Topics in Ship Structures

07 Linear Fracture Mechanics

Reference : Fracture Mechanics by T.L. Anderson

Lecture Note of Eindhoven University of Technology

2017. 10
by Jang, Beom Seon

OPen INteractive Structural Lab

0. INTRODUCTION
Historical Overview 2

Griffith theory for brittle material (1920’s)

 “A crack in a component will propagate if the total energy of the
system is lowered with crack propagation.”
 “ if the change in elastic strain energy due to crack extension > the
energy required to create new crack surfaces, crack propagation will
occur”

Irwin (1940’s) extended the theory for ductile materials.

 “ the energy due to plastic deformation must be added to the surface
energy associated with creation of new crack surfaces”
 “ local stresses near the crack tip are of the general form”

OPen INteractive Structural Lab

0. Introduction
Minimum total potential energy principle 3

 Fundamental concept used in physics, chemistry, biology, and engineering.

 It dictates that a structure or body shall deform or displace to a position
that (locally) minimizes the total potential energy, with the lost potential
energy being converted into kinetic energy (specifically heat).

 The potential energy of an elastic body, Π, is defined as follows:

𝜫=𝑼−𝑭
Here, 𝑼 : strain energy stored in the body
𝑭 : the work done by external loads

 Examples
 A rolling ball will end up stationary at the bottom of a hill, the point of
minimum potential energy. It rolls downward under the influence of gravity,
friction produced by its motion transfers energy in the form of heat of the
surroundings with an attendant increase in entropy.
 Deformation of spring under gravity stretches and vibrates and finally stops
due to the structural damping.

OPen INteractive Structural Lab

0. Introduction
Minimum total potential energy principle 4

 This energy is at a stationary position when an infinitesimal variation from

such position involves no change in energy.
𝛿𝛱 = 𝛿(𝑈 − 𝐹)
 The principle of minimum total potential energy may be derived as a special
case of the virtual work principle for elastic systems subject to conservative
forces.
 The equality between external and internal virtual work (due to virtual
displacements) is:

𝑆𝑡
𝛿u𝑇 T𝑑𝑆 + 𝑆𝑡
𝛿u𝑇 f𝑑𝑉= 𝑉
𝛿𝜀 𝑇 𝜎𝑑𝑉
𝑢 = vector of displacements
𝑇= vector of distributed forces acting on the part of the surface
𝑓= vector of body forces
 In the special case of elastic bodies
𝛿U = 𝑉 𝛿𝜀 𝑇 𝜎𝑑𝑉,𝛿F = 𝑆𝑡 𝛿𝑢𝑇 𝑇𝑑𝑆 + 𝑉 𝛿𝑢𝑇 𝑓𝑑𝑉
∴ 𝛿𝑈 = 𝛿𝐹
 The basis for developing the finite element method.

OPen INteractive Structural Lab

1. An Atomic View of Fracture
An atomic view of fracture 5

Physical meaning of fracture

 A material fractures when sufficient stress
and work are applied at the atomic level to
break the bonds that hold atoms together.
 The bond strength is supplied by the
attractive forces between atoms.
 A tensile force is required to increase the
separation distance from the equilibrium
value, this force must exceed the cohesive
force to sever the bond.
 The bond energy is given by

 x0 : the equilibrium spacing

 P : the applied force

OPen INteractive Structural Lab

1. An Atomic View of Fracture
Calculation of cohesive stress 6

 Idealizing the interatomic force-displacement relationship as a sine wave.

 For small displacements,

⇒ 𝑃 = 𝑘𝑥

 Assume n atomics within unit area A(=1), 𝒏𝒌 → stiffness

n kxo nPc xo      xo 
 Left side E Right side     c  ⇒
A( 1) A( 1)      
𝑷𝑳 𝑬𝑨 𝑬𝑨 𝒌𝑳
𝜹 = 𝑬𝑨 → 𝑷 = 𝜹, 𝒌= →𝑬=
𝑳 𝑳 𝑨

 When λ is assumed to be equal to the atomic spacing

OPen INteractive Structural Lab

1. An Atomic View of Fracture
Surface energy 7

 Surface energy

force per unit area distance

 Cohesive stress in terms of surface energy

E  x  s   c2
xo
c  ,  c o
 xo E E

γs : the surface energy per unit area,

= one-half of the fracture energy

 because tension should be applied two atoms in opposite direction

and surface energy of 2γs is required to split the atoms.

OPen INteractive Structural Lab

2. Stress Concentration Effect of Flaws
Stress concentration effect of flaws 8

Effect of flaws
 The theoretical cohesive strength of a material is approximately c=E/π
 Experimental fracture strengths for brittle materials are typically three
or four orders of magnitude below this value.
 The discrepancy between the actual strengths of brittle materials and
theoretical estimates was due to flaws.
 Fracture occurs when the stress at the atomic level > cohesive strength
of material.
 Flaws lower the global strength by magnifying the stress locally.
 The first quantitative evidence for the stress concentration effect of flaws
was provided by Inglis

OPen INteractive Structural Lab

2. Stress Concentration Effect of Flaws
Stress concentration effect of flaws 9

Stress concentration
 The hole is not influenced by the plate B.C.
 i.e. plate width >> 2a, plate height >>2b

 Where the ratio σA/σ is defined as stress concentration factor kt. When
a=b, the hole is circular and kt=3.0.

 As the major axis a↑, relative to b, the elliptical hole → a sharp crack.

