Contemporary Fatigue Analysis

Methods
(basics concepts and assumptions)

9/30/2018 1
 Information Path for Strength and Fatigue Life Analysis

Material Component Loading

Properties Geometry History

Stress-Strain
Analysis

Damage Analysis

Fatigue Life

9/30/2018 2
 Stress Parameters Used in Fatigue Analyses
Sn – net nominal stress; S – gross nominal stress
oe
peak – local linear-elastic notch-tip stress K – stress intensity factor
oa
peak – local actual elastic-plastic notch-tip stress K  S  a Y
Kt = o e /S – stress concentration factor
peak n

S M S

  
 epeak  peak  K
 epeak 2 a
 apeak Sn crack
Sn

  y 
y

0 y 0 0
a
dn dn
T T T

a) S
b) c)
9/30/2018 3
 What stress parameter is needed for the Fracture
Mechanics based (da/dN-OK) fatigue analysis?

S
The Stress Intensity Factor K characterizing
the stress field in the crack tip region is
K
 x   needed!
2 x
Stress o(x)

The K factor can be obtained from:

-ready made Handbook solutions (easy to use

x
but often inadequate in practice)
0
a -from the near crack tip stress o(x)
distribution or the displacement data obtained
T from FE analysis of a cracked body (tedious)
-from the weight function by using the FE
stress analysis data of un-cracked body
S (versatile and suitable for FCG analysis)

9/30/2018 4
 Loads and stresses in a structure

Load F

opeak
on
9/30/2018 5
 The Most Popular Methods for
Fatigue Life Analysis - outlines

• Stress-Life Method or the S - N approach;

uses the nominal or simple engineering stress ‘ S ‘ to
quantify fatigue damage

• Strain-Life Method or the s - N approach;

uses the local notch tip strains and stresses to quantify
the fatigue damage

• Fracture Mechanics or the da/dN - OK approach;

uses the stress intensity factor to quatify the fatigue
crack growth rate

9/30/2018
. 6
 Information path for fatigue life estimation based on
the S-N method
LOADING GEOMETRY, Kf MATERIAL
F o
E
PSO

0 s
t

Stress-Strain
Analysis

MATERIAL
on
Damage Analysis A
oe D
F
No N

Fatigue Life

9/30/2018 7
 The Similitude Concept in the S-N Method
a) Structure

F
Q

Stress amplitude, n/2or hs/2

n
K0
K1
K5 K2
K3
K4
K5

0
Number of cycles,N

The Similitude Concept states that if the nominal stress histories in the structure and in the test
specimen are the same, then the fatigue response in each case will also be the same and can be
described by the generic S-N curve. It is assumed that such an approach accounts also for the stress
concentration, loading sequence effects, manufacturing etc.

9/30/2018
. 8
Steps in Fatigue Life Prediction Procedure Based on the
S-N Approach
a) Structure

The S – N method

b) Component

F
e) Standard S-N curves Q
c) Section with welded joint
Stress amplitude, n/2or hs/2

Weld
d) Standard welded joints

R
P
K0 V
K1
K2
K3
K4
K5

Number of cycles,N

5
9/30/2018
. 9
Steps in Fatigue Life Prediction Procedure Based
on the S-N Approach (continued)
f) h) Fatigue damage:

Stress,  n
 1 
m5
1
D1  
K5 N1 C5
 2 m5
n t D2 
1

N2 C5
 3 
m5
1
D3  
N3 C5
g)
D4 
1  4  m5


N4 C5
Stress

3 4  5 

m5
1
D5  
N5 C5
i) Total damage:

D  D1  D2  D3  D4  D5 ;
t

1

j) Fatigue life: N blck=1/D


2



9/30/2018
5

10
 The linear hypothesis of Fatigue Damage Accumulation
(the Miner rule)

 
  Ni( i)m = A or Ni=A/(i)m


 
Stress range, 

e
 
N5 ‹ œ


N1 N4 N2 N3 N0 Cycles
nR
5

D1 
1

 1  m

; D4 
1

 4 m
;
D  Di  D1  D2  D3  D4  D 5
i1
N1 A N4 A
1 1 1 1 1
1  2  m
1 1

  3 
m
     ;
D2   ; D5   0; D 3  ; N1 N2 N3 N4 N5
N2 A  N3 A
if D 1  Failure!!
1 1
LR   ; N f  LR n R
9/30/2018 D 1 N1  1 N 2  1 N 3  1 N 4  1 N 5
11
The FALSN fatigue life estimation software – Typical input and output data

