Fatigue Failure Resulting From Variable Loading

BITSPilani
Pilani Campus
VINAYAK KALLURI
BITSPilani
Pilani Campus
Fat i gue Fai l ur e Res ul t i ng
f r om Var i a bl e Loadi ng
BITS Pilani, Pilani Campus
Variable Loading
Variable loading results when the applied load or
the induced stress on a component is not constant
but changes with time
In reality most mechanical components experience
variable loading due to
-Change in the magnitude of applied load
Example: Extrusion process
-Change in direction of load application
Example: a connecting rod
-Change in point of load application
Example: a rotating shaft
Fatigue
Fatigue is a phenomenon associated with variable
loading or more precisely to cyclic stressing or
straining of a material
ASTMDefinition of fatigue
The process of progressive localized permanent
structural changes occurring in a material subjected
to conditions that produce fluctuating stresses at
some point or points and that may result in cracks or
complete fracture after a sufficient number of
fluctuations.
Fracture Failure- Mechanism
Three stages are involved in fatigue failure
-Crack initiation
-Crack propagation
-Fracture / Rupture
Fracture Failure- Mechanism
Crack propagation modes.
fatigue failure of the crankshaft under pure bending load.
Fatigue failure
Crack initiation, propagation and rupture in a shaft subjected to repeated bending
Introduction to Fatigue in Metals
Crack initiation at
the outer surface
Beach marks
showing the nature
of crack propagation
Final rupture occurs
over a limited area,
characterizing a very
small load required
to cause it
Crack initiation at the root of
keyway at B
Final failure over the small
area at C due to sudden
rupture
Crack propagation occurs
over a time period
Connecting rod failed by fatigue failure
The crack got initiated at the flash line of forging.
Flash line of
forging
Fatigue failure of a steam engine connecting rod due to PURE TENSION load.
No surface crack.
Crack may initiate
anywhere that is the
weakest or unknown
source of weakness.
In this rod, the crack
initiated due to forging
flake slightly below the
centre line.
The crack propagated radially outward
until some time after which the sudden
rupture occurred.
Radial direction of
crack propagation
Approach to Fatigue Failure in Analysis and Design
Fatigue life methods
Fatigue strength and endurance limit
Endurance limit modifying factors
Stress concentration and notch sensitivity
Fluctuating stresses
Combination of loading modes
Variable, fluctuating stresses, cumulative fatigue
damage
Fatigue Life Methods
predict the failure in number of cycles N to failure for a specific type of loading
3 3
10 : (HCF) fatigue cycle High ; 10 1 : (LCF) fatigue cycle Low > N N
Stress life methods
Based on stress levels only
Least accurate of the three, particularly for LCF
It is the most traditional because easiest to implement for a wide range of applications
Has ample supporting data
Represents high cycle fatigue adequately
Strain life methods
Involves more detailed analysis of plastic deformation at localized regions
Good for LCF
Some uncertainties may exist in results because several idealizations get compounded
Hence normally not used in regular practice but only for completeness and special occasions
Linear elastic fracture mechanics methods (LEFM)
Assumes that crack is already present and detected
The crack location is then employed to predict crack growth and sudden rupture with respect to
the stress nature and intensity
Most practical when applied to large structures in conjunction with computer codes and periodic
inspection
Stress Life Method
R. R. Moore high-speed rotating
beam machine.
Fatigue tester
Courtesy: Revolution Industries
pure reversed bending without transverse shear
SFD
BMD
M
b
Stress Life Method
Stress Life Method
Pure bending by means of weights and no transverse shear.
The specimen shown is very carefully machined and polished with
a final polishing in the axial direction to void circumferential
scratches.
Number of revolutions of the specimen required for failure are
recorded.
The first test is made at a stress that is some what under the
ultimate strength of the material.
Next, the test is repeated for a lower load, and so on.
The results are plotted in the S-N diagram, which is either semi-log
or log-log.
The S-N Diagram for steel (UNS G41300), normalized, S
ut
=812 MPa.
Endurance Limit,
It is the stress at which the
component can sustain
infinite number of cycles
Stress Life Method: S-N Diagram
The plot in the S-N diagram never becomes horizontal
for non-ferrous metals and alloys
For non-ferrous metals and alloys, stress at a specific
number of cycles, normally at 5*10
8
cycles, must be
used as fatigue strength
Endurance limit for non-ferrous metals and alloys
For different aluminium alloys (which is non-ferrous)
For non-ferrous metals and alloys, the S-N diagram never becomes horizontal and
hence they do not have endurance limit. Therefore, a stress at a specific number
of cycles, normally at 5*10
8
cycles, must be used as fatigue strength
Stress Life Method: S-N Diagram
Instead of referring to experimental data-bank each time,
it should be possible to quickly estimate the value of
endurance limit using some kind of formula
To enable that, data has been generated for different
types of steels, for endurance limit with respect to the
ultimate tensile strength
This plot seemed to closely follow a combination of two
straight lines, of which the second being almost
horizontal at S
ut
=1460 MPa
Estimation of Endurance Limit
Estimation of Endurance Limit
For steels, Endurance limit :
conditions loading actual in the limit Endurance
bending reverse in obtained limit Endurance
1460 700
1460 5 0
'
'
=
=

