1.

1 Linear Elastic Fracture Mechanics (LEFM)

Fracture mechanics is concerned with understanding the behavior of

materials and structures that contain cracks or defects. Linear Elastic Fracture
Mechanics (LEFM) is a branch of fracture mechanics that assumes materials to be
linearly elastic and deformable under loading conditions. This assumption allows
for the development of simple mathematical relationships to analyze the initiation
and propagation of cracks in materials. LEFM is widely used in engineering and
materials science to assess the safety and integrity of structures subjected to
mechanical loading. LEFM was first introduced by Griffith [38], he found that the
product of the square root of the flaw length 𝑎 and the applied stress 𝜎 is constant
in brittle pre-cracked plates but this approach showed some problems as the stress
at the crack tip was infinite. Strain energy release G was defined. Later, this theory
was explained taking energy considerations through thermodynamic approach.
Irwin extended Griffith's theory by including plastic behavior to address fracture in
metals. He defined the stress intensity factor (SIF) to describe the state of stress in
the vicinity of the crack tip [39].

1.1.1 Stress Intensity Factor (SIF):

The SIF was developed by Irwin to describe the stress state at a crack tip, is
related to the rate of crack growth, and is used as a failure criterion. Irwin defined
SIF as a near-crack-tip approximation to Westergaard's solution for the stress field
surrounding a crack [40, 41].

If the applied stress at the crack tip exceeds a critical value known as the fracture
toughness (KIC), crack propagation occurs. The fracture toughness represents the
material's resistance to crack propagation and is a material property that
characterizes its toughness. For a simple case, the failure in an infinite plate with
central crack can be expressed in terms of critical energy release rate in relation
with critical stress (σc) [39].

σc2 ∗𝜋𝑎
Gc = (2.1)
𝐸

Which can be further related in

KIC = √𝐺c ∗ E = 𝜎c √𝜋 ∗ 𝑎 (2.2)

Where: 𝐾𝐼𝐶 denotes the fracture toughness of the specimen.

Griffith's energy release rate (G) and Irwin's SIF (K) is related at the plane stress
K2
condition by the relationship G =
𝐸

K2 𝐸
And for the plane strain condition by G = where E′=
𝐸′ (1−ʋ2 )

1.1.2 Fracture Modes

LEFM distinguishes three primary modes of crack propagation, often

referred to as Mode I, Mode II, and Mode III. Mode I is the most common mode,
also known as tensile or opening mode. In Mode I, the crack surfaces are pulled
apart by the applied tension. Mode II called the in-plane shear mode, this mode
occurs when the applied forces cause sliding of the crack surfaces parallel to the
crack plane (perpendicular to the crack front). In mode III the crack surface slides
apart parallel to the crack front and known as the anti-plane shear mode as shown
in figure 2.1 [42].
 Fig2.1. Modes of fracture a) mode I, b) mode II, and c)mode III [42]

From the fundamental definition, the potential energy P is related to the crack
𝑑𝑃
growth 𝑑𝑎 proposed by Irwin G = - [43]. From the previous equations, for
𝑑𝑎

different crack modes the following equations is valid for plane stress and plane
strain conditions [44].

1 (1+ʋ)
G = (KI2 + KII2) + KIII2 (2.3) for the plane stress.
𝐸 𝐸

(1−ʋ2 ) (1+ʋ)
G= (KI2 + KII2) + KIII2 (2.4) for the plane strain.
𝐸 𝐸

1.2 Fatigue crack propagation regime

Paris and Erdogan studied the crack growth behavior experimentally to find a
relation between the crack growth rate 𝑑𝑎/𝑑𝑁 and the stress intensity factor range
𝛥𝐾 [45]. The fatigue crack growth rate in metals can be described by the following
empirical relationship:

da
= C(∆K)m (2.5)
dN
Where: da/dN is the crack growth rate, ∆K is the stress intensity range,

and C and m are material constants.

