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

net/publication/258397854

The Cohesive Zone Model for Fatigue Crack Growth

Article  in  Advances in Mechanical Engineering · October 2013

DOI: 10.1155/2013/737392

CITATIONS READS

9 1,239

3 authors, including:

Jinxiang Liu
Beijing Institute of Technology
36 PUBLICATIONS   391 CITATIONS   

SEE PROFILE

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

micro combustion View project

All content following this page was uploaded by Jinxiang Liu on 28 October 2015.

The user has requested enhancement of the downloaded file.

Hindawi Publishing Corporation
Advances in Mechanical Engineering
Volume 2013, Article ID 737392, 16 pages
http://dx.doi.org/10.1155/2013/737392

Review Article
The Cohesive Zone Model for Fatigue Crack Growth

Jinxiang Liu,1 Jun Li,1 and Bo Wu2

1
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China
2
China North Engine Research Institute, Tianjin 300400, China

Correspondence should be addressed to Jinxiang Liu; liujx@bit.edu.cn

Received 12 May 2013; Accepted 27 August 2013

Academic Editor: Indra Vir Singh

Copyright © 2013 Jinxiang Liu et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

In the past decade, the cohesive zone model has been receiving increasing attention as a powerful tool for the simulation of fatigue
crack growth. When applying cohesive zone model to fatigue fracture problem, three aspects should generally be taken into account,
that is, unloading-reloading path, damage evolution during cyclic loading, and crack surface contact and friction behavior. This
paper addresses the critical views of these aspects. Before that, the formulation of cohesive zone model and identification of cohesive
zone model parameters and its numerical implementation have been reviewed.

1. Introduction employed to treat fracture problems involving large-scale

Fatigue fracture is one of the failure modes in engineering
where crack initiates, propagates, and finally results in the
failure of components under cyclic loading. It is widely
spread in the fields of aerospace, transportation, machinery,
and other industries, bringing on great economic loss and
casualties every year. Therefore, it is extraordinary significant
to avoid the occurrence of fatigue fracture. To do this, the
factors influencing fatigue fracture have to be clear. The
accurate and efficient prediction of fatigue crack growth is
helpful for making sure how the influencing factors work on
fatigue fracture.
For fatigue crack growth prediction, fracture mechanics
has been applied and a great number of investigations can be
found in the literature. The well-known Paris law in the linear
fracture mechanics successfully predicts the fatigue crack
growth for the conditions of small-scale yielding, constant
amplitude loading, and long cracks. The modified ones have
also been proposed to incorporate variable amplitude loading
[1], stress ratio effect [2, 3], crack closure [4], small cracks
[5, 6], and so on. However, this method is only applicable
to the conditions of small-scale yielding. In addition, while
calculating fatigue crack growth with this method, the stress
intensity factor range Δ𝐾 keeps constant in every increment,
which is not consistent with the real conditions. The con-
cept of 𝐽-integral in nonlinear fracture mechanics has been
yielding [7–9], but it cannot be used under cyclic loading.
Although recently researchers propose cyclic 𝐽-integral [10,
11] to characterize the fatigue crack growth, it is difficult to be
evaluated because of the integral operation.
As an alternative method, the cohesive zone model
(CZM) has been receiving increasing attention. The concept
of CZM can be traced back to the strip yield models
of Dugdale [12] and Barenblatt [13]. Without the crack
nucleation criteria used in classical fracture mechanics, the
CZM removes the unreal stress singularity at the crack
tip and regards crack growth as a progressive process of
material degeneration. An appealing feature of this approach
is that it can be easily implemented in various computational
methods, such as finite element method (FEM). CZM is
incipiently used to deal with fracture problem and then
extended to predict fatigue fracture. Recently, this method
has been successfully employed to predict fatigue crack
growth in metals [14–17], composite materials [18], fiber-
metal laminate [19], weld specimen [20, 21], and so on.
In this paper, we mainly summarize the CZM for fatigue
crack growth. Comparing the CZM for fracture, three addi-
tional aspects should be generally included, incorporating
loading-unloading path, damage evolution during cyclic
loading, and crack surface contact and friction behavior.
Critical views of these aspects are addressed in Section 3.
Before that, the formulation of CZM, identification of CZM

Downloaded from ade.sagepub.com by guest on September 2, 2015

2 Advances in Mechanical Engineering
parameters, and implementation of CZM are reviewed in
Section 2. With this paper, we expect to help readers capture
an insight into the concept of CZM and its important aspects
in fatigue crack prediction, so that a better use of it can be
achieved.
2. Cohesive Zone Model
2.1. Traction-Separation Law. During crack propagation,
there exists a narrow layer termed the cohesive zone where
microvoids initiate, grow, and finally coalesce with crack
formation. It consists of two fictitious cohesive surfaces,
which superpose with each other in the undeformed con-
figuration. Under external loading, the cohesive surfaces
can be separated but connected by cohesive traction. The
relation between cohesive traction and displacement jump
of cohesive surfaces is usually called traction-separation law.
As a phenomenological model, there is no evidence which
form for traction-separation law to follow exactly. There-
fore, a lot of traction-separation laws have been proposed
by researchers according to the fracture characteristics of
material, such as ductile or brittle and pure or mixed mode. In
this section, traction-separation laws are reviewed according
to the fracture mode as follows.
2.1.1. Pure Mode. With regard to mode I fracture, the relation
between normal traction and normal separation should be
defined in the traction-separation law. For the relation,
Needleman [27] proposed a polynomial form in which the
traction first increases to a peak value 𝑇0 and then decreases
to zero at a critical separation 𝛿0 , which fractures the cohesive
zone. Hillerborg et al. [28] suggested a linear decreasing
form as displayed in Figure 1(c). The traction starts with an
infinite stiffness until the cohesive stress 𝑇0 is reached. Other
typical traction-separation laws used by authors for mode I
fracture include exponential form proposed by Needleman
[29], bilinear form by Geubelle and Baylor [30], constant
form by Yuan et al. [31], and multilinear form by Tvergaard
and Hutchinson [32], as shown in Figure 1.
For mode II fracture, linear and trigonometric relation
between tangential traction and tangential separation were,
respectively, used by Needleman [27, 35]. The trigono-
metric form describes a periodic dependence of the trac-
tion on tangential separation. Geubelle and Baylor [30]
employed a bilinear form. Tvergaard [36] applied a polyno-
mial (quadratic) form to mode II crack. Xu and Needleman
[37] utilized the exponential form to predict shear separation,
while the linear decreasing form was adopted by Camacho
and Ortiz [38]. The mentioned traction-separation laws have
been shown in Figure 2. Recently, Chen and Linzell [39]
introduced the trapezoidal form to simulate mode II fracture.
Dourado et al. [40] used another bilinear form instead.
For mode III fracture, a bilinear form like the one in
Figure 1(d) was widely used by researchers [15, 23]. A possible
form was also formulated by Zhang and Deng [22] from
elastic stress and displacement fields around the crack tip. The
normalized plot of the traction-separation relation is shown
Downloaded from ade.sagepub.com by guest on September 2, 2015
in Figure 3. Actually, less traction-separation laws used for
mode III fracture can be found in the published literature.
2.1.2. Mixed Mode. As for mixed mode crack growth, Xu and
Needleman [37] extended the exponential form to predict
combined normal and shear separation. In the formulation,
the traction components are regarded as the function of
uncoupled separations in normal and tangential direction.
Unlike this model, Tvergaard [36] proposed a quadratic CZM
with a coupling parameter in which the traction components
are expressed as the function of the relative separation and
this parameter. The parameter is introduced to characterize
the interaction of normal and shear separation. Similar
CZMs for mixed mode have also been proposed in [30, 43].
Camacho and Ortiz [38] rebuilt the linear decreasing form
with the concept of effective quantity instead and then applied
it to brittle material under mixed mode. With this concept,
Yuval and Leslie [44] recently put forward a new cohesive
zone model for mixed mode interface fracture in bimaterials.
Other similar CZMs can be also found in [45].
2.2. Identification of CZM Parameters. Generally, CZM
involves three important parameters. They are the critical
separation 𝛿𝑐 , the cohesive strength 𝑇0 , and the cohesive
energy Γ0 , that is, the area under traction-separation law.
The cohesive strength 𝑇0 is usually regarded as a material
parameter not only for brittle material but also for ductile
material. The selection of the second material parameter is
up to the preferred local failure criterion and cannot be
determined in advance. However, it is usual to take Γ0 and
𝑇0 to be the cohesive constitutive parameters in the research.
The cohesive parameters of the traction-separation law in
each fracture mode should be determined for application of
this technique [46]. Hence, the methods used to determine
cohesive parameters for ductile material and brittle material
are reviewed, respectively, according to the fracture mode.
It should be noted that some methods described below have
also been applied to the parameter identification of laminated
composites, adhesive and spot-weld joints though.
2.2.1. Ductile Material. For mode I crack, both experimental
method and numerical method are used for determining
the cohesive parameters. For the determination of 𝑇0 , it
can be obtained by the common way of fitting the crack
growth resistance curve of fracture mechanics specimens
after determining the cohesive energy Γ0 . Lin et al. [47]
determined 𝑇0 through fitting the load-deformation curve of
a notched tensile bar as well. Cornec et al. [48] employed
a hybrid technique combining experimental and numerical
simulation. In this method, 𝑇0 equals the maximum axial
stress of the notch section at the instant of crack initiation
in FEM analysis, which coincides well with final fracture
in experiment. For the determination of Γ0 , three different
methods have been introduced to evaluate Γ0 , including
potential drop method, multiple specimen method, and the
blunting line. These methods are all based on the assumption
that Γ0 equals the 𝐽-integral at the initiation of stable crack
Advances in Mechanical Engineering 3