>
=
e
e
ut
ut ut
e
S
S
MPa S for MPa
MPa S for S .
S
Endurance limit (Se) is only for rotational bending of
round bar at idealistic conditions (prepared very
carefully and tested under closely controlled conditions).
Endurance strength (S
e
) is for all other types of loading,
geometry and operating conditions.
Endurance limit Vs. Endurance strength
Endurance limit Endurance strength
Endurance limit modifying factors
'
e e d c b a e
S k k k k k S =
strength endurance i.e. use of condition and geometry
the in part machine a of location critical at the limit endurance S
limit endurance specimen test beam - otary S
factor on modificati m
factor r
factor on modificati
factor on modificati load
factor on modificati size
factor on modificati condition surface
e
'
e
=
=
=
=
=
=
=
=
r
effects us iscellaneo k
eliability k
e temperatur k
k
k
k
f
e
d
c
b
a
b
ut a
aS k =
Table 6.2; page:288
Surface condition modification factor (k
a
)
The surface modification factor depends on the quality of the finish of
the actual part surface and on the tensile strength of the part material.
Size modification factor, k
b
( )
1. effect, size no loading axial For
254 51 000837 . 0 859 . 0
51 79 . 2 24 . 1 62 . 7 /
: only torsion and bending in bars CS circular rotating For
107 . 0
107 . 0
=

=
=

b
b
k
mm d if d
mm d if d d
k
etc. section cross channel section, - I r, rectangula circular,
but rotating - non are that bars about What
( ) | |
d d
d A
d d d A
Case
e
e e e
37 . 0 (2) and (1) Equation
), 2 ( 01046 . 0
i.e 0.95d of spacing a having chords parallel two of
outside area the twice is area stress percent 95 the
rounds, solid g nonrotatin For
), 1 ( 0766 . 0 95 . 0
4
circular rotating for : 1
2
95 . 0
2
2
2
95 . 0
=
=
= =
K
K

K
b
for non-rotating shapes
Effective dimension is used
Effective dimension d
e
obtained by equating the volume of
material stressed at and above 95 percent of the maximum stress
to the same volume in the rotating-beam specimen
Table 6-3; page:290
Load modification factor, k
c
=
torsion
axial
bending
k
c
, 59 . 0
, 85 . 0
, 1
Actually the k
c
is dependent on
the S
ut
of the material. Tables 6-
11 to 6-14 (page no. 333) in
Text Book give the details. The
above values are average
values.
Temperature modifying factor, k
d
When operating temperatures are below room
temperature, brittle fracture is a strong possibility and
should be investigated first.
When the operating temperatures are higher than room
temperature, yielding should be investigated first because
the yield strength drops off so rapidly with temperature.
Any stress will induce creep in a material operating at
high temperatures
d
( ) ( ) ( ) ( )
F T
where
T T T T k
o
F
F F F F d
1000 70
10 595 . 0 10 104 . 0 10 115 . 0 10 432 . 0 975 . 0
4 12 3 8 2 5 3

+ + =

145 tests of 21 different carbon and alloy steels results a
fourth-order polynomial curve fit to the data underlying
d
( ) ( ) ( ) ( )
F T
where
T T T T k
o
F
F F F F d
1000 70
10 595 . 0 10 104 . 0 10 115 . 0 10 432 . 0 975 . 0
4 12 3 8 2 5 3

