The Basic Nature of Vehicle Understeer-Oversteer

Downloaded from SAE International by University of Edinburgh, Thursday, August 09, 2018
650085
THE WORDS "UNDERSTEER" and "oversteer" have been
used for years by automotive engineers. Numerous papers
pertaining to the subject of vehicle handling and stability
refer to understeer-oversteer as generally known terms. A
The Basic Nature of
few authors (1-4)* define these terms, but their definitions
are limited to the steady-state condition and are not appli
cable to transient conditions.
Vehicle Understeer-
This paper defines understeer-oversteer in the transient
as well as in the steady-state condition under various types Oversteer
of inputs. It shows the relationship between understeer-over
steer and directional and oscillatory stability. It employs
the computer program evolved from the equations of mo Walter Bergman
tions shown in the Appendix, to analyze the basic nature of Ford Motor Co.
understeer-oversteer. It develops new methods which per
mit the quantitative determination of vehicle understeer and
vided by the tum velocity. Turn velocity is equal to the
stability. It discusses the direct and indirect understeering
algebraic sum of the yaw velocity w and side-slip velocity
and oversteering effects produced by individual design fac
<jJ . Therefore, the radius of the curvature of the path of
tors.
travel is
VEHICLE STEER TERMINOLOGY
w + oj
(1)
Vehicle steer is the control of vehicle course by means
of steering application. At any given instance the vehicle
where v is equal to the forward velocity.
course is determined by the course angle, which is an angle
Side-slip velocity in cornering has about the same phys
between the velocity vector at any point of a vehicle and
the reference axis fixed in the plane of the road. As a re ical meaning as wheel slip during acceleration or braking
sult of the lateral flexibility of pneumatic tires, an angle and modifies vehicle behavior in the transient condition
exists between the velocity vector and the longitudinal axis only.
of a vehicle which is called side-slip angle. The side-slip In steady-state comering the side-slip angle remains
angle determines the "nose-in" or "nose-out" attitude of constant and therefore the side-slip velocity isequalto zero
a vehicle in comering and therefore is often called attitude
RESULTANT
angle. The angle between the longitudinal axis of a ve SIDE-SUP ANGLE OR VELOCITY
hicle and the fixed reference axis is called yaw angle ATTITUDE
ANGLE
(Fig. 1).
Derivatives of the course angle, side-slip angle, and yaw
angle with respect to time are called turn velocity, side
slip velocity, and yaw velocity, respectively. Tumvelocity
determines the actual tum of a vehicle with respect to the
road and describes the curvature of the path of travel, the
radius of which is equal to the vehicle forward velocity di-
REFERENCE AXIS
'Numbers in parentheses designate References at end of Fig. 1 - Angles determining motion of a vehicle in plane
paper. of the road surface
ABSTRACT ploys a computer program evolved from the equations of

motion, shown in the comprehensive Appendix, to analyze
This paper gives a comprehensive analysis of vehicle un the basic nature of understeer-oversteer. New procedures
dersteer-oversteer, utilizing theoretical and experimental are outlined for quantitative determination of vehicle un
approaches. It departs from the conventional by defining dersteer and stability. The paper discusses the relationships
understeer-oversteer in the transient as well as in the steady- between understeer, oscillatory stability, and the subjective
state condition under various types of Inputs. The relation " feel" of passenger automobiles. Understeer rate and yaw
ship befween understeer-oversteer and directional and oscil damping are stated to be the major factors controlling vehi
latory stability is stated within this concept. The paper cle handling characteristics under transient conditions. The
introduces new definitions, based on transient condition anal paper discusses, in detail, the direct and indirect under-
ysis, for direct and indirect understeer, directional and os steering-oversteering effects produced by individual vehicle
cillatory stability, side-slip velocity, and others. It e m design factors.
387
388 W. BERGMAN
and the turn velocity is equal to the yaw velocity. Hence, Rearranging terms in the above expression we obtain:
the radius of curvature of the path of travel in steady-state
cornering is O) V
(5')
V f
(2)
The transitional response of a vehicle is determined pri

The center of turn in steady-state cornering is determined marily by the side-slip velocity, which can be either posi
as a point of intersection of the perpendiculars to the veloc - tive or negative. It is positive if the tum velocity is larger
ity vectors at each tire contact point. The steady-state ra than the yaw velocity.
dius of the curvature of the path of travel can be determined Since side-slip velocity is usually negative during an in
from the geometrical relationship shown in Fig. 2 as crease of steer angle, the radius of curvature at the path of
travel becomes:
R (3)
R= (6)
where: W + (-OJ ) LO
s
jf = Wheelbase
6 = Wheelbase angle
Therefore, the radius of curvature of the path of travel
The wheelbase angle is an angle between the lines con
in a transient condition during an increase of steer angle is
necting the steady-state turn center with the centers of the
larger than the corresponding steady-state radius, and the
front and rear axles. Combining Eqs. 2 and 3, steady-state
four turn centers of the tires are located on the extension of
yaw velocity can be expressed as
the perpendiculars to their respective velocity vectors, be
yond the point of their intersection. Hence, in a transient
(4) condition each tire is turning about its own individual turn
center, as shown in Fig. 4.
It is seen that, in steady-state cornering, yaw velocity Assuming that the vehicle body is rigid, one might sup
is proportional to the wheelbase angle. pose that every point of the body is turning about a common
Since in steady-state cornering, side-slip velocity is equal center. The only turn center which is common for all points
to zero, the yaw velocity becomes equal to the turn velocity. of the car body is the point of intersection of the perpen
dicular to the velocity vectors at each tire contact point.
Therefore, turn velocity is also proportional to the wheel-
In a transient maneuver then, a vehicle body is turning about
base angle. In the case of a hypothetical vehicle equipped
the same tum center as in a steady-state condition but each
with an ideal Ackermann steer geometry, the wheelbase
tire negotiates a turn about individual tum centers which
angle 6 is equal to the front steer angle 5 (Fig. 3). There
do not coincide with the vehicle turn center.
fore, steady-state yaw velocity for a vehicle with Acker
When the steady-state condition is reached, turn centers
mann steer geometry becomes:
for all four tires merge into one common center, which co
incides with the vehicle body turn center.
CJ = — 5 (5)
i f
8 . = FRONT STEER ANGLE
x = R(jtan. 02
£-x = Rotan.;9| 0 = WHEELBASE ANGLE
^ I
ton./3|+tan.;32 0, +$2
"°-T
TUHN CKNTKIi
Fig. 2 - Determination of position of radius of turn in Fig. 3 - Ackermann

steady-state cornering TURN CENTER steer geometry
BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 389
GENERAL DEFINITIONS OF UNDERSTEER-OVERSTEER AND put can be presented as a yaw moment and lateral force ap
STABILITY plied at the center of gravity. This lateral force generates
a secondary yaw moment which tends to turn and yaw the
Understeer and oversteer are transient responses of a ve vehicle away from the force in the case of an understeering
hicle and are defined as changes of vehicle yaw produced response or toward this force in the case of an oversteering
by any change of lateral acceleration or by any external response. Therefore, understeer assists yaw produced by an
lateral force applied at the center of gravity. aerodynamic force and oversteer resists it. If a lateral force
A vehicle is said to have a neutral steer response if the applied at the center of gravity does not produce a change
change of lateral acceleration or the change of the external of yaw, the vehicle has a neutral response.
lateral force produces no change in its yaw. When the lat Understeer- Oversteer Produced by Steering Inputs - Steer
eral acceleration is increa.sed on a moving vehicle, if its ing input can be expressed similarly to wind-force input in
yaw velocity increases, the vehicle is said to oversteer; if terms of lateral force and yawing moment. The yawing mo
the yaw velocity decreases, the vehicle understeers. The ment generated by the steering input tends to turn the ve
understeering-oversteering properties of a vehicle vary widely hicle in the direction in which the front wheels are turning.
with changes in operating conditions and cannot be expressed The lateral force originated from a steering application pro
quantitatively unless these conditions are specified. Under duces a secondary yawing moment which tends to resist or
steer and oversteer can be originated by disturbance inputs assist the turn of a vehicle. A vehicle has an understeering
and by steer inputs. Disturbance inputs are unbalanced lat response if the secondary yawing moment tends to resist the
eral components of inertia forces or aerodynamic forces ap turn of the vehicle and an oversteering response if it tends
plied to a vehicle in a cornering maneuver or in a straight to assist it. A vehicle which develops no secondary yaw mo
ahead motion. ment is said to have a neutral steer response. Steering in
Understeer-Oversteer Due to Inertia-Force Input - Un puts are usually applied in order to prevent uncontrollable
balanced lateral components of inertia forces originate by change of vehicle course caused by disturbance inputs, and
a change of lateral acceleration, which can be accomplished control vehicle course.
by a change of the forward velocity of a vehicle initially Understeer-Oversteer Due to Steering Inputs Applied to
negotiating a steady-state cornering maneuver. Understeer Control Disturbances - In the case when steering input is
and oversteer are defined as a resultant change of the radius employed to counteract a disturbance input expressed in
of the vehicle path of travel. An increase or decrease of terms of lateral acceleration, understeer-oversteer are ex
the path radius indicates that the vehicle response is under- pressed in terms of the relationship between the lateral ac
steering or oversteering, respectively. No change of path celeration and the steer angle. A vehicle is said to have an
radius occurs if the response is neutral. understeering or oversteering response when an increase of
Understeer-Oversteer Produced by Aerodynamic Forces - lateral acceleration requires an increase or decrease of the
Since the center of pressure on most modern automobiles is steer angle, respectively, and a neutral steer response when
located in front of the center of gravity, the lateral wind- a change of lateral acceleration requires no change of steer
force component produces a moment which tends to turn a angle.
vehicle away from the wind force. Lateral wind-force in- Understeer-Oversteer Due to Steering Control Inputs - In
the case when the steering input is used to control vehicle
course, understeer-oversteer are defined in terms of the re
lationship between the steer angle and the side-slip velocity.
A vehicle is said to have a neutral steer response if the side
slip velocity is equal to zero, and an understeering response
if the side-slip velocity is negative during an increase of
steer angle and positive during a decrease of steer angle.
The side-slip velocity of an understeering vehicle becomes
zero when the steady-state condition is reached. A change
of vehicle side-slip velocity during a lane changing maneu
ver is shown in Fig. 5.
A vehicle is said to have an oversteering response if the
side-slip velocity is negative during an increase as well as
a decrease of steer angle. The side-slip velocity of an over-
VEHICLE BODY
TURN CENTER steering vehicle shows a progressive increase until it reaches
its maximum value, equal to yaw velocity at breakaway.
Fig. 4 - Vehicle and Breakaway is defined as the yawing of a vehicle without
tires turn centers in turning, a condition existing when the side-slip velocity
transient maneuver reaches 100%.
TIRES TURN
during entry of a turn CENTERS General Definitions of Vehicle Stability - Understeer and
oversteer are closely associated with vehicle stability. There
390 W. BERGMAN
are two types of stability which should be considered in defin or continuous steering or disturbance inputs generates force
ing vehicle behavior: directional stability and oscillatory and moments tending to restore the initial steady-state con
stability. Generally speaking, an understeering vehicle is dition or to reach a new steady-state condition within a fi
a directionally stable vehicle and an oversteering vehicle nite time interval. A vehicle is directionally unstable if
is a directionally unstable vehicle. Under most environ any deviation from the initial condition produced by t e m
mental and operating conditions a commercially acceptable porary or continuous steering or disturbance inputs generates
understeering vehicle is an oscillatorily stable vehicle. How forces and moments which tend to increase an initial devia
ever, an understeering vehicle might become oscillatorily tion and cause an ever increasing response until the vehicle
unstable due to an excessive rear axle roll understeer "built" is spun-out. Therefore, the directionally unstable vehicle
into it or due to unfavorable environmental or operating con develops only a transient response and never can reach the
ditions. An oversteering vehicle does not develop an oscil steady-state condition.
latory yaw response and therefore criteria of oscillatory sta Vehicle directional stability or instability is expressed
bility are not applicable in this case. A neutral steer ve mathematically in terms of directional stability rate. The
hicle represents a borderline case between oscillatory and directionally stable vehicle has a positive and the direc
nonoscillatory yaw response and therefore can be treated as tionally unstable vehicle a negative rate.
a vehicle with critical yaw damping. A vehicle is oscillatorily stable if any temporary or con
A vehicle is directionally stable if any deviation from tinuous steer or disturbance input produces an oscillatory
the prescribed steady-state condition produced by temporary yaw response of progressively decreasing amplitude and r e
stores the initial steady-state condition or reaches a new
steady-state condition within a finite time interval. A v e
hicle is oscillatorily unstable if any temporary or continuous
^ \ V SIDE-SIJP
\ ^ VELOCITY steer or disturbance input produces an oscillatory yaw re
- -
sponse of ever increasing amplitude until breakaway occurs.
- - Vehicle oscillatory stability and instability are expressed
^ ^ ^—TURN y ^ ^
SIDE-SUP VELOCITY •^ VELOCITY / ^ ^ ^ mathematically in terms of yaw damping. (See section on
TURN VELOCITY, ,
YAW VELOCITY- "Oscillatory Stability".) A vehicle with positive yaw damp
DEG/SEC \ «
ing is said to be oscillatorily stable, whereas with negative
- -
- yaw damping it is oscillatorily unstable.
-
YAW '— ^ ^ ~
VELOCITY ^ DIRECTIONAL STABILITY AND STEADY-STATE UNDER
- - STEER
-
1 1 1 _l L^ * ^ 1 1 1
1.00 1.25 Directional Stability - The directional stability rate is
TIME - SEC
the measure of vehicle resistance to a change from one
Fig. 5 - Side-slip velocity of an understeering vehicle steady-state condition to another, a change caused by unit
during lane changing maneuver input lateral force brought about by wind force or road ir-
28 ■
• • •
24
• • •
20
• / •
•
1
16 ji * •
LATERAL
COMPONENT
• ••
O F WIND •
VELOCITY-MPH
12
• • •
'
8
• >
•
4 ~ '^ *
Fig. 6 - Lateral component of wind velo
0 1 1 1 1 ' 1 1 J
city versus steering wheel angle vehicle H 12 IG 20 24
4-door sedan equipped with manual steering S T E E R I N G W H E E L ANGLE - DEGREES

