**preview**has

**blurred**sections. Sign up to view the full version! View Full Document

**Unformatted text preview: **Mechanical Engineering
Shigleys Mechanical Engineering Design,
Eighth Edition
BudynasNisbett
=>?
McGraw-Hill
McGrawHill Primis
ISBN: 0390764876
Text:
Shigleys Mechanical Engineering Design,
Eighth Edition
BudynasNisbett
This book was printed on recycled paper.
Mechanical Engineering
http://www.primisonline.com
Copyright ©2006 by The McGrawHill Companies, Inc. All rights
reserved. Printed in the United States of America. Except as
permitted under the United States Copyright Act of 1976, no part
of this publication may be reproduced or distributed in any form
or by any means, or stored in a database or retrieval system,
without prior written permission of the publisher.
This McGrawHill Primis text may include materials submitted to
McGrawHill for publication by the instructor of this course. The
instructor is solely responsible for the editorial content of such
materials.
111
0192GEN
ISBN: 0390764876
Mechanical
Engineering
Contents
BudynasNisbett Shigleys Mechanical Engineering Design, Eighth Edition
Front Matter
1
Preface
List of Symbols
1
5
I. Basics
8
Introduction
1. Introduction to Mechanical Engineering Design
2. Materials
3. Load and Stress Analysis
4. Deflection and Stiffness
8
9
33
72
1 45
II. Failure Prevention
208
Introduction
5. Failures Resulting from Static Loading
6. Fatigue Failure Resulting from Variable Loading
208
2 09
2 60
III. Design of Mechanical Elements
349
Introduction
7. Shafts and Shaft Components
8. Screws, Fasteners, and the Design of Nonpermanent Joints
9. Welding, Bonding, and the Design of Permanent Joints
10. Mechanical Springs
11. RollingContact Bearings
12. Lubrication and Journal Bearings
13. Gears General
14. Spur and Helical Gears
15. Bevel and Worm Gears
16. Clutches, Brakes, Couplings, and Flywheels
17. Flexible Mechanical Elements
18. Power Transmission Case Study
349
3 50
3 98
4 60
IV. Analysis Tools
928
Introduction
19. FiniteElement Analysis
20. Statistical Considerations
928
9 29
9 52
iii
5 01
5 50
5 97
6 52
7 11
7 62
8 02
8 56
9 09
Back Matter
978
Appendix A: Useful Tables
Appendix B: Answers to Selected Problems
Index
iv
9 78
1 034
1039
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
Preface
© The McGrawHill
Companies, 2008
1
Preface
Objectives
This text is intended for students beginning the study of mechanical engineering
design. The focus is on blending fundamental development of concepts with practical specification of components. Students of this text should find that it inherently
directs them into familiarity with both the basis for decisions and the standards of
industrial components. For this reason, as students transition to practicing engineers,
they will find that this text is indispensable as a reference text. The objectives of the
text are to:
Cover the basics of machine design, including the design process, engineering mechanics and materials, failure prevention under static and variable loading, and characteristics of the principal types of mechanical elements.
Offer a practical approach to the subject through a wide range of real-world applications and examples.
Encourage readers to link design and analysis.
Encourage readers to link fundamental concepts with practical component specication.
New to This Edition
This eighth edition contains the following signicant enhancements:
New chapter on the Finite Element Method. In response to many requests from
reviewers, this edition presents an introductory chapter on the nite element method.
The goal of this chapter is to provide an overview of the terminology, method, capabilities, and applications of this tool in the design environment.
New transmission case study. The traditional separation of topics into chapters
sometimes leaves students at a loss when it comes time to integrate dependent topics
in a larger design process. A comprehensive case study is incorporated through standalone example problems in multiple chapters, then culminated with a new chapter
that discusses and demonstrates the integration of the parts into a complete design
process. Example problems relevant to the case study are presented on engineering
paper background to quickly identify them as part of the case study.
Revised and expanded coverage of shaft design. Complementing the new transmission case study is a signicantly revised and expanded chapter focusing on issues relevant to shaft design. The motivating goal is to provide a meaningful presentation that
allows a new designer to progress through the entire shaft design process from general shaft layout to specifying dimensions. The chapter has been moved to immediately follow the fatigue chapter, providing an opportunity to seamlessly transition
from the fatigue coverage to its application in the design of shafts.
Availability of information to complete the details of a design. Additional focus is
placed on ensuring the designer can carry the process through to completion.
xv
2
xvi
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
Preface
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
By assigning larger design problems in class, the authors have identied where the
students lack details. For example, information is now provided for such details as
specifying keys to transmit torque, stress concentration factors for keyways and retaining ring grooves, and allowable deections for gears and bearings. The use of internet catalogs and engineering component search engines is emphasized to obtain
current component specications.
Streamlining of presentation. Coverage of material continues to be streamlined to
focus on presenting straightforward concept development and a clear design procedure for student designers.
Content Changes and Reorganization
A new Part 4: Analysis Tools has been added at the end of the book to include the new
chapter on nite elements and the chapter on statistical considerations. Based on a survey of instructors, the consensus was to move these chapters to the end of the book
where they are available to those instructors wishing to use them. Moving the statistical chapter from its former location causes the renumbering of the former chapters 2
through 7. Since the shaft chapter has been moved to immediately follow the fatigue
chapter, the component chapters (Chapters 8 through 17) maintain their same numbering. The new organization, along with brief comments on content changes, is given
below:
Part 1: Basics
Part 1 provides a logical and unied introduction to the background material needed for
machine design. The chapters in Part 1 have received a thorough cleanup to streamline
and sharpen the focus, and eliminate clutter.
Chapter 1, Introduction. Some outdated and unnecessary material has been removed.
A new section on problem specication introduces the transmission case study.
Chapter 2, Materials. New material is included on selecting materials in a design
process. The Ashby charts are included and referenced as a design tool.
Chapter 3, Load and Stress Analysis. Several sections have been rewritten to improve clarity. Bending in two planes is specically addressed, along with an example
problem.
Chapter 4, Deection and Stiffness. Several sections have been rewritten to improve
clarity. A new example problem for deection of a stepped shaft is included. A new
section is included on elastic stability of structural members in compression.
Part 2: Failure Prevention
This section covers failure by static and dynamic loading. These chapters have received
extensive cleanup and clarication, targeting student designers.
Chapter 5, Failures Resulting from Static Loading. In addition to extensive cleanup
for improved clarity, a summary of important design equations is provided at the end
of the chapter.
Chapter 6, Fatigue Failure Resulting from Variable Loading. Confusing material on
obtaining and using the S-N diagram is claried. The multiple methods for obtaining
notch sensitivity are condensed. The section on combination loading is rewritten for
greater clarity. A chapter summary is provided to overview the analysis roadmap and
important design equations used in the process of fatigue analysis.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
Preface
3
© The McGrawHill
Companies, 2008
Preface
xvii
Part 3: Design of Mechanical Elements
Part 3 covers the design of specic machine components. All chapters have received
general cleanup. The shaft chapter has been moved to the beginning of the section. The
arrangement of chapters, along with any signicant changes, is described below:
Chapter 7, Shafts and Shaft Components. This chapter is signicantly expanded and
rewritten to be comprehensive in designing shafts. Instructors that previously did not
specically cover the shaft chapter are encouraged to use this chapter immediately
following the coverage of fatigue failure. The design of a shaft provides a natural progression from the failure prevention section into application toward components. This
chapter is an essential part of the new transmission case study. The coverage of
setscrews, keys, pins, and retaining rings, previously placed in the chapter on bolted
joints, has been moved into this chapter. The coverage of limits and ts, previously
placed in the chapter on statistics, has been moved into this chapter.
Chapter 8, Screws, Fasteners, and the Design of Nonpermanent Joints. The section on setscrews, keys, and pins, has been moved from this chapter to Chapter 7.
The coverage of bolted and riveted joints loaded in shear has been returned to this
chapter.
Chapter 9, Welding, Bonding, and the Design of Permanent Joints. The section on
bolted and riveted joints loaded in shear has been moved to Chapter 8.
Chapter 10, Mechanical Springs.
Chapter 11, Rolling-Contact Bearings.
Chapter 12, Lubrication and Journal Bearings.
Chapter 13, Gears General. New example problems are included to address design
of compound gear trains to achieve specied gear ratios. The discussion of the relationship between torque, speed, and power is claried.
Chapter 14, Spur and Helical Gears. The current AGMA standard (ANSI/AGMA
2001-D04) has been reviewed to ensure up-to-date information in the gear chapters.
All references in this chapter are updated to reect the current standard.
Chapter 15, Bevel and Worm Gears.
Chapter 16, Clutches, Brakes, Couplings, and Flywheels.
Chapter 17, Flexible Mechanical Elements.
Chapter 18, Power Transmission Case Study. This new chapter provides a complete
case study of a double reduction power transmission. The focus is on providing an example for student designers of the process of integrating topics from multiple chapters. Instructors are encouraged to include one of the variations of this case study as a
design project in the course. Student feedback consistently shows that this type of
project is one of the most valuable aspects of a rst course in machine design. This
chapter can be utilized in a tutorial fashion for students working through a similar
design.
Part 4: Analysis Tools
Part 4 includes a new chapter on nite element methods, and a new location for the
chapter on statistical considerations. Instructors can reference these chapters as needed.
Chapter 19, Finite Element Analysis. This chapter is intended to provide an introduction to the nite element method, and particularly its application to the machine
design process.
4
xviii
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
Preface
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Chapter 20, Statistical Considerations. This chapter is relocated and organized as a
tool for users that wish to incorporate statistical concepts into the machine design
process. This chapter should be reviewed if Secs. 513, 617, or Chap. 11 are to be
covered.
Supplements
The 8th edition of Shigleys Mechanical Engineering Design features McGraw-Hills ARIS
(Assessment Review and Instruction System). ARIS makes homework meaningfuland
manageablefor instructors and students. Instructors can assign and grade text-specic
homework within the industrys most robust and versatile homework management system. Students can access multimedia learning tools and benet from unlimited practice
via algorithmic problems. Go to aris.mhhe.com to learn more and register!
The array of tools available to users of Shigleys Mechanical Engineering Design
includes:
Student Supplements
TutorialsPresentation of major concepts, with visuals. Among the topics covered
are pressure vessel design, press and shrink ts, contact stresses, and design for static
failure.
MATLAB® for machine design. Includes visual simulations and accompanying source
code. The simulations are linked to examples and problems in the text and demonstrate
the ways computational software can be used in mechanical design and analysis.
Fundamentals of engineering (FE) exam questions for machine design. Interactive
problems and solutions serve as effective, self-testing problems as well as excellent
preparation for the FE exam.
Algorithmic Problems. Allow step-by-step problem-solving using a recursive computational procedure (algorithm) to create an innite number of problems.
Instructor Supplements (under password protection)
Solutions manual. The instructors manual contains solutions to most end-of-chapter
nondesign problems.
PowerPoint® slides. Slides of important gures and tables from the text are provided
in PowerPoint format for use in lectures.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
List of Symbols
© The McGrawHill
Companies, 2008
5
List of Symbols
This is a list of common symbols used in machine design and in this book. Specialized
use in a subject-matter area often attracts fore and post subscripts and superscripts.
To make the table brief enough to be useful the symbol kernels are listed. See
Table 141, pp. 715716 for spur and helical gearing symbols, and Table 151,
pp. 769770 for bevel-gear symbols.
