You are on page 1of 96

MAAE 4012

Strength & Fracture Analysis


Chapter 4
Fatigue
Professor R. Bell
Department of Mechanical & Aerospace Engineering
Carleton University
2013
Chapter 4 - Fatigue

Department of Mechanical
& Aerospace Engineering

Crack Growth and Fracture


Crack Growth takes place by:
Fatigue due to cyclic loading
Stress Corrosion due to sustained loading
C
Creep
- constant
t t lloading
di att high
hi h ttemperature
t
Hydrogen induced cracking - Delayed hydride
cracking
Liquid metal induced cracking. (of little interest in load
bearing structures. e.g. Hg in contact with Al)

Items 1 and 2 are of most interest in general design

Chapter 4 - Fatigue

Definitions

Department of Mechanical
& Aerospace Engineering

Fatigue
F
Fatigue
ti
is
i th
the mostt important
i
t t crackk growth
th mechanism
h i
Fatigue is a process of crack growth which causes damage
to a component subjected to repeated loading. This damage
may result in final failure
failure.
Fatigue is a complex metallurgical process which is difficult
to accurately describe
Assessment for fatigue in design must be made
Fatigue can be divided into two periods
Initiation
Propagation
Total Life (NT) = Initiation Life (Ni) + Propagation Life (Np)

Chapter 4 - Fatigue

Fatigue Design Criteria

Department of Mechanical
& Aerospace Engineering

Infinite Life design


Safe Life Design
Fail - Safe Design
Damage Tolerant Design

Chapter 4 - Introduction

Fatigue Design Criteria

Department of Mechanical
& Aerospace Engineering

Infinite Life Design


Unlimited safety - requires design stresses to be safely
below the fatigue limit
Safe Life Design
Designing for a finite life - Failure will not occur during
the life of the component.
Safe life design must include a margin for the scatter of
fatigue results.
Margin of safety = 20 x desired life
e.g. pressure vessels and jet engines
Landing gear are designed for safe life

Chapter 4 - Introduction

Fatigue Design Criteria

Department of Mechanical
& Aerospace Engineering

Fail
F il - Safe
S f Design
D i
designed by aircraft engineers who could not tolerate
extra weight required by large safety factors nor the
danger resulting from small safety factors
Unlimited safety - requires design stresses to be safely
below the fatigue limit
multiple load paths and crack arresters
Aero engines
g
are fail - safe designs
g only
y in multi-engine
g
aircraft

Chapter 4 - Introduction

Fatigue Design Criteria

Department of Mechanical
& Aerospace Engineering

Damage
D
T
Tolerant
l
Design
D i
Refinement of fail - safe philosophy
Assumes that cracks exist and uses a Fracture
Mechanics analysis to check whether such cracks will
grow large enough (critical size) to produce failures
before they are sure to be detected and repaired
In pressure vessels "leak before burst" is an example of
this philosophy
In aircraft multiple load paths, crack arresters etc.

Chapter 4 - Introduction

Damage Tolerant Design

Department of Mechanical
& Aerospace Engineering

Philosophy Based on:


Damage will occur despite all precautions taken
Inspection so that cracks can be detected and repairs
made at proper time
Adequate residual strength is maintained in the
damaged structure so that during the period between
inspections when the damage is undetected ultimate
failure of the structure is not possible.

Chapter 4 - Introduction

Department of Mechanical
& Aerospace Engineering

Fatigue Design Methods

Stress Life Approach. Total Life


Strain Life Approach. - Initiation Life
Fracture Mechanics Approach Propagation Life

Chapter 4 - Fatigue

Stress Life Approach

Department of Mechanical
& Aerospace Engineering

Stress Life Approach


First fatigue method to be developed
Used mainly for long life applications were stresses and
strains are elastic
It does not distinguish between initiation and propagation, but
with the total life of a component. It deals with the life to
failure of a component.

Chapter 4 - Fatigue

10

Strain Life Approach

Department of Mechanical
& Aerospace Engineering

Strain Life Approach


Developed in the 1960's . This method is considered
an initiation approach.
It is used when the strain is no longer totally elastic, but
has a plastic component.
Short low cycle fatigue lives generally occur under
these conditions.

Chapter 4 - Fatigue

11

Department of Mechanical
& Aerospace Engineering

Fracture Mechanics Approach


Fracture Mechanics Approach

This method is based upon linear elastic fracture


mechanics (LEFM) principles which are adapted
for cyclic loading.
This method is used to predict propagation life
f
from
an initial
i iti l crack
k or d
defect.
f t
This method is also used in combination with the
strain life approach to predict the total life
(initiation and propagation).