regularities; the latter is not treated in this paper. This re both the wheelbase angle and the change of yaw velocity
sistance is expressed as a ratio of: per unit input of steer angle are smaller than these corre
sponding values in a neutral steer vehicle.
change of the lateral component of wind force
Substituting Eq. 7 into Eq. 4 we obtain an expression for
change of vehicle steady-state condition
the directional controllability rate of an understeering ve
the change of condition being expressed in terms of course hicle:
angle, or since under actual driving conditions steering cor 5 + a -a +6
U) V
rections are made to prevent a change of course angle, in u^ t f r r'
(8)
terms of the vehicle steer or steering wheel angle. Thisrate
f \ f
change can be determined as a slope to a curve of "wind
and the steady-state understeer rate is the difference be
force" versus "steering wheel angle" as shown in Fig. 6.
tween Eq. 8 and Eq. 5';
The scatter of the individual points obtained from the road
tests illustrate changes of aerodynamic yawing moment
caused by changes of the angle of attack. u = —2-j =-T- (5 + a - a + 5 ) (9)
6 «6 ' t f r r
Directional Controllability - The directional controlla
bility rate is, like the directional stability rate, a measure where w and w are the steady-state yaw velocities of a
of vehicle change from one steady-state condition to an o u '
other, a change caused by unit input steer angle. This change neutral and an understeering vehicle, respectively.
is expressed as a ratio of: Steady-State Understeer - Steady-state understeer, be
sides being present in responses as the above, exhibits itself
change of vehicle steady-state condition in vehicle steady-state response to a change of lateral a c
change of steer angle celeration with the front wheels 'fixed." Steady-state un
the change of condition being expressed in terms of course dersteer rate in such a case is defined as:
angle or yaw velocity. change of vehicle steady-state condition
The understeer properties of the vehicle affect both the change of lateral acceleration
above rates, but while it is very difficult to isolate the pre
cise effect of understeer in the directional stability rate, its
effect in the directional controllability rate can be deter
mined readily as the difference between the response of an
understeering and of a neutral steer vehicle, as produced by
a unit input of steer angle.
A neutral steer vehicle always has an ideal Ackermann
steer geometry. Eq. 5' expresses this response mathemati
cally:
(5-)
which states that a vehicle has a neutral response if the di C'llAXGt OF
sricEKixc:
rectional controllability rate is equal to the forward velocity WHEEL ANGLK-
DKGKKEb
divided by the wheelbase (4).
In the case of the understeering vehicle shown in Fig. 8,
the wheelbase angle becomes:
6 = 5 - 6 - a +a - 6 C^)
f t f r r
where:
6 = Front steer angle
5 = Front steer angle due to the geometrical and com

pliance toe change
a = Front slip angle
a = Rear slip angle

LATERAL ACUELERATION - G's
6 = Rear steer angle (axle roll steer)
Fig. 7 - Handling characteristics of Ford and competitive
Since in steady-state cornering, an understeering vehicle station wagons in steady-state cornering. Full Load. Left
develops larger slip angles on the front than on the reartires, turn at 82 ft radius
392 W. BERGMAN
The change in the vehicle condition can be expressed of gravity of a linearly moving vehicle. This force can be
as a change in yaw velocity, in a case where such change either an inertia force or an external force, but while such
is brought about by a change of forward velocity, but if a forces are readily applied on a computer simulated vehicle,
steering correction is made to prevent a change of the path in actuality they can only be applied insome artificial way,
of travel, the change in vehicle condition can be expressed such as by means of a rocket installed atthe center of grav
in terms of a change of steer angle or steering wheel angle ity of a vehicle (5). Under the action of this force, a change
(2, 3). In cases like these, the steady-state understeer rate of vehicle lateral acceleration results because of the change
can be obtained as a tangent to the curves "steering wheel 2
from linear to curvilinear travel and is equal to a = v / R
angle or yaw velocity" versus " lateral acceleration," as in
developed when the curvilinear steady-state condition is
Fig. 7. Such curves allow comparison of different vehicles,
reached at radius R. The disturbing lateral force F is also
providing their overall steering ratios do not differ appre
expressed in terms of lateral acceleration: o£ = F / m . The
ciably.
The steady-state understeer rate can best be expressed, steady-state understeer rate u, then, is a ratio of a / a ,
' o
however, as a dimensionless ratio where both the vehicle
which is convenient to express in per cent.
steady-state condition and the lateral acceleration are stated
When a = a (100% static understeer rate) itfoUowsthat
in terms of acceleration units. Let us determine this rate o
by considering a lateral force suddenly applied at the center
V (10)
:— from which: -—=
m R
mv
which states that a vehicle exhibits a 100% steady-state un
dersteer rate when a lateral force continuously applied at its
center of gravity produces a change of curvature of its path
of travel (1/R) equal to this lateral force divided by the prod
uct, ' m a s s times velocity squared."
Since a neutral steer vehicle continues a linear motion
despite the force application, the radius R equals<», which
in the expression a = v / R results in a = o. Hence a v e
hicle has a steady-state neutral steer response if its steady-
state understeer rate is equal to zero.
The dimensionless ratio u will be used as the term for
the steady-state understeer rate in the oscillatory stability
discussion that follows and for the quantitative analysis of
the effect of individual design factors producing vehicle
understeer in section on "Quantitative Analysis of Under
Fig. 8 - Steer geometry of understeering vehicle steer Produced by Individual Design Factors."
YAW
iCChJLERATION
DKGREE/SEC^
Fig. 9 - Vehicle dynamic yaw re

sponse. Calculated values
OSCILLATORY STABILITY tion of yaw damping is caused by an excessive increase of

rear axle roll understeer. The results show agreement even
Oscillatory Stability - Oscillatory stability is distinguished
though the inputs are different.
from directional stability by being oscillatory and transient
Vehicle velocity has a considerable effect on the values
in nature, and is a measure of the "fishtailing" which a c
of yaw damping, as shown in Fig. 1 1 .
companies rapid maneuvers. This transitional, oscillatory
Close agreement is shown in Fig. 12 between an increase
response is best expressed in terms of yaw damping, which
of yaw damping and progressive improvement in vehicle
is a measure of how quickly the vehicle stabilizes from a
"feel" during lane changing maneuvers at 60 mph. By plot
sudden steer or disturbance input.
ting the subjective rating versus the steady-state understeer
Because of the nonlinearity of vehicle yaw, the yaw os
rate of these vehicles, as shown in Fig. 13, an optimum rat
cillations do not show ideal linear or exponential decay.
ing is reached at a steady-state understeer rate of about
But, as an approximation, yaw damping can be obtained
70%. It is interesting to note that an increase of rear axle
from the average ratio of two consecutive amplitudes:
roll understeer (change from vehicle A to vehicle B, Figs.
9, 10, 12, and 13) produced a drastic decrease of yaw damp
n-l\
expressed in per cent ing with only a negligible increase of steady-state under
steer rate, and resulted in a car rated as commercially un
acceptable.
where: Oscillatory Wind Stability - Oscillatory wind stability,
a = Amplitude in terms of yaw damping, is a measure of the car's ability
n = Number of half cycles
Fig. 9 shows curves plotted from the calculated values
of yaw acceleration after the force input is released, for an
experimental vehicle subjected to a momentary lateral force
of 0.2 g at the center of gravity held until maximum yaw
acceleration was reached, then gradually reduced to zero.
The calculated value of yaw dampingforvehicle Ais88.3<yo
and for vehicle B is 36.2<7o. and illustrates how an excessive
amount of rear axle roll understeer can have an adverse
effect on vehicle transitional behavior. YAW
DAMPING
Oscillatory Steering Stability - Oscillatory steering sta PKH CENT
bility, in terms of yaw damping, is a measure of the car's

ability to damp out a momentary steer input. The follow
ing test procedure was developed to measure the values of
this response. Yaw oscillations were generated in a lin
early moving vehicle by applying a steering input and then
quickly releasing the steering wheel to let the vehicle and
the steering wheel oscillate freely. Fig. 10 shows the os
cillographic recordings of the yaw oscillations after the
steering wheel is released. They have good correlation with
the calculated results of Fig. 9, in that a significant reduc Fig. 11 - Effect of velocity on yaw damping. Vehicle D
VEHICLK A ORIGINAL DESIGN -

YAW DAMPING 73 PERCENT
REAR AXLE STEER INCREASED

FROM . 058 TO . 171 DEGREE PER
DEGREE OF ROLL - YAW
DAMPING 36.3 PERCENT
ACCKl.KIUTION —
DKt; si-:c-
Fig. 10 - Vehicle dynamic yaw re

TIMb;-St:CONDS sponse for experimental vehicle
394 W. BERGMAN
to damp out a momentary wind-force input. Because yaw

EACH O F T i l t : T E S T VEEIICLEE
A . B , C , D . E . HAS 3 RATINGS
• damping is usually not sufficient to keep the vehicle's lat
GIVEN BY INDIVIDUAL • eral displacements within the lane width, oscillatory steer
EVALUATIONS • /•
C ing corrections are necessary to maintain vehicle course.
• / 1 The resultant yaw oscillations, therefore, are caused by var
_ B E T - 'ER
A iable side-wind and steering inputs which occur randomly,
" making mathematical analysis difficult without drastically
SL'BJKCTIVK nfinni.ri T INI-
R)
RATING simplifying the assumptions.
In view of these difficulties, an empirical approach ap
• -
WOI SE • pears to be more suitable. A simple test method was de
• vised to determine quantitatively vehicle oscillatory sta
bility in cross winds. This method employs a cumulative
~ ^^ • ~
E steering counter consisting of 20 electric contacts at 3 deg
^^^* increments concentrically mounted with the steering shaft.
1 1 A slider attached to the steering shaft slides over the con
tacts and actuates the counters. Tests were conducted by
YAW DAMPING
IN P E R C E N T
driving a vehicle an arbitrary distance of 10 or 4 miles and
back, the counters continuously counting the steering cor
Fig. 12 - Subjective rating versus yaw damping - Ford rections made by the driver during each run. The results
and competitive vehicles of a test run are presented in Fig. 14 by a curve which shows
I 1 1
c
• A
" •
• ^ ^ ^ — ■ • -
•
BETTKR
~
SUBJECTIVE BOUDKK LINK ^
•
RATING
E yT
- • X _
uoRst; y
\•
X *
•
1 1 Fig. 13 - Subjective rating versus static
understeer rate - Ford and competitive
STEADY STATE UNDERSTEER RATE
IN PERCENT vehicles
Ol"
OCCniRKNCKS
Fig. 14 - Sample record of directional

