2020-10-05

Lecture Slides

Chapter 5

Failures Resulting from

Static Loading

The McGraw-Hill Companies © 2012

Static Failure Theories

 Introduction
◦ Predictability of permanent distortion or separation
 Failure Theories
◦ Ductile materials under static loading
 Maximum-Shear-Stress (MSS) theory, Sec. 6-4
 Distortion-Energy (DE) theory, Sec. 6-5
 Ductile Coulomb-Mohr (DCM) theory, Sec. 6-6
◦ Failure of brittle materials under static loading
 Maximum-Normal-Stress (MNS) theory, Sec. 6-8
 Brittle Coulomb-Mohr (BCM) theory and Modifications, Sec.
6-9
◦ Selection
 Fracture Mechanics
 Stochastic Analysis

1
 2020-10-05

Chapter Outline

Shigley’s Mechanical Engineering Design

Failure Examples

Fig. 5–1

 Failure of truck driveshaft spline due to corrosion fatigue

Shigley’s Mechanical Engineering Design

2
 2020-10-05

Failure Examples

Fig. 5–2

 Impact failure of a lawn-mower blade driver hub.

 The blade impacted a surveying pipe marker.

Shigley’s Mechanical Engineering Design

Failure Examples

Fig. 5–3

 Failure of an overhead-pulley retaining bolt on a weightlifting

machine.
 A manufacturing error caused a gap that forced the bolt to take
the entire moment load.

Shigley’s Mechanical Engineering Design

3
 2020-10-05

Failure Examples

Fig. 5–4

 Chain test fixture that failed in one cycle.

 To alleviate complaints of excessive wear, the manufacturer decided to
case-harden the material
 (a) Two halves showing brittle fracture initiated by stress concentration
 (b) Enlarged view showing cracks induced by stress concentration at
the support-pin holes
Shigley’s Mechanical Engineering Design

Failure Examples

Fig. 5–5

 Valve-spring failure caused by spring surge in an oversped engine.

 The fractures exhibit the classic 45 degree shear failure

Shigley’s Mechanical Engineering Design

4
 2020-10-05

Static Strength

 Property of a mechanical element

 Results from material identity, treatment, and processing
 Usually necessary to design using published strength values
 Experimental test data is better, but generally only warranted
for large quantities or when failure is very costly (in time,
expense, or life)
 Methods are needed to safely and efficiently use published
strength values for a variety of situations

Shigley’s Mechanical Engineering Design

Static load & Failure

• Static load

 Unchanged in magnitude, points of application, and direction

 Axial tension or compression, shear, bending, torsion or any
combination of these

• Failure
 a part has separated into two or more pieces
 a part has become permanently distorted, thus ruining its geometry
 a part has had its reliability downgraded
 a part has had its function compromised whatever the reason

Shigley’s Mechanical Engineering Design

5
 2020-10-05

Failure

 Why do parts fail ?

◦ Parts fail because their stresses exceed their strength
 What kind of stresses cause the failure ?
◦ Tensile? Compressive? Shear?
 It depends on
◦ the material
 relative strengths in compression, tension, and shear
◦ the character of the loading
 Static: slowly applied, remain constant w/ time, Chapter 6
 Dynamic: impact loads / fatigue loads, Chapter 7
◦ the presence or absence of cracks in the material

Stress Concentration

 Localized increase of stress near discontinuities

 Kt is Theoretical (Geometric) Stress Concentration Factor

 K t   max  nom

 K ts   max  nom

Shigley’s Mechanical Engineering Design

6
 2020-10-05

Theoretical Stress Concentration Factor

 Graphs available for

standard configurations
 See Appendix A–15 and
A–16 for common
examples
 Many more in Peterson’s
Stress-Concentration
Factors
 Note the trend for higher
Kt at sharper discontinuity
radius, and at greater
disruption

Shigley’s Mechanical Engineering Design

Stress Concentration for Static and Ductile Conditions

 With static loads and ductile materials

◦ Highest stressed fibers yield (cold work)
◦ Load is shared with next fibers
◦ Cold working is localized
◦ Overall part does not see damage unless ultimate strength is
exceeded
◦ Stress concentration effect is commonly ignored for static
loads on ductile materials
 Stress concentration must be included for dynamic loading (See
Ch. 6)
 Stress concentration must be included for brittle materials, since
localized yielding may reach brittle failure rather than cold-
working and sharing the load.

