Finite Elements in Analysis and Design 63 (2013) 98106

Contents lists available at SciVerse ScienceDirect

Finite Elements in Analysis and Design

journal homepage: www.elsevier.com/locate/nel

An analytical approach for design and performance evaluation of torsion

beam rear suspension
Dongchan Lee a, Chulho Yang
a
b

b,n

Metariver Technology Co. Ltd., B-801, Garden 5 Works, 52 Chungmin-ro, Songpa-gu, Seoul 138-961, South Korea
Mechanical Engineering Technology, Oklahoma State University, 385 Cordell S., Stillwater, OK 74078, USA

articleinfo

abstract

Article history:
Received 1 November 2011
Received in revised form
13 August 2012
Accepted 2 September 2012

Torsion beam rear suspension systems have been widely used in small passenger cars owing to their
compactness, light weight, and cost efciency. To study the roll behavior of torsion beam suspension
systems, analytical equations to obtain roll center height, roll steer, and roll camber have been
developed in terms of geometry points through the kinematic consideration of the system on the
assumption of rigid body motion. In most cases, commercial software for nite element methods or
experimental equations for individual parts are utilized for the design and evaluation of the suspension
systems. This paper, however, proposes an analytical method to calculate the torsional stiffness of a
torsion beam for various loading conditions based on the assumption that a torsion beam is in the range
of linear torsional angles with a constant cross section. The proposed method is useful for exploring
suspension system design, especially in the early stage of vehicle system development in which detail
design parameters such as layout, cross sections, and materials are not yet determined. It is shown that
the torsional stiffness and roll stiffness predicted using the proposed method has sufcient precision
with around four percent difference from the results of nite element analysis and bench tests.
& 2012 Elsevier B.V. All rights reserved.

Keywords:
Rear suspension
Torsion beam
Section prole
Shear center
Roll steer
Roll camber
Roll center height
Roll stiffness
Warping constant

1. Introduction
Torsion beam rear suspension systems are commonly used for
small passenger vehicles because of various advantages including
a reduced weight, lower cost, and compactness. These systems
consist of two trailing arms interconnected by a exible beam
with the simple conguration shown in Fig. 1. This exible beam
has open or closed tubular cross sections that are rigidly connected to each trailing arm in order to provide the torsional and
bending stiffness required for kinematic performance and compliance of the system. A V- or U-shaped prole is typically used
for the open cross sections. The geometric prole of the torsion
beam consists of a constant section area, a transition area in
which the section prole is changed along its distance, and a
junction area at each end that is welded to the trailing arms.
Despite these relatively simple congurations, many design
parameters need to be determined during the early stage of
system design considering the elastic deformation of the whole
system. The torsion beam must allow large torsional displacement over its length without failure of the material and must
have enough stiffness to support lateral forces during cornering. It
must also have enough exibility to allow each wheel to displace

Corresponding author. Tel.: 1 405 744 3033.

E-mail address: chulho.yang@okstate.edu (C. Yang).

0168-874X/$ - see front matter & 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.nel.2012.09.002

independently over a bump. This suspension shows a non-linear

relationship between the wheel toe-in and the wheel vertical
displacement when subjected to an anti-symmetrical roll loading.
This non-linear relationship contributes to the overall handling
response of the vehicle. Therefore, engineers must consider a
balance between the stiffness and the exibility of the components when they design a torsion beam suspension. In the
concept design stage it is necessary to predict changes in
performance that result from changes in the alignment and
compliance steer to see how the stiffness of each member has
an impact on the overall performance of the suspension. Most
researches [15] have been conducted for in-house design to
examine deformation behavior and roll properties based on the
kinematic motions. In addition, commercial nite element analysis (FEA) software or experimental equations for each member is
mainly used for the evaluation of the system. However, in general,
detail geometry (layout, cross section, and materials) of each part
is not available for simulation or test in this stage.
This research proposes design guidelines for a torsion beam
rear suspension that can be used for determining design parameters such as mounting positions, torsion beam cross section,
and material through theoretical analysis of kinematic and elastic
behavior of the system. The proposed design approach is conrmed through the feasibilities of simulation using the commercial FEA software Adams and bench tests of real automotive
suspension systems.