Chapter 4 - Fatigue

12

Department of Mechanical &


Aerospace Engineering

Fatigue Limit Ferrous Materials


F
Failure
il
att stresses
t
wellll b
below
l
th
the
nominal tensile strength
50-90% of al structural failures occur
by fatigue
Wohler carried out the first systematic
Endurance No of cycles to failure
test of railway axles under repeated
Most ferrous materials exhibit a Fatigue
loads in Germany in 1840s
Limit
Analysis based on Wohler S-N
SN
Fatigue Limit
diagram
Stress level below which failure will
Showed that fatigue life decreased
not occur by fatigue
with higher stress amplitudes
Fatigue Limit can disappear due to
Periodic overloads
Corrosive environments
g temperatures
p
High
Chapter 4 - Fatigue

13

Department of Mechanical &


Aerospace Engineering

Fatigue Limit Non-Ferrous Materials


Non-Ferrous Materials

Effect of Notches

Most Non-Ferrous materials have no


fatigue / endurance limit

The presence of notches reduces the


fatigue performance of material and
structures

A Pseudo endurance limit is taken as the


stress corresponding to a life of 5 x 108
cycles
Chapter 4 - Fatigue

14

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface
S f
Treatment
T t
t
Plating
Thermal
Mechanical
Temperature
Environment
Stress Concentration Kt

Chapter 4 - Fatigue

15

Stress Fluctuation

Department of Mechanical &


Aerospace Engineering

Variable Amplitude
p
Loading
g
However:
Most tests are carried out at
Constant Amplitude loading

Chapter 4 - Fatigue

16

Department of Mechanical &


Aerospace Engineering

Constant Amplitude Loading

Stress ratio R = Smin / Smax


Mean stress Sm
Stress Amplitude Sa
Stress Range = Smax Smin
Show diagram of Mean loads
Chapter 4 - Fatigue

17

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface
S f
Treatment
T t
t
Plating
Thermal
Mechanical
Temperature
Environment
Stress Concentration Kt

Chapter 4 - Fatigue

18

Department of Mechanical &


Aerospace Engineering

Effect of Mean Stress

Mean stress has an


effect on fatigue life
The effect can be
portrayed
p
y as a function
of Smean and Sa

m 4 m 3 m 2 m1
Chapter 4- Fatigue

19

Department of Mechanical &


Aerospace Engineering

Mean Stress Equations


Goodman :
Gerber

a
Se

m
SU

1
2


m 1
:
S e SU

Soderberg :

Morrow

Se
Se

SY
Sf

1
1

a = fatigue strength in terms of stress amplitude where m 0


Se = fatigue strength in terms of stress amplitude where m = 0
m = mean stress
Su = Ultimate strength
Sy = Yield strength
f = true fracture strength
Chapter 4 - Fatigue

20

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
Temperature
Environment
Stress Concentration Kt

Chapter 4 - Fatigue

21

Department of Mechanical &


Aerospace Engineering

Size Effect

Other things being equal, an


iincrease iin section
ti size
i will
ill
result in a decrease in fatigue
life
g is controlled by
y
Fatigue
weakest link of the material
Probability of weak link
increases with volume
A larger component will have a
larger surface volume and
therefore a larger surface area
subjected to high stress

Chapter 4 - Fatigue

22

Department of Mechanical &


Aerospace Engineering

Size Effect in Large and Small Specimens

Larger volume subjected to


higher stress

Cracks initiate earlier and


propagate faster

Chapter 4 - Fatigue

23

Department of Mechanical &


Aerospace Engineering

Size Effect in Welded Specimens

Weld toe radius does not


scale up

Therefore hot spot has a


higher local stress

Chapter 4 - Fatigue

24

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
e pe atu e
Temperature
Environment
Stress Concentration Kt
Chapter 4 - Fatigue

25

Surface Finish

Department of Mechanical &


Aerospace Engineering

Machining marks, scratches and surface scale lead to the


initiation of cracks
The condition of the surface is more important for higher
strength steels
Residual surface stresses caused by machining operation
can be important. An example can be the tensile residual
stresses left by some grinding operations
At shorter lives
lives, where crack propagation dominates
dominates, the
condition of the surface finish has less effect on the fatigue
life.
Localized surface irregularities such as stamping marks can
serve as very effective stress concentrations and should not
be ignored.

Chapter 4 - Fatigue

26

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
e pe atu e
Temperature
Environment
Stress Concentration Kt
Chapter 4 - Fatigue

27

Department of Mechanical &


Aerospace Engineering

Surface Treatment

Cracks initiate at the free surface therefore any surface treatment


has a significant effect
Plating
Chrome and Nickel plating can cause up to a 60% reduction in
the endurance limit. Primarily due to the high residual tensile
stresses caused by the plating process
process.

To alleviate the problem: Nitride part before plating


2.
2 Shot peen part before or after plating
3. Bake or anneal part after plating

Chapter 4 - Fatigue

28

Department of Mechanical &


Aerospace Engineering

Surface Treatment
Thermal Treatment

Diffusion processes such as carburizing or nitriding.


This process produces a higher strength material at the surface and
causes a volumetric change which produces residual compressive
stresses.

Flame or Induction Hardening


causes a phase transformation which produces a volumetric change
which causes residual compressive stresses

Manufacturing processes such as welding, grinding and flame


cutting
can set up detrimental residual tensile stresses and shot peening can
undo some of this damage.

