World Academy of Science, Engineering and Technology

International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

Fatigue Life of an Anti-Roll Bar of a Passenger

Vehicle
J. Marzbanrad, A. Yadollahi

Abstract—In the present paper, Fatigue life assessment of an That is, bent extended ends which are attached to each
anti-roll bar component of a passenger vehicle, is investigated by swing-arm via short vertical link-rods by means of rubber-
ANSYS 11 software. A stress analysis is also carried out by the bush joints. Thus the function of the antiroll bar is to relieve
finite element technique for the determination of highly stressed the main suspension springs of some of their load every time
regions on the bar. Anti-roll bar is a suspension element used at the the body rolls. An antiroll bar stiffens the suspension
front, rear, or at both ends of a car that reduces body roll by resisting
springing when the body rolls or one wheel goes over a bump
any unequal vertical motion between the pair of wheels to which it is
connected. As a first stage, fatigue damage models proposed by some
or dip in the road. The antiroll bar therefore permits softer
well-known references and the corresponding assumptions are suspension springing to be used when the vehicle is moving
discussed and some enhancements are proposed. Then, fracture straight ahead, as this responds far better than hard springing
analysis of an anti-roll bar of an automobile is carried out. The to small ripples or irregularities in the road, hence it improves
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

analysed type of the anti-roll bar is especially important as many the ride comfort of the passengers. In the extreme, some
cases are reported about the fracture after a 100,000 km of travel antiroll bars are tuned to take as much as 30–40% of the total
fatigue and fracture conditions. This paper demonstrates fatigue life vertical load imposed on the suspension when subjected to
of an anti-roll bar and then evaluated by experimental analytically severe body roll [2]. Fig. 1, shows the anti-roll bar in case of
results from other researcher.
the installation. The anti-roll bar picture analysed in this study
is given in Fig. 2.Recently many cases of fractured anti-roll
Keywords—Anti-roll bar, Fracture, Fatigue life, Random loading
bars after a 100,000 km of travel for a special passenger car
are reported. All of the bars are fractured at nearly the same
I. INTRODUCTION
location (Fig. 3). In this study, fracture analysis of this anti-

G ROUND vehicle design typically represents a trade-off

between performance and safety. Design parameters
affecting lateral dynamics can influence maneuvering ability,
roll bar is carried out. Mechanical characteristics of the
material are obtained first. Then the microstructure and
chemical compositions are determined. Some fractographic
but also have some influence on the dynamic stability studies are carried out to access the fatigue and fracture
including spinout and rollover. In steering maneuvers, vertical conditions. A stress analysis is also carried out by the finite
loads on tires at the outer track increase and those on the inner element technique for the determination of highly stressed
track decrease, which is called lateral load transfer. When regions on the bar.Many engineering components are
moment equilibrium is broken at some conditions, a vehicle subjected to complex combinations of loads that fluctuate with
loses roll stability. Geometric dimensions, suspension time in a non-proportional manner. In some cases, time
characteristics as well as maneuvering conditions influence the variations of the loads are of random nature. Component that
dynamic roll behavior of a car. To improve the roll are used in airplanes, ground vehicles, marine and offshore
characteristics of a car, the customary approach is to increase structures, railways, and bridges are subjected to loads that
the roll stiffness by using a stabilizer bar which affects the ride vary with time in a very complicated manner. Many fatigue
comfort with respect to high frequency isolation induced by damage models have been proposed in the literature, most of
road excitation. In order to enhance the vehicle performance, them are limited to specific materials or simple loading
several experiences are studied. At present one of the most conditions. Multiaxial fatigue criteria are presented based on
efficient methods is the roll control with a semi-active different concepts: stress invariants, energy, critical plane,
suspension system in order to isolate the driver from roadway fracture mechanics, etc. Fatigue damage increases with the
noise, road holding on irregular road surfaces and safe turning number of cycles of the applied load in a cumulative manner.
through steering. As is the case with any vehicle system, an Many cumulative damage theories are reviewed by Fatemi and
actual car is expected to operate in a highly variable Yang [3]. A theory is proposed in the present paper that is
environment. For instance, parameter variations resulting from suitable for life assessment in three-dimensional stress fields.
loading pattern and driving condition will influence vehicle Generally, the aim of the fatigue damage models proposed so
dynamics [1]. far was checking the fatigue failure strength rather than
The antiroll bar is a circular sectioned torsion bar mounted calculating the number of cycles or number of history blocks
transversely on spaced out rubber bush swivel bearings to the that may lead to the fatigue failure. In the current paper, an
body and having cranked arms. algorithm is developed to predict the fatigue life based on the
damage models that are proposed in the paper. Result obtained
J. Marzbanrad is with the Iran University of Science and Technology,
based on FEM analysis by ANSYS 11 software is compared
School of Automotive Engineering, Iran, Tehran (phone: 9821-7720360; fax: with some well-known damage theories, the modified theories
9821-77491224; e-mail: marzban@ iust.ac.ir). and the experimental results prepared by another author.
A. Yadollahi is with the Islamic Azad University, Semnan Branch,
Automotive Department, Iran, Semnan (e-mail: [email protected]).

