Introduction to Fracture

and Damage Mechanics

Wolfgang Brocks

Five Lectures
at

Politecnico di Milano

Milano, March 2012

Contents
I.

Linear elastic fracture mechanics (LEFM)

II.

Stress field at a crack tip

Stress intensity approach (IRWIN)
Energy approach (GRIFFITH)
J-integral
Fracture criteria fracture toughness
Terminology

Plasticity

III.

Fundamentals of incremental plasticity

Finite plasticity (deformation theory)
Plasticity, damage, fracture
Porous metal plasticity (GTN Model)

Small Scale Yielding

IV.

Plastic zone at the crack tip

Effective crack length (Irwin)
Effective SIF
Dugdale model
Crack tip opening displacement (CTOD)
Standards: ASTM E 399, ASTM E 561

Elastic-plastic fracture mechanics (EPFM)

V.

Deformation theory of plasticity

J as energy release rate
HRR field, CTOD
Deformation vs incremental theory of plasticity
R curves
Energy dissipation rate

Damage Mechanics

Deformation, damage and fracture

Crack tip and process zone
Continuum damage mechanics (CDM)
Micromechanisms of ductile fracture
Micromechanical models
Porous metal plasticity (GTN Model)

Concepts of Fracture Mechanics

Part I: Linear Elastic Fracture Mechanics

W. Brocks
Christian Albrecht University
Material Mechanics

Overview

I.

Linear elastic fracture mechanics

(LEFM)

II.

Small Scale Yielding

III. Elastic-plastic fracture mechanics (EPFM)

Milano_2012

Seite 1

LEFM

Stress Field at a crack tip

Stress intensity approach (Irwin)
Energy approach (Griffith)
J-integral
Fracture criteria fracture toughness
Terminology

Milano_2012

Stress Field at a Crack Tip

boundary conditions

yy ( x 0, y = 0) = ( r, = ) = 0
xy ( x 0, y = 0) = r ( r, = ) = 0
Hookes law of linear elasticity

ij = 2G ij +

1 2

kk ij
Inglis [1913], Westergaard
[1939], Sneddon [1946], ...

Williams series [1957]

= 0

Airys stress function

=r

rr =

+2

A cos + B cos ( + 2 ) +
+C sin + D sin ( + 2 )

A1 5
C1 5
2

3
3
3
1
4 cos 2 4 cos 2 +
4 sin 2 + 4 cos 2 + 4 A0 cos +
r
r

+ A1 r 94 cos 2 + 34 cos 52 + C1 r 94 sin 2 + 154 cos 52 + O (r )

Milano_2012

Seite 2

Fracture Modes

Milano_2012

LEFM: Stress Intensity Approach

Irwin [1957]: mode I

ij ( r , ) =

KI
fijI ( ) + T i11 j
2 r

KI = stress intensity factor

K I = a Y (geometry)

T = non-singular T-stress
Rice [1974]: effect on plastic zone
General asymptotic solution
stresses

ij (r , ) =

displacements

ui (r , ) =

1
K I fijI ( ) + K II f ijII ( ) + K III fijIII ( )
2 r

1
r
K I giI ( ) + K II giII ( ) + K III giIII ( )
2G 2

Milano_2012

Seite 3

Angular Functions in LEFM

Milano_2012

LEFM: Energy Approach

Elastic strain energy of a panel of thickness
B under biaxial tension in a circular
domain of radius r
U 0e =

A.A. Griffith [1920]

Br 2 2
2
2
( 1)(1 + ) + 2 (1 )
16G

insert hole: fixed grips

> energy release
e
U e = U 0e U rel

elliptical hole, axes a,b

e
U rel
=

B 2
32G

(1 + ) (1 ) ( a + b )
2

+2 (1 2 )( a 2 b 2 ) + (1 + ) ( a 2 + b 2 )

crack

e
=
U rel

a 2 B 2
8G

(1 + )

Milano_2012

Seite 4

Fracture of Brittle Materials

e
Urel
Usep 0
B ( 2a )

Crack extends if

U e

U e

rel
G e =
=

(2
)
(2
B
a
B
a)

Energy release rate

Usep

Separation energy (SE)

(energy per area)

B(2a )

= 2 = c

G e (a ) = c = 2

fracture criterion

E c
a

c =

fracture stress

Ge=

Irwin [1957]:

K I2
E

Milano_2012

Path-Independent Integrals
( xi )

is some (scalar, vector, tensor) field quantity being

steadily differentiable in domain B and divergence free

,i :=

= 0 in B
xi

dv = n da = 0

Gau theorem

,i

B0 = B BS

singularity S in B :

(.) = ? (.) + v (.) + > (.) + v (.) = 0

B0

? (.) = > (.)

B +

BS

and

v (.) = v (.)