+ + =

Effect of Operating Temperature on the Tensile Strength of Steel.
(S
T
= tensile strength at operating temperature; S
RT
= tensile strength at room temperature)
Table 6.4; page :291
RT
T
d
S
S
k =
a e
z k 08 . 0 1 =
Reliability factor, k
e
Table 65
Accounts for
Residual stress due to shot-peening, hammering
etc.
Corrosion
Coating failure
Spraying etc.
Miscellaneous effects factor, k
f
Estimate the endurance strength of the given material for the following two cases
1. A32 mm diameter shaft made of hot rolled AISI 1040 steel. The shaft surfaces are grinded. The shaft is
subjected to reverse torque and to be used for a part that sees 250C in service at 99.99%reliability.
2. A solid 20 mm side-square rod is cantilevered at one end. The rod supports a completely reverse axial load at
the other end. The material is AISI 1015 cold-drawn steel. The expected reliability of the rod is 90%.
Problem
Two steels are being considered for manufacture of forged
connecting rods. One is AISI 4340 Cr-Mo-Ni steel capable of
being heat-treated to a tensile strength of 1820 MPa. The other
is a plain carbon steel AISI 1040 with an attainable S
ut
of 790
MPa. If each rod is to have a size giving an equivalent diameter
of 20 mm, is there any advantage to using the alloy steel for this
fatigue application?
Problem
For 4340 Cr-Mo-Ni steel, S
ut
= 1820 MPa
For 1040 HR steel, S
ut
= 790 MPa
899 . 0 ) 20 ( 24 . 1 24 . 1
107 . 0 107 . 0
= = =

d k
b
155 . 0 ) 1820 ( 272
995 . 0
= = =
b
ut a
S a k
385 . 0 ) 730 ( 272
995 . 0
= = =
b
ut a
S a k
. 1 ; 1 ; 1 ; 899 . 0 = = = =
e d c b
k k k k
. 1 ; 1 ; 1 = = =
e d c
k k k
MPa S
e
54 . 97 ) 700 ( ) 1 ( ) 1 ( ) 1 ( ) 899 . 0 ( ) 155 . 0 ( = =
MPa S
e
75 . 136 ) 395 ( ) 1 ( ) 1 ( ) 1 ( ) 899 . 0 ( ) 385 . 0 ( = =
The single most influential factor leading to high
possibility of crack initiation
Stress concentration can be due to
Function of geometry (sudden change in
size/diameter; holes in the structure etc.
and surface texture (surface finish, presence of
disintegrations etc.)
Stress concentration
K
t
=Theoretical stress concentration factor
Stress concentration (K
t
)-revised
stress Nominal
stress Maximum
=
t
K
( )
FEM as such simulation numerical
or s experiment through Determined
stress Nominal
max
=
=

=
t
nom t
K
K
t d w
P

d w
K
f
is a reduced value of K
t
and it is also called as fatigue
strength reduction factor
Actual / Fatigue stress concentration factor, K
f
factor) (geometric factor ion concentrat stress l Theoretica
21) - 6 & 20 - 6 Fig. (from y value sensitivit notch
=
=
t
K
q
( ) ( ) 1 1 1 1 + = + =
ts shear fs t f
K q K or K q K
specimen free - notch in stress
specimen notched in stress maximum
=
f
K
Stress-concentration factors for a variety of geometries under different
loading conditions can be found in Table A15 (page:1026-1032)
Notch Sensitivity plot for Steels and UNS A92024-T wrought Al alloys
Fig: 6-20 ; page : 295
(Reverse bending or reverse axial loads)
Fig: 6-21 ; page : 296
Notch Sensitivity plot for Steels and UNS A92024-T wrought Al alloys
(Reversed torsion condition)
Estimation of K
f
K
f
= 1+q(K
t
-1).
When q=0, the material has no sensitivity to notches, and hence K
f
=1.
When q=1, or when notch radius is large for which q is almost equal
to 1, the material has full notch sensitivity, and K
f
= K
t
.
For all grades of cast iron, use q=0.20.
Use the different graphs as given to obtain q for bending/axial and
torsional loading.
Whenever the graphs do not give values of q for certain combinations
of data, use either Neuber equation or Heywood equation.
Use the Neuber equation when the notch is circular/cylindrical.
Estimation of K
f
( )
radius notch
strength. ultimate of function i.e ), (
constant material a is and constant Neuber is a where
1 1
1
1
=
=
+ =
+
=
r
S f a
K q K and
r
a
q
ut
t f
For steel, with S
ut
in kpsi, the Neuber constant can be approximated
by a third-order polynomial fit of data as
3 8 2 5 3
3 8 2 5 3
) 10 ( 67 . 2 ) 10 ( 35 . 1 ) 10 ( 51 . 2 19 . 0 :
) 10 ( 67 . 2 ) 10 ( 51 . 1 ) 10 ( 08 . 3 246 . 0 :
ut ut ut
ut ut ut
S S S a Torsion
S S S a axial or Bending

