Fracture Mechanics

Fracture Mechanics
Complex stresses, Complex Strains, Stress- strain curves of metals,
Introduction to fracture mechanics, strength of structures in the
presence of cracks, Elastic- Plastic Fracture Mechanics ,LEFM
with small crack tip Plasticity, Irwin model (1960) and strip yield
model, Plastic zone size and shape, Crack tip opening
displacement (CTOD) Wells in 1961, J-Integral (Rice 1958),
Mixed mode crack propagation , Energy criterion for crack
growth, Maximum Circumferential Tensile Stress, Maximum
Energy Release Rate, Minimum Strain Energy Density Criteria,
Fatigue Crack Growth Rate, Time dependent crack growth,
Dynamic fracture mechanics(DFM),Dynamic crack propagation.
 Fracture Mechanics

Lecture 1
Complex stresses and Theories of
yielding under complex stresses

Course Postgraduate PhD

Asst. Prof. Dr. Alaa Dahham Younis
 Complex stresses
In a practical design, a structure is usually subjected to a combination of different types of
loading that generate different types of stresses within the structure. For example, the stress
field in a beam–column joint is very complex, with combinations of bending, shearing and
contact stresses (Figure 1.1(a)). If a cut (plane) is taken through a point, the stress on the plane
is usually different to the stress on a different plane through the same point, not just in terms
of magnitude but also direction. On an arbitrary plane through a point, a general () stress can
always be resolved into three independent components that are perpendicular to each other
(Figure 1.1(b)). The three components include a normal stress (n), which is perpendicular to the
plane, two shear stresses (τ1 and τ2), which are parallel to the plane and perpendicular to each
other. The stresses at a point inside the joint are best presented by the stresses acting on an
infinitesimal cubic element taken around the point. The element has six faces (planes) that are

Figure 1.1
either perpendicular or parallel to each other. On each of the faces there are three independent
stresses, including two shear stresses and a normal stress.
Figure 1.2 shows all the stresses at a point in a material, which is sufficient and necessary to
represent the state of stress at the point.

• When a structure is subjected to external

loads, the state of stress is, in general,
different at different points within the structure.
• At a point in the structure, the stress in one
direction is usually different to the stress
in a different direction
• State of stress (Figure 1.1(c)) shows stresses
acting on six different planes at a point.
Therefore, when we say that we know a stress, Figure 1.2
it means that we know not only the
magnitude and direction of the stress, but also
the plane on which the stress acts.
• Since the cubic element has infinitesimal dimensions in the three co-ordinate directions, the
normal stresses acting on any two faces that are parallel to each other are equal but in opposite
directions. On any two planes that are perpendicular to each other, the shear stresses
perpendicular to the intersection of the two planes are equal, but in an opposite sense, i.e., are
either towards or away from the intersection line. At a point, therefore, there are only six
independent stresses, i.e., xx , yy , zz , τxy = τyx , τxz = τzx and τyz = τzy (Figure 1.2).
• At any point within a material, if the two shear stresses are zero on a plane, this plane is called
principal plane. The normal stress acting on the principal plane is called principal stress, and its
direction is called principal direction. If a cubic element is chosen such that all the faces of the
cube are free of shear stresses, the element is called principal element.

1.1 Two-dimensional state of stress

In some cases, the stresses relative to a particular direction are sufficiently small compared to
the stresses relative to the other two directions. Typical example problems include stresses in
a thin plate subjected to in-plane loadings (Figure 1.3(a)) and in a thin-walled vessel under
internal pressure (Figure 1.3(b)) or torsion (Figure 1.3(c)). Suppose that the small stress is
related

Figure 1.3
To the z direction and is ignored, the three-dimensional state of stress can be reduced to a
two-dimensional one. Since the remaining stresses lie in a plane, the simplified two-dimensional
problems are called plane problems. For the thin plate subjected to in-plane loads, Figure 1.3(a)
shows the two-dimensional state of stress. For a thin-walled cylinder subjected to internal
pressure or torsion, the states of stress are shown by Figures 1.3(b) and 1.3(c).
In Figure 1.3, the normal stresses (x and y have a single subscript index that indicates the
coordinate axis the stresses are parallel to. The first subscript index of a shear stress (τxy or τyx )
denotes the direction of the normal of the plane on which the stress acts, while the second
index denotes the axis to which the shear stress is parallel. Since the two-dimensional element
is infinitesimal, τxy is numerically equal to τyx .