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Nomenclature
A
C
L

Longitudinal distance from body mounting pivot to

shear center
Lateral span of body mounting bush pivot
Longitudinal distance from body mounting pivot to
wheel center
Arc radius of torsion beam prole

T
e
hp
he

f
y

99

Tread, lateral span between two wheel centers

Shear center of cross section of torsion beam
Height of body mounting bush pivot
Height from body mounting bush pivot to shear
center
Roll angle of torsion beam suspension
Angular displacement of torsion beam

2. Design approach for torsion beam rear suspension

For the conguration of a torsion beam rear suspension, many
design parameters, such as position, cross section, material
properties, and geometric dimensions of each component, need
to be considered. The hierarchical design process provides better
performance in achieving efciency and simple operability. If
design engineers need to develop one of the components in detail,
they must estimate the object of design for that component,
which will be incorporated into the upper level of the entire
suspension system. In torsion beam rear suspension, it is important to select the design requirements that have a direct effect on
the performance of the vehicle to show the appropriate results,
including weight, durability, roll center height, roll rate, and
lateral stiffness. Design parameters should be considered in
relation to thickness and shape dimensions of each component
for the given vehicle specication. In particular, the vehicle
specications for torsion beam rear suspension design are its
total weight, pivot point and wheel center of trailing arm, shear
center of torsion beam, and roll center height. Engineers should
determine the fundamental requirements (roll steer, roll camber,
roll stiffness) and layout of the torsion beam rear suspension
through the following considerations.

For discussion of minimal variations of roll axis, the following

denitions are described. In Fig. 2, the vectors from the body
mounting pivot point (point P) to shear center (point Q), wheel
center (point W), and tire patch point (point G) can be given by
! n !!
PQ A u C=2 v

o! T
he w ,

! n !
!
PW L u TC=2 v
o! n !
!! T
PG L u TC=2 v hp w

oT
!
,
hw hp w

and

!!!
where u , v , and w are the unit vectors in x, y, and z directions,
respectively. A is the distance from the point P to shear center Q in
x-direction, C is the span of the body mounting points in ydirection, L is the distance from the point P to wheel center W
in x-direction, and T is the tread between the left and right wheel
centers.
!The vector of rotation axis between
point P and point Q, DaPQ ,
!
can be calculated from its unit vector ( u PQ ) and magnitude

DaPQ

):

DaPQ
2.1. Kinematic characteristics of torsion beam suspension
The roll motion of torsion beam suspension occurs on the roll
axis of the rear suspension and the shear center of the torsion
beam. The shear center is very important in determining the
kinematic characteristics of the suspension. If the shear center of
the torsion beam is given or calculated, kinematic modeling can
be achieved. Under the roll motion, the shear center becomes the
rotation center of the beam section, and if the right- and left-hand
wheels are displaced differently over a bump, the shear center
axis and the center plane of the vehicle make a point of instant
center (point Q). The axis between the body mounting point
(point P) of the trailing arm and point Q becomes the roll axis of
the rear suspension. The design parameters, including the roll
steer and roll camber, can be estimated from the minimal
variations of roll axis [36].

!9DaPQ 9 n !
! !Au 9DaPQ 9 u PQ !
9 PQ 9

!!
C=2 v he w

oT
4

The roll angle (f) of the vehicle is equal to the portion of the
vertical displacement (DZ W ) divided by the half tread (T=2), and
the total torsion angle (y) of the torsion beam is equal to two
!times the y-direction component in DaPQ :
!9DaPQ 9
f DZ W =T=2 ! ATC CL
T9 PQ 9

and

!9DaPQ 9
y ! C:
9 PQ 9
The displacement of wheel center (point W) can be given by
! !DaPQ PW and the displacement of wheel center in z-direction

Fig. 1. Descriptions of torsion beam axles (a) torsion beam rear suspension (b) suspension system for a typical front wheel drive vehicle.

100

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Fig. 2. Kinematic model of torsion beam rear suspension (a) side view of suspension (b) top view of suspension (c) torsion angle of torsion beam (d) roll angle of
suspension.

can be derived from Eq. (2) and Eq. (4):

_! _9DaPQ 9TC CL
A !229 PQ 9

DZ W

The roll steer angle (bf ) and roll camber angle (gf ) of the wheel
under the roll angle of the rear suspension can be given by

bf

bThe

and

f TCA CL

gf

g
f

TA
TCA CL

The toe angle (b) and camber angle (g) are equal to the
z-direction component and x-direction component in Eq. (4),
respectively,