Inglis ⇒

When a>>b, ⇒

OPen INteractive Structural Lab

2. Stress Concentration Effect of Flaws
Calculation of global strength 10

 Calculation of remote stress at failure (by Inglis)

 An infinite stress at the tip of an infinitely sharp crack, where ρ =0.
 Paradox: A material containing a sharp crack should theoretically fail
upon the application of an infinitesimal(극소의) load.
⇒ this paradox of a sharp crack motivated Griffith to develop a
fracture theory based on energy rather than local stress.
 Metals deform plastically, which causes an initially sharp crack to blunt.
 In the absence of plastic deformation, the minimum radius a crack tip
can have is on the order of atomic radius ⇒ ρ = x0
 Local stress concentration at the tip of an atomically sharp crack:

Stress at A : Cohesive stress :

 Crack/fracture occurs when

= ⇒

OPen INteractive Structural Lab

3. The Griffith Energy Balance
The Griffith energy balance 11

Griffith’s idea
 First law of thermodynamics : when a system goes from a non
equilibrium state to equilibrium, there is a net decrease in energy.
 The crack is formed by the sudden annihilation (소멸) of the tractions
acting on its surface.
 The new state is not under equilibrium, then by the theorem of
minimum potential energy, the potential energy is reduced by the
attainment of equilibrium.
 A crack can form only if the process causes the total energy to
decrease or remain constant.
 The critical conditions for fracture can be defined as the point where
crack growth occurs under equilibrium conditions, with no net change
in total energy.

Traction

OPen INteractive Structural Lab

3. The Griffith Energy Balance
The griffith energy balance 12

Griffith’s idea

Traction Crack occurs

Potential energy Π1 Potential energy Π2

Then where the energy : ΔΠ=Π1 - Π2 (Π1 > Π2 )gone?

 The reduced potential energy is used to create new crack

surfaces.

OPen INteractive Structural Lab

3. The Griffith Energy Balance
The griffith energy balance 13

The Griffith energy balance for an incremental increase in the crack

area dA,
Area of
energy
Release.

  a2

 2 a 2 B
  0  A  2aB
E
d da d  d  A  d  1  d    2 a
       
dA dA da dA  2 B   da  2 B  da  E

dWs 1  dWs 
    2 s
dA 2 B  da 

 2f a  2 E s 
1/2

 2 s f   A through-thickness crack in an
E  a  infinitely wide plate subjected to
a remote tensile stress
OPen INteractive Structural Lab
3. The Griffith Energy Balance
Comparison of two approaches 14

 Inglis’s approach

 Griffith’ approach

 When crack-tip radius is equaling to atomic space order, these two

approaches are consistent with one another, at least in the sharp
crack in an ideally brittle solid.

 But, when the crack-tip radius (ρ)is significantly greater than the
atomic space, there is an apparent contradiction between two
approaches. However,  a B 2 2
  0 
E
 Griffith model is insensitive to the notch radius as long as a>>b
 According to the Inglis stress analysis, in order for σc to be attained at
the tip of notch, σf must vary with (1/ρ)1/2.

OPen INteractive Structural Lab

3. The Griffith Energy Balance
Comparison of two approach 15

Inglis’s approach v.s. Griffith’ approach

Example:

Consider a crack with ρ = 5 × 10-6m. Such a crack would appear sharp

under a light microscope, but ρ would be four orders of magnitude larger
than the atomic spacing in a typical crystalline solid. Thus the local stress
approach would predict a global fracture strength 100 times larger than
the Griffith equation. The actual material behavior is somewhere between
these extremes; fracture stress does depend on notch root radius, but
not to the extent implied by the Inglis stress analysis.

⇒ ρ = x0 = ⇒

1/2
 E s  
f  
 4 ax 0 
OPen INteractive Structural Lab
3. The Griffith Energy Balance : Lecture Note of Eindhoven University of Technology
Energy balance 16

 Energy types
E : total energy
 : potential energy supplied by the internal strain
energy and external forces

Ws : work required to create new free surfaces

U k : kinetic energy resulted from material velocity
Plate with a line crack of length a.
U d : dissipated energy due to friction and plastic
deformation

Energy balance in terms of time

E   Ws  U d  U k derivative

d da d d Tran from time derivative to

   a  
dt dt da da state variable derivative
dE d  dWs dU d dU k Energy balance in terms of derivative
   
da da da da da of state variable, a

OPen INteractive Structural Lab

3. The Griffith Energy Balance : Lecture Note of Eindhoven University of Technology
Griffith energy balance 17

 Energy types
E : total energy
 : potential energy supplied by the internal strain
energy and external forces

Ws : work required to create new free surfaces

U k : kinetic energy resulted from material velocity
Plate with a line crack of 0
length a.
U d : dissipated energy due to friction and plastic
0 deformation
 Energy balance equation with all energy terms
dE d  dWs dU d dU k
    Assumption for Griffith energy balance ①
da da da da da
0 0

dE d  dWs Assumption for Griffith energy balance ②

 
da da da
0

 Griffith energy balance equation

d  dWs
 
da da OPen INteractive Structural Lab
3. The Griffith Energy Balance : Lecture Note of Eindhoven University of Technology
Griffith energy balance 18

dWs
d  dWs  4 B s
  Griffith energy balance equation da
da da
d   2 a
Left side :  
G
1 d
R
1  dWs  Right side : dA E
B da energy release rate 
B  da 

crack resistance force

Crack growth criterion by the relation of 𝑎 with

critical crack length 𝑎𝑐

 d  dWs 
a  ac 
 da
 
da 
: no crack growth

 d  dWs 
A plate loaded in tension and fixed at its edges a  ac    : unstable crack growth
 da da 
energy Ws
 d  dWs 
a  ac    : critical crack length
 da da 

a ⇒ For a crack increase in size, sufficient potential energy

must be available in the plate to overcome the surface
energy


Surface and internal energy vs. crack length OPen INteractive Structural Lab
3. The Griffith Energy Balance
Modified Griffith equation 19

Original Griffith equation

 This equation is valid only for ideally brittle solids.
 A good agreement is obtained between the equation and the
experimental fracture strength of glass.
 However, severely underestimates the fracture strength of metals.