1.1.1 Sign convention of stresses

The stresses shown in Figure 1.3 are all defined as positive in the chosen coordinate system,
where the following sign conventions are followed

• Tensile and compressive stresses are always defined, respectively, as positive and negative
• Positive shear stresses are defined the same as the positive shear forces

1.1.2 Analytical method

Once the stress components that align with a typical x–y coordinate system are found
(Figure 1.2), transforming the stresses into another coordinate system is sometimes necessary.
Two key reasons that we may want to calculate stresses in a different coordinate system include
the following
• To determine the stress in an important direction, for example, stresses normal and parallel to
the plane of a weld (Figure 1.4(b));
• To determine the maximum normal stress or maximum shear stress at a point. These stresses
may not necessarily align with the chosen coordinate directions
 Figure 1.4

(a) Stresses on an arbitrarily inclined plane

To further understand the stresses on an arbitrarily inclined plane (cut) through a point in
a material, consider the two elements taken around the point in Figure 1.4.
Since the elements are taken at the same point, we might take them for the same state of stress,
but measured in a different coordinate system. The new coordinate system, x`–y`, is defined by a
rotation θ of both the coordinate axes from their original directions (Figure 1.5). The
relationship between the states of stress in terms of the original system and the rotated system can
be best presented by considering the equilibrium of the wedge of material (Figure 1.6) taken from
Figure 1.5(b). The normal stress,θ , and shear stress,τθ , acting on the inclined plane are,
respectively, the normal stress, x`, and shear stress, τx`y`, in the rotated coordinate system, while
the stresses acting on the vertical and horizontal sides of the wedge are identical to those acting
on the vertical and horizontal sides of the element (Figure 1.5(a)) in the original coordinate
system.
 Figure 1.5

The equilibrium of the wedge yields:

(1.1)

Where an anticlockwise angle from the x-axis is defined as positive. Figure 1.6
(b) Principal stresses
From Equation (1.1), the stresses on an inclined plane change as the value of changes.
It means that on different planes taken by cutting through the point the stresses are generally
different. It is natural to think that there are special planes on which the normal stress reaches
either maximum or minimum (maximum compressive stress) algebraically.
The maximum and minimum normal stresses are both called principal stresses. When a normal
stress is either maximum or minimum, the plane on which the stress acts is always free of shear
stress. In a two-dimensional stress system, there are two principal stresses, that is, the maximum
and the minimum normal stresses at a point, as shown in Figure 1.7.
The principal stresses can be calculated as follows:

Figure 1.7 (1.2)

(c) The directions of principal stresses
The angle between a principal stress and the x-axis can be calculated as follows:

(1.3)

Figure 1.8
Since the two principal stresses are perpendicular to each other, the direction of the second
principal stress is θ+90.
In a plane problem, there are two principal stresses (Equation (1.2)) and two associated
directions. The easiest way to relate the stresses to their respective directions is based on the
following simple observation.

In Figure 1.8, the shear stresses τxy generates tension in one diagonal direction and
compression in the other, which suggests that combined with actions of x and y , the
normal stress in the direction of the tension diagonal is more tensile or larger than that in
the direction of the compression diagonal. Hence, it can be concluded that the direction of
1 is related to where the shear stresses are pointing to.
At a point in a material, a normal stress is a principal stress if:
• the stress is either the maximum tensile stress or the maximum compressive stress at the point;
Or
• the plane on which the normal stress acts is free of shear stresses. The plane is one of the
principal planes.

(d) Maximum shear stress

Following the same argument as for the existence of maximum normal stresses, there exist special
planes on which shear stress reaches maximum or minimum. (They have equal magnitudes in an
opposite sense.) Figure 1.9 shows a concrete cylinder under compression.
The cylinder fails due to maximum shearing at about 45 to the axial of compression. The cylinder
may fail along the other diagonal direction under the same compression due to an equal shear
stress of opposite sense. By observation, the plane perpendicular to the axial direction is a
principal plane since there is no shear stress acting on this plane. The maximum shear stress acts
on the plane that is 45 away from the principal plane. From Equation (1.1), the maximum shear
stress can be obtained by replacing x and y with 1 and 2, respectively:

(1.4a)

And from Equation (1.2):

(1.4b)