DaPQ
b

Fig. 3. Typical section proles (a) thin-walled open section (b) thin-walled closed
section.

he

PQ

and

DaPQ

10

fS

2K

DZ

f2

AyS C=2xS 2
2
2K S
r2 C 2

14

PQ

11

2.2. Cross-sectional property of torsion beam with respect to torsion

The total roll stiffness of a torsion beam rear suspension is
mainly determined by the roll stiffness of the torsion beam (K
and spring (K fS ),
KfK

fT K fS

fT

The torsion stiffness of the beam is given by the twisting

moment per total twist angle using the shear modulus (G), torsion
constant (J), and length (LT ) of the beam as shown in Eq. (15a)
[68],
Mt

y
J

12

GJ
,
LT

15a

n1X

b t 3,
3i1 i i

15b

and

0
12 ,jmax kmaxX X!!!r jD s k i ADsk =t n
2
The roll stiffness due to the torsional stiffness of the torsion
jmax kmaxX X
4A
beam, K fT , can be given by the lever ratio (r) of the twist angle of
J H
@
the torsion beam to the roll angle. The roll stiffness due to the
j1k1
ds=ts
left- and right-hand side springs,K f S , can be given by the
j1k1
15c
condition that the stored potential energy due to the roll motion
2(1=2K fS f ) is equal to the stored potential energy of two springs
J in Eq. (15b) and (15c) represents the torsion constant along the
(2 1=2K S DZ s 2 ):
center line of wall thickness of a thin-walled open section
(Fig. 3(a)) and along the center line of wall thickness of a thinwalled closed section (Fig. 3(b)), respectively. r k denotes the
KT
f
direction vector of thickness element (sk , sk 1 ). Dsk is the
K fT ,
r
13
y
r2

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

!
circumferential vector of thickness element (sk , sk 1 ) and i is the
unit vector in the longitudinal direction of the torsion beam.
The vertical deformation of springs under the roll motion, DZ S ,

101

! !

DaPQ PS

is

x2 dA
"Z

f0

2t

__2 #3l cos2 f0lcosf0

l Rsinf0 Rsinf tRdf
122
20

DZ S

A yS C=2xS

16

rC
The roll center height of torsion beam suspension, HRCH , can be
given by the geometric behavior of the tire patch point in the roll
motion as shown in Fig. 4
HRCH

TA hp he LT DY

17

TCA CL2 DZ G
! !
where the vector of the tire patch point is DaPQ PG ,
DY G A hp he L, and DZ G TC=2 A CL=2.
2.3. Cross-sectional property of torsion beam with respect to
shear force
Under the roll motion, the total longitudinal shear force is
transmitted across the plane of cross section along the beam and
is called the shear ow [810]. It is the resultant of the shearstress distribution across the thickness. The relation for shear ow
in the presence of a shear force can be given by
VQ
q I

If it is desired to bend a beam by transverse forces without

twisting the beam, each transverse force should pass through the
shear center of the cross section of application. When the section
has an axis of symmetry, the shear center lies on this axis, and
only loading in the plane perpendicular to this axis needs to be
considered. The location of the shear center is independent of the
direction and magnitude of the transverse forces. For the symmetrical sections, the shear center lies on the axis of symmetry,
while for a beam with two axes of symmetry, the shear center
coincides with their point of intersection. In general, it is not
necessary for the shear center to lie on a principal axis and it may
be located outside the cross section of the beam. For thin-walled
sections, the shear stresses are taken to be uniformly distributed
over the thickness of the wall and parallel to the boundary of the
cross section. In Fig. 5, the moment M z of the shear stress due to
the shear ows about arbitrary point O is the external moment
(eV) attributable to V about O. The distance between O and the
shear center is given as Eq. (21).
__iR h2Mz2 f0Re V V 0 qcosf cosf Rcosf qRsin f Rdf0
"#
___ _
R2 tR2f0 sin2f0E
sinf0 Ef0 , 21
I cosf0 2cosf04

18

Q and I are the rst moment and the area moment of inertia of
the shaded area about the Y axis in Fig. 5 and can be expressed as
shown in Eqs. (19) and (20):
"Z__#Zf0l cosf0
QxdA tl19R2 sin fdf Rsinf0
2f

where
E l2 =2R2 cosf0 Rl sinf0
_
IR

and

_ 3__2

f0 sin2f0l cos2 f0lcosf0

l Rsinf0
24122

2.4. Cross-sectional property of the torsion beam with respect to

warping
Warping of the torsion beam cross section can seriously affect
the durability by increasing the maximum stress in the junction
area between the torsion beam and trailing arm. It also affects the
kinematic performance of the suspension system causing changes
in toe angle. Therefore, warping must be carefully considered
when the suspension system is designed.