Weldment

9/30/2018 12
 The scatter in fatigue:Fatigue S-N curves for assigned probability
of failure; P-S-N curves
S45 Steel tempered
at 600o C (W1)
500
 a  A N  m
f(N)
Stress amplitude Sa [N/mm2]

450 P=99%

P=90%
f(S) P=50% 8
P=10%
6
P=1% 48
400
Probability of failure
P(%)

105 106 107 108

(source: S. Nishijima, ref. 39) Number of cycles N
9/30/2018 13
 Probabilistic fatigue life assessment
LOADING STRUCTURAL COMPONENT MATERIAL

S
S
Rw
PW L
No N
t
fS fK fR

S Smax K Kf R Rw

SAMPLING A SET OF RANDOM DATA

COMPUTING of Tj

j<L
T
N

Failure probability calculation

Pf = P[T(X)  Tr]:
L f (T Tr )
Pf 
L

9/30/2018 14
 Characteristic regions of cumulative
probability of the fatigue life distribution

Un-satisfactory Most
Over designed
frequent

9/30/2018
. 15
 Information path for fatigue life
estimation based on the s-N method
LOADING GEOMETRY, Kf MATERIAL
F o

PSO E
0 s
t

Stress-Strain
Analysis

MATERIAL
½ Os
Damage Analysis

2Nf

Fatigue Life

. 19
9/30/2018
 The Similitude Concept in the s – N Method

log (/2)
a) Specimen f
 f
' j)
    2N c
b
x
speak  2N '
speak 2 E l) f
f f

y
z

speak log
0 (2Nf)

b) Notched component
The Similitude Concept states that if
the local notch-tip strain history in the
notch tip and the strain history in the
test specimen are the same, then the
fatigue response in the notch tip region
and in the specimen will also be the
same and can be described by the
material strain-life (s-N) curve.

9/30/2018
. 17
 Steps in fatigue life prediction procedure based on
the - N approach
a) Structure
peak

hs
n
pe
ak

h 
n
s

b) Component

d)

c) Section with welded joint

9/30/2018
. 18
 f)
e)
1' 
1

pea
3 5

k
1
7     ' 
'
n
tt
E K  
0
2 8 0
6 

1,1'
4   ,   
 
5,5'
7,7'  'f  , 

3
    
' '
2 f f

  , 
peak f
Neuber :

'
 E  'f 'f

8 N f ?,?
6 2,2'
4
log (/2)

 Fatigue damage:
f  f
'

    2N 
b c
 1 1 1 1
D ;D ;D ;D ;
'
2N f f f
2 E 1
N1 2
N2 3
N3 4
N4
 Total damage:



  f/E
D  D1  D2  D3  D4 ;
 e/E

 p e= 
 0 Fatigue life: N blck=1/D
2N log(2Nf) 2Ne

(continued). Steps in fatigue life prediction procedure based on the -Napproach

9/30/2018 19
9/30/2018
© 2008 Grzegorz Glinka. All rights reserved. 23
9/30/2018
© 2008 Grzegorz Glinka. All rights reserved. 24
9/30/2018
© 2008 Grzegorz Glinka. All rights reserved. 25
 Three basic sets of input data for the evaluation of the Fatigue Crack
Initiation Life and Reliability (the s – Napproach)
LOADING COMPONENT MATERIAL
S K5 2
 2
,f K'

S  ,f
S
E p 2
t 2N 1

f(k) f(Kt)
fM1 fM3

e ’f M K’
Scaling factor k SCF Kt
fM2

SAMPLING: k , Kt , o f , sf' , K’ s ’f

LIFE CALCULATION: T i

i <L
Y

Computing of failureprobabilities
Lf (T  Tr )
 X T  r 
Pf  P T
L

f(T) R(T) = 1 - F(T)

1

9/30/2018 T
26 T
 Probability of failure

9/30/2018
 Information path for fatigue life estimation based on
the da/dN-OK method

LOADING GEOMETRY, Kt MATERIAL

F o
E
PSO

0 s
t

Stress-Strain
Analysis

MATERIAL
da
n
dN
Damage Analysis

Kth K

Fatigue Life
9/30/2018
 The Similitude Concept in the da/dN – OK Method
10-6
a) Structure

10-7

Crack Growth Rate, m/cycle

H

F 10-8
Q

10-9
b) Weld detail
a
10-10

10-11

K
10-12
1 10 100
K,MPa m
a c) Specimen

The Similitude Concept states that if the stress

P intensity K for a crack in the actual component
and in the test specimen are the same, then the
fatigue crack growth response in the component
and in the specimen will also be the same and
can be described by the material fatigue crack
P growth curve da/dN - OK.