International Scholarly and Scientific Research & Innovation 6(2) 2012 407 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

Fig. 1 Schematic presentation of an anti-roll in action

International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

Fig. 2 The picture of the undamaged anti-roll bar analysed in this study

Fig. 3 Two different fractured anti-roll bar

II.FATIGUE ASSESSMENT CRITERIA where J 2 and J 3 are the second and the third invariants of the
Stress-based multi-axial fatigue failure criteria have already deviatoric stress tensor, respectively, σ 1 , σ 2 , and σ 3 are the
been developed based on experimental observations. Some of
these criteria as well as some recently proposed criteria are principal stresses, and τ f and σ f are the fatigue failure
mentioned in the current section. stresses. In many applications, the stress state can be reduced
to a two-dimensional state including a normal stress and a
A. Von Mises and Tresca criteria
shear stress components only (e.g. for a component under
The analytical forms of these criteria are respectively [4]: combined bending and torsion). In these situations, (1), can be
Tresca criterion: rewritten as:

J 2 sin(α + π / 3)τ f , σ max + (ϕτ max ) 2 = σ f (3)

(1)
§ 3 3. J 3 · π
1
α = cos −1 ¨ ¸:− <α < π
3 ¨ 2 J 23 / 2 ¸ 6 6 where the value of ϕ is 2 and 3 for Tresca and Von- Mises
© ¹
criteria, respectively.
Von Mises criterion:
B. Ellipse quadrant criterion
Gough and pollard, proposed the following modified von
J2 =
1
6
[ ]
(σ 1 − σ 2 ) 2 + (σ 2 − σ 3 ) 2 + (σ 3 − σ 1 ) 2 = σ 3f / 3 (2) Mises criterion for in-phase fully reversed loadings [12]:

International Scholarly and Scientific Research & Innovation 6(2) 2012 408 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

2 2 2 2
§ σa · § τa · § τ a · § σ n ,max ·
¨¨ ¸¸ + ¨¨ ¸¸ = 1 (4) ¨¨ ¸¸ + ¨¨ ¸¸ = 1 (9)
© σ R = −1 ¹ © τ R = −1 ¹ © τ R = −1 ¹ © σ R = − 1 ¹

where σ R =−1 and τ R=−1 are the fatigue strength values for the The weighted mean principal stress method is used to
determine the critical plane.
specified number of cycles to failure. In the present paper, the
following abbreviations are used, for simplicity: G.Ninic criterion
Ninic [12] has presented a criterion that modifies Gough’s
σ R ≡ (σ a ) R , τ R ≡ (τ a ) R (5) criterion as follows:

C.Sines criterion [5] 2 2

§ τa · § σ + σ R = −1 (σ m / σ n ) 2 ·
Sines suggested that fatigue failure is achieved when the ¨¨ ¸¸ + k ¨¨ a ¸¸ = 1 (10)
following condition is satisfied [6]: © τ R = −1 ¹ © σ R = −1 ¹
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