B +

path independence

> n da = > n da
i

BS

Milano_2012

Seite 5

10

Energy Momentum Tensor

Eshelby [1965]: energy momentum tensor

Pij = w ij

w
u k ,i
uk , j

w
ui ( x j )

energy density
displacement field

with

Pij , j = 0

Material forces acting on singularities (defects) in the continuum,

e.g. dislocations, inclusions, ...
Fi = > Pij n j da
B

and
the J-vector

J i = > wni jk nk u j ,i ds

w=

 d

ij ij

=0

Milano_2012

11

J-Integral
The J-integral of Cherepanov [1967] and Rice [1986]
is the 1st component of the J-vector
J = > w dx2 ij n j ui ,1 ds

Conditions:

Equilibrium

ij ,i = 0

Small (linear) strains

ij = 12 ( ui , j + u j ,i )

Hyperelastic material

ij =

9
9
9
9
9

w
ij

time independent processes

no volume forces
homogeneous material
plane stress and strain fields, no dependence on x3
straight and stress free crack faces parallel to x1
Milano_2012

Seite 6

12

J as Energy Release Rate

U e
J = Ge =

Acrack vL

mode I

Ba

=
Acrack
2 Ba

C(T), SE(B)
M(T), DE(T)

mixed mode
J = G e = GIe + GIIe + GIIIe =

K I2 K II2 K III2
+
+
E E 2G

Milano_2012

13

Fracture Criteria
Brittle fracture:
predominantly elastic catastrophic failure
Mode I
stress Intensity factor

K I = K Ic

energy release rate

G I e = G Ic

J-integral

J = J Ic

K Ic2
E
K2
J Ic = G Ic = Ic
E

G Ic = ce =

E = E
1 2

Milano_2012

Seite 7

plane stress
plane strain

14

ASTM E 1823
Standard Terminology Relating to Fatigue and Fracture Testing
Crack extension, a an increase in crack size.
Crack-extension force, G the elastic energy per unit of new
separation area that is made available at the front of an ideal
crack in an elastic solid during a virtual increment of forward
crack extension.
Crack-tip plane strain a stress-strain field (near the crack tip)
that approaches plane strain to a degree required by an
empirical criterion.
Crack-tip plane stress a stress-strain field (near the crack tip)
that is not in plane strain.
Fracture toughness a generic term for measures of resistance to
extension of a crack.

Milano_2012

15

ASTM E 1823 ctd.

Plane-strain fracture toughness, KIc the crack-extension
resistance under conditions of crack-tip plane strain in Mode I for
slow rates of loading under predominantly linear-elastic
conditions and negligible plastic-zone adjustment. The stress
intensity factor, KIc, is measured using the operational procedure
(and satisfying all of the validity requirements) specified in Test
Method E 399, that provides for the measurement of crackextension resistance at the onset (2% or less) of crack extension
and provides operational definitions of crack-tip sharpness, onset
of crack-extension, and crack-tip plane strain.
Plane-strain fracture toughness, JIc the crack-extension
resistance under conditions of crack-tip plane strain in Mode I
with slow rates of loading and substantial plastic deformation.
The J-integral, JIc, is measured using the operational procedure
(and satisfying all of the validity requirements) specified in Test
Method E 1820, that provides for the measurement of crackextension resistance near the onset of stable crack extension.
Milano_2012

Seite 8

16

ASTM E 399
Standard Test Method for Linear-Elastic Plane-Strain Fracture
Toughness KIc of Metallic Materials
Characterizes the resistance of a material to fracture in a neutral environment in
the presence of a sharp crack under essentially linear-elastic stress and severe
tensile constraint, such that (1) the state of stress near the crack front
approaches tritensile plane strain, and (2) the crack-tip plastic zone is small
compared to the crack size, specimen thickness, and ligament ahead of the
crack;
Is believed to represent a lower limiting value of fracture toughness;
May be used to estimate the relation between failure stress and crack size for a
material in service wherein the conditions of high constraint described above
would be expected;
Only if the dimensions of the product are sufficient to provide specimens of the
size required for valid KIc determination.

Specimen size

K
W a 2.5 Ic
YS

YS 0.2 % offset yield strength

Milano_2012

17

ASTM E 399 ctd.

Specimen configurations
SE(B): Single-edge-notched and fatigue precracked beam loaded in
three-point bending; support span S = 4 W, thickness B = W/2,
W = width;
C(T): Compact specimen, single-edge-notched and fatigue precracked
plate loaded in tension; thickness B = W/2;
DC(T): Disk-shaped compact specimen, single-edge-notched and
fatigue precracked disc segment loaded in tension; thickness
B = W/2;
A(T): Arc-shaped tension specimen, single-edge-notched and fatigue
precracked ring segment loaded in tension; radius ratio unspecified;
A(B): Arc-shaped bend specimen, single-edge-notched and fatigue
precracked ring segment loaded in bending; radius ratio for
S/W = 4 and for S/W = 3;

Milano_2012

Seite 9

18

Specimen Configurations

Bend Type

SE(B)

C(T)
W
B
a
b = W-a

width
thickness
crack length
ligament width

Milano_2012

Seite 10

19

Inelastic Deformation and Damage

Part I: Plasticity

W. Brocks
Christian Albrecht University
Material Mechanics

Outline

Fundamentals of incremental plasticity

Finite plasticity (deformation theory)
Plasticity, damage, fracture
Porous Metal Plasticity (GTN Model)

Milano_2012

Seite 1

Plasticity

Inelastic deformation of metals at low temperatures and slow

(= quasistatic) loading, i. e. time and rate independent material
behaviour

Microscopic mechanisms: motion of dislocations, twinning.

Phenomenological theory on macro-scale in the framework of

continuum mechanics.