Modified Griffith equation

 Irwin and Orowan independently modified the Griffith equation to
account for materials that are capable of plastic flow.

 γp : plastic work per unit area of surface created.

 γs : the total energy of broken bonds in a unit area
(ideally brittle solid)
 γp >> γs

OPen INteractive Structural Lab

3. The Griffith Energy Balance
Modified Griffith equation 20

A Generalized Griffith model

 Account for any type of energy dissipation
 wf : Fracture energy which could include plastic, viscoelastic,
viscoplastic effects, crack meandering and branching (which increase
the surface area), etc.

wf = γp + γs
wf = γp
quasi-brittle elastic-plastic material

brittle material with crack meandering (구불

ideally brittle material 구불한 길)and branching
OPen INteractive Structural Lab
3. The Griffith Energy Balance
Modified Griffith equation 21

A Caution of applying Griffith equation

 The Griffith model only could apply to linear elastic material behavior.
 The global behavior of structure must be elastic.
 Nonlinear effects such as plasticity, must be confined to a small region
near the crack tip
 2 a 2 B
  0  A  2aB
E

 Wf is assumed as constant, however, in many ductile materials fracture

energy increases with crack growth.

OPen INteractive Structural Lab

4. Energy Release Rate
Modified Griffith equation 22

Energy release rate (by Irwin)

 a measure of the energy available for an increment of crack
extension

 the rate of change in potential energy with the crack area (Not
with time).
 crack extension force or the crack driving force
 The potential energy of an elastic body, Π, is defined as follows:

 U  F
 U : strain energy stored in the body
 F : the work done by external loads

OPen INteractive Structural Lab

4. Energy Release Rate
Cracked Plate at a fixed load P 23

P
F  P U
2
P P
 U  F   P  
2 2
   U
1  d  1  dU  P  d 
G        
B  da  P B  da  P 2 B  da  P

Crack extension da results in a net increase in strain energy

From   
P From left figure
2 (dF ) p  Pd 
Pd 
(d  ) p   P(  d  ) P Pd 
2 (dU ) p   
2 2 2
(dU ) p  dF  d  Pd  Pd 
(d  ) p   Pd   
2 2
Pd  Pd 
(dU ) p  Pd    P(  d ) P Pd 
2 2 i.e.    P(  d )   
2 2 2

OPen INteractive Structural Lab

4. Energy Release Rate
Cracked Plate at a fixed displacement  24

P
F 0 U
2
P P
 U  F  0    U
2 2
1  d  1  dU    dP 
G           
B  da   B  da   2 B  da  

Crack extension da results in a net decrease in strain

energy since dP<0
P From left figure
From  
2 (dF )   0
dP
(d  )  ( P  dP ) P dP
2 (dU )     0
2 2 2
(dU )   dF  d  dP dP
(d  )   0 
2 2
dP dP ( P  dP ) P dP
(dU )   0   i.e.    
2 2 2 2 2

OPen INteractive Structural Lab

4. Energy Release Rate
Cracked Plate at a fixed load P 25

 The absolute values of these energies differ by the amount dPd/2

Pd  dP
(dU ) p  (dU )  
2 2
Pd   dP d dP 
    (dU ) p  (dU ) 
2  2 2 
d dP
2

OPen INteractive Structural Lab

4. Energy Release Rate
Comparison of two approaches 26

 introduce the compliance, which is the inverse of the plate stiffness


C
P

Fixed Load Fixed Displacement

P  d    dP 
GP    G    
2 B  da  P 2 B  da  

1
1 P2
Energy Release rate   PC  d   PdC P  C  dP   dC
 

P 2 dC
G  GP  G 
2 B da

Pd  dP
(dU ) p  ( 0) (dU )   ( 0)
2 2
Strain Energy
(dU )   (dU ) p

“에너지 해방률은 결국 하중의 제곱에 비례하고, Crack이 커지면서 유연성이 커지는 비율에 비례한다”

OPen INteractive Structural Lab

4. Energy Release Rate
Calculation of energy release rate from experiment 27

 Calculation of energy release rate by Load-displacement curve obtained

from a experiment

(1) (2) (3)

 Step 1 : Load – displacement curves are plotted with increasing crack

length 𝑎 (1)
∆
 Step 2 : Calculate compliance 𝑐 = 𝑃 for different 𝑎.
 Step 3 : Plot compliance 𝑐 versus crack length 𝑎. (2)
𝑑𝐶
 Step 4 : Calculate 𝑑𝑎 by differentiating compliance curve 𝑐 in terms of 𝑎 (3).
 Step 5 : For specific load 𝑃 and crack length 𝑎, calculate energy release.
P 2 dC
G
2 B da
OPen INteractive Structural Lab
4. Energy Release Rate
EX 2.2 : Beam with a central crack loaded in Mode I 28

Determine the energy release rate for a double cantilever beam (DCB)
specimen

 From Beam theory :  Pa 3 Bh3

 , I
2 3EI 12
 Compliance  8a 3 dC 24a 2
C  
P EBh3 da EBh3

 Energy Release Rate

P 2 dC P 2 24a 2 12 P 2 a 2
G   
2 B da 2 B EBh3 EB 2 h3
OPen INteractive Structural Lab
5. Instability and the R Curve
Calculation of energy release rate from experiment 29

 Driving force curve : G vs. crack size

 Resistance curve : R vs. crack size.

d G dR
 Stable condition : G= R and da  da
d G dR
 Unstable crack growth : 
da da
 A flat R curve : the material resistance is constant with crack growth.
 Fracture occurs when the stress reaches 𝜎2 . Unstable propagation.
 A rising R curve : for larger 𝜎4 , the plate is unstable with further crack growth.