( J 2 ) a + α (I1 ) m = β where k is a normal stress sensitivity factor that is calculated

(6) based on a maximization procedure, as:
α = (3τ R = −1 − 3σ R = −1 ) / σ m β + τ R = −1
­§ σ · ª § σ R = −1 · º
2 2
τ 2
the first stress invariant ( σ h is the hydrostatic stress)
I 1 = σ h is °¨¨ R = −1
¸¸ «1 − ¨¨ ¸¸ » 0 ≤ R = −1 ≤
° τ 2τ σ R = −1
k = ®© R =−1 ¹ «¬ © R =−1 ¹ »¼
2 (11)
and α and beta are material parameters.
° 2 τ R = −1
D.Findley criterion ° 1 ≤
¯ 2 σ R = −1
Findley criterion may be described as [7]:

(τ a + kσ n ) max(θ ,ψ ,ϕ ) = f H.Liu criterion

(7)
2 − σ R =−1 / τ R =−1 Liu et al. [13] proposed the following characteristic plane
k= based criterion:
2 σ R =−1 / τ R =−1 − 1
(σ R =−1 ) 2
σm º
2
f = ª
4 σ R =−1 / τ R =−1 − 1 « σ a (1 + η σ » § τ ·2 § σH ·
2
(12)
« R = −1
» + ¨¨ a ¸¸ + k ¨ a ¸ = μ
« σ R = −1 » © τ R = −1 ¹ ¨σ ¸
© R = −1 ¹
According to this criterion, the critical plane is a plane « »
¬ ¼
where max θ ,ϕ ,ψ ,t (τ a + kσ n ) occurs θ , ϕ , and ψ are the Eulerian
angles and t denotes the time.
where σ aH is the hydrostatic stress amplitude. This criterion
E. McDiarmid criterion may be considered as a modified version of Gough’s criterion.
McDiarmid used the concept of type A and type B cracks The main aim was to substitute the 3D stress field variation
introduced by Brown and Miller [8] to develop his linear effect by the effect of the amplitude value of the normal and
criterion: the shear stresses acting on a characteristic plane, considering
the time variation of the hydrostatic stress amplitude.
τ a + k (σ a ) max( t ) = f I. Von Mises and Tresca criteria
k = (τ f ) AorB / 2σ u (8)
Tresca and Von Mises criteria, (1) and (2), are suitable for
f = (τ f ) AorB proportional fully reversed loadings. Although the effect of the
mean shear stress is often neglected in HCF analysis [18],
where σ u is the ultimate strenght. According to this criterion, Goodman and Gerber curves reveal that the shear mean stress
at the critical plane maxθ ,ϕ ,ψ ,t (τ a ) is achieved. Therefore, two effect may be accounted for. Fatigue damage of a normal uni-
limit cases (upper and lower limits) are predicted. axial stress that varies with a mean value σ m and an amplitude

F. Carpinteri–Spagnoli theory σ R is related through the following equations to the fully

Carpinteri and Spagnoli proposed the following nonlinear reversed one with the same fatigue damage:
relationships to be evaluated to determine the critical plane [9-
11]:

International Scholarly and Scientific Research & Innovation 6(2) 2012 409 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

Modified Goodman line equation: 2

α= (σ R =−1 / σ R=0 − 1)
(17)
3
­ § σ · 2
°σ R =−1 ¨¨1 − m ¸¸ 0 ≤ σ m β= σ R=−1
σR = ® © σR ¹
(13) 3
° σ R=−1 σm ≤ σR
¯ K.Findley criterion
Two limit cases: a hydrostatic stress field situation and a
Gerber equation: pure torsion loading may be considered to calibrate Findley
and McDiarmid criteria. Therefore, one may write:
σ R = σ R =−1[1 − (σ m / σ u ) 2 ] (14)
­ f = 1 + k 2τ R
°
Similar equations can be written for the shear stress. Test ®ψ = k τR (18)
=
results tend to fall between Goodman and Gerber curves. The ° 1+ k 2 (σ k ) max R , f
material parameter ϕ appeared in (3), may be calibrated using ¯
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

