See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/235278120

Continuum damage mechanics modeling for fatigue life of

elastomeric materials

Article  in  International Journal of Structural Integrity · March 2010

DOI: 10.1108/17579861011023801

CITATIONS READS
6 225

3 authors, including:

Aidy Ali Mina Hosseini

National Defence University of Malaysia 20 PUBLICATIONS   331 CITATIONS   
162 PUBLICATIONS   2,194 CITATIONS   
SEE PROFILE
SEE PROFILE

Some of the authors of this publication are also working on these related projects:

NDT thermography for defect detection of kenaf/PU composites View project

Save World from COVID-19 View project

All content following this page was uploaded by Aidy Ali on 26 July 2018.

The user has requested enhancement of the downloaded file.

 The current issue and full text archive of this journal is available at
www.emeraldinsight.com/1757-9864.htm

Fatigue life of
Continuum damage mechanics elastomeric
modeling for fatigue life of materials
elastomeric materials
63
Aidy Ali, Maryam Hosseini and Barkawi Sahari
Department of Mechanical and Manufacturing Engineering,
Faculty of Engineering, Universiti Putra Malaysia, Serdang, Malaysia

Abstract
Purpose – The purpose of this paper is to investigate the fatigue behavior of rubber using dumb-bell
test specimens under uniaxial loading.
Design/methodology/approach – The material used is a vulcanized natural rubber with a
formulation typical for engine mounts and an international rubber hardness degree of 60. Fatigue tests
are conducted under the displacement controlled condition with a sine waveform of 0.1 Hz and the load
ratio of zero.
Findings – In modeling fatigue damage behavior, a continuum damage model is presented based on
the function of the strain range under cyclic loading. The Ogden strain energy potential is used to
define the constitutive relation of the natural rubber. A good agreement is obtained between fatigue
experimental data and theoretical predictions.
Originality/value – Fatigue analysis and lifetime evaluation are very important in design to ensure
the safety and reliability of rubber components. The design of rubber against fatigue failure is an
important topic that must be considered for safety during operation.
Keywords Elastomers, Rubbers, Fatigue, Modelling
Paper type Research paper

1. Introduction
Rubber as an engineering material is used in many industries, such as the automotive
industry and in a wide range of applications, including engine mountings, tires,
vibration isolators, seals, hoses, belts, impact bumpers, medical devices and structural
bearings. Rubber is an ideal material for many applications because it can withstand
very large strains with no permanent deformation or fracture (Mars, 2002). In addition
to elastic recovery, elastomers have special physical properties (flexibility,
extensibility, resiliency and durability) that are unmatched by other types of
materials (Coran, 2006). These notable characteristics are affected over time by a
number of variables, including fatigue, light, heat, oxygen and ozone (Nagdi, 1993).
Elastomeric parts impose static and dynamic loading condition, which cause failure
due to fatigue (Andriyana and Verron, 2005). Mechanical fatigue of elastomer is
defined as a progressive reduction of the physical properties as a result of crack growth
with continuous dynamic excitation (Abraham et al., 2005). Short- and long-term