Fig. 4. Roll center height of torsion beam rear suspension.

Fig. 5. Shear ow and shear force of torsion beam section.

22

Fig. 6. Warping shape of cross section of torsion beam.

23

102

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Assuming that the cross sections of the torsion beam (Fig. 6)

rotate with respect to an axis passing through point A and parallel
to the longitudinal axis while it is twisted, we nd that any
longitudinal position N in the middle surface of the wall becomes
inclined to the axis of rotation by the angle y. [9,10] The
longitudinal point N is dened by the distance s measured along
the middle line of the cross section. The tangent to the middle line
at N remains perpendicular to the longitudinal layer, and the
small angle between this tangent and the xy-plane is rycos a r y
after torsion.
Having w denote the displacement of the middle line of the
cross section in z direction and considering the twist moment to
be positive, the following expression can be obtained:

3. Application to design and validation of a suspension system

Fig. 7 shows the verication and validation ows through the

entire process of designing a rear suspension from the conceptual
description of the system to the nal design. The predictive
capability of models is limited to specic load ranges and has
been proven by showing the correlation through experiments.
Table 1 shows the geometric parameters of a given rear
suspension. Table 2 shows the property of each section prole
for the torsion beam suspension using the numerical section
property equations. Fig. 8 shows the typical nite element model
of the torsion beam rear suspension for the verication of
numerical equations. The simulations and experiments are made
for four section proles of the torsion beam rear suspension.
The suspension system is modeled as a single exible body with
the origin located at the vehicle center line, and hence all
deformation is described in this coordinate system. Large rotation
@w
of the suspension can occur at the origin of this coordinate
24
r y
system; therefore, the deformation is superimposed onto the
@s
large rotation at the vehicle center line. The toe and camber are
By integrating Eq. (24), we can obtain the warping displacegoverned mainly by linear elastic deformation ranges that can be
ment of the point from which s is measured in z direction:
superimposed.
Zs
In reality, the rotation at the left and right suspension bushes
can be considered independently due to the exibility of the
rds25w w0 y
torsion beam itself. The deformation needs to be described
o
separately relative to each body mounting bush. A model is
created by separating the suspension at the center of the torsion
The average value w of the warping displacement can be
beam and dening a exible body for each half. Each body is
calculated from Eq. (25) as follows:
described relative to its mounting bush and both bodies are
_ZZ _Z s1 my m
joined together using a xed joint. The linear deformation can
wwds w0 rds ds26
be superimposed onto the large rotation of each body of its own
m om oo
mounting bush.
Numerical validation has been achieved by comparing the
results obtained from commercial software and experiments.
Subtracting this value from the warping displacement given by
Fig. 9 shows the test set-ups for evaluating kinematic and
Eq. (25), the warping of the cross section can be found with
compliance performances of the torsion beam rear suspension.
respect to the plane of average warping:
Table 3 shows roll steer, roll camber, roll center height, and roll
_Z _Z sZ s
stiffness compared with the quasi-static results obtained from the
y mdc
commercial multi-body dynamic software, MSC/Adams, and
,27wrds dsyrds yws ws
experiments. From the results of quasi-static analysis and experiws ws
ments, it is known that the numerical equations can be veried
m odzoo
for their feasibilities of estimations of kinematic characteristics of
where the warping function ws represents the doubled sectional
a torsion beam rear suspension. Fig. 10 shows the average force
area corresponding to the arc s of the middle line of the cross
and vertical displacement measured at the left and right wheel
section, while ws represents the average value of ws . For the case
center under the roll motions. Note that the estimated values of
of non-uniform torsion, the constant angle of twist per unit
kinematic performance are very close to the results of nite
length, y, is replaced by the variable rate of change of the angle
element analysis and tests.
of twist dc=dz for the angle of rotation of any cross section, c.
When a new product is to be designed, various solutions are
Since the rate of change in twist angle varies along the length
often possible to be embodied in concept designs. A concept
of the torsion beam, adjacent cross sections will not be warped
design is an early representation of a product, incorporating only
equally and, as a result, there will be axial strain ez of the
a minimum of details in shape, material, and manufacturing
longitudinal layer of the beam. We can obtain the axial strain
process. While the concept denes the maximum achievable
and the normal stresses produced during non-uniform torsion
performance of a product, the detail design determines the actual
from Hookes law:
performance. During detail design, each design parameter is
rened through an optimization process.
It is known that torsional stiffness of a torsion beam controls
7085% of the roll stiffness of a vehicle and reinforcement about
1530%. For the optimization of a beam cross section, roll
2
2
@wd c
dc
stiffness, lateral stiffness, and weight are considered as they are
ez ws ws 2 ,
sz Eez Ews ws dz2
28
primarily affected by the cross section [11]. Durability of the
@zdz
beam and junction area is additionally considered when the
The twist moment is applied only to the ends of the beam and
topological optimization for the junction and reinforcement
its cross sections are free to warp. Under such conditions, warping
panels is conducted. To obtain the best solution, it is necessary to
is the same for all cross sections and takes place without any axial
select major design parameters that affect the performance of the
strain of the longitudinal direction. The case of non-uniform
torsion beam suspension system. The shape of the beam cross
torsion occurs if any cross sections are not free to warp or if the
section is an important design parameter related with roll stiffness,
twist moment varies along the length of the beam. The differential equation for non-uniform torsion is given by