Material behaviour is non-linear and plastic (permanent)

deformations depend on loading history.
Constitutive equations are established incrementally for
small changes of loading and deformation

ij ij t , ij ij t
Incremental Theory of Plasticity
t > 0 is no physical time but a scalar loading parameter, and hence

ij and ij

are no velocities but rates

Milano_2012

Uniaxial Tensile Test

R0 : E

R0 :

linear elasticity: Hooke

true stress-strain curve

plasticity: nonlinear --curve

permanent strain

e p
yield condition

RF ( p ) , RF (0) R0
R0 yield strength (0, Y)
RF(p) uniaxial yield curve

loading / unloading

0 , p 0 loading
0 , p 0 unloading

Milano_2012

Seite 2

3D Generalisation

Additive decomposition of strain rates

Total plastic strains

ijp

ij ije ijp

p
ij

Plastic incompressibility:
kkp 0
plastic deformations are isochoric
plastic yielding is not affected by hydrostatic stress

ij ij h ij
h kk
1
3

Yield condition

deviatoric stress
hydrostatic stress

( ij , ijp ) ij ij ij ij 2 ( p ) 0

Hardening:
ij
kinematic: tensorial variable back stresses
isotropic: scalar variable accumulated plastic strain p
Milano_2012

Yield Surface

0
elastic
0
elastic
0 inadmissible
Milano_2012

Seite 3

Elasto-Plasticity

Hookes law of elasticity

ije

ij
kk ij

2G
1

ijp

Associated flow rule normality rule

Loading condition

Consistency condition

ij

ij 0 , ijp 0
ij

loading

ij 0,
ij

unloading

ijp 0

ij p ijp 0
ij
ij

ij ijp p 0

Equivalence of dissipation rates

Stability
Drucker [1964]
Milano_2012

von Mises Yield Condition

Yield condition - isotropic hardening

( ij , ijp ) 2 RF2 ( p ) 0
von Mises [1913, 1928]
equivalent stress

3 J 2 ( ij )

3
2

ij ij

J2-theory

1
2

2
11 22 2 22 33 2 33 11 2 3 122 23
132

Loading condition

Flow rule

ij ij 0 loading
ij ij 0 unloading

ijp
ij

plastic multiplier
from uniaxial test

Milano_2012

Seite 4

Prandtl-Reuss

ij ije ijp

Hookes law of elasticity + associated flow rule

p ij ijp
ij ij

Equivalence of dissipation rates

equivalent plastic strain

2
3

ijpijp d p

plastic multiplier

3 p 3 p

2 2 RF ( p )

Total strain rates (Prandtl [1924], Reu [1930])

ij ije ijp ije 13 kke ij ijp

1
1
3
ij
h ij
ij
2G
3K
2Tp RF

Milano_2012

Finite Plasticity
additive decomposition of total strains
plastic strains
total strains

ij ije ijp

Hencky [1924]

ijp ij
1
3
1

h ij
ij
2
G
2
S
3
K
p

ij

power law of Ramberg & Osgood [1945]

uniaxial

0 0
0
1 n

3D

ijp 3

0 2 0

ij
0

Milano_2012

Seite 5

10

Deformation Theory
ij (t ) (t ) ij0
For radial (proportional) loading,
the Hencky equations can be derived by integration of the
Prandtl-Reu equations.
This has do hold for every point of the continuum and excludes
stress redistribution,
unloading.
Finite plasticity actually describes a hyperelastic material having
t
a strain energy density
w ij ij d
0

so that

w
ij
ij

In elastic-plastic fracture mechanics (EPFM), finite plasticity +

Ramberg-Osgood Power law are adressed as
Deformation Theory of plasticity.
Milano_2012

11

Plasticity and Fracture

uniaxial tensile test

Milano_2012

Seite 6

12

Fracture of a Tensile Bar

Milano_2012

13

Fracture surface

round tensile specimen of Al 2024 T 351

Milano_2012

Seite 7

14

Deformation, Damage, Fracture

Deformation
Cohesion of matter is conserved
Elastic: atomic scale
Reversible change of atomic distances
Plastic: crystalline scale
Irreversible shift of atoms, dislocation movement

Damage
Laminar or volumetric discontinuities on the micro scale (microcracks, microvoids, micro-cavities)
Damage evolution is an irreversible process, whose
micromechanical causes are very similar to deformation
processes but whose macroscopic implications are much
different

Fracture
Laminar discontinuities on the macro scale leading to global
failure (cleavage fracture, ductile rupture)
Milano_2012

15

Damage Models

Damage models describe evolution of degradation phenomena

on the microscale from initial (undamaged or predamaged) state
up to creation of a crack on the mesoscale (material element)
Damage is described by means of internal variables in the
framework of continuum mechanics.
Phenomenological models
Change of macroscopically observable properties are
interpreted by means of the internal variable(s);
Concept of effective stress: Kachanov [1958, 1986], Lemaitre
& Chaboche [1992], Lemaitre [1992].
Micromechanical models
The mechanical behaviour of a representative volume element
(RVE) with defect(s) is studied;
Constitutive equations are formulated on a mesoscale by
homogenisation of local stresses and strains in the RVE.
Milano_2012

Seite 8

16

Ductile Damage

Nucleation, growth and coalescence of microvoids at inclusions or

second-phase particles
Void growth is strain controlled, and depends on hydrostatic stress

Milano_2012

17

Porous Metal Plasticity

Plastic Potential of Gurson, Tvergaard & Needleman (GTN model)

including scalar damage variable f*(f), (f = void volume fraction)

( ij , p , f * )

3
h
2
*2
2q1 f * cosh q2
1 q3 f 0
2

RF ( p )
2
R
(

)
F
p

SIII

Pores are assumed

to be present from the beginning, f0,

or nucleate as a function of plastic

equivalent strain, fn, n, sn
Evolution equation of damage

f 1 f kkp

SII

SI

Volume dilatation caused by void growth

p
kk

Milano_2012

Seite 9

18

Concepts of Fracture Mechanics

Part II: Small Scale Yielding

W. Brocks
Christian Albrecht University
Material Mechanics

Overview

I.