When: 𝑑𝑎/𝑑𝑁 is plotted against the 𝛥𝐾 in log-log scale it was found the two
extreme asymptotes indicate the start and the end of the crack life. The left value
occurs at 𝛥𝐾 = 𝛥𝐾𝑡ℎ signifies that the values of K below this value cause no crack
growth. The second value occurs at 𝐾𝑚𝑎𝑥 = 𝐾𝑐 signifies for the complete fracture as
shown in figure 2.2.

𝐝𝐚
𝐝𝐍

G
Gthresh Gpl Gc

Fig2.2. Fatigue crack growth governed by Paris law [45].

1.3 The Finite Element Method (FEM)

FEM is a numerical technique provides an approximate solution to complex

problems by dividing the domain into smaller subdomains, known as finite
elements. These elements are geometrically simple shapes that collectively cover
the entire domain of interest. Each element with a set of nodes (points) at which
the unknown variable will be approximated. This process is known as
discretization.

The next step is to define shape functions, also known as interpolation

functions, which represent how the unknown u(x) varies within each element.
These functions are defined such that they satisfy the following conditions:

1) They take the value of 1 at the nodes corresponding to the element and 0 at all
other nodes within the element.

2) They satisfy the continuity requirements between neighboring elements.

Once the shape functions are defined for each element, the global system of
equations is assembled. This involves combining the element equations into a
global matrix equation that represents the entire domain. The assembly process
involves mapping the element-level degrees of freedom to global degrees of
freedom.
With the assembled and modified global system of equations, we can now solve
for the unknown nodal values using numerical methods like direct solvers or
iterative techniques. After obtaining the nodal values, we can interpolate the
solution within each element using the previously defined shape functions.

Post-processing involves analyzing and visualizing the results of interest, such as

stress distributions, temperature profiles, or fluid velocity fields, depending on the
nature of the problem.

1.4 The extended finite element method (XFEM)

While FEM is very effective for many problems, it faces challenges when
dealing with discontinuities in the solution, such as cracks as mentioned above.
Multiple strategies have been put forward for dealing crack-related problems.
These include methods like the quarter-point finite element method, the enriched
finite element method, the boundary collocation method, the integral equation
method, the body force method, the boundary elements method, the dislocation
method, as well as mesh-free techniques like the element-free Galerkin method. In
order to eliminate the need for re-meshing during crack modeling, various methods
have been suggested, including the incorporation of a discontinuous mode at the
element level, the utilization of a moving mesh technique, and an enrichment
approach based on the partition-of-unity X-FEM.

Using automatic re-meshing as a potential solution may face challenges in

achieving convergence and have high cost of computational expenses The XFEM
approach, on the other hand, uses the partition of unity, where crack propagation is
governed by the enrichment of finite element nodes includes the asymptotic near-
crack tip field and a Heaviside function H(x). The first one represents to the
singularity located at the tip of the crack, while the second one represents the
separation between the crack surfaces. In the case of branched cracks (multiple), a
new discontinuous function J(x) is employed to account for branching. This
methodology has been applied to model complex geometries, including cases
involving multiple branched cracks, voids, and cracks originating from holes.
These functions allow for the enrichment of the finite element space as the fracture
process evolves. The formulation utilized in ABAQUS for the approximation for a
displacement vector function 𝑢(x) is detailed in equation3.1.
𝑁 4

𝑢(𝑥) = ∑ 𝑁𝑖 (𝑥) ( 𝑢𝑖 + 𝐻(𝑥) 𝑎𝑖 + ∑ (𝐹𝛼 (𝑥)𝑏𝑖 𝛼 )) (3.1)

𝑖=1 𝛼=1

Where: the first term is relevant to all nodes within the model, the second term
applies to nodes enriched by the discontinuous jump function (crack interior), and
the third term associates to crack-tip.

𝑁𝑖 (𝑥) is the shape function associated with the nodal points.

𝑢𝑖 is the nodal displacement vector related to the finite element solution.