A
A
a
a
ˆ
a
B
Bhn
B
b
ˆ
b
b
C
c
CDF
COV
c
D
d
E
e
F
f
fom
G
g
H
HB
HRC
h
hC R
¯
I
i
i
Area, coefcient
Area variate
Distance, regression constant
Regression constant estimate
Distance variate
Coefcient
Brinell hardness
Variate
Distance, Weibull shape parameter, range number, regression constant,
width
Regression constant estimate
Distance variate
Basic load rating, bolted-joint constant, center distance, coefcient of
variation, column end condition, correction factor, specic heat capacity,
spring index
Distance, viscous damping, velocity coefcient
Cumulative distribution function
Coefcient of variation
Distance variate
Helix diameter
Diameter, distance
Modulus of elasticity, energy, error
Distance, eccentricity, efciency, Naperian logarithmic base
Force, fundamental dimension force
Coefcient of friction, frequency, function
Figure of merit
Torsional modulus of elasticity
Acceleration due to gravity, function
Heat, power
Brinell hardness
Rockwell C-scale hardness
Distance, lm thickness
Combined overall coefcient of convection and radiation heat transfer
Integral, linear impulse, mass moment of inertia, second moment of area
Index
Unit vector in x-direction
xxiii
6
xxiv
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
List of Symbols
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
J
j
K
k
k
L
LN
l
M
M
m
N
N
n
nd
P
PDF
p
Q
q
R
R
r
r
S
S
s
T
T
t
U
U
u
V
v
W
W
w
w
X
x
x
Y
y
y
Z
z
z
Mechanical equivalent of heat, polar second moment of area, geometry
factor
Unit vector in the y-direction
Service factor, stress-concentration factor, stress-augmentation factor,
torque coefcient
Marin endurance limit modifying factor, spring rate
k variate, unit vector in the z-direction
Length, life, fundamental dimension length
Lognormal distribution
Length
Fundamental dimension mass, moment
Moment vector, moment variate
Mass, slope, strain-strengthening exponent
Normal force, number, rotational speed
Normal distribution
Load factor, rotational speed, safety factor
Design factor
Force, pressure, diametral pitch
Probability density function
Pitch, pressure, probability
First moment of area, imaginary force, volume
Distributed load, notch sensitivity
Radius, reaction force, reliability, Rockwell hardness, stress ratio
Vector reaction force
Correlation coefcient, radius
Distance vector
Sommerfeld number, strength
S variate
Distance, sample standard deviation, stress
Temperature, tolerance, torque, fundamental dimension time
Torque vector, torque variate
Distance, Students t-statistic, time, tolerance
Strain energy
Uniform distribution
Strain energy per unit volume
Linear velocity, shear force
Linear velocity
Cold-work factor, load, weight
Weibull distribution
Distance, gap, load intensity
Vector distance
Coordinate, truncated number
Coordinate, true value of a number, Weibull parameter
x variate
Coordinate
Coordinate, deection
y variate
Coordinate, section modulus, viscosity
Standard deviation of the unit normal distribution
Variate of z
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Front Matter
List of Symbols
© The McGrawHill
Companies, 2008
List of Symbols
α
β
δ
ǫ
ε
Ŵ
γ
λ
L
µ
ν
ω
φ
ψ
ρ
σ
σ
S
σ
ˆ
τ
θ
¢
$
7
xxv
Coefcient, coefcient of linear thermal expansion, end-condition for
springs, thread angle
Bearing angle, coefcient
Change, deection
Deviation, elongation
Eccentricity ratio, engineering (normal) strain
Normal distribution with a mean of 0 and a standard deviation of s
True or logarithmic normal strain
Gamma function
Pitch angle, shear strain, specic weight
Slenderness ratio for springs
Unit lognormal with a mean of l and a standard deviation equal to COV
Absolute viscosity, population mean
Poisson ratio
Angular velocity, circular frequency
Angle, wave length
Slope integral
Radius of curvature
Normal stress
Von Mises stress
Normal stress variate
Standard deviation
Shear stress
Shear stress variate
Angle, Weibull characteristic parameter
Cost per unit weight
Cost
8
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
PART
I. Basics
Introduction
1
Basics
© The McGrawHill
Companies, 2008
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
9
1
Introduction to Mechanical
Engineering Design
Chapter Outline
11
Design
12
Mechanical Engineering Design
13
Phases and Interactions of the Design Process
14
Design Tools and Resources
15
The Design Engineers Professional Responsibilities
16
Standards and Codes
17
Economics
18
Safety and Product Liability
19
Stress and Strength
4
5
5
8
10
12
12
15
15
110
Uncertainty
111
Design Factor and Factor of Safety
112
Reliability
113
Dimensions and Tolerances
114
Units
115
Calculations and Signicant Figures
116
Power Transmission Case Study Specications
16
17
18
19
21
22
23
3
10
4
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Mechanical design is a complex undertaking, requiring many skills. Extensive relationships need to be subdivided into a series of simple tasks. The complexity of the subject
requires a sequence in which ideas are introduced and iterated.
We rst address the nature of design in general, and then mechanical engineering
design in particular. Design is an iterative process with many interactive phases. Many
resources exist to support the designer, including many sources of information and an
abundance of computational design tools. The design engineer needs not only to develop
competence in their eld but must also cultivate a strong sense of responsibility and
professional work ethic.
There are roles to be played by codes and standards, ever-present economics, safety,
and considerations of product liability. The survival of a mechanical component is often
related through stress and strength. Matters of uncertainty are ever-present in engineering design and are typically addressed by the design factor and factor of safety, either
in the form of a deterministic (absolute) or statistical sense. The latter, statistical
approach, deals with a designs reliability and requires good statistical data.
In mechanical design, other considerations include dimensions and tolerances,
units, and calculations.
The book consists of four parts. Part 1, Basics, begins by explaining some differences between design and analysis and introducing some fundamental notions and
approaches to design. It continues with three chapters reviewing material properties,
stress analysis, and stiffness and deection analysis, which are the key principles necessary for the remainder of the book.
Part 2, Failure Prevention, consists of two chapters on the prevention of failure of
mechanical parts. Why machine parts fail and how they can be designed to prevent failure are difcult questions, and so we take two chapters to answer them, one on preventing failure due to static loads, and the other on preventing fatigue failure due to
time-varying, cyclic loads.
In Part 3, Design of Mechanical Elements, the material of Parts 1 and 2 is applied
to the analysis, selection, and design of specic mechanical elements such as shafts,
fasteners, weldments, springs, rolling contact bearings, lm bearings, gears, belts,
chains, and wire ropes.
Part 4, Analysis Tools, provides introductions to two important methods used in
mechanical design, nite element analysis and statistical analysis. This is optional study
material, but some sections and examples in Parts 1 to 3 demonstrate the use of these tools.
There are two appendixes at the end of the book. Appendix A contains many useful tables referenced throughout the book. Appendix B contains answers to selected
end-of-chapter problems.
11
Design
To design is either to formulate a plan for the satisfaction of a specied need or to solve
a problem. If the plan results in the creation of something having a physical reality, then
the product must be functional, safe, reliable, competitive, usable, manufacturable, and
marketable.
Design is an innovative and highly iterative process. It is also a decision-making
process. Decisions sometimes have to be made with too little information, occasionally with just the right amount of information, or with an excess of partially contradictory
information. Decisions are sometimes made tentatively, with the right reserved to adjust
as more becomes known. The point is that the engineering designer has to be personally
comfortable with a decision-making, problem-solving role.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
11
5
Design is a communication-intensive activity in which both words and pictures are
used, and written and oral forms are employed. Engineers have to communicate effectively and work with people of many disciplines. These are important skills, and an
engineers success depends on them.
A designers personal resources of creativeness, communicative ability, and problemsolving skill are intertwined with knowledge of technology and rst principles.
Engineering tools (such as mathematics, statistics, computers, graphics, and languages)
are combined to produce a plan that, when carried out, produces a product that is functional, safe, reliable, competitive, usable, manufacturable, and marketable, regardless
of who builds it or who uses it.
12
Mechanical Engineering Design
Mechanical engineers are associated with the production and processing of energy and
with providing the means of production, the tools of transportation, and the techniques
of automation. The skill and knowledge base are extensive. Among the disciplinary
bases are mechanics of solids and uids, mass and momentum transport, manufacturing processes, and electrical and information theory. Mechanical engineering design
involves all the disciplines of mechanical engineering.
Real problems resist compartmentalization. A simple journal bearing involves uid
ow, heat transfer, friction, energy transport, material selection, thermomechanical
treatments, statistical descriptions, and so on. A building is environmentally controlled.
The heating, ventilation, and air-conditioning considerations are sufciently specialized
that some speak of heating, ventilating, and air-conditioning design as if it is separate
and distinct from mechanical engineering design. Similarly, internal-combustion engine
design, turbomachinery design, and jet-engine design are sometimes considered discrete entities. Here, the leading string of words preceding the word design is merely a
product descriptor. Similarly, there are phrases such as machine design, machine-element
design, machine-component design, systems design, and uid-power design. All of
these phrases are somewhat more focused examples of mechanical engineering design.
They all draw on the same bodies of knowledge, are similarly organized, and require
similar skills.
13
Phases and Interactions of the Design Process
What is the design process? How does it begin? Does the engineer simply sit down at
a desk with a blank sheet of paper and jot down some ideas? What happens next? What
factors inuence or control the decisions that have to be made? Finally, how does the
design process end?
The complete design process, from start to finish, is often outlined as in Fig. 11.
The process begins with an identification of a need and a decision to do something
about it. After many iterations, the process ends with the presentation of the plans
for satisfying the need. Depending on the nature of the design task, several design
phases may be repeated throughout the life of the product, from inception to termination. In the next several subsections, we shall examine these steps in the design
process in detail.
Identication of need generally starts the design process. Recognition of the need
and phrasing the need often constitute a highly creative act, because the need may be
only a vague discontent, a feeling of uneasiness, or a sensing that something is not right.
The need is often not evident at all; recognition is usually triggered by a particular
12
6
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Mechanical Engineering Design
Figure 11
Identification of need
The phases in design,
acknowledging the many
feedbacks and iterations.
Definition of problem
Synthesis
Analysis and optimization
Evaluation
Iteration
Presentation
adverse circumstance or a set of random circumstances that arises almost simultaneously.
For example, the need to do something about a food-packaging machine may be indicated by the noise level, by a variation in package weight, and by slight but perceptible
variations in the quality of the packaging or wrap.
There is a distinct difference between the statement of the need and the denition
of the problem. The denition of problem is more specic and must include all the specications for the object that is to be designed. The specications are the input and output quantities, the characteristics and dimensions of the space the object must occupy,
and all the limitations on these quantities. We can regard the object to be designed as
something in a black box. In this case we must specify the inputs and outputs of the box,
together with their characteristics and limitations. The specications dene the cost, the
number to be manufactured, the expected life, the range, the operating temperature, and
the reliability. Specied characteristics can include the speeds, feeds, temperature limitations, maximum range, expected variations in the variables, dimensional and weight
limitations, etc.
There are many implied specifications that result either from the designers particular environment or from the nature of the problem itself. The manufacturing
processes that are available, together with the facilities of a certain plant, constitute
restrictions on a designers freedom, and hence are a part of the implied specifications. It may be that a small plant, for instance, does not own cold-working machinery. Knowing this, the designer might select other metal-processing methods that
can be performed in the plant. The labor skills available and the competitive situation also constitute implied constraints. Anything that limits the designers freedom
of choice is a constraint. Many materials and sizes are listed in suppliers catalogs,
for instance, but these are not all easily available and shortages frequently occur.
Furthermore, inventory economics requires that a manufacturer stock a minimum
number of materials and sizes. An example of a specification is given in Sec. 116.
This example is for a case study of a power transmission that is presented throughout
this text.
The synthesis of a scheme connecting possible system elements is sometimes
called the invention of the concept or concept design. This is the rst and most important step in the synthesis task. Various schemes must be proposed, investigated, and
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Introduction to Mechanical Engineering Design
13
7
quantied in terms of established metrics.1 As the eshing out of the scheme progresses,
analyses must be performed to assess whether the system performance is satisfactory or
better, and, if satisfactory, just how well it will perform. System schemes that do not
survive analysis are revised, improved, or discarded. Those with potential are optimized
to determine the best performance of which the scheme is capable. Competing schemes
are compared so that the path leading to the most competitive product can be chosen.
Figure 11 shows that synthesis and analysis and optimization are intimately and
iteratively related.
We have noted, and we emphasize, that design is an iterative process in which we
proceed through several steps, evaluate the results, and then return to an earlier phase
of the procedure. Thus, we may synthesize several components of a system, analyze and
optimize them, and return to synthesis to see what effect this has on the remaining parts
of the system. For example, the design of a system to transmit power requires attention
to the design and selection of individual components (e.g., gears, bearings, shaft).
However, as is often the case in design, these components are not independent. In order
to design the shaft for stress and deection, it is necessary to know the applied forces.
If the forces are transmitted through gears, it is necessary to know the gear specications in order to determine the forces that will be transmitted to the shaft. But stock
gears come with certain bore sizes, requiring knowledge of the necessary shaft diameter. Clearly, rough estimates will need to be made in order to proceed through the
process, rening and iterating until a nal design is obtained that is satisfactory for each
individual component as well as for the overall design specications. Throughout the
text we will elaborate on this process for the case study of a power transmission design.
