Escolar Documentos
Profissional Documentos
Cultura Documentos
400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-5760
The appearance of this ISSN code at the bottom of this page indicates SAE’s consent that copies of the
paper may be made for personal or internal use of specific clients. This consent is given on the condition,
however, that the copier pay a per article copy fee through the Copyright Clearance Center, Inc. Operations
Center, 222 Rosewood Drive, Danvers, MA 01923 for copying beyond that permitted by Sections 107 or
108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying such as copying for
general distribution, for advertising or promotional purposes, for creating new collective works, or for
resale.
Quantity reprint rates can be obtained from the Customer Sales and Satisfaction Department.
To request permission to reprint a technical paper or permission to use copyrighted SAE publications in
other works, contact the SAE Publications Group.
No part of this publication may be reproduced in any form, in an electronic retrieval system or otherwise, without the prior written
permission of the publisher.
ISSN 0148-7191
Copyright © 2002 Society of Automotive Engineers, Inc.
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely
responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in
SAE Transactions. For permission to publish this paper in full or in part, contact the SAE Publications Group.
Persons wishing to submit papers to be considered for presentation or publication through SAE should send the manuscript or a 300
word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.
Printed in USA
2002-01-0804
1 1
U= k s ( xt − x s ) 2 + k t ( x r − xt ) 2 (4)
2 2
The non-linear characteristics of the damper
have been lumped into the damping
coefficient. This includes the effects of
suspension geometry as well as the
Figure 1 - Quarter Car Model differences in bump and rebound damping.
The quarter car model is used to study the The Rayleigh dissipation function of the
behavior of the system around its dampers and damping of the tire is given by
equilibrium state; therefore, the eqn 5.
displacements are taken to be zero at static
equilibrium. 1 . . 1 . .
R= c s ( x t − x s ) 2 + ct ( x r − x t ) 2 (5)
2 2
The variables of the model are defined as:
Substituting eqns 3,4 and 5 into 1 and 2
xs - Displacement of Sprung mass from yields the equations of motion for the quarter
equilibrium position car.
xt – Displacement of unsprun mass from
equilibrium position .. . .
ks – suspension stiffness m s x s − c s ( x t − x s ) − k s ( xt − x s ) = 0 (6)
cs – suspension damping
kt – tire vertical stiffness .. . . . .
cs – tire damping mt x t + c s ( x t − x s ) − c t ( x r − x t )
The equations of motion can be derived + k s ( x t − x s ) − k t ( x r − xt ) = 0 (7)
using Lagrange’s Equations for a holonomic
system without external forces. Root Locus Analysis of Damping Coefficient
eqn 3.
Figure 2 shows the root locus for a quarter
car model of a typical Champcar in street
course configuration. The parameters used where the tire stiffness is much higher than
for the root locus are for a front corner and the suspension stiffness.
are shown in Table 1 in the appendix.
Figure 2
The root locus shows the effects of varying Figure 3
the damping coefficient of the quarter car Figure 3 shows the root locus for a typical
model. The pole trajectory for the unsprung road car suspension, the parameters used
mass shows that adequate damping, a are shown in Table 2 in the appendix. The
damping ratio of 1, can be achieved by root locus shows that the damping ratio for
increasing the damping coefficient. This both modes moves closely together as the
occurs when the pole related to the damping coefficient is increased. The
unsprung mass motion is on the real axis. square marker shows the current roots for a
The markers on figure 2 show the damping damping value of 65 lb/(in/s) [11383
coefficient of 49 lb/(in/s) [8581 N/(m/s)] that N/(m/s)]. For this amount of damping the
results in well damped unsprung mass mode unsprung mass mode is overdamped, the
(the marker is on the real axis), however, pole is on the real axis, and the sprung
this amount of damping only results in a mass mode is almost as well damped with a
damping ratio of about .2 for the sprung .8 damping ratio. Thus the damper
mass. Conventional wisdom would indicate dampens both modes relatively equally.
that increasing the damping coefficient
would improve the situation and dampen the Unlike the racecar, the tires of a passenger
body mode. In fact the opposite results. car are quite stiffer than the suspension
The root locus analysis shows that springs, thus increasing the damping does
increasing the damping coefficient will not increase the displacement of the tire
decrease the damping of the sprung mass substantially. However, this brings about
mode. other difficulties, since increasing the
damping coefficient to dampen the sprung
The reason for this limitation is that the and unsprung masses will result in
stiffness of the tire is the same order of increased transmissibility of road vibrations
magnitude as the spring stiffness. to the passenger compartment.