H(x) is the discontinuous jump function. The Heaviside jump function exhibits a
discontinuous behavior along the crack surface, maintaining a constant value on
each side of the crack, with a value of +1 on one side and -1 on the other.

𝑎𝑖 is the nodal enriched degree of freedom vector.

𝐹𝛼 (𝑥) is the asymptotic crack-tip functions that assigned to nodes of the element
where the crack tip locates.

𝑏𝑖 𝛼 is the nodal degree of freedom vector (crack-tip).

𝜃 𝜃 𝜃 𝜃
𝐹𝛼 (𝑥) = (√𝑟 sin , √𝑟cos , √𝑟 sin θ sin , √𝑟 sin θ cos )
2 2 2 2
Where 𝑟and 𝜃are the polar coordinates found at the crack tip at 𝜃= 0 tangent to the
crack face.

Fig3.1. Schematic of the mesh nodes showing the enriched nodes.

ABAQUS field outputs resulting from XFEM analysis include:

STATUSXFEM:

A scalar variable indicates the degree of damage within an element. A fully

cracked element is represented by a value of 1, while an element with no cracks
inside it is assigned a value of 0. Values between 0 and 1 denote partial damage or
cracking.

PSI-LSM and PHI-LSM (Ψ and Φ functions):

The XFEM technology was introduced through the development of the level set
method (LSM), which led to the establishment of nodal signed distance functions
Ψ and Φ. These serve to identify the crack surface and the orthogonal plane, as
shown in Figure 3.2. Φ defines the geometric distance from a node to the crack
surface, while Ψ describes the plane that is perpendicular to the crack surface.
Nodes with Φ (Ψ) values approaching zero correspond to the crack surface (the
plane orthogonal to the crack surface).

Fig3.2. The signed distance functions Φ and Ψ.

Low cycle fatigue (LCF) analysis with XFEM is used to simulate crack
propagation can be conducted in ABAQUS. It's worth noting that, as far as the
authors are aware, only a limited number of applications have employed this
method. This could be attributed to the fact that combining LCF and XFEM isn't
readily available in the user-interface of ABAQUS/CAE interface and requires
manual modeling using key commands.

Many numerical simulations of fatigue crack growth (FCG) in metallic

components have been carried out using finite element (FE) analysis. One of the
most efficient numerical tools for this purpose is the extended finite element
method (XFEM) as evidenced by various studies

Melson conducted FE analysis to simulate the fatigue behavior of center-cracked

plate (CCP) specimens made of aluminum. Various FCG laws were considered,
and multiple modeling techniques, including the virtual crack closure technique
(VCCT) and XFEM, were employed.

In ABAQUS, XFEM currently focuses on modeling the propagation of pre-

existing cracks. Consequently, this approach needs the presence of a pre-existing
crack (or the creation of one as part of the analysis). To implement this, a
specialized XFEM crack must be assigned to the pre-cracked part or element,
along with a surface behavior that allows the crack to grow following the desired
propagation model.

XFEM in ABAQUS is used with the direct cyclic approach in combination with
Paris law to simulate crack propagation. The direct cyclic defines the cyclic
loading condition by defining of a periodic function, which will be used by the
software to apply the amount of load at its corresponding time. The crack growth is
defined using Paris law, which is based on relative fracture energy release rates.
The initiation of the crack propagation condition;
𝑁
f= >1
C1 ∆𝐺 C2

Where: ΔG is the relative fracture energy rate between its maximum and minimum
values.

C1 and C2 are material constants

Three common mixed mode model for evaluating the equivalent fracture energy
release rate GequivC : the BK law, the Power law and the Reeder law.

The power law model is described in Wu and Reuter (1965) by

Gequiv GI a G G
=( ) m + ( II )an + ( III )a0
GequivC GIC GIIC GIIIC

To define this model, provide GIC , GIIC , GIIIC , am, an and ao.