Shigley’s Mechanical Engineering Design

7
 2020-10-05

Need for Static Failure Theories

 Uniaxial stress element (e.g. tension test)

Strength S
n 
Stress 
 Multi-axial stress element
◦ One strength, multiple stresses
◦ How to compare stress state to single strength?

Shigley’s Mechanical Engineering Design

Tensile / Torsion Tests

 Which stress failed the part, the normal stress or the shear
stress ?

ductile brittle Mohr’s circle

Tensile
test

Torsion
test

8
 2020-10-05

Ductile vs. Brittle

Ductility Brittleness
Tensile Test Significant deformation No significant deformation
Significant plastic region No plastic region
Paper clip, low-C steel Matchstick, cast iron
Necking, torn surface No necking, finer surface
% elongation to fracture > 5% No clear yield point
Compression Test No fracture, even material Rough, angled fracture
(tension = compression) (compression > tension)
Bending Test yielding fracture
Torsion Test Wound up for several No significant plastic
revolutions deformation

Identification Sy (= Syt = Syc) Sut , Suc

▶ Same metals can be either ductile or brittle depending on the way

they are manufactured, worked, and heat treated.

9
 2020-10-05

Need for Static Failure Theories

 Failure theories propose appropriate means of comparing multi-

axial stress states to single strength
 Usually based on some hypothesis of what aspect of the stress
state is critical
 Some failure theories have gained recognition of usefulness for
various situations

==> No universal rule

Shigley’s Mechanical Engineering Design

Maximum Normal (Principal) Stress Theory

 Theory: Yielding begins when the maximum principal stress in

a stress element exceeds the yield strength.
 For any stress element, use Mohr’s circle to find the principal
stresses.
 Compare the largest principal stress to the yield strength.
 Often the first theory to be proposed by engineering students.
 Is it a good theory?

Shigley’s Mechanical Engineering Design

10
 2020-10-05

Maximum Normal (Principal) Stress Theory

 Experimental data
shows the theory is
unsafe in the 4th
quadrant.
 This theory is not safe
to use for ductile
materials.

Shigley’s Mechanical Engineering Design

Maximum Shear Stress Theory (MSS)

 Theory: Yielding begins when the maximum shear stress in a

stress element exceeds the maximum shear stress in a tension
test specimen of the same material when that specimen begins to
yield.
 For a tension test specimen, the maximum shear stress is 1 /2.
 At yielding, when 1 = Sy, the maximum shear stress is Sy /2 .
 Could restate the theory as follows:
◦ Theory: Yielding begins when the maximum shear stress in a
stress element exceeds Sy/2.

C.A.Coulomb(propose)Tresca(1864) 
J.J.Guest(experiment)

Shigley’s Mechanical Engineering Design

11
 2020-10-05

Maximum Shear Stress Theory (MSS)

 For any stress element, use Mohr’s circle to find the maximum
shear stress. Compare the maximum shear stress to Sy/2.
 Ordering the principal stresses such that 1 ≥ 2 ≥ 3,

 Incorporating a design factor n

 Or solving for factor of safety

Sy / 2
n
 max

Shigley’s Mechanical Engineering Design

Maximum Shear Stress Theory (MSS)

 To compare to experimental data, express max in terms of

principal stresses and plot.
 To simplify, consider a plane stress state
 Let A and B represent the two non-zero principal stresses, then
order them with the zero principal stress such that 1 ≥ 2 ≥ 3
 Assuming A ≥ B there are three cases to consider
◦ Case 1: A ≥ B ≥ 0
◦ Case 2: A ≥ 0 ≥ B
◦ Case 3: 0 ≥ A ≥ B

Shigley’s Mechanical Engineering Design

12
 2020-10-05

Maximum Shear Stress Theory (MSS)

 Case 1: A ≥ B ≥ 0
◦ For this case, 1 = A and 3 = 0
◦ Eq. (5–1) reduces to A ≥ Sy
 Case 2: A ≥ 0 ≥ B
◦ For this case, 1 = A and 3 = B
◦ Eq. (5–1) reduces to A − B ≥ Sy
 Case 3: 0 ≥ A ≥ B
◦ For this case, 1 = 0 and 3 = B
◦ Eq. (5–1) reduces to B ≤ −Sy

Shigley’s Mechanical Engineering Design

Maximum Shear Stress Theory (MSS)