Both analysis and optimization require that we construct or devise abstract models
of the system that will admit some form of mathematical analysis. We call these models mathematical models. In creating them it is our hope that we can nd one that will
simulate the real physical system very well. As indicated in Fig. 11, evaluation is a
signicant phase of the total design process. Evaluation is the nal proof of a successful design and usually involves the testing of a prototype in the laboratory. Here we
wish to discover if the design really satises the needs. Is it reliable? Will it compete
successfully with similar products? Is it economical to manufacture and to use? Is it
easily maintained and adjusted? Can a prot be made from its sale or use? How likely
is it to result in product-liability lawsuits? And is insurance easily and cheaply
obtained? Is it likely that recalls will be needed to replace defective parts or systems?
Communicating the design to others is the nal, vital presentation step in the
design process. Undoubtedly, many great designs, inventions, and creative works have
been lost to posterity simply because the originators were unable or unwilling to
explain their accomplishments to others. Presentation is a selling job. The engineer,
when presenting a new solution to administrative, management, or supervisory persons,
is attempting to sell or to prove to them that this solution is a better one. Unless this can
be done successfully, the time and effort spent on obtaining the solution have been
largely wasted. When designers sell a new idea, they also sell themselves. If they are
repeatedly successful in selling ideas, designs, and new solutions to management, they
begin to receive salary increases and promotions; in fact, this is how anyone succeeds
in his or her profession.
1
An excellent reference for this topic is presented by Stuart Pugh, Total DesignIntegrated Methods for
Successful Product Engineering, Addison-Wesley, 1991. A description of the Pugh method is also provided
in Chap. 8, David G. Ullman, The Mechanical Design Process, 3rd ed., McGraw-Hill, 2003.
14
8
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Mechanical Engineering Design
Design Considerations
Sometimes the strength required of an element in a system is an important factor in the
determination of the geometry and the dimensions of the element. In such a situation
we say that strength is an important design consideration. When we use the expression
design consideration, we are referring to some characteristic that inuences the design
of the element or, perhaps, the entire system. Usually quite a number of such characteristics must be considered and prioritized in a given design situation. Many of the
important ones are as follows (not necessarily in order of importance):
1
2
3
4
5
6
7
8
9
10
11
12
13
Functionality
Strength/stress
Distortion/deection/stiffness
Wear
Corrosion
Safety
Reliability
Manufacturability
Utility
Cost
Friction
Weight
Life
14
15
16
17
18
19
20
21
22
23
24
25
26
Noise
Styling
Shape
Size
Control
Thermal properties
Surface
Lubrication
Marketability
Maintenance
Volume
Liability
Remanufacturing/resource recovery
Some of these characteristics have to do directly with the dimensions, the material, the
processing, and the joining of the elements of the system. Several characteristics may
be interrelated, which affects the conguration of the total system.
14
Design Tools and Resources
Today, the engineer has a great variety of tools and resources available to assist in the
solution of design problems. Inexpensive microcomputers and robust computer software packages provide tools of immense capability for the design, analysis, and simulation of mechanical components. In addition to these tools, the engineer always needs
technical information, either in the form of basic science/engineering behavior or the
characteristics of specic off-the-shelf components. Here, the resources can range from
science/engineering textbooks to manufacturers brochures or catalogs. Here too, the
computer can play a major role in gathering information.2
Computational Tools
Computer-aided design (CAD) software allows the development of three-dimensional
(3-D) designs from which conventional two-dimensional orthographic views with automatic dimensioning can be produced. Manufacturing tool paths can be generated from the
3-D models, and in some cases, parts can be created directly from a 3-D database by using
a rapid prototyping and manufacturing method (stereolithography)paperless manufacturing! Another advantage of a 3-D database is that it allows rapid and accurate calculations of mass properties such as mass, location of the center of gravity, and mass moments
of inertia. Other geometric properties such as areas and distances between points are
likewise easily obtained. There are a great many CAD software packages available such
2
An excellent and comprehensive discussion of the process of gathering information can be found in
Chap. 4, George E. Dieter, Engineering Design, A Materials and Processing Approach, 3rd ed.,
McGraw-Hill, New York, 2000.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Introduction to Mechanical Engineering Design
15
9
as Aries, AutoCAD, CadKey, I-Deas, Unigraphics, Solid Works, and ProEngineer, to
name a few.
The term computer-aided engineering (CAE) generally applies to all computerrelated engineering applications. With this denition, CAD can be considered as a subset of CAE. Some computer software packages perform specic engineering analysis
and/or simulation tasks that assist the designer, but they are not considered a tool for the
creation of the design that CAD is. Such software ts into two categories: engineeringbased and non-engineering-specic. Some examples of engineering-based software for
mechanical engineering applicationssoftware that might also be integrated within a
CAD systeminclude nite-element analysis (FEA) programs for analysis of stress
and deection (see Chap. 19), vibration, and heat transfer (e.g., Algor, ANSYS, and
MSC/NASTRAN); computational uid dynamics (CFD) programs for uid-ow analysis and simulation (e.g., CFD++, FIDAP, and Fluent); and programs for simulation of
dynamic force and motion in mechanisms (e.g., ADAMS, DADS, and Working Model).
Examples of non-engineering-specic computer-aided applications include software for word processing, spreadsheet software (e.g., Excel, Lotus, and Quattro-Pro),
and mathematical solvers (e.g., Maple, MathCad, Matlab, Mathematica, and TKsolver).
Your instructor is the best source of information about programs that may be available
to you and can recommend those that are useful for specic tasks. One caution, however:
Computer software is no substitute for the human thought process. You are the driver here;
the computer is the vehicle to assist you on your journey to a solution. Numbers generated
by a computer can be far from the truth if you entered incorrect input, if you misinterpreted
the application or the output of the program, if the program contained bugs, etc. It is your
responsibility to assure the validity of the results, so be careful to check the application and
results carefully, perform benchmark testing by submitting problems with known solutions, and monitor the software company and user-group newsletters.
Acquiring Technical Information
We currently live in what is referred to as the information age, where information is generated at an astounding pace. It is difcult, but extremely important, to keep abreast of past
and current developments in ones eld of study and occupation. The reference in Footnote
2 provides an excellent description of the informational resources available and is highly
recommended reading for the serious design engineer. Some sources of information are:
Libraries (community, university, and private). Engineering dictionaries and encyclopedias, textbooks, monographs, handbooks, indexing and abstract services, journals,
translations, technical reports, patents, and business sources/brochures/catalogs.
Government sources. Departments of Defense, Commerce, Energy, and Transportation;
NASA; Government Printing Ofce; U.S. Patent and Trademark Ofce; National
Technical Information Service; and National Institute for Standards and Technology.
Professional societies. American Society of Mechanical Engineers, Society of
Manufacturing Engineers, Society of Automotive Engineers, American Society for
Testing and Materials, and American Welding Society.
Commercial vendors. Catalogs, technical literature, test data, samples, and cost
information.
Internet. The computer network gateway to websites associated with most of the
categories listed above.3
3
Some helpful Web resources, to name a few, include www.globalspec.com, www.engnetglobal.com,
www.efunda.com, www.thomasnet.com, and www.uspto.gov.
16
10
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
This list is not complete. The reader is urged to explore the various sources of
information on a regular basis and keep records of the knowledge gained.
15
The Design Engineers Professional Responsibilities
In general, the design engineer is required to satisfy the needs of customers (management, clients, consumers, etc.) and is expected to do so in a competent, responsible, ethical, and professional manner. Much of engineering course work and practical
experience focuses on competence, but when does one begin to develop engineering
responsibility and professionalism? To start on the road to success, you should start
to develop these characteristics early in your educational program. You need to cultivate your professional work ethic and process skills before graduation, so that
when you begin your formal engineering career, you will be prepared to meet the
challenges.
It is not obvious to some students, but communication skills play a large role here,
and it is the wise student who continuously works to improve these skillseven if it
is not a direct requirement of a course assignment! Success in engineering (achievements, promotions, raises, etc.) may in large part be due to competence but if you cannot communicate your ideas clearly and concisely, your technical prociency may be
compromised.
You can start to develop your communication skills by keeping a neat and clear
journal/logbook of your activities, entering dated entries frequently. (Many companies
require their engineers to keep a journal for patent and liability concerns.) Separate
journals should be used for each design project (or course subject). When starting a
project or problem, in the denition stage, make journal entries quite frequently. Others,
as well as yourself, may later question why you made certain decisions. Good chronological records will make it easier to explain your decisions at a later date.
Many engineering students see themselves after graduation as practicing engineers
designing, developing, and analyzing products and processes and consider the need of
good communication skills, either oral or writing, as secondary. This is far from the
truth. Most practicing engineers spend a good deal of time communicating with others,
writing proposals and technical reports, and giving presentations and interacting with
engineering and nonengineering support personnel. You have the time now to sharpen
your communication skills. When given an assignment to write or make any presentation, technical or nontechnical, accept it enthusiastically, and work on improving your
communication skills. It will be time well spent to learn the skills now rather than on
the job.
When you are working on a design problem, it is important that you develop a
systematic approach. Careful attention to the following action steps will help you to
organize your solution processing technique.
Understand the problem. Problem denition is probably the most signicant step in the
engineering design process. Carefully read, understand, and rene the problem statement.
Identify the known. From the rened problem statement, describe concisely what
information is known and relevant.
Identify the unknown and formulate the solution strategy. State what must be determined, in what order, so as to arrive at a solution to the problem. Sketch the component or system under investigation, identifying known and unknown parameters.
Create a owchart of the steps necessary to reach the nal solution. The steps may
require the use of free-body diagrams; material properties from tables; equations
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Introduction to Mechanical Engineering Design
17
11
from rst principles, textbooks, or handbooks relating the known and unknown
parameters; experimentally or numerically based charts; specic computational tools
as discussed in Sec. 14; etc.
State all assumptions and decisions. Real design problems generally do not have
unique, ideal, closed-form solutions. Selections, such as choice of materials, and heat
treatments, require decisions. Analyses require assumptions related to the modeling
of the real components or system. All assumptions and decisions should be identied
and recorded.
Analyze the problem. Using your solution strategy in conjunction with your decisions
and assumptions, execute the analysis of the problem. Reference the sources of all
equations, tables, charts, software results, etc. Check the credibility of your results.
Check the order of magnitude, dimensionality, trends, signs, etc.
Evaluate your solution. Evaluate each step in the solution, noting how changes in
strategy, decisions, assumptions, and execution might change the results, in positive
or negative ways. If possible, incorporate the positive changes in your nal solution.
Present your solution. Here is where your communication skills are important. At
this point, you are selling yourself and your technical abilities. If you cannot skillfully explain what you have done, some or all of your work may be misunderstood
and unaccepted. Know your audience.
As stated earlier, all design processes are interactive and iterative. Thus, it may be necessary to repeat some or all of the above steps more than once if less than satisfactory
results are obtained.
In order to be effective, all professionals must keep current in their elds of
endeavor. The design engineer can satisfy this in a number of ways by: being an active
member of a professional society such as the American Society of Mechanical
Engineers (ASME), the Society of Automotive Engineers (SAE), and the Society of
Manufacturing Engineers (SME); attending meetings, conferences, and seminars of
societies, manufacturers, universities, etc.; taking specic graduate courses or programs
at universities; regularly reading technical and professional journals; etc. An engineers
education does not end at graduation.
The design engineers professional obligations include conducting activities in an
ethical manner. Reproduced here is the Engineers Creed from the National Society of
Professional Engineers (NSPE)4:
As a Professional Engineer I dedicate my professional knowledge and skill to the
advancement and betterment of human welfare.
I pledge:
To give the utmost of performance;
To participate in none but honest enterprise;
To live and work according to the laws of man and the highest standards of professional conduct;
To place service before prot, the honor and standing of the profession before
personal advantage, and the public welfare above all other considerations.
In humility and with need for Divine Guidance, I make this pledge.
4
Adopted by the National Society of Professional Engineers, June 1954. The Engineers Creed. Reprinted
by permission of the National Society of Professional Engineers. This has been expanded and revised by
NSPE. For the current revision, January 2006, see the website www.nspe.org/ethics/ehl-code.asp, or the pdf
le, www.nspe.org/ethics/code-2006-Jan.pdf.
