Fracture mechanics:

Validity and use

Concepts and Parameters
• LEFM = Linear-elastic fracture mechanics
• EPFM = Elastic-plastic fracture mechanics

• K = Stress intensity factor (LEFM)

• G = Energy release rate (LEFM)
• CTOD = Crack tip opening displacement (EPFM)
• J-contour integral = Crack tip condition under loading (EPFM)

7.10.2016
2
http://en.wikipedia.org/wiki/SS_Schenectad
y
 Fracture Mechanics I

Transition behavior and brittle

fracture
 Food for thought:
• Why are cracks dangerous?
• Why a cracked component does not
crack with a relatively small load?
• What is G? Unit?
• What is K? Unit?
• How is fracture mechanics used?
Fracture mechanics: When is it
valid and how should we apply
it?
 Fracture mechanics can be
used:
• When brittle fracture is the limiting factor:
– For materials that are susceptible to brittle
fracture
• Analysis of a cracked component
– Assumed or observed crack
 FM can give answer to:
• What is the largest allowable stress for a
cracked component?
• What is the largest allowable crack size
for a given stress level?
• Is the material usable with a given crack
size and stress level?
 The analysis
• Stress analysis for an intact component
• Add K-solution and calculate K-value for
the stress state
• Compare the K-value for materials critical
K-value
 Limitations
• KIC can only be determined to brittle
materials
• KIC determination is expensive
Elastic strain energy from a
crack
Factors impacting fracture
 Limitations
• K is only valid, when
– Plasticity is restricted to a small area ahead
of the crack
– Components in plane strain condition
– Materials exhibit brittle behavior
 Validation
• Experimental requirement from standard:
æ K I ö2
a,B,(W - a) ³ 2.5ç ÷
è s YS ø
• ry < specimen size / 50
– Plane strain condition
• LEFM is valid, as long as plane dimensions are
large compared to the area dominated by
singularity
– Experimental work should be conducted with full
component thickness
 For more ductile materials:
• Material in front of the crack deforms
plastically
• Crack tip blunts
• Crack may grow with a ductile
mechanism for a while before initiating a
brittle fracture
– > EPFM, but we’ll get to this later…
Transition temperature in steels
 Transition temperature in
steels
• Tendency for brittle fracture increases
with decreasing temperature
• High temperatures -> no brittle fracture
• Simple criterion => avoid using steels in
temperatures in which brittle fracture may
occur
 Design
• Avoid brittle fracture
– Conservative
– The price: leave out materials which could
work
 Ferritic steels
• Brittle cleavage fracture in low
temperatures
– Design according to FM parameters
• Ductile fracture in high temperatures
– Design according to strength
 Transition temperature
• Above TT: Avoid determining expensive
KIc
• TT is determined by performing tests at
various temperatures
 Charpy V-Notch Impact
Fracture Test
• Notched, not cracked specimen
• Impact loading
• Output: Energy consumed by fracture
 Applications
• QC and classification for welds
• QC after heat treatments
• Studying embrittlement phenomena
• Transition temperature for ferritic steels
Charpy V-Notch Test
CVN-Test

26
CVN-Test
 TT Criterion
• TT Criterion
– Impact energy 27J (sometimes 40J)
– Ductile fracture 50% of fracture surface
– "lateral expansion"
 Trantisiokäyrä

140

120

100
Iskuenergia (J)

80

60

40

20

0
-120 -100 -80 -60 -40 -20 0 20 40 60
Lämpötila (oC)
 Challenges of CVN
• Numerous empirical criteria for various
purposes
• No clear technical foundation
• Sometimes unnecessary limit the use of
materials
Brittle Fracture (cleavage fracture)
 Nucleation of a brittle fracture
• Statistic nature
• Sample size effect
• Size of singularity
• Ductile fracture before brittle fracture
 Crack
nucleator

A508 Cl.3
T = - 30°C
KJC = 280 MPam
Da = 0.1 mm
 s s s

KW952D

s s s
Local stress produces a Cracking of the carbide Advancing microcrack
dislocation pile-up which introduces a microcrack encounters the first large
impinges on a grain which propagates into angle boundary.
boundary carbide. the matrix.
Further analysis on CVN data
 Some questions
• Why does cleavage fracture
occur in ferritic steels?
• Why does the tendency for
cleavage fracture increase with
decreasing temperature?

41
Propagation of a brittle fracture
on an atomic scale
• In a continuum model, the crack tip
experiences an infinite stress
concentration
• In an atomic model, the radius of the
crack tip does not reach zero
• Upon loading a crack, the atoms at the crack tip move
to a maximum Peierls potential after which:
– A wild dislocation may appear OR
– Crack may propagate
• Depends on the surface energy
• If dislocation requires more energy than crack
propagation then fracture advances
• A simplified solution, in reality other dislocation sources
as well, but the basic principle is the same
 Transition Temperature
• All that decreases surface energy or
increase the energy required by
dislocations, increase the tendency for
brittle fracture
• Decrease in temperature and increase in
loading speed for BCC materials
Rail Fracture - Revisited
 Background
• Rail in use for 30 years
• Fracture nucleated from a poorly finished hole, from which
a 0.9 mm hole was found
• Low impact energy (<10J) low KIC were measured for the
rail
• Measured values where similar to a new, unused rail
• The base of the rail was suspected to have softened thus
increasing the stress levels in the rail
• Large residual stresses were also measured from the rail
• Low temperature (-7 C) induced a tensile stress (Normal
temperature app. +15°C)
 Conclusions
• Direct cause for the accident was the failure of the rail underneath
a train car
• Rail failure was probably due to a combination of the illfinished
hole and local properties of the rail steel. Wearing of the rail base,
residual stresses and tensile stresses caused by the low
temperature may also have contributed to the failure.
• Rail failure lead to this accident because it was not addressed
correctly. Two drivers reported a large noise but traffic control
could not identify this as a serious problem.
Fracture Mechanics II
LEFM => EPFM
LEFM Limitations
K is valid when:
• Plasticity is limited to a relatively small area ahead of crack tip
• Measurements in plane stress conditions
• For brittle materials
Validation
æ K I ö2
Experimental requirement from standard: a,B,(W - a) ³ 2.5ç ÷
è s YS ø
ry < specimen size / 50
- Plane stress condition
LEFM is valid, as long as plane dimensions are large compared
to the area dominated by singularity
• Experimental work should be conducted with full
component thickness
For more ductile materials:
• Plasticity at the crack tip
• Blunting of the crack tip
• Crack growth with ductile mechanism before the nucleation
of brittle fracture
Effect of crack tip blunting
Size of a plastic area ~ distance where stress equals to yield stress