Tn Tn

T0

T0

𝛿n 𝛿n
𝛿c 𝛿c

(a) (b)
Tn Tn
T0
T0

𝛿n 𝛿n
𝛿c 𝛿c

(c) (d)
Tn Tn

T0 T0

𝛿n 𝛿n
𝛿c 𝛿1 𝛿2 𝛿c

(e) (f)

Figure 1: Various cohesive laws used by several authors for normal separation: (a) polynomial form, (b) exponential form, (c) linear decreasing
form, (d) bilinear form, (e) constant form, and (f) trilinear form.

extension and fracture mechanics specimens, such as side-
grooved compact tension specimen as shown in Figure 4(a),
are usually used.
For modes II and III, there are no generally accepted
test procedures available for the determination of crack
initiation [48]. In addition, it may readily lead to local mode
I separation under global mode II loading, where the crack
extends perpendicularly to the principal normal stress and
deviates from the original crack extension direction. Due to
these difficulties, few investigations have been made for the
determination of cohesive parameters for shear separation so
far and more work is needed to be done in the future.
2.2.2. Brittle Material. For brittle material, the determination
of the cohesive parameters is focused on mode I fracture

Downloaded from ade.sagepub.com by guest on September 2, 2015

4 Advances in Mechanical Engineering
Tt
𝛿t
(a)
Tt
𝛿t
(d)
Figure 2: Various cohesive laws used by several authors for shear separation: (a) linear form, (b) polynomial form, (c) exponential form, (d)
trigonometric form, (e) linear decreasing form, and (f) bilinear form.
Tt /T0
𝛿t /𝛿0
Figure 3: The traction-separation law (in normalized form) for
mode III fracture [22].
and few investigations about other modes can be found in
the literature. Therefore, only the determination of mode I
cohesive parameters is considered in this section.
For mode I crack, there are direct methods and indirect
methods for the identification of CZM parameters. In early
research, uniaxial tensile test was widely used to determine
the cohesive parameters of mode I fracture [49]. Though it is
regarded as the most direct way to characterize the fracture
properties for brittle material, it is generally accepted that
Downloaded from ade.sagepub.com by guest on September 2, 2015
Tt
Tt
𝛿t
𝛿t
(b) (c)
Tt Tt
𝛿t
𝛿t
(e) (f)
the test is difficult to perform [50–52]. Therefore, Rocco et
al. [53] used a simple splitting tensile test to determine the
cohesive strength. Linsbauer and Tschegg [54] proposed a
wedge splitting test which was designed to measure some
mode I fracture properties, specifically the cohesive energy Γ0 .
In Figures 4(b) and 4(c), two different specimens for wedge
splitting test are displayed. The three-point bend test was also
employed to measure the cohesive energy.
Indirect method, typically through an inverse analysis,
is often introduced to obtain parameters for brittle material.
Roelfstra and Wittmann [55] proposed a finite element
scheme for characterizing a bilinear law through fitting the
experimental records. Mihashi and Nomura [56] developed
the moving-window data fitting technique. In these works,
optimization procedures were employed, which involves the
minimization of a chosen objective function expressed as
the norm of the difference between the calculated and
experimental data. Bolzon et al. [57, 58] have employed a
general formulation of the softening law for inverse analysis
in complementarity format and formulated the parameter
identification problem as a mathematical programming prob-
lem. Other indirect methods include compliance technique
employed by Hu and Mai [59] and general bilinear fit by
Guinea et al. [60].
It should be noted that the methods mentioned above
for brittle material are for the assumed form of traction-
separation law, with the cohesive parameters only to be
determined. Methods for determining the complete shape of
traction-separation law have also been used by researchers,
including 𝐽-integral method proposed by Li et al. [61], saw
Advances in Mechanical Engineering
(a)
Figure 4: Specimens for the determination of CZM parameters: (a) fracture mechanics specimen, (b) cubical wedge splitting specimen, and
(c) cylindrical wedge splitting specimen.
T FEM. However, the classic boundary integral equation can
𝛿 the combination of CZM and XFEM, the displacement jump
Figure 5: Cohesive law with unloading-reloading behavior for
brittle material [23].
cutting technique developed by Hu and Wittmann [62–64],
and back analysis method introduced by Nanakorn and Horii
[65].
2.3. Implementation of CZM. As mentioned above, CZM fits
naturally within the conventional framework of finite ele-
ment analysis. For finite element implementation, researchers
have embedded CZM into cohesive finite elements [66–
68]. These elements are surface-like and are compatible
with general bulk finite element discretization of the solid,
bridging nascent surfaces and governing their separation in
accordance with a cohesive law. An alternative approach used
is to embed CZM as a mixed boundary condition [25, 69–
71]. Compared with the former one, there is no need for the
element stiffness matrix to be defined here.
Another method for CZM implementation is the com-
bination with boundary element method (BEM) [72]. The
method is attractive because the boundary alone has to be
discretized and the dimensionality of the stiffness matrix
formed in BEM can be then reduced in comparison to
Downloaded from ade.sagepub.com by guest on September 2, 2015
5
(b) (c)
not be applied to crack problems, which will lead to an ill-
conditioned stiffness matrix. The reason is that crack surfaces
are coincident while modelling a cracks with this method. To
deal with this problem, many approaches have been proposed
by researchers, such as the subdomains method [73], the
dual boundary element method [74], and the single-domain
traction boundary element method [42, 75]. For a discussion
of some of these approaches, one may refer to relative paper
[76].
In recent years, researchers have implemented CZM
into extended finite element method (XFEM) [77] as well.
The XFEM approximates the discontinuity by introducing
enriched degrees of freedom and an additional enriched
function into the node of conventional finite element. With
can be incorporated into the conventional finite elements and
the path of the discontinuity is completely independent of the
mesh structure [78, 79]. In other words, the algorithm allows
introducing a new crack surface at arbitrary locations and
directions in a finite element mesh, without remeshing [80].
3. Fatigue Crack Growth
Applying CZM to the specific field of fatigue to characterize
the process of crack growth under cyclic loading, some
aspects have to be dealt with in advance. First of all, one
has to define the behavior of CZM under local unloading
to account for the irreversible damage process in cases of
global unloading, crack branching, or multiple cracks. Then
in order to capture finite life effects in fatigue, CZM is to
embrace the constitutive relation for damage accumulation
under cyclic loading. In addition, the potential contact and
friction behavior of the crack surfaces should be taken into
account.
3.1. Unloading-Reloading Path of CZM. In order to study
the structural behavior such as its load-deflection curve,
it is usually sufficient for CZM to deal only with loading
conditions [81]. This is perhaps the main reason why CZMs
proposed in the initial period are reversible and history
6 Advances in Mechanical Engineering

independent, without the unloading and reloading behaviors However, unloading is often separately treated for normal
being considered. The study of crack propagation, therefore, and tangential separation by researchers [38]. This treatment
is limited to cases under monotonic loading and without means that unloading and loading can occur in different
complicated crack behaviors. directions at one time.