LEl-T stability measurements by means of steer
STEERING WHEEL ANGLE - DEGREES ing counters
the number of occurrences versus steering wheel ampli affecting vehicle oscillatory wind stability. It has been
tude. found from numerous tests conducted on a vehicle driven
The area bounded by the curve represents the total angu by the same driver at various cross wind velocities ranging
lar travel of the steering wheel during the duration of the from 2-19 mph, that the total angular travel of the steering
test run and can be used as an indicator of vehicle oscilla wheel employed per four miles of vehicle travel varies from
tory wind stability. However, this area is greatly affected 800-3000 deg (Fig. 16).
by such uncontrollable factors as the driver and changes of Another criterion of vehicle oscillatory wind stability is
wind velocity. In order to determine the effect of the driver the so-called, wide-center-feel angle, which quantitatively
on vehicle oscillatory wind stability, 10 consecutive test expresses the subjective wide-center-feel at the steering
runs were conducted along a four mile long section of high wheel. The wide-center-feel angle can be determined from
way. The drivers were changed after each mn so that every the slope of the curve, "steering wheel angle" versus "num
test run was conducted by a different driver. Fig. 15 shows ber of occurrences," as illustrated in Fig. 14.
that the number of steering corrections employed by indi It has been found that both the wide-center-feel angle
vidual drivers varied considerably. It appears thatthe num and the area under the curve are important criteria of ve
ber of steering corrections is somewhat related to psycho hicle oscillatory wind stability. However, neither of these
logical type, since a phlegmatic type of driver usually made stability criteria can be used numerically by itself unless
considerably less corrections than an excitable type. the uncontrollable effects of driver and wind are eliminated,
A change of wind velocity is another important variable so oscillatory wind stability measurements are conducted on
140 - -
VEHICLE B 1 SPEED - 65 MPll
[ CROSSWI.ND - 10 MPll
1 DISTANCE OF TRAVEL 4 MILES
\2U — —
100 - -
y / \ DRIVER B
80
NORTHBOUND / \ / \ 1 \ SOUTHBOUND
TRAVEL / \ / \ / \ TRAVEL
60 -
NUMBER
OF
OCCURENCES
to
/ X'
A ^
A
A i''^-'^ V -
/1 ''
/<ît'/
/ •'
» DRIVER A
\
«■<
y^
If i
«\ '
•V, 1 •»\\
\
0
1 1 , ,1 sVr- "\, 1 1 !-•
Fig. 15 - Effect of driver on vehicle
S T E E R I N G W H E E L ANGLE — D E G R E E S stability
VEHICLE A ( S P E E D - 6 0 MPH
DRIVER A 1 DISTANCE O F TRAVEL 4 MILES
NUMBER
OF
OCCURRENCES
RIGHT
Fig. 16 - Effect of crosswind on vehicle
STEERING WHEEL ANGLE - DEGREES stability
396 W. BERGMAN
a relative basis. Two vehicles were used simultaneously to zero, and the vehicle will not turn but assume a new
during testing, one of which was a test and the other a base linear path of travel at an angle to its original path (Fig.
line vehicle. To avoid the effect of the driver, the drivers 19), indicating that the vehicle has a neutral steer. There
were rotated between the test and the baseline vehicle. fore, the location of the resultant lateral tire force with re
Each test consisted of 10 repetitive runs, each four miles spect to the horizontal location of the e.g. of the vehicle
long in one direction. is the most important factor determining understeer-over
In this way, the determination of oscillatory wind sta steer, although it will be shown later in this paper that the
bility comparisons between vehicles and of the effect of in resultant lateral tire force is not the sole source of vehicle
dividual components was achieved, by using the ratio of the understeer; other forces and moments contribute to it.
areas under the curve "steering wheel angle" versus "num An understeering or oversteering vehicle, then, due to
ber of occurrences," or of the wide-center-feel angles. the effect produced by the disturbing force, describes a
curved path of travel, as a result of which an inertia force
MECHANICS OF UNDERSTEER-OVERSTEER is generated which naturally acts toward the outside of the
curvature. At the same time, the disturbing force acts to
Let us analyze the mechanics of understeer-oversteer ward the inside of the curvature on an understeering vehicle
produced by a lateral force continuously applied at the cen and toward the outside of the curvature on an oversteering
ter of gravity of a vehicle moving straight ahead at a con vehicle. Therefore, an understeering vehicle develops an
stant speed. We assume that the driver holds the steering inertia force which opposes the disturbing force with the ef
wheel in a fixed, straight ahead position and allows the car fect that the resultant lateral force at the e.g. of the ve
to change its direction of motion under the action of the hicle, composed of the disturbing force and the inertia
disturbing lateral force. According to the commonly a c force, is reduced, which produces a corresponding reduction
cepted definition, an understeering vehicle tums away from of the lateral tire forces (Fig. 20). The yawing moment pro-
and an oversteering vehicle runs toward the disturbing lat
eral force (1). We assume that at first, the vehicle does not
yaw, and therefore an angle is developed between its longi
DISTUR1!1,N(; FORCK
tudinal axis and the new direction of motion. If the wheels
remain in their initially straight ahead position, it is not
difficult to see that the front and rear tires will operate un
der slip angles.
The slip angles generate lateral forces on the front and
rear tires which act in a direction opposite to the disturb
PATH O F
ing lateral force at the center of gravity of the vehicle. The T H E VKIIICLK
HKSLLTANT
sum of the front and rear lateral tire forces is equal to the TIHK r O R C E
resultant lateral tire force, which together with the disturb Fig. 18 - Oversteering response of vehicle
ing lateral force, creates a couple that generates a yawing
moment that turns the vehicle either away from or toward
the disturbing lateral force. If the resultant lateral tire force DISTURBING FORCE
is applied behind the center of gravity, the yawing moment

PATH O F
tums the vehicle away from the disturbing lateral force, that THE VEHICLE
is, the vehicle shows an understeering response (Fig. 17); if
this force is applied in front, the yawing moment turns the
vehicle toward the disturbing lateral force, which indicates
oversteer (Fig. 18). If the force coincides with the horizon
tal location of the e.g., the yawing moment will be equal
RESULTANT TIRE FORCE
Fig. 19 - Neutral response of vehicle

DISTURBING F O R C E PATH O F
THE VEHICLE
DISTURBING F O R C E
CG
RESULTANT RESULTANT INERTLA

TIRE FORCE TIRE FORCE FORCE
Fig. 17 - Understeering response of vehicle Fig. 20 - Effect of inertia force on understeering response
duced by the couple composed of these two forces is also A change of distribution of the slip angle values between
reduced, which tends to reduce the yaw of the vehicle and front and rear tires results in the forward shift of the result
resist the increase of the curvature described by the vehicle. ant lateral tire force, which tends to reduce the yawing mo
The inertia force developed during an understeering response ment. Therefore, a change of slip angles distribution tends
of the vehicle tends to stabilize its motion and the transient to reduce understeering response.
understeering response ultimately evolves into a steady-state An oversteering vehicle shows a progressive decrease of
condition. An understeering vehicle, then, is a directionally the front as regards to the rear slip angle, which also tends
stable vehicle. to reduce its response. However, this reduction is insignifi
An oversteering vehicle develops an inertia force which cant compared to the self-energizing effect of the inertia
acts in the same direction as the disturbing force, which in force. It is interesting to note that in the first instant after
creases the resultant lateral force at the e.g. of the vehicle, the application of the disturbing lateral force, all three ve
resulting in a corresponding increase of the resultant lateral hicles (understeering, oversteering, and neutral steer) show
tire force (Fig. 21). The yawing moment increases, which about the same response by turning away from the disturbing
increases the yaw of the vehicle and results in a progres force and describe almost the same radius of turn. Despite
sively increasing curvature of the vehicle's travel. The ve this similarity, the yaw of these vehicles is different. A neu
hicle moves more and more sharply into a turn until a com tral steer vehicle turns without yawing; an understeering
plete breakaway occurs. The inertia force developed during vehicle develops yaw in the direction of turn; however, an
an oversteering response of the vehicle produces instability oversteering vehicle shows a yaw in the direction opposite
of the vehicle and explains why an oversteering vehicle is to that of turn. An oversteering vehicle being directionally
directionally an unstable vehicle. unstable, always shows transient response by increasing the
Inertia force increases not only with an increase of the curvature of the path of travel until it spins out.
curvature of vehicle travel but also with an increase of ve An understeering and neutral steer vehicle reach steady-
locity. Therefore, the effect of inertia force on vehicleun- state conditions at the end of the response; however, the type
dersteering-oversteering response is more pronounced athigh of steady-state response depends on the type of input e m
speeds. Hence, it is essential to conduct comparative eval ployed. Steady-state responses of an understeering and neu
uation of two vehicles at the same speed. tral steer vehicle to various types of inputs are shown in Ta
It is very common to define understeer and oversteer in ble 1.
terms of the relationship between the front and rear slip
angles; when the front is larger than the rear slip angle, the BASIC SOURCE OF UNDERSTEER-OVERSTEER
vehicle is said to be understeering; conversely, when the
rear is larger than the front slip angle, the vehicle is said
The basic source of understeer-oversteer is a yawing mo
to be oversteering (1). However, the vehicle designer, to
ment produced by tire forces about the vertical axis through
increase total vehicle understeer, increases the rear slip
the vehicle center of gravity, which can be broken down
angle by increasing rear axle roll understeer. This contra
into three components:
diction can be explained by the fact that the above defini
tion was originated by observation of vehicle behavior in 1. Yawing moment produced by the resultant lateral tire
steady-state cornering and is not applicable to transient un force.
dersteering response. During the first phase of a transient 2. Yawing moment produced by tire aligning torques.
response, the understeering vehicle develops a larger slip 3. Yawing moment produced by fore-and-aft forces.
angle on the rear tires than on the front tires (Fig. 22). In
approaching the steady-state condition, the trend reverses
itself, and the front slip angle becomes larger than the rear
one. The latter condition is usually maintained during a
steady-state condition following an understeering response .
DlFFFllKNCK BF'rWFFN
I N E l i T U I-OUCE FUONT AND UFAIi -
SLIP ANGLES-DEG. -1.0
PATH OF TRAVEL
DISTURBING FORCE
.50 1.00 1.50 2.00 2.50 3.00
TIME - SEC
RESULTANT
TIRE FORCE
Fig. 22 -.Calculated transient response of understeering
Fig, 21 - Effect of inertia force on oversteering response and oversteering vehicle to lateral force input
398 W. BERGMAN
The first of these sources is the most important; the sec of the rear lateral tire force to the resultant tire force is lar
ond is significant but less important than the first; and the ger than the ratio of the rear weight to the total weight of
third has little significance and is of academic interest only. the vehicle, the resultant tire' force is located behind the
Understeer-Oversteer Produced by Resultant Lateral Tire e.g., which indicates that the vehicle is understeering. Con
Forces - As mentioned previously, the location of the re versely, if the ratio of tire forces is smaller than the weight
sultant lateral tire force with respect to the horizontal lo ratio, the resultant tire force is located in front of the e.g.
cation of the e.g. of the vehicle is the most important factor and the vehicle is oversteering. Equality of these two ratios
determining understeering or oversteering properties. The indicates a neutral steer vehicle. To determine if a vehicle
distance between the point of application of this force and is understeering or oversteering, it is necessary to know the
the centerline of the front axle is determined by the fore- values of the front and rear lateral tire forces.
and-aft distribution of the lateral tire forces. This distance Lateral tire forces are produced by the slip angle and dy
rear lateral tire force namic camber angle, and are modified by lateral and fore-
is equal to ■ ■ times the wheelbase.
resultant lateral tire force and-aft weight transfer forces resulting from traction and
The horizontal location of the e.g. is determined in a sim- braking application (6). The distribution of lateral tire for
rear weight . , , „ ,r , ces can be changed by changing the distribution of slip an
liar manner: ; 7^;— times the wheelbase. If the ratio gles and cornering powers between front and rear tires, and
total weight
Table 1 - Steady-State Responses of Vehicle to Various Inputs
Original
Input Condition Understeering Neutral
Lateral force step Linear Steady state corner Change of course

input (continuous ing angle with linear
application) path path of travel
Lateral force of Change of course Lateral displacement

pulse input angle linear path of path of travel.
(temporary travel of travel New path of travel
application) is parallel to original
at one
Steering step constant Steady state corner Steady state cornering

input (continuous ing
application) speed
Steering pulse Change of course Lateral displacement

input angle with linear of path of travel.
(temporary path of travel New path of travel is
application) parallel to original
one
An increase of Steady- A new steady state No change

forward velocity state cornering. An in
to a new constant cornering crease of turn ra-
value (continuous) dius, displacement
of center of turn
A temporary increase Steady- The same steady No change