Flat R curve Rising R curve

OPen INteractive Structural Lab
5. Instability and the R Curve
Reasons for the R Curve Shape 30

 The shape of the R curve depends on the

material behavior and, to a lesser extent, on the
configuration of the cracked structure.
 Ideally brittle material : flat R curve because the
surface energy is an invariant material property.
 Nonlinear material behavior : R curve can take
on a variety of shapes.
 ex) ductile fracture in metals : rising R curve
 a plastic zone at the tip of the crack Flat R curve
increases in size as the crack grows. The
driving force must increase to maintain the
crack growth.
 If the cracked body is infinite (i.e., if the
plastic zone is small compared to the body)
the plastic zone size and R eventually reach
steady-state values  flat with further
growth.

Rising R curve
OPen INteractive Structural Lab
5. Instability and the R Curve
Reasons for the R Curve Shape 31

 Ideally, the R curve, as well as other measures of fracture toughness,

should be a property only of the material and not depend on the size or
shape of the cracked body.
 Much of fracture mechanics is predicated on the assumption that fracture
toughness is a material property.
 However, the size and geometry of the cracked structure can exert some
influence on the shape of the R curve.
 A crack in a thin sheet  a steeper R curve than a crack in a thick plate
because of a low degree of stress triaxiality (3축) at the crack tip in the thin
sheet, while the material near the tip of the crack in the thick plate may be in
plane strain.
 The R curve can also be affected if the growing crack approaches a free
boundary in the structure. A wide plate may exhibit a somewhat different
crack growth resistance behavior than a narrow plate of the same material.

OPen INteractive Structural Lab

5. Instability and the R Curve
Load control VS. Displacement Control 32

 The stability of crack growth depends on the rate of change in i.e., the
second derivative of potential energy. Although the driving force G is the
same for both load control and displacement control, the rate of change of
the driving force curve depends on how the structure is loaded.
 Displacement control tends to be more stable than load control. The driving
force actually decreases with crack growth in displacement control.

 Ex) A cracked structure subjected to a load

P3 and a displacement Δ3.
 Load controlled, it is at the point of
instability where the driving force curve is
tangent to the R curve.
 Displacement control : the structure is
stable because the driving force decreases
with crack growth; the displacement must
be increased for further crack growth.

Driving force/R curve diagram for

load(P) and displacement control (Δ)

OPen INteractive Structural Lab

5. Instability and the R Curve
Load control VS. Displacement Control 33

 Ex 2.3) Evaluate the relative stability of a DCB specimen in load control and
displacement control.
P 2 dC P 2 24a 2 12 P 2 a 2  8a 3 dC 24a 2
G    C  
2 B da 2 B EBh 3
EB 2 3
h P EBh 3
da EBh3
Sol)
 Load Control : the slope of the driving force is given by (Ex.2.2)

P  d  P 6 Pa 2 P 2 a 2  Pa 3 Bh3
G       , I
2 B  da  P 2 B 3EI BEI 2 3EI 12
 Displacement control :
express G in terms of Δ and a.

  dP    3EI  9 2 EI
G        3 4  
2 B  da   2B  2a  4 Ba 4

Double cantilever beam (DCB)

specimen.

OPen INteractive Structural Lab

5. Instability and the R Curve
Structures with Finite Compliance 34

 Most real structures are subject to conditions between load control

and pure displacement control.
 Pure displacement control : infinite spring, Cm = 0.
 Load control : an infinitely soft spring, Cm =.
 At the moment of instability.

G= R

A cracked structure with finite

compliance, represented
schematically by a spring in
series.

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Introduction 35

 For certain cracked configurations subjected to external forces, it is possible

to derive closed-form expressions for the stresses in the body, assuming
isotropic linear elastic material behavior
 the stress field in any linear elastic cracked body is given by

leading term
high order term

where
𝜎ij : stress tensor
K : constant
fij : dimensionless function of θ
Am : amplitude
gij(m) : dimensionless function of θ for the mth term

Definition of the coordinate axis

 Stress near the crack tip varies with 1/ 𝑟 ahead of a crack tip. The z
 Stress singularity direction is normal to the page.

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
The Stress Intensity Factor 36

 Stress intensity factor : KI, KII, KIII

The stress fields ahead of a crack tip i(n
an isotropic linear elastic material are

Stress Fields Ahead of a Crack Tip(Linear Elastic, Isotropic Material)

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
The Stress Intensity Factor 37

 In a mixed-mode problem (i.e., when more than one loading mode is

present), the individual contributions to a given stress component
are additive from the principle of linear superposition.

Crack-Tip Displacement Fields (Linear Elastic, Isotropic Material)

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
The Stress Intensity Factor 38

 Consider the Mode I singular field on the crack plane, where θ = 0.

 The shear stress is zero, which means that the crack plane is a principal
plane for pure Mode I loading.
 Near the crack tip, where the singularity dominates the stress field.
 Stresses far from the crack tip are governed by the remote boundary
conditions.

 K is known, it is possible to solve for all

components of stress, strain, and
displacement as a function of r and θ.

 This single-parameter description of

crack tip conditions turns out to be one
of the most important concepts in
fracture mechanics

Stress normal to the crack plane in Mode I

OPen INteractive Structural Lab
6. Stress Analysis of Cracks
Relationship between K and Global Behavior 39

 The crack-tip stresses must be proportional to the remote stress  KI ∝ σ .