Figure 1.9
1.1.3 Graphic method
Mohr’s circle illustrates principal stresses and stress transformations via a graphical format,
that is, a graphic representation of Equations (1.1–1.4). The circle is plotted in a plane
coordinate system where the horizontal axis denotes normal stress. The vertical coordinate
denotes the shear stress on the same plane (Figure 1.10). While plotting a Mohr’s circle a sign
convention, for example,, must be followed.
Here we take tensile stresses and shear stresses that would turn an element clockwise as positive.
Figure 1.10 shows how the stresses acting on an element are related to a Mohr’s circle

Figure 1.10
The two principal stresses are shown by the  coordinates of the two intersections of the circle
with the horizontal axis (where shear stresses are zero). The vertical coordinates of either the
highest or the lowest point on the circle denote the maximum magnitude of shear stress that is
also equal to the radius of the circle. If the state of stress at a point is known, that is, x , y and
τxy are known, the following steps can be followed to plot a Mohr’s circle:
• Set up a co-ordinate system where the horizontal axis is the normal stress axis and the vertical
axis is the shear stress axis; positive directions of the axes take upwards and to the right.
• Locate two points, A and B, related to, respectively, the stresses on the right and upper faces of
a state-of-stress element, with respective coordinates (x`, τxy) and (y`, −τxy ), in the – co-
ordinate system and connect the two points by a straight line. The sign convention defined in
Section 1.1.1 must be followed to locate the two points. The intersection of the straight line with
the -axis is marked ‘O’ and is at a distance of ave = x+y /2 away from the origin.
• With its centre at ‘O’, draw a circle passing through points A and B.
• Measure coordinates of the two intersections of the circle with the -axis to obtain the two
principal stresses.
• Measure the radius of the circle to obtain the maximum shear stress.
• To determine the magnitudes of the stresses acting on an inclined plane,θ away from the right-
hand-side face, measure an angle of 2θ from OA and take the coordinates of the intersection with
the circle.
• The horizontal and vertical coordinates of the intersection are, respectively, the normal and
shear stresses on the inclined plane.
1.2 Key points review

1.2.1 Complex stress system

• At a point in a material, there are six independent stress components, including three normal
stresses and three shear stresses.
• In a two-dimensional case, there are three independent stresses, two normal stresses and one
shear stress, at a point of the material.
• A stress usually varies from point to point.
• A stress is uniquely defined by the following three properties:
– magnitude
– direction
– plane (cross-section/cut) on which the stress acts.
• Without knowing any of the three, the stress is not completely defined.
• Principal stresses are normal stresses and include both maximum and minimum compressive
stresses.
• In a three-dimensional stress system, there are three principal stresses; while in a two-
dimensional system, there are two principal stresses.
• Principal stresses are always perpendicular to each other.
• The plane on which a principal stress acts is free of shear stresses.
• Maximum shear stress is equal to half of the maximum difference between principal stresses.
• Maximum shear stress is always 45 away from a principal stress.
1.2.2 Mohr’s circle
• The largest and smallest horizontal coordinates of the circle are, respectively, the two principal
stresses 1 and 2.
• The maximum shear stress is numerically equal to the length of the radius of the circle and also
equal to 1−2/2.
• An angle difference between two planes through a point is represented by a difference of 2θ
between the two locations relative to the stresses on the two planes along the Mohr’s circle.
• A normal stress equal to x+y /2 acts on the planes of maximum shear stresses.
• If 1 = 2, Mohr’s circle degenerates into a point, for which no shear stresses develop at the
point

EXAMPLE 1.1
Consider the bar shown in Figure E1.2 under torsional loading. Determine the principal stresses
and their orientation.

Figure E1.2
[Solution]
The state of stress of the shaft is determined by the shear stress caused by the torque.
On the element taken, there are no normal stresses.
Thus:

Calculating the magnitudes of 1 and 2 (Equation (1.2)):

Calculate the orientation of these principal stresses (Equation (1.3)):

The direction of 1 is 45 away from the x-axis clockwise, which can be judged by the above
sketched deformation, where a tensile strain is observed in this direction

EXAMPLE 1.2
A thin-walled cylinder has an internal diameter of 60mm and a wall thickness of 1.5 mm.
Determine the principal stresses at a point on the outside surface of the generator when the
cylinder is subjected to an internal pressure of 6 MPa and a torque, about its longitudinal axis, of
1.0 kN m. If the cylinder is made from plates that are welded along the 45 seams, calculate the
normal and shear stresses along the seams.