9/30/2018
 Steps in the Fatigue Life Prediction Procedure Based
on the da/dN-K Approach
a) Structure

e)

Stress, S
S3 S4

b) Component

S
t

1
H

S
2
F

S
Q

5
c) Section with welded joint

Weld

R
P
V d)

9/30/2018
Steps in Fatigue Life Prediction Procedure Based on the
da/dN-K Approach (cont’d)
i)

af
f) Stress intensity factor, K
(indirect method)

Crack depth, a
Weight function, m(x,y)
(x, y)
K   x, y m x, y dxdy Fatigue Life
A

K
Y
 n a
ai
Stress intensity factor, K
Number of cycles , N
(direct method)
g)
K I   yFE 2 xFE h) Integration of Paris’ equation
a
or
ai  C K i  N i
m
dU
K E  EG
da
N

K a f  a0   ai
Y
 n a i1

N   N i
9/30/2018
9/30/2018
9/30/2018
9/30/2018
 Probabilistic analysis using MC simulation
LOADING STRUCTURAL COMPONENT MATERIAL

da n
S dN

PW L
Kth K
t

fS fa fK
fC

S Smax a a K Kt
C C

fKth
SAMPLING RANDOM VARIABLES
Kth Kth

CALCULATION of Tj fKc

Kc KIc
j<L
Yes
No

Failure probability calculation

Pf = P[T(X) Tr]:

L f(T  Tr)
Pf 
L

The FALPR statistical simulation flow chart for the analysis of fatiguecrack growth
9/30/2018
 Irregular geometrical shape of a real fatigue crack

9/30/2018
 Welded Joint with Load = 4000 lb (ai/ci=0.286)

300000

250000 Exp. (Specimen #11)

Exp. (Specimen #12)
Exp. (Specimen #14)
Fatigue life N f (Cycle)

200000 Exp. (Specimen #15)

Exp. (Specimen #16)
Exp. (Specimen #17)
Exp. (Specimen #18)
150000 Pre. Life (Loc. 1) No Rs
Pre. Life (Loc. 1) With Rs

100000

50000

0
0 0.5 1 1.5 2 2.5

Crack Length 2c (in)

9/30/2018
 Global and Local Approaches to Stress
Analysis and Fatigue

9/30/2018
 Stress state near the notch tip (on the symmetry line)
22
a

23a 21a=
0 12a= 0
32a
2 =0
11a
31a 13 = 0
a

33a
1
3
 11a 0 0 
 
 ij   0  22 0 
a a

 0 0  a 
33 


9/30/2018 36
 Stress state in the disk at the blade-disk interface
22
a

23a 21a=
0 12a= 0
2 32
a
11a= 0

1
31a 13a= 0
33a
3

11a 0 0 
 a 
 ij   0  22  23 
a a

 0  32 a 

a
 33 

9/30/2018
. 37
Stresses concentration in axis-symmetric notched body

 22 F

C 2
 peak C
F
 22 n 
 22
n Anet
 33

A, B
 11
and
 peak  K t n
A D B
1
 22 3
 22
D

 11

F
 33
9/30/2018
. 38
 Stresses concentration in a prismatic notched body

 22

F
A, B, C
2

F
 22
 peak
 22
C
n 
D Anet
 11 n
11 and
A
E  peak  K t  n
D B  33 1
 22
E
3

F
 33

9/30/2018 39
 How to get the nominal stress from the Finite
Element Method stress data?
r

The smallest optimum element

size b < ¼r !!

9/30/2018 40
 Loads and stresses in a structure
Fi
F

0 t
Fi-1
Fi+1

 peak,i  f  F i ; f  F  ?
 
 

 n,i  g Fi  ; g F ?