3.1.1. Pure Mode. Under pure mode, researchers employed 3.2. Damage Evolution. Using the configuration mentioned
different unloading-reloading paths for brittle and ductile above to simulate fatigue cracks will bring infinite life
fracture, respectively. For brittle material, Camacho and Ortiz prediction, which is inconsistent with the real conditions.
[38] employed a linear unloading/reloading within CZM, To capture finite life effects in fatigue, damage evolution
as shown in Figure 5. The unloading and reloading follow during fatigue loading should be taken into account. A widely
the same path and separation decreases to zero when the used method is to track unloading-reloading and degrading
stress vanishes with reduced stiffness. This local unloading- stiffness.
reloading configuration has been also used by Li et al. [81]
for mode I, Chen and Linzell [39] for mode II, and Yang and 3.2.1. Partial Unloading-Reloading. On the foundation of
Ravi-Chandar [23] for mode III fracture problem. CZM with the linear unloading-reloading behavior, De-
For ductile material, Scheider [24] and Roe [82] intro- Andrés et al. [25] introduced a partial unloading-reloading
duced a distinct unloading-reloading path, as shown in configuration and characterized the accumulated damage
Figure 6. The reduction of separation behaves elastically with with a damage variable 𝐷, which is defined as
the initial stiffness at the origin. In this configuration, there
exists the possibility for the presence of a residual displace- Φ (𝛿max )
ment continuity. It may be explained by the fact that the 𝐷= , (2)
inelastic separation is irreversible since the work for inducing 𝐺𝑐
damage is totally dissipated through microcrack nucleation,
where 𝛿max , Φ(𝛿max ), and 𝐺𝑐 are the maximum attained
void growth, and coalescence. This unloading-reloading path
separation, corresponding dissipated energy to 𝛿max , and
was also used by Deshpande et al. [83, 84] to predict the near-
fracture energy, respectively. 𝐷 ranges from 0 to 1, with these
threshold fatigue crack growth in metal single crystals. In
limits referring to an uncracked solid and a fully formed new
the paper, a microstructurally based approach was proposed
fracture surface. A point on the cohesive zone at the crack tip
using discrete dislocation dynamics in the bulk material in
is considered for CZM to account for accumulated damage
combination with a CZM. The CZM followed a universal
process as depicted in Figure 7. Suppose that the forward
exponential form, which was first suggested by Rose et al. [85,
of the loading curve leads to the increasing separation of
86] for the normal traction versus normal separation relation
cohesive surfaces (Figure 7(a)). Upon unloading, the cohe-
based on fit atomistic calculations and then was widely used
sive zone cannot close completely due to plastic deformation
as a traction-separation law for metal single crystals [37, 41].
of the surrounding materials (Figure 7(b)). Then the damage
The crack growth was exclusively triggered by the irreversible
locus is reached at the forward part of subsequent loading and
motion of dislocation within the continuum elements, not
further damage accumulates (Figure 7(c)). After sufficient
within the cohesive elements.
loading cycles, material in the cohesive zone will degenerate
It is noted that the unloading and the subsequent reload-
completely and form new fracture surfaces, predicating the
ing mentioned above are assumed to follow the same path.
propagation of the fatigue crack. The crack fronts of alu-
The fracture process zone (the cohesive zone) eventually
minum shafts subjected to axial loading have been predicted
stabilizes with no further damage, which causes shakedown
using this partial unloading-reloading configuration. The
effects and crack arrest under cyclic loading [23]. Hence this
computed results reproduce the experimentally observed
linear configuration of unloading and reloading behaviors is
progression of beachmark crack fronts as shown in Figure 8.
incapable of modeling the subcritical crack growth. Damage
The calculation error between the computed cycles and
evolution during fatigue loading should be introduced which
experimental cycles ranges from 14% to 18%.
will be described in Section 3.2.

3.1.2. Mixed Mode. For mixed mode, an important problem 3.2.2. Hysteretic Unloading-Reloading. CZM, which incor-
concerns the definition of unloading. Tvergaard [36] defined porates the differences between unloading and reloading
it based on a nondimensional parameter 𝜆, which is formu- path, was proposed by Yang et al. [87] and Nguyen et al.
lated as [88], respectively. In the model, the damage accumulation is
accounted for on a cycle-by-cycle basis. The irreversible dam-
2 age occurs not only along the damage locus but also along
𝛿𝑛 𝛿𝑡 2
𝜆 = √( ) + ( ). (1) an unloading-reloading path underneath it. Linear unloading
𝛿𝑛𝑐 𝛿𝑡𝑐 and nonlinear reloading make it possible to take dissipa-
tive mechanisms into account, such as crystallographic slip
In this case, unloading occurs while 𝜆 decreases, even
and frictional interactions between asperities. In addition,
if separation in one direction increases. Another similar
shakedown and crack arrest can be prevented, thus allowing
definition is also used in which whether unloading or not
for steady crack growth. Figure 9 illustrates the behavior
depends on the change of the total separation 𝛿 = √𝛿𝑛2 + 𝛿𝑡2 . of the model. Material degradation accumulates below the

Downloaded from ade.sagepub.com by guest on September 2, 2015

Advances in Mechanical Engineering
Tn
A
𝛿A 𝛿n
Figure 6: Cohesive law with unloading-reloading behavior for
ductile material [24].
damage locus prior to failure. The applied global loading
range controls the local upper and lower loading levels.
Yang et al. [87] analyzed a single-edge-notched rectangu-
lar specimen in pure mode I and in mixed mode, demonstrat-
ing the capability of the present model to predict fatigue crack
initiation and growth in quasibrittle materials. Since some
ductility due to bridging and roughness effects of particles
or aggregates at the fracture surfaces can be observed for
quasibrittle materials, a so-called line spring model was used
as CZM in which the tail of descending traction-separation
curve is very long after an initially steep softening response.
Nguyen et al. [88] modeled M(T) specimen under cyclic
mode I loading with up to 6000 load cycles and compared the
results with experiments in the Paris regime. The calculations
demonstrated that the theory was capable of treating long
cracks under constant-amplitude loading, short cracks, and
overloads. Serebrinsky and Ortiz [26] also assessed the ability
of this hysteretic CZM to predict the number of cycles
for fatigue failure. CZMs with different threshold 𝜎th were
used for the analysis of two different ductile materials.
Comparisons between computed results and experimental
data as shown in Figures 10 and 11 indicate the approach
capture salient aspects of the observed behavior, such as the
threshold stresses for fatigue failure, the overall shape of the
𝑆-𝑁 curve, and the effect of stress ratio. The error in Figure 10
is less than 15% while relatively significant deviation can be
found in Figure 11. A similar CZM, which also incorporates
the differences between unloading and reloading paths was
used by Maiti and Geubelle [89] to simulate fatigue crack
growth of polymers and then to investigate the effect of
fatigue crack retardation induced by crack closure [90].
3.2.3. Linear Varying Slope Unloading-Reloading. Since dam-
age evolution is a nonlinear process for inelastic deformation,
CZM can be established in analogy to the principles of
plasticity but allowing for strain softening. The well-known
characteristics of typical elastic-plastic damage evolution laws
Downloaded from ade.sagepub.com by guest on September 2, 2015
7
include the following [91]: (i) damage begins to accumulate
once a deformation measure, accumulated or current, is
greater than a critical magnitude; (ii) the increment of dam-
age is related to the increment of deformation as weighted
by the current load level; (iii) there exists an endurance limit
which is a stress level below which cyclic loading can proceed
infinitely without failure. Based on this consideration, Roe
and Siegmund [33] proposed the evolution equation for dam-
age of the cohesive zone under cyclic loading. Its increment
form is written as
|Δ𝛿| 𝑇 𝜎𝑓
Δ𝐷𝑐 = [ − ] 𝐻 (𝛿 − 𝛿0 ) (3)
𝛿Σ 𝜎max 𝜎max,0
with
Δ𝐷𝑐 ≥ 0, (4)
where Δ𝛿 and 𝑇 are the effective cohesive zone quantities,
with 𝐻 designating the Heaviside function. In the expression,
two additional parameters are introduced, that is, the cohe-
sive zone endurance limit 𝜎𝑓 and the accumulated cohesive
length 𝛿Σ , which determines the amount of accumulated
effective separation necessary to fail the cohesive zone.
𝛿Σ is a multiple of the cohesive length 𝛿0 , which is the
material separation across the crack surfaces in the cohesive
zone corresponding to the cohesive strength under normal
loading. The magnitude of the incremental damage is then
dependent on the two additional material parameters and
the proportional to the scaled and normalized incremental
resultant separation, |Δ𝛿|/𝛿Σ , weighted by a measure of
current traction reduced by the endurance limit.
Then the damage is translated as the degradation of the
cohesive properties in the CZM constitutive relation by
𝜎max = 𝜎max,0 (1 − 𝐷) ,
(5)
𝜏max = 𝜏max,0 (1 − 𝐷) ,
where 𝜎max,0 and 𝜏max,0 are the current cohesive normal and
tangential strengths used to substitute the initial ones. The
unloading and reloading path in the investigation follow a
linear relationship with a slop equal to that of the current
traction-separation curve at zero separation. In the current
model, the accumulated damage has been accounted for
explicitly and incrementally. The model is capable of pre-
dicting complicated loading conditions. The unloading and
reloading path (assuming 𝐷 = 0) are depicted in Figure 12
for normal and shear separation cases and Figure 13 depicts
a typical model response under load-controlled conditions.
Using a double-cantilever beam specimen, several key fea-
tures found in experiments of adhesives are successfully
reproduced. These include negligible differences in crack ini-
tiation times for different mode cracks as shown in Figure 14.
In addition, as displayed in Figure 15, the predicted fatigue
crack growth rate is in a power law dependence on the applied
energy release rate range.
The damage accumulation model as expressed in (3)
has been applied to the researches of transient fatigue crack
growth under shield of crack bridging [92], the influence of
8 Advances in Mechanical Engineering