18
12
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
16
Standards and Codes
A standard is a set of specications for parts, materials, or processes intended to
achieve uniformity, efciency, and a specied quality. One of the important purposes
of a standard is to place a limit on the number of items in the specications so as to
provide a reasonable inventory of tooling, sizes, shapes, and varieties.
A code is a set of specications for the analysis, design, manufacture, and construction of something. The purpose of a code is to achieve a specied degree of safety,
efciency, and performance or quality. It is important to observe that safety codes do
not imply absolute safety. In fact, absolute safety is impossible to obtain. Sometimes
the unexpected event really does happen. Designing a building to withstand a 120 mi/h
wind does not mean that the designers think a 140 mi/h wind is impossible; it simply
means that they think it is highly improbable.
All of the organizations and societies listed below have established specications
for standards and safety or design codes. The name of the organization provides a clue
to the nature of the standard or code. Some of the standards and codes, as well as
addresses, can be obtained in most technical libraries. The organizations of interest to
mechanical engineers are:
Aluminum Association (AA)
American Gear Manufacturers Association (AGMA)
American Institute of Steel Construction (AISC)
American Iron and Steel Institute (AISI)
American National Standards Institute (ANSI)5
ASM International6
American Society of Mechanical Engineers (ASME)
American Society of Testing and Materials (ASTM)
American Welding Society (AWS)
American Bearing Manufacturers Association (ABMA)7
British Standards Institution (BSI)
Industrial Fasteners Institute (IFI)
Institution of Mechanical Engineers (I. Mech. E.)
International Bureau of Weights and Measures (BIPM)
International Standards Organization (ISO)
National Institute for Standards and Technology (NIST)8
Society of Automotive Engineers (SAE)
17
Economics
The consideration of cost plays such an important role in the design decision process that
we could easily spend as much time in studying the cost factor as in the study of the
entire subject of design. Here we introduce only a few general concepts and simple rules.
5
In 1966 the American Standards Association (ASA) changed its name to the United States of America
Standards Institute (USAS). Then, in 1969, the name was again changed, to American National Standards
Institute, as shown above and as it is today. This means that you may occasionally nd ANSI standards
designated as ASA or USAS.
6
Formally American Society for Metals (ASM). Currently the acronym ASM is undened.
7
In 1993 the Anti-Friction Bearing Manufacturers Association (AFBMA) changed its name to the American
Bearing Manufacturers Association (ABMA).
8
Former National Bureau of Standards (NBS).
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Introduction to Mechanical Engineering Design
19
13
First, observe that nothing can be said in an absolute sense concerning costs.
Materials and labor usually show an increasing cost from year to year. But the costs
of processing the materials can be expected to exhibit a decreasing trend because of
the use of automated machine tools and robots. The cost of manufacturing a single
product will vary from city to city and from one plant to another because of overhead, labor, taxes, and freight differentials and the inevitable slight manufacturing
variations.
Standard Sizes
The use of standard or stock sizes is a rst principle of cost reduction. An engineer who
species an AISI 1020 bar of hot-rolled steel 53 mm square has added cost to the product, provided that a bar 50 or 60 mm square, both of which are preferred sizes, would
do equally well. The 53-mm size can be obtained by special order or by rolling or
machining a 60-mm square, but these approaches add cost to the product. To ensure that
standard or preferred sizes are specied, designers must have access to stock lists of the
materials they employ.
A further word of caution regarding the selection of preferred sizes is necessary.
Although a great many sizes are usually listed in catalogs, they are not all readily available. Some sizes are used so infrequently that they are not stocked. A rush order for
such sizes may mean more on expense and delay. Thus you should also have access to
a list such as those in Table A17 for preferred inch and millimeter sizes.
There are many purchased parts, such as motors, pumps, bearings, and fasteners,
that are specied by designers. In the case of these, too, you should make a special
effort to specify parts that are readily available. Parts that are made and sold in large
quantities usually cost somewhat less than the odd sizes. The cost of rolling bearings,
for example, depends more on the quantity of production by the bearing manufacturer
than on the size of the bearing.
Large Tolerances
Among the effects of design specications on costs, tolerances are perhaps most signicant. Tolerances, manufacturing processes, and surface nish are interrelated and
inuence the producibility of the end product in many ways. Close tolerances may
necessitate additional steps in processing and inspection or even render a part completely impractical to produce economically. Tolerances cover dimensional variation
and surface-roughness range and also the variation in mechanical properties resulting
from heat treatment and other processing operations.
Since parts having large tolerances can often be produced by machines with
higher production rates, costs will be significantly smaller. Also, fewer such parts will
be rejected in the inspection process, and they are usually easier to assemble. A plot
of cost versus tolerance/machining process is shown in Fig. 12, and illustrates the
drastic increase in manufacturing cost as tolerance diminishes with finer machining
processing.
Breakeven Points
Sometimes it happens that, when two or more design approaches are compared for cost,
the choice between the two depends on a set of conditions such as the quantity of production, the speed of the assembly lines, or some other condition. There then occurs a
point corresponding to equal cost, which is called the breakeven point.
20
14
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Mechanical Engineering Design
Figure 12
Costs, %
Cost versus tolerance/
machining process.
(From David G. Ullman, The
Mechanical Design Process,
3rd ed., McGraw-Hill, New
York, 2003.)
400
380
360
340
320
300
280
260
240
220
200
180
160
140
120
100
80
60
40
20
Material: steel
0.030
0.015
0.010
0.005
0.003
0.001
0.0005
0.00025
0.063
0.025
0.012
0.006
Semifinish
turn
Finish
turn
Grind
Nominal tolerances (inches)
0.75
0.50
0.50
0.125
Nominal tolerance (mm)
Rough turn
Hone
Machining operations
Figure 13
140
A breakeven point.
Breakeven point
120
Cost, $
100
Automatic screw
machine
80
60
Hand screw machine
40
20
0
0
20
40
60
Production
80
100
As an example, consider a situation in which a certain part can be manufactured at
the rate of 25 parts per hour on an automatic screw machine or 10 parts per hour on a
hand screw machine. Let us suppose, too, that the setup time for the automatic is 3 h and
that the labor cost for either machine is $20 per hour, including overhead. Figure 13 is
a graph of cost versus production by the two methods. The breakeven point for this
example corresponds to 50 parts. If the desired production is greater than 50 parts, the
automatic machine should be used.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
21
15
Cost Estimates
There are many ways of obtaining relative cost gures so that two or more designs
can be roughly compared. A certain amount of judgment may be required in some
instances. For example, we can compare the relative value of two automobiles by
comparing the dollar cost per pound of weight. Another way to compare the cost of
one design with another is simply to count the number of parts. The design having
the smaller number of parts is likely to cost less. Many other cost estimators can be
used, depending upon the application, such as area, volume, horsepower, torque,
capacity, speed, and various performance ratios.9
18
Safety and Product Liability
The strict liability concept of product liability generally prevails in the United States.
This concept states that the manufacturer of an article is liable for any damage or harm
that results because of a defect. And it doesnt matter whether the manufacturer knew
about the defect, or even could have known about it. For example, suppose an article
was manufactured, say, 10 years ago. And suppose at that time the article could not have
been considered defective on the basis of all technological knowledge then available.
Ten years later, according to the concept of strict liability, the manufacturer is still
liable. Thus, under this concept, the plaintiff needs only to prove that the article was
defective and that the defect caused some damage or harm. Negligence of the manufacturer need not be proved.
The best approaches to the prevention of product liability are good engineering in
analysis and design, quality control, and comprehensive testing procedures. Advertising
managers often make glowing promises in the warranties and sales literature for a product. These statements should be reviewed carefully by the engineering staff to eliminate
excessive promises and to insert adequate warnings and instructions for use.
19
Stress and Strength
The survival of many products depends on how the designer adjusts the maximum
stresses in a component to be less than the components strength at specic locations of
interest. The designer must allow the maximum stress to be less than the strength by a
sufcient margin so that despite the uncertainties, failure is rare.
In focusing on the stress-strength comparison at a critical (controlling) location,
we often look for strength in the geometry and condition of use. Strengths are the
magnitudes of stresses at which something of interest occurs, such as the proportional
limit, 0.2 percent-offset yielding, or fracture. In many cases, such events represent the
stress level at which loss of function occurs.
Strength is a property of a material or of a mechanical element. The strength of an
element depends on the choice, the treatment, and the processing of the material.
Consider, for example, a shipment of springs. We can associate a strength with a specic spring. When this spring is incorporated into a machine, external forces are applied
that result in load-induced stresses in the spring, the magnitudes of which depend on its
geometry and are independent of the material and its processing. If the spring is
removed from the machine unharmed, the stress due to the external forces will return
9
For an overview of estimating manufacturing costs, see Chap. 11, Karl T. Ulrich and Steven D. Eppinger,
Product Design and Development, 3rd ed., McGraw-Hill, New York, 2004.
22
16
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
to zero. But the strength remains as one of the properties of the spring. Remember, then,
that strength is an inherent property of a part, a property built into the part because of
the use of a particular material and process.
Various metalworking and heat-treating processes, such as forging, rolling, and
cold forming, cause variations in the strength from point to point throughout a part. The
spring cited above is quite likely to have a strength on the outside of the coils different
from its strength on the inside because the spring has been formed by a cold winding
process, and the two sides may not have been deformed by the same amount.
Remember, too, therefore, that a strength value given for a part may apply to only a particular point or set of points on the part.
In this book we shall use the capital letter S to denote strength, with appropriate
subscripts to denote the type of strength. Thus, Ss is a shear strength, Sy a yield
strength, and Su an ultimate strength.
In accordance with accepted engineering practice, we shall employ the Greek letters σ (sigma) and τ (tau) to designate normal and shear stresses, respectively. Again,
various subscripts will indicate some special characteristic. For example, σ1 is a principal stress, σ y a stress component in the y direction, and σr a stress component in the
radial direction.
Stress is a state property at a specic point within a body, which is a function of
load, geometry, temperature, and manufacturing processing. In an elementary course in
mechanics of materials, stress related to load and geometry is emphasized with some
discussion of thermal stresses. However, stresses due to heat treatments, molding,
assembly, etc. are also important and are sometimes neglected. A review of stress analysis for basic load states and geometry is given in Chap. 3.
110
Uncertainty
Uncertainties in machinery design abound. Examples of uncertainties concerning stress
and strength include
Composition of material and the effect of variation on properties.
Variations in properties from place to place within a bar of stock.
Effect of processing locally, or nearby, on properties.
Effect of nearby assemblies such as weldments and shrink ts on stress conditions.
Effect of thermomechanical treatment on properties.
Intensity and distribution of loading.
Validity of mathematical models used to represent reality.
Intensity of stress concentrations.
Inuence of time on strength and geometry.
Effect of corrosion.
Effect of wear.
Uncertainty as to the length of any list of uncertainties.
Engineers must accommodate uncertainty. Uncertainty always accompanies change.
Material properties, load variability, fabrication delity, and validity of mathematical
models are among concerns to designers.
There are mathematical methods to address uncertainties. The primary techniques
are the deterministic and stochastic methods. The deterministic method establishes a
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
23
17
design factor based on the absolute uncertainties of a loss-of-function parameter and a
maximum allowable parameter. Here the parameter can be load, stress, deection, etc.
Thus, the design factor n d is dened as
nd =
loss-of-function parameter
maximum allowable parameter
(11)
If the parameter is load, then the maximum allowable load can be found from
Maximum allowable load =
loss-of-function load
nd
(12)
EXAMPLE 11
Consider that the maximum load on a structure is known with an uncertainty of ±20 percent, and the load causing failure is known within ±15 percent. If the load causing failure is nominally 2000 lbf, determine the design factor and the maximum allowable load
that will offset the absolute uncertainties.
Solution
To account for its uncertainty, the loss-of-function load must increase to 1/0.85, whereas
the maximum allowable load must decrease to 1/1.2. Thus to offset the absolute uncertainties the design factor should be
Answer
nd =
1/0.85
= 1.4
1/1.2
From Eq. (12), the maximum allowable load is found to be
Answer
Maximum allowable load =
2000
= 1400 lbf
1.4
Stochastic methods (see Chap. 20) are based on the statistical nature of the design
parameters and focus on the probability of survival of the designs function (that is, on
reliability). Sections 513 and 617 demonstrate how this is accomplished.