The authors acknowledge the Malaysian Rubber Board for providing rubber materials, International Journal of Structural
Integrity
dumb-bell test specimens and supporting experiments for this paper. The authors thank the Vol. 1 No. 1, 2010
Department of Mechanical Engineering, International Islamic University Malaysia for pp. 63-72
q Emerald Group Publishing Limited
supporting this research through permission to use the fatigue test machine in their 1757-9864
laboratory. All of this supports is highly appreciated. DOI 10.1108/17579861011023801
IJSI durabilities should be determined for the development of new industrial products
1,1 (Verron et al., 2006). Many factors are recognized to influence the mechanical fatigue
life of rubber parts, which are categorized according to mechanical loading history,
environmental conditions, formulation of the rubber compound and dissipative
constitutive behavior (Mars and Fatemi, 2004). The nucleation and growth of cracks
are important factors in the long-term durability of rubber and are considered in the
64 modeling and design of mechanical fatigue in rubber components at the early design
stage (Mars and Fatemi, 2002).
In recent years, there has been increasing interest in the fatigue life prediction of the
rubber parts because it is essential for extension of the warranty period of automotive
components. The design of rubber parts versus fatigue failure is important in
preventing damage or failure during operation (Woo et al., 2008).
To improve the durability and the analytical forms applied for predicting the
fatigue lives of rubber, a better understanding of the effects of loading conditions on
fatigue behavior is necessary for the case of cyclic loading (Harbour et al., 2008).
To the best of the author’s knowledge, fatigue life prediction in rubber is still
limited, in comparison to the fatigue behaviors of linear materials (metals and
ceramics) and despite the increasing application of rubber in industry (Abraham et al.,
2005; Saintier et al., 2006). Fatigue life research for elastomeric materials is at an early
stage, because of the complex reactions among polymers, fillers, softeners and other
additives (Abraham et al., 2005). Most studies in this field concern the growth of a
precracked specimen based on fracture mechanics, but there has been little discussion
about fatigue crack initiation in unprecracked specimens. The paper investigates
fatigue behavior using rubber dumb-bell test specimens under uniaxial loading
conditions based on fatigue nucleation. Tests were conducted, and the results are
presented as a Wöhler S-N curve.

2. Experiment procedure
2.1 Sample preparation
A typical engine mount formulation and hardness, IRHD 60 ^ 1, for natural rubber
was selected as the test material, as shown in Table I.
The Type 2 dumb-bell specimen of ISO 37 with a thickness of 2 mm was selected
(Brown, 2006; ISO 37, 1998). The geometry and test length of dumb-bell test specimen
are shown in Figure 1.

Ingredient Parts/100 rubber

SMR 10 100
N550, FEF black 45
Processing oil 10
Zinc oxide 5
Stearic acid 2
Antioxidant, TMQ 2
Antidegradant, 6PPD 2
Table I. Sulfur 2.25
Formulation for the MBTS 1
natural rubber compound DPG 0.2
 25 Fatigue life of
elastomeric
materials
12.5 4

65
8 12.5

75

Figure 1.
Dimensions and test
length of dumb-bell test
specimen in millimetres
20

2.2 Tensile test

Tensile testing was carried out using an Instron 5500 Series IX with 1-kN capacity and
self-tightening roller grips at room temperature. The Type 2 dumb-bell specimen of
ISO 37 with a gauge length of 20 mm was used, and the displacement was measured
with a contact extensometer suitable for soft specimens. The extensometer was
attached to the test piece, and its movement was measured as the test specimen
was elongated. A speed of 500 mm/min was set. The elongation of the specimen was
allowed to continue until rupture. Five samples were used for the tensile test, and an
average of results was taken as the resultant value.

2.3 Fatigue test

Fatigue tests were conducted with a servo-hydraulic fatigue machine (Shimadzu Co.,
EHF-EM100KN-20L) at room temperature. In this experiment, the specimen design
was the same as the tensile stress-strain test samples. The tests were performed under
displacement-controlled loading conditions and a sinusoidal waveform. The strains
were measured from the displacements and the initial gauge length between two grips.
For all test samples, a frequency of 0.1 Hz was selected due to limitations of the fatigue
test machine for the displacement of the grips. Furthermore, this frequency was
applied for failure caused by the initiation and growth of cracks and to avoid large
increases in temperature on the specimen surface, which would consequently cause
thermal break-down.

3. Experimental results
3.1 Material characterization
In this study, the mechanical behavior of the material is defined by the Ogden model.
The experimental methodology for determining the material properties is presented
here using uniaxial stress-strain data. In the following, the Ogden model is presented
and reformulated in terms of the uniaxial stress-strain relation. The relation of the
Ogden strain energy potential is given by:
IJSI Xn
2mi