F 2
t
2

1
1

T 𝛿

(a)

F 2
t 2

3
1
1 3

T 𝛿

(b)
2 4
F t
2

3
1
1 3

T 𝛿

(c)

Figure 7: Schematic representation of the process of damage accumulation during cyclic loading: (a) opening of cohesive surface, (b) partial
unloading, and (c) reloading and subsequent damage [25].

220,000 255,000

270,000 317,500

330,000
285,000

297,000 345,000

310,000 355,000
(a) (b)

Figure 8: Progression of beachmark crack fronts with the number of cycles: (a) experiment; (b) calculation [25].

Downloaded from ade.sagepub.com by guest on September 2, 2015

Advances in Mechanical Engineering 9

T 1
Damage locus 1
T0 2 3 4 5 R = 0.5
0.8
R = 0.1

0.6

∑max = 𝜎max /𝜎c

1

2 0.4
3
4
5

0.2

𝛿c 𝛿
102 103 104 105 106 107 108
Figure 9: Cohesive law with accumulating damage under cyclic N
loading: the unloading-reloading hysteresis loops accounting for Solid line: calc, 𝜎th = 0.4𝜎c , Nc = 2000
material degradation are enclosed by the limiting curve for mono- Symbols: exp, Al. alloy 2048-T851
tonic loading.
Figure 11: Comparison of normalized maximum stress Σmax versus
cycles between experiments (symbols: aluminum alloy 2048-T851,
1 uniaxial tensile strength 𝜎𝑐 ; open symbols: long transverse, 𝜎𝑐 =
+ + 465 MPa; closed symbols: longitudinal, 𝜎𝑐 = 457 MPa) and calcu-
+ R = 0.05
0.8 +++ +++++ lations (solid lines: CZM with a threshold 𝜎th = 0.4𝜎𝑐 , characteristic
+ R = 0.2 parameter 𝑁𝑐 = 2000) [26].
+ +++++ +
+++ ++ +
0.6 + + ++
+ +
∑max = 𝜎max /𝜎c

shot peening on the fatigue crack growth and relaxation of

0.4
0.2
102 103 104
Solid lines: calc, 𝜎th = 0.5𝜎c , Nc = 333
Symbols: exp, steel 300 M
Figure 10: Comparison of normalized maximum stress Σmax versus
cycles between experiments (symbols: steel 300 M, uniaxial tensile
strength 𝜎𝑐 = 1930 MPa) and calculations (solid lines: CZM with a
threshold 𝜎th = 0.5𝜎𝑐 , characteristic parameter 𝑁𝑐 = 333) [26].
constraint effect [93], and the effect of strain gradient plas-
ticity [94]. Fatigue crack growth rate in complex stress states
[95] and at plastically mismatched bi-material interfaces [96]
were also investigated using this model. In addition, Jiang
et al. [97] extended this irreversible cohesive zone model
to three-dimensional conditions and predicted the influence
of overload and loading mode on fatigue crack growth. Xu
and Yuan [98, 99] have combined this irreversible CZM with
XFEM to simulate fatigue crack growth under mixed mode
for brittle material and ductile material, respectively. Liu et
al. [100, 101] utilized the model to investigate the influence of
residual stress under cyclic load.
Although some experimental phenomena have been
reproduced in numerical simulation, the model cannot
obtain the satisfied results in predicting uniaxial fatigue test
as shown in [33]. The well-known features of the Goodman
criterion and Geber criterion can only be reproduced in very
low cycles. For aforementioned reasons, Xu and Yuan [80]
proposed a new evolution equation for the damage variable,
which neglected the damage induced by the shear stress based
105 106 107 108
on the experimental observation that the crack propagation
N
is dominated by mode I mechanism [102]. Following sugges-
tions in [91], the evolution of damage indicator is defined as
𝑚
|Δ𝛿| 𝑇𝑛
Δ𝐷𝑐 = [ + 𝑛] 𝐻 (𝜎eq − 𝑓0 ) . (6)
𝛿Σ 𝜎max,0
Variables in the previous equation agree with those in (3).
The Heaviside function, 𝐻, prescribes nucleation of damage
once the nonlocal equivalent principal stress, 𝜎eq , ahead of
the cohesive zone tip is greater than the material fatigue
limit 𝑓0 . 𝑚 and 𝑛 are additional parameters of the evolution
equation. The material damage is caused in both unloading
and reloading processes, except for the penetration. With the
above damage evolution definition, the fatigue crack growth
of the cracked C(T) specimen under different stress ratios
was analyzed as shown in Figure 16. However, the comparison
between experimental data and computational prediction
reveals significant deviation for high stress ratios. To improve
the prediction of the CZM, Li and Yuan [34] incorporated the
effect of stress ratio 𝑅 on the accumulated cohesive length 𝑑Σ.
The computational predictions have been shown in Figure 17.

Downloaded from ade.sagepub.com by guest on September 2, 2015

10 Advances in Mechanical Engineering

Tn /𝜎max, 0

Tt /𝜏max, 0
𝛿t /𝛿0

𝛿n /𝛿0

(a) (b)

Figure 12: Normal (a) and shear (b) separation behavior including unloading and reloading [33].

0.5 0.5
0.45 0.4
0.4 0.3
0.35 0.2
Tn /𝜎max, 0

0.3 0.1
Tt /𝜏max, 0

0.25 0
0.2 0.1
0.15 0.2
0.1 0.3
0.05 0.4
0 0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.4 0.2 0 0.2 0.4 0.6
𝛿n /𝛿0 𝛿t /𝛿0

(a) (b)

Figure 13: Predicted CZM response under load-controlled conditions: normal (a) and shear traction (b) separation behavior [33].