Linear elastic fracture mechanics (LEFM)

II.

Small Scale Yielding (SSY)

III. Elastic-plastic fracture mechanics (EPFM)

Milano_2012

Seite 1

SSY

Plastic zone at the crack tip

Effective crack length (Irwin)
Effective SIF
Dugdale model
Crack tip opening displacement (CTOD)
Standards: ASTM E 399, ASTM E 561

Milano_2012

Yielding in the Ligament (I)

Irwin [1964]: extension of LEFM to small plastic zones
Small Scale Yielding, SSY
K dominated stress field mode I
Yielding in the ligament
perfectly plastic material

=0
RF ( p ) = R0

3 = zz = 0

(a) plane stress

1 = xx = 2 = yy =

KI
2 r

Yield condition (both: von Mises and Tresca)

rp =

1
2

KI

R0

yy = R0 , 0 r rp

rp = radius of plastic zone

Milano_2012

Seite 2

Yielding in the Ligament (II)

(b) plane strain 3 = zz = 0

3 = zz = ( 1 + 2 ) = 2 yy =

2 K I
2 r

Yield condition (both: von Mises and Tresca)

(1 2 ) yy = R0 ,
rp

(1 2 )
=
2

0 r rp

KI

R0

smaller plastic zone due

to triaxiality of stress state

Cut-off of largest principle stress at R0 (plane stress)

rp

rp

yy (r ) dr =

KI
2
dr =
r = 2 R0 rp
p
2 r

equilibrium ?
Milano_2012

Effective Crack Length

aeff = a + rp

rp  a

effective SIF
a
K Ieff = K I ( aeff ) = aeff Y eff
W

effective J

J ssy = G ssy =

2
K Ieff
E

no singularity
at the crack tip

total diameter of plastic zone

KI
d p = 2rp =

2 R0

1
, =
2
(1 2 )

Milano_2012

Seite 3

plane stress
plane strain

Example SSY (I)

a
KI = a Y
W

F
BW

C(T)

a
Y
W

ASTM E 399

Milano_2012

Example SSY (II)

plane stress

dp
W

a
Y

2 W R0 W

plane strain

plane stress

plane strain

a
K Ieff = aeff Y eff
W
Milano_2012

Seite 4

CTOD (Irwin)

KI
E

u y ( r, ) = 4

4 K I2
ER0

plane stress
plane strain

t = 2u y ( rp , )

Wells [1961]

tIrwin =

2
1

r
2

1
plane stress

2
(1 ) (1 2 ) plane strain

(1 ) (1 2 ) 0.36
2

tplane strain 0,36 tplane stress

t = c

Criterion for crack initiation:

Milano_2012

Shape of Plastic Zone

ij (r , ) =

LEFM:

KI
fijI ( )
2 r

plane stress

33 =
( 11 + 22 ) plane strain
Yield condition (von Mises)
2
2
2
2 = 12 ( 11 22 ) + ( 22 33 ) + ( 33 11 ) + 3 ( 122 + 232 + 132 )

rp

= R0

d p ( ) =

1
2

KI

R0

1 + 23 sin 2 + cos
plane stress
3 2
2
2 sin + (1 2 ) (1 + cos ) plane strain

Milano_2012

Seite 5

10

Plastic Zone

plane stress
plane strain

plane strain
plane stress

3D: dog bone model

Hahn & Rosenfield [1965]

Milano_2012

11

Dugdale Model

strip
yield
model

2c = 2 a + 2d p

yy ( r,0) = R0 , 0 r d p
Superposition
no singularity

K I(1) = c , K I(2) =
K

(1)
I

+K

(2)
I

=0

R0 c arccos

a
= cos

c
2 R0

d p = c 1 cos
= a sec
1
2 R0

2 R0

R0  1:

a
c
no restriction
with respect to
plastic zone size!

Irwin
dp
c
1.23 a
= 1.23 d p
8 R0
R0
Milano_2012

Seite 6

plane stress!
12

CTOD (Dugdale)

u y ( x = a , y = 0) = 4

Crack opening profile

t = 2u y ( x = a, y = 0)

Definition of CTOD

tDugdale =

tIrwin =

R0
a ln sec

E
2 R0

8 R0
a ln sec

E
2 R0

no dependence
on geometry!

4a 2
ER0

for Griffith crack, plane stress

Milano_2012

13

Barenblatt Model
Idea:
Singularity at the crack tip is unphysical
Griffith [1920]: Energy approach
Irwin [1964]: Effective crack length
Dugdale [1960]: Strip yield model
Barenblatt [1959]: Cohesive zone

Stress distribution (x) is unknown and cannot be measured

Cohesive model: traction-separation law ()

Milano_2012

Seite 7

14

Energy Release Rate

U

rel
G =
= Ba

B
a

v
L

= G ssy =
Usep
Ba

a + rp KI2
K

=G e +G p

E
a E
2
Ieff

= c > ce = 2

Fracture criterion:

G = c

Cohesive model:

c = ( ) d

Separation energy

local criterion!