results of the pure bending (or pure tension) and pure torsion
fatigue experiments. Therefore, one may write: where (σ k )max R,f is the hydrostatic pressure value
corresponding to the fatigue failure at the specified ‘R’ value
§ τ max
2
· § σ n ,max ·
2
(if it is R-dependent). The function ψ may be determined
¨¨ ¸¸ + ¨¨ ¸¸ = 1 (15)
using experimental data. Therefore, both ‘f’ and ‘k’ parameters
©τ m +τ R ¹ ©σm +σ R ¹
may be considered as functions of ‘R’. Then, using the mean
stress correction, one has:
or

2 1/ 2
ª§ τ 2
· § σ n ,max · º τ eq = max (τ R + k σ n ) / 1 + k 2 τ R = − 1 / τ R (19)
σ eq = «¨¨ max ¸¸ + ¨¨ ¸¸ » σ R =−1 (16) θ ,ϕ ,ψ ,t
«¬© τ m + τ R ¹ ©σ m +σ R ¹ »¼
On the other hand, if the ‘f’ parameter is considered as a
where σ eq is the equivalent fully reversed normal stress. The function that is independent of ‘R’, then, from the uni-axial
above equations are valid for proportional loadings. For non- cyclic loading, one has:
proportional loadings, σ n ,max and τ max do not occur
σ R2 + k 2 (σ m + σ R ) 2 + k .( σ m + σ R ) = 2 1 + k 2 τ R
simultaneously. Generally, for non-proportional loading, a (20)
criterion whose expressions change with time is more suitable. σ k + 1+ k 2
Ÿ R =0 =
Therefore, the corresponding critical plane may be found by a σ R = −1 2 k + 1 + 4 k 2
time marching method. Thus, equations such as (3), (4), (9),
(10) and (12) that contain stationary parameters, may not seem This equation can be used to determine the ‘k’ parameter.
to be useful for a life assessment under a non-proportional
random loading. Furthermore, (3), (4) and (11) suffer from the L. McDiarmid criterion
fact that their results show no difference between the positive Using the same limit cases that are mentioned in the
and the negative σ m values. It is an evident that compressive previous section, one has:
normal stresses, in contrast to the tensive ones, tend to close
the crack enclosure and subsequently, enhance the fatigue f = (τ a ) AorB = τ R
(21)
strength. (σ a ) R , f = 2σ u
J. Sines criterion
Generally, the second identity does not hold. If we use the
Since for non-proportional loadings, Δ(σ 1 − σ 2 ) ≠ function ψ that is defined in (19), we have:
Δσ 1 − Δ σ 2 ) difficulties arise when calculating (J 2 )a . However, τ eq = [τ a + (σ n ) max(t ) /ψ ].τ R =−1 / τ R (22)
in the present research, in contrast to the other references,
Δ J 2 is not computed substituting the principal stresses by This equation must be calculated on a plane with maximum
their increments in the J 2 expression. Δ J 2 and ( J 2 )a shear stress. Using the limit uni-axial cyclic tension case
where τ a = σ R / 2 and σ n ,max = (σ m + σ R ) / 2 one has:
expressions are derived by a time marching procedure in the
current paper. Besides, according to the uniaxial fatigue
experiments data, it is easily verified that [18]: 2τ R − σ R
k= (23)
σm +σR

International Scholarly and Scientific Research & Innovation 6(2) 2012 410 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

If the ‘f’ parameter is considered as a function that is Specifying:

independent of ‘R’, then, from the uni-axial cyclic tension 1-Load Components time histories.
case, one has: 2-S-N-p or T-N-p curves and material parameters.