Chapter 4 - Fatigue

29

Department of Mechanical &


Aerospace Engineering

Surface Treatment
Mechanical Treatment
Cold working the surface can set up residual compressive
stresses.
Cold Rolling of screw threads
Shot Peening - Steel or glass beads fired at the surface to cause
yielding. Shot Peening has the greatest effect at long lives.
Advantage - it can be applied to oddly shaped parts
Disadvantage - it dimples the surface
At high temperatures the residual stresses can fade 500oF for steel and 250oF
for aluminium.
Steels with yield stresses less than 80ksi are seldom cold rolled or shot
peened because with such low yield stresses it is possible to load to yield and
wipe out the residual stresses.
Shot peening has the greatest effect when it is applied to areas of high stress
gradients ((i.e. at stress concentrations).
g
)
Chapter 4 - Fatigue

30

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
e pe atu e
Temperature
Environment
Stress Concentration Kt
Chapter 4 - Fatigue

31

Department of Mechanical &


Aerospace Engineering

Temperature
At low temperature
p
there is a tendency
y for the endurance
limits of steels to increase. However a more important
design consideration is that the there is a significant
decrease in fracture toughness at low temperatures
At high temperatures the endurance limit in steels
disappears
pp
At temperatures greater than 0.5 x the melting temperature
creep becomes important. In this range the stress-life
stress life
approach is no longer applicable.

Chapter 4- Fatigue

32

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
e pe atu e
Temperature
Environment
Stress Concentration Kt
Chapter 4 - Fatigue

33

Department of Mechanical &


Aerospace Engineering

Environment
When fatigue loading takes place in a corrosive
environment the net detrimental effect is more significant
that if corrosion and fatigue were considered separately.
This is Corrosion Fatigue
Corrosion fatigue is not clearly understood - involves many
variables
This topic will be considered later in course

Chapter 4 - Fatigue

34

Department of Mechanical &


Aerospace Engineering

Factors that Influence Fatigue


Stress Fluctuation
Mean Stress Effect
Size Effect
Surface Finish
Surface Treatment
Plating
Thermal
Mechanical
e pe atu e
Temperature
Environment
Stress Concentration Kt
Chapter 4 - Fatigue

35

Department of Mechanical &


Aerospace Engineering

Stress Concentration Factor Kt


This is the ratio of the local to nominal stress
Kt is too severe a factor where fatigue
using the fatigue stress concentration factor Kf

S e( un notched )
K f ( notched )
Se
q

K f 1
Kt 1

K f Kt
q is the notch sensitivity
factor which takes account of
the sensitivity of the material
to stress raisers
Chapter 4 - Fatigue

36

Department of Mechanical &


Aerospace Engineering

Stress Concentration Factor Kt


Relationships for determining q
Peterson : q

1
a
1
r

Kt 1
Kf 1
a
1
r
a = material constant
r = notch root radius
= material constant

N b : q
Neuber

Kt 1

Kf 1

Approximations for a for ferrous materials

300
a 0 .001

S
k
ksi
i
U

1 .8

in
1 .8

2070
a 0.025
mm

S
MPa
U

Chapter 4 - Fatigue

37

Department of Mechanical &


Aerospace Engineering

Stress Concentration Factor Kt


Approximations for a for ferrous materials
300
a 0 .001

S U ksi

1 .8

1 .8

in

2070
a 0.025
mm

S
MPa
U

Using approximation Su 0.5 BHN


a for normalized or annealed steels 0.010 in (BHN 170)
a for quenched and tempered steels 0.0025 in (BHN 360)
a for highly hardened steels 0.001
0 001 in (BHN 600)

Chapter 4 - Fatigue

38

Department of Mechanical &


Aerospace Engineering

Notch Sensitivity Curves

Peterson

Neuber
Chapter 4 - Fatigue

39

Department of Mechanical &


Aerospace Engineering

Notch Sensitivity
y for Two Geometrically
y Similar Plates
Kt 1
a
a
1
1
r
r
Approximations for a for ferrous materials
Peterson : q

300
a 0 .001

S
ksi
U

K f 1

1 .8

in

Plate 1

Plate 2

W1 = 5.0 in
d1 = 0.5 in
Su = 100 ksi
Kt = 2.7
q = 0.97
Kf1 = 2.65
2 65

W2 = 0.5 in
d2 = 0.05 in
Su = 100 ksi
Kt = 2.7
q = 0.78
Kf2 = 2.32
2 32
Chapter 4 - Fatigue

40

Department of Mechanical &


Aerospace Engineering

Stress History or Cumulative Damage


In most practical situations constant amplitude loading does not exist.
The loading is variable amplitude

Stress Cycle Counting


Cumulative Damage Miner Summation
This approach does not take account of the
order of stress cycles
The order does have an effect on fatigue life

Chapter 4 - Fatigue

41

Department of Mechanical &


Aerospace Engineering

Miner Summation

ni
1
Ni

n1 n2
n3
ni

Damage
D
N1 N 2 N 3 N i

Chapter 4 - Fatigue

42

Department of Mechanical &


Aerospace Engineering

Weld Detail Example


Offshore Detail F2 Classification
L
Long
term S
Stress R
Range Di
Distribution
ib i
Block

Stress Range
)Fi (MPa)

No of
Cycles
Ni

455

400

345

30

300

70

250

300

210

700

170

3,000

130

7,000

90

30,000

10

55

70,000

11

30

30,000

Chapter 4 - Fatigue

43

Department of Mechanical &


Aerospace Engineering

Weld Detail Example


Offshore Detail F2 Classification
The S-N curve for the F2 classification is:

m
t
log N log a log
m log
4
22
where log a 1163
. and m 3.0
Determine the damage accumulation
and hence estimate the cycles to failure
Chapter 4 - Fatigue