Figure E1.2
[Solution]
The state of stress is defined by the axial and hoop stresses caused by the internal pressure and
the shear stress due to the applied torque. The pressure p acting on the end plate area A is
equivalent to an axial force of:

And the hoop force, Py , is equal to:

For the thin-walled cylinder, the cross-sectional area can be calculated approximately by dt.
For the infinitesimal element taken at A (Figure E1.2), the hoop stress is:
 The axial stress is:

Since the wall of the cylinder is thin, that is, t << d, the torsional constant is calculated
approximately as:

Where D is the outside diameter of the cylinder. Considering D ≈ d, the shear stress due to the
torque is then:

From Equation (1.2):

The maximum shear stress is (Equation (1.4a)):

Along the 45 seams (Equation (1.1)):

Complex Strain
Consider a rectangular block of material OLMN as shown in the xy plane (Fig. 1.11). The strains
along Ox and Oy are εx, and εy, and γxy is the shearing strain.

Fig. 1.11. Strains on an inclined plane

Let the diagonal OM be of length a; then ON = a cos θ and OL = a sin θ, and the increases
in length of these sides under strain are εx a cos θ and εy asin θ (i.e. strain x original length).
If M moves to M', the movement of M parallel to the x axis is
and the movement parallel to the y axis is

Thus the movement of M parallel to OM, which since the strains are small is practically
coincident with MM', is

Then

…. (1)

This is identical in form with the equation defining the direct stress on any inclined plane θ
with εx and εy replacing x and y and ½ γxy, replacing τxy, i.e. the shear stress is replaced by
HALF the shear strain.
Shear strain
To determine the shear strain in the direction OM consider the displacement of point P at the foot
of the perpendicular from N to OM (Fig. 1.12)

In the strained condition this point moves to P'

Since

During straining the line PN rotates

Fig. 1.12 Enlarged view of part of Fig. 1.11
counterclockwise through a small angle α
The line OM also rotates, but clockwise, through a small angle

Thus the required shear strain γθ in the direction OM, i.e. the amount by which the angle
OPN changes, is given by

Substituting for εθ from eqn. (1) gives

which again is similar in form to the expression for the shear stress τ on any inclined plane θ.
For consistency of sign convention, because OM’ moves clockwise with respect to OM it is
considered to be a negative shear strain, ie.
Principal Strain- Mohr’s strain circle
Since the equations for stress and strain on oblique planes are identical in form, as noted
above, it is evident that Mohr’s stress circle construction can be used equally well to represent
strain conditions using the horizontal axis for linear strains and the vertical axis for half the
shear strain. It should be noted, however, that angles given by Mohr’s stress circle refer to the
directions of the planes on which the stresses act and not to the direction of the stresses
themselves. The directions of the stresses and hence the associated strains are therefore
normal (i.e. at 90º) to the directions of the planes. Since angles are doubled in Mohr’s circle
construction it follows therefore that for true similarity of working a relative rotation of the
axes of 2 x 90 = 180” must be introduced. This is achieved by plotting positive shear strains
vertically downwards on the strain circle construction as shown in Fig. 1.13

The sign convention adopted for strains is as follows:

Linear strains: extension positive
compression negative

Shear strains:
The convention for shear strains is a little more
difficult. The first subscript in the symbol
γxy usually denotes the shear strain associated with
that direction, i.e. with Ox. Similarly, γyx is
the shear strain associated with Oy
Fig. 1.13. Mohr’s strain circle.
If, under strain, the line associated with the first subscript moves counterclockwise with respect
to the other line, the shearing strain is said to be positive, and if it moves clockwise it is said to
be negative. It will then be seen that positive shear strains are associated with planes carrying
positive shear stresses and negative shear strains with planes carrying negative shear stresses

Thus

Mohr’s circle for strains εx, εy and shear strain γxy (positive referred to x direction) is therefore
constructed as for the stress circle with ½ γxy replacing τxy and the axis of shear reversed, as
shown in Fig. 1.14.

The maximum principal strain is then ε1 at an angle θ1 to εx in the same angular direction as
that in Mohr’s circle (Fig. 1.14).
Again, angles are doubled on Mohr’s circle

Strain conditions at any angle α to εx are found

as in the stress circle by marking off an angle
2u from the point representing the x direction,
i.e. x’. The coordinates of the point on the
circle thus obtained are the strains required.
Alternatively, the principal strains may be Fig. 1.14. Strain system at a point, including
determined analytically from eqn. (1). the principal strains and their inclination
 i.e.