ai a a  X n
1
U¼ l1 þ l
2 i þ l
3 i 2 3 þ ð J 2 1Þ2i : ð1Þ
1,1 i¼1
a 2
i i¼1
D i

where:
l
i ¼ J 21=3 li , J ¼ l1 l2 l3 and li ¼ the principal stretches.
66 J ¼ the Jacobean determinant.
n ¼ the number of terms in the series.
mi and ai ¼ the shear behavior of the material (constants).
Di ¼ the compressibility.
When the material is assumed to be incompressible, J ¼ l1 l2 l3 ¼ 1 (Forni et al., 1999).
In the uniaxial stress state, the principal stretches in the uniaxial loading condition
are presented by:
1
l1 ¼ lU ; l2 ¼ l3 ¼ pffiffiffiffiffiffi : ð2Þ
lU
where:
lU ¼ the stretch in the loading direction.
l2 and l3 ¼ the principal stretches on the planes perpendicular to the loading
direction.
The stretch ratio is described as the ratio of the extended length of a specimen to the
original length. Therefore, the nominal strain can be defined based on the principal
stretch ratio as 1U ¼ lU 2 1.
To determine the stress and strain relation under uniaxial tension, the uniaxial
nominal stress (sU) can be considered as follows (Kim et al., 2004; Wang et al., 2002):
›U
sU ¼ : ð3Þ
›lU
Using the Ogden strain energy potential of order 1 (n ¼ 1), the strain energy potential
(U ) for the uniaxial stress state can be obtained in the form of:
2m1  a1 2a1 =2
 2m
1
U¼ 2
lU þ 2 l U 2 3 ¼ 2 ½ð1 þ 1U Þa1 þ 2ð1 þ 1U Þ2a1 =2 2 3
: ð4Þ
a1 a1
Therefore, the nominal stress and strain relation under uniaxial tension (sU) will be
defined as:
2m1  a1 21 2ða =2þ1Þ
 2m
1
sU ¼ lU 2 lU 1 ¼ ½ð1 þ 1U Þa1 21 2 ð1 þ 1U Þ2ða1 =2þ1Þ
: ð5Þ
a1 a1
where material parameters m1 and a1 can be obtained by fitting the experimental
stress-strain curve into equation (5).
 The coefficients m1 and a1 can be obtained by using the least square method Fatigue life of
available in Excel software, and these coefficients are 1.2 and 2.9 MPa, respectively.
Figure 2 shows the nominal stress-strain curve and data fit with Ogden model. The
elastomeric
results of tensile testing indicate that the developed Ogden model is suitable for materials
explaining the correlation between stress and strain for the material used in this work.

3.2 The fatigue properties of natural rubber material 67

The fatigue life versus strain range (1 2 n) curve in double logarithmic coordinates is
shown as a line in Figure 3 with the related 1 2 n equation and regression.
The general fatigue life equation for natural rubber can be defined as follows
(Li et al., 2009):
1 m nf ¼ C: ð6Þ
where m and C are constants and can be calculated by the fatigue test results and data
fit curve using the least square method in Excel software. Based on the 1 2 n curve of
the natural rubber by data fit, m and C were found to be 4.54 and 439,940, respectively.
Therefore, the fatigue life equation can be presented as:
1 4:54 nf ¼ 439; 940: ð7Þ
Equation (7) can be rewritten in double logarithmic coordinates as follows:
log nf ¼ 5:64 2 4:54 log 1: ð8Þ

4. Modeling using continuum damage mechanics theory

A theoretical equation was developed to relate the fatigue life as a function of the strain
range under cyclic loading. The comparison between the predicted fatigue lives of the
rubber test specimen with the experimental life was done.

30
Experiment
Ogden model
25

20
Stress (Mpa)

15

10

5
Figure 2.
Nominal stress-strain
0 behavior with
0 1 2 3 4 5 6 Ogden model
Strain (mm/mm)
IJSI 10
1,1 Experimental
Power law fit

Strain range, ∆ε (mm/mm)

68

∆ε = 17.50 (Nf)–0.22
R2 = 0.99

Figure 3.
The fatigue life versus
1
strain range in the double 1.E+01 1.E+02 1.E+03 1.E+04 1.E+05 1.E+06
logarithmic coordinates
Fatigue life, Nf (cycle)