44

Weld Detail Example


Solution
log N 1163
.

log N log a

Total Damage D = 0.3151


No of cycles to failure

m
t
log m log
4
22

3.0
40
log 3.0 log
4.0
22

l N 1144
log
. 3.0 log
l

411110
1.3 106
0.3151

Department of Mechanical &


Aerospace Engineering

Block

ni

Ni

ni/Ni

455

2924

0 0014
0.0014

400

4303

0.0014

30

345

6707

0.0045

70

300

10200

0.0069

300

250

17627

0.0170

700

210

29740

0.0235

3000

170

56060

0.0535

7000

130

125363

0.0558

30,000

90

377809

0.0794

10

70,000

55

1655434

0.0423

11

300,000

30

10200847

0.0294

ni = 411110

Chapter 4 - Fatigue

D = 0.3151

45

Department of Mechanical &


Aerospace Engineering

High / Low Cycle Fatigue


High Cycle Fatigue

Low Cycle Fatigue

Low stress levels

High stress levels

Stress much less than


yield

Cycles < 104

Cycles > 105

Chapter 4 - Fatigue

46

Department of Mechanical &


Aerospace Engineering

Strain Life Approach


Strain life approach is based on observations of a components
response in critical regions (i
(i.e.
e notches)
At high load levels in low cycle fatigue (LCF) regime the cyclic
stress-strain
i response and
d the
h material
i lb
behaviour
h i
iis b
best
modelled by strain controlled conditions.
This method takes account of both elastic and plastic strains.
pp
assumes that smooth specimens
p
tested
The strain-life approach
under strain controlled conditions can simulate the fatigue
damage at the root of a notch.

Chapter 4 - Fatigue

47

Department of Mechanical &


Aerospace Engineering

Equally Stressed Volume of Material


Equivalent fatigue damage is assumed
to occur in the material at the notch
root and in the smooth specimen when
both are subjected to identical stressstrain histories.
Crack growth is not accounted for in
this method.
Failure of the component is assumed
to occur when the equally stressed
volume fails.
This method is considered an
initiation life estimate.

Chapter 4 - Fatigue

48

Department of Mechanical &


Aerospace Engineering

Fatigue Life Prediction


To predict fatigue life using the Strain Life approach
the following Information is required:
properties obtained from smooth
specimen strain controlled fatigue data
Stress-strain history at the critical
location e.g. at the notch
Cycle counting information
Methods to incorporate mean stress
effects
Damage summation techniques Miner's rule

Chapter 4 - Fatigue

49

Department of Mechanical &


Aerospace Engineering

Review of Monotonic Stress- Strain Behaviour


Engineering
g
g Stress

P
Ao

Engineering Strain

l
lo

True Stress

True Strain

dl
lo l
l

ln

l
lo

ln((1 e)
Chapter 4 - Fatigue

S (1 e)
50

Department of Mechanical &


Aerospace Engineering

Cyclic Stress Strain Behaviour


is the total strain range
is the total stress range
The total strain is the sum of
th elastic
the
l ti and
d plastic
l ti strains
t i

e p
in terms of amplitudes
e p

2
2
2
The area within the loop is the energy dissipated during a cycle

Chapter 4 - Fatigue

51

Department of Mechanical &


Aerospace Engineering

Cyclic Strain Hardening and Softening


Depending
D
di on th
the iinitial
iti l conditions
diti
off a metal
t l and
d th
the ttestt
conditions, a metal may:

y
y harden
Cyclically
Cyclically soften
Be cyclically stable
Have mixed behaviour
(harden or soften depending on strain range)

Chapter 4- Fatigue

52

Department of Mechanical &


Aerospace Engineering

Cyclic Strain Hardening and Softening


Cyclic Hardening:
"soft"
so t material
ate a - initially
t a y dislocation
d s ocat o
density is low - density increases
due to cyclic plastic straining
resulting in strain hardening

Cyclic Softening:
"Hard" material - strain cycling
y
g
causes a rearrangement of
dislocations which offers the
least resistance to deformation
- therefore the material softens

Chapter 4 - Fatigue

53

Department of Mechanical &


Aerospace Engineering

Cyclic Strain Hardening and Softening


Strain Hardening and softening occurs during early fatigue life
- then the material stabilizes - usually at about 20 - 40 % of the fatigue life
Therefore fatigue
g p
properties
p
are usually
y specified
p
at half life
(at 50% of life when the material has stabilized)
Will Material Cyclically Harden Or Soften?

sult 1.4

Material will cyclically harden

ult
12
.
y

Material will cyclically soften

The monotonic strain hardening exponent n can also be used

n 0.2

Material will cyclically harden

n 01
.