of forward velocity state state cornering,
to a new constant cornering Displacement of
value and then a re center of turn
duction to original
velocity
BASIC NATUttE OF VEHICLE UNDERSTEER-OVERSTEER 399
also by changing wheel camber. The slip angle distribution dersteering-oversteering effect from an indirect one. By
is controlled by toe-change, rear axle steer, and compliance means of testing, we can determine only the resultant com
steer. Inflation pressure, dynamic wheel load, and power bined effect produced by an individual factor.
and braking applications are major factors controlling tire We will illustrate the direct and indirect effects by se
cornering power. The effects of some of these factors on lecting a front stabilizer bar as an example. The front sta
the distribution of lateral tire forces will be discussed later. bilizer bar increases front roll rate, thus increasing lateral
Understeer-Oversteer Fhroduced by Aligning Torque - Al weight transfer on the front, which produces an understeering
though, so far, we have treated front and rear lateral tire effect; this is the direct effect. At the same time, the front
force as applied at the centerline of the wheel, the force stabilizer bar increases the overall roll rate and thus reduces
is actually applied behind the centerline of the wheel. The the roll of the vehicle. On a vehicle with roll understeer
moment produced by the lateral tire force about the vertical geometry, the reduction of roll results in a reduction of un
axis through the center of the wheel is defined as aligning dersteer, which can be termed an oversteering effect. There
torque. It is not difficult to see that the aligning torque at fore, the stabilizer bar produces a combined understeering-
each tire tends to turn the vehicle in a direction away from oversteering effect, which consists of a direct understeering
the disturbing force, thus producing an understeering response. effect and an indirect oversteering effect. On modem pas
Understeer produced by aligning torques is additive to senger vehicles which are equipped with low roll-rate sus
the understeer produced by the yawing moment generated pensions and have a considerable amount of roll understeer,
by a couple composed of the resultant lateral tire force and the indirect oversteering effect produced by the front sta
the disturbing lateral force. bilizer bar is very significant. On the other hand, the wide
Understeer-Oversteer Produced by Fore-And-Aft Forces - tires employed on modem cars tend to reduce the sensitiv
The yawing moment generated by fore-and-aft forces such ity of the tire cornering power to lateral weight transfer and
as tire rolling resistance, aerodynamic drag, traction, brak therefore reduce the direct understeering effect produced
ing, and fore-and-aft inertia force produced during accel by the stabilizer bar. The indirect effect produced by the
eration and braking is determined by the distance between front stabilizer bar is often larger than its direct effect, and
the resultant force acting forward and the resultant force the final combined effect produced by the front stabilizer
acting rearward, and is affected by vehicle roll and lateral bar on modern passenger cars is generally an oversteering
weight transfer. The direction of this yawing moment can effect.
be different in various vehicles and also can change with a Before World War II, the tires used on passenger cars were
change of operating conditions. much narrower than modem tires and generated more un
dersteer due to lateral weight transfer. At the same time,
DIRECT AND INDIRECT UNDERSTEERING-OVERSTEERING suspensions were much stiffer and had a higher roll rate,
EFFECTS PRODUCED BY INDIVIDUAL DESIGN FACTORS which, combined with a smaller amount of roll understeer
" built* into the vehicle, resulted in the stabilizer bar hav
Direct and Indirect Understeering-Oversteering Effects - ing a rather insignificant effect on roll understeer. On pre
Individual design factors affect vehicle understeer-oversteer war vehicles, the combined effect produced by the front sta
both directly and indirectly. The direct effect of an indi bilizer bar was an understeering effect (1).
vidual factor on vehicle understeer-oversteer is defined as Due to the fact that most design factors produce both di
a change of vehicle understeering-oversieering response pro rect and indirect understeering-oversteering effects which
duced by this factor only. To determine the direct under- very often oppose each other, the combined effect of these
steering-oversteering effect of an individual factor is pos factors can be understeering, oversteering, or even neutral.
sible only when this factor is completely isolated from other It is also possible that the combined effect of a certain de
factors, which can be achieved by considering this factor sign factor is an understeering effect when it is applied to
applied to a neutral steer vehicle. The indirect effect of one vehicle but can reverse itself and become an oversteer
an individual factor on vehicle understeer-oversteer is de ing effect in the case of another vehicle. It is important
fined as an additional change of vehicle understeering-over- to consider such implications in the case when certain de
steering response generated by other factors as a result of sign features are adapted from other vehicles into a new ve
the direct effect produced by the initial factor. In other hicle design.
words, an individual factor not only produces a direct under- In the following discussion, we will consider the effects
steering-oversteering effect on the vehicle but also affects of individual factors on vehicle understeer-oversteer. Di
other factors. As a result of this effect, other factors pro rect and indirect effects produced by each factor will be
duce an understeering-oversteering effect which is additive discussed separately. Due to space limitation, only a few
to the initial direct understeering-oversteering effect. factors can be included in this discussion.
An indirect understeering-oversteering effect produced Effect of Rear Axle Roll-Steer -
by an individual factor either assists or opposes the direct 1. Direct Effect. Rear axle understeer geometry results
effect; and, therefore, it either increases or decreases the in an increase of the rear slip-angle and therefore in an in
resultant understeering-oversteering effect produced by this crease of lateral force on the rear tires, which causes a rear
factor. It is practically impossible to separate a direct un ward shift of the point of application of the resultant lat-
400 W. BERGMAN
eral tire force, which, in turn, results in a yawing moment negative camber change in jounce and positive camber
that tends to turn the vehicle away from the disturbing force. change in rebound. Hence, geometrical change opposes the
2. Indirect Effect. An increase of the slip-angle result camber change due to roll and therefore produces an over-
ing from rear axle roll steer produces an increase of align steering effect. Reduction of the camber roll understeer
ing torque on the rear tires, which results in an additional produced by geometrical camber change is relatively small
increase of understeer produced by rear axle roll steer. An on most modern passenger cars equipped with an S. L. A.
increase of lateral tire force on the rear tires, resulting from front suspension.
an increase of the rear slip-angle, produces an increase of 2. Indirect Effect. Camber thmst develops a small tor
lateral weight transfer on the rear tires, which results in an que which tends to oppose the aligning torque, thus produc
indirect oversteering effect. This indirect oversteering ef ing an oversteering response. Therefore, the indirect effect
fect is very insignificant and only slightly reduces the di produced by camber thmst opposes its direct effect. How
rect understeering effect produced by rear axle roll-steer. ever, this indirect effect is so insignificant that it can be
Effect of Toe-Change - ignored for all practical purposes.
1. Direct Effect. Toe-change consists of both geometri Camber thmst reduces the lateral force on the front tires,
cal and compliance toe-change; the latter is the major fac which causes a reduction of lateral weight transfer on the
tor. Since compliance toe-change produces toe-out in jounce front and results in an indirect oversteering effect. This in
and toe-in in rebound, it reduces the front slip-angle and direct effect of camber thrast is of the second order of im
reduces the front lateral tire force, which results in an un portance
dersteering effect. Steering gear compliance and flexibil Effect of Weight Distribution -
ity of the front-end structure are the major factors control 1. Direct Effect. Change of wheel load affects tire cor
ling compliance toe-change. nering power and therefore lateral tire force. Due to the
Geometrical toe-change modifies the compliance tbe- nonlinearity of tire characteristics, an increase of wheel load
change. If the geometrical toe-change follows the same produces a less-than-proportional increase of lateral tire
pattern as the compliance toe-change, it is additive and force. If the weight distribution of the car is changed to
increases the understeering effect. In the case when geo- increase the per cent of the weight on the front tires, the
metricaltoe-change has a reverse pattern (toe-in injounce- distribution of the lateral tire force is also changed, but less
-toe-out in rebound), it tends to reduce the front slip-angle, than proportionally to the change in weight. The resultant
reducing the understeering effect due to compliance toe- lateral force therefore shifts rearward away from the c. g.
change. and results in an understeering effect. Conversely, an in
2. Indirect Effect. Regular toe-change (toe-out in jounce crease of the per cent of weight on the rear tires produces
- - t o e - i n in rebound), produces a decrease of front slip-angle an oversteering effect.
and a decrease of aligning torque on the front tires. Thus, 2. Indirect Effect. In discussing an indirect understeer
a direct understeering effect is accompanied and opposed ing - oversteering effect produced by the fore-and-aft weight
by an indirect oversteering effect, the reduction of the distribution, one should distinguish between two cases: de
aligning torque. Therefore, understeer accomplished by sign weight distribution, and change of weight distribution
means of toe-change on the front is less effective than rear on an existing vehicle produced by a change of loading.
axle roU-understeer, because in the latter case the indirect The design weight distribution produces practically no in
effect due to aligning torque is additive to the direct effect. direct effect on vehicle understeer-oversteer. A very small
Another indirect effect produced by toe-change results effect produced by a change of the distribution of lateral
from the decrease of lateral weight transfer on the front tires weight transfer can be ignored.
due to a decrease of the lateral tire forces produced by toe- A change of weight distribution on an existing vehicle
change, which is an indirect oversteering effect that oppos obtained by a change of loading produces a significant in
es the direct understeering effect produced by toe-change. direct understeering-oversteering effect on the vehicle, which
The indirect oversteering effects produced by toe-change results from a change in static design height that produces
are very insignificant and can be ignored in design practice. a change of vehicle roll-steer geometry (rear axle steer,
Effect of Dynamic Camber Change - dynamic camber, and geometrical toe-change). This can
1. Direct Effect. Dynamic camber change on a vehicle sometimes be more important than a direct effect. There
equipped with an independent front suspension is composed fore, the resultant effect produced by a change of weight
of camber change due to the roll of the vehicle and the geo distribution might be just the opposite to the direct effect;
metrical camber change. Camber change due to roll re that is, an increase of load on the rear can result in an in
sults in a lean of the wheels toward the outside of a turn, crease of understeer instead of the generally expected de
crease. An increase of rear axle roll understeer due to an
which produces camber thrast opposing the lateral tire force
increase of passenger load is the major factor responsible
component due to the slip angle, and reduces lateral tire
for this phenomenon (Table 2).
force on the front, thus producing an understeering effect.
Geometrical camber change is defined as camber change Therefore, the effect of design weight distribution on
with respect to the car body. This camber change is deter vehicle understeer-oversteer cannot be determined simply
mined by suspension geometry and usually has a pattern of by the testing of an existing vehicle under various load con-
ditions. The results of such testing might be erroneous and Effect of Power Application -
improper to use in future vehicle design practice. 1. Direct Effect. Power application reduces tire corner
Effect of Lateral Weight Transfer - ing power and the effective lateral coefficient of friction,
1. Direct Effect. The direct effect of lateral weight and therefore causes a reduction of the lateral forces on the
transfer is determined entirely by its distribution between driven tires (6). Power application on a rear-wheel-drive
front and rear, which is determined mainly by roll rates and vehicle reduces lateral tire force on the rear, which results
roll center heights. in a shift of the resultant lateral tire force forward away
Due to the nonlinearity of tire characteristics, lateral from the center of gravity. This produces a yawing m o m
weight transfer results in the reduction of lateral tire forces ent which turns the vehicle toward the disturbing lateral tire
on the front as well as on the rear tires. If the lateral weight force, which is an oversteering response. Conversely, power
transfer on the front is larger than that on the rear, the la application on a front-wheel-drive vehicle results in an un
teral force on the front tires experiences more decrease than dersteering trend. The effect of power application is part
that on the rear tires, and the point of application of the icularly severe when the vehicle is approaching skid condi
resultant lateral tire force shifts away from the e . g . toward tions. This can be observed in severe cornering and also dur
the rear, which produces an understeering response. Con ing normal driving on slippery roads.
versely, an oversteering response will result if the lateral 2. Indirect Effect. Power application has a multiple in
weight transfer on the rear tires is larger than that on the direct effect on a vehicle.
front tires. (a) It produces fore-and-aft weight transfer, which re
2. Indirect Effect. Larger lateral weight transfer on dri sults in either an understeering or oversteering response due
ven tires than on free-rolling tires results in a more rapid to change in lateral tire forces.
reduction of comering power, due to the power application, (b) It produces pitch of the vehicle, which results in a
and produces an oversteering effect on a rear-wheel-drive change of roll-steer and a consequent change of understeer-
vehicle and an understeering effect on a front-wheel-drive oversteer.
vehicle. The indirect effect of lateral weight transfer is (c) On a vehicle equipped with a solid rear axle, an in
usually larger than the direct effect; however, it first be crease of driveshaft torque resulting from power application
comes significant in a severe comering maneuver. In a produces roll of the body and a consequent roll-steer effect.
moderately transient maneuver, such as lane changing, the (d) It produces lateral weight transfer due to driveshaft
resultant effect of lateral weight transfer is very small and torque. This lateral weight transfer takes place only on ve
cannot be detected subjectively even by experienced dri hicles equipped with a soUd rear axle.
vers. This conclusion is derived from a series of road tests Effect of Tread Change -
on various vehicles, where a change of distribution of the 1. Direct Effect. In order to understand the understeer
lateral weight transfer had been accomplished by remov ing - oversteering effect of tread change due to a change of
ing the stabilizer bar from the front and installing it at the lateral weight transfer, it will be appropriate to remind the
rear. No change in the behavior of the vehicle could be no reader of the relationship between tread change and roll cen
ticed by the driver after a series of fore-and-aft shiftings of ter height. The tread change determines the roll center
the stabilizer bar. height. The roll center height can be determined in terms
Contrary to this, a change of the overall roll rate accom of tread change by using the following expression: roll cen
plished by an installation of an additional stabilizer bar or ter height = half the tread times the tangent of half the an
by removal of a stabilizer bar, has always been unmistaka gle between the tangent to the tread change curve and the
bly detected by a driver. This illustrates the significance vertical at design position. Therefore, if the suspension ge
of the overall roll stiffness and the insignificance of its dis ometry were changed so that there was as increase of the
tribution under moderate driving conditions. front roll center height, it would cause an increase of
Table 2 - Values of Understeer Produced By Individual Design

Factors at Various Passenger Loadings
Understeer Values in Per Cent of

Total Understeer of the Original Vehicle
Factor 1 pass, load 3 pass, load 5 pass, load
Design camber 0.20 -0.53 -1.06