600 is built in a thermodynamic framework. The incremental

damage evolution is expressed as Δ𝐷𝑓 = Δ𝐷𝑓 (𝐷𝑓 , 𝑌, 𝛿)
500
with the following features: (i) damage only begins if the
400 thermodynamic force 𝑌 is higher than a threshold 𝑌0 ; (ii) the
damage increment is related to the opening increment; (iii)
Nini

300 damage occurs only under loading conditions. It is defined as

200 𝑚 𝑛
Δ𝐷𝑓 = 𝐴(1 − 𝐷𝑓 ) (√𝑌 − √𝑌0 ) |Δ𝑢| , (7)
100

0
where 𝐴, 𝑚, and 𝑛 are parameters controlling the damage
0.1 0.2 0.3 0.4 0.5 rate. Because the material damage is irreversible, the damage
ΔG/𝜙n, 0 increment 𝐷𝑓 should not be less than zero. To take into
account the time effect observed in loadings at low frequency,
Figure 14: The relationship between cycle number to fatigue crack as shown in Figure 18, the damage variable 𝐷𝑓 is replaced by
initiation and applied energy release rate range [33], for mode I (X) the total damage defined by
and mode II (󳵳).
𝐷 = 𝐷𝑓 + 𝐷𝑐 . (8)
Creep damage in the cohesive zone model is chosen as
Comparing results in Figure 16, the predictions obviously
agree well with the experimental results. The maximum error |𝑇| − 𝑇𝑐 𝑟
can be controlled within 5%. Δ𝐷𝑐 = (1 − 𝐷)−𝑝 ⟨ ⟩, (9)
𝐶
Bouvard et al. [41] proposed another irreversible CZM
used not only for fatigue conditions but also for creep-fatigue where 𝐶, 𝑝, and 𝑟 are parameters, 𝑇 is the traction force, and
conditions. Comparing the presentation in [33], the model 𝑇𝑐 is a traction threshold for creep damage.

Downloaded from ade.sagepub.com by guest on September 2, 2015

Advances in Mechanical Engineering 11

1E + 3
Static failure
1E + 2 d∑ = p ln R + q

10−6
1E + 1 Region II

da/dN
d(a/𝛿0 )/dN

1E + 0
Region III

1E − 1
Region I
1E − 2
Threshold 10−7
30 40 50 60 70 80 90
1E − 3
1E − 3 1E − 2 1E − 1 1E + 0 1E + 1 ΔK
ΔG/𝜙n, 0
R = 0.1 (exp.) R= 0.1 (CZM)
R = 0.2 (exp.) R= 0.2 (CZM)
Figure 15: Predicted dependence of the normalized fatigue crack
R = 0.4 (exp.) R= 0.4 (CZM)
growth rate on the normalized applied energy release rate range for
R = 0.7 (exp.) R= 0.7 (CZM)
mode I, mode II, and mixed-mode loading [33].
Figure 17: The numerical results about the fatigue crack growth in
comparison with experiment for the cracked C(T) specimen, 𝑑Σ =
𝑝 ln 𝑅 + 𝑞 [34].
dΣ = 40𝛿n

𝜎
10−6
da/dN

th
𝜎max

10−7 𝜎min
30 40 50 60 70 80 90
ΔK

R = 0.1 (exp.) R= 0.1 (CZM)

T
R = 0.2 (exp.) R= 0.2 (CZM)
R = 0.4 (exp.) R= 0.4 (CZM) Figure 18: Loading cycles with a hold time that may vary from 0
R = 0.7 (exp.) R= 0.7 (CZM) (pure fatigue) to ∞ (creep tests) [41].
Figure 16: The numerical results about the fatigue crack growth in
comparison with experiment for the cracked C(T) specimen, 𝑑Σ =
40𝛿𝑛 [34].
parameters, bringing the linear varying slop unloading-
reloading. With this damage model, Yang et al. [103] suc-
cessfully predicted the low cycle fatigue life of solder joints
Unlike the damage model introduced by Roe and Sieg- plastically deformed under mode II cyclic loading.
mund, Yang et al. [103] regarded the damage as a function of Recently, Gong et al. [104] proposed a CZM coupled
accumulated plastic strain. The accumulated damage follows with damage for interface fatigue problems. The cumulative
a simple form as damage due to fatigue is taken into account in terms of the
degradation of interfacial cohesive properties as Roe and
Siegmund [33] did. In the model, damage is related to the
𝛾
Δ𝐷 = 𝛼(𝜀𝑝 ) , (10) opening displacement in the cohesive zone and its evolution
is formulated as
where 𝜀𝑝 is the plastic shear strain 𝛼 and 𝛾 are parameters 𝛾
𝛿𝑛ave
determined by fitting the experimental data. Then the damage Δ𝐷 = (𝛼 ) , (11)
accumulation is translated as the evolvement of cohesive (1 − 𝐷) 𝛿0