Material will cyclically soften

Chapter 4 - Fatigue

54

Department of Mechanical &


Aerospace Engineering

Cyclic and Monotonic Stress-Strain Curves

Chapter 4 - Fatigue

55

Department of Mechanical &


Aerospace Engineering

Strain Life Curve


Basquin Equation
Basquin (1910) observed that S-N data could be plotted on a log-log plot
using true stress amplitude as:

f 2 N f
2

Where:
/2 = true stress amplitude
2Nf = reversals to failure ( 2 rev = 1 cycle)
'f = fatigue strength coefficient
b = ffatigue
ti
strength
t
th exponentt (Basquins's
(B
i ' exponent)
t)
'f is approximately equal to the true fracture strength f
b is usually between -0.05 and -0.12

Manson Coffin Equation


M
Manson
and
dC
Coffin
ffi (19
(1950)
0) ffound
d that
h the
h Plastic
Pl i strain
i - Life
Lif
data could also be linearized on a log-log plot.

f 2 N f
2

Where:
/2 = true strain amplitude
2Nf = reversals
l tto failure
f il
( 2 rev = 1 cycle)
l )
'f = fatigue ductility coefficient
c = fatigue ductility exponent
'f is approximately equal to the true fracture ductility f
c is usually
y between -0.5 and -0.7
Chapter 4 - Fatigue

56

Department of Mechanical &


Aerospace Engineering

Strain-Life Relationship
Total Strain in terms of strain amplitude

The elastic term can be written as:

e p

2
2
2
e

2
2E
f

2N f
2
E

In terms of life to failure (Basquin):

Also the Plastic term is: (Manson-Coffin)

f 2 N f

Therefore the Total Strain can be written as:

2N f
2
E

f 2 N f

This is termed the Strain Life Relationship


at large strains or short lives plastic strain predominates
at small strains or short lives elastic strain predominates

Chapter 4 - Fatigue

57

Department of Mechanical &


Aerospace Engineering

Strain-Life Relationship

Chapter 4 - Fatigue

58

Mean Stress Effects

Department of Mechanical &


Aerospace Engineering

MORROW :
Modify the elastic term

( f o )

2N f
2
E

2 N
b

-tensile mean stress


-mean stress effect is significant at low values of plastic strain
-predicts that the ratio of elastic to plastic strain depends on
the stress ratio - this is not correct
Chapter 4 - Fatigue

59

Department of Mechanical &


Aerospace Engineering

Mean Stress Effects

SMITH, WATSON & TOPPER (SWT)


for completely reversed loading modified
the strain life equation by the term

MANSON & HALFORD:

max

f 2 N f
2

multiplying the strain life equation by this term

( f o )

2N f
2
E

c
b

f o
2 N f
f
f

Modified both the elastic and plastic terms to maintain


the independence of the elastic / plastic ratio with mean
Stress This approach tends to predict too much mean
stress effect at short lives where the plastic strains
predominate

2
( f )
max

2N f
2
E

max is evaluated

as

2b

f f 2 N f

o
2

since max N f
when max < 0 this is undefined
Therefore this approach assumes that
no fatigue damage occurs when max < 0

Chapter 4 - Fatigue

60

b c

Strain Life Prediction

Department of Mechanical &


Aerospace Engineering

Fatigue Analysis Procedure

Material
Properties

Component
Geometry

Loading
History

Stress-Strain
Analysis
Damage
Analysis

Fatigue Life
Chapter 4 - Fatigue

61

Department of Mechanical &


Aerospace Engineering

Strain Life Prediction

The basic assumption of the strain-life (N) fatigue analysis

approach is that fatigue damage accumulation and fatigue life


to crack initiation at the notch tip, are the same as in a
smooth material specimen, if subjected to the same stress-strain
history.
Thus,
Thus if the stress-strain
stress strain history at the notch tip of a machine
component is the same as that one applied to a smooth specimen
the fatigue life of the smooth laboratory specimen will be equal to
the fatigue crack initiation life at the notch tip
tip.

Chapter 4 - Fatigue

62

Department of Mechanical &


Aerospace Engineering

Smooth Specimen
p
Notch Tip
p
a)

peak

Stress

Stress

peak

peak

peak

Chapter 4 - Fatigue

63

Department of Mechanical &


Aerospace Engineering

Smooth Specimen
p
Weld Toe
b)

Chapter 4 - Fatigue

64

Department of Mechanical &


Aerospace Engineering

Material Properties for -N Approach


Stress Strain Curve

Strain Life Curve

f
2 N f

2
E
/

/f 2 N f

n'
'
E K

where:
E is the modulus of elasticity,
K cyclic strength coefficient,
n strain hardening exponent

where:
f - fatigue
f ti
strength
t
th coefficient,
ffi i t
b - fatigue strength exponent,
f - fatigue ductility coefficient,
c - fatigue
at gue duct
ductility
ty e
exponent
po e t
Chapter 4 - Fatigue

65

Department of Mechanical &


Aerospace Engineering

Steps in a Fatigue Life Prediction


Analysis of external forces acting on the structure and the
component in question (Fig. a, b),
Analysis of internal loads in chosen cross section of a component (
Fig. b),
Selection of particular welded joints in the structure (Fig. c),
Calculation of the peak stress, peak, at the critical point (Fig. d)
Assembling of the stress peak history in the form of appropriate
sequence of stress peaks and valleys (Fig. e-f),
Determination of the elastic-plastic response at the critical location
(Fig. h),
Identification (extraction) of cycles represented by closed stressstrain hysteresis loops (Fig. i),
Calculation of fatigue damage (Fig. j-k),
Fatigue damage summation (Miner- Palmgren hypothesis, Fig. 4l),
Determination of fatigue life (Fig. 4m) in terms of number of stress
history repetitions, Nblock, (No. of blocks) or the number of cycles to
fatigue crack initiation, N.
Chapter 4 - Fatigue