Rear axle roll steer 1.15 12.49 23.92
Geometrical toe change 6.59 4.78 3.31
402 W. BERGMAN
lateral weight transfer and result in an understeering ef ment which was caused by vehicle understeering response.
fect. This reduction was brought about by two effects:
2. Indirect Effect. An increase of roll center height pro 1. The inertia force was reduced as a result of a change
duced by an increase of the rate of tread change results in of curvature of the path of travel.
a decrease of the effective moment-arm of the overturning 2. The resultant tire force was relocated toward the cen
couple. Therefore, an increase of tread change causes a ter of gravity, caused by a change of slip angle distribution.
reduction of the moment producing roll of the car body and (Refer to section on " Mechanics of Understeer-Oversteer.)
thus a decrease of roll of the car body. Hence, on a vehicle To determine only the direct understeer effect, it is nec
with roll understeer geometry, an increase of tread change essary to isolate a factor from all other factors. To do this,
produces a decrease of understeer which can be termed an the values of all factors summarized in Table 3 were set
oversteering response. Conversely, a decrease of the rate equal to zero, which reduced the vehicle steady-state un
of tread change on a vehicle with roll understeer results in dersteer rate, as computed in section, "Vehicle Steer Ter
an understeering response. Obviously, any modification of minology," from its original value of 59.93% to 7.23%,
the rate of tread change on a neutral steer vehicle will have which represents the understeer produced by all other fac
no indirect effect on its behavior. tors not listed in Table 3 (for practical purpose, this can
The indirect oversteering-understeering effect produced be considered a neutral vehicle). The understeer value
by the rate of tread change is an essential factor controlling of each design factor was then determined as the difference
vehicle handling behavior. On modern vehicles with their between the steady-state understeer rate of the near-neu
large amount of roll understeer, the indirect oversteering tral vehicle after and prior to addition of this factor. These
effect due to increase of the rate of tread change is a pre values are tabulated in the right hand column of Table 3 .
dominant factor which is usually greater than the direct un Since the indirect understeering effect and the reduction of
dersteering effect resulting from an increase of lateral weight yaw moment are small at 7.23% steady-state understeerrate,
transfer. the understeer values in the right hand column are results
primarily from the direct understeering effect. Table 3 shows
that direct understeer is reduced considerably by indirect
QUANTITATIVE ANALYSIS OF UNDERSTEER PRODUCED
understeer and the reduction of yaw moment; also that fac
BY INDIVIDUAL DESIGN FACTORS
tors become more influential as understeer decreases. Since
indirect understeer and the reduction of yaw moment usually
It is essential for the vehicle designer to determine the
increases as the total understeer increases, understeer values
understeer value produced by each individual design factor
of individual factors are not constants, but are dependent
and establish its relative significance. A method for doing
variables of the total understeer of a vehicle.
this is as follows: the steady-state understeer rate is com
puted for an understeering vehicle. One of the factors is Therefore, the characteristic of an individual compon
then set to zero and the rate computed again. The effect ent, like rear axle roll-steer geometry or toe-change, has
of this factor is considered the difference in rates. The un no meaning in itself, but only as regard the whole. Con
dersteer values for the six most important factors are shown sequently, whether the specifications of a component are
in the left hand column of Table 3. Each value represents " good" or "bad" has validity only in the light of complete
the sum of the direct and indirect understeering-oversteer vehicle consideration. However, the method described above
ing effects. This value also contains a reduction of yaw m o - permits the determination of the relative significance of
individual components or design factors, as long as they are
incorporated in the vehicle. The characteristics of indivi
dual components obtained from two different vehicles are
Table 3 - Understeer Values of Various Design Factors not comparable even on the relative basis. Hence, it is i m
proper to assume that, just because a certain vehicle hand
Steady-State Understeer Rate of les well, a component of that vehicle may be removed bodi
Individual Design Factors in Per ly and incorporated into another vehicle with the same re
Cent, Computed at Vehicle sult it may have produced in its original role.
Steady-state Understeer Rate of
Vehicle handling should be expressed, not in individual
Factor 59.93% 7.23% component characteristics, but as a whole, expressed in terms
of the steady-state understeer rate and yaw damping.
Aligning torque 2.87 7.49
Fore-and-aft CONCLUSIONS
weight distribution 4.02 10.54
Dynamic camber 6.22 19.05 1. Vehicle handling characteristics are to be expressed
Geometrical toe change 2,39 8.30 quantitatively in terms of steady-state understeer rate and
Compliance toe change 9.98 26.06 yaw damping, whose values can be either computed or meas
Rear axle roll steer 5.89 14.57 ured.
2. The subjective feel of a vehicle improves as the yaw
damping increases, also as the steady-state understeer rate hicles. Many thanks to T . A. Layher for development of
increases up to about 70%. the computer program.
3 . Dynamic wind stability can be measured by means
of cumulative steering counters, and expressed relatively REFERENCES
in terms of the total angular travel of the steering wheel or
of the "wide center feel" angle. 1 . Maurice Olley, "Road Manners of the Modern Car, "
4 . The location of the resultant lateral tire force with "Proceedings of the Institution of Automobile Engineers,"
respect to the vehicle center of gravity is the major factor Vol. XLI, (1947), 147.
determining the amount of vehicle understeer-oversteer. 2 . J. R. Ellis, "Understeer and Oversteer," "Automobile
5. Individual design factors produce direct and indirect Engineer, " (May 1963), 178.
understeering-oversteering effects. 3 . J. R. Ellis, "Tyre Mechanics and Vehicle Handling,"
6. The value of understeer produced by an individual "Automotive Design Engineering," (September 1963), 6 1 .
design factor varies with a change of understeer of the whole 4 . R. T . Bundorf and D. L. Nordeen, "Terminology For
vehicle. Therefore, whether the characteristic of an in Vehicle Directional Control and Tire Characteristics" pro
dividual component, like rear axle roll steer or tread change, posed in the article "Vehicle Directional Control Behavior
is "good" or "bad" has no meaning in itself, but only in the Described in More Precise T e r m s , " SAE Journal, (February
light of a complete vehicle consideration. 1964), 26.
5. R. T . Bundorf, D. E, Pollock, a n d M . C . Hardin,
ACKNOWLEDGMENTS "Vehicle Handling Response To Aerodynamic Inputs," P a
per presented at SAE Summer Meeting, Montreal, Canada.
The author is greatly indebted to H . J. Kozicki for in June 10-14. 1963.
valuable technical assistance in the preparation of this pa 6. W. Bergman, "Theoretical Prediction of the Effect
per, and for many helpful suggestions and critical discus of Traction On Cornering Force. " SAE Transactions, Vol.
sions on various parts of the manuscript. The author also 69, (1961), 614.
wishes to thank O. D. Dillman and R. R. Peterson for prac 7. W. Bergman, E. Fox, and E. Saibel, "Dynamics of
tical guidance and encouragement throughout the execution an Automobile In a CorneringManeuver On and Off the High
of the project pertaining to the subject of this paper. way. " Paper presented to first International Conference on
Further acknowledgment is made to A. T . Gale for dis the Mechanics of Soil-Vehicle Systems, June 1961, Torino,
cussions on car handling problems and evaluation of test v e Italy.
Appendix appears on next page.

404 W. BERGMAN
APPENDIX
EQUATIONS O F VEHICLE MOTION
L i s t of Symbols
Vehicle Parameters
/r = Vehicle total weight ( l b . )
IVs = Vehicle s p r u n g weight ( l b . )
a , b , c = C o o r d i n a t e s of C. G. in d o u b l e - p r i m e d coordinate s y s t e m (ft.)
/ = Wheelbase (ft.)
hj = F r o n t roll c e n t e r height (ft.)
h, = R e a r r o l l c e n t e r height (ft.)
tf = F r o n t t r e a d (ft.)
tr - R e a r t r e a d
ho = Static C . G . height
2
/x = Moment of i n e r t i a of the vehicle about X axis (in roll) ( l b . / f t . / s e c . )
2
/„ = Moment of i n e r t i a of the vehicle about Y a x i s (in pitch) ( l b . / f t . / s e c . )
2
h - Moment of i n e r t i a of the vehicle about Z axis (in yaw) ( l b . / f t . / s e c . )
Is. = P r o d u c t of i n e r t i a of the vehicle about X and Z axis
k, - Body r o l l coefficient due to engine t o r q u e (nondimensional)
ke = 1.0 for s y m m e t r i c a l rigid d r i v e n axle s u s p e n s i o n s
ke = 0 for suspended c a r r i e r s u s p e n s i o n s
khj = Independent r e a r s u s p e n s i o n coefficient (nondimensional)
k),d = 1 . 0 for independent r e a r and swing axle
khd - 0 for solid axle s u s p e n s i o n s
kfd = F r o n t - w h e e l - d r i v e coefficient (nondimensional)
ktd = 1 . 0 for front- o r f o u r - w h e e l - d r i v e vehicle
kfd = 0 for r e a r - w h e e l - d r i v e vehicle
krd - R e a r - w h e e l - d r i v e coefficient (nondimensional)
krd = 1 . 0 for r e a r - o r f o u r - w h e e l - d r i v e vehicle
krd = 0 for f r o n t - w h e e l - d r i v e vehicle
(, = R e a r s p r i n g b a s e (ft.)
Subscripts
//. Jr. ri. rr = F r o n t left, front r i g h t , r e a r left
/. r = F r o n t , r e a r
ir°f, w°;r. fv°ri. ^°rr = Statlc whccl loads (lb.)
kf = F r o n t r i d e r a t e ( l b . / f t . )
kr = R e a r r i d e r a t e ( l b . / f t . )
L-'f - F r o n t s t a b i l i z e r b a r r a t e at the wheel ( l b . / f t . )
Vr = R e a r s t a b i l i z e r b a r r a t e ( l b . / f t . )
k„, = F r o n t shock a b s o r b e r r a t e —jf '
<„r - R e a r shock a b s o r b e r r a t e — *
ft.
Ax,. K^-2 = S t e e r i n g c o m p l i a n c e coefficients
^xr = R e a r c o m p l i a n c e coefficient (rad. /lb.)
k„ = Effective m a s s factor (nondimensional)

I = Axle r a t i o (nondimensional)
va = Axle efficiency (nondimensional)
i"/. s ^ = B a c k l a s h , front, r e a r ( r a d . )
*/, *°r = Static c a m b e r angle front, r e a r ( r a d . )
i°f, A°r = Dynamic t o e - i n (in.)
r. = T i r e r o l l i n g r a d i u s u n d e r load (ft.)
r = R e a r s p r i n g t w i s t coefficient
(nondimensional)
)io= Static coefficient of friction between t i r e and r o a d s u r f a c e (nondimensional)
2
Jf = F r o n t a l a r e a of the vehicle (ft. )
lb / s e c
p = A i r density —'/, 4 '
<-x = A e r o d y n a m i c d r a g coefficient (nondimensional)
cy - A e r o d y n a m i c l a t e r a l force coefficient (nondimensional)
fz = A e r o d y n a m i c lift coefficient (nondimensional)
fx, = A e r o d y n a m i c r o l l coefficient (nondimensional)
f„„ = A e r o d y n a m i c pitch coefficient (nondimensional)
c,z - A e r o d y n a m i c yaw coefficient (nondimensional)
., rr, u rr- .. f ■ u inches i n c h e s \
A'— a„ a,, a, - Toc changc coefficients I i n c h e s , —r\— , ~fr^)
X, — a„. ax. a, = R e a r axle s t e e r coclficicnls (It. , nondimensional, ft. ~ )
*' — ao.a,. a, = C a m b e r change coefficients (rad. , r a d . / f t . , r a d . / f t . -)
dt\ = Rate of t r e a d changc coefficients at the t i r e contact point
- j - a„, a,, a. (nondimcnsional, ft. " ^ f t . - ^ )
<i(\' = R a t e of t r e a d change coefficient at the wheel c e n t e r
Tf)i~ "'' "'■ ' (nondimensional, ft."-'^, ft."'^)
406 W. BERGMAN
—) — a„, a,, <jj = Rate of wheelbase change coefficient (nondimensional, ft. , ft. )
2
^s-' — a„, a,, a, = Jounce bumpcr load coefficient (lb. , lb./ft. , lb./ft. )
2
fTs" —fl«.ai, ai = Rebound bumper load coefficient (lb. , lb./ft. , lb./ft. )
f^BJ = Wheel displacement at which jounce bumper contact takes.place (ft.)
i = f,r
f.BK zz Wheel displacement at which rebound bumper contact takes place (ft.)
' = /.'•
// = Front wheel vertical displacement due to load change (ft.)
to" = Rear wheel vertical displacement due to load change (ft.)
Tire P a r a m e t e r s
r = Tire rolling radius without load (in.)

^' 1 = Tire tread radius (in.)
X. = Tire rolling resistance coefficient (nondimensional)
K = Tire rolling resistance speed factor J^'2
^» = Tire power transmission loss factor (nondimensional)

Svo = Apparent tire deflection at zero load (in.)
*s = Tire contact area compression coefficient (nondimensional)
^> = Tire rate -r—*
in.
'■'ô ~ Lateral unit rate at zero load (Ib./in.^)
_3
ks = Interaction coefficient lateral unit rate — load (in. )
_3
*,< == Interaction coefficient lateral unit rate — traction (in. )
io - Rear elastic trail coefficient (nondimensional)
' - Rear elastic trail traction factor (nondimensional)
max = Maximum width of the t i r e contact area (in.)
Independent and Dependent Variables (2 Dependent, 6 Independent)
5- = Steering wheel angle (rad.)

Ta = Axle torque (lb./ft.)
x,y = Coordinates of the geometrical center of the right r e a r wheel
'f' = Yaw angle (rad.)
//(. f/r. fri. trr = Vcrtlcal whecl displacements (ft.)
Vehicle Auxiliary Variables
e = Roll angle (rad.)

"' = Pitch angle (rad.)
!.((, 8/r, 5,;, !,. = Dynamic steer angle (rad.)
^fi, fi/r, 0ri, f)rr = Attltude angle (rad.)
"//, "/r, <»w, ar, = Slip angle (rad.)
^'c, y\, ic = Coordinates of C.G. in X'Y'Z' system (ft.)
"" = Vehicle velocity or velocity at the center of the right r e a r wheel
(ft./sec.)
2
1 = Aerodynamic p r e s s u r e (lb./ft. )
'f/i, fi^fr, i^ru ly.r = Vertical wheel reactions (lb.)
Rfi, Rfr, Rr!, Krr = RoUlng reslstancc (lb.)
Ty, Tr = Traction force (lb.)
Qfi, Qfr, 0,1, Orr = Camber thrust (lb.)
Myi, Mfr, Mr,, Mrr = Tlrc aligning torque (lb./ft.)
lyyiK /j'y,', iy,i\ ii,„>' = Components of the vertical wheel reactions due to compression
of the rubber bumpers
fri = Vertical wheel displacement (ft.)
Tire Auxiliary Variables
'o - Length of the contact area (in.)

s- = Width of the contact area (in.)
z'.' = Effective tractive coefficient (nondimensional)
^'<- = Length of adhesion region (in.)
"h = Distance between the front end of the contact area and the point of
intersection of the elastic line with the lateral deflection curve
produced by friction
^,« = Tire coefficients
cs = Lateral rate ( l b . / i n . )
o
's = Lateral unit rate (lb./in. )
2
/>»,„x = Maximum p r e s s u r e between tire and road surface (lb./in. )
M = Effective coefficient of friction (nondimensional)
I = Elastic trail coefficient (nondimensional)
2r = Vertical tire deflection (in.)
408 W. BERGMAN
Equations of Inertia Forces and Moments
To write dynamic equations of motion, a triple coordinate system has been

selected. (See Fig. 23) The first coordinate system, which is designated as XYZ
system, allows us to describe motion of the automobile in space. This coordinate
system is fixed on the earth and constitutes our inertial frame. For purposes of
simplification, we neglect the contour of the earth within the area in which the ve
hicle moves and consider it a flat, indeformable surface.
The second coordinate system, X'Y'Z' system, is attached to the unsprung

mass of a vehicle and is designated as the single-primed system. The center of the
r e a r right wheel is arbitrarily selected as the origin of the primed system. The X'
axis is parallel to the longitudinal axis of the vehicle, and the Z' axis is perpendicular
to the earth's surface. Therefore, the axes Z and Z', of the fixed and movable s y s
t e m s , are always parallel to each other.
The third coordinate system, X"Y"Z", is called the double-primed system.