Downloaded from ade.sagepub.com by guest on September 2, 2015

12
+ zone material has to be as high as possible, at least as high
−
(a)
f monotonic loading. The same technique was also employed
f0
𝛿td |𝛿t |
(b)
Figure 19: (a) Illustration of the two originally coincident points of a
cohesive zone in the contact/sliding mode (the tangential interaction
between the two points is modeled using Coulomb frictional force
and the cohesive force if the cohesive line spring is not already
broken); (b) sketch of the variation of the frictional coefficient of
the contact surface [42].
where 𝛿𝑛ave is the average opening displacement across the
interface and 𝛿0 is the initial critical displacement, 𝛼 and
𝛾 are material parameters. During each cyclic load, the
cohesive properties of the cohesive zone gradually reduce as
the opening displacement increases.
Similar to the damage variable 𝐷 used above, Brinck-
mann and Siegmund [105] introduced a so-called dislo-
cation stress enhancement factor 𝐸𝜎 into CZM. Based on
the micromechanics of dislocations, the factor is used to
account for the influence of dislocations at the crack tip on
the material separation process. Then the instant cohesive
strength 𝜎max is calculated by
𝜎max = 𝜎max,0 (1 − 𝐸𝜎 ) ,
where the 𝐸𝜎 is defined as the ratio of the stress from
a given dislocation density distribution to an appropriate
reference stress. In the paper, the crack growth for constant
amplitude loading and overload were computed and the
results compared well with the experimental findings.
3.3. Crack Surface Contact and Friction Behavior. To com-
plete the formulation of CZM, the potential conditions of
contact and friction behavior under cyclic loading should be
taken into account. For negative normal separation which
means the interpenetration of the cohesive zone, it is phys-
ically not admissible. Therefore, the stiffness of the cohesive
Downloaded from ade.sagepub.com by guest on September 2, 2015
Advances in Mechanical Engineering
as the initial elastic stiffness, to prevent the appearance of
penetration. Roe and Siegmund [33] employed a penalization
technique in which a high penalty stiffness was prescribed for
the case of 𝛿𝑛 < 0. The traction-separation relation under
contact conditions followed a modification of that under
in [93, 94, 97].
Based on the line-spring model as shown in Figure 19(a),
Yang et al. [42, 87] managed to describe the frictional
interaction simply through a Coulomb type law. Then the
tangential interaction of the crack surfaces in contact is given
by
𝑇𝑡 = 𝑘𝛿𝑡 − sgn (𝛿𝑡 ) 𝑓𝑇𝑛 , (13)
where 𝑓 is frictional coefficient assumed to be
󵄨󵄨 󵄨󵄨
{ 󵄨𝛿 󵄨 󵄨󵄨 󵄨󵄨
{𝑓0 󵄨 𝑡 󵄨 , 󵄨󵄨𝛿𝑡 󵄨󵄨 ≤ 𝛿𝑡𝑑
𝑓 = { 𝛿𝑡𝑑 (14)
{ 󵄨󵄨 󵄨󵄨
{𝑓0 , 󵄨󵄨𝛿𝑡 󵄨󵄨 > 𝛿𝑡𝑑 ,
for the smooth transition of frictional force near 𝛿𝑡 =
0 (𝑓0 and 𝛿𝑡𝑑 are both nonnegative material constants).
Function 𝑓 is plotted in Figure 19(b). In (13), the normal
component of the traction 𝑇𝑛 is evaluated by taking the
contact condition 𝛿𝑛 = 0 and 𝑘 is the stiffness of the cohesive
zone material which depends on the current damage. The first
term represents the tangential traction contribution due to
the line spring (nontrivially if it is not broken completely) and
the second term represents the Coulomb friction component.
Note that the description of contact and friction above is
a very simple one based on the assumptions of small relative
sliding and simple geometry of the contact crack surfaces.
In fact, the friction behavior has been neglected in a lot of
investigations [33], only the contact behavior is considered.
4. Conclusions
This paper provides an overview of CZM, with an emphasis
on its important aspects for fatigue crack growth. Three
general aspects have been discussed, including unloading-
reloading path, damage evolution during cyclic loading, and
crack surface contact and friction behavior. As a phenomeno-
logical model, lots of traction-separation laws have been
(12) reviewed according to the fracture mode. Methods for the
determination of cohesive parameters were also reviewed for
ductile material and brittle material, respectively. To simulate
fatigue crack growth, CZM can be flexibly implemented
within the framework of FEM, BEM, or XFEM.
With its specific advantages, CZM has successfully pre-
dicted some experimental behaviors during crack initiation
and subsequent propagation process. However, the present
applications of CZM are still far from practical engineering
employment. To apply CZM to practical engineering, several
problems have to be solved, including the unclear physical
meaning behind CZM and uncertain influencing factors on
damage evolution. Thus, there are a lot of research works
needed to be done in the future.
Advances in Mechanical Engineering
Acknowledgment
This study is based upon work supported by the National
Natural Science Foundation of China (Grant no. 51005020).
References
[1] O. E. Wheeler, “Spectrum loading and crack growth,” Jouranl of
Basic Engineering, vol. 94, no. 1, pp. 181–186, 1972.
[2] J. Weertman, “Rate of growth of fatigue cracks calculated from
the theory of infinitesimal dislocations distributed on a plane,”
International Journal of Fracture, vol. 26, no. 4, pp. 308–315,
1984.
[3] A. H. Noroozi, G. Glinka, and S. Lambert, “A two parameter
driving force for fatigue crack growth analysis,” International
Journal of Fatigue, vol. 27, no. 10–12, pp. 1277–1296, 2005.
[4] R. Foreman, V. Keary, and R. Engle, “Numeral analysis of
crack propagation in cyclic-loaded structures,” Journal of Basic
Engineering, vol. 89, no. 3, pp. 459–463, 1967.
[5] W. Elber, “Fatigue crack closure under cyclic tension,” Engineer-
ing Fracture Mechanics, vol. 2, no. 1, pp. 37–44, 1970.
[6] M. H. El Haddad, N. E. Dowling, T. H. Topper, and K. N. Smith,
“J integral applications for short fatigue cracks at notches,”
International Journal of Fracture, vol. 16, no. 1, pp. 15–30, 1980.
[7] E. Budzakoska, D. G. Carr, P. A. Stathers et al., “Predicting
the J integral fracture toughness of Al 6061 using the small
punch test,” Fatigue and Fracture of Engineering Materials and
Structures, vol. 30, no. 9, pp. 796–807, 2007.
[8] I. Haryanto and O. Takahashi, “Ductile fracture assessment of
Indonesian wearing course mixtures using critical J integral
and crack tip opening angle,” International Journal of Pavement
Engineering, vol. 9, no. 3, pp. 165–176, 2008.
[9] C.-C. Lee, T.-C. Huang, C.-C. Hsia, and K.-N. Chiang, “Inter-
facial fracture investigation of low-k packaging using J-integral
methodology,” IEEE Transactions on Advanced Packaging, vol.
31, no. 1, pp. 91–99, 2008.
[10] N. Dowling and J. Begley, “Fatigue crack growth during gross
plasticity and the J-integral,” American Society for Testing and
Materials STP, vol. 590, pp. 80–103, 1976.
[11] K. Tanaka, “The cyclic J-integral as a criterion for fatigue crack
growth,” International Journal of Fracture, vol. 22, no. 2, pp. 91–
104, 1983.
[12] D. S. Dugdale, “Yielding of steel sheets containing slits,” Journal
of the Mechanics and Physics of Solids, vol. 8, no. 2, pp. 100–104,
1960.
[13] G. I. Barenblatt, “The mathematical theory of equilibrium
cracks in brittle fracture,” Advances in Applied Mechanics, vol.
7, pp. 55–129, 1962.
[14] F. Moroni and A. Pirondi, “A procedure for the simulation of
fatigue crack growth in adhesively bonded joints based on the
cohesive zone model and different mixed-mode propagation
criteria,” Engineering Fracture Mechanics, vol. 78, no. 8, pp.
1808–1816, 2011.
[15] A. Pirondi and F. Moroni, “Simulation of mixed-mode I/II
fatigue crack propagation in adhesive joints with a modified
cohesive zone model,” Journal of Adhesion Science and Technol-
ogy, vol. 25, no. 18, pp. 2483–2499, 2011.
[16] P. Beaurepaire and G. I. Schuëller, “Modeling of the variability of
fatigue crack growth using cohesive zone elements,” Engineering
Fracture Mechanics, vol. 78, no. 12, pp. 2399–2413, 2011.
Downloaded from ade.sagepub.com by guest on September 2, 2015
13
[17] A. Ural, V. R. Krishnan, and K. D. Papoulia, “A cohesive zone
model for fatigue crack growth allowing for crack retardation,”
International Journal of Solids and Structures, vol. 46, no. 11-12,
pp. 2453–2462, 2009.
[18] P. W. Harper and S. R. Hallett, “A fatigue degradation law for
cohesive interface elements—development and application to
composite materials,” International Journal of Fatigue, vol. 32,
no. 11, pp. 1774–1787, 2010.
[19] T. Yamaguchi, T. Okabe, and S. Yashiro, “Fatigue simulation
for titanium/CFRP hybrid laminates using cohesive elements,”