66

Department of Mechanical &


Aerospace Engineering

Steps in a Fatigue Life Prediction


a)
Structure

Analysis of external
f
forces
acting
ti on the
th
structure

b) Component

H
Q
c) Section with welded
joint

Selection of
particular welded
joints

Weld

R
peak

Analysis of
internal loads in
chosen
h
cross
section of a
component

hs
n
d)

Chapter
- Fatigue
Chapter
6 -4Fatigue
peak

67

Department of Mechanical &


Aerospace Engineering

Steps in a Fatigue Life prediction


F
Q

c) Section with welded joint

Weld

R
V

peak

hs
n
d)

peak
hs
h

Chapter 4 - Fatigue

Calculation of the peak


stress, peak, at the
point using
g FE
critical p
analysis

68

Department of Mechanical &


Aerospace Engineering

Steps in a Fatigue Life prediction


e)

g)

Assembling of the
stress peak
history

'
K

'


Neuber :
peak

f)

h)

1'

1,1'

peak

5,5'
3

7,7'
5

Determination of
the elastic-plastic
response at the
critical location

2,2'
6
4

j)

2N f
2
E
Chapter
- Fatigue
Chapter
6 -4Fatigue
log (/2)

i)

'

'f 2 N f c

69

peak

Department of Mechanical &


Aerospace Engineering

Steps in a Fatigue Life prediction


Neuber :

h)

1'

1,1'

5,5'
5,5
3

Identification of
cycles represented
b closed
by
l
d stresst
strain hysteresis
loops
i)

7,7'

peea

f)

2,2'
6

log (/2)

j)

2N f
2
E
'

'f 2 N f c

Calculation of
fatigue damage

f/E
e/E

e=

Fatigue damage
by Miner
summation

k) Fatigue damage:
g
l)) Total damage:

D1

2N

log (2Nf)

2Ne

1
1
1
1
1
; D2
; D3
; D4
; D5
;
N1
N2
N3
N4
N5

D D1 D2 D3 D4 D5 ;

j) Fatigue life: N blck=1/D

Fig. 4. (continued). Steps in fatigue life prediction procedure based on the -N


approach

Chapter 4 - Fatigue

70

Department of Mechanical &


Aerospace Engineering

Steps
p in a Fatigue
g Life Prediction
Analysis of external forces acting on the structure and the
component in question (Fig. a, b),
Analysis of internal loads in chosen cross section of a component
( Fig. b),
Selection of particular welded joints in the structure (Fig. c),
Calculation of the peak stress, peak, at the critical point (Fig. d)
Assembling of the stress peak history in the form of appropriate
sequence of stress peaks and valleys (Fig. e-f),
Determination of the elastic-plastic response at the critical location
(Fig. h),
Identification (extraction) of cycles represented by closed stressstrain hysteresis loops (Fig. i),
Calculation of fatigue damage (Fig. j-k),
Fatigue damage summation (Miner- Palmgren hypothesis, Fig. l),
Determination of fatigue life (Fig. m) in terms of number of stress
history repetitions, Nblock, (No. of blocks) or the number of cycles
to fatigue crack initiation, N.
Chapter 4 - Fatigue

71

Department of Mechanical &


Aerospace Engineering

Factors Affecting Fatigue Performance


1.

2
2.

3.

STRESS/LOAD

State of Stress

Stress Range
g

Stress Ratio

Constant or Variable Loading

Frequency

Maximum Stress
GEOMETRY

Stress Raisers

Size

Stress Gradient

Metallurgical & Mechanical


Properties of Base Materials and
Weldments
EXTERNAL ENVIRONMENT
Temperature
Aggressiveness
Corrosion Fatigue

Chapter 4 - Fatigue

3 STAGES

Fatigue Crack Initiation


Fatigue Crack Propagation
Final Failure - Fracture

NT Ni N p

72

Department of Mechanical &


Aerospace Engineering

Crack Initiation
CRACK INITIATION
at a surface
- stress concentrations
- surface
f
scratches
t h
- dents
- machining marks
- fillets
SURFACE TREATMENTS
1. MECHANICAL - Shot Peening
- Cold Rolling
- Grinding
- Polishing
2. THERMAL
- Flame Hardening
- Induction Hardening
3. SURFACE COATING

- Nitriding
g
- Plating
Chapter 4 - Fatigue

73

Crack Growth

Department of Mechanical &


Aerospace Engineering

Crack Growth through loading and unloading cycle

Chapter 4 - Fatigue

74

Department of Mechanical &


Aerospace Engineering

Crack Growth and Stress Intensity Factor


R

min max

1
max
max
max

1 R max

K min min a
K max max a
K

K min
K max

K max

K
1 R

This indicates that crack growth rate is a rising


function of K and R.

da
f ( K , R )
dN
The process of crack growth is a complicated one,
and
d it d
depends
d
on material behaviour
therefore it is not possible to obtain this function
through rigorous mathematical description.
It must be obtained from material tests,, i.e. empirically.
p
y
Chapter 4 - Fatigue

75

Department of Mechanical &


Aerospace Engineering

Measurement of Growth Rate Function


For a wide plate with a small centre crack (a/W < 0.4, 1),

K a

If the da/dN vs K results are plotted on a log-log plot the results are
approximately a straight line of the form y= mx + C, i.e.

da
log
m logK logC
dN
da
C K m
dN
This is known as the Paris Equation

The values of C and m are obtained by curve fitting.