 Plot three cases on

principal stress axes
 Case 1: A ≥ B ≥ 0
◦ A ≥ Sy
 Case 2: A ≥ 0 ≥ B
◦ A − B ≥ Sy
 Case 3: 0 ≥ A ≥ B
◦ B ≤ −Sy
 Other lines are
symmetric cases
 Inside envelope is
predicted safe zone

Fig. 5–7
Shigley’s Mechanical Engineering Design

13
 2020-10-05

Maximum Shear Stress Theory (MSS)

 Comparison to
experimental data
 Conservative in all
quadrants
 Commonly used for
design situations

Shigley’s Mechanical Engineering Design

Distortion Energy (DE) Failure Theory

 Also known as:

◦ Octahedral Shear Stress
◦ Shear Energy
◦ Von Mises
◦ Von Mises – Hencky

Shigley’s Mechanical Engineering Design

14
 2020-10-05

Distortion Energy (DE) Failure Theory

 Originated from observation that ductile materials stressed

hydrostatically (equal principal stresses) exhibited yield
strengths greatly in excess of expected values.
 Theorizes that if strain energy is divided into hydrostatic
volume changing energy and angular distortion energy, the
yielding is primarily affected by the distortion energy.

Fig. 5–8
Shigley’s Mechanical Engineering Design

Distortion Energy (DE) Failure Theory

 Theory: Yielding occurs when the distortion strain energy per

unit volume reaches the distortion strain energy per unit volume
for yield in simple tension or compression of the same material.

Fig. 5–8

Shigley’s Mechanical Engineering Design

15
 2020-10-05

Deriving the Distortion Energy

 Hydrostatic stress is average of principal stresses

 Strain energy per unit volume,

 Substituting Eq. (3–19) for principal strains into strain energy
equation,

Shigley’s Mechanical Engineering Design

Deriving the Distortion Energy

 Strain energy for producing only volume change is obtained by

substituting av for 1, 2, and 3

 Substituting av from Eq. (a),

 Obtain distortion energy by subtracting volume changing

energy, Eq. (5–7), from total strain energy, Eq. (b)

Shigley’s Mechanical Engineering Design

16
 2020-10-05

Deriving the Distortion Energy

 Tension test specimen at yield has 1 = Sy and 2 = 3 =0

 Applying to Eq. (5–8), distortion energy for tension test
specimen is

 DE theory predicts failure when distortion energy, Eq. (5–8),

exceeds distortion energy of tension test specimen, Eq. (5–9)

Shigley’s Mechanical Engineering Design

Von Mises Stress

 Left hand side is defined as von Mises stress

 For plane stress, simplifies to

 In terms of xyz components, in three dimensions

 In terms of xyz components, for plane stress

Shigley’s Mechanical Engineering Design

17
 2020-10-05

Distortion Energy Theory With Von Mises Stress

 Von Mises Stress can be thought of as a single, equivalent, or
effective stress for the entire general state of stress in a stress
element.
 Distortion Energy failure theory simply compares von Mises
stress to yield strength.

 Introducing a design factor,

 Expressing as factor of safety,

Sy
n


Shigley’s Mechanical Engineering Design

Octahedral Stresses

 Same results obtained by evaluating octahedral stresses.

 Octahedral stresses are identical on 8 surfaces symmetric to the
principal stress directions.
 Octahedral stresses allow representation of any stress situation
with a set of normal and shear stresses.

Principal stress element with single

All 8 octahedral planes showing
octahedral plane showing
Shigley’s Mechanical Engineering Design

18
 2020-10-05

Octahedral Shear Stress

 Octahedral normal stresses are normal to the octahedral

surfaces, and are equal to the average of the principal stresses.
 Octahedral shear stresses lie on the octahedral surfaces.

Fig. 5–10
Shigley’s Mechanical Engineering Design

Octahedral Shear Stress Failure Theory

 Theory: Yielding begins when the octahedral shear stress in a

stress element exceeds the octahedral shear stress in a tension
test specimen at yielding.
 The octahedral shear stress is

 For a tension test specimen at yielding, 1 = Sy , 2 = 3 = 0.