Continuum damage mechanic (CDM) has been used in different areas such as metals,
composites and elastomers. The damage at different sites can merge and make a
microscopic crack causing the failure. The theory was used in the study of fatigue
damage phenomena such as the nucleation and initial defect growth (microvoids and
microcracks) in elastomers by Mahmoud et al. (2007) and Wang et al. (2002). They have
verified that the theory of CDMs is suitable for the study of fatigue damage behavior
for elastomers.
According to the CDM theory, the effective nominal normal stress is defined as
s
U ¼ ðsU Þ=ð1 2 DÞ using the effective nominal stress concept through the damage
variable D. Therefore, the nominal stress-strain relation of a damaged material
resembles a material with no damage and is expressed by the following using the
Ogden model as a constitutive model:
sU 2m1  a1 21 2ða =2þ1Þ

s
U ¼ ¼ lU 2 lU 1
12D a1
ð9Þ
2m1
¼ ½ð1 þ 1U Þa1 21 2 ð1 þ 1U Þ2ða1 =2þ1Þ
:
a1
By considering the hypothesis of isotropic damage, the constitutive equation for
damage evaluation is given by:
›f*
D_ ¼ 2 : ð10Þ
›y
where f* and y are the dissipation potential and the damage strain energy rate,
respectively:
b 2yaþ1
f* ¼ : ð11Þ
aþ1 b
In this equation, a and b are material parameters. The damage strain energy release Fatigue life of
rate ( y) in the uniaxial stress state is expressed by the following, in which the damage
strain energy is a function of the effective nominal normal stress:
elastomeric
materials
›U ›U ðs
U Þ
2y ¼ ¼ : ð12Þ
›D ›D
It can be rewritten as the following by using equation (4): 69
›U ðs
U Þ ›U ›lU ðs
U Þ 2m1  a1 21 
2ða =2þ1Þ ›lU ðs
U Þ
2y ¼ ¼ ¼ lU 2 lU 1 : ð13Þ
›D ›lU ›D a1 ›D
Taking partial derivative of equation (9) with respect to D:
›lU ðs
U Þ a1 s U 1
¼ 2
h i: ð14Þ
›D 2m1 ð1 2 DÞ ða1 2 1Þla1 22 þ ða1 =2 þ 1Þl2ða1 =2þ2Þ
U U

By substituting equation (14) into equation (13), the damage strain release rate is
defined as:
h i2
2ða =2þ1Þ
2m1 laU1 21 2 lU 1
2y ¼ h i: ð15Þ
a1 ð1 2 DÞ ða1 2 1Þla1 22 þ ða1 =2 þ 1Þl2ða1 =2þ2Þ
U U

Equation (10) can be rewritten by using equations (11) and (15):

8 h i2 9a
>
< l a1 21
2 l
2ða1 =2þ1Þ >
=
_ ›w* 2m1 U U
D¼2 ¼ h i : ð16Þ
›y :a1 bð1 2 DÞ ða1 2 1Þla1 22 þ ða1 =2 þ 1Þl2ða1 =2þ2Þ >
> ;
U U

The fatigue damage evolution per cycle is defined as follows, where the principal
stretch range ðDlU Þ is used to replace (Dl):
8 h i2 9a
>
< a1 21
lU 2 lU
2ða1 =2þ1Þ >
=
dD 2m1
¼ h i : ð17Þ
dn > :a1 bð1 2 DÞ ða1 2 1ÞlaU1 22 þ ða1 =2 þ 1ÞlU 2ða1 =2þ2Þ >
;

The damage value at any cycle can be determined by integration equation (17),
assuming that the damage variable (D) is zero when n ¼ 0, at the beginning of the cycle
loading:
Z D
d½1 2 ð1 2 DÞ1þa

0
8 h i2 9a
Z n>
<2m la1 21
2 l
2ða1 =2þ1Þ >
= ð18Þ
1 U U
¼ ð1 þ aÞ h i dn:
0 >
2ða1 =2þ2Þ >
: a1 b ða1 2 1ÞlaU1 22 þ ða1 =2 þ 1ÞlU ;
IJSI The relation between the damage variable (D) and the number of cycles (n) is given by:
1,1
8 h i2 9a
>
<2m 2ða =2þ1Þ
laU1 21 2 lU 1 >
=
1
1 2 ð1 2 DÞ1þa ¼ ð1 þ aÞ h i n: ð19Þ
> a b
: 1 ða1 2 1ÞlaU1 22 þ ða1 =2 þ 1Þl2ð a1 =2þ2Þ >
;
U
70
where the condition  the fatigue rupture is D ¼ Dc ¼ 1. At the moment of failure, the

of
number of cycle nf is presented by:

8 8 h 9a 921
i2
>
< >
<2m 2ða1 =2þ1Þ
laU1 21 2 lU >
=> =
1
nf ¼ ð1 þ aÞ h i : ð20Þ
>
: > 2ða1 =2þ2Þ > >
: a1 b ða1 2 1ÞlaU1 22 þ ða1 =2 þ 1ÞlU ;;

By using DlU ¼ 1 þ D1U , the fatigue life can be expressed as a function of the strain
range ðD1U Þ for elastomeric materials under cyclic loading:

( ( )a )21
2m1 ½ð1 þ D1u Þa1 21 2 ð1 þ D1u Þ2ða1 =2þ1Þ
2
nf ¼ ð1 þ aÞ :
a1 b ½ða1 2 1Þð1 þ D1u Þa1 22 þ ða1 =2 þ 1Þð1 þ D1u Þ2ða1 =2þ2Þ

ð21Þ

In equation (21), a and b are determined by the experimental fatigue life as a function of
the strain range. Based on the experimental results and curve fitting, two material
properties including a and b are obtained as 2 and 2,163 MPa, respectively.
Equation (21) can be written after substituting unknown parameters m1, a1, a and b:

 22
½ð1 þ D1Þ1:9 2 ð1 þ D1Þ22:45
2
nf ¼ 3:1 £ 106 : ð22Þ
1:9ð1 þ D1Þ0:9 þ 2:45ð1 þ D1Þ23:45

Figure 4 shows correlation between the experimental fatigue life and the theoretical
prediction based on the proposed damage model. This relation indicates good
agreement within a factor of 1.3.

5. Conclusions
A theoretical equation for the fatigue life of natural rubber was developed and is
presented based on strain range. The CDMs approach was used to predict fatigue life.
The correlation of the predicted life with the experimental data is strong. This supports
the idea that the CDMs-based model can be applied to study fatigue life for natural
rubber.
 6 Fatigue life of
Experimental elastomeric
5 Prediction materials
Strain range, ∆ε (mm/mm)

4
71
3

1
Figure 4.
Relation between the
0 experimental and
1.E+01 1.E+02 1.E+03 1.E+04 1.E+05
predicted fatigue life
Fatigue life, Nf (cycle)