The BK law model is described in Benzeggagh and Kenane (1996) by the

following formula:
G +GIII
GequivC = GIC + (GIIC − GIC ) ( II
G +G
)η
I II + GIII

To define this model, some values must be provided GIC , GIIC and η. This model
provides a power law relationship combining energy release rates in Mode I, Mode
II, and Mode III into a single scalar fracture criterion.

The Reeder law model is described in Reeder et al. (2002) by

G +GIII
GequivC = GIC + (GIIC − GIC ) ( II
G +G + G
)η + (GIIIC −
I II III

GIII G +GIII
GIIC ) (
GII + GIII
) (G +IIG )η
I II + GIII
 1.5 Contour integral method

The J-integral is a very important fracture parameter for linear and nonlinear
material response. It is related to the energy release associated with crack
extension. If the material behavior is linear, the J-integral can be linked to the
stress intensity factor. ABAQUS provides a procedure to get the J-integral based
on studies [147, 148], with the virtual crack extension/domain. The method is very
attractive due to its simplicity, its low cost for the analysis, and high accuracy,
even if a coarse mesh is used. The J integral is defined as the rate of energy release
per unit step for crack propagation. The equation representing the energy release
rate is given as follows:

𝐽 = lim ∮ 𝑛. 𝐻. 𝑞 𝑑г (3.2)
г⇾0

Where:

Г is a contour beginning from the bottom to the top of the crack surface, as shown
in figure 3.3a.

The limit indicates that shrinks onto the crack tip

q is a unit vector in the direction of the virtual crack propagation

n is the normal vector to Г*H .

H is given by H= WI – σ * ∂u/∂x

Where W is the elastic strain energy if the material behavior is linear; and is
defined as the elastic strain energy density plus the plastic dissipation for elastic-
plastic or elastic-viscoplastic materials.

Following Shih et al. (1986), equation 3.2 can be written in the form
 . .
𝜕𝑢
𝐽= ∫ 𝑚. 𝐻. 𝑞′ 𝑑г − ∫ t. . 𝑞′ 𝑑г (3.3)
𝐶+ 𝐶++ г+ 𝐶− 𝐶+ + 𝐶− 𝜕𝑥

Where:

𝑞 ′ is a sufficiently smooth weighting function within the region enclosed by the

closed contour C+ C++ г+ C- and has the value of q on г and zero on C.

m is the outward normal to the domain enclosed by the closed contour, as shown in
Figure 3.3b.

m = -n on г

t=m.σ is the surface traction on the crack surfaces C+ C

Using the divergence theorem, the closed contour integral can be converted into
the domain integral
. .
𝜕 ′
𝜕𝑢
𝐽 = ∫ ( ) . (𝐻. 𝑞 ) 𝑑г − ∫ t. . 𝑞′ 𝑑г
𝐴 𝜕𝑥 𝐶+ + 𝐶− 𝜕𝑥

Where A is the domain enclosed by the closed contour C+ C++ г+ C-

(a) (b)
Fig3.3. (a) Contour for evaluation of the J-integral and (b) Closed contour C+ C++
г+ C- encloses a domain A that includes the crack-tip region as г=0

To evaluate these integrals, ABAQUS defines the domain in terms of rings

of elements surrounding the crack tip. Different “contours” (domains) are created.
The first contour consists of those elements directly connected to crack-tip nodes.
The next contour consists of the ring of elements that share nodes with the
elements in the first contour as well as the elements in the first contour. Each
subsequent contour is defined by adding the next ring of elements that share nodes
with the elements in the previous contour. 𝑞 ′ is chosen to have a magnitude of zero
at the nodes on the outside of the contour and to be one (in the crack direction) at
all nodes inside the contour except for the midside nodes (if they exist) in the outer
ring of elements. These midside nodes are assigned a value between zero and one
according to the position of the node on the side of the element.

Once the J integral is calculated which is the energy release rate so the SIF can be
evaluated by:

K= √𝐽 ∗ 𝐸 for plane stress condition

For the plane strain condition: by K= √𝐽 ∗ 𝐸′

𝐸
Where: E′=
(1−ʋ2 )