Increasing the damper stiffness by
increasing the damping coefficient will limit SEVEN DEGREE OF FREEDOM (DOF)
the movement of the suspension, and will
CAR MODEL
increase the deflection of the tire, so that the
sprung mass then bounces on the tire. The
Although the heave response of the car can
manifestation of this phenomenon is known
be easily understood with the quarter car
in the racing community as locking the
model, a seven DOF model shows the
shock. In contrast to this situation we can
interactions of the different modes. The car
examine a passenger car suspension,
is modeled as an independently sprung,
rigid body with pitch and roll motions about
its principal axes. The suspension geometry tr – rear track width
effects are neglected, except for motion ratio
considerations. The motion ratio is used to Again we use Lagrange’s Equations to
generate an equivalent vertical spring derive the equations of motion.
stiffness and damping at the wheel. Figure
4 shows the geometry of this model. The kinetic energy of the system consists of
the translational energy of the masses as
well as the rotational kinetic energy of roll
and pitch of the sprung mass. The kinetic
energy is given by eqn 8.
1 .2 1 .2 1 .2 1 .2
T= ms x s + I p Θ p + I r Θ r + mtfl x fl
2 2 2 2
(8)
1 .2 1 .2 1 .2
+ mtfr x fr + mtrl x rl + mtrr x rr
Figure 4 –Seven DOF Car Model 2 2 2
The variables of the model are defined as: The potential energy of the system is given
by eqn 9.
xs – vertical displacement of sprung mass
Θp – pitch angle of sprung mass 1 tf
Θr – roll angle of sprung mass U = k fr ( x fr − a Θ p − x s + Θ r )2
2 2
xfl – vertical displacement of front left tire 1 tf
xfr – vertical displacement of front right tire + k fl ( x fl − a Θ p − x s − Θ r )2
2 2
xrl – vertical displacement of rear left tire 1 tr
+ k rr ( x rr + b Θ p − x s + Θ r )2
xrr – vertical displacement of rear right tire 2 2 (9)
yfl – vertical input at front left tire 1 tr
+ k rl ( x rl + b Θ p − x s − Θ r )2
yfr – vertical input at front right tire 2 2
yrl – vertical input at rear left tire 1 1
+ k tfr ( y fr − x fr ) 2 + k tfl ( y fl − x fl ) 2
yrr – vertical input at rear right tire 2 2
ms – sprung mass +
1 1
k trr ( y rr − x rr ) 2 + k trl ( y rl − x rl ) 2
Ip – pitch inertia 2 2
Ir – roll inertia
mtfl – front left tire mass The Rayleigh dissipation function of the
mtfr – front right tire mass dampers and damping in the tires is given
mtrl – rear left tire mass by eqn 10.
mtrr – rear right tire mass
kfl – front left suspension stiffness
kfr – front right suspension stiffness
krl – rear left suspension stiffness . . .
1 tf .
krr – rear right suspension stiffness R= c fr ( x fr − a Θ p − x s + Θ r ) 2
cfl – front left damping coefficient 2 2
cfr – front right damping coefficient 1 . . . tf .
crl – rear left damping coefficient + c fl ( x fl − a Θ p − x s − Θ r ) 2
2 2
crr – rear right damping coefficient . . .
1 tr .
ktfl – front left tire stiffness + c rr ( x rr + b Θ p − x s + Θ r ) 2
ktfr – front right tire stiffness 2 2 (10)
ktrl – rear left tire stiffness 1 . . . tr .
+ c rl ( x rl + b Θ p − x s − Θ r ) 2
ktrr – rear right tire stiffness 2 2
ctfl – front left tire damping coefficient 1 . . 1 . .
ctfr – front right tire damping coefficient + ctfr ( y fr − x fr ) 2 + ctfl ( y fl − x fl ) 2
ctrl – rear left tire damping coefficient 2 2
1 . . 1 . .
ctrr – rear right tire damping coefficient + k trr ( y rr − x rr ) 2 + k trl ( y rl − x rl ) 2
a – distance from cg to front axle 2 2
b – distance from cg to rear axle
tf – front track width
iii
Substituting into Lagrange’s equations output Y of the system. In the case of the
results in the equations of motion for the 7 seven post rig, the input signals are created
DOF model. The equations of motion can by the control system so therefore can be
be found in Appendix B. assumed to be noise free.