Substituting into Eq. (5–16),

 The theory predicts failure when Eq. (5–16) exceeds

Eq. (5–17). This condition reduces to

Shigley’s Mechanical Engineering Design

19
 2020-10-05

Failure Theory in Terms of von Mises Stress

 Equation is identical to Eq. (5–10) from Distortion Energy

approach
 Identical conclusion for:
◦ Distortion Energy
◦ Octahedral Shear Stress
◦ Shear Energy
◦ Von Mises
◦ Von Mises – Hencky

Sy
n


Shigley’s Mechanical Engineering Design

DE Theory Compared to Experimental Data

 Plot von Mises stress on

principal stress axes to
compare to experimental
data (and to other failure
theories)
 DE curve is typical of data
 Note that typical equates to
a 50% reliability from a
design perspective
 Commonly used for
analysis situations
 MSS theory useful for
design situations where Fig. 5–15
higher reliability is desired
Shigley’s Mechanical Engineering Design

20
 2020-10-05

Shear Strength Predictions

 For pure shear loading, Mohr’s circle shows that A = −B = 
 Plotting this equation on principal stress axes gives load line for
pure shear case
 Intersection of pure shear load line with failure curve indicates
shear strength has been reached
 Each failure theory predicts shear strength to be some fraction of
normal strength

Fig. 5–9 Shigley’s Mechanical Engineering Design

Shear Strength Predictions

 For MSS theory, intersecting pure shear load line with failure
line [Eq. (5–5)] results in

Fig. 5–9
Shigley’s Mechanical Engineering Design

21
 2020-10-05

Shear Strength Predictions

 For DE theory, intersection pure shear load line with failure
curve [Eq. (5–11)] gives

 Therefore, DE theory predicts shear strength as

Fig. 5–9 Shigley’s Mechanical Engineering Design

Example 5-1

Shigley’s Mechanical Engineering Design

22
 2020-10-05

Example 5-1

Shigley’s Mechanical Engineering Design

Example 5-1

Shigley’s Mechanical Engineering Design

23
 2020-10-05

Example 5-1

Shigley’s Mechanical Engineering Design

Example 5-1

Shigley’s Mechanical Engineering Design

24
 2020-10-05

Example 5-1

Shigley’s Mechanical Engineering Design

Example 5-1

Fig. 5−11 Shigley’s Mechanical Engineering Design

25
 2020-10-05

Example 5-1

Fig. 5−11 Shigley’s Mechanical Engineering Design

Even / Uneven Materials

Even material Uneven material

strength compressive  tensile compressive  (shear)  tensile
Tension: microscopic flaw  nuclei for crack
Sys  0.5Sy Compression: flaws are pressed  resistance
for slippage

Many cast materials, gray cast iron,

example fully hardened tool steel
ceramics
f(max)  maximum shear-
failure f(, )
stress theory

Mohr’s
circles

26
 2020-10-05

Mohr Theory

 Some materials have compressive strengths different from

tensile strengths
 Mohr theory is based on three simple tests: tension, compression,
and shear
 Plotting Mohr’s circle for each, bounding curve defines failure
envelope

Fig. 5−12
Shigley’s Mechanical Engineering Design

Coulomb-Mohr Theory

 Curved failure curve is difficult to determine analytically

 Coulomb-Mohr theory simplifies to linear failure envelope using
only tension and compression tests (dashed circles)

Fig. 5−13
Shigley’s Mechanical Engineering Design

27
 2020-10-05

Coulomb-Mohr Theory

 From the geometry, derive

the failure criteria

Fig. 5−13

Shigley’s Mechanical Engineering Design

Coulomb-Mohr Theory

◦ Plane stress problems

 One of principal stress is zero

  1 0
  1   3  0 :   1   1  St
  St S c
     
 1   3   1  0   3 : 1  3  1  1  3  1
  S t S c St Sc
  0 3
 0   1   3 : S  S  1   3   S c
  t c

 1   3

Shigley’s Mechanical Engineering Design

28
 2020-10-05

Coulomb-Mohr Theory

 Incorporating factor of safety

 For ductile material, use tensile and compressive yield strengths

 For brittle material, use tensile and compressive ultimate
strengths

Shigley’s Mechanical Engineering Design

Coulomb-Mohr Theory

 To plot on principal stress axes, consider three cases

 Case 1: A ≥ B ≥ 0 For this case, 1 = A and 3 = 0
◦ Eq. (5−22) reduces to

 Case 2: A ≥ 0 ≥ B For this case, 1 = A and 3 = B

◦ Eq. (5-22) reduces to

 Case 3: 0 ≥ A ≥ B For this case, 1 = 0 and 3 = B

◦ Eq. (5−22) reduces to

Shigley’s Mechanical Engineering Design

29
 2020-10-05

Coulomb-Mohr Theory
 Plot three cases on principal stress axes
 Similar to MSS theory, except with different strengths for
compression and tension

Fig. 5−14
Shigley’s Mechanical Engineering Design

Coulomb-Mohr Theory
 Intersect the pure shear load line with the failure line to
determine the shear strength
 Since failure line is a function of tensile and compressive
strengths, shear strength is also a function of these terms.