Milano_2012

15

ASTM E 399: Size Condition

Standard Test Method for Linear-Elastic Plane-Strain Fracture
Toughness KIc of Metallic Materials
Characterizes the resistance of a material to fracture in a neutral
environment in the presence of a sharp crack under essentially linearelastic stress and severe tensile constraint, such that
(1) the state of stress near the crack front approaches tritensile plane
strain, and
(2) the crack-tip plastic zone is small compared to the crack size,
specimen thickness, and ligament ahead of the crack;

Specimen size

dp =

(1 2 )

K
W a 2.5 Ic
YS

KI

YS

YS 0.2 % offset yield strength

Milano_2012

Seite 8

dp
W a

(1 2 )
2.5

B
0.02
(W a )
16

ASTM E 561
Standard Test Method for K-R Curve Determination
Determination of the resistance to fracture under Mode I loading
using M(T), C(T), or crack-line wedge-loaded C(W) specimen;
continuous record of toughness development in terms of KR plotted
against crack extension.
Materials are not limited by strength, thickness or toughness, so long as
specimens are of sufficient size to remain predominantly elastic.
Plot of crack extension resistance KR as a function of effective
crack extension ae.
Measurement of physical crack size by direct observation and then
calculating the effective crack size ae by adding the plastic zone
size;
Measurement of physical crack size by unloading compliance and then
calculating the effective crack size ae by adding the plastic zone size;
Measurement of the effective crack size ae directly by loading
compliance.
Milano_2012

17

FM Test Specimens (Tension Type)

M(T)

DE(T)

Milano_2012

Seite 9

18

Concepts of Fracture Mechanics

Part III: Elastic-Plastic Fracture

W. Brocks
Christian Albrecht University
Material Mechanics

Overview

I.

Linear elastic fracture mechanics (LEFM)

II.

Small Scale Yielding

III. Elastic-plastic fracture mechanics

(EPFM)

Milano_2012

Seite 1

EPFM

Deformation theory of plasticity

J as energy release rate
HRR field, CTOD
Deformation vs incremental theory
of plasticity
R curves
Energy dissipation rate

Milano_2012

EPFM
Analytical solutions and analyses in

Elastic-Plastic Fracture Mechanics,

i.e. fracture under large scale yielding (LSY)conditions
are based on Deformation Theory of Plasticity
which actually describes hyperelastic materials

ij =

w
ij

w=

 d

ij ij

=0

 d

p
ij ij

=0

in the following, the superscripts

e stands for linear elastic

p stands for nonlinear

U = U e + U p = F dvLe + F dvLp

J = Je + Jp =

Milano_2012

Seite 2

K I2 U p

E Acrack v
L

J as Stress Intensity Factor

Power law of Ramberg & Osgood [1945]

uniaxial

e p
= +
=
+
0 0 0 0
0

3D

ijp 3
=
0 2 0

1 n

ij
0

Hutchinson [1868],
Rice & Rosengren [1968]
singular stress and strain fields at the crack tip (HRR field) mode I

ij = K r

1
n +1

 ij ( )
n

K n
ij = 0 r n+1ij ( )
0
p
ij

n+1
J
K = 0

0 0 I n

Milano_2012

ij ijp = O ( r 1 )
5

HRR Angular Functions

 xx

 yy

Milano_2012

Seite 3

CTOD
n

HRR displacement field

n+1 n1+1
J
ui = 0
r ui ( )
0 0 I n

Crack Tip Opening Displacement, t , Shih [1981]

t = 2u y (rt , ) , r u x (rt , ) = u y ( rt , )

t = dn

0
1

d n = ( 0 ) n Dn

Milano_2012

J = J el + J pl

Path Dependence of J

FE simulation
C(T) specimen
plane strain
stationary crack
incremental theory
of plasticity
ASTM E 1820: reference value
far field value
J =J e + J p

Je =

K I2
E

KI = a Y ( a W )
Jp =

Milano_2012

Seite 4

U p

B (W a )

Stresses at Crack Tip

incremental theory of plasticity

large strain analysis

HRR

ij r

0 J 0

1
n +1

r

t

Milano_2012

1
n +1

R-Curves in FM
Different from quasi-brittle fracture, ductile crack extension is
deformation controlled: R-curves J(a), (a)
R curve a plot of crack-extension resistance as a function of
stable crack extension (ASTM E 1820)
JR-curve: J(a)
measure F, VL, a

J ( a( i ) ) = J e + J (pi ) =

U (pi )
a ( i )
K I2 p
+ J ( i 1) + ( i 1)
1 ( i 1)

E
b( i 1) B
b( i 1)
recursion formula
Milano_2012

Seite 5

10

ASTM E 1823
Standard Terminology Relating to Fatigue and Fracture Testing
Crack extension, a an increase in crack size.
Crack-extension resistance, KR, GR of JR a measure of the resistance
of a material to crack extension expressed in terms of the stressintensity factor, K; crack-extension force, G ; or values of J derived
using the J-integral concept.
Crack-tip opening displacement (CTOD), the crack displacement
resulting from the total deformation (elastic plus plastic) at variously
defined locations near the original (prior to force application) crack tip.
J-R curve a plot of resistance to stable crack extension, ap.
R curve a plot of crack-extension resistance as a function of stable
crack extension, ap or ae.
Stable crack extension a displacement-controlled crack extension
beyond the stretch-zone width. The extension stops when the applied
displacement is held constant.