111
Design Factor and Factor of Safety
A general approach to the allowable load versus loss-of-function load problem is the
deterministic design factor method, and sometimes called the classical method of
design. The fundamental equation is Eq. (11) where nd is called the design factor. All
loss-of-function modes must be analyzed, and the mode leading to the smallest design
factor governs. After the design is completed, the actual design factor may change as
a result of changes such as rounding up to a standard size for a cross section or using
off-the-shelf components with higher ratings instead of employing what is calculated
by using the design factor. The factor is then referred to as the factor of safety, n. The
factor of safety has the same denition as the design factor, but it generally differs
numerically.
Since stress may not vary linearly with load (see Sec. 319), using load as the
loss-of-function parameter may not be acceptable. It is more common then to express
24
18
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Mechanical Engineering Design
the design factor in terms of a stress and a relevant strength. Thus Eq. (11) can be
rewritten as
nd =
S
loss-of-function strength
=
allowable stress
σ (or τ )
(13)
The stress and strength terms in Eq. (13) must be of the same type and units. Also, the
stress and strength must apply to the same critical location in the part.
EXAMPLE 12
Solution
A rod with a cross-sectional area of A and loaded in tension with an axial force of P
2000 lbf undergoes a stress of σ = P/A. Using a material strength of 24 kpsi and a
design factor of 3.0, determine the minimum diameter of a solid circular rod. Using
Table A17, select a preferred fractional diameter and determine the rods factor of safety.
Since A = πd 2/4, and σ = S / n d , then
σ=
S
P
2 000
24 000
=
=
=
nd
3
A
π d 2/4
or,
Answer
d=
4 Pn d
πS
1/2
=
4(2000)3
π (24 000)
1/2
= 0.564 in
From Table A17, the next higher preferred size is 5 in 0.625 in. Thus, according to
8
the same equation developed earlier, the factor of safety n is
Answer
n=
π (24 000)0.6252
π Sd 2
=
= 3.68
4P
4(2000)
Thus rounding the diameter has increased the actual design factor.
112
Reliability
In these days of greatly increasing numbers of liability lawsuits and the need to conform to
regulations issued by governmental agencies such as EPA and OSHA, it is very important
for the designer and the manufacturer to know the reliability of their product. The reliability method of design is one in which we obtain the distribution of stresses and the distribution of strengths and then relate these two in order to achieve an acceptable success rate.
The statistical measure of the probability that a mechanical element will not fail in
use is called the reliability of that element. The reliability R can be expressed by a number having the range 0 R 1. A reliability of R = 0.90 means that there is a 90 percent chance that the part will perform its proper function without failure. The failure of
6 parts out of every 1000 manufactured might be considered an acceptable failure rate
for a certain class of products. This represents a reliability of
R =1
or 99.4 percent.
6
= 0.994
1000
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
25
19
In the reliability method of design, the designers task is to make a judicious selection of materials, processes, and geometry (size) so as to achieve a specic reliability
goal. Thus, if the objective reliability is to be 99.4 percent, as above, what combination
of materials, processing, and dimensions is needed to meet this goal?
Analyses that lead to an assessment of reliability address uncertainties, or their
estimates, in parameters that describe the situation. Stochastic variables such as
stress, strength, load, or size are described in terms of their means, standard deviations, and distributions. If bearing balls are produced by a manufacturing process in
which a diameter distribution is created, we can say upon choosing a ball that there
is uncertainty as to size. If we wish to consider weight or moment of inertia in rolling,
this size uncertainty can be considered to be propagated to our knowledge of weight
or inertia. There are ways of estimating the statistical parameters describing weight
and inertia from those describing size and density. These methods are variously called
propagation of error, propagation of uncertainty, or propagation of dispersion. These
methods are integral parts of analysis or synthesis tasks when probability of failure is
involved.
It is important to note that good statistical data and estimates are essential to perform an acceptable reliability analysis. This requires a good deal of testing and validation of the data. In many cases, this is not practical and a deterministic approach to the
design must be undertaken.
113
Dimensions and Tolerances
The following terms are used generally in dimensioning:
Nominal size. The size we use in speaking of an element. For example, we may specify a 1 1 -in pipe or a 1 -in bolt. Either the theoretical size or the actual measured size
2
2
may be quite different. The theoretical size of a 1 1 -in pipe is 1.900 in for the outside
2
diameter. And the diameter of the 1 -in bolt, say, may actually measure 0.492 in.
2
Limits. The stated maximum and minimum dimensions.
Tolerance. The difference between the two limits.
Bilateral tolerance. The variation in both directions from the basic dimension. That
is, the basic size is between the two limits, for example, 1.005 ± 0.002 in. The two
parts of the tolerance need not be equal.
Unilateral tolerance. The basic dimension is taken as one of the limits, and variation
is permitted in only one direction, for example,
1.005
+0.004
0.000
in
Clearance. A general term that refers to the mating of cylindrical parts such as a bolt
and a hole. The word clearance is used only when the internal member is smaller than
the external member. The diametral clearance is the measured difference in the two
diameters. The radial clearance is the difference in the two radii.
Interference. The opposite of clearance, for mating cylindrical parts in which the
internal member is larger than the external member.
Allowance. The minimum stated clearance or the maximum stated interference for
mating parts.
When several parts are assembled, the gap (or interference) depends on the dimensions and tolerances of the individual parts.
26
20
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Mechanical Engineering Design
EXAMPLE 13
A shouldered screw contains three hollow right circular cylindrical parts on the screw
before a nut is tightened against the shoulder. To sustain the function, the gap w must
equal or exceed 0.003 in. The parts in the assembly depicted in Fig. 14 have dimensions and tolerances as follows:
a = 1.750 ± 0.003 in
b = 0.750 ± 0.001 in
c = 0.120 ± 0.005 in
d = 0.875 ± 0.001 in
Figure 14
a
An assembly of three
cylindrical sleeves of lengths
a, b, and c on a shoulder bolt
shank of length a. The gap w
is of interest.
b
c
d
w
All parts except the part with the dimension d are supplied by vendors. The part containing the dimension d is made in-house.
(a) Estimate the mean and tolerance on the gap w .
(b) What basic value of d will assure that w 0.003 in?
Solution
(a) The mean value of w is given by
w = a b c d = 1.750 0.750 0.120 0.875 = 0.005 in
¯
¯¯¯¯
Answer
For equal bilateral tolerances, the tolerance of the gap is
Answer
tw =
all
t = 0.003 + 0.001 + 0.005 + 0.001 = 0.010 in
Then, w = 0.005 ± 0.010, and
¯
wmax = w + tw = 0.005 + 0.010 = 0.015 in
wmin = w tw = 0.005 0.010 = 0.005 in
¯
Thus, both clearance and interference are possible.
¯
(b) If wmin is to be 0.003 in, then, w = wmin + tw = 0.003 + 0.010 = 0.013 in. Thus,
¯¯¯¯¯
d = a b c w = 1.750 0.750 0.120 0.013 = 0.867 in
Answer
The previous example represented an absolute tolerance system. Statistically, gap
dimensions near the gap limits are rare events. Using a statistical tolerance system, the
probability that the gap falls within a given limit is determined.10 This probability deals
with the statistical distributions of the individual dimensions. For example, if the distributions of the dimensions in the previous example were normal and the tolerances, t, were
10
See Chapter 20 for a description of the statistical terminology.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
27
21
given in terms of standard deviations of the dimension distribution, the standard devia¯
tion of the gap w would be tw =
t 2 . However, this assumes a normal distribution
all
for the individual dimensions, a rare occurrence. To nd the distribution of w and/or the
probability of observing values of w within certain limits requires a computer simulation
in most cases. Monte Carlo computer simulations are used to determine the distribution
of w by the following approach:
Generate an instance for each dimension in the problem by selecting the value of
each dimension based on its probability distribution.
2 Calculate w using the values of the dimensions obtained in step 1.
3 Repeat steps 1 and 2 N times to generate the distribution of w . As the number of
trials increases, the reliability of the distribution increases.
1
114
Units
In the symbolic units equation for Newtons second law, F
ma,
2
F = M LT (14)
F stands for force, M for mass, L for length, and T for time. Units chosen for any three
of these quantities are called base units. The rst three having been chosen, the fourth
unit is called a derived unit. When force, length, and time are chosen as base units, the
mass is the derived unit and the system that results is called a gravitational system of
units. When mass, length, and time are chosen as base units, force is the derived unit
and the system that results is called an absolute system of units.
In some English-speaking countries, the U.S. customary foot-pound-second system
(fps) and the inch-pound-second system (ips) are the two standard gravitational systems
most used by engineers. In the fps system the unit of mass is
FT 2
(pound-force)(second)2
M=
=
= lbf · s2 /ft = slug
(15)
L
foot
Thus, length, time, and force are the three base units in the fps gravitational system.
The unit of force in the fps system is the pound, more properly the pound-force. We
shall often abbreviate this unit as lbf; the abbreviation lb is permissible however, since
we shall be dealing only with the U.S. customary gravitational system. In some branches
of engineering it is useful to represent 1000 lbf as a kilopound and to abbreviate it as
kip. Note: In Eq. (15) the derived unit of mass in the fps gravitational system is the
lbf · s2 /ft and is called a slug; there is no abbreviation for slug.
The unit of mass in the ips gravitational system is
(pound-force)(second)2
FT 2
=
= lbf · s2/in
M=
(16)
L
inch
The mass unit lbf · s2 /in has no ofcial name.
The International System of Units (SI) is an absolute system. The base units are the
meter, the kilogram (for mass), and the second. The unit of force is derived by using
Newtons second law and is called the newton. The units constituting the newton (N) are
F=
ML
(kilogram)(meter)
=
= kg · m /s2 = N
T2
(second)2
(17)
The weight of an object is the force exerted upon it by gravity. Designating the weight
as W and the acceleration due to gravity as g, we have
W = mg
(18)
28
22
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
In the fps system, standard gravity is g 32.1740 ft/s2. For most cases this is rounded
off to 32.2. Thus the weight of a mass of 1 slug in the fps system is
W = mg = (1 slug)(32.2 ft /s2 ) = 32.2 lbf
In the ips system, standard gravity is 386.088 or about 386 in/s2. Thus, in this system,
a unit mass weighs
W = (1 lbf · s2 /in)(386 in/s2 ) = 386 lbf
With SI units, standard gravity is 9.806 or about 9.81 m/s. Thus, the weight of a 1-kg
mass is
W = (1 kg)(9.81 m/s2 ) = 9.81 N
A series of names and symbols to form multiples and submultiples of SI units has
been established to provide an alternative to the writing of powers of 10. Table A1
includes these prexes and symbols.
Numbers having four or more digits are placed in groups of three and separated by
a space instead of a comma. However, the space may be omitted for the special case of
numbers having four digits. A period is used as a decimal point. These recommendations avoid the confusion caused by certain European countries in which a comma
is used as a decimal point, and by the English use of a centered period. Examples of
correct and incorrect usage are as follows:
1924 or 1 924 but not 1,924
0.1924 or 0.192 4 but not 0.192,4
192 423.618 50 but not 192,423.61850
The decimal point should always be preceded by a zero for numbers less than unity.
115
Calculations and Signicant Figures
The discussion in this section applies to real numbers, not integers. The accuracy of a real
number depends on the number of signicant gures describing the number. Usually, but
not always, three or four signicant gures are necessary for engineering accuracy. Unless
otherwise stated, no less than three signicant gures should be used in your calculations.
The number of signicant gures is usually inferred by the number of gures given
(except for leading zeros). For example, 706, 3.14, and 0.002 19 are assumed to be numbers with three signicant gures. For trailing zeros, a little more clarication is necessary. To display 706 to four signicant gures insert a trailing zero and display either
706.0, 7.060 × 102 , or 0.7060 × 103. Also, consider a number such as 91 600. Scientic
notation should be used to clarify the accuracy. For three signicant gures express the
number as 91.6 × 103. For four signicant gures express it as 91.60 × 103.
Computers and calculators display calculations to many signicant gures. However,
you should never report a number of signicant gures of a calculation any greater than
the smallest number of signicant gures of the numbers used for the calculation. Of
course, you should use the greatest accuracy possible when performing a calculation. For
example, determine the circumference of a solid shaft with a diameter of d = 0.40 in. The
circumference is given by C = πd . Since d is given with two signicant gures, C should
be reported with only two signicant gures. Now if we used only two signicant gures
for π our calculator would give C = 3.1 (0.40) = 1.24 in. This rounds off to two significant gures as C = 1.2 in. However, using π = 3.141 592 654 as programmed in the
calculator, C = 3.141 592 654 (0.40) = 1.256 637 061 in. This rounds off to C = 1.3
in, which is 8.3 percent higher than the rst calculation. Note, however, since d is given
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Introduction to Mechanical Engineering Design
29
23
with two signicant gures, it is implied that the range of d is 0.40 ± 0.005. This means
that the calculation of C is only accurate to within ±0.005/0.40 = ±0.0125 = ±1.25%.