 
 

on o pea
k

9/30/2018
 Loads and Stresses
The load, the nominal stress, the local peak stress and the stress concentration factor

Axial load – linear elastic analysis

F

 peak n  F ;
Anet
Stress

n

 n  kF F;  peak  hFF;

y Analytical, FEM Hndbk
0

kF  1 ; hF  peak
; hF  kF Kt ;
dn Anet F
T

 n, i  kF Fi ;  peak, i  hF Fi ;
F

9/30/2018
 Loads and Stresses
The load, the nominal stress, the local peak stress and the stress concentration factor

M Bending load – linear elastic analysis

M cnet
n 
Stress

 peak ;
I net
n

 y  n  kM M;
0
c
dn k M  net ;   n,i  kM M i;
T I net

 peak,i  hM M i or  peak,i  k M M i K bt
b) M

9/30/2018
Stress Concentration Factors in Fatigue Analysis
The nominal stress and the stress concentration factor in simple load/geometry configurations
Simple axial load
Tension Bending
P P
S  n or S 
M Anet Agross
S
Pure bending load

Stress
peak  peak M cnet M cgross
Stress

n  or S 
n I net I gross
n
 peak  peak
r y r y Kt  or K t 
n S
0 0
net Kt gross Kt
dn dn
T T Kt – stress concentration factor
(net or gross, net Kt S gross Kt !! )
opeak – stress at the notch tip
S
M o n - net nominal stress
S
S - gross nominal stress
9/30/2018
 Stress concentration factors for notched machine components

3.0

2.8
opeak /on

2.6
Stress concentration factor Kt=

2.4

2.2

2.0

1.8 P d b P
1.6
h
1.4 o peak

1.2 o n =P/[(b-d)h]
1.0
0 0.1 0.2 0.3 0.4 0.5 0.6

Diameter-to-width ratio d/b

(B.J. Hamrock et. al., ref.(26)

9/30/2018
 Stress concentration factors for notched machine components

r
3.0

2.8
M
Stress concentration factor Kt=opeak/o

M
2.6 H h

2.4 b
n

2.2 Mc 6M
 peak n  
I bh2
2.0

1.8

1.6 H/h=6
H/h=2
1.4 H/h=1.2
H/h=1.05
1.2
H/h=1.01
1.0
0 0.05 0.10 0.15 0.20 0.25 0.30
Radius-to-height ratio r/h
(B.J. Hamrock et. al.
9/30/2018
Stress concentration & stress distributions in weldments

t1
 pea
k
E
r D
 hs
 
n
M A
t B
C
P
F
C
Various stress distributions in a T-butt weldment with transverse fillet welds;

• Normal stress distribution in the weld throat plane (A),

• Through the thickness normal stress distribution in the weld toe plane (B),
• Through the thickness normal stress distribution away from the weld (C),
• Normal stress distribution along the surface of the plate (D),
• Normal stress distribution along the surface of the weld (E),
• Linearized normal stress distribution in the weld toe plane (F).
9/30/2018 47
Stress concentration factor for a butt weldment
under axial loading
l = hp

g=h
0

P P

t
 W   
0.65
1  exp  0.9 
2h  h 
K tten  1  2 1

 W   2.8  2 r 
W
1  exp  0.45    t  
 2h    

where : W  t  2h  0.6hp
Range of application - reasonably designed weldments, (K.Iida and T. Uemura, ref. 14)
9/30/2018
. 48
 Stress concentration factor for a T-butt
weldment under tension load; (non-load carrying fillet weld)

t1 = tp h
Validated for : 0.02 Š r/t Š0.16 and p
30o Š 8 Š 60o source
, [14]

h
&
x

P P

t
y
 W   
0.65
1  exp  0.9 
2h  h 
K tt  1    1

 W   2.8 W  2 r 
1  exp  0.45    t  
 2h    

where : W  t  2h   0.3 t p  2hp 
9/30/2018
. 49
 Cyclic Loads and Cyclic Stress Patterns
(histories) in Engineering Objects

9/30/2018 50
 Loads and Stresses
The load, the nominal stress, the local peak stress and the stress concentration factor