Milano_2012

11

ASTM E 1820
Standard Test Method for Measurement of Fracture Toughness
Determination of fracture toughness of metallic materials using the parameters K,
J, and CTOD ().
Assuming the existence of a preexisting, sharp, fatigue crack, the material fracture
toughness values identified by this test method characterize its resistance to
(1) Fracture of a stationary crack
(2) Fracture after some stable tearing
(3) Stable tearing onset
(4) Sustained stable tearing
This test method is particularly useful when the material response cannot be
anticipated before the test.
Serve as a basis for material comparison, selection and quality assurance;
rank materials within a similar yield strength range;
Serve as a basis for structural flaw tolerance assessment; awareness of
differences that may exist between laboratory test and field conditions is
required.

Milano_2012

Seite 6

12

ASTM E 1820 (ctd.)

Cautionary statements
Fracture after some stable tearing is sensitive to material
inhomogeneity and to constraint variations that may be induced
to planar geometry, thickness differences, mode of loading, and
structural details;
J-R curve from bend-type specimens, SE(B), C(T), DC(T), has
been observed to be conservative with respect to results from
tensile loading configurations;
The values of c, u, Jc, and Ju, may be affected by specimen
dimensions

Milano_2012

13

CTOD R-Curve
Schwalbe [1995]: 5(a)
ASTM E 2472
particularly for thin panels

Milano_2012

Seite 7

14

ASTM E 2472
Standard Test Method for Determination of Resistance to
Stable Crack Extension under Low-Constraint Conditions

Determination of the resistance against stable crack extension

of metallic materials under Mode I loading in terms of critical
crack-tip opening angle (CTOA) and/or crack opening
displacement (COD) as 5 resistance curve.
Materials are not limited by strength, thickness or toughness, as
long as and , ensuring low constraint conditions in M(T) and
C(T) specimens.

Milano_2012

15

Limitations
For extending crack
J becomes significantly path dependent
J looses its property of being an energy release rate
J is a cumulated quantity of global dissipation
JR curves are geometry dependent

tension

BAM Berlin
bending

Milano_2012

Seite 8

16

Dissipation Rate
Wex

=
U e + U p + Usep
Ba Ba

Energy balance
Dissipation rate
Turner (1990)

R=

Udiss U p Usep
=
+
= Rp + c
Ba
Ba Ba

Rp =

U p
Ba

c =
commonly:

Rp  c

global plastic dissipation rate

Usep

local separation rate

Ba

geometry dependence of JR curves

dJ p
(W a ) da

R ( a ) =
p
(W a ) dJ + J p

da

M(T), DE(T)
C(T), SE(B)

Milano_2012

17

R(a)

accumulated
plastic work

stationary
value

Milano_2012

Seite 9

18

Inelastic Deformation and Damage

Part II: Damage Mechanics

W. Brocks
Christian Albrecht University
Material Mechanics

Outline

Deformation, Damage and Fracture

Crack Tip and Process Zone
Damage Models
Micromechanisms of Ductile Fracture
Micromechanical Models
Porous Metal Plasticity (GTN Model)

Milano_2012

Seite 1

Damage - Definition

l'endommagement, comme le diable, invisible mais redoutable

Surface or volume-like discontinuities on the materials microlevel (microcracks, microvoids)

Damage evolution is irreversible (dissipation!)
Damage causes degradation (reduction of performance)
Examples for processes involving damage phenomena:
ductile damage in metals, creep damage, fiber cracking or
fiber-matrix delamination in reinforced composites, corrosion,
fatigue

Milano_2012

Observable Effects

Physical Appearance of Damage

volume defects (microvoids, microcavities)

surface defects (microcracks)

Mascroscopic Effects of Damage

decreases elasticity modulus

decreases yield stress

decreases hardness

increases creep strain rate

decreases sound-propagation velocity

decreases density

increases electrical resistance

Milano_2012

Seite 2

Damage Models

Damage models describe evolution of degradation phenomena

on the microscale from initial (undamaged or predamaged) state
up to creation of a crack on the mesoscale (material element)
Damage is described by means of internal variables in the
framework of continuum mechanics.
Phenomenological models
Change of macroscopically observable properties are
interpreted by means of the internal variable(s);
Concept of effective stress: Kachanov [1958, 1986], Lemaitre
& Chaboche [1992], Lemaitre [1992].
Micromechanical models
The mechanical behaviour of a representative volume element
(RVE) with defect(s) is studied;
Constitutive equations are formulated on a mesoscale by
homogenisation of local stresses and strains in the RVE.
Milano_2012

Continuum Damage Mechanics (CDM)

J. Lemaitre, R. Desmorat
Engineering Damage Mechanics
Springer, 2005

Kachanov [1958]
Hult [1972]
Lemaitre [1971]
Lemaitre & Chaboche [1976]

D. Krajcinovic
Damage Mechanics
Elsevier, 1996

A

Milano_2012

Seite 3

Effective Area
DV =

Volume density of microvoids

Vvoids
= fV
VRVE

Surface density of microcracks or intersections

of microvoids with plane of normal n

D(n) =

Acracks
ARVE

A = A AD

"Effective" area

Isotropic Damage Scalar Damage Variable

if D(n) does not depend on n

D=

AD
A

A = (1 D ) A

Milano_2012

Anisotropic Damage
Tensorial Damage Variables
n A = (1 D ) n A

rank 2 tensor D

( mn ) A = ( I D) ( mn ) A

rank 4 tensor D

Dijkl = Dijlk = D jikl = Dklij , most general case

with symmetries

metric tensor (mn) defines the reference configuration

n

unchanged

n

A

Milano_2012

Seite 4

Effective Stress (I)

Isotropic Damage

1D

 =

3D

S
S =
1 D

1 D
or ij =

ij
1 D

Anisotropic Damage
rank 4 tensor D
m S n A = m S n A

projection of stress vector on m

S mn A = S mn A = S ( I D ) mn A
1
1
S = S ( I D)
or ij = (ik jl Dijkl ) kl

Milano_2012

Effective Stress (II)

Rank 2 tensor D requires additional conditions:

Symmetry of the effective stress:
1
S = S (1 D )

is not symmetric!