The calculation could also be one in a series of calculations, and rounding each calculation separately may lead to an accumulation of greater inaccuracy. Thus, it is considered
good engineering practice to make all calculations to the greatest accuracy possible and
report the results within the accuracy of the given input.
116
Power Transmission Case Study Specications
A case study incorporating the many facets of the design process for a power transmission speed reducer will be considered throughout this textbook. The problem will be
introduced here with the denition and specication for the product to be designed.
Further details and component analysis will be presented in subsequent chapters.
Chapter 18 provides an overview of the entire process, focusing on the design sequence,
the interaction between the component designs, and other details pertinent to transmission of power. It also contains a complete case study of the power transmission speed
reducer introduced here.
Many industrial applications require machinery to be powered by engines or electric motors. The power source usually runs most efciently at a narrow range of rotational speed. When the application requires power to be delivered at a slower speed than
supplied by the motor, a speed reducer is introduced. The speed reducer should transmit
the power from the motor to the application with as little energy loss as practical, while
reducing the speed and consequently increasing the torque. For example, assume that a
company wishes to provide off-the-shelf speed reducers in various capacities and speed
ratios to sell to a wide variety of target applications. The marketing team has determined
a need for one of these speed reducers to satisfy the following customer requirements.
Design Requirements
Power to be delivered: 20 hp
Input speed: 1750 rev/min
Output speed: 85 rev/min
Targeted for uniformly loaded applications, such as conveyor belts, blowers,
and generators
Output shaft and input shaft in-line
Base mounted with 4 bolts
Continuous operation
6-year life, with 8 hours/day, 5 days/wk
Low maintenance
Competitive cost
Nominal operating conditions of industrialized locations
Input and output shafts standard size for typical couplings
In reality, the company would likely design for a whole range of speed ratios for
each power capacity, obtainable by interchanging gear sizes within the same overall
design. For simplicity, in this case study only one speed ratio will be considered.
Notice that the list of customer requirements includes some numerical specics, but
also includes some generalized requirements, e.g., low maintenance and competitive cost.
These general requirements give some guidance on what needs to be considered in the
design process, but are difcult to achieve with any certainty. In order to pin down these
nebulous requirements, it is best to further develop the customer requirements into a set of
product specications that are measurable. This task is usually achieved through the work
of a team including engineering, marketing, management, and customers. Various tools
30
24
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
may be used (see Footnote 1) to prioritize the requirements, determine suitable metrics to
be achieved, and to establish target values for each metric. The goal of this process is to
obtain a product specication that identies precisely what the product must satisfy. The
following product specications provide an appropriate framework for this design task.
Design Specications
Power to be delivered: 20 hp
Power efciency: >95%
Steady state input speed: 1750 rev/min
Maximum input speed: 2400 rev/min
Steady-state output speed: 8288 rev/min
Usually low shock levels, occasional moderate shock
Input and output shaft diameter tolerance: ±0.001 in
Output shaft and input shaft in-line: concentricity ±0.005 in, alignment
±0.001 rad
Maximum allowable loads on input shaft: axial, 50 lbf; transverse, 100 lbf
Maximum allowable loads on output shaft: axial, 50 lbf; transverse, 500 lbf
Base mounted with 4 bolts
Mounting orientation only with base on bottom
100% duty cycle
Maintenance schedule: lubrication check every 2000 hours; change of lubrication every 8000 hours of operation; gears and bearing life >12,000 hours;
innite shaft life; gears, bearings, and shafts replaceable
Access to check, drain, and rell lubrication without disassembly or opening of
gasketed joints.
Manufacturing cost per unit: < $300
Production: 10,000 units per year
Operating temperature range: 10 to 120 F
Sealed against water and dust from typical weather
Noise: < 85 dB from 1 meter
PROBLEMS
11
Select a mechanical component from Part 3 of this book (roller bearings, springs, etc.), go to your
universitys library or the appropriate internet website, and, using the Thomas Register of
American Manufacturers, report on the information obtained on ve manufacturers or suppliers.
12
Select a mechanical component from Part 3 of this book (roller bearings, springs, etc.), go to the
Internet, and, using a search engine, report on the information obtained on ve manufacturers or
suppliers.
13
Select an organization listed in Sec. 16, go to the Internet, and list what information is available
on the organization.
14
Go to the Internet and connect to the NSPE website (www.nspe.org). Read the full version of the
NSPE Code of Ethics for Engineers and briey discuss your reading.
15
Highway tunnel trafc (two parallel lanes in the same direction) experience indicates the average
spacing between vehicles increases with speed. Data from a New York tunnel show that between
15 and 35 mi/h, the space x between vehicles (in miles) is x = 0.324/(42.1 v) where v is the
vehicles speed in miles per hour.
(a) Ignoring the length of individual vehicles, what speed will give the tunnel the largest volume
in vehicles per hour?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
1. Introduction to
Mechanical Engineering
Design
Introduction to Mechanical Engineering Design
31
25
(b) Does including the length of the vehicles cut the tunnel capacity prediction signicantly?
Assume the average vehicle length is 10 ft.
(c) For part (b), does the optimal speed change much?
16
The engineering designer must create (invent) the concept and connectivity of the elements that
constitute a design, and not lose sight of the need to develop ideas with optimality in mind. A useful design attribute can be cost, which can be related to the amount of material used (volume or
weight). When you think about it, the weight is a function of the geometry and density. When the
design is solidied, nding the weight is a straightforward, sometimes tedious task. The gure
depicts a simple bracket frame that has supports that project from a wall column. The bracket supports a chain-fall hoist. Pinned joints are used to avoid bending. The cost of a link can be approximated by $ = ¢ Al γ , where ¢ is the cost of the link per unit weight, A is the cross-sectional area
of the prismatic link, l is the pin-to-pin link length, and γ is the specic weight of the material used.
To be sure, this is approximate because no decisions have been made concerning the geometric
form of the links or their ttings. By investigating cost now in this approximate way, one can detect
whether a particular set of proportions of the bracket (indexed by angle θ ) is advantageous. Is there
a preferable angle θ ? Show that the cost can be expressed as
$=
1 + cos2 θ
sin θ cos θ
γ ¢W l2
S
where W is the weight of the hoist and load, and S is the allowable tensile or compressive stress
in the link material (assume S = | Fi / A| and no column buckling action). What is the desirable
angle θ corresponding to the minimal cost?
l1
Problem 16
F1
(a) A chain-hoist bracket frame.
(b) Free body of pin.
F2
l2
W
(b)
(a )
17
When one knows the true values x1 and x2 and has approximations X 1 and X 2 at hand, one can
see where errors may arise. By viewing error as something to be added to an approximation to
attain a true value, it follows that the error ei , is related to X i , and xi as xi = X i + ei
(a) Show that the error in a sum X 1 + X 2 is
(x1 + x2 ) ( X 1 + X 2 ) = e1 + e2
(b) Show that the error in a difference X 1 X 2 is
(x1 x2 ) ( X 1 X 2 ) = e1 e2
(c) Show that the error in a product X 1 X 2 is
x1 x2 X 1 X 2 = X 1 X 2
e1
e2
+
X1
X2
(d ) Show that in a quotient X 1 / X 2 the error is
x1
X1
X1
=
x2
X2
X2
e2
e1
X1
X2
32
26
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
1. Introduction to
Mechanical Engineering
Design
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
18
Use the true values x1 = 5 and x2 = 6
(a) Demonstrate the correctness of the error equation from Prob. 17 for addition if three correct
digits are used for X 1 and X 2 .
(b) Demonstrate the correctness of the error equation for addition using three-digit signicant
numbers for X 1 and X 2 .
19
Convert the following to appropriate SI units:
(a) A stress of 20 000 psi.
(b) A force of 350 lbf.
(c) A moment of 1200 lbf in.
(d) An area of 2.4 in2 .
(e) A second moment of area of 17.4 in4 .
( f ) An area of 3.6 mi2 .
(g ) A modulus of elasticity of 21 Mpsi.
(h) A speed of 45 mi/h.
(i) A volume of 60 in3 .
110
Convert the following to appropriate ips units:
(a) A length of 1.5 m.
(b) A stress of 600 MPa.
(c) A pressure of 160 kPa.
(d) A section modulus of 1.84 (105 ) mm3 .
(e) A unit weight of 38.1 N/m.
( f ) A deection of 0.05 mm.
(g) A velocity of 6.12 m/s.
(h) A unit strain of 0.0021 m/m.
(i) A volume of 30 L.
111
Generally, nal design results are rounded to or xed to three digits because the given data cannot justify a greater display. In addition, prexes should be selected so as to limit number strings
to no more than four digits to the left of the decimal point. Using these rules, as well as those for
the choice of prexes, solve the following relations:
(a) σ = M / Z , where M = 200 N · m and Z = 15.3 × 103 mm3.
(b) σ = F / A , where F = 42 kN and A = 600 mm2 .
(c) y = Fl 3 /3 E I , where F = 1200 N, l = 800 mm, E = 207 GPa, and I = 64 × 103 mm4 .
(d ) θ = T l / G J , where J = π d 4 /32, T = 1100 N · m, l = 250 mm, G = 79.3 GPa, and d =
25 mm. Convert results to degrees of angle.
112
Repeat Prob. 111 for the following:
(a) σ = F /wt , where F = 600 N, w = 20 mm, and t = 6 mm.
(b) I = bh 3 /12, where b = 8 mm and h = 24 mm.
(c) I = π d 4 /64, where d = 32 mm.
(d ) τ = 16T /π d 3 , where T = 16 N m and d = 25 mm.
113
Repeat Prob. 111 for:
(a) τ = F / A , where A = π d 2 /4, F = 120 kN, and d = 20 mm.
(b) σ = 32 Fa /π d 3 , where F = 800 N, a = 800 mm, and d = 32 mm.
(c) Z = (π/32d )(d 4 di4 ) for d = 36 mm and di = 26 mm.
(d ) k = (d 4 G )/(8 D 3 N ) , where d = 1.6 mm, G = 79.3 GPa, D = 19.2 mm, and N = 32 (a
dimensionless number).
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
33
2
Materials
Chapter Outline
21
Material Strength and Stiffness
22
The Statistical Signicance of Material Properties
23
Strength and Cold Work
24
Hardness
25
Impact Properties
26
Temperature Effects
27
Numbering Systems
28
Sand Casting
41
29
Shell Molding
42
28
32
33
36
37
39
40
210
Investment Casting
211
Powder-Metallurgy Process
212
Hot-Working Processes
213
Cold-Working Processes
214
The Heat Treatment of Steel
215
Alloy Steels
216
Corrosion-Resistant Steels
217
Casting Materials
218
Nonferrous Metals
219
Plastics
220
Composite Materials
221
Materials Selection
42
42
43
44
44
47
48
49
51
54
55
56
27
34
28
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
The selection of a material for a machine part or a structural member is one of the most
important decisions the designer is called on to make. The decision is usually made
before the dimensions of the part are established. After choosing the process of creating the desired geometry and the material (the two cannot be divorced), the designer can
proportion the member so that loss of function can be avoided or the chance of loss of
function can be held to an acceptable risk.
In Chaps. 3 and 4, methods for estimating stresses and deections of machine
members are presented. These estimates are based on the properties of the material
from which the member will be made. For deections and stability evaluations, for
example, the elastic (stiffness) properties of the material are required, and evaluations
of stress at a critical location in a machine member require a comparison with the
strength of the material at that location in the geometry and condition of use. This
strength is a material property found by testing and is adjusted to the geometry and condition of use as necessary.
As important as stress and deection are in the design of mechanical parts, the
selection of a material is not always based on these factors. Many parts carry no loads
on them whatever. Parts may be designed merely to ll up space or for aesthetic qualities. Members must frequently be designed to also resist corrosion. Sometimes temperature effects are more important in design than stress and strain. So many other factors
besides stress and strain may govern the design of parts that the designer must have the
versatility that comes only with a broad background in materials and processes.