 Stress intensity has units of stress 𝑙𝑒𝑛𝑔𝑡ℎ.


 This means, the amplitude of the crack-tip singularity for this configuration is
proportional to the remote stress and the square root of the crack size.
 Through Crack

 The edge crack

Through crack of elliptical shape.

 The edge crack opens more because it

is less restrained than the through
crack.

Edge crack of V shape.

OPen INteractive Structural Lab
6. Stress Analysis of Cracks
Relationship between K and Global Behavior 40

 A through crack in an infinite plate where the normal to the crack plane is
oriented at an angle β with the stress axis.
 Redefine the coordinate axis to coincide with the crack orientation, the
applied stress can be resolved into normal and shear components.
By using Mohr’s circle and
formulas of trigonometric
function
1  cos 2
 cos 2 
2
sin 2  2sin  cos 

Through crack in an infinite plate for the general case

where the principal stress is not perpendicular to the
crack plane.

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Relationship between K and Global Behavior 41

 The penny-shaped crack in an infinite medium

A penny-shaped crack Elliptical cracks Semielliptical cracks

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Effect of Finite Size 42

 The crack dimensions are small compared to the size of the plate;
the crack-tip conditions are not influenced by external boundaries.
 As the crack size increases, or as the plate dimensions decrease,
the outer boundaries begin to exert an influence on the crack tip.

Collinear cracks in an infinite plate

Stress concentration effects infinite plate
and finite plate

As a / W  0, K I    a (Infinite plate)
As a / W  1, K I  
OPen INteractive Structural Lab
6. Stress Analysis of Cracks
Effect of Finite Size 43

 More accurate solutions for a through crack in a finite plate from

finite-element analysis and fit to a polynomial expression.

(2.46)

(2.45)

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Effect of Finite Size 44

 KI Solutions for Common Test Specimens

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Effect of Finite Size 45

 Plot of stress intensity solutions

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Effect of Finite Size 46

 K can always be related to the through crack with an appropriate correction

factor.

 EX. 2.4) Show that the KI solution for the single edge notched tensile panel
reduces to 𝐾𝐼 = 1.12𝜎 𝜋𝑎 when a << W.

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Principle of Superposition 47

 For linear elastic materials, individual components of stress, strain, and

displacement are additive.
 Stress intensity factors are additive as long as the mode of loading is
consistent.

 Ex) An edge-cracked panel subject to combined membrane (axial) loading

Pm, and three-point bending Pb.

OPen INteractive Structural Lab

6. Stress Analysis of Cracks
Principle of Superposition 48

 Ex. 2.5) Determine the stress intensity factor for a semielliptical

surface crack subjected to an internal pressure p.

KI = 0 because the crack faces

close, and the plate behaves
as if the crack were not
KI for a semielliptical surface crack under present

internal pressure p by means of the principle

of superposition.
OPen INteractive Structural Lab
6. Stress Analysis of Cracks
Principle of Superposition 49

 Stresses acting on the boundary (i.e., tractions) can

be replaced with tractions that act on the crack face.
 Mode I is assumed and no shear stresses act on
Plane A-B.

Uncracked body subject to

an arbitrary boundary
traction P(x), which results
in a normal stress
distribution p(x) acting on
Plane A-B.

KI = 0 because the crack faces

close, and the plate behaves
as if the crack were not
present

OPen INteractive Structural Lab

7. Relationship between K and G
Relationship between K and G 50

 G : the energy release rate, the net change in potential energy that
accompanies an increment of crack extension.
 K : quantity characterizes the stresses, strains, and displacements
near the crack tip.
 For a through crack

 For plane strain conditions, E must be replaced by .

OPen INteractive Structural Lab

7. Relationship between K and G
Relationship between K and G 51

 Is it general relationship that applies to all configurations?

 The work required to close the crack at the tip is related to the energy
release rate.

 ΔU is the work of crack closure, which is equal to the

sum of contributions to work from x = 0 to x = Δa.

Application of closure stresses

 The factor of 2 on the work is required because both which shorten a crack by a
crack faces are displaced an absolute distance uy(x).

OPen INteractive Structural Lab

7. Relationship between K and G
Relationship between K and G 52

 The normal stress required to close the crack is related to KI for the
shortened crack.

(  1) K I2
4(1  v) K I2 K I2 K I2   (3  4v),   G 
E
   2(1  v)
 8
E E E
2(1  v) (1  v )2

 Energy release rate, like energy, is a scalar quantity.

OPen INteractive Structural Lab

7. Crack-Tip Plasticity
Relationship between K and G 53

 Linear elastic stress analysis of sharp cracks predicts

infinite stresses at the crack tip.
 In real materials, however, stresses at the crack tip are finite
because the crack-tip radius must be finite.
 The elastic stress analysis becomes increasingly inaccurate
as the inelastic region at the crack tip grows.
 Simple corrections to linear elastic fracture mechanics
(LEFM) are available when moderate crack-tip yielding
occurs.
 The size of the crack-tip-yielding zone can be estimated by
two methods:
 the Irwin approach, where the elastic stress analysis is
used to estimate the elastic-plastic boundary,
 the strip-yield model.

OPen INteractive Structural Lab

7. Crack-Tip Plasticity
The Irwin Approach 54

 On the crack plane (θ = 0), the normal stress σyy in a linear elastic material
is given by

 the boundary between elastic and plastic behavior is assumed to occur

when the stresses satisfy a yield criterion, σyy =σYS.