1+ k σ σ − σ R =0 (24)
= R =0 Ÿ k = R =−1 Solving the model for the specified loads time histories
1 + 2k σ R = −1 2σ R =0 − σ R =−1 and boundary conditions to extract the stress
components time historis.
M. Carpinteri–Spagnoli theory
Mean stress effect is not included in (9). It is an evidence
Choosing regions of the component that are more
that when the mean shear stress value of a component reaches likely to fail in fatigue.
the ultimate shear stress value, the component will fail with
(τ a = 0, σ n ,max(t ) = 0) . Therefore, it seems that (15) is more
Determination of the equivalent stress time history
general than (9). (including application of rain flow method on the time
histories of the stress components).
N.Ninic equation
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

σ a and τ a used in Ninic equation belong to a plane with

Calculating the cumulated damage
maximum damage and are different from those parameters
appeared in Gough’s equation. The ‘k’ parameter is extracted
from a simple fully reversed uniaxial loading. Ultimate stress Repeating the calculations for the other presecribed
planes (for critical plane approaches)
values are not considered in the maximization procedure.
Besides, there exists noifference between the positive and
negative values of σ m in this formulation. It is proven that the Predicting the component life based on the minimum
of the computed lives.
results of Ninic equation are more accurate than the results of
Fig. 4 The proposed numerical life assessment scheme
the previous mentioned theories [18]. However, the criterion
has some limitations.
IV. INSPECTION OF THE FRACTURE
III. THE PROPOSED LIFE ASSESSMENT PROCEDURE For the analysis of fractured region, a finite element
analysis is carried out for the determination of the highly
In the present paper, equivalent stresses are used to predict
stressed region.
the components fatigue lives. Miner cumulative damage
criterion is used in the present numerical scheme. It is proven A. Finite element stress analysis
that results of Miner’s theory for random loadings are more The anti-roll bar is modeled via ANSYS 11/Mechanical
reliable than those obtained for strain-controlled or stress- module. After the definition of the geometry, the material
controlled load cases [14]. The numerical stages are illustrated properties developed by the mentioned experimental studies
in compact form in Fig. 4. are entered into the program and a static stress analysis is
In the critical plane approaches proposed so far, the critical carried out. Material properties are [15]:
plane orientation is considered to be fixed. Cumulative
damage that is occurred on this fixed plane is considered to σ u = 1300Mpa,
lead to failure. In the present paper, it is assumed that the
critical plane orientation vary with time (Fig. 4). Among all σ R=−1 = 570Mpa,
planes, one with the greatest cumulative fatigue damage is σ R=0 = 960Mpa,
considered as the critical plane. σ yp = 700Mpa,
(τ a ) R =−1 = 400 Mpa,
(τ a ) R =0 = 640Mpa,
E = 206.8Mpa,
υ = 0.3

where σ yp is the yield stress of the material.

The corresponding FEM model in ANSYS software
environment is illustrated in Fig. 6. Support bushings are
modeled using spring elements with six (translational and
rotational) stiffness coefficients. Opposite displacements are
applied to the ends of the anti-roll bar. The time history of the
mentioned displacement input is assumed to be constructed
from a repeated block shown in Fig. 6, and the primary subject

International Scholarly and Scientific Research & Innovation 6(2) 2012 411 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

of the analysis is to determine the highly stressed region of the σ yp . This location is loaded by bending and torsion because
cross-section. This determination allows us to assign the of the geometry and boundary conditions. The highest stress is
beginning point of the fracture. obtained at the inner point of this curved location. The fatigue
The finite element mesh is generated with SOLID 95 crack is initiated and propagated from this location.
elements from the element library. This element is a
tetrahedron with 20 nodes and 30 degrees of freedom and very
sensitive for this type of three-dimensional problems [16, 17].
Smart meshing technique is applied for the mesh generation to
allow the program to determine the most appropriate meshing.
As a result, the element and node numbers of the solution
domain are 18266 and 30514, respectively.
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