 1   3   ,  2  0

S sy  S sy S yt S yc
  1  S sy 
S yt S yc S yt  S yc

Shigley’s Mechanical Engineering Design

30
 2020-10-05

Example 5-2

Shigley’s Mechanical Engineering Design

Example 5-2

Shigley’s Mechanical Engineering Design

31
 2020-10-05

Example 5-3

Fig. 5−16 Shigley’s Mechanical Engineering Design

Example 5-3

Shigley’s Mechanical Engineering Design

32
 2020-10-05

Example 5-3

Shigley’s Mechanical Engineering Design

Example 5-4

Fig. 5−17
Shigley’s Mechanical Engineering Design

33
 2020-10-05

Example 5-4

Fig. 5−17 Shigley’s Mechanical Engineering Design

Example 5-4

Shigley’s Mechanical Engineering Design

34
 2020-10-05

Example 5-4

Shigley’s Mechanical Engineering Design

Failure Theories for Brittle Materials

 Experimental data indicates some differences in failure for brittle
materials.
 Failure criteria is generally ultimate fracture rather than yielding
 Compressive strengths are usually larger than tensile strengths

Fig. 5−19 Shigley’s Mechanical Engineering Design

35
 2020-10-05

Maximum Normal Stress Theory

 Theory: Failure occurs when the maximum principal stress in a

stress element exceeds the strength.
 Predicts failure when

 For plane stress,

 Incorporating design factor,

Shigley’s Mechanical Engineering Design

Maximum Normal Stress Theory

 Plot on principal stress axes

 Unsafe in part of fourth quadrant
 Not recommended for use
 Included for historical comparison

Fig. 5−18 Shigley’s Mechanical Engineering Design

36
 2020-10-05

Brittle Coulomb-Mohr

 Same as previously derived, using ultimate strengths for failure

 Failure equations dependent on quadrant
Quadrant condition Failure criteria

Fig. 5−14

Shigley’s Mechanical Engineering Design

Brittle Failure Experimental Data

 Coulomb-Mohr is
conservative in 4th quadrant
 Modified Mohr criteria
adjusts to better fit the data
in the 4th quadrant

Fig. 5−19

Shigley’s Mechanical Engineering Design

37
 2020-10-05

Modified-Mohr
Quadrant condition Failure criteria

Shigley’s Mechanical Engineering Design

Example 5-5

Fig. 5−16
Shigley’s Mechanical Engineering Design

38
 2020-10-05

Example 5-5

Shigley’s Mechanical Engineering Design

Example 5-5

Shigley’s Mechanical Engineering Design

39
 2020-10-05

Selection of Failure Criteria

 First determine ductile vs. brittle

 For ductile
◦ MSS is conservative, often used for design where higher
reliability is desired
◦ DE is typical, often used for analysis where agreement with
experimental data is desired
◦ If tensile and compressive strengths differ, use Ductile
Coulomb-Mohr
 For brittle
◦ Mohr theory is best, but difficult to use
◦ Brittle Coulomb-Mohr is very conservative in 4th quadrant
◦ Modified Mohr is still slightly conservative in 4th quadrant, but
closer to typical
Shigley’s Mechanical Engineering Design

Selection of Failure Criteria in Flowchart Form

Fig. 5−21
Shigley’s Mechanical Engineering Design

40
 2020-10-05

Introduction to Fracture Mechanics

 Linear elastic fracture mechanics (LEFM) analyzes crack

growth during service
 Assumes cracks can exist before service begins, e.g. flaw,
inclusion, or defect
 Attempts to model and predict the growth of a crack
 Stress concentration approach is inadequate when notch radius
becomes extremely sharp, as in a crack, since stress
concentration factor approaches infinity
 Ductile materials often can neglect effect of crack growth, since
local plastic deformation blunts sharp cracks
 Relatively brittle materials, such as glass, hard steels, strong
aluminum alloys, and steel below the ductile-to-brittle transition
temperature, benefit from fracture mechanics analysis

Shigley’s Mechanical Engineering Design

Quasi-Static Fracture

 Though brittle fracture seems instantaneous, it actually takes

time to feed the crack energy from the stress field to the crack
for propagation.
 A static crack may be stable and not propagate.
 Some level of loading can render a crack unstable, causing it to
propagate to fracture.