References
Abraham, F., Alshuth, T. and Jerrams, S. (2005), “The effect of minimum stress and stress
amplitude on the fatigue life of non strain crystallizing elastomers”, Mater. Des., Vol. 26,
pp. 239-45.
Andriyana, A. and Verron, E. (2005), “Effect of the hysteretic response of elastomers on
the fatigue life”, in Austrell, P.E. and Kari, L. (Eds), Constitutive Models for Rubber IV,
A.A. Balkema, Rotterdam, pp. 31-6.
Brown, R. (2006), Physical Testing of Rubber, Springer, New York, NY.
Coran, A.Y. (2006), “Elastomers”, in Harper, C.A. (Ed.), Handbook of Plastics Technologies,
McGraw-Hill, New York, NY, pp. 4.1-4.111.
Forni, M., Martelli, A. and Dusi, A. (1999), “Implementation and validation of hyperelastic finite
element models of high damping rubber bearings”, in Dorfmann, A. and Muhr, A. (Eds),
Constitutive Models for Rubber, A.A. Balkema, Rotterdam, pp. 237-47.
Harbour, R.J., Fatemi, A. and Mars, W.V. (2008), “Fatigue life analysis and predictions for NR and
SBR under variable amplitude and multiaxial loading conditions”, Int. J. Fatigue, Vol. 30
No. 7, pp. 1231-47.
ISO 37 (1998), “Rubber, vulcanized or thermoplastic – determination of tensile stress-strain
properties”, International Organization for Standardization, Geneva.
Kim, W.D., Lee, H.J., Kim, J.Y. and Koh, S.K. (2004), “Fatigue life estimation of an engine rubber
mount”, Int. J. Fatigue, Vol. 26 No. 5, pp. 553-60.
Li, Q., Zhao, J. and Zhao, B. (2009), “Fatigue life prediction of a rubber mount based on test of
material properties and finite element analysis”, Eng. Fail. Anal., Vol. 16, pp. 2304-10.
Mahmoud, W.E., Mansour, S.A., Hafez, M. and Salam, M.A. (2007), “On the degradation and
stability of high abrasion furnace black (HAF)/acrylonitrile butadiene rubber (NBR) and
IJSI high abrasion furnace black (HAF)/graphite/acrylonitrile butadiene rubber (NBR) under
cyclic stress strain”, Polym. Degrad. Stab., Vol. 92, pp. 2011-15.
1,1 Mars, W.V. (2002), “Cracking energy density as a predictor of fatigue life under multiaxial
conditions”, Rubber Chem. Technol., Vol. 75 No. 1, pp. 1-17.
Mars, W.V. and Fatemi, A. (2002), “A literature survey on fatigue analysis approaches for
rubber”, Int. J. Fatigue, Vol. 24 No. 9, pp. 949-61.
72 Mars, W.V. and Fatemi, A. (2004), “Factors that affect the fatigue life of rubber: a literature
survey”, Rubber Chem. Technol., Vol. 76, pp. 391-412.
Nagdi, K. (1993), Rubber as an Engineering Material, Hanser, Munich.
Saintier, N., Cailletaud, G. and Piques, R. (2006), “Multiaxial fatigue life prediction for a natural
rubber”, Int. J. Fatigue, Vol. 28, pp. 530-9.
Verron, E., Le Cam, J.B. and Gornet, L. (2006), “A multiaxial criterion for crack nucleation in
rubber”, Mech. Res. Commun., Vol. 33, pp. 493-8.
Wang, B., Lu, H. and Kim, G. (2002), “A damage model for the fatigue life of elastomeric
materials”, Mech. Mater., Vol. 34, pp. 475-83.
Woo, C.S., Kim, W.D. and Kwon, J.D. (2008), “A study on the material properties and fatigue life
prediction of natural rubber component”, Mater. Sci. Engng. A, Vols 483-484, pp. 376-81.

About the authors

Aidy Ali is a Lecturer of Mechanical Engineering at Universiti Putra Malaysia (UPM).
He received his first degree in Mechanical Engineering from Putra University in 1999; then
continued studies at National University Malaysia in Manufacturing System Engineering,
leading to Master degree. He is working on simulating crack propagation of elastic plastic
material under mixed mode loading conditions. He was awarded with a ME degree in 2002. After
that, he pursued his PhD degree in Improving Fatigue Life of Aircraft Components by using
Surface Engineering at Sheffield University, UK. He was awarded a PhD degree in March 2006.
During his PhD training, he was attached with Airbus UK in Chester and Manchester to develop
a treatment for an ageing aircraft wing structures. His expertise is related to mechanical fatigue
and fracture of metallic materials, failure assessment, failure prevention analysis and reliability
engineering prediction.
Maryam Hosseini was born in Iran on 20 February 1974. She holds a Bachelors’ degree in
Mechanical Engineering from Mashhad Ferdosi University, Iran in 1998. Later on, she continued
her education in Mechanical Engineering as Master student at UPM in December 2007.
Barkawi Sahari is a Professor in Mechanical Engineering, Director of Institute of Advanced
Technology and Leader for CNGDI Engine and Transmission Research Program at UPM.
He holds a BSc (Hons) in 1981 and a PhD in 1984, degrees from Nottingham University, UK.
Since graduation, he serves as a Lecturer in the Department of Mechanical and Manufacturing
Engineering, UPM. He specialised in stress analysis and his works are related to elastic plastic
analysis, crash analysis and natural gas vehicle. His current research interest is in compressed
natural gas direct injection vehicle structural body design.

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints

View publication stats