21
Material Strength and Stiffness
The standard tensile test is used to obtain a variety of material characteristics and
strengths that are used in design. Figure 2l illustrates a typical tension-test specimen
and its characteristic dimensions.1 The original diameter d0 and the gauge length l0 ,
used to measure the deections, are recorded before the test is begun. The specimen is
then mounted in the test machine and slowly loaded in tension while the load P and
deection are observed. The load is converted to stress by the calculation
σ=
P
A0
(21)
2
where A0 = 1 π d0 is the original area of the specimen.
4
d0
P
P
l0
Figure 21
A typical tension-test specimen. Some of the standard
dimensions used for d0 are 2.5, 6.25, and 12.5 mm
and 0.505 in, but other sections and sizes are in use.
Common gauge lengths l0 used are 10, 25, and 50 mm
and 1 and 2 in.
1
See ASTM standards E8 and E-8 m for standard dimensions.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
35
© The McGrawHill
Companies, 2008
2. Materials
Materials
29
The deection, or extension of the gage length, is given by l l0 where l is the
gauge length corresponding to the load P. The normal strain is calculated from
l l0
l0
ǫ=
(22)
At the conclusion of, or during, the test, the results are plotted as a stress-strain diagram. Figure 22 depicts typical stress-strain diagrams for ductile and brittle materials.
Ductile materials deform much more than brittle materials.
Point pl in Fig. 22a is called the proportional limit. This is the point at which the
curve rst begins to deviate from a straight line. No permanent set will be observable
in the specimen if the load is removed at this point. In the linear range, the uniaxial
stress-strain relation is given by Hookes law as
(23)
σ = Eǫ
where the constant of proportionality E, the slope of the linear part of the stress-strain
curve, is called Youngs modulus or the modulus of elasticity. E is a measure of the
stiffness of a material, and since strain is dimensionless, the units of E are the same as
stress. Steel, for example, has a modulus of elasticity of about 30 Mpsi (207 GPa)
regardless of heat treatment, carbon content, or alloying. Stainless steel is about
27.5 Mpsi (190 GPa).
Point el in Fig. 22 is called the elastic limit. If the specimen is loaded beyond this
point, the deformation is said to be plastic and the material will take on a permanent set
when the load is removed. Between pl and el the diagram is not a perfectly straight line,
even though the specimen is elastic.
During the tension test, many materials reach a point at which the strain begins to
increase very rapidly without a corresponding increase in stress. This point is called the
yield point. Not all materials have an obvious yield point, especially for brittle
materials. For this reason, yield strength Sy is often dened by an offset method as
shown in Fig. 22, where line a y is drawn at slope E. Point a corresponds to a denite
or stated amount of permanent set, usually 0.2 percent of the original gauge length
(ǫ = 0.002), although 0.01, 0.1, and 0.5 percent are sometimes used.
The ultimate, or tensile, strength Su or Sut corresponds to point u in Fig. 22 and
is the maximum stress reached on the stress-strain diagram.2 As shown in Fig. 22a,
Su
Sf
Sy
u
f
y
el
pl
Sut
Sy
u, f
y
Stress
Stress-strain diagram obtained
from the standard tensile test
(a) Ductile material; (b) brittle
material.
pl marks the proportional limit;
el, the elastic limit; y, the
offset-yield strength as dened
by offset strain a; u, the
maximum or ultimate strength;
and f, the fracture strength.
= P/A0
Figure 22
Oa
y
u
f
a
Strain
2
Strain
(a )
(b)
Usage varies. For a long time engineers used the term ultimate strength, hence the subscript u in Su or Sut .
However, in material science and metallurgy the term tensile strength is used.
36
30
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
some materials exhibit a downward trend after the maximum stress is reached and fracture at point f on the diagram. Others, such as some of the cast irons and high-strength
steels, fracture while the stress-strain trace is still rising, as shown in Fig. 22b, where
points u and f are identical.
As noted in Sec. 19, strength, as used in this book, is a built-in property of a material, or of a mechanical element, because of the selection of a particular material or
process or both. The strength of a connecting rod at the critical location in the geometry and condition of use, for example, is the same no matter whether it is already an element in an operating machine or whether it is lying on a workbench awaiting assembly
with other parts. On the other hand, stress is something that occurs in a part, usually as
a result of its being assembled into a machine and loaded. However, stresses may be
built into a part by processing or handling. For example, shot peening produces a compressive stress in the outer surface of a part, and also improves the fatigue strength of
the part. Thus, in this book we will be very careful in distinguishing between strength,
designated by S, and stress, designated by σ or τ .
The diagrams in Fig. 22 are called engineering stress-strain diagrams because the
stresses and strains calculated in Eqs. (21) and (22) are not true values. The stress
calculated in Eq. (21) is based on the original area before the load is applied. In reality, as the load is applied the area reduces so that the actual or true stress is larger than
the engineering stress. To obtain the true stress for the diagram the load and the crosssectional area must be measured simultaneously during the test. Figure 22a represents
a ductile material where the stress appears to decrease from points u to f. Typically,
beyond point u the specimen begins to neck at a location of weakness where the area
reduces dramatically, as shown in Fig. 23. For this reason, the true stress is much higher than the engineering stress at the necked section.
The engineering strain given by Eq. (22) is based on net change in length from the
original length. In plotting the true stress-strain diagram, it is customary to use a term
called true strain or, sometimes, logarithmic strain. True strain is the sum of the incremental elongations divided by the current gauge length at load P, or
l
ε=
l0
dl
l
= ln
l
l0
(24)
where the symbol ε is used to represent true strain. The most important characteristic
of a true stress-strain diagram (Fig. 24) is that the true stress continually increases all
the way to fracture. Thus, as shown in Fig. 24, the true fracture stress σ f is greater than
the true ultimate stress σu . Contrast this with Fig. 22a, where the engineering fracture
strength S f is less than the engineering ultimate strength Su .
Compression tests are more difcult to conduct, and the geometry of the test specimens differs from the geometry of those used in tension tests. The reason for this is that
the specimen may buckle during testing or it may be difcult to distribute the stresses
evenly. Other difculties occur because ductile materials will bulge after yielding.
However, the results can be plotted on a stress-strain diagram also, and the same
strength denitions can be applied as used in tensile testing. For most ductile materials
the compressive strengths are about the same as the tensile strengths. When substantial
differences occur between tensile and compressive strengths, however, as is the case with
Figure 23
Tension specimen after
necking.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
37
© The McGrawHill
Companies, 2008
2. Materials
Materials
31
Figure 24
f
f
True stress-strain diagram
plotted in Cartesian
coordinates.
u
True stress
u
u
f
True strain
the cast irons, the tensile and compressive strengths should be stated separately, Sut ,
Suc , where Suc is reported as a positive quantity.
Torsional strengths are found by twisting solid circular bars and recording the torque
and the twist angle. The results are then plotted as a torque-twist diagram. The shear
stresses in the specimen are linear with respect to radial location, being zero at the center of the specimen and maximum at the outer radius r (see Chap. 3). The maximum shear
stress τmax is related to the angle of twist θ by
τmax =
Gr
θ
l0
(25)
where θ is in radians, r is the radius of the specimen, l0 is the gauge length, and G is
the material stiffness property called the shear modulus or the modulus of rigidity. The
maximum shear stress is also related to the applied torque T as
τmax =
Tr
J
(26)
where J = 1 π r 4 is the polar second moment of area of the cross section.
2
The torque-twist diagram will be similar to Fig. 22, and, using Eqs. (25) and
(26), the modulus of rigidity can be found as well as the elastic limit and the torsional
yield strength Ssy . The maximum point on a torque-twist diagram, corresponding to
point u on Fig. 22, is Tu . The equation
Ssu =
Tu r
J
(27)
denes the modulus of rupture for the torsion test. Note that it is incorrect to call Ssu
the ultimate torsional strength, as the outermost region of the bar is in a plastic state at
the torque Tu and the stress distribution is no longer linear.
All of the stresses and strengths dened by the stress-strain diagram of Fig. 22 and
similar diagrams are specically known as engineering stresses and strengths or nominal stresses and strengths. These are the values normally used in all engineering design
calculations. The adjectives engineering and nominal are used here to emphasize that
the stresses are computed by using the original or unstressed cross-sectional area of the
specimen. In this book we shall use these modiers only when we specically wish to
call attention to this distinction.
38
32
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
The Statistical Signicance of Material Properties
22
There is some subtlety in the ideas presented in the previous section that should be pondered carefully before continuing. Figure 22 depicts the result of a single tension test
(one specimen, now fractured). It is common for engineers to consider these important
stress values (at points pl, el, y , u , and f) as properties and to denote them as strengths
with a special notation, uppercase S, in lieu of lowercase sigma σ , with subscripts
added: Spl for proportional limit, Sy for yield strength, Su for ultimate tensile strength
( Sut or Suc , if tensile or compressive sense is important).
If there were 1000 nominally identical specimens, the values of strength obtained
would be distributed between some minimum and maximum values. It follows that the
description of strength, a material property, is distributional and thus is statistical in
nature. Chapter 20 provides more detail on statistical considerations in design. Here we
will simply describe the results of one example, Ex. 20-4. Consider the following table,
which is a histographic report containing the maximum stresses of 1000 tensile tests on
a 1020 steel from a single heat. Here we are seeking the ultimate tensile strength Sut .
The class frequency is the number of occurrences within a 1 kpsi range given by the
class midpoint. Thus, 18 maximum stress values occurred in the range of 57 to 58 kpsi.
Class Frequency f i
Class Midpoint
xi , kpsi
2
18
23
31
83
109 138 151 139 130
82
49
28
11
4
2
56.5 57.5 58.5 59.5 60.5 61.5 62.5 63.5 64.5 65.5 66.5 67.5 68.5 69.5 70.5 71.5
The probability density is dened as the number of occurrences divided by the total
sample number. The bar chart in Fig. 25 depicts the histogram of the probability density. If the data is in the form of a Gaussian or normal distribution, the probability
density function determined in Ex. 20-4 is
f (x ) =
1
1
exp
2
2.594 2π
x 63.62
2.594
2
where the mean stress is 63.62 kpsi and the standard deviation is 2.594 kpsi. A plot
of f (x ) is included in Fig. 25. The description of the strength Sut is then expressed
in terms of its statistical parameters and its distribution type. In this case
Sut = N(63.62, 2.594) kpsi.
Note that the test program has described 1020 property Sut, for only one heat of
one supplier. Testing is an involved and expensive process. Tables of properties are
often prepared to be helpful to other persons. A statistical quantity is described by its
mean, standard deviation, and distribution type. Many tables display a single number,
which is often the mean, minimum, or some percentile, such as the 99th percentile.
Always read the foonotes to the table. If no qualication is made in a single-entry table,
the table is subject to serious doubt.
Since it is no surprise that useful descriptions of a property are statistical in nature,
engineers, when ordering property tests, should couch the instructions so the data generated are enough for them to observe the statistical parameters and to identify the distributional characteristic. The tensile test program on 1000 specimens of 1020 steel is a
large one. If you were faced with putting something in a table of ultimate tensile
strengths and constrained to a single number, what would it be and just how would your
footnote read?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
39
© The McGrawHill
Companies, 2008
2. Materials
33
Materials
0.2
Figure 25
Histogram for 1000 tensile
tests on a 1020 steel from a
single heat.
Probability density
f(x )
0.1
0
50
60
70
Ultimate tensile strength, kpsi
23
Strength and Cold Work
Cold working is the process of plastic straining below the recrystallization temperature
in the plastic region of the stress-strain diagram. Materials can be deformed plastically
by the application of heat, as in blacksmithing or hot rolling, but the resulting mechanical properties are quite different from those obtained by cold working. The purpose of
this section is to explain what happens to the signicant mechanical properties of a
material when that material is cold-worked.
Consider the stress-strain diagram of Fig. 26a. Here a material has been stressed
beyond the yield strength at y to some point i, in the plastic region, and then the load
removed. At this point the material has a permanent plastic deformation ǫ p . If the load
corresponding to point i is now reapplied, the material will be elastically deformed by
Figure 26
u
Pu
u
i
Pi
i
i
f
Sy
f
Py
y
y
Load, P
Nominal stress,
(a) Stress-strain diagram
showing unloading and
reloading at point i in the
plastic region; (b) analogous
load-deformation diagram.