As for the derivation of the principal stress equations, the principal strains, i.e. the maximum and
minimum values of strain, occur at values of 8 obtained by equating dεθ/dθ to zero.

i.e. once again identical in form to the principal stress equation with ε replacing  and ½ γ
replacing τ.

Similarly
Stress – Strain curve
the Stress vs Strain curve for a material is unique, irrespective of the geometric dimensions of
the material specimen. Thus, for studying engineering properties of a material, Stress- Strain
curve is commonly used.
The following is an example of a Stress-Strain curve for mild steel specimen obtained by
performing simple tension test.

Further Observations from Stress-Strain

Curve Strain Hardening:
Strain Hardening is a phenomenon occurs in
certain materials, particularly ductile metals,
when it is strained beyond the yield point. An
increased stress is required to produce
additional plastic deformation and the metal
apparently becomes stronger and more difficult
to deform.
Thus, Strain Hardening, also known as Work
hardening or cold working, is defined as the Figure 1.15 Stress-Strain curve for mild steel
strengthening of a metallic material by plastic
deformation. This strengthening occurs
because of dislocation movements and
dislocation generation within the crystal
structure of the material.
In the Stress-Strain curve of Figure 1.15, the zone between the lower yield point (Y2) and the
Ultimate Stress (U) is called as the strain-hardening zone.

In Figure 1.16, the phenomenon of strain-

hardening is more clearly illustrated
For work hardening materials the yield stress
increases with increasing plastic deformation.
The strain can be decomposed into a
recoverable elastic strain, εe and an inelastic
strain, εp. The stress at initial yield is 0. But,
once the materials is stressed beyond 0 and up
to the point 3, at stress y, and unloaded, it has
a permanent strain p at zero stress. This
permanent strain at zero stress develop due to
onset of plasticity in the material is also called
as the permanent set or residual strain. If once Figure 1.16: Illustration of the
again the material is gradually stressed its Phenomenon of Strain hardening
stress-strain curve follows the straight line path
till it reaches y after which it again deforms
plastically. This means that the yield stress is
y for the second cycle of loading, which is
greater than the yield stress in the first cycle of
loading, 0. This phenomenon is called as
strain-hardening.
strain-hardening exponent
An empirical mathematical relationship was advanced by Holloman in 1945 to describe the
shape of the engineering stress-strain curve.

where is the σ true stress, ε is true strain, K is a strength coefficient (equal to the true stress at ε =
1.0), and n is the strain-hardening exponent. Thus, one can obtain n from a log-log plot of σ
versus ε.

Strain-hardening exponent

n = 0 for perfectly plastic solids

n = 1 for perfectly elastic solids
n = 0.1 – 0.5 for most metals

Strain-hardening rate
Necking
Necking is a mode of tensile deformation in ductile materials where relatively large amounts of
strain localize disproportionately in a small region of the material. This results in prominent
decrease in local cross-sectional area and the material specimen assumes a shape in the form of
a "neck". This phenomenon is called necking.

In the Stress-Strain curve of Figure 1.15, the zone between the Ultimate Stress (U) and the point
of Rupture (B) is called as the necking zone.

Figure 1.17: The phenomenon of necking

 Criteria for Necking
Let us start by considering the amount of force (dF) that is required to deform a specimen by dε.

The slope of the stress strain curve is:

(d/dε) is the Work Hardening Rate. It is the slope of the stress-strain curve. It is always positive

(dA/dε) is the Rate of Geometrical Softening. It is the rate at which the cross-sectional area of the
specimen decreases with increasing strain due to constancy of volume. It is always negative

Local ↓ in A (i.e., deformation) causes that region to strain harden locally (relative to the rest of
the cross section). The remainder of the cross section then deforms until a uniform cross-section
is re-established.
The rates balance at the UTS [(dA/dε) = (d/dε)].

When (dA/dε) > (d/dε), deformation becomes unstable. The material cannot strain harden fast
enough to inhibit necking

The criteria for instability is defined by the condition where the slope of the force distance curve
equals zero (dF = 0):

….. *
Recall that deformation is a constant volume process.
Thus:

If we invoke the instability criteria from above (*) then we get

Thus, at the point of tensile instability

If we incorporate engineering strain e, into the equation presented above, we can develop a more
explicit expression