It is fixed in the sprung mass of the vehicle. In the normal design position of the
vehicle, the origin of the double-primed system is located directly under the origin
of the single-primed system. Thus, X' and X" axes will be parallel, as will the
Y' and Y" axes. The Z' and Z" axes will coincide.
Since the vertical distance between the axes of the primed and double-primed
systems affects the coordinate C of the centroid of the sprung m a s s , this distance
cannot be selected arbitrarily. It is chosen in such a way that, in the normal design
position, when the planes X'O'Y' and X"0"Y" are parallel to each other, X"0"Y"
intersects the roll axis at the point of intersection of the latter with a perpendicular
through the C.G. Therefore, the vertical distance between planes X'O'Y' and X"0"Y"
in the normal design position becomes: z„ = -r„ + ' ^ ~ "-- + — (1)
(See Fig. 24.) The motion of the vehicle's unsprung mass is described in a fixed
XYZ system by instantaneous xrr, yr, coordinates which determine the position of the
center of the right r e a r wheel by the yaw angle and by the first and second derivatives
of all these quantities. Motions of the other three wheels of the vehicle are determined
subject to the assiunption of a rigid chassis frame.
i h e motion of the sprung mass in relation to the unsprung mass is described

in a movable X'Y'Z' coordinate system. This motion is completely specified by the
deflections of the springs employed in our mathematical model. These will be found
from the z., i = i, 2, 3, 4 and will be considered small quantities.
The -i are the vertical distances from the centers of the wheels to the points
on the car body located on X" axis and Y" axis. In the final equations, the quantities
^' are replaced by quantities U, ' = //, fr, ri, rr , since the latter a r e more commonly
used in automotive design practice. The geometrical relationship between the quan
tities 2< and /. is shown in Fig. 25. The equations of motion have been written first
with respect to the fixed coordinate system XYZ and then transformed to the more
convenient system, the primed system, for calculational purposes. In each system
of axes, the unit vectors a r e designated by ;, y and k respectively.
z'
2"
F,.*«,
Fig. 26 - Forces and moments acting on vehicle in plan

Fig. 23 - Inertial frame view
*f/**r/
Fig. 24 - Determination of coordinate "Z " of vehicle

Fig. 27 - Forces and moments acting on vehicle in front
equipped with solid rear axle view
•- WfZ+Wfr
Fig. 28 - Forces and moments acting on vehicle in side

Fig, 25 - Geometrical relationship between terms Z. and f.
view
410 W, BERGMAN
The derivation of the dynamic equation of motion (inertia forces) has been
presented in the previous paper.'^ Therefore, here we limit ourselves to writing
final equations, after they have been transformed to the primed axes.
The first three dynamic equations of motion a r e given by f = WT„„ (2)

where T;^ is the vector from the origin of the fixed XYZ coordinate system to the
C. G. of the vehicle (point C on Fig. 23).
The quantity ?i7 , representing acceleration of the center of m a s s , can be

expressed as a sum of its components in the primed coordinate system,
roc = i l (v, cui ^ + y. sin ^ — a^' + cip — k//)

+ 2 I ( — ■»■' ■''" ^ + y, cos <!/ — bij/'^ + dif, — c'e) (3)
+ X. ('-' + *^ — "f)
where
e = fJiJzJii (4)
, = i [-2/,,+ (,+,^);;, + {> -;.)/„] (5)

h,. (I — a) . , h/a ,^,
f = A. - - ^ - av + bB - frr - -J- (6)
where
h, = — {ho — r„) 4- Ty, (7)
since
2l = Zo — /rr (8)
We obtain the value V, by differentiating equation (8)

2l = —/rr
Substituting quantities -'}„ for il and *„, 'y„ for x,, j'l ,
we have Jt = Ji (iVr cos ^ + '<>„ sin 4> — t^' + c!^ — i4)

+ y_( — x'rr sin ^ + y„ cos ^ - Ixj,' + a4> - c'e) (3')
+ K ( - frr + he - a^)
The other three dynamic equations of motion describing the sum of inertia
moments about the primed axis a r e
T„ = UJ- l.^)_i;_ + W + UJ'' ~ lx.e)± (9)
Basic Equations of Motion
By equating the inertia t e r m s described above to those describing external

forces, we arrive at the following system of equations, (10) to (15). (See Figs.
26-28.)
2 / - , = -(F/i + Qft) sin 8/, + (-J- - R A COS 8/1 - (/"/, + Q/r) sin 8/, + ( y - -R/r) cos 8/,
- (/"n + Qri) sin tri + ( y - *'') '^"^ '-■' - C " + Q"^ ^»« S"--- + ( y - •*'••■) "^ ' " + ''»?^/ (10)
= IXr, COS \l> + yrr / t n ^ — O^' H ^ "^V" ~ '^]
2/"v = (/"/I + Qn) COS 8/, + ( y - -'f/i) "" »/i + C/r + Qtr) COS I,, + ( y - -RA) sin 8/,
+ (fn + Qrl) COS In + ( y - R r \ sin irl + iFrr + Qrr) COS l „ + ( y " Rr!\ sin i „ + CyqAf ( H )
W I .. . . . IV...\
= ~ - (— x,r stn \f> + y„ COS }l> — b<t>' -{- a ^ »'''*/
W
^F, = fffl + W,r + ffrt + IVrr - IV + CqA, = — ( -)'„ + hB - a^) (12)
g
ILM, = \(,Fii + Q,i) COS 8/, + ( y - Rfl) -"" */' + C"/-- + «?/') cos 5/r + ( y - Rfr\ sin 8/r
+ {Frl + Qrl) COS Irl + ( y - R r \ sin 8,, + (frr + !?rr) COS Srr + ( y - Rrr\ siu 8..1 (z,' + r„)
+ c^qA, W + r j + ;r,, 0, - yd + c,.qA/l = Ij- I,.f
SA/, = (IVrt + JVrr) X,' - (fVfl + Wfr) (/ - X,') - f- {Ffl + Qf,) siu Ijl + ( y - IÂ COS 8/1
- iF,r + Q/r) sin Ifr + ( y " * / ' ) <""■' '/>• " (''-•' + !?>•') ^'« ^'l "•" ( T ~ ^'■') " ^ ' ' ' ^^^^
- (Frr + Qrr) sin 8rr + ( y — *rrj COS irr (Ze' + ''») + Cy„qAfl — CrqA/ ( z / + û) = /yi>'
•LM. = [(/•/, + (?/,) cos 8/, + ( y - ■'J/A ^'« 8/, + (/"/r + !?/r) COS l „ + ( y " ■'^/r) ^''' 8/r] (' - */)
- ["(fn + !?W) fOJ 8w + ( y - *rl) Jin 8,1 + (,F„ + (?rr) fOJ Srr + ( y " *'■'■) ^''' '■''•l * ' '
- [ ( y- /?/l) fo/ 8/, - (/-/, + (?/,) sin S/,J ( ^ ^ ^ y - ^ - ^ c ' )
+ I ( y — RiA cos 8/, — {Ffr + !?/r) sin 8/,J fy.' ^^^ j ^ g .
- r ( y - Rrl\ COS irl - (Frl + Qrl) sin irll Or - Vc')
+ [(y - Rrr) COS Irr - (Frr + (?,r) sin 8 „ 1 y,'

- Msi - M,r - Mrl - Mrr + <r.W^/' " ^ H' - /x.'S
The left parts of equations (10) to (15) describe external forces and moments;
however, the right parts express inertia forces and moments broken down into the
individual components, acting about the respective coordinate axes in a primed system.
In this system, the coordinates of the centroid of the total mass of the vehicle, x'„ y'c,
and z'c , may be expressed by the following relationships:
IV,
v / = a + — c^ (16)
IV
IV.
- v ^ ' ^ * - - ÎV^ " (17)
IV
j;(c + =«) (18)
412 W. BERGMAN
where
e =^ ^ ^ (19)
tr
and * =Yi[r ^f" + (' + ,7) f" + (' - T) ^^'] (20)
The external forces and moments shown in equations (10) to (15) are composed
of tire forces and moments, aerodynamic forces and moments, and moment produced
by driveshaft torque. Tire forces a r e broken down into the following components:
lateral t i r e force generated by slip angle, camber thrust, traction force, rolling r e
sistance, and dynamic wheel load.
The explicit expressions necessary for calculation of these forces will be

shown in the auxiliary equations below:
_ aii rc,i Xai |a,| ifii pmax U , , /■
2
Xbi' + Xbi Xai + Xai'\ £»i_A„, |a, | (.V6, — Xai) (21)
3/„,- / " {,- / „ , (I,- /„,• - Xai)

Xbi + XaiS
O""-^-)]
i = / / , / r , rl, rr
<t>i C',i lai' fi (22)
144 r„.
Ri = fTi (X„ + X, v\r) + —-—

(23)
Ri = /F.- (\„ + \ , f^r) + —~
i = rl, rr
k't (24)
IVf, = fT",, + k,f„ - Y (//r - ffl) + kaljn
[(/■/I + Q/i) COS 8/, + (^ - Rfi'^ sin 8/,J '^'^'

(25)
- [ - (^// + Qn) -'in ^si + ( y - î^ "" 8/(] ^ + ^"ii
Wl, = Wtr" + k,ftr + y (/A - ill) + *«///r + \i.Fh + Qtr) COS 8/r
+ ( y - RS^ sin « / r ] ^ - [ - f.FSr + QSr) sin h!r (26)
+ [ - (/"rf + Sr;) sin Irl + ( y - Rr^ COS 8 r ( J ^ ^ + kjrl
+ [*r/w - y (/rr - / w ) ] */,d - [(/"w + Gr/) r0.t 8w

(27)
+ ( y - / ? r , ) . " « 8., 1 ^ */,rf - [(/"W + !2rl) f".' Sr,
+ ( y - Rrl\ sin irl + (Frr + Qrr) COS irr
+ (^ - R„) sin i.r']— (1 - k,„) + - ^ + IVr,"

BASIC NATURE OF VECHICLE UNDERSTEER-OVERSTEER 413
+ ["- (/"rr + Qrr) si„ S„ + ( y " '^") '^"■' 8 - ' l ^ + *"' f"
+ r*r/rr + y (/rr - /r()l *W + ["(/■., + Qrr) COS i„

(28)
+ ( y - lirr\ sin 8„"1 ^ **<, + [(/•,, + Qrl) COS Sri
+ ( y - Krl) sin ir, + (Frr + Qrr) COS irr
iTr \ -1 Ar Tak,
+ (^ - Rrr) sin irr " ^ " **■') ^ + '*'""
V2 / J tr trt-na
Tire aligning torques are calculated from the following equation:
w "• V^* Xai\a<\ , 7/ I T \

Mi = -ri—r \ rz (2f,/„i — 3/„ + 2Ar„.)
, Mi fmait Si (X6i — ATai) P i , / , , i 1 \
H ^T 2/„ (A;'6, + Xji Xoi + Xâi)
^ oi L
— {Xhi + -âi) («'6t + **„,) — P„i (Xbi + Xai) ( 2 9 )
C,i Xai \ai\ (xbi — Xai) f . , , , . .

Uiiloi (iiloi — Xai) L
— 4 (x»6,- + Xbi Xai + X'ai) + 3/„. (xbi + Xai) — 6{,/'„, \ i
The derivation of the equations of lateral tire force and aligning torque
(equations 21 and 29) is shown in (6).
Equations 21, 22 and 29 contain the auxiliary variables f'.. , /». and
^' , which are calculated from the following equations:
(30)
c'.i = f'.oi - {k„ IM'.I + k,) Wi
' = /'. A. '■/, rr. (31)
h = 2*5 -yjlrlyi
(32)
i = fl, fr, rl, rr.
414 W. BERGMAN
and
2*s V2rX7
(33)
where
(34)
fl,fr
\2 7 /^.
1 = r l . rr
(35)
Tf = i/d —
(36)
r - . ^" (37)
The auxiliary variables x„,, ATI,, <-,i, i,, M;, 8„,, «,
are calculated from the following equations:
c.< hi \ai\ (38)
if
-V..- > hi («.• - Vi.- («i - 1)
Compute /". and .w, from the following equations:
-W, - - r r ^ ["' ' ; ; '•"' (2i,/., + 2,.,)] , . //. I,. ,1. „ (40)
1/ lot,-1 L ii J
instead of using equations 21 and 29,
if .v„, < o
set "" = ^
Mi = 0
and compute /", from the following equation
'"•■ = "• '^■- I S T (41)
instead of using equation 21.

loi r , ^' // Bi V 4z/,S,1 (42)
where
4<Xi pmaxi Si
Ai =
loi
(43)
Cgi -tai |g,-|

5.
It hi — X„
(44)
where l.S fVi

Pmaxi (45)
hi U
ii = V/i»' - (M'.)' - - ^ (46)
!■ = lo + Jlsl M'I (47)
, Wi
5tit — 8«o + , (48)
ai = Si — &\ (49)
The quantities 3< are calculated from the following relationships:
(See Fig. 29.)
Fig. 29 - Position of vehicle in plane of road surface

416 W. BERGMAN
li + v„ sin 0,r
P/i = P/r = /an-' (50)
Vrr COS firr V""/
Vrr hr — hf
0,1 = 0rr = tan-^-. lA H e (51)
The quantities $, contained in equation (49) a r e calculated from the follow
ing e x p r e s s i o n s :
8/, = I'ji - l°n - S'/i - —--"-^'/i -\K,o + K,, {M,i + M,r) + K,, {M,i + M/rY] (52)
2 \ptn\
hr = h'fr + SV + S'/r + -^7^--'SVr-[if^ + AT^rl (M/, + il/^,) + iST,, (M/, + Tl//.)^) (53)
2 \otfr\
Sw = S'-, (1 - hi) - fs%i + Vri + - ' ; J ! - ± - ' ^ U M + A'.. {FH + F„) (54)
S.. = l\ (1 - *w) + (S%, + 5'.. + ^ ^ ~ ^ - T - ' ) ha + Axr (/"ri + F„) (55)
s^ = ^ (56)
(57)
,■ = fljr,rl,ir
The r e a r axle steer angle s/ contained in expressions (54) and (55) i s
computed from the values of the fore-and-aft displacements of the r e a r axle.
8/ = - — 7 ^ (58)
The values of dynamic camber angle «, contained in equation (22) a r e
computed from the following expressions:
*/i = 9 - *°/i - *'/i (59)
«/r = # + *V -I- *Vr (60)