F Simultaneous axial and bending load

M
 n,i  kFFi  kM Mi;
Stress

 peak
 peak,i  kF Fi K tt  k M M i K tb;
n
or
 y  peak,i  hF Fi  hM Mi;
0

dn kF , kM ; From simple analytical stress analysis

T

K tt , K bt ; From stress concentration handbooks

b) M
F hF , hM ; From detail FEM analysis

9/30/2018
. 51
 The load W and the nominal stress o n in an railway axle
32M b
 n,min  
d3
W a) 1
b)

d N.A.
y 2
W/2
W/2 W/2
B A x 3
RB RA
32M b
Moment Mb
 n,min  
d3 1 cycle
c)
l 1
L
o n,max
o n,a

Stress
S   n   peak  
Mb c b 32M
 ; 3
I d 2 time

W d d3 o n,min
Mb   L  l ; c  ; I  ; 3
4 2 64
Note! In the case of smooth components,
such as the railway axle, the nominal stress and the local peak stress are the same!
9/30/2018
. 52
 Fluctuations and complexity of the stress state at the
notch tip
x3 x3
 33
T
  F  22
F   x2  22 x2

 
2R

T
 23

t
  33

Non-proportional loading path Proportional loading path

o 23 AB C o 23 A B C
o 23
A o 23
A
0 t 0 t
o2
B o 22
2
o 22 0 B o 22 0
C
0 t C 0 t

9/30/2018 53
 How to establish the nominal stress history?
a) The analytical or FE analysis should be carried out for one characteristic load magnitude, i.e.
P=1, Mb =1, T=1 in order to establish the proportionality factors, kP, kM, and kT such that:

  kP P;
P
n   kM Mb;   kT T
M
n
T
n

b) The peak and valleys of the nominal stress history on,,i are determined by scaling the peak and
valleys load history Pi, Mb,I and Ti by appropriate proportionality factors kP, kM, and kT such that:

 n,iP  kP Pi  n,iM  kM Mb,i ; n,iT  kT Ti

c)In the case of proportional loading the normal peak and valley stresses can be added and the
resultant nominal normal stress history can be established. Because all load modes in proportional
loading have the same number of simultaneous reversals the resultant history has also the same
number of resultant reversals as any of the single mode stress history.

n,i  kPPi  kM Mb,i;

d)In the case of non-proportional loading the normal stress histories (and separately the shear
stresses) have to be added as time dependent processes. Because each individual stress history
has different number of reversals the number of reversals in the resultant stress history can be
established after the final superposition of all histories.

 n ti  kP P ti  kM Mb ti 

9/30/2018
. 54
 Superposition of nominal stress histories induced by two
independent loading modes
Two proportional modes of loading Two non-proportional modes of loading

Mode a Mode a
Stress on,a

Stress on,a
time
0 0
Mode b
Mode b
Stress on,b

Stress on,b
time
0 0

Resultant stress: on= on,a+ on,b

Resultant stress:
on(ti)= on,a(ti)+ on,b(ti)
Stress o n

Stress o n

0 0

9/30/2018
© 2008 Grzegorz Glinka. All rights reserved. 55
 Wind load and stress fluctuations in a wind
turbine blade
In-plane bending Out of plane bending
14 14

Wind speed [m/s]

Wind speed [m/s]

12 12

10 10

8 - 28
Load-lag stress [MPa]

Load-lag stress [MPa]

5

- 30
0
0
- 40

-5
- 50
605 610 615 620 600 605 610 615 620
600
Time [s] Time [s]
Source [43]
Note! One reversal of the wind speed results in several stress reversals

Wind speed fluctuations + Blade vibrations ‹ Stress fluctuations

9/30/2018
. 56
Characteristic load/stress history in the aircraft wing skin

a)

b)
0

Flying
Stress

c)

time
0
Taxiing Landing Source [9]

a) Ground loads on the wings, b) Distribution of the wing bending moment induced by the ground
load, c) Stress in the lower wing skin induced by the ground and flight loads

9/30/2018
 Loads and stresses in a structure
Fi
F

0 t
Fi-1
Fi+1

 peak,i  f  F i ; f  F  ?
 
 

 n,i  g Fi  ; g F ?