dcd c
29
C 2 3 ,Mt C 1
dzdz
where C 1 is the torsional rigidity of the beam, C 1 GJ, and C 2 is
the warping rigidity referred to as the portion of the twist
moment due to non-uniform torsion and non-uniform warping
of the cross section, C 2 EC w . C w is the warping constant given by
Rm
C w 0 ws ws 2 tds and has units of length to the sixth power.
The constraints of ange warping of the torsion beam can
increase the stress in the zones connected to the trailing arm, as
well as torsion stiffness of the beam.

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

103

Fig. 7. Product verication plans (a) Modeling and simulation phases for product design and (b) design ows for torsion beam rear suspension.

Table 1
Geometric parameters of torsion beam suspension.
Parameter

Value

hp (mm)
he (mm)
A (mm)
C (mm)
L (mm)
T (mm)

263.6
13
188.7
1147.6
423.56
1410

lateral stiffness, durability, and weight. Therefore, thickness and

shape of the beam cross section are optimized to satisfy various
design requirements for the torsion beam suspension system as
follows:
Objective: Minimize maximum stress (smax)
Design Variables: Thickness (t) and shape of the beam cross
section (a, R, R2)
Dimensions in transition zone (Tr1, Tr2, Tr3)
Constraints: Roll Center Height r Upper limit
Weight r Upper limit (or initial value)

Roll Stiffness Z Lower limit

W/C Lateral Stiffness Z Lower limit
Note that the upper and lower limits are not described with detail
values because they are proprietary values dened by each automotive maker for the performance of its own vehicles. iSIGHT,
developed by DASSAULT Systems, is used for the optimization based
on an SQP (Sequential Quadratic Programming) optimizer interoperating with Hypermesh and NX.NASTRAN. The shape of the torsion
beam is calculated and determined with respect to each design
parameter using morphing functions given in Hypermesh. Linear
analysis of stress, roll stiffness, and lateral stiffness are performed
using NX.NASTRAN, and the stiffness of each part is calculated using
numerical equations while roll behavior of a vehicle is analyzed.
In the design phase focusing on the kinematic roll performances,
roll strength under the roll motions of a 4.5 degree angle can be
determined. Fig. 11 shows the design parameters of the cross section
prole and the geometric parameters of the transition zone in the
torsion beam. Fig. 12 shows the design plans for the roll strength in
the roll durability. Fig. 13 shows the stress contours for the initial
and optimum torsion beam suspension under roll motion. Changes
in system characteristics and performances for the initial and
optimum designs are compared in Table 4.

104

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Table 2
Section properties and proles for torsion beam suspension.
Section

t (mm)

Area (mm2)

I1 (mm4)

I2 (mm4)

J (mm4)

Stiffness (N m/rad)

2.6

806.11

574,321

274,941

33,979

28.7

2.6

805.90

630,351

259,312

34,246

27.9

2.6

806.39

639,279

259,013

24,602

24.6

2.6

806.42

692,674

244,369

23,692

23.4

Section prole

Fig. 8. Typical tubular torsion beam suspension (a) CAD model (b) nite element model.