Fig. 7 The effective stresses, based on the greatest stress amplitude of

the time history illustrated in Fig. 5
B. Finite element fatigue analysis
Fig. 5 FEM model of the anti-roll bar
Fatigue life assessment of an anti-roll bar component of a
passenger vehicle, is investigated. Experimental data reveal
70
that applying opposite fully reversed vertical displacements
60 with amplitude of 50 mm at the ends of the anti-roll bar, will
50
lead to fatigue failure after approximately 49,500 cycles. It
Vertical Displacement (mm)

has been used the experimental data from reference [18] for
40 validation of FEM results. The FEM model is validated by
30
comparing the static stiffness results obtained from the FEM
analysis with the experimental ones. Number of cycles to
20
failure calculated using FEM model via software ANSYS 11
10
is 55,040 cycles and Number of cycles to failure is calculated
using various theories. Results are given in Table I.
0
To see the mean stress effects, another input is also applied
-10 and also number of cycles to failure calculated using FEM
model via software ANSYS 11 is 140,870 cycles. As before,
-20
0 5 10 15 20 25 the corresponding results are also given in Table I.
Time (seconds) For the special cases studied, according to results shown in
Table I, modified Sines criterion offers the lowest life values.
Fig. 6 The second block signal applied on the anti-roll bar component
For fully reversed displacement inputs, the more accurate
[15]
criteria may be ordered based on the accuracy level as: the
As a result of the study, the maximum stresses obtained at
general criterion, Gouph and Pollard criterion, Ninic criterion
the third bent location of the bar. To insure that no yielding
and the modified von Mises theory, respectively. Since more
occurs during loading, so that the fatigue failure occurs in the
stress components may be considered in the general criterion,
HCF regime, the effective (Von Mises) stresses of the
lower life is predicted by the general criterion in comparison
component are determined based on the greatest stress
with Gough–Pollard or Ninic criteria. Significant
amplitude of the time history illustrated in Fig. 7. The
discrepancies are noticeable in the results of the second type
corresponding results are illustrated in Fig. 7 in Pa. As it may
input. Results of von Mises theory are also computed. They
be noticed from Fig. 6, the maximum value of the effective
are omitted because they were unjustifiably lower than the
stress is 583 MPa which is greater than σ R = −1 and less than
other results.

International Scholarly and Scientific Research & Innovation 6(2) 2012 412 scholar.waset.org/1999.8/14510
 World Academy of Science, Engineering and Technology
International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering Vol:6, No:2, 2012

For the second type of loading, the more accurate criteria [11] A. Spagnoli, “A new high-cycle fatigue criterion applied to out-ofphase
biaxial stress state,” Int J Mech Sci, vol. 43, pp. 2581–95, 2001.
may be ordered based on the accuracy level as: the general [12] HJ. Gough, HV. Pollard, “The strength of metals under combined
criterion, the modified von Mises theory, and the modified alternating stresses,” Proc Inst Mech Eng, vol. 131, pp. 1–103, 1935.
Findley criterion, respectively. Remarkable enhancements are [13] Y. Liu, S. Mahadevan, “A unified multiaxial fatigue damage model for
noticed due to using the modifications proposed in the present isotropic and anisotropic materials,” Int J Fatigue,vol. 29, pp. 347–59,
2007.
paper. Furthermore, there exist good agreements between the [14] DH. Wright, “Testing automotive materials and components,” McGraw-
results of the general criterion and the experimental results. Hill Publishing Co., 1993.
[15] M. Shariyat, “A fatigue model developed by modification of Gough’s
theory, for andom non-proportional loading conditions and three-
V.CONCLUSIONS dimensional stress fields,” Int J Fatigue, vol. 30, pp. 1248–1258, 2008.
In the present paper, as a first stage, some modifications are [16] J. Schijve, “Fatigue of structures and materials,” Secaucus, (NJ, USA):
Kluwer Academic Publishers, 2001.
proposed to some existing fatigue failure models. Many hints [17] ANSYS 5.4 versions users manual.
that may be considered to develop general fatigue failure [18] M. Shariyat, A. Ganjidoust, “Fatigue failure and damage analysis of an
models for three-dimensional stress fields with random, non- anti-roll bar under random fatigue tests,” Journal of Solid Mechanics,
pp. 3–20, 2010.
proportional loadings are mentioned. Then, fatigue life of an
anti-roll bar component of a passenger vehicle, is investigated
International Science Index, Mechanical and Mechatronics Engineering Vol:6, No:2, 2012 waset.org/Publication/14510