 
 

on o pea
k

9/30/2018
 Loading and stress histories
Nominal stress history, o n Notch tip stress history, opeak
Load history, F
opeak,max

Stress opeak
on,max

Stress o n
Fmax o peak,
Load F

o n,i i
Fi

o n,i+1 o peak,i+1
0 Fi+1
Time Time Time
0 0
Fi-1 o n,i-1
o peak,i-1

+1

˜n,i i
Characteristic non-dimensional load/stress history

 n,i  kF Fi
i+1
 peak ,i  K t n,i  Kt kF Fi
0
Fi  n,i 
i-1 ˜n,i    peak ,i
F max  n,max  peak ,max
 1  ˜n,i 1
9/30/2018
-1
 How to get the nominal stress o n from the
Finite Element method stress data?
Notched shaft under axial, bending and torsion load

x3

M T M
  F
F   x2
d


D

T
Discrete cross section stress distribution
t obtained from the FE analysis
r x3
on o pea
k

a) Run each load case

separately for an unit o 22
load d 0
b) Linearize the FE stress
field for each load case

9/30/2018
 0 0

   y dy 6   y ydy
   m  b  t
 t
t1 n n n
t t2
n n

y    y y i 6    y yi yi

(y1)  1
1
t t2
(yi)
n  peak x
yi
Determination of
yi nominal stresses
t yn
from discrete FE data
by the linearization
yn method
(yn)
0 0

   y 1dy    ydy    y y

n

i
P
 nm   t
 t
 1
;
1t 1t t 0
t 0
t n
   y 1y dy 6    y y dy 6   y yi yi
cM 2 t
 nb    t
2
 1
;
I 1 t 3 t t2
9/30/2018 12
64
 How to get the resultant stress distribution from the
Finite Element stress data? (Notched shaft under axial, bending load)
x3 x3
o nb

Bending M M
t 23 P
o 22 P o 22 x

d
d

2
0

D
o 33

t
r
x3 x3
o nm o pea
k

Axial Resultant

o 22 o(x3) o 22
0
d
d

9/30/2018 62
Cyclic nominal stress and corresponding fluctuating stress distribution

o n, max Stress on

on, 0 time

on, min
x3

o22(x3,on,min) o 22(x3,o n,0) o22(x3,on,max)

Resultant

o2
0 2
d

9/30/2018 63
 The stress state at the weld toe

• Multiaxial state of
stress at weld toe
• One shear and two  zz
normal stresses  xx
• Due to stress
concentration, oxx is
the largest component
– Predominantly responsible
for fatigue damage xz
zx
 xx
zz

9/30/2018
. 67
 Determination of the nominal, on, and the hot spot
stress , ohs, from 3D-FE stress analysis data
a)
y
b) y

x L L
P x
P P P
 (x,y)
t t
 peak

(x,y)

L/2 0  hs

    x, y dxdy P
n  L/2  t
 - depends on L and is constant along the weld toe line
t L t L
0 0
  x  0, y dy 6   x  0, y ydy Independent of L but it changes
 hs     
m
hs
b
hs
t  t along the weld toe line
t t2
a)Stress distribution in the critical cross section near the cover plate ending and the nominal or the
hot spot stress  n (independent of length L ) and  hs (independent of length L),
b)Stress distribution in the critical plane near the ending of a vertical attachment (gusset) and the
nominal or the hot spot stress  n (dependent on length L ) or  hs (independent of length L)
9/30/2018 65
The Nominal Stress o n versus the local Hot Spot Stress ohs

 hs, A   m
 b ;    m  b ;
hs, A hs,A hs,B hs,B hs,B

 n,A  P ;
t L
 n,B  ??; A
 hs,A y

P
P
B
 hs,B (x,y)

y L
x x

9/30/2018 66
 Example:
Preparation and Analysis of Representative
Stress/Load History:

The Rainflow Cycle Counting Procedure

9/30/2018
.
 Stress/Load Analysis - Cycle Counting Procedure
and Presentation of Results
The measured stress, strain, or load history is given usually in the form of a time series, i.e. a sequence of
discrete values of the quantity measured in equal time intervals. When plotted in the stress-time space the
discrete point values can be connected resulting in a continuously changing signal. However, the time effect
on the fatigue performance of metals (except aggressive environments) is negligible in most cases.
Therefore the excursions of the signal, represented by amplitudes or ranges, are the most important
quantities in fatigue analyses. Subsequently, the knowledge of the reversal point values, denoted with large
diamond symbols in the next Figure, is sufficient for fatigue life calculations. For that reason the intermediate
values between subsequent reversal points can be deleted before any further analysis of the loading/stress
signal is carried out. An example of a signal represented by the reversal points only is shown in slide no. 141.
The fatigue damage analysis requires decomposing the signal into elementary events called ‘cycles’.
Definition of a loading/stress cycle is easy and unique in the case of a constant amplitude signal as that one
shown in the figures. A stress/loading cycle, as marked with the thick line, is defined as an excursion starting
at one point and ending at the next subsequent point having exactly the same magnitude and the same sign
of the second derivative. The maximum, minimum, amplitude or range and mean stress values characterise
the cycle. 1 cycle