æ
1 K ö 2

rp = ç I ÷
p è s YS ø
Irwin’s Plasticity Correction
Differs form the correct one because it is based on elasticity
• Yielding rearranges stresses in order to maintain balance
• Correcting the aforementioned solution with a balance equation:

æ
1 KI ö 2

ry = ç ÷
2p è s YS ø
Irwins correction: crack behaves like the crack tip is in the middle of plastic zone.
Keff depends on K => iterative solution

aeff = a + ry
K eff = Y(aeff )s paeff
Elastic-plastic Fracture
Mechanics
(EPFM)
Crack-Tip-Opening Displacement
Wells - KIc values cannot be measured for ductile materials.
More ductility, more crack tip blunting
Blunting describes the ductility of the material
CTOD
CTOD

Connected to K in case of LEFM

2
4 KI
d= Plastic zone
p s YS E
Experimental determination of CTOD

KI
2
rp (W - a)Vp
d = del + d pl = +
ms YS E' rp (W - a) + a
J-integral
Non-linear G
dP
J =- =U -F
dA
Path independent integral
• Does not consider internal stresses or æ ¶u ö
• Temperature gradients J = ò ç wdy - Ti i ds÷
G
è ¶x ø
Path Independency
J is a stress potential integral æ ¶ui ö
Ti is the stress normal to the road line J = ò çè wdy - Ti ¶x ds÷ø
G
J-integral over a random, closed loop
is equal to 0.
Around crack tip J≠0
Path independency
Closed loop J1+J2+J3+J4 = 0
J2 = J4 = 0
Ti = 0 on a free surface
J1 = -J3 for a random J3
=> J is path independent
J describes crack properties
J-integral has been proven to be:

dP
J=-
dA
J-integral
Expanding LEFM by idealization of elastic-plastic to nonlinear-
elastic
G and J
In linear-elastic region: G = J
J is valid in the nonlinear-elastic region
In an increasing, monotonic loading, nonlinear-elastic and
elastic-plastic behavior do not differ from one another
J describes material’s loading condition outside G’s validity
region.
Energy release rate and driving force
"Energy release rate" should be interpreted as the change in
energy state in a small crack increase increment.
J-integral as a stress parameter
J unanimously describes the stress state at the crack tip.
Description of crack tip
J describes the conditions of the plastic zone

• LE =>
1
r
• EP => 1
-
r n +1
Singularities
In LEFM region, stress singularity applies. It describes the
condition at the crack tip when plasticity is minimal and it is
limited to a small region ahead of crack tip.
With increasing load, HRR singularity applies. It describes crack
tip condition when plasticity is limited to the crack tip.
 Notch Crack

s yy s yy

s KI 

x x
J vs CTOD

J = sYSd
K-J
K and J are connected in LE region
In EP region, an equivalent K value can be estimated
J can be measured (small specimen) and K can be calculated
=> KJc
Similitude
One parameter is enough
A single FM parameter (G, K, J, CTOD) is sufficient alone to
describe the requirements for fracture (in their own validity
regions)
This applies regardless of loading, crack size, geometry,
material, environment, etc.
-> "Similitude concept"
LEFM vai EPFM
Ductile Fracture
Ductile Fracture
Void nucleation
Void growth
Void coalescence
J-R -curve
Final fracture does no always occur at a specific J-value
Jc increases as the crack propagates
J-R-curves for ductile metals

450
Mode I, N=3.33
400
Mode I, N=6.67
350 Mode I, N=20
Mode II, N=3.33

J-integral [kN/m]
300
Mode II, N=6.67
250
Mode II, N=20
200
150
100
50
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Da [mm]
Void Sheets

Dimples
 560 650
TL TL

560 LT 650 LT

Splitting
 Initiation/
propagation Splitting
energy
Quasi- effect
cleavage,
cleavage

Propagation
region Ductile
regions

Initiation
region
Mechanism (schematic)

Mode I Brittle Fracture:

• Weakest Link
• Principal and/or tensile
stresses

Mode I Ductile Fracture:

• Plastic and/or
equivalent strain/s
• Nucleation, growth and
coalescence
Step-by-step

al986.dsf
Blunting
KIC and Ductile fracture
For some materials, ductile fracture requires only little energy
and valid KIC values may be obtained
KIC derivation is mechanism independent

7.10.2016
100
Aluminum
Not susceptible to cleavage
Typically ductile fracture behavior
Fracture resistance for ductile fracture is low
Measured KIC values are low ≈ 25 MPa·m½

Fracture control:
• Aircraft
• Rockets

7.10.2016
101
Mechanism vs. Behavior

Mechanism Ductile Fracture Brittle Fracture

Cleavage X

Grain boundary X

Ductile X X

7.10.2016
102