Compatibility with the thermodynamics framework: existence of

strain potentials and principle of strain equivalence,
Symmetrisation (not derived from a potential)
1
1
S = 12 S (1 D ) + (1 D ) S

Different effect of the damage on the hydrostatic and deviatoric

stress

S = ( H S H) +

h
1 2
1 with H = (1 D )
1 Dh

Milano_2012

Seite 5

10

Example: Rank 2 Tensor

Concept of effective stress (III)

Transversal isotropic damage (2=3):
Uniaxial loading 1,

D1 0
D= 0 D2

0 0

D1
1

0 ei e j ; H = 0

D2

1 =

4 1
2 1
1 1
+
+
9 1 D1 9 1 D2 3 Dh

 =

2 1
1 1
+
1
3 1 D1 31 D2

0
1
1 D2
0

0 ei e j

1
1 D2
0

effective von Mises stress

11

Milano_2012

11

Thermodynamics of Damage

1. Definition of state variables, the actual value of each defining

the present state of the corresponding mechanism involved
2. Definition of a state potential from which derive the state laws
such as thermo-elasticity and the definition of the variables
associated with the internal state variables
3. Definition of a dissipation potential from which derive the laws
of evolution of the state variables associated with the dissipative
mechanism

Check 2nd Principle of Thermodynamics !

Milano_2012

Seite 6

12

Variables

Mechanism

State variable
observable

Thermoelasticity

Temperature/Entropy

conjugate
variable

internal
S
s
p

Plasticity

-S

Isotropic hardening

Kinematic hardening

Damage isotropic

D
D

-Y

Damage anisotropic

-Y
p p , R

Milano_2012

13

State Potential
Helmholtz specific free energy

( Ee , D or D, p, A, ) = e + p +
Gibbs specific free enthalpy taken as state potential
1

* = sup SE
E

1
1
= sup SEe e + SEp p
e
E

State laws of thermoelasticity can be deducted

E=
s=

*
*
= e + Ep = Ee + Ep
S
S

Milano_2012

Seite 7

14

Dissipation Potential

Definition of conjugate variables

R =

*
p

*
A
*
Y =
D
X =

or Y =

*
D

2nd Principle of Thermodynamics (Clausius-Duhem inequality)

 ) + Y D q grad 0
 p ( R p + X A
S E

Evolution equations for internal variables (kinetic laws) are derived

from a dissipation potential , which is a convex function of the
conjugate variables

( S, R, X,Y or Y, )
Milano_2012

15

Normality Rule

Normality rule of generalised standard materials

 p =  = 
E
( S )
S

p = 
R

 = 
A
X

= 
D = 
( Y )
Y

flow rule

= 
or D = 
( Y )
Y

Nice and consistent theoretical framework but

wherefrom to get the dissipation potential ?

Milano_2012

Seite 8

16

Principle of Strain Equivalence

Strain constitutive equations of a damage material are derived
from the same formalism as for a non-damaged material except
that the stress is replaced by the effective stress
Example: State potential for linear isotropic elasticity

e* =

(1 + ) ij ij kk2

2E (1 D ) 2E (1 D )

Elastic strain

ije =

e* 1 +

ij  kkij
=
E
E
ij

Energy density release rate Y

2
*  2
h
Y = e =
(1 + ) + 3(1 2 )
D 2E 3

h = 13 kk
=

2
3

ij ij

Milano_2012

17

Local and Micromechanical Approaches

Cleavage (brittle fracture)
Microcrack formation and coalescence
Stress controlled
Ritchie, Knott & Rice [1973]: RKR model, Beremin [1983]
Ductile tearing
Nucleation, growth and coalescence of microvoids at inclusions or
second-phase particles
Strain controlled, void growth dependent on hydrostatic stress
Rice & Tracey [1973], Gurson [1977], Beremin [1983], Tvergaard &
Needleman [1982, 1984, ...], Thomason [1985, 1990], Rousselier
[1987]
Creep damage
Nucleation, growth and coalescence of micropores at grain boundaries
Stress or strain controlled
Hutchinson [1983], Rodin & Parks [1988], Sester & Riedel [1995]
Milano_2012

Seite 9

18

Cleavage

Mechanisms of microcrack initiation

Broberg [1999]

Coalescence of microcracks

Milano_2012

19

Ductile Fracture
fracture surface of Al 2024

Failure mechanisms: Nucleation, growth and coalescence of voids

Voids nucleate at secondary phase particles due to particle/matrix
debonding and/or particle fracture
Localisation of plastic deformation is prior to failure
Milano_2012

Seite 10

20

Void Nucleation

Void nucleation at coarse particles in Al 2024 T 351

Milano_2012

21

Ductile Crack Extension (I)

Ductile crack extension
in an Al alloy

Schematic view of process

zone with unit cells
Broberg [1999]
Milano_2012

Seite 11

22

Ductile Crack Extension (I)

Milano_2012

23

Models of Void Growth (I)