   max   min  stressrange o max

 max   min oa
a 
Stress
 stress amplitude om

O
o
2 time
  0
 m  max min  mean stress omin
2
Unfortunately, the cycle definition is not simple in the case of a variable amplitude signal. The only non-
dubious quantity, which can easily be defined, is a reversal, example of which is marked with the thick line in
the Figures below.
9/30/2018
.
 Removing material from a clay mine in Tennessee

9/30/2018
.
9/30/2018
.
 Bending Moment Time Series
20

15

10
Bending moment [10kNm]

0
1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 81
-5

-10

-15

-20

-25

-30

-35
Load point No.

Bending Moment measurements obtained at constant time intervals

9/30/2018
.
 Bending Moment History - Peaks and Valleys
20
Bending moment value [10kNm]
15

10

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
-5

-10

-15

-20

-25

-30

-35
Load point No.

Bending Moment signal represented by the reversal point values

9/30/2018
.
 Constant and Variable Amplitude Stress Histories;
Definition of the Stress Cycle & Stress Reversal

a) Constant amplitude stress history

ma

   max   min ;
x
Stress

mea
n

  max   min
One cycle mi a  
n 2 2

0 Time
   min
b)
Variable amplitude stress history
 m max ;
2
Stress

R
 min
 max
One
reversal Time
0
0
9/30/2018
 Stress Reversals and Stress Cycles in a Variable
Amplitude Stress History

The reversal is simply an excursion between two-consecutive reversal

points, i.e. an excursion between subsequent peak and valley or valley
and peak.

In recent years the rainflow cycle counting method has been accepted
world-wide as the most appropriate for extracting stress/load cycles for
fatigue analyses. The rainflow cycle is defined as a stress excursion,
which when applied to a deformable material, will generate a closed
stress-strain hysteresis loop. It is believed that the surface area of the
stress-strain hysteresis loop represents the amount of damage induced
by given cycle. An example of a short stress history and its rainflow
counted cycles content is shown in the following Figure.

9/30/2018
.
 Stress History and the “Rainflow” Counted Cycles

Stress history Rainflow counted cycles

i
Stress

+1

i-1

i
-2 i
+2

Time

i

A rainflow counted cycle is identified when any two adjacent reversals in the
stress history satisfy the following relation:

ABS  i1   i ABS  i   i1

9/30/2018
The Mathematics of the Cycle Rainflow Counting Method
for Fatigue Analysis of Fluctuating Stress/Load Histories

A rainflow counted cycle is identified when any two adjacent reversals in

thee stress history satisfy the following relation:

ABS  i1   i ABS  i   i1

The stress amplitude of such a cycle is: The stress range of such a cycle is:

ABS  i1   i   ABS  i1   i

a 
2
The mean stress of such a cycle is:

 i1   i
m 
2
9/30/2018
.
 The rainflow cycle counting procedure - example

Determine stress ranges, OSi, and corresponding mean stresses, Smi for the stress history
given below. Use the ‘rainflow’ counting procedure.

Si= 0, 4, 1, 3, 2, 6, -2, 5, 1, 4, 2, 3, -3, 1, -2 (units: MPa•102)

6
5
Stress Si (MPa•102)

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
-1
-2

-3
Reversing point number, i

9/30/2018 77
 The ASTM rainflow counting procedure

1. Find the reversing point with highest absolute stress magnitude,

2. The part of the stress history before the maximum absolute attach to the end of
the history,
3. Perform the rainflow counting on the re-arranged stress history, i.e. from
maximum to maximum

Original stress history

Absolute maximum !

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5

9/30/2018 78
 The ASTM modification of the Stress History
The original stress history

Absolute maximum

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 1 2 3 4 5

6
5
4
The modified 3
stress history 2
1
0
4 6 7 9 10 11 12 13 14 15
-1 1 2 3 5 8
-2
-3
9/30/2018 79
 The Modified Stress History according to the ASTM

6
5

Stress (MPa)x102
4
3
2
1
0
-1 1 2 3 4 5 6 6 8 9 11 12 13 14 15

-2
-3

Start counting from point No. 2 !