SEVEN POST RIG DATA ANALYSIS For the discussions in this case study we
are only concerned with heave and pitch
The rig instrumentation can include up to 40 motions since the car is symmetric in the roll
channels of vehicle data and it is easy to axis, thus only two accelerometers are
become overwhelmed with the analysis. In sufficient to describe the motions of the car.
order to simplify the analysis, only a few One accelerometer is placed along the
parameters are tracked, and when further centerline of the car on the front axle line,
analysis is warranted the remaining and another one is placed on the centerline
channels are analyzed. Most of the data along the rear axle line. From the geometry
reduction occurs in the frequency domain. of the locations of the accelerometers we
Computer software has been written in can define the following heave and pitch
Matlab using the signal processing toolbox given by eqns 13 - 14
in order to automate the analysis procedure.
The analysis consists of obtaining the b * FrontAccel + a * Re arAccel
frequency response function (FRF) of the a+b
Heave = (13)
major modes of the vehicle; specifically the InputAccel eration
heave, pitch and roll modes of the sprung
mass as well as the vertical modes of the
FrontAccel − Re arAccel
unsprung mass.
Pitch = a+b (14)
InputAcceleration
This section discusses how these FRF’s are
obtained. In the subsequent section the
Eqn 16, the contact patch load magnification
equivalent responses will be obtained for the
transfer function, shows how well the tire is
7 DOF model.
kept in contact with the road. A load cell in
each wheelpan measures the vertical force
The FRF is given by H(f) in eqn 11
at each contact patch. Since the
longitudinal and lateral force generation of
Y( f ) = H( f )* X( f ) (11) the tire is proportional to the vertical load,
minimizing the variation of the contact patch
where Y(f) is the system response and X(f) load leads to a more constant lateral and
is the input signal, both of these signals are longitudinal force generation. Looking at
measured and thus eqn 11 can be solved for this graph for the front and rear pairs is
H(f). Using the transfer function estimate useful when making adjustments to improve
function (TFE) in Matlab’s signal processing particular areas of handling. For example,
toolbox solves for H(f). This function divides to improve traction of the rear wheels the
the data into overlapping blocks, windows contact load variation at the rear would need
the data with a Hanning window to reduce to be minimized.
leakage and finally forms the transfer
function using the H1 estimate of the FRF. ContactLoa d
ContactPat chLoadMagn ification = (16)
The H1 estimate is given by eqn 12 InputAccel
Θp S ( f ) = C sp ( f / V ) − N (22)
Pitch = (18)
Y fl
Recall that to find the velocity PSD from the
displacement PSD it is only necessary to
Since the input is a pure heave multiply by the frequency in the frequency
displacement all four wheel displacements domain, thus if we assume N=1 the velocity
will be the same and thus it is sufficient to PSD becomes:
use the displacement of one of the corners,
in this case we chose the front left SV ( f ) = C spV (23)
wheelpam, Yfl.
Table 1
.. . . . tr .
m trl x rl + c rl ( x rl + b Θ p − x s − Θr)
2
. .
− c trl ( y rl − x rl )
tr
+ k rl ( x rl + b Θ p − x s − Θr)
2
− k trl ( y rl − x rl ) = 0
.. . . . tr .
m trr x rr + c rr ( x rr + b Θ p − x s + Θr)
2
. .
− c trr ( y rr − x rr )
tr
+ k rr ( x rr + b Θ p − x s + Θr)
2
− k trr ( y rr − x rr ) = 0
i
Chalasani, R. M., “Ride Performance
Potential of Active Suspension Systems-
Part 1:Simplified Analysis Based on a
Quarter Car Model”. ASME Symposium on
Simulation of Ground Vehicles and
Transport Systems CA, 1986
ii
Genta, G., “Motor Vehicle Dynamics:
Modeling and Simulation”
iii rd
Randall, R.B., “Frequency Analysis”, 3
ed., Brüel & Kjaer 1987
iv
Wong, J.Y., “Theory of Ground Vehicles”,
nd
2 ed., Wiley Interscience 1993
v
Kasprzak, J.L., Floyd, R. S., “Use of
Simulation to Tune Race Car Dampers” SAE
942504
vi
Classnotes ME653:Vehicle Dynamics, Dr.
E.H. Law, Clemson University – Motorsports
Engineering Program