Shigley’s Mechanical Engineering Design

41
 2020-10-05

Quasi-Static Fracture

 Foundation work for fracture mechanics established by Griffith

in 1921
 Considered infinite plate with an elliptical flaw
 Maximum stress occurs at (±a, 0)

Fig. 5−22
Shigley’s Mechanical Engineering Design

Quasi-Static Fracture

 Crack growth occurs when energy release rate from applied

loading is greater than rate of energy for crack growth
 Unstable crack growth occurs when rate of change of energy
release rate relative to crack length exceeds rate of change of
crack growth rate of energy

Shigley’s Mechanical Engineering Design

42
 2020-10-05

Crack Modes and the Stress Intensity Factor

 Three distinct modes of crack propagation

◦ Mode I: Opening crack mode, due to tensile stress field
◦ Mode II: Sliding mode, due to in-plane shear
◦ Mode III: Tearing mode, due to out-of-plane shear
 Combination of modes possible
 Opening crack mode is most common, and is focus of this text

Fig. 5−23
Shigley’s Mechanical Engineering Design

Mode I Crack Model

 Stress field on dx dy element at crack tip

Fig. 5−24

Shigley’s Mechanical Engineering Design

43
 2020-10-05

Stress Intensity Factor

 Common practice to define stress intensity factor

 Incorporating KI, stress field equations are

Shigley’s Mechanical Engineering Design

Stress Intensity Modification Factor

 Stress intensity factor KI is a function of geometry, size, and

shape of the crack, and type of loading
 For various load and geometric configurations, a stress intensity
modification factor b can be incorporated

 Tables for b are available in the literature

 Figures 5−25 to 5−30 present some common configurations

Shigley’s Mechanical Engineering Design

44
 2020-10-05

Stress Intensity Modification Factor

 Off-center crack in plate in

longitudinal tension
 Solid curves are for crack tip
at A
 Dashed curves are for tip at B

Fig. 5−25

Shigley’s Mechanical Engineering Design

Stress Intensity Modification Factor

 Plate loaded in longitudinal

tension with crack at edge
 For solid curve there are no
constraints to bending
 Dashed curve obtained with
bending constraints added

Fig. 5−26

Shigley’s Mechanical Engineering Design

45
 2020-10-05

Stress Intensity Modification Factor

 Beams of rectangular cross

section having an edge crack

Fig. 5−27

Shigley’s Mechanical Engineering Design

Stress Intensity Modification Factor

 Plate in tension containing circular hole with two cracks

Fig. 5−28
Shigley’s Mechanical Engineering Design

46
 2020-10-05

Stress Intensity Modification Factor

 Cylinder loaded in axial tension having a radial crack of depth a

extending completely around the circumference

Fig. 5−29
Shigley’s Mechanical Engineering Design

Stress Intensity Modification Factor

 Cylinder subjected to internal

pressure p, having a radial crack
in the longitudinal direction of
depth a

Fig. 5−30

Shigley’s Mechanical Engineering Design

47
 2020-10-05

Fracture Toughness

 Crack propagation initiates when the stress intensity factor

reaches a critical value, the critical stress intensity factor KIc
 KIc is a material property dependent on material, crack mode,
processing of material, temperature, loading rate, and state of
stress at crack site
 Also know as fracture toughness of material
 Fracture toughness for plane strain is normally lower than for
plain stress
 KIc is typically defined as mode I, plane strain fracture toughness

Shigley’s Mechanical Engineering Design

Typical Values for KIc

Shigley’s Mechanical Engineering Design

48
 2020-10-05

Brittle Fracture Factor of Safety

 Brittle fracture should be considered as a failure mode for

◦ Low-temperature operation, where ductile-to-brittle transition
temperature may be reached
◦ Materials with high ratio of Sy/Su, indicating little ability to
absorb energy in plastic region
 A factor of safety for brittle fracture