Chapter 4 - Fatigue

76

Department of Mechanical &


Aerospace Engineering

Measurement of Growth Rate Function


Point A : log( 1.6 10 7 ) m log 6.3 log C
Point B : log( 4 10 5 ) m log 40 log C
6.8 0.8 m log C
4.4 1.6 m log C
2.4 0.8 m

m3

log C 9.3

C 6.3 10 10

da
6.3 10 10 K 3
dN
where da/dN is in m/s and MPam
NOTE:
The values of C and m depend on the units of the test data and conversion
from one set of units to another should only be attempted with great caution.

Chapter 4 - Fatigue

77

Department of Mechanical &


Aerospace Engineering

Crack Propagation Equations


da
C K m
dN

PARIS EQUATION

FORMAN EQUATION
This equation shows the growth rate depends on R and
also at fracture when K =Kc the growth rate goes to 4
KLESNIL & LUKAS

SCHULTZ

COLLIPRIEST

da
C K m

dN 1 R K C K

da
C K m K thm
dN
m
m
da C K K th

dN 1 R K C K

ln K 2 1 R K C K th
Kc 2
d
da
2
arctan
C K th K exp ln
dN
K th
ln 1 R K C K th

Chapter 4 - Fatigue

78

Department of Mechanical &


Aerospace Engineering

Crack Propagation Equations


STRESS INTENSITY FACTOR
MPa m

ksi in

3/2
Nmm-3/2

3/2
kp mm-3/2

MPa m

0.9101

31.623

3.2235

Ksi in

1.0988

34.747

3.5420

N mm-3/2

0.031623

0.02878

0.10194

kp mm-3/2

0.31022

0.28233

9.8067

STRESS
MPa

ksi

N/mm2

kp/mm2

0.1449

0.1019

6 904
6.904

6 9033
6.9033

0 7037
0.7037

N/mm2

0.1449

0.1019

kp/mm2

9.8067

1.4211

9.8067

MPa
ksi

Chapter 4 - Fatigue

79

Department of Mechanical &


Aerospace Engineering

Crack Propagation Equations


The complete da/dN -K curve is sigmoidal in shape.

3 Regions
Threshold
Propagation
P
ti
Fracture

Chapter 4 - Fatigue

80

Department of Mechanical &


Aerospace Engineering

Fatigue Crack Threshold


Threshold Influenced by:
Yield strength, grain size mean stress
Stress History - residual stress
Crack
C k ti
tip opening
i mode
d
Temperature
Environment
A lot of scatter in existing
g Kth data
Stress Ratio most important Factor
Estimates of Kth
Steels:
martensitic, bainetic, ferrite
ferrite-pearlite,
pearlite, austenitic
Kth = 6.4 (1 - 0.85 R)ksi in
R>+1
Kth = 7 (1 - 0.85 R)

MN/m3/2

Kth = 5.5 ksi in

constant

R > 0.1
Kth = 6 MN/m3/2
Chapter 4 - Fatigue

81

Department of Mechanical &


Aerospace Engineering

Life Prediction Constant Amplitude Loading

Paris Equation

dda
C K m
dN
Nf

integrating

1
dN
C

af

da
a K m
i

K m Y a
Nf

1
dN
C

af

da

ai

Chapter 4 - Fatigue

82

Department of Mechanical &


Aerospace Engineering

Life Prediction Constant Amplitude


p
Paris Equation
q
Closed form approach
pp
da
C K m
dN
af

Nf

SIF of an edge
g crack

K m 1.12 a
Nf

Integrating (for m 2)

da
a K m
i

1
dN
C

af

da

1.12

ai

2
1
1
Nf

m2
m2
(m 2) C (1.12 ) m 2
ai
a f 2
Chapter 4 - Fatigue

83

Department of Mechanical &


Aerospace Engineering

Life Prediction Constant Amplitude


Paris Equation
q
Closed form approach
pp

2
1
1
Nf

m2
m2
(m 2) C (1.12 a ) m 2
ai
a f 2

Before this equation can


be evaluated the final
crack size must be calculated:

K f (g ) a
1 KC
af
f g
Chapter 4 - Fatigue

84

Department of Mechanical &


Aerospace Engineering

Life Prediction Procedure


1. Estimate the initial flaw size a0
1
2. Obtain SIF relationship
3. Knowing KC or KIC and the nominal design stress Calculate the critical flaw size af
4. Obtain the relationship relating da/dN with K e.g
da
2.25
0.66 10 8 K I
dN
da
is in in/cycle and K is in ksi in
dN

5. Determine KI for expression in (2)


5
6. Integrate da/dN expression between limits a0 to af (at KIC)
to obtain life

Chapter 4 - Fatigue

85

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


1. Assume
A514 steel
(689 MN/m2)
YS =100 ksi
KIC = 150 ksiin
(165 MNm3/2)
a0 = 0.3 in
(7.6 mm)
Edge
g crack in tension
max = 45 ksi
(310 MN/m2)
(172 MN/m2)
max = 25 ksi
= 20 ksi

k i
(138 MN/m
MN/ 2)