McClintock [1968]
damage

zx = d zx =

d [ln ( rx A x )]
ln ( A 0x rx0 )

coalescence

2rx = A x

power law

= n

void growth

d zx
1
=

0
d
ln A 0x r x

fracture strain

3 (1 n ) ( xx + yy ) 3 ( xx yy )
3

+
sinh

2
2 (1 n )

f =
sinh

(1 n ) ln ( A0x rx0 )
(1 n ) ( xx + yy ) ( 2

Milano_2012

Seite 12

))

24

Models of Void Growth (II)

Rice & Tracey [1969]

r
2
D =
= 0.283exp h
r
3

h
=T

triaxiality

void-volume fraction for tensile test

Milano_2012

25

Void Nucleation

Particle cracking
Particle-matrix debonding
in Al-TiAl MMC
Milano_2012

Seite 13

26

Representative Volume Element (Unit Cell)

In-situ observation and FE

simulation of void
nucleation by particle
decohesion and fracture

Milano_2012

27

Unit Cell Simulations

Debonding of matrix at a particle

Cracking of particle

Milano_2012

Seite 14

28

Representative Volume Element (Unit Cell)

0.10
f

0.30

T=2

necking
Volume fraction

-2 E1

0.20
0.06
0.15
fc

0.10

0.02

0.00
0.0

0.05

0.1

0.2

0.00
0.3

Ev

Evolution of void volume fraction can be computed from

simple geometrical RVEs
Critical volume fractions can be obtained from plastic collapse
of the cell (function of triaxiality!)
Procedure can be applied independently of the aggregate
(void, particle, evolving object)

Milano_2012

29

Mesoscopic Response
FE simulation of void growth:
Mesoscopic stress-strain curves

Milano_2012

Seite 15

30

Coupled / Uncoupled Models

Uncoupled models: Example

Rice & Tracey [1969]

round tensile bar:

dr
3
= 0.283 d p exp T
r
2
35

notched bar
30
(2)

25

F [kN]

20
15
10
von Mises
coupled model

5
(1)

0
0,0

0,2

GTN

0,4

0,6

0,8

1,0

u2/2 [mm]
31

Milano_2012

31

Porous Metal Plasticity

Additional scalar internal variable in the yield potential, which is a
function of porosity f

( ij , p ) = 0

( ij , p , f ) = 0

Porosity equals the void volume fraction in an RVE:

f =

Vvoids
VRVE

Yield potential formulation is obtained from homogenisation

ij =
ij =

1
VRVE

1
VRVE

ij

dV =

VRVE

ij

dV =

1
ij n j dS
( VRVE ) (
VRVE )

1
2 VRVE

(u

i, j

+ u j ,i ) dV

mesoscopic
stresses and
strains

Evolution equation of void growth is derived from plastic

incompressibility of matrix
volume dilatation
 kkp 0
f = (1 f )  p
due to void growth
kk

32

Milano_2012

Seite 16

32

Gurson and Rousselier Model

Gurson [1977], Tvergaard & Needleman [1984]

3
2

+ 2q1 f * cosh q2 h 1 q3 f *2 = 0

R( p )
2 R( p )

damage variable f*(f)

p = Ep =

2
3

ijp ijp

Rousselier [1987]

1
D f exp
1 = 0
(1 f ) R( p ) R( p )
(1 f ) 1

33

Milano_2012

33

Comparison

1.0

R0
Gurson

0.8

0.6

Rousselier
0.4

0.2

0.0
0.5

1.0

1.5

2.0

2.5

h R0
34

Milano_2012

Seite 17

34

Extensions
Tvergaard & Needleman
damage function
f
for

f* =
for
f

f
+

f
(
c)
c

f fc
f fc

Chu & Needleman [1980]

void nucleation
f = fgrowth + fnucl = (1 f ) kkp + Anp

1 p n 2
fn

An =
exp
2 sn
sn 2

35

Milano_2012

35

Effect of Triaxiality

f0 = 0
fc = 0.12
fn = 0.05

n = 0.05
sn = 0.15
q1 = 1.5
q2 = 1.0
q3 = q12 = 2.25

36

Milano_2012

Seite 18

36

Tensile Test: GTN model

Simulation of deformation and damage in a round tensile bar

Milano_2012

37

SE(B): GTN model

Milano_2012

Seite 19

38

Punch Test

600
1

Experiment

F [kN]

Simulation
400

punch force
x-axis

2
200

1
axial force

0
0

50

100

150
u3 [mm]

Simulation with the GTN model

39

Milano_2012

39

Punch Test

Milano_2012

Seite 20

40

Summary (I)
Ductile crack extension and fracture can be modelled on various
length scales:
(1) Micromechanics: void nucleation, growth and coalescence
(2) Continuum mechanics: constitutive equations with damage
(3) Cohesive surfaces: traction-separation law
(4) Elastic-plastic FM: R-curves for J or CTOD
The models require determination of respective parameters:
(1) Microstructural characteristics: volume fraction, shape, distance of
particles, ...
(2) Initiation: f0, fn, n, sn, coalescence: fc, final fracture: ff, ....
(3) Shape of TSL, cohesive strength c , separation energy c
(4) J(a) or (a)

Milano_2012

41

Summary (II)

The models have specific favourable and preferential applications:

(1) Effects of nucleation mechanism, stress triaxiality, void /
particle shape, void / particle spacing, ...
(2) Constraint effects, inhomogeneous materials, damage
evolution, ...
(3) Large crack growth, residual strength of structures
(4) Standard FM assessment of engineering structures

Acknowledgement:
FE simulations by Dr. Dirk Steglich,
Helmholtzzentrum Geesthacht

Milano_2012

Seite 21

42