First-order and second-order

estimates of plastic zone size
OPen INteractive Structural Lab
7. Crack-Tip Plasticity
The Irwin Approach 55

 The simple analysis in the preceding paragraph is

not strictly correct because it was based on an
elastic crack-tip solution. When yielding occurs,
stresses must redistribute in order to satisfy
equilibrium.
 The cross-hatched region represents forces that
would be present in an elastic material but cannot
be carried in the elastic-plastic material because the
stress cannot exceed the yield. First-order and second-order
 The plastic zone must increase in size in order to estimates of plastic zone size
accommodate these forces. A simple force balance leads
to a second-order estimate of the plastic zone size rp :

ry
KI KI KI 1 KI 1 K I2
 2r 1/2
 2ry 
1/2
2 
2 0 2 2 2  YS   YS

OPen INteractive Structural Lab

7. Crack-Tip Plasticity
The Irwin Approach 56

 effective crack length that is slightly longer than the actual crack size

 For plane stress For plane strain

 The effective stress intensity

 Since the effective crack size is taken into account in the geometry
correction factor Y, an iterative solution is usually required to solve
for Keff.

Keff

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
The Irwin Approach 57

 In certain cases, this iterative procedure is unnecessary because a closed-form

solution is possible. For example, the effective Mode I stress intensity factor for a
through crack in an infinite plate in plane stress is given by

 Elliptical and semielliptical flaws also have an approximate closed-form plastic zone
correction, provided the flaw is small compared to the plate dimensions.
 In the case of the embedded elliptical flaw, Keff is given by Elliptical cracks

 Effective compliance

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
The Irwin Approach 58

 It should be noted that the author does not recommend using the
Irwin plastic zone adjustment for practical applications.
 It was presented here primarily to provide a historical context to the
development of both linear and nonlinear fracture mechanics.

Definition of the effective compliance to

account for crack-tip plasticity

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
The Strip-Yield Model, Dugdale [24] and Barenblatt [25]
59

 A long, slender plastic zone at the crack tip is assumed in a nonhardening material in
plane stress.
 The strip-yield plastic zone is modeled by assuming a crack of length 2a + 2ρ, where
ρ is the length of the plastic zone, with a closure stress equal to σYS applied at each
crack tip.

The strip-yield model.

 This model approximates elastic-plastic behavior by superimposing two elastic

solutions: a through crack under remote tension and a through crack with closure
stresses at the tip.
 Since the stresses are finite in the strip-yield zone, there cannot be a stress
singularity at the crack tip.
0

 The plastic zone length ρ must be chosen such that the stress intensity factors from
the remote tension and closure stress cancel one another.

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
The Strip-Yield Model, Dugdale [24] and Barenblatt [25]
60

 The stress intensity due to the closure stress can be estimated by considering a
normal force P applied to the crack at a distance x from the centerline of the crack.
 The stress intensities for the two crack tips are given by assuming the plate is of unit
thickness.

Crack-opening force applied at a

distance x from the center-line.

 The closure force at a point within the strip-yield zone is equal to

 The total stress intensity at each crack tip resulting from the closure stresses is
obtained by replacing a with a + ρ in Equation

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
The Strip-Yield Model, Dugdale [24] and Barenblatt [25]
61

 The stress intensity from the remote tensile stress

must balance with Kclosure.

a  a 
K  K closure  0   (a   )  2 YS cos 1  
 a 
 ρ approaches infinity as σ σYS. Let us explore the strip-yield model further
by performing a Taylor series expansion.

 Neglecting all but the first two terms and solving for the plastic zone size
gives. the Irwin and strip-yield approaches predict similar plastic zone sizes.
Irwin approach

𝜋 1
= 0.392 , = 0.318
8 𝜋
OPen INteractive Structural Lab
8 Crack-Tip Plasticity
The Strip-Yield Model, Dugdale [24] and Barenblatt [25]
62

 One way to estimate the effective stress intensity with the strip-yield model is to set
equal to a + ρ.

  
(a   )  a sec  
 2 YS 

 It tends to overestimate Keff; the actual aeff is somewhat less than a + ρ because the
strip-yield zone is loaded to σYS.
 Burdekin and Stone [26] obtained a more realistic estimate of Keff for the strip-yield
model.

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
Comparison of Plastic Zone Corrections 63

 A pure LEFM

 The Irwin correction for plane stress

 The strip-yield correction on stress intensity

 Both the Irwin and strip-yield corrections

Comparison of plastic zone corrections
deviate from the LEFM theory at stresses
for a through crack in plane strain.
greater than 0.5σYS.
 The two plasticity corrections agree with
each other up to approximately 0.85σYS.

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
Plastic Zone Shape 64

 The estimates of plastic zone size that have been presented so far consider only the
crack plane θ = 0.

 For plane stress, Setting σe = σYS

𝜎3 =0 Crack tip stress by principal stress

(𝜎1 −𝜎2 )2 + (𝜎2 )2 + (𝜎1 )2 =2𝜎𝑌𝑆
2
𝐾𝐼 𝜃 𝜃
𝜎1 = cos {1 + 𝑠𝑖𝑛 }
2𝜋𝑟 2 2

𝐾𝐼 𝜃 𝜃
𝜎2 = cos {1 − 𝑠𝑖𝑛 }
2𝜋𝑟 2 2

𝜎3 =0 for plane stress

𝐾𝐼2 2 𝜃 𝜃 2
𝑐𝑜𝑠 {6𝑠𝑖𝑛2 + 2}=2𝜎𝑌𝑆 =
2𝜐𝐾𝐼
cos
𝜃
for plane strain
2𝜋𝑟𝑦 2 2
2𝜋𝑟 2

𝐾𝐼2 𝜃 𝜃 𝐾𝐼2 3
𝑟𝑦 =2𝜋𝜎2 𝑐𝑜𝑠 2 2 {3𝑠𝑖𝑛2 2 + 1}=4𝜋𝜎2 [1 + cos 𝜃 + 2 𝑠𝑖𝑛2 𝜃 ]
𝑌𝑆 𝑌𝑆

𝜃 1 𝜃 𝜃 1 HW #2 : Prove this.
2
𝑐𝑜𝑠 = 1 + cos 𝜃 , 𝑐𝑜𝑠 2 𝑠𝑖𝑛2 = 𝑠𝑖𝑛2 𝜃
2 2 2 2 4
OPen INteractive Structural Lab
8 Crack-Tip Plasticity
Plastic Zone Shape 65

