CHAPTER 4: THEORY OF

NUR AIN BT MOHD ZAINUDDIN FAILURE

nurain1465@uitm.edu.my
0166601492
Faculty of Chemical Engineering
Credit to: NFAK
 COURSE LEARNING OUTCOME

• At the end of this chapter, student should be able to:

1. Explain the concept of failure for materials.

2. Determine the failure for ductile and brittle materials.
3. Determine the failure of materials due to Creep and
fatigue.

Chapter 5 – Theory of Failure 2

5.1 INTRODUCTION TO FAILURE
5.1.1 Introduction
• Failure for ductile material is by yielding, while brittle material
is by fracture.
• Failure depend on temperature, rate of loading, chemical
environment and material of construction.
Fracture Very Moderately Brittle
behavior: Ductile Ductile

%AR or %EL Large Moderate Small

• Ductile Ductile: Brittle: No
fracture is usually warning before warning
desirable! fracture
Chapter 5 – Theory of Failure 3
• Metals can fail by brittle or ductile fracture.
• Ductile fracture is better than brittle fracture » ductile fracture
occurs over a period of time, where as brittle fracture is fast
and can occur (with flaws) at lower stress levels than a ductile
fracture.

Stress strain curve for brittle and ductile material

Chapter 5 – Theory of Failure 4
5.1.2 Ductile Failure
• Fracture results in separation of stressed solid into two or
more parts.
• Ductile fracture: High plastic deformation & slow crack
propagation.
• Three steps:
 Specimen forms neck and cavities within neck.
 Cavities form crack and crack propagates towards
surface, perpendicular to stress.
 Direction of crack changes to 45 resulting in cup-cone
fracture.

Chapter 5 – Theory of Failure 5

5.1.3 Brittle Failure
• No significant plastic deformation before fracture.
• Common at high strain rates and low temperature.
• Three stages:
 Plastic deformation concentrates dislocation along slip
planes.
 Microcracks nucleate due to shear stress where
dislocations are blocked.
 Crack propagates to fracture.
Example: HCP Zinc ingle crystal under high stress along
{0001} plane undergoes brittle fracture.

Chapter 5 – Theory of Failure 6

• Brittle fractures are due to defects like
 Folds
 Undesirable grain flow
 Porosity
 Tears and Cracks
 Corrosion damage
 Embrittlement due to atomic hydrogen

• At low operating temperature, ductile to brittle transition

takes place.

Chapter 5 – Theory of Failure 7

• Example: failure of a Pipe

• Ductile failure:
 one piece
 large deformation

• Brittle failure:
 Many pieces
 small deformation

Chapter 5 – Theory of Failure 8

• Moderately Ductile Failure

void void growth shearing

necking fracture
nucleation and linkage at surface
s

• Resulting 50
50mm
mm
fracture
surfaces
(steel)
100 mm
particles
Analysis of Fracture surface of
serve as void
Metallurgical Failures tire cord wire
nucleation
loaded in tension.
sites.
Chapter 5 – Theory of Failure 9
• Ductile Vs Brittle Fracture

Ductile fracture brittle fracture

(cup-and-cone fracture)

Chapter 5 – Theory of Failure 10

5.2 FAILURE OF DUCTILEMATERIAL
5.2.1 Maximum Shear StressTheory
• For yielding of a ductile material, it occurs along the contact
planes.
• Maximum-shear-stress theory or Tresca yield criterion is used
to predict the failure stress of a ductile material subjected to
any type of loading.
• Maximum shear stress theory states that yielding of
material begin when absolute maximum shear stress reaches
the shear stress that cause the same material to yield when
subjected to only axial tension.

Chapter 5 – Theory of Failure 11

• With reference from plane stress, the
maximum-shear-stress theory for plane stress can be
expressed for any two in-plane principal stresses.

1  Y 
  1 , 2 have same signs
 2   Y 
 1   2   Y   1, 2 have oppositesigns

Chapter 5 – Theory of Failure 12

• To avoid failure, max shear stress theory, abs max < or
equal to σy/2, where σy is obtain from test.
• In application, absolute maximum shear stress will be express
in terms of principal stress.
• Failure occur if in plane principal stress are opposite sign.

 max   min
 abs 
max 2

• The failure occur out of the plane if they are same sign.

 max
 abs 
max 2

Chapter 5 – Theory of Failure 13

5.2.2 Maximum Distortion Energy Theory
• The energy per unit volume of material is called the strain-
energy density.
1
u  
2
• Refer to 10.31a, each principal stress contributes a portion to
a total strain-energy density.