*ri = (e - *%i - «Vi) k,., (61)
*„ = (.e + *rr + *'rr) *»d (62)
The values
" dfi'dfi , representing changes in suspension and steering geometry due to

change of vertical wheel displacements, are expressed by the polynomial
a„ + a, fi + 02 Pi (63)
where the coefficients a». "i. a. a r e determined from the corresponding curves by using
the curve fitting procedure.
The values of the wheel load components ^*, produced by compression of the
rubber bumpers a r e also expressed by a polynomial
fFî = ao + ai (/.• - /») + a2 (Ji - / " ) ' (64)
where /* = the wheel displacement at which bumper contact takes place.
The aerodynamic forces and moments contained in the basic equations of motion
(equations 10-15) a r e described in a conventional manner as a function of air p r e s s u r e
The empirical coefficients employed in aerodynamic terms a r e obtained from

wind tunnel t e s t s .
DISCUSSION
A. G. FONDA static stability, and there is no recognition of the fact that

Giannini Controls Corp. a statically stable oversteer vehicle can even exist. There
is less recognition of the fact that it is considerably more
THE PAPER PRESENTED by Mr. Bergman is most creative and likely to exist than the statically unstable oversteer car. In
provocative, especially in the various definitions given for effect, Mr. Bergman has collapsed the spectrum of possible
understeer, neutral steer, and oversteer. However, one dif- automobiles by ignoring all cases between neutral steer and
ficulty is that Mr. Bergman has so many competing " defini- critical oversteer, and has ignored all distinctions between
tions" that one is finally led to ask, "Will the real defini- the neutral steer case and the critical oversteer case,
tion please stand up?" Mr. Bergman's cmcial error occurs in his fourth paragraph
Answering my own question, the best accepted definition under "Mechanics of Understeer-Oversteer," where he states
is that which distinguishes negative, zero, and positive di- that the oversteer vehicle subjected to side force on its mass
rections of centrifugal force induced by applied side force, (as per the definition) always " moves more and more sharply
as shown by Fig. 17-19 in the paper. This definition is based into a turn until a complete breakaway occurs." The belief
on aircraft concepts, but, apparently was first stated for the that the curvature must increase indefinitely in case of over-
automobile (1). Bergman's remaining "definitions" can steer is a once-common fallacy which has been often re-
best be described as useful descriptions of the natural be- futed. The same vehicle in the same test at a lower speed
havior of an already defined vehicle. will still move more and more sharply into a turn, but will
Even so, some of the descriptions disagree with the def- do so not at a constant or increasing rate but at a contin-
inition or with each other. This is especially true when neu- ually decreasing rate, until further change in negligible and
tral steer and oversteer are being described by Mr. Bergman, a steady path curvature results.
Without disputing his discussions of understeer, there isriiuch An automobile behaves like a feedback system. Imagine
too close an identification between neutral steer and neutral a power boost system in which the output force is added to
41E W. BERGMAN
the input force. The output might either aid or oppose the critical oversteer. For fixed V, both terms vary as l/R,
input, and if aiding, might aid so strongly that the system which is the way a driver follows a winding road at fixed
would run away. Thus, there is a zero-feedback case sep 2
speed. The 5 per unit curvature 1/R increases with V for
arating positive from negative feedback, and a unity-feed
2
back case separating stable positive feedback from unstable understeer, decreases with V for oversteer, and becomes
positive feedback. zero at a specific speed called the critical velocity, if the
In perfect analogy, the automobile' s centrifugal force vehicle is oversteer (and can go fast enough).
may aid or oppose an applied side force, and if aiding, may In solution, we find that
aid so strongly as to cause a runaway response. (If strongly
opposing, there might be some hunting for the final value:
however, this is rare.) The definitions are that for under
V,
CR f C f) L/xM
steer. The centrifugal force opposes the applied force. For

neutral steer it is zero, for oversteer it aids, and for critical At no lesser speed is the oversteer vehicle directionally un
oversteer it aids at a unity rate, so that the total force in stable, and it is unfair criticism to say that it is.
creases uniformly with time, a runaway response. For the nonlinear vehicle, critical oversteer occurs when
there is a peak of 6 versus 1/R, hence, a zero rate of change
Most cars are understeer, and most of the remainder are
of 6 with 1/R, for fixed V. This was shown on Nordeen' s
subcritically oversteer even at their top speed. The main Fig. 4 in the SAE Journal for December 1964. In such a ve
objection to subcritical oversteer is the long time required hicle, V has not held constant, but has gradually descended
to develop the large response: but the time and the final CR
response both become infinite only at critical oversteer, at to the fixed value of V. If instead R is held fixed for a non
some high speed. linear vehicle and V varied, critical oversteer (due to rising
Most cars are nonlinear. They do not maintain fixed V and perhaps descending V ) is reached, not when 5 peaks,
amounts of under/oversteer, as shown by Fig. 7. But it is CR
quite worthwhile to know how the linear automobile behaves. (this only denotes neutral steer), but rather, when 5 returns
The steering angle change for holding a steady radius is to zero. This is the same R and V = V„ which would have
CR
2 been found for the same vehicle by beginning with and hold
A5 = X M V /R a' C
ing the same V while gradually turning sharper. However,
2 at some other fixed value of R, the 6 might reach zero at
due to a balance between the centrifugal force MV /R act some other V, denoting a variation of V with R due to non-
ing at a lever arm x, opposed by the front tire side force CR
C A5 acting at a lever arm a ' . Both moments are taken Unearities. There has been little, if any study of this or any
2
other variation in 5 versus V /R as V and R both vary.
about the neutral steer point, which is where M would have
The transient response of the linear automobile is also
to have been located to make the steering correction zero.
quite simple, using the Fonda method of including initial
The angle 5 may be measured at the front wheels, or in the
conditions while excluding second order transients. When a
linkage, or at the steering wheel. The value of C changes step input of yaw moment (or steering) disturbs the automo
accordingly, and the value of x also accordingly changes if bile, it initially accelerates in yaw according to its yaw in-
2
2 ertia Mk , but eventually reaches a yaw velocity according
some part of the force MV /R is able to cause front wheel to the following total of the yaw dampings due to front tires,
steering. and centrifugal force:
This equation shows that 5 will change in direct propor-
2, 2 (a' f (C^/V) + (b- fiCÎV) + xMV^/Rw^.^^^
tion to the side acceleration V /R, or as V for constant R.
The change will be positive for understeer; the noseheavy
(a'b'C + b ' a ' C ) / V + xMV
car wants to nose out of the turn. It will be negative for over
steer: the tailheavy car wants to move into the turn, or " over
. . . steer" itself. = ( a ' b ' C + xMV ) / V
The total steering angle for the linear car will be
where the identities a'C_ = b'C„ C = C + C„, and w_. ,
'1 ' 2 1 2 final
xM V
= V/R have been substituted. The exponential rise time of
a' C^ R
a mass damper response to torque is the ratio of its mass to
its damping; in the case of the automobile we find that
where 1 is the steering ratio from the front wheels to wher
ever 5 is being measured. The second term of this equation
Mk V
is positive for understeer, zero for neutral steer, negative for t =
e
oversteer, and equals the first term, giving a zero sum, for a ' b ' C + xMV
BASIC NATUI^ OF VEHICLE UNDERSTEER-OVERSTEER 419
This is the exact time to the graphical intersection be A. D. BOSLEY