Chapter 4 - Fatigue

86

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


p

2. Edge crack in tension KI = 1.1.2 a


3. Calculate the critical crack size aC at = 45 ksi
2


K IC
150


acr
2.81 in

1.12 max 1.12 45 1.77

4. Assume an interval of crack growth say a =


0.1 in (smaller intervals give slightly better
results
5. Determine KI where aavg is the average crack
size between ai and aj
Chapter 4 - Fatigue

87

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


p
6. Using the da/dN expression solve for N for
each interval of crack growth a/N
da
2.25
0.66 10 8 K i
dN
a
N
0.66 10 8 1.98 20 aavgg

2.25

For crack growth from 0.3 to 0.4


N 12,500 cycles

7. Repeat for growth from 0.4


7
0 4 0.5
0 5 and so on.
on
8. Total Life 86,700 cycles

Chapter 4 - Fatigue

88

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


p
(ksi% in)

)N

GN

(cycles)

(cycles)

0.35

23.5

12,500

12,500

0.5

0.45

26.7

9,750

22,250

0.5

0.6

0.55

29.4

7,550

29,800

0.6

0.7

0.65

32.2

6,150

35,950

0.7

0.8

0.75

34.6

5,200

41,150

0.8

0.9

0.85

36.6

4,600

45,750

0.9

1.0

0.95

38.8

4,100

49,850

1.0

1.1

1.05

40.5

3,700

53,550

1.1

1.2

1.15

42.5

3,300

56,850

1.2

1.3

1.25

44.5

2,950

59,800

1.3

1.4

1.35

46.1

2,700

62,500

ao
(in)

af
(in)

aavg
(in)

0.3

0.4

0.4

)K

Chapter 4 - Fatigue

Total Life
86,700
86
700
cycles

89

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


p

Chapter 4 - Fatigue

90

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


p
If the required Life is 100,000 cycles:
Options:
1.
1
2.
3.

IIncrease af by
b using
i
a tougher
t
h material
t i l i.e.
i Higher
Hi h KIC
Lower the design stress max i.e. Increase acrit at failure
Lower stress
1.
2.

4.

Decrease rate of crack growth


Slightly lower has a significant effect on life due to
power law relationship

Improve fabrication quality and inspection capability


so that the initial flaw size is reduced

Chapter 4 - Fatigue

91

Department of Mechanical &


Aerospace Engineering

Life Prediction Example


If a0 were 0.2 in
N = 18,000 cycles
for a = 0.2 0.3

Total life = 104,700 cycles

Chapter 4 - Fatigue

92

Department of Mechanical &


Aerospace Engineering

Load Interaction - Retardation

Next Section

Chapter 4 - Fatigue

93

Department of Mechanical &


Aerospace Engineering

References
N.E. Dowling, Mechanical Behavior of Metals, Prentice Hall, New
J
Jersey,
1999.
1999
A. Baumel and T. Seeger, Materials Data for Cyclic Loading,
Elsevier, Amsterdam, 1990.
H.
H Neuber
Neuber, Theory of Stress Concentration of Shear Strained
Prismatic Bodies with Arbitrary Non Linear Stress-Strain Law,
ASME Journal of Applied Mechanics, vol. 28, 1961, pp. 544-550.
K. Molski and G. Glinka, A Method of ElasticPlastic Stress and
Strain Calculation at a Notch Root, Material Science and
Engineering, vol. 50, 1981, pp. 93-100.
A. Moftakhar, A. Buczynski and G. Glinka, Calculation of ElastoPlastic Strains and Stresses in Notches under Multiaxial
Loading, International Journal of Fracture, vol. 70, 1995, pp.
357-373.
Chapter 4 - Fatigue

94

References

Department of Mechanical &


Aerospace Engineering

J.M. Barsom and S.T. Rolfe,


Fatigue & Fracture Control in Structures, Prentice Hall, 1987.
P.A. Ross-Ross,
Th Investigation
The
I
ti ti into
i t the
th Cracking
C
ki off Pressure
P
Tubes
T b in
i Pickering
Pi k i
Units 3 and4, From Steam to Space, CSME 1996.
Standard Method for Plane-Strain Fracture Toughness of Metallic
Materials. ASTM Specification
p
E-399-83
Brock, D.
"Elementary Engineering Fracture Mechanics"
((2nd Edition,, Martinus Nijhoff
j
Publishers,, 1982))
Brock, D.
"The Practrical Use of Fracture Mechanics
((Kluwer Academic Publishers,, 1988))
Anderson, T.L
Fracture Mechanics - Fundamentals and Applications
((2 nd
d Edition,
d t o , CRC
C C Press,
ess, Boca
oca Raton,
ato , 1995)
995)
Chapter 4 - Fatigue

95

References

Department of Mechanical &


Aerospace Engineering

Bannantine, J.A., Comer, J.J. and Handrock, J.L. "Fundamental of


Metal Fatigue Analysis (Prentice Hall, 1990)

Almar-Naess, A. "Fatigue Handbook - Offshore Steel structures.


(Tapir Publishers, Trondheim, Norway, 1985)

Chapter 4 - Fatigue

96

You might also like