Fig. 10. Force-displacement curves for roll motion of torsion beam rear
suspension.

4. Conclusion

Fig. 9. Test set-ups for torsion beam rear suspension.

Table 3
Kinematic performances of torsion beam suspension.
Performances

Numerical equation

Adams

Test

Roll
Roll
Roll
Roll

0.026
0.31
109
12.17

0.029
0.42
112
12.8

0.031
0.45
117
12.65

steer (%)
camber (%)
center height (mm)
stiffness (N m/rad)

A new analytical method for designing a torsion beam rear

suspension system is proposed in this paper. This method utilizes
theoretical descriptions for kinematic and elastic behavior of a
torsion beam suspension. The analytical investigations can give
engineers an efcient and easy way to design a torsion beam rear
suspension by enabling design engineers to quickly perform basic
design and property evaluations in the suspension planning stage.
Quasi-static nite element analyses and experiments are included in
the paper to validate the feasibility of the proposed approach

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Fig. 11. Design parameters for the durability strength (a) design parameters of cross section prole (b) design parameters in transition zone mounted to trailing arm (c)
shape prole of design parameter Tr1.

Fig. 12. Design plans for the roll durability strength.

105

106

D. Lee, C. Yang / Finite Elements in Analysis and Design 63 (2013) 98106

Fig. 13. Roll motion strength of torsion beam suspension (a) roll motion strength of initial model (b) roll motion strength of optimized model.

Table 4
Comparison of the values of objective function constraints between initial and
optimum designs.

Performances

Lower limit

Initial

Optimum

Upper limit

Objective function (MPa)

Roll center height (mm)
Weight (kg)
Roll stiffness (N m/rad)
W/C lateral stiffness (N m)

12.00
390

478
109
1020
12.17
345

370
116
1085
13.5
412

130
1130

with condence in predicting the behavior of the exible body. The

simulation and experiments show that the numerical calculations of
suspension properties are sufciently precise to allow engineers to
judge the acceptability of design plans in the development stage.
Several processes necessary for analysis of the suspension are
integrated and automated, thus the analysis and design works for
a suspension system are efciently performed. Design optimization
considering kinematic performances, compliance, and strength of
the torsion beam axle suspension is also successfully performed by
applying a deterministic optimization procedure and discrete design
solution. The characteristics of the exible body are also affected by
the lateral stiffness of the bushings. Therefore, further studies need to
be carried out to predict the behavior of non-linear elastic bushes

under different loading conditions, such as braking and cornering, to

complete the evaluation of the torsion beam suspension system.

References
[1] H. Horntrich, Rear suspension design with front wheel drive vehicles, SAE
810421 (1981).
[2] H. Shimatani, S. Murata, K. Watanabe, T. Kaneko, H. Sakai, Development of torsion
beam rear suspension system with toe control links, SAE (1999) 1999-01-0045.
[3] D.C. Lee, J.H. Byun, A study on the structural characteristics and roll behavior
of suspension for the section prole of torsion beam, Trans. KSAE 7 (9) (1999)
195202.
[4] T.L. Satchell, The design trailing twist axles, SAE 810420 (1981).
[5] J.M. Lee, J.R. Yun, J.S. Kang, S.W. Bae, A study on the steady-state cornering of
a vehicle considering roll motion, Trans. KSAE 5 (6) (1997) 89102.
[6] J. Kang, Kinematic analysis of torsion beam rear suspension, Trans. KSAE 12
(5) (2004) 146153.
[7] S.H. Crandall, N.C. Dahl, T.J. Lardner, An Introduction to the Mechanics of
Solids, second ed., McGraw-Hill, New York, USA, 1978.
[8] H. Sugiura, Y. Mizutani, H. Nishigaki, First-order analysis for automotive
suspension design, R&D Rev. Toyota CRDL 37 (1) (2002) 2530.
[9] A.C. Ugural, S.K. Fenster, Advanced Strength and Applied Elasticity, third ed.,
Prentice-Hall, New Jersey, USA, 1995.
[10] S.P. Timoshenko, J.M. Gere, Theory of Elastic Stability, second ed., McGrawHill, New York, USA, 1963.
[11] D.C. Lee, C.S. Han, A frequency response function-based updating technique
for the nite element model of automotive structures, Proc. Inst. Mech. Eng.
Part D J. Automob. Eng. 222 (10) (2008) 17811791.