 Curvature gradient, 114
Damping rate, 76 Deviation, 13 Disturbance-reaction problem, 84 Disturbing force lever, 9 Down forces, 96 Downhill capacity, 97 Drag link, 57, 58 Drive pitch angle, 103 Driver, 3 Driving force distribution, 100 Driving safety, 72 Dynamic axle load, 95 Dynamic force elements, 63 Dynamic wheel loads, 94 Eective value, 13 Eigenvalues, 128 Environment, 4 First harmonic oscillation, 63 Fourier-approximation, 64 Frequency domain, 63 Friction, 97 Front wheel drive, 98, 118 Generalized uid mass, 70 Grade, 95 Hydro-mount, 69 Kingpin, 7 Kingpin Angle, 8 Lateral acceleration, 121, 132 Lateral force, 126 Lateral slip, 126
Vehicle Dynamics
Ljapunov equation, 84 Load, 4 Maximum acceleration, 97, 98 Maximum deceleration, 97, 99 Mean value, 13 Natural frequency, 76 Optimal damping, 81, 87 Oversteering, 133 Overturning limit, 118 Parallel track model, 10 Parallel tracks, 138 Pinion, 57 Pivot pole, 109 Power spectral density, 139 Quarter car model, 87, 90 Rack, 57 Random road prole, 138 Rear wheel drive, 98, 118 Reference frames Ground xed, 5 Inertial, 5 Vehicle xed, 5 Relative damping rate, 77 Ride comfort, 83 Ride safety, 83 Road, 10, 23 Roll axis, 124 Roll center, 124 Roll steer, 141 Roll stiness, 120 Roll support, 121, 124 Rolling condition, 126 Safety, 72 Side slip angle, 109 Sky hook damper, 87 Space requirement, 110 Spring rate, 78 Stability, 128 State equation, 128 State matrix, 88
State vector, 88 Steady state cornering, 117, 136, 141 Steering activity, 140 Steering angle, 114 Steering box, 57, 58 Steering lever, 58 Steering oset, 9 Steering system Drag link steering system, 58 Lever arm, 57 Rack and pinion, 57 Steering tendency, 125, 132 Step steer input, 137, 143 Suspension model, 75 Suspension spring rate, 78 System response, 63 Tilting condition, 97 Tire Bore slip, 52 Bore torque, 20, 50 Camber angle, 24 Camber inuence, 48 Characteristics, 52 Circumferential direction, 24 Composites, 19 Contact forces, 20 Contact patch, 20 Contact point, 23 Contact point velocity, 31 Contact torques, 20 Cornering stiness, 41 Deection, 26 Deformation velocity, 31 Development, 19 Dynamic oset, 41 Dynamic radius, 32, 33 Friction coecient, 45 Lateral direction, 24 Lateral force, 20 Lateral force characteristics, 41 Lateral force distribution, 40 Lateral slip, 40 Lateral velocity, 31
ii
Lift o, 88 Linear model, 126 Loaded radius, 24, 32 Longitudinal force, 20, 38, 39 Longitudinal force characteristics, 39 Longitudinal force distribution, 39 Longitudinal slip, 39 Longitudinal velocity, 31 Model, 52 Normal force, 20 Pneumatic trail, 41 Radial damping, 35 Radial direction, 24 Radial stiness, 34, 121 Rolling resistance, 20, 36, 37 Self aligning torque, 20, 41 Sliding velocity, 40 Static radius, 24, 32, 33 Tilting torque, 20 Track normal, 24, 26 Transport velocity, 33 Tread deection, 38 Tread particles, 37 Unloaded radius, 32 Vertical force, 34 Wheel load inuence, 41 Tire Model Kinematic, 109 Linear, 133 TMeasy, 52 Toe angle, 6 Toe-in angle, 6 Track, 23 Track curvature, 114 Track normal, 5 Track radius, 114 Track width, 109, 121 Trailer, 112, 115 Understeering, 133 Variance, 13 Vehicle, 3 Vehicle comfort, 72 Vehicle dynamics, 2
Vehicle model, 75, 90, 94, 103, 112, 121, Vertical dynamics, 72
Angular velocity, 50 Wheel base, 109 Wheel load, 20 Wheel loads, 94 Wheel rotation axis, 5 Wheel Suspension Semi-trailing arm, 143 Single wishbone, 143 Trailing arm, 143 Wheel suspension Central control arm, 56 Double wishbone, 55 McPherson, 55 Multi-Link, 55 Semi-trailing arm, 56 SLA, 56 Yaw angle, 112, 115 Yaw velocity, 126
iii