 For plane strain, Setting σe = σYS

𝜎3 = ν(𝜎1 +𝜎2 )
(ν2 − ν + 1)(𝜎1 2 + 𝜎2 2 )+(2ν2 − 2ν − 1) 𝜎1 𝜎2 =2𝜎𝑌𝑆
2

Crack tip stress by principal stress

𝐾𝐼 𝜃 𝜃
2
𝐾𝐼2 𝜃 𝜃 𝜎1 = cos {1 + 𝑠𝑖𝑛 }
(ν − ν + 1) 𝑐𝑜𝑠 2 {2 + 2𝑠𝑖𝑛2 } 2𝜋𝑟 2 2
2𝜋𝑟𝑦 2 2
𝐾𝐼2 𝜃 𝜃 𝐾𝐼 𝜃 𝜃
+ (2ν2 − 2ν − 1) 2𝜋𝑟 𝑐𝑜𝑠 2 2 {2 − 2𝑠𝑖𝑛2 }= 2𝜎𝑌𝑆
2 𝜎2 = 2𝜋𝑟
cos 2
{1 − 𝑠𝑖𝑛 2
}
𝑦 2

𝐾𝐼2
𝜎3 =0 for plane stress
2 𝜃 𝜃
𝑐𝑜𝑠 6𝑠𝑖𝑛2 2 + 2(1 − 2𝑣) 2
=2𝜎𝑌𝑆
2
2𝜋𝑟𝑦 2 2𝜐𝐾𝐼 𝜃
= 2𝜋𝑟
cos 2
for plane strain
𝐾𝐼2 3
𝑟𝑦 = 2 [ 1 − 2υ 2 1 + cos 𝜃 + 2 𝑠𝑖𝑛2 𝜃 }
4𝜋𝜎𝑌𝑆

𝜃 1 𝜃 𝜃 1
𝑐𝑜𝑠 2 = 1 + cos 𝜃 , 𝑐𝑜𝑠 2 𝑠𝑖𝑛2 = 𝑠𝑖𝑛2 𝜃 HW #2 : Prove this.
2 2 2 2 4

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
Plastic Zone Shape 66

 Crack-tip plastic zone shapes estimated from the elastic solutions.

 Not strictly correct because they are based on a purely elastic analysis.
 The Irwin plasticity correction, which accounts for stress redistribution by
means of an effective crack length, is also simplistic and not totally correct.

 Plane stress
𝐾𝐼2 3
𝑟𝑦 =4𝜋𝜎2 [1 + cos 𝜃 + 2 𝑠𝑖𝑛2 𝜃 ]
𝑌𝑆

 Plane strain
𝐾𝐼2 3
𝑟𝑦 =4𝜋𝜎2 [ 1 − 2υ 2
1 + cos 𝜃 + 2 𝑠𝑖𝑛2 𝜃 }
𝑌𝑆

OPen INteractive Structural Lab

8 Crack-Tip Plasticity
Plastic Zone Shape 67

 The latter, which was published by Dodds et al. [27], assumed a material with the
following uniaxial stress-strain relationship:

 The definition of the elastic-plastic boundary is somewhat arbitrary, since materials

that can be described by the above Eq do not have a definite yield point. σe = σYS (the
0.2% offset yield strength).

Contours of constant effective stress Effect of strain-hardening on the

in Mode I, obtained from finite element Mode I plastic zone; n = 5 : a high strain
analysis, n=50 hardening, n = 50 : very low hardening
OPen INteractive Structural Lab
9. K-CONTROLLED FRACTURE
Introduction 68

 The stresses near the crack tip in a linear

elastic material vary as 1 𝑟; the stress intensity
factor defines the amplitude of the singularity.
 If K> Kcrit , cack extension must occur.
 Kcrit : a measure of fracture toughness, is a
material constant that is independent of the size
and geometry of the cracked body
 G also provides a single-parameter description
of crack-tip conditions, and Gc is an alternative
measure of toughness.
 Under certain conditions, K still uniquely
characterizes crack-tip conditions when a
plastic zone is present. In such cases, Kcrit is a
geometry-independent material constant.
 Assume that the plastic zone is small compared
to all the length dimensions in the structure and
test specimen. Schematic test specimen and structure
loaded to the same stress intensity. The
 KI characterizes crack-tip conditions. As the
crack-tip conditions should be identical in
load is increased, both configurations both configurations as long as the plastic
(specimen and structure) will fail at the same zone is small compared to all relevant
critical stress intensity, provided the plastic zone dimensions. Thus, both will fail at the
remains small in each case. same critical K value.
OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Introduction 69

 The NASA data exhibited an apparent effect of specimen thickness on

the critical stress intensity for fracture, Kcrit.
 The explanation that was originally offered was that thin specimens
are subject to plane stress loading at the crack tip, while thick
specimens experience plane strain conditions.
 The biaxial stress state associated with plane stress results in a higher
measured toughness than is observed in the same material when
subject to a triaxial stress state.
 This section presents an updated perspective on the interrelationship
between specimen dimensions, crack-tip triaxiality, and fracture
toughness.