Shigley’s Mechanical Engineering Design

Example 5-6

Shigley’s Mechanical Engineering Design

49
 2020-10-05

Example 5-6

Fig. 5−25
Shigley’s Mechanical Engineering Design

Example 5-6

Shigley’s Mechanical Engineering Design

50
 2020-10-05

Example 5-7

Shigley’s Mechanical Engineering Design

Example 5-7

Shigley’s Mechanical Engineering Design

51
 2020-10-05

Example 5-7

Fig. 5−26 Shigley’s Mechanical Engineering Design

Example 5-7

Shigley’s Mechanical Engineering Design

52
 2020-10-05

Stochastic Analysis

 Reliability is the probability that machine systems and

components will perform their intended function without failure.
 Deterministic relations between stress, strength, and design
factor are often used due to simplicity and difficulty in acquiring
statistical data.
 Stress and strength are actually statistical in nature.

Shigley’s Mechanical Engineering Design

Probability Density Functions

 Stress and strength are statistical in nature

 Plots of probability density functions shows distributions
 Overlap is called interference of  and S, and indicates parts
expected to fail

Fig. 5−31 (a)

Shigley’s Mechanical Engineering Design

53
 2020-10-05

Probability Density Functions

 Mean values of stress and strength are m and mS

 Average factor of safety is

 Margin of safety for any value of stress  and strength S is

 The overlap area has negative margin of safety

Fig. 5−31 (a)

Shigley’s Mechanical Engineering Design

Margin of Safety

 Distribution of margin of safety is dependent on distributions of

stress and strength
 Reliability R is area under the margin of safety curve for m > 0
 Interference is the area 1−R where parts are expected to fail

Fig. 5−31 (b)

Shigley’s Mechanical Engineering Design

54
 2020-10-05

Normal-Normal Case

 Common for stress and strength to have normal distributions

 Margin of safety is m = S – , and will be normally distributed

 Reliability is probability p that m > 0

 To find chance that m > 0, form the transformation variable of m

and substitute m=0 [See Eq. (20−16)]

 Eq. (5−40) is known as the normal coupling equation

Shigley’s Mechanical Engineering Design

Simple Algebraic Operations

55
 2020-10-05

Normal-Normal Case

 Reliability is given by

 Get R from Table A−10

 The design factor is given by

where

Shigley’s Mechanical Engineering Design

Lognormal-Lognormal Case

 For case where stress and strength have lognormal distributions,

from Eqs. (20−18) and (20−19),

 Applying Eq. (5−40),

Shigley’s Mechanical Engineering Design

56
 2020-10-05

Lognormal-Lognormal Case

 The design factor n is the random variable that is the quotient of

S/
 The quotient of lognormals is lognormal. Note that

 The companion normal to n, from Eqs. (20−18) and (20−19), has

mean and standard deviation of

Shigley’s Mechanical Engineering Design

Lognormal-Lognormal Case

 The transformation variable for the companion normal y

distribution is

 Failure will occur when the stress is greater than the strength,
when n  1 , or when y < 0. So,

 Solving for mn,

Shigley’s Mechanical Engineering Design

57
 2020-10-05

Example 5-8

Shigley’s Mechanical Engineering Design

Example 5-8

Shigley’s Mechanical Engineering Design

58
 2020-10-05

Example 5-8

Shigley’s Mechanical Engineering Design

Example 5-9

Shigley’s Mechanical Engineering Design

59
 2020-10-05

Example 5-9

Shigley’s Mechanical Engineering Design

Interference - General

 A general approach to
interference is needed to
handle cases where the
two variables do not have
the same type of
distribution.
 Define variable x to
identify points on both
distributions

Fig. 5−32
Shigley’s Mechanical Engineering Design

60
 2020-10-05

Interference - General

 Substituting 1– R2 for F2 and

–dR1 for dF1,

 The reliability is obtained by

integrating x from – ∞ to ∞ which
corresponds to integration from 1
to 0 on reliability R1.

Shigley’s Mechanical Engineering Design

Interference - General

where

Shigley’s Mechanical Engineering Design

61
 2020-10-05

Interference - General

 Plots of R1 vs R2
 Shaded area is equal to 1– R, and is obtained by numerical
integration
 Plot (a) for asymptotic distributions
 Plot (b) for lower truncated distributions such as Weibull

Fig. 5−33 Shigley’s Mechanical Engineering Design

62