1
u 1 1  1 2 2  1 3 3
2 2 2
• Knowing that

 x  1 x vy   z  ,  y  1  y vx   z  ,  z  1  z v x   y 

E E E
Chapter 5 – Theory of Failure 14
• Substitute into above resulting

u
1
2E
 1 2   1 3   3 2 )
12   22   23  2v(  

• 1904-M. Huber proposed that yielding in a ductile material

occurs;
• When the distortion per unit volume of the same material
equal or exceeded the distortion energy per unit volume of the
same material when it subjected to yielding in a simple
tension test.

Chapter 5 – Theory of Failure 15

• Then knowing that
 1  (1   avg )
 2  ( 2   avg )
 3  ( 3   avg )
and
 avg  ( 1   2   3 ) / 3
• Therefore

ud 
1v
6E

(1   2 )2  ( 3   1 )2  ( 2   3 ) 2 
• If plane stress

ud 
1 v 2
3E

1  21   2
2 2

Chapter 5 – Theory of Failure 16
• Unaxial tension test
1   Y , 2   3  0
• Therefore
(u d )Y 
1 v
3E
 
Y 2
• With

(ud )Y ud
• So

 1
2
  2 12   2   Y2
2

• This method is 15% than Tresca theory
Chapter 5 – Theory of Failure 17
5.3 FAILURE FOR BRITTLEMATERIAL
5.3.1 Brittle Materials
• Maximum-normal stress theory states that a brittle material
will fail when maximum principal stress is equal to the
ultimate normal stress.
• Brittle material will fail when the maximum principal stress in
the material reaches a limiting value that is equal to ultimate
normal stress the material can sustain when it is subjected to
simple tension.

1   ult
 2   ult

Chapter 5 – Theory of Failure 18

5.3.2 Maximum Normal StressTheory
• Brittle material such as gray iron, tend to fail by fracture with
no apparent yielding.
• Fractures will occur when normal stress reaches the ultimate
stress, σult.
• Brittle material will fail when the maximum principal stress σ1
in the material reaches a limiting value that is equal to
ultimate normal stress.

1  ult
 2  ult 10 - 31

Chapter 5 – Theory of Failure 19

• Figure shows how brittle materials fail.

Chapter 5 – Theory of Failure 20

Example 5.1:
• The solid shaft has a radius of 0.5 cm and is made of steel
having a yield stress of σ = 360 MPa. Determine if the loadings
cause the shaft to fail according to the maximum-shear-stress
theory and the maximum-distortion-energy theory.

Solution:
• Since maximum shear stress caused by the torque,
• we have

P 15
x    19.10 kN/cm2  195 MPa
A  0.5

 xy  Tc  3.250.5  16.55 kN/cm2  165.5 MPa

J  0.54
2

Chapter 5 – Theory of Failure 21

• Principal stresses can also be obtained using the stress-
transformation equations,

 x   y  2  xy 2
1,2   x   y  
 
2  2 
 σ1  95.6 MPa and σ 2  286.6 MPa
• Since the principal stresses have opposite signs, the absolute
maximum shear stress will occur in the plane,

1   2   Y
95.6   286.6  360
382.2  360

Chapter 5 – Theory of Failure 22

• Thus, shear failure of the material will occur according
to this theory.
• Using maximum-distortion-energy theory,

 2
1  
1 2   2
2   2
Y 
95.6  95.6 286.6   286.6  360
2 2 2

118677.9 129600

• Using this theory, failure will not occur.

Chapter 5 – Theory of Failure 23

5.3 CREEP AND FATIGUE
5.3.1 Creep
• When material support a load for long period of time, it will
deform until a sudden fracture occurs.
• This time-dependent permanent deformation is known as
creep.
• Both stress and/or temperature play a significant role in the
rate of creep.
• Creep strength will decrease for higher temperatures or
higher applied stresses.
• Creep strength represents the highest initial stress the
material can withstand during given time without causing
specified creep strain.

Chapter 5 – Theory of Failure 24

• Simple method to
determine creep
strength
• Measure time taken to
produce allowable
strain or rupture
strain for each
specimen
• Plot stress vs. strain
• Creep strength
inversely proportional
to temperature and
applied stresses

Chapter 5 – Theory of Failure 25

• Creep test determines the effect of temperature and stre ss on
creep rate.
• Metals are tested at constant stress at different temperature
or constant temperature with different stress.