tween the tangents to the initial and final yaw responses. Chrysler Corp.
Neglecting intervening transients, the response will reach
63% of its final value in the time t , or 95% in the time 31 , THIS PAPER brings to mind the need for the use of stand
e e ardized terminology of the people dealing with car handling.
and so on. In the last few weeks 1 have read four papers, all of which
This response time rises linearly with V in case of neutral have the same characteristics; -that is, heading angle de
steer, at an increasing rate with V in case of oversteer (be scribed or titled with different words. The use of a stand
coming infinite at V = V ), and at a decreasing rate with ardized terminology will make all of our jobs easier and
V for understeer. Compared with neutral steer or oversteer, communication between people working in the area of car
the prime advantages of understeer are: handling much easier. At this point, I would like to ask the
1. As speed increases, the understeer automobile needs author some questions concerning his paper.
more input in order to reach the same path curvature, thus In Fig. 7, he describes the steering characteristics of sev
tending to meet the driver's need for reduced steering sen eral different vehicles when rounding an 82 ft radius corner
sitivity (in effect, higher steering ratio) at high speed. at various lateral forces. Could he comment on the very
2. Especially at higher speeds, the greater initial input sm.all steering angle required by vehicle F and the very large
needed for the same steady state causes the understeer auto discrepancy between the characteristics of vehicle F and the
mobile to'accelerate more rapidly in yaw, and thus more other vehicles shown on this chart? It would appear that the
quickly reach its steady state. The tendency of the response steering angle for vehicle F is very, very small. Secondly,
time to rise with speed is thus reduced. (This is exactly the in this test, since 3 / 1 0 G is only a little above 20 mph, does
same as the main reason for using negative feedback - - t o he not feel that the radius of the circle is somewhat small?
reduce the response time of the system.) Secondly, in the discussion of primary and secondary effects
There is no particular disadvantage of understeer in mod of component changes, the author suggests that the addition
erate amounts, although an "excessively negative feedback" of a sway bar or stabilizer bar in modern car generally pro
oscillation (not shown by the above analysis) can develop duces an oversteering effect. 1 wonder if he really believes
for large amounts of understeer. (Noticeable oscillation be this and if the development people in his group agree.
gins when 6 has been more than doubled relative to >) L/R. There is one portion of the paper that Mr. Bergman pre
This occurs at a "characteristic" speed i/+ a' C 1 L/xM.) sented which, by its inclusion indicates that we all may have
a common problem; that is, the transfer of one component
The usual disadvantage of understeer occurs instead when it
from one vehicle to another under the assumption that the
is of moderate degree, but is obtained through heavy use of
effect of the component in the first vehicle will be dupli
roll steer.
cated in the second. To Mr. Bergman's statement that the
The desired amount of understeer can be obtained in sus
effect of a component in a particular vehicle must be eval
pension design by the use of either roll steer (in phase with
uated in that vehicle and that no assumption can be made
body roll) or deflection steer (in phase with side force). The
because it was satisfactory in another vehicle, (it could be
lag with which roll steer occurs is certainly not enjoyed by
satisfactory in this one), I would certainly be in agreement.
the driver, and is especially noticeable in an emergency,
particularly to a young driver. He can respond before the
body does, in the belief that the vehicle response is already
DONALD L. NORDEEN
complete. Assuming he is executing an S-maneuver or as
General Motors Corp.
suming he has already overcorrected,his next correction is
simultaneous with the delinquent roll understeer effect, and
THE CHOICE of any given chassis parameter depends upon
in the same direction. The roll steer thus makes his over
the values of other parameters. At least 20 parameters in
corrections progressively larger, each attempting to over
fluence the response of the vehicle for lateral accelerations
come the one before. The police report would only say that 2
"the driver lost control of his vehicle." Yet who was a c less than 8 ft/sec . Sixteen independent numbers are needed
tually responsible? Mr. Bergman has not raised this or any in order to characterize the response for the low lateral a c
related objection to roll steer. celeration maneuvers. Even more parameters influence the
Mastery of the basics of under/oversteer allows the de 2
response for lateral accelerations greater than 8 ft/sec .
signer to proceed to the question of how best to provide a There is no magic set of numbers for the chassis parameters
selected amount of steady state understeer. Speaking in the which will guarantee that the handling response will be suit
interest of public safety, roll steer should always be min able Each must be chosen by considering the entire chassis
imized in vehicle design. design and handling concept. The chassis parameters to ob
tain a particular handling response are interrelated as Mr.
REFERENCES
Bergman points out.
1. Lind Walker, "The Automobile Engineer," Vol. 40 There are several aspects of Mr. Bergman's paper which
n. 530, 533 (1950). require clarification. The first of these is the relationship
420 W. BERGMAN
between the center of rotation of the vehicle and the center distinguishing between general definitions and partial def
of curvature of the vehicle path. Mr. Bergman is not con initions pertaining only to specific conditions, Mr. Fonda
sistent in his terminology which describe these two aspects. finally chose one of the partial definitions to be the "real"
He uses the term "tire turn center" to denote center of cur one. He claimed that this "real" definition was first for
vature of the path of the tire. He uses the term " vehicle mulated for an automobile by Lind Walker in 1950. Exam
body turn center" to denote the center of rotation of the ve ination of literature pertaining to this subject reveals that
hicle. These ambiguities make understanding of the paper Huber was the first who introduced the basic concept of this
very difficult. This discussion refers to Fig. 4 in Mr. Berg definition in Europe in 1940 and Maurice OUey was the first
m a n ' s paper. For the vehicle to be rigid as Mr. Bergman to employ a similar definition in this country in 1947. Care
assumes, both the center of rotation of the vehicle and the ful study of Walker's publication reveals that contrary to
center of curvature of the vehicle path are instantaneous cen' Mr. Fonda's claims, it contains no illustrations similar to
ters. These centers become stationary and the same in a my Figs. 17-19. Referring to the remaining definitions as
steady state turning maneuver. The centers of rotation of " descriptions," Mr. Fonda makes a crucial statement that
the sprung and unsprung masses will be different: therefore, "some of the descriptions disagree with the definition or with
it is my opinion that the center of rotation of the unsprung each other." Without even attempting to substantiate his
mass is a common center for all points in the unsprung mass. general criticism, Mr. Fonda concentrated his remarks on
The centers of rotation of the individual tires are the same my single statement that an oversteering vehicle is a di-
point rather than four individual centers as described in Mr. rectionally unstable vehicle. Apparently, Fonda did not re
Bergman's paper. If the tires are rotating at a given instant alize that his " real" definition formulated by Walker also
about different centers, the distance (wheelbase or tread) be states that an oversteering vehicle is unstable (1). There
tween these two points must be changing which violates the fore, Mr. Fonda criticizes the concept which he himself se
rigid nature of the vehicle. The important thing is that the lected to be the only " real" one.
center of rotation of the vehicle is not the same as the cen
Mr. Fonda feels that it is very important to discuss the
ter of the path curvature. Generally, during the initial por
behavior of a statically stable oversteer vehicle. He says
tion of the transient response, the radius to the center of ro
that such a vehicle "is considerably more likely to existthan
tation will be smaller than the radius to the center of path
a statically (directionally) unstable oversteer car." I would
curvature.
agree with Fonda if we were discussing sport cars. However,
The concept of understeer, neutral steer and oversteer are in my paper I was concerned primarily with the handling be
qualitative terms in the sense of positive, zero, and nega havior of passenger cars, in which case a discussion of the
tive. No rating scale is generally given these terms. Mr. concept of a stable oversteer vehicle is completely irrelevant.
Bergman's understeer rate is quantitative. Quantitative meas
Extensive tests conducted on a fleet of 1965 passenger
ures of understeer are desired: however, understeer rate as
cars manufactured in the United States showed that under
defined by Mr. Bergman ishighly influenced by vehicle speed,
typical operating conditions (dry road surfaces, no excessive
and hence may be difficult to measure for some kinds of ve
power application), all vehicles displayed an understeering
hicles. It also is an insensitive measure of vehicle under
behavior within a range of lateral acceleration up to 0.5 g.
steer for vehicles possessing a large amount of understeer.
Understeer showed a progressive increase as the vehicle ve
A comparison of experimental and computed responses locity in a constant radius of turn was increasing. Some of
is shown in Fig. 9 and 10 in the paper. It is important to those vehicles, which possessed sufficient power and did not
validate analytical models, if the results obtained from these overflood the carburetor in a severe cornering maneuver,
models are to be useful. The mathematical model appar could be deliberately broken loose and put into an oversteer
ently used in Mr. Bergman's paper is a position control model condition by an excessive power application. The transition
where the steering wheel is held fixed, but the steering link from understeer to oversteer was very rapid, and the vehicles
age compliance is included. The experimental results shown moved more and more sharply into a turn. None of those
in Fig. 10 are for a free control vehicle where the steering vehicles displayed "stable subcritical oversteer" as described
wheel is not fixed. The motions of the steering wheel due by Mr. Fonda. Therefore, in the case of modern passenger cars,
to the dynamics of the steering system will change the forces the spectrum between neutral steer and the critical over
developed at the front of the vehicle, and therefore, will steer condition completely collapsed.
alter the measured yaw acceleration and the vehicle. The
responses shown in Fig. 9 and 10 are similar by coincidence. After completing his criticism, Mr. Fonda attempts to con
The mathematical model used in the analysis does not de template the vehicle's handling behavior by using an ex
scribe the system which was measured in Fig. 10. tremely oversimplified linear mathematical model. Such
an imaginary "vehicle" will naturally display subcritical
AUTHOR' S CLOSURE oversteer behavior, if the lever arm X is negative and the
TO DISCUSSION absolute value of the second term in his equation is larger
than the value of the first term. This condition could be
APPARENTLY MR. FONDA failed to comprehend the broad achieved by a simple "design modification" change of
definition of understeer-oversteer given in my paper. Not weight distribution. Therefore, Mr. Fonda's imaginary "car"
understeers when it is noseheavy and oversteers when it is vehicle and not with the center of curvature of the tire path.
tailheavy. Although he recognizes the existence of two instantaneous
Unfortunately, the behavior of a real vehicle is not as centers, center of rotation of the vehicle and the center of
simple as Mr. Fonda contemplates on his model. Understeer curvature of the vehicle path, he does not define the latter.
characteristics of a typical passenger car are controlled not just The curvature of the vehicle path of travel is determined
by weight distribution, but also by many other vehicle and tire by the curvatures of the path of travel of each individual
parameters (over 100 parameters). This characteristic might tire. An instantaneous center of the curvature of the path of
change appreciably with a change of operating and environ travel of each individual tire is called the " instantaneous
mental conditions such as power application, and coefficient tire turn center" or "tire turn center." Each tire turn cen
of friction. The influence of these conditions on vehicle be ter is located on the perpendicular to the velocity vector
havior was completely ignored by Fonda. The tailheavy car at tire contact point.
shows an understeering behavior because other understeering
factors overrule the oversteering tendency produced by fore- The vehicle body turn center (center of rotation of the
vehicle) is located on the intersection of the perpendiculars
and-aft weight distribution. All six vehicles shown on my
to the velocity vectors at each tire contact point. In the
Fig. 7 are tailheavy vehicles, but all of them display an un
transient condition during entry of a turn, the radii of the
dersteering behavior.
path of curvature of the tires are larger than the distances
In discussing roll .steer, one must distinguish between front
between the vehicle body turn center and corresponding wheel
and rear roll steer. The rear roll steer has a more signifi
centers. Since the vehicle body turn center and the tire turn
cant effect on vehicle transient response than the front roll
centers are located on the perpendiculars to the velocity vec
steer. The rear roll steer is one of the major factors con
tors, the tire turn centers are located beyond the point of in
trolling vehicle oscillatory stability. An adverse effect of
tersection of these perpendiculars. Therefore, these centers
an excessive rear roll understeer was discussed on page 7 of
must lie apart from each other and consequently, each tire
my paper and was illustrated in Fig. 9 and 10. It is impor has its own individual turn center which does not coincide
tant to note that the decrease in yaw damping with increas with the vehicle body turn center.
ing rear roll steer is due to phase change between yaw and
roll. The above statement is based on the rigorous mathemat
I am in full agreement with Mr. Bosley' s plea for usage ical analysis of vehicle motion. For purpose of illustration
of standardized terminology. The reply to Mr. Bosley's ques positions of instantaneous tire turn centers and vehicle turn
tions and comments follow the order set by h i m : center have been calculated and shown on Fig. A. Calcula
tions were performed for an experimental vehicle entering
1. Vehicle F represents a heavily loaded competitive sta a turn at 61.45 mph at the time 0.4 sec after 3.5 deg step
tion wagon which I cannot identify. It is true that this ve input of the front steer angle. The wheel base of a vehicle
hicle shows much less understeer than any other vehicle is £ = 10.45 feet and the instantaneous values of attitude
tested. Its behavior on the road during lane changing m a
neuver was also quite different from the behavior of other
vehicles evaluated under similar conditions. I agree with
Mr. Bosley that the radius of the circle is rather small, but
unfortunately, this radius is limited by the size of our skid
path.
Orr(Xrr.V)
2. The addition of a front stabilizer bar to a modern pas
senger car generally results in an oversteering effect, when
the vehicle is evaluated during a light lane changing ma
neuver. This behavior was established by subjective evalua
tion conducted by several Ford development engineers and
also substantiated by tests conducted on an instrument ve
hicle. Surprisingly, no change in vehicle behavior was de
tected subjectively, when the stabilizer bar was removed
from the front and installed on the rear. Evaluation of the
same vehicle in a severe cornering maneuver produced quite
different results. Addition of a front stabilizer bar resulted
in an increase of understeer; however, a shift of the stabi
lizer bar from the front to the rear produced a noticeable
oversteering tendency.
I believe that Mr. Nordeen' s criticism of my Fig. 4 was
caused primarily by his misinterpretation of the meaning of
the term "tire turn center." Mr. Nordeen apparently as Fig. A - Positions of instantaneous tire turn centers and ve
sociated tire turn center with the center of rotation of the hicle turn center in transient maneuver (calculated values)
422 W. BERGMAN
angles 0. and radii p. of curvatures of the paths of front right The above example illustrates that during transitional
maneuver, tire turn centers lie apart from each other. Cal
and rear right tires are:
culated distance between the front right and rear right tire
6^ = 0.340 deg mm centers is 37.34 ft. We might ask Mr. Nordeen which
fr °
of these centers constitute his " center of curvature of the
p , = 876.98 ft vehicle path?"
fr
Understeer rate described in my paper permits the meas
6 = -0.997 deg urement of understeer quantitatively and, therefore, it rep
rr °
resents a very useful tool for analysis of vehicle handling
p = 912.89 ft behavior. By using understeer rate, it was possible to prove
rr
that the effectiveness of understeer produced by individual
components diminished as the understeer of a vehicle in
The radii of curvature of tire path were calculated from the
creases (See Table 3 in the paper). Therefore, a vehicle
equation
possessing a large amount of understeer has very low sen
sitivity to a further increase of understeer. This illustrates
3/2
the nonlinear nature of understeer and should not be inter
P = preted that understeer rate "is an insensitive measure of ve
XY - YX hicle understeer for vehicles possessing a large amount of un
dersteer. " I agree with Mr. Nordeen that understeer rate is
Where X, Y are instantaneous coordinates of geometrical influenced by vehicle velocity, but so does the vehicle un
wheel centers, which motion is described in an XYZ .coor dersteer. Therefore, this illustrated that the understeer rate
dinate system fixed in space. The values of XYX and Y is an adequate quantitative measure of understeer within the
were calculated from the equations of motion described in
wide range of speeds.
the Appendix to the paper.
I completely agree with Mr. Nordeen's discussion per
Coordinates of instantaneous tire turn centers were calcu taining to Fig. 9 and 10. Fig. 9 shows calculated values for
lated from the values of p , 0 and i. Coordinate values ex-
1 1 a fixed control vehicle. However, Fig. 10 shows test values
pressed in Table A. The geometrical center of the right for a free control vehicle. The purpose of these illustrations
rear wheel was used as the origin of coordinate system. was not to validate the mathematical model, but to show
that an excessive increase of rear axle roll understeer pro
duces a significant reduction of yaw damping.
Table A - Coordinate Values Both calculated results on a fixed control vehicle and test
results on a free control vehicle illustrate this phenomenon.
Turn Center X Y The paper clearly states that, " The results show agreement
even though the inputs are different."
Rear right tire - O 15.89 912.76
rr REFERENCES
Front right tire - O^ 05.23 876.98
fr 1. Lind Walker, "The Automobile Engineer, Vol. 40,
Vehicle body - O 07.79 447.51
1950 p. 282-283.

The Basic Nature of Vehicle Understeer-Oversteer

Uploaded by

Copyright:

Available Formats

The Basic Nature of Vehicle Understeer-Oversteer

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Basic Nature of Vehicle Understeer-Oversteer

Uploaded by

Copyright:

Available Formats

What is the difference between understeer and oversteer?

What is the difference between understeer and oversteer?

What are the angles used to determine vehicle motion?

What are the angles used to determine vehicle motion?

Downloaded from SAE International by University of Edinburgh, Thursday, August 09, 2018

ABSTRACT ploys a computer program evolved from the equations of

The transitional response of a vehicle is determined pri­

8 . = FRONT STEER ANGLE

Fig. 2 - Determination of position of radius of turn in Fig. 3 - Ackermann

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 389

4-door sedan equipped with manual steering S T E E R I N G W H E E L ANGLE - DEGREES

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 391

5 = Front steer angle due to the geometrical and com­

a = Rear slip angle

Fig. 9 - Vehicle dynamic yaw re­

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 393

OSCILLATORY STABILITY tion of yaw damping is caused by an excessive increase of

bility, in terms of yaw damping, is a measure of the car's

VEHICLK A ORIGINAL DESIGN -

REAR AXLE STEER INCREASED

Fig. 10 - Vehicle dynamic yaw re­

to damp out a momentary wind-force input. Because yaw

Fig. 14 - Sample record of directional

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 395

is applied behind the center of gravity, the yawing moment

Fig. 19 - Neutral response of vehicle

RESULTANT RESULTANT INERTLA

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 397

I N E l i T U I-OUCE FUONT AND UFAIi -

SLIP ANGLES-DEG. -1.0

Table 1 - Steady-State Responses of Vehicle to Various Inputs

Lateral force step Linear Steady state corner­ Change of course

Lateral force of Change of course Lateral displacement

Steering step constant Steady state corner­ Steady state cornering

Steering pulse Change of course Lateral displacement

An increase of Steady- A new steady state No change

A temporary increase Steady- The same steady No change

BASIC NATUttE OF VEHICLE UNDERSTEER-OVERSTEER 399

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 401

Table 2 - Values of Understeer Produced By Individual Design

Understeer Values in Per Cent of

Factor 1 pass, load 3 pass, load 5 pass, load

Design camber 0.20 -0.53 -1.06

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 403

Appendix appears on next page.

EQUATIONS O F VEHICLE MOTION

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 405

k„, = F r o n t shock a b s o r b e r r a t e —jf '

Ax,. K^-2 = S t e e r i n g c o m p l i a n c e coefficients

^xr = R e a r c o m p l i a n c e coefficient (rad. /lb.)

k„ = Effective m a s s factor (nondimensional)

r = Tire rolling radius without load (in.)

^» = Tire power transmission loss factor (nondimensional)

Independent and Dependent Variables (2 Dependent, 6 Independent)

5- = Steering wheel angle (rad.)

BASIC NATURE OF VEHICLE UNDERSTEER-OVERSTEER 407

Vehicle Auxiliary Variables

e = Roll angle (rad.)

Tire Auxiliary Variables

'o - Length of the contact area (in.)

The transitional response of a vehicle is determined pri

5 = Front steer angle due to the geometrical and com

Fig. 9 - Vehicle dynamic yaw re

Fig. 10 - Vehicle dynamic yaw re

Lateral force step Linear Steady state corner Change of course

Steering step constant Steady state corner Steady state cornering

+ [r/w - y (/rr - / w ) ] /,d - [(/"w + Gr/) r0.t 8w

+ rr/rr + y (/rr - /r()l W + ["(/■., + Qrr) COS i„

The quantities $, contained in equation (49) a r e calculated from the follow

/i = 9 - °/i - *'/i (59)

«/r = # + V -I- Vr (60)

ri = (e - %i - «Vi) k,., (61)

„ = (.e + rr + 'rr) »d (62)