Composites Science and Technology, vol. 69, no. 11-12, pp. 1968–
1973, 2009.
[20] S.-Y. Yang, “Analysis of fatigue crack growth in weld specimen
using three-dimensional cohesive zone model,” Key Engineering
Materials, vol. 321, pp. 674–677, 2006.
[21] C. Lequesne, A. Plumier, H. Degee, and A. M. Habraken,
“Numerical study of the fatigue crack in welded beam-to-
column connection using cohesive zone model,” Key Engineer-
ing Materials, vol. 324, pp. 847–850, 2006.
[22] W. Zhang and X. Deng, “Formulation of a cohesive zone model
for a Mode III crack,” Engineering Fracture Mechanics, vol. 72,
no. 12, pp. 1818–1829, 2005.
[23] B. Yang and K. Ravi-Chandar, “On the role of the process zone in
dynamic fracture,” Journal of the Mechanics and Physics of Solids,
vol. 44, no. 12, pp. 1955–1976, 1996.
[24] I. Scheider, Bruchmechanische Bewertung von Laserschweißver-
bindungen durch numerische Rissfortschrittsanalysen mit dem
Kohasivzonenmodell [Ph.D. thesis], Technical University Ham-
burg, Harburg, Germany, 2000.
[25] A. De-Andrés, J. L. Pérez, and M. Ortiz, “Elastoplastic finite
element analysis of three-dimensional fatigue crack growth
in aluminum shafts subjected to axial loading,” International
Journal of Solids and Structures, vol. 36, no. 15, pp. 2231–2258,
1999.
[26] S. Serebrinsky and M. Ortiz, “A hysteretic cohesive-law model
of fatigue-crack nucleation,” Scripta Materialia, vol. 53, no. 10,
pp. 1193–1196, 2005.
[27] A. Needleman, “A continuum model for void nucleation by
inclusion debonding,” Journal of Applied Mechanics, vol. 54, no.
3, pp. 525–531, 1987.
[28] A. Hillerborg, M. Modéer, and P.-E. Petersson, “Analysis of
crack formation and crack growth in concrete by means of
fracture mechanics and finite elements,” Cement and Concrete
Research, vol. 6, no. 6, pp. 773–781, 1976.
[29] A. Needleman, “An analysis of decohesion along an imperfect
interface,” International Journal of Fracture, vol. 42, no. 1, pp.
21–40, 1990.
[30] P. H. Geubelle and J. S. Baylor, “Impact-induced delamination
of composites: a 2D simulation,” Composites B, vol. 29, no. 5, pp.
589–602, 1998.
[31] H. Yuan, G. Lin, and A. Cornec, “Verification of a cohesive zone
model for ductile fracture,” Journal of Engineering Materials and
Technology, vol. 118, no. 2, pp. 192–200, 1996.
[32] V. Tvergaard and J. W. Hutchinson, “The relation between crack
growth resistance and fracture process parameters in elastic-
plastic solids,” Journal of the Mechanics and Physics of Solids, vol.
40, no. 6, pp. 1377–1397, 1992.
[33] K. L. Roe and T. Siegmund, “An irreversible cohesive zone
model for interface fatigue crack growth simulation,” Engineer-
ing Fracture Mechanics, vol. 70, no. 2, pp. 209–232, 2003.
14
[34] H. Li and H. Yuan, “Cohesive zone modelling of low cycle
fatigue cracks in cracked and notched specimens,” Fatigue &
Fracture of Engineering Material & Structures, 2013.
[35] A. Needleman, “An analysis of tensile decohesion along an
interface,” Journal of the Mechanics and Physics of Solids, vol. 38,
no. 3, pp. 289–324, 1990.
[36] V. Tvergaard, “Effect of fibre debonding in a whisker-reinforced
metal,” Materials Science and Engineering A, vol. 125, no. 2, pp.
203–213, 1990.
[37] X.-P. Xu and A. Needleman, “Void nucleation by inclusion
debonding in a crystal matrix,” Modelling and Simulation in
Materials Science and Engineering, vol. 1, no. 2, pp. 111–132, 1993.
[38] G. T. Camacho and M. Ortiz, “Computational modelling of
impact damage in brittle materials,” International Journal of
Solids and Structures, vol. 33, no. 20–22, pp. 2899–2938, 1996.
[39] C.-C. Chen and D. G. Linzell, “Modeling end notched flexure
tests to establish cohesive element Mode II fracture parameters,”
Engineering Fracture Mechanics, vol. 77, no. 8, pp. 1338–1347,
2010.
[40] N. Dourado, M. F. S. F. de Moura, A. B. de Morais, and A. B.
Pereira, “Bilinear approximations to the mode II delamination
cohesive law using an inverse method,” Mechanics of Materials,
vol. 49, pp. 42–50, 2012.
[41] J. L. Bouvard, J. L. Chaboche, F. Feyel, and F. Gallerneau, “A
cohesive zone model for fatigue and creep-fatigue crack growth
in single crystal superalloys,” International Journal of Fatigue,
vol. 31, no. 5, pp. 868–879, 2009.
[42] B. Yang and K. Ravi-Chandar, “A single-domain dual-
boundary-element formulation incorporating a cohesive zone
model for elastostatic cracks,” International Journal of Fracture,
vol. 93, no. 1–4, pp. 115–144, 1998.
[43] V. Tvergaard and J. W. Hutchinson, “The influence of plasticity
on mixed mode interface toughness,” Journal of the Mechanics
and Physics of Solids, vol. 41, no. 6, pp. 1119–1135, 1993.
[44] F. Yuval and B. S. Leslie, “A new cohesive zone model for mixed
mode interface fracture in bimaterials,” Engineering Fracture
Mechanics, vol. 75, no. 15, pp. 4583–4593, 2008.
[45] L. N. Lens, E. Bittencourt, and V. M. R. d’Avila, “Constitutive
models for cohesive zones in mixed-mode fracture of plain
concrete,” Engineering Fracture Mechanics, vol. 76, no. 14, pp.
2281–2297, 2009.
[46] M. J. Lee, T. M. Cho, W. S. Kim, B. C. Lee, and J. J. Lee,
“Determination of cohesive parameters for a mixed-mode
cohesive zone model,” International Journal of Adhesion and
Adhesives, vol. 30, no. 5, pp. 322–328, 2010.
[47] G. Lin, A. Cornec, and K.-H. Schwalbe, “Three-dimensional
finite element simulation of crack extension in aluminium alloy
2024FC,” Fatigue and Fracture of Engineering Materials and
Structures, vol. 21, no. 10, pp. 1159–1173, 1998.
[48] A. Cornec, I. Scheider, and K.-H. Schwalbe, “On the practi-
cal application of the cohesive model,” Engineering Fracture
Mechanics, vol. 70, no. 14, pp. 1963–1987, 2003.
[49] Z. Li, S. M. Kulkarni, and S. P. Shah, “New test method for
obtaining softening response of unnotched concrete specimen
under uniaxial tension,” Experimental Mechanics, vol. 33, no. 3,
pp. 181–188, 1993.
[50] H. W. Reinhardt, H. A. W. Cornelissen, and D. A. Hordijk,
“Tensile tests and failure analysis of concrete,” Journal of
Structural Engineering, vol. 112, no. 11, pp. 2462–2477, 1986.
Downloaded from ade.sagepub.com by guest on September 2, 2015
Advances in Mechanical Engineering
[51] S. Cattaneo and G. Rosati, “Effect of different boundary condi-
tions in direct tensile tests: experimental results,” Magazine of
Concrete Research, vol. 51, no. 5, pp. 365–374, 1999.
[52] J. G. M. van Mier and M. R. A. van Vliet, “Uniaxial tension test
for the determination of fracture parameters of concrete: state
of the art,” Engineering Fracture Mechanics, vol. 69, no. 2, pp.
235–247, 2002.
[53] C. Rocco, G. V. Guinea, J. Planas, and M. Elices, “Review of
the splitting-test standards from a fracture mechanics point of
view,” Cement and Concrete Research, vol. 31, no. 1, pp. 73–82,
2001.
[54] H. N. Linsbauer and E. K. Tschegg, “Fracture energy determi-
nation of concrete with cube-shaped specimens,” Zement und
Beton, vol. 31, pp. 38–40, 1986.
[55] P. E. Roelfstra and F. H. Wittmann, “Numerical method to link
strain softening with failure of concrete,” in Fracture Toughness
and Fracture Energy of Concrete, pp. 163–175, Elsevier Science,
Amsterdam, The Netherlands, 1986.
[56] H. Mihashi and N. Nomura, “Correlation between characteris-
tics of fracture process zone and tension-softening properties
of concrete,” Nuclear Engineering and Design, vol. 165, no. 3, pp.
359–376, 1996.
[57] G. Bolzon and G. Maier, “Identification of cohesive crack
models for concrete on the basis of three-point bending tests,”
Computational Modelling of Concrete Structures, vol. 1, pp. 301–
310, 1998.
[58] G. Bolzon, R. Fedele, and G. Maier, “Parameter identification
of a cohesive crack model by Kalman filter,” Computer Methods
in Applied Mechanics and Engineering, vol. 191, no. 25-26, pp.
2847–2871, 2002.
[59] X. Hu and Y.-W. Mai, “Crack-bridging analysis for alumina
ceramics under monotonic and cyclic loading,” Journal of the
American Ceramic Society, vol. 75, no. 4, pp. 848–853, 1992.
[60] G. V. Guinea, J. Planas, and M. Elices, “A general bilinear fit for
the softening curve of concrete,” Materials and Structures, vol.
27, no. 2, pp. 99–105, 1994.
[61] V. C. Li, C.-M. Chan, and C. K. Y. Leung, “Experimental deter-
mination of the tension-softening relations for cementitious
composites,” Cement and Concrete Research, vol. 17, no. 3, pp.
441–452, 1987.
[62] X. Z. Hu and F. H. Wittmann, “Experimental method to deter-
mine extension of fracture-process zone,” Journal of Materials
in Civil Engineering, vol. 2, pp. 15–23, 1990.