+ =
+ =
Use Heywood equation when the notch is NOT circular/cylindrical
but is a tranverse hole or shoulder or groove.
Estimation of K
f
( )
335 page 15; - 6 Table the in given are values
1 2
1
a
where
r
a
K
K
K
K
t
t
t
f
+
=
r= hole/ shoulder/groove size
Fluctuating Stress
2
stress e or variabl amplitude
2
stress mean or midrange
stress of range
stress maximum
stress minimum
min max
min max
min max
max
min

= =
+
= =
= =
=
=
a
m
r
Reversed: mean stress is zero;
equal reversals on both sides;
useful in conducting experiments
Repeated: minimum stress is zero;
mean stress equal to half of the
range stress
Fluctuating: maximum, minimum
and mean stress are all non-zero
and arbitrary
Three specific types of cyclic loading:
Design for Infinite life under cyclic loading
f yt
m
e
a
n S S
1
= +

1 = +
yt
m
e
a
S
S
S
S
1 = +
ut
m
e
a
S
S
S
S
1
2
=
|
|
\
|
+
ut
m
e
a
S
S
S
1
2
2
=
|
|
\
|
+
|
|
\
|
yt
m
e
a
S
S
S
S
Different fatigue failure models
yielding) static for
checking for (only line Langer
1
line Elliptic ASME
1
line Gerber
1
line Goodman Modified
1
line Soderberg
1
2 2
2
2
K
K
K
K
K
y yt
m
yt
a
f yt
m
e
a
f ut
m
f
e
a
f ut
m
e
a
f yt
m
e
a
n S S
n S S
n S
n
S
n S S
n S S
= +
|
|
\
|
=
|
|
\
|
+
|
|
\
|
=
|
|
\
|
+
= +
= +

Where
ao f a mo f m
K and K = =
Modified Goodman and
Langer Failure Criteria
Amplitude and Steady Coordinates of Strength and Important
Intersections in First Quadrant
Gerber and Langer
Failure Criteria
ASME-Elliptic and Langer
Failure Criteria
Results based on Smiths tests (72 tests on torsion
strength) and confirmation by Joerres of Associated
Spring-Barnes Group, the following relation is
used to identify the shear ultimate strength from
tensile strength.
From distortion energy theory,
ut su
S 0.67 S =
yt sy
S 0.577 S =
Torsional Fatigue Strength
An AISI 1020 cold drawn steel shaft of diameter 30
mm is welded to a fixed support at one end and
another end is subjected to a bending moment
varying from 25 Nm to 100 Nm. Fatigue stress
concentration factor under bending (K
f
) is 1.6. Find
the factor of safety using the Gerber failure criteria.
Problem:
( )
( )
( )
( )
.
38 . 6
1
470
73 . 37
196.8
64 . 22
1
S S
equation; failure fatigue Gerber
MPa 196.8 = S k k k = S
1 = k
0.958 30) 1.24(0.37 = ) 1.24(d = k
drawn) cold for -0.265 = b and 4.51 = (a 0.874 = aS = k
MPa 235 = 0.5S = S
64 . 22 15 . 14 6 . 1
73 . 37 58 . 23 6 . 1
15 . 14
58 . 23
43 . 9
30
10 25 32 32
72 . 37
30
10 100 32 32
390 470 S ; Steel CD 1020 AISI For
2
2
ut
m
e
a
1
e c b a e
c
0.107 - 0.107 -
e b
b
ut a
ut
1
e
3
3
3
min
3
3
3
max
ut
safety factor high with designed over is member The
n
n
n
n
n
MPa K
MPa K
MPa stress bending Variable
MPa stress bending Mean
MPa
d
M
stress bending Minimum
MPa
d
M
stress bending Maximum
MPa S and MPa
f
f
f
f
f
ba f a
bm f m
ba
bm
y
= =
|
\
|
+
=
|
|
\
|
+
=
= = =
= = =
= =
= =
=
= =
=
= =
= =