Plastic crack tip zone for a thin and a thick plate

OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Crack-Tip Triaxiality 70

 If there was no crack, the plate would be in a state of plane stress. Regions of the
plate that are sufficiently far from the crack tip must also be loaded in plane stress.
 Because of the large stress normal to the crack plane, the crack-tip material tries to
contract in the x and z directions, but is prevented from doing so by the surrounding
material. This constraint causes a triaxial state of stress near the crack-tip.
 In the interior of the plate, the z stress, and therefore the level of triaxiality is high.
 The central region : plane strain.
 Near the free surface : low stress triaxiality.
 The free surface : pure plane stress.

Three-dimensional deformation at the tip of a crack. Schematic variation of transverse stress and
strain through the thickness
OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Crack-Tip Triaxiality 71

 These results were obtained from a three-dimensional elastic-plastic finite element

analysis performed by Narasimhan and Rosakis
 Under plane strain loading, the quantity σzz /(σxx + σyy) is equal to Poisson’s ratio for
elastic material behavior and is equal to 0.5 for incompressible plastic deformation.
 Material near the crack tip experiences high triaxiality, but σzz = 0 when x is a
significant fraction of the plate thickness.

 Consider a point on the crack plane (θ

= 0) just ahead of the crack-tip.
 σxx = σyy under linear elastic conditions.

 Plane stress : σzz = 0

 Plane strain : σzz = 2νσyy (ν = 0.3)
HW #2 : Prove this.

Transverse stress through the thickness as a

function of distance from the crack tip
OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Effect of thickens on Apparent Fracture toughness 72

 The measured Kcrit values decrease with specimen thickness until a plateau
is reached.
 This apparent asymptote in the toughness vs. thickness trend is designated
by the symbol KIc, and is referred to as “plane strain fracture toughness”.
 A decrease in apparent toughness with specimen thickness, generally
correspond to materials in which the crack propagation is ductile.

Variation of measured fracture toughness with Variation of measured fracture toughness with
specimen thickness for an unspecified alloy specimen thickness for 7075-T6 Aluminum
OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Effect of thickens on Apparent Fracture toughness 73

 In such tests, the crack “tunnels” through the center of the specimen.
 The crack grows preferentially in the region of high triaxiality. Crack growth
on the outer regions of the specimen lags behind, and occurs at a 45° angle
to the applied load. The resulting fracture surface exhibits a flat region in the
central region and 45° shear lips on the edges.
 Fracture toughness tests on very thin plates or sheets typically result in a
45° shear fracture. At larger thicknesses, there is generally some mixture of
shear fracture and flat fracture. The thickness effect on the apparent
fracture toughness is due to the relative portions of flat and shear fracture.
 In the limit of a very thick specimen, the flat fracture mechanism dominates,
and further increases in thickness have relatively little effect on the
measured toughness.

Effect of specimen thickness on fracture surface

morphology for materials that exhibit ductile crack growth.
OPen INteractive Structural Lab
10. Plane strain fracture : Fact vs. Fiction
Effect of thickens on Apparent Fracture toughness 74

 For very thin sections, plane strain conditions do not exist at x = xc. As the thickness
increases, the size of the plane strain zone increases relative to the low triaxiality
zone near the free surfaces.
 This trend are not indicative of a transition from “plane stress fracture” to “plane
strain fracture.”
 Rather, this trend reflects the differing relative contributions of two distinct fracture
mechanisms.
 In fact, there is no such thing as “plane stress fracture” except perhaps in very thin
foil. There is nearly always some level of triaxiality along the crack front.

Fracture process zone at the tip of a

crack tip Effect of thickness on stress—the crack-tip
stress state in the fracture process zone.

OPen INteractive Structural Lab

10. Plane strain fracture : Fact vs. Fiction
Plastic Zone Effects 75

 3D elastic plastic finite element analyses of fracture toughness specimens  a high

degree of triaxiality near the crack tip exists even when the entire cross-section has
yielded.
 Although K is not valid as a characterizing parameter under fully plastic conditions, a
single-parameter description of fracture toughness is still possible using the J integral,
or cracktip-opening displacement.
 The evolution of the Mode I plastic zone at mid-thickness in a plate containing an
edge crack. The plastic zone boundary is defined at σe = σYS.
 As the quantity increases relative to plate thickness B, the plastic zone size increases.
 At low KI values, the plastic zone has a typical plane strain shape, but evolves into a
plane stress shape at higher KI values.

Effect of KI, relative to thickness, of the

OPen plastic
INteractive Structural
zone size Lab
and shape.
10. Plane strain fracture : Fact vs. Fiction
Plastic Zone Effects 76

 Although the stress state at the plastic zone boundary is plane stress, the material
close to the crack tip is subject to a triaxial stress state. The figure depicts a plastic
zone in the center of an edge-cracked plate.
 Because the plastic zone size is of the same order of magnitude as the plate
thickness, the plastic zone has a plane stress shape.
 At the crack tip, however, there is a zone of high triaxiality. As stated above, the zone
of high triaxiality at the crack tip can persist even in the presence of large-scale
plasticity.
 When performing laboratory KIc tests on standard specimens, the following size
requirements have been adopted.

 Recall that the quantity (KIC/YS)2 is proportional

to the plastic zone size.
 The minimum requirements on the crack length
and ligament length (W – a) are designed to
ensure that the plastic zone is sufficiently small
for fracture to be K-controlled. => ensure plane
Cracked plate in which the plastic zone
strain conditions along the crack front. size is of the same order of magnitude as
the plate thickness
OPen INteractive Structural Lab