High temperature
or stress

Medium temperature
or stress
Low temperature Creep strength: Stress to produce
or stress Minimum creep rate of 10 -5
%/h at a given temperature.

Chapter 5 – Theory of Failure 26

• Creep is progressive deformation under constant stress.
• Important in high temperature applications.

• Primary creep: creep rate

decreases with time due
to strain hardening.
• Secondary creep: Creep
rate is constant due to
simultaneous strain hard-
ening and recovery process.
• Tertiary creep: Creep rate
increases with time leading
to necking and fracture.

Chapter 5 – Theory of Failure 27

• Sample deformation at a constant stress (s)
 Primary Creep: slope (creep rate) decreases with time.
 Secondary Creep: steady-state i.e., constant slope.
 Tertiary Creep: slope (creep rate) increases with time, i.e.
acceleration of rate.



0 t

Chapter 5 – Theory of Failure 28

• Creep rupture test is same as creep test but aimed at
failing the specimen.
• Plotted as log stress versus log rupture time.
• Time for stress rupture
decreases with increased
stress and temperature.

Chapter 5 – Theory of Failure 29

• Larsen Miller parameter is used to represent creep-stres s
rupture data.
P (Larsen-Miller) = T [ log tr + C ]
T = temperature(K), tr = stress-rupture time h
C = Constant (order of 20)

• Also, P(Larsen-Miller) = [T (0C) + 273 (20+log tr)

or P(Larsen-Miller) = [T (0F) + 460 (20+log tr)

• At a given stress level, the log time to stress rupture plus

constant multiplied by temperature remains constant for a
given material.

Chapter 5 – Theory of Failure 30

• If two variables of time to
rupture, temperature and
stress are known, 3rd parameter
that fits L.M. parameter can
be determined.

Example 5.2:
• For alloy CM, at 207 MPa,
LM parameter is 27.8 x 103 K
Then if temperature is known,
time to rupture can be found.

Chapter 5 – Theory of Failure 31

Example 5.3:
• Calculate time to cause 0.2% creep strain in gamma
Titanium aluminide at 40 KSI and 1200F.

Solution:
• From fig, p = 38000
38000 = (1200 + 460) (log t0.2% + 20) t = 776 h
Chapter 5 – Theory of Failure 32
5.3.2 Fatigue
• Defined as a metal subjected to repeated cycles of stress and
strain, breaking down structurally, before fracturing.
• Needs to be accounted for in design of connecting rods (e.g.
steam/gas turbine blades, connections/supports for bridges,
railroad wheels/axles and parts subjected to cyclic loading).
• Fatigue occurs at a stress lesser than the material’s yield
stress.
• Also referred to as the endurance limit (el) or fatigue limit.

5.3.2 (a) Fatigues Testing

• Alternating compression and tension load is applied on
specimens towards center.

Chapter 5 – Theory of Failure 33

• Failure » specimens under cyclic stress (subjected to
dynamic and fluctuating stresses).
• Plot stress (S) against number of cycles-to-failure N (S-N
diagram) on logarithmic scale.
• Key parameters are stress amplitude (S), mean stress (m) and
frequency.
specimen compression on top

motor counter
bearing bearing

flex coupling
tension on bottom

Chapter 5 – Theory of Failure 34

• Fatigue...
» can cause part failure, even though smax < sc.
» causes ~ 90% of mechanical engineering failures.

Chapter 5 – Theory of Failure 35

• Fluctuating stress cycles can be categorized by the
below parameters:
 max   min
Mean stress = m 
2

Stress range =  r   max   min

 max   min
Stress amplitude = a 
2

 min
Stress ratio = R 
 max

Chapter 5 – Theory of Failure 36

5.3.2 (b) Fatigue Design Parameter
S = stress amplitude
case for
• Fatigue limit, Sfat : unsafe steel (typ.)
» no fatigue if S < Sfat
Sfat
safe

10 105 107 109

N = Cycles to failure

• Sometimes, the S = stress amplitude case for

unsafe Al (typ.)
fatigue limit is zero!

safe

103 105 107 109

N = Cycles to failure
Chapter 5 – Theory of Failure 37
• Metals often fail at much lower stress at cyclic loading
compared to static loading.
• Crack nucleates at region of stress concentration and
propagates due to cyclic loading.
• Failure occurs when cross sectional area of the metal too
small to withstand applied load.

Fracture started
here

Fatigue fractured surface of keyed

shaft
Chapter 5 – Theory of Failure 38
Final rupture