6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

starting point
9/30/2018 80
 Start counting from the point No. 2 !!
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 
 6   7  3  2  2.5;
 67   6   7  3  2  1; m,67 2 2

6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 
 4   5  1 4  2.5;
 45   4   5  1  4  3; m,45 2 2
9/30/2018 81
 6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 
 2   3  2  5  1.5;
 23   2   3  2  (5)  7; m,67 2 2

6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 910   9   10  1  (2)  3;  m,910   9  10  1  2  0.5;

2 2
9/30/2018 82
 6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 1314   13   14  3  2  1;  m,1314  13  14  3  2  2.5;

2 2
6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 11   12  4 1  2.5;
 1112   11   12  4  1  3;  m,1112 
2 2
9/30/2018 83
 6
5
4
3
2
1
0
1 2 3 4 5 6 7 8 9 11 12 13 14 15
-1
-2
-3

 
 1   8  6  3  1.5;
 18   1 8  6  (3)  9; m,18 2 2

Cycles counted –ASTM method

1. Oo6- 7 =1; om,6- 7 = 2.5;
2. Oo4- 5 =3; om,4- 5 = 2.5;
3. Oo13- 14=1; om,10- 11= 2.5;
4. Oo11- 12=3; om,11- 12= 2.5;
5. Oo2- 3 =7; om,2- 3 = 1.5;
6. Oo9- 10 =3; om,9- 10 =-0.5;
7. Oo1- 8 =9; om,1- 8 = 1.5;

9/30/2018 84
 Extracted rainflow cycles, Oo- Oom
Mean stress, o m
 m -32 -22 -13 -3.2 6.44 16.1 25.7 35.3 45 54 64.1 73.7 83.3 92.9 103 112 122 131 141 151 
298.8 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1
283.9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
268.9 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1
254 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1
239 0 0 0 0 0 0 0 1 2 2 0 0 0 0 0 0 0 0 0 0 5
Stress range, Oo

224.1 0 0 0 0 0 0 0 0 2 2 1 0 0 0 0 0 0 0 0 0 5
209.2 0 0 0 0 0 0 0 3 4 5 2 0 0 0 0 0 0 0 0 0 14
194.2 0 0 0 0 0 1 0 1 7 2 0 0 0 0 0 0 0 0 0 0 11
179.3 0 0 0 0 0 0 1 0 4 4 0 0 0 0 0 0 0 0 0 0 9
164.3 0 0 0 0 1 0 0 0 3 1 0 0 0 0 0 0 0 0 0 0 5
149.4 0 0 0 0 0 0 1 0 0 0 0 2 1 0 0 0 0 0 0 0 4
134.5 0 0 0 0 0 0 0 0 0 0 0 4 1 1 0 0 0 0 0 0 6
119.5 0 1 1 0 0 0 0 0 0 0 3 1 5 1 2 0 0 0 0 0 14
104.6 0 0 1 2 1 0 0 0 0 2 4 3 7 3 2 1 2 1 0 0 29
89.64 0 1 2 3 7 2 0 0 0 1 2 8 10 7 5 6 2 1 0 0 57
74.7 1 1 3 4 3 5 0 1 2 2 10 18 23 20 17 11 4 1 0 0 126
59.76 2 1 5 7 4 1 4 5 1 2 11 20 34 31 31 28 9 7 1 1 205
44.82 1 6 9 7 9 7 10 3 3 8 15 37 49 64 62 41 16 11 2 1 361
29.88 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
14.94 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
854
Total number of cycles, N=854

9/30/2018 85
 Extracted rainflow cycles, Oo- Oom

9/30/2018 86
 Total number of cycles in the entire history, NT

a) The stress range

max
exceedance diagram
(stress spectrum)

Stress range j

nj=6 - number of cyclesOoj=6
in the class j=6

1 max/max
Relative stress range j/ max

12 3 j=4 j=6 j=7

j=4/max 0
Number cycles N
nj=6/NT
0.5

b) the stress range

12 3 j=4 j=6 j=7 frequency distribution
diagram
0 0.5 1.0
Relative number of cycles N/NT
9/30/2018 87