[63] X. HU and F. H. Wittmann, “An analytical method to determine
the bridging stress transferred within the fracture process zone.
I: general theory,” Cement and Concrete Research, vol. 21, no. 6,
pp. 1118–1128, 1991.
[64] X. Z. Hu and F. H. Wittmann, “An analytical method to
determine the bridging stress transferred within the fracture
process zone. II: application to mortar,” Cement and Concrete
Research, vol. 22, no. 4, pp. 559–570, 1992.
[65] p. Nanakorn and H. Horii, “Back analysis of tension-softening
relationship of concrete,” Journal of Materials, Concrete Struc-
tures and Pavements, vol. 32, pp. 265–275, 1996.
[66] X.-P. Xu and A. Needleman, “Numerical simulations of dynamic
interfacial crack growth allowing for crack growth away from
the bond line,” International Journal of Fracture, vol. 74, no. 3,
pp. 253–275, 1995.
[67] X.-P. Xu and A. Needleman, “Numerical simulations of dynamic
crack growth along an interface,” International Journal of Frac-
ture, vol. 74, no. 4, pp. 289–324, 1996.
Advances in Mechanical Engineering
[68] M. Ortiz and S. Suresh, “Statistical properties of residual stresses
and intergranular fracture in ceramic materials,” Journal of
Applied Mechanics, vol. 60, no. 1, pp. 77–84, 1993.
[69] V. Tvergaad and J. W. Hutchinson, “Effect of strain dependent
cohesive zone model on prediction of interface crack,” Journal
de Physique IV, vol. 6, pp. 165–172, 1996.
[70] A. Needleman, “Micromechanical modelling of interfacial
decohesion,” Ultramicroscopy, vol. 40, no. 3, pp. 203–214, 1992.
[71] V. Tvergaard and J. W. Hutchinson, “Effect of strain-dependent
cohesive zone model on predictions of crack growth resistance,”
International Journal of Solids and Structures, vol. 33, no. 20–22,
pp. 3297–3308, 1996.
[72] M. H. Aliabadi, “Boundary element formulations in fracture
mechanics,” Applied Mechanics Reviews, vol. 50, no. 2, pp. 83–
96, 1997.
[73] G. E. Blandford, A. R. Ingraffea, and J. A. Liggett, “Two-
dimensional stress intensity factor computations using the
boundary element method,” International Journal for Numerical
Methods in Engineering, vol. 17, no. 3, pp. 387–404, 1981.
[74] A. Portela, M. H. Aliabadi, and D. P. Rooke, “Dual boundary
element method. Effective implementation for crack problems,”
International Journal for Numerical Methods in Engineering, vol.
33, no. 6, pp. 1269–1287, 1992.
[75] A. Young, “A single-domain boundary element method for
3-D elastostatic crack analysis using continuous elements,”
International Journal for Numerical Methods in Engineering, vol.
39, no. 8, pp. 1265–1293, 1996.
[76] W. H. Chen and T. C. Chen, “An efficient dual boundary
element technique for a two-dimensional fracture problem with
multiple cracks,” International Journal of Numerical Methods in
Engineering, vol. 38, no. 10, pp. 1739–1756, 1995.
[77] T. Belytschko and T. Black, “Elastic crack growth in finite
elements with minimal remeshing,” International Journal for
Numerical Methods in Engineering, vol. 45, no. 5, pp. 601–620,
1999.
[78] R. de Borst, “Numerical aspects of cohesive-zone models,”
Engineering Fracture Mechanics, vol. 70, no. 14, pp. 1743–1757,
2003.
[79] G. N. Wells and L. J. Sluys, “A new method for modelling
cohesive cracks using finite elements,” International Journal for
Numerical Methods in Engineering, vol. 50, no. 12, pp. 2667–
2682, 2001.
[80] Y. J. Xu and H. Yuan, “On damage accumulations in the cyclic
cohesive zone model for XFEM analysis of mixed-mode fatigue
crack growth,” Computational Materials Science, vol. 46, no. 3,
pp. 579–585, 2009.
[81] Y. N. Li, A. P. Hong, and W. K. Binienda, “Theory of cohesive
crack model with interactive cracks,” International Journal of
Solids and Structures, vol. 35, no. 11, pp. 981–994, 1998.
[82] K. L. Roe, A cohesive zone model for fatigue crack growth
simulation [M.S. thesis], Purdue University, West Lafayette, Ind,
USA, 2001.
[100] J. Liu, C. Xiang, and H. Yuan, “Prediction of 3D small fatigue
[83] V. S. Deshpande, A. Needleman, and E. van der Giessen, “A
discrete dislocation analysis of near-threshold fatigue crack
growth,” Acta Materialia, vol. 49, no. 16, pp. 3189–3203, 2001.
[84] V. S. Deshpande, A. Needleman, and E. van der Giessen,
“Discrete dislocation modeling of fatigue crack propagation,”
Acta Materialia, vol. 50, no. 4, pp. 831–846, 2002.
Downloaded from ade.sagepub.com by guest on September 2, 2015
15
[85] J. H. Rose, J. Ferrante, and J. R. Smith, “Universal binding energy
curves for metals and bimetallic interfaces,” Physical Review
Letters, vol. 47, no. 9, pp. 675–678, 1981.
[86] J. Ferrante, J. R. Smith, and J. H. Rose, “Diatomic molecules
and metallic adhesion, cohesion, and chemisorption: a single
binding-energy relation,” Physical Review Letters, vol. 50, no. 18,
pp. 1835–1845, 1983.
[87] B. Yang, S. Mall, and K. Ravi-Chandar, “A cohesive zone model
for fatigue crack growth in quasibrittle materials,” International
Journal of Solids and Structures, vol. 38, no. 22-23, pp. 3927–
3944, 2001.
[88] O. Nguyen, E. A. Repetto, M. Ortiz, and R. A. Radovitzky, “A
cohesive model of fatigue crack growth,” International Journal
of Fracture, vol. 110, no. 4, pp. 351–369, 2001.
[89] S. Maiti and P. H. Geubelle, “A cohesive model for fatigue failure
of polymers,” Engineering Fracture Mechanics, vol. 72, no. 5, pp.
691–708, 2005.
[90] S. Maiti and P. H. Geubelle, “Cohesive modeling of fatigue
crack retardation in polymers: crack closure effect,” Engineering
Fracture Mechanics, vol. 73, no. 1, pp. 22–41, 2006.
[91] J. Lemaitre, Course on Damage Mechanics, Springer, Berlin,
Germany, 1996.
[92] T. Siegmund, “A numerical study of transient fatigue crack
growth by use of an irreversible cohesive zone model,” Interna-
tional Journal of Fatigue, vol. 26, no. 9, pp. 929–939, 2004.
[93] B. Wang and T. Siegmund, “A numerical analysis of constraint
effects in fatigue crack growth by use of an irreversible cohesive
zone model,” International Journal of Fracture, vol. 132, no. 2, pp.
175–196, 2005.
[94] S. Brinckmann and T. Siegmund, “Computations of fatigue
crack growth with strain gradient plasticity and an irreversible
cohesive zone model,” Engineering Fracture Mechanics, vol. 75,
no. 8, pp. 2276–2294, 2008.
[95] D. Jha and A. Banerjee, “A cohesive model for fatigue failure in
complex stress-states,” International Journal of Fatigue, vol. 36,
no. 1, pp. 155–162, 2012.
[96] B. Wang and T. Siegmund, “Simulation of fatigue crack growth
at plastically mismatched bi-material interfaces,” International
Journal of Plasticity, vol. 22, no. 9, pp. 1586–1609, 2006.
[97] H. Jiang, X. Gao, and T. S. Srivatsan, “Predicting the influ-
ence of overload and loading mode on fatigue crack growth:
a numerical approach using irreversible cohesive elements,”
Finite Elements in Analysis and Design, vol. 45, no. 10, pp. 675–
685, 2009.
[98] Y. Xu and H. Yuan, “Computational analysis of mixed-mode
fatigue crack growth in quasi-brittle materials using extended
finite element methods,” Engineering Fracture Mechanics, vol.
76, no. 2, pp. 165–181, 2009.
[99] Y. Xu and H. Yuan, “Computational modeling of mixed-mode
fatigue crack growth using extended finite element methods,”
International Journal of Fracture, vol. 159, no. 2, pp. 151–165,
2009.
crack propagation in shot-peened specimens,” Computational
Materials Science, vol. 46, no. 3, pp. 566–571, 2009.
[101] J. Liu, H. Yuan, and R. Liao, “Prediction of fatigue crack
growth and residual stress relaxations in shot-peened material,”
Materials Science and Engineering A, vol. 527, no. 21-22, pp.
5962–5968, 2010.
 16 Advances in Mechanical Engineering

[102] M. Sander and H. A. Richard, “Experimental and numerical

investigations on the influence of the loading direction on the
fatigue crack growth,” International Journal of Fatigue, vol. 28,
no. 5-6, pp. 583–591, 2006.
[103] Q. D. Yang, D. J. Shim, and S. M. Spearing, “A cohesive zone
model for low cycle fatigue life prediction of solder joints,”
Microelectronic Engineering, vol. 75, no. 1, pp. 85–95, 2004.
[104] B. Gong, M. Paggi, and A. Carpinteri, “A cohesive crack
model coupled with damage for interface fatigue problems,”
International Journal of Fracture, vol. 173, no. 2, pp. 91–104, 2012.
[105] S. Brinckmann and T. Siegmund, “A cohesive zone model
based on the micromechanics of dislocations,” Modelling and
Simulation in Materials Science and Engineering, vol. 16, no. 6,
Article ID 065003, 2008.

Downloaded from ade.sagepub.com by guest on September 2, 2015

View publication stats