by numerical method and finally comparison is made among Javad Marzbanrad is an associate professor in
Automotive Engineering at Iran University of Science
the results of the FEM analysis, results of the existing theories, and Technology. He was born in 1964 in Tehran, Iran.
results of the modified versions of the theories, as well as the He got his Ph.D’s Degree in 2001 from Tarbiat
experimental results. The presented results confirm the Modarres University. He was in the Clarkson
University, NY, U.S. in 2000-01 as visiting scholar.
accuracy of numerical fatigue analysis. His research interests are Automotive, Control and
Vibration and Solid Mechanics. Now, he is the faculty
TABLE I
of School of Automotive Engineering in Iran
COMPARISONS AMONG THE RESULTS OF THE EXISTING THEORIES, RESULTS OF
University of Science and Technology. He has more than 150 published
THE MODIFIED OR PROPOSED CRITERIA, AND THE EXPERIMENTAL RESULTS
papers in the conferences and journals. He also published three books in
Result Title Fully reversed input Input shown in Fig.5 mechanical software fields. He has also registered six inventions in Iran.

This study (ANSYS) 55,040 140,870 Aref Yadollahi was born in 1984 in Semnan, Iran. He
Experimental [18] 49,500 116,820 got his MSc’s Degree in 2010 from Iran University of
Gough-Pollard 70,301 328,371 Science and Technology. His research interests are
Modified sines 23,275 58,901 Automotive, NVH and Solid Mechanics. Now, he is the
Fidley 90,146 280,892 tutor of Automotive Department in the Islamic Azad
McDiarmid 91,713 296,172 University, Semnan Branch. He has more than 10
Carpinteri-Spagnoli 76,799 276,855 published papers in the conferences and journals. He
Ninic 70,301 213,673 also published one book in mechanical software field.
Modified Fidley 82,614 196,563
Modified McDiarmid 78,346 226,389
Modified Von Mises 81,713 186,407
The general criterion 67,168 162,623

REFERENCES
[1] H. Heisler, “Vehicle and engine technology,” 2nd ed. London: SAE
International, 1999.
[2] TD. Gillespie, “Fundamentals of vehicle dynamics,” SAE Publication,
1992.
[3] A. Fatemi, L. Yang, “Cumulative fatigue damage and life prediction
theories: a survey of the state of the art for homogeneous materials,” Int
J Fatigue, vol. 20, no. 1, pp. 9–34, 1998.
[4] P. Haupt, “Continuum mechanics and theory of materials,” Springer-
Verlag, 2002.
[5] G. Sines, “Behavior of metals under complex stresses,” In Metal fatigue,
G. Sines, JL. Waisman, Ed. New York: McGraw-Hill, 1959.
[6] YC. Fung, P. Tong, “Classical and computational solid mechanics,”
World Scientific Publishing Co. Inc., 2001.
[7] WN. Findley, “A theory for effect of mean stress on fatigue of metals
under combined torsion and axial load or bending,” J Eng Ind, pp. 301–
6, 1959.
[8] MW. Brown, KJ. Miller, “A theory for fatigue failure under multiaxial
stress–strain conditions,” Proc Inst Mech Eng, vol. 187, no. 65, pp. 745–
55, 1973.
[9] A. Carpinteri, R. Brighenti, and A. Spagnoli, “A fracture plane
approach in multiaxial high-cycle fatigue of metals,” Fatigue Fract Eng
Mater Struct, vol. 23, no. 4, pp. 355–64, 2000.
[10] A. Carpinteri, A. Spagnoli, “Multiaxial high-cycle fatigue criterion for
hard metals,” Int J Fatigue,vol. 23, pp. 135–45, 2001.

International Scholarly and Scientific Research & Innovation 6(2) 2012 413 scholar.waset.org/1999.8/14510