Different types of cyclic loads may be applied in
combination, for example, bending, axial and torsional
on machine components
When the loads and in-phase, the maximum values of
loads occurs at the same time and so are the minimum
values.
Hence in such cases, we can estimate the maximum and
minimum von-Mises stress values and then estimate the
mean and amplitude von-Mises stresses. Then fatigue
criterion may be applied.
Combination of loading modes
Combined loading
For the common case of a shaft with bending stresses, torsional shear
stresses, and axial stresses, the von Mises stress is
Considering that the bending, torsional, and axial stresses have
alternating and midrange components, the von Mises stresses for the
two stress elements can be written as
For plane stress
| |
2
1
2
2 2
'
3
xy y y x x
+ + =
| |
2
1
2
2
'
3
xy x
+ =
( ) ( ) ( )
( )
( ) ( ) | |
( ) ( ) ( ) ( ) | | ( ) ( ) | | { }
2 / 1
2 2
'
2 / 1
2
2
'
3
3
85 . 0
torsion
mo
torsion
fs
axial
mo
axial
f
bending
mo
bending
f m
torsion
ao
torsion
fs
axial
ao
axial
f
bending
ao
bending
f a
K K K
K K K

+ + =
+
(
+ =
Calculate von Mises stresses for alternating and
midrange stress states,
a
and
m
.
Apply stresses to fatigue criterion i.e Soderberg,
Modified-Goodman, Gerbers or ASME Elliptic criteria
by replacing
a
and
m
with
a
and
m
respectively
Conservative check for localized yielding using von
Mises stresses i.e
Design for Combined loading
Fig shows clutch-testing machine. The steel shaft rotates at a constant speed
. An axial load is applied to the shaft and is cycled from zero to P. The
torque T induced by the clutch face onto the shaft
where D and d are defined in the figure and f is the coefficient of friction of
the clutch face. The shaft is machined with S
y
= 800 MPa and S
ut
= 1000
MPa. The theoretical stress concentration factors for the fillet are 3.0 and 1.8
for the axial and torsional loading, respectively.
Assume the load variation P is synchronous
with shaft rotation. With f = 0.3, find the
maximum allowable load P such that the shaft
will survive a minimum of 10
8
cycles with a
factor of safety of 3. Use the modified
Goodman criterion.
Prob:6-57
Solution
Solution
Practice problem
In the figure shaft A, made of AISI 1010 hot-rolled steel, is welded to a
fixed support and is subjected to loading by equal and opposite forces F via
shaft B. In addition, the shaft is also subjected to a study compressive force
of P. The length of shaft A from the fixed support to the connection at shaft
B is 1 m. If the load F cycles from 0.5 to 2kN and P= 1 kN, For shaft Afind
the factor of safety for infinite life using (a) the modified Goodman fatigue
failure criterion. (b) Gerber fatigue failure criterion.
Strain-Life method is used to estimate fatigue
strength for high cycle fatigue (i.e. N>10
3
cycles)
Fatigue strength
If a completely reversed stress(
rev
) in given, the
number of cycles to failure can be expressed as
( )
|
|
\
|
= =
=
e
ut
e
ut
b
f
S
fS
b
S
fS
a
where N a S
log
3
1
&
2
Fatigue strength calculation / Design for finite life
b
rev
a
N
1
|
\
|
=

Where f is fatigue strength
fraction depends on S
ut
from
Figure 6-18; page:285
Fatigue strength fraction
Figure 6-18; page:285
A 25-mm-diameter solid round bar has a groove 2.5-mm
deep with a 2.5-mm radius machined into it. The bar is
made of AISI 1020 CD steel and is subjected to a purely
reversing torque of 200 N-m. For the S-N curve of this
material, let f = 0.9.
(a) Estimate the number of cycles to failure.
(b) If the bar is also placed in an environment with a
temperature of 450
0
F, estimate the number of cycles to
failure.
Problem : 6-12
D= 25 mm ; d= 20 mm ; r= 2.5 mm
r/d =2.5/20 = 0.125; D/d= 1.25
From figure: K
ts
= 1.4
Solution
MPa S and kpsi MPa
y
390 ) 68 ( 470 S ; Steel CD 1020 AISI For
ut
= =
Solution
( )
cycles
a
N
S
fS
b
S
fS
a
MPa
b
a
se
su
se
su
22472
169 . 0 log
3
1
94 . 913
315 S 0.67 S
MPa 87.94 = S k k k = S
0.59 = k
0.726 0) 2 1.24( = ) 1.24(d = k
drawn) cold for -0.265 = b and 4.51 = (a 0.874 = aS = k
MPa 235 = 0.5S = S
1
2
ut su
1
e c b a e
c
0.107 - 0.107 -
e b
b
ut a
ut
1
e
=
|
\
|
=
=
|
|
\
|
=
= =
= =
=
Solution

Fatigue Failure Resulting From Variable Loading

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Fatigue Failure Resulting From Variable Loading

Uploaded by

Copyright:

Available Formats

BITSPilani

You might also like