Su
O
p
Unit strain,
e
A0
Ai
Af
Ai
Area deformation (reduction)
(a )
(b)
40
34
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
the amount ǫe . Thus at point i the total unit strain consists of the two components ǫ p and
ǫe and is given by the equation
ǫ = ǫ p + ǫe
(a)
This material can be unloaded and reloaded any number of times from and to point i,
and it is found that the action always occurs along the straight line that is approximately parallel to the initial elastic line Oy. Thus
ǫe =
σi
E
(b)
The material now has a higher yield point, is less ductile as a result of a reduction in
strain capacity, and is said to be strain-hardened. If the process is continued, increasing
ǫ p , the material can become brittle and exhibit sudden fracture.
It is possible to construct a similar diagram, as in Fig. 26b, where the abscissa is
the area deformation and the ordinate is the applied load. The reduction in area corresponding to the load Pf , at fracture, is dened as
R=
A0 A f
Af
=1
A0
A0
(28)
where A0 is the original area. The quantity R in Eq. (28) is usually expressed in percent and tabulated in lists of mechanical properties as a measure of ductility. See
Appendix Table A20, for example. Ductility is an important property because it measures the ability of a material to absorb overloads and to be cold-worked. Thus such
operations as bending, drawing, heading, and stretch forming are metal-processing
operations that require ductile materials.
Figure 26b can also be used to dene the quantity of cold work. The cold-work
factor W is dened as
W=
A0 Ai
A 0 Ai
A0
A0
(29)
where Ai corresponds to the area after the load Pi has been released. The approximation in Eq. (29) results because of the difculty of measuring the small diametral
changes in the elastic region. If the amount of cold work is known, then Eq. (29) can
be solved for the area Ai . The result is
Ai = A0 (1 W )
(210)
Cold working a material produces a new set of values for the strengths, as can
be seen from stress-strain diagrams. Datsko3 describes the plastic region of the true
stresstrue strain diagram by the equation
σ = σ0 εm
3
(211)
Joseph Datsko, Solid Materials, Chap. 32 in Joseph E. Shigley, Charles R. Mischke, and Thomas H.
Brown, Jr. (eds.), Standard Handbook of Machine Design, 3rd ed., McGraw-Hill, New York, 2004. See also
Joseph Datsko, New Look at Material Strength, Machine Design, vol. 58, no. 3, Feb. 6, 1986, pp. 8185.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
41
© The McGrawHill
Companies, 2008
2. Materials
35
Materials
where
σ = true stress
σ0 = a strength coefcient, or strain-strengthening coefcient
ε = true plastic strain
m = strain-strengthening exponent
It can be shown4 that
(212)
m = εu
provided that the load-deformation curve exhibits a stationary point (a place of zero
slope).
Difculties arise when using the gauge length to evaluate the true strain in the
plastic range, since necking causes the strain to be nonuniform. A more satisfactory
relation can be obtained by using the area at the neck. Assuming that the change in volume of the material is small, Al = A0 l0 . Thus, l / l0 = A0 / A , and the true strain is
given by
ε = ln
l
A0
= ln
l0
A
(213)
Returning to Fig. 26b, if point i is to the left of point u, that is, Pi < Pu , then the
new yield strength is
Sy =
Pi
= σ0 εim
Ai
Pi Pu
(214)
Because of the reduced area, that is, because Ai < A0 , the ultimate strength also
changes, and is
Su =
Pu
Ai
(c)
Since Pu = Su A0 , we nd, with Eq. (210), that
Su =
Su
Su A0
=
A0 (1 W )
1W
εi εu
(215)
which is valid only when point i is to the left of point u.
For points to the right of u, the yield strength is approaching the ultimate strength,
and, with small loss in accuracy,
.
.
Su = Sy = σ0 εim
εi εu
(216)
A little thought will reveal that a bar will have the same ultimate load in tension after
being strain-strengthened in tension as it had before. The new strength is of interest to
us not because the static ultimate load increases, butsince fatigue strengths are correlated with the local ultimate strengthsbecause the fatigue strength improves. Also
the yield strength increases, giving a larger range of sustainable elastic loading.
4
See Sec. 52, J. E. Shigley and C. R. Mischke, Mechanical Engineering Design, 6th ed., McGraw-Hill,
New York, 2001.
42
36
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
EXAMPLE 21
Solution
An annealed AISI 1018 steel (see Table A22) has Sy = 32.0 kpsi, Su = 49.5 kpsi,
σ f = 91.1 kpsi, σ0 = 90 kpsi, m = 0.25, and ε f = 1.05 in/in. Find the new values of
the strengths if the material is given 15 percent cold work.
From Eq. (212), we nd the true strain corresponding to the ultimate strength to be
εu = m = 0.25
The ratio A0 / Ai is, from Eq. (29),
A0
1
1
=
= 1.176
=
Ai
1W
1 0.15
The true strain corresponding to 15 percent cold work is obtained from Eq. (213). Thus
εi = ln
A0
= ln 1.176 = 0.1625
Ai
Since εi < εu , Eqs. (214) and (215) apply. Therefore,
Answer
Sy = σ0 εim = 90(0.1625)0.25 = 57.1 kpsi
Answer
Su =
24
49.5
Su
=
= 58.2 kpsi
1W
1 0.15
Hardness
The resistance of a material to penetration by a pointed tool is called hardness. Though
there are many hardness-measuring systems, we shall consider here only the two in
greatest use.
Rockwell hardness tests are described by ASTM standard hardness method E18
and measurements are quickly and easily made, they have good reproducibility, and the
test machine for them is easy to use. In fact, the hardness number is read directly from
a dial. Rockwell hardness scales are designated as A , B , C , . . . , etc. The indenters are
1
described as a diamond, a 16 -in-diameter ball, and a diamond for scales A, B, and C,
respectively, where the load applied is either 60, 100, or 150 kg. Thus the Rockwell B
1
scale, designated R B , uses a 100-kg load and a No. 2 indenter, which is a 16 -in-diameter
ball. The Rockwell C scale RC uses a diamond cone, which is the No. 1 indenter, and
a load of 150 kg. Hardness numbers so obtained are relative. Therefore a hardness
RC = 50 has meaning only in relation to another hardness number using the same scale.
The Brinell hardness is another test in very general use. In testing, the indenting
tool through which force is applied is a ball and the hardness number HB is found as
a number equal to the applied load divided by the spherical surface area of the indentation. Thus the units of HB are the same as those of stress, though they are seldom
used. Brinell hardness testing takes more time, since HB must be computed from the
test data. The primary advantage of both methods is that they are nondestructive in
most cases. Both are empirically and directly related to the ultimate strength of the
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
43
© The McGrawHill
Companies, 2008
2. Materials
37
Materials
material tested. This means that the strength of parts could, if desired, be tested part
by part during manufacture.
For steels, the relationship between the minimum ultimate strength and the Brinell
hardness number for 200 HB 450 is found to be
Su =
0.495 HB
3.41 HB
kpsi
MPa
(217)
Similar relationships for cast iron can be derived from data supplied by Krause.5
Data from 72 tests of gray iron produced by one foundry and poured in two sizes of test
bars are reported in graph form. The minimum strength, as dened by the ASTM, is
found from these data to be
Su =
0.23 HB 12.5 kpsi
1.58 HB 86 MPa
(218)
Walton6 shows a chart from which the SAE minimum strength can be obtained. The
result is
Su = 0.2375 HB 16 kpsi
(219)
which is even more conservative than the values obtained from Eq. (218).
EXAMPLE 22
Solution
It is necessary to ensure that a certain part supplied by a foundry always meets or
exceeds ASTM No. 20 specications for cast iron (see Table A24). What hardness
should be specied?
From Eq. (218), with (Su)min = 20 kpsi, we have
Answer
HB =
20 + 12.5
Su + 12.5
=
= 141
0.23
0.23
If the foundry can control the hardness within 20 points, routinely, then specify
145 < HB < 165. This imposes no hardship on the foundry and assures the designer
that ASTM grade 20 will always be supplied at a predictable cost.
25
Impact Properties
An external force applied to a structure or part is called an impact load if the time of
application is less than one-third the lowest natural period of vibration of the part or
structure. Otherwise it is called simply a static load.
5
D. E. Krause, Gray IronA Unique Engineering Material, ASTM Special Publication 455, 1969,
pp. 329, as reported in Charles F. Walton (ed.), Iron Castings Handbook, Iron Founders Society, Inc.,
Cleveland, 1971, pp. 204, 205.
6
Ibid.
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
A mean trace shows the effect
of temperature on impact
values. The result of interest is
the brittle-ductile transition
temperature, often dened as
the temperature at which the
mean trace passes through the
15 ft · lbf level. The critical
temperature is dependent on
the geometry of the notch,
which is why the Charpy
V notch is closely dened.
60
40
20
0
400
200
0
200
400
Temperature, °F
Figure 28
100
Inuence of strain rate on
tensile properties.
100
80
80
Ratio, Sy /Su
Strength, kpsi
Ultimate
strength, Su
60
Ratio, Sy /Su , %
Figure 27
The Charpy (commonly used) and Izod (rarely used) notched-bar tests utilize bars of
specied geometries to determine brittleness and impact strength. These tests are helpful
in comparing several materials and in the determination of low-temperature brittleness. In
both tests the specimen is struck by a pendulum released from a xed height, and the
energy absorbed by the specimen, called the impact value, can be computed from the
height of swing after fracture, but is read from a dial that essentially computes the result.
The effect of temperature on impact values is shown in Fig. 27 for a material
showing a ductile-brittle transition. Not all materials show this transition. Notice the
narrow region of critical temperatures where the impact value increases very rapidly. In
the low-temperature region the fracture appears as a brittle, shattering type, whereas the
appearance is a tough, tearing type above the critical-temperature region. The critical
temperature seems to be dependent on both the material and the geometry of the notch.
For this reason designers should not rely too heavily on the results of notched-bar tests.
The average strain rate used in obtaining the stress-strain diagram is about
0.001 in/(in · s) or less. When the strain rate is increased, as it is under impact conditions,
the strengths increase, as shown in Fig. 28. In fact, at very high strain rates the yield
strength seems to approach the ultimate strength as a limit. But note that the curves show
little change in the elongation. This means that the ductility remains about the same.
Also, in view of the sharp increase in yield strength, a mild steel could be expected to
behave elastically throughout practically its entire strength range under impact conditions.
60
Total elongation
40
40
Yield strength, Sy
20
0
106
20
104
102
1
1
Strain rate, s
10 2
0
10 4
Elongation, %
38
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
Charpy, ft lbf
44
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
45
© The McGrawHill
Companies, 2008
2. Materials
Materials
39
The Charpy and Izod tests really provide toughness data under dynamic, rather than
static, conditions. It may well be that impact data obtained from these tests are as dependent on the notch geometry as they are on the strain rate. For these reasons it may be better to use the concepts of notch sensitivity, fracture toughness, and fracture mechanics,
discussed in Chaps. 5 and 6, to assess the possibility of cracking or fracture.
26
Temperature Effects
Strength and ductility, or brittleness, are properties affected by the temperature of the
operating environment.
The effect of temperature on the static properties of steels is typied by the strength
versus temperature chart of Fig. 29. Note that the tensile strength changes only a small
amount until a certain temperature is reached. At that point it falls off rapidly. The yield
strength, however, decreases continuously as the environmental temperature is increased.
There is a substantial increase in ductility, as might be expected, at the higher temperatures.
Many tests have been made of ferrous metals subjected to constant loads for long
periods of time at elevated temperatures. The specimens were found to be permanently
deformed during the tests, even though at times the actual stresses were less than the
yield strength of the material obtained from short-time tests made at the same temperature. This continuous deformation under load is called creep.
One of the most useful tests to have been devised is the long-time creep test under
constant load. Figure 210 illustrates a curve that is typical of this kind of test. The
curve is obtained at a constant stated temperature. A number of tests are usually run
simultaneously at different stress intensities. The curve exhibits three distinct regions.
In the rst stage are included both the elastic and the plastic deformation. This stage shows
a decreasing creep rate, which is due to the strain hardening. The second stage shows
a constant minimum creep rate caused by the annealing effect. In the third stage the
specimen shows a considerable reduction in area, the true stress is increased, and a
higher creep eventually leads to fracture.
When the operating temperatures are lower than the transition temperature
(Fig. 27), the possibility arises that a part could fail by a brittle fracture. This subject
will be discussed in Chap. 5.
Figure 29
1.0
Sut
0.9
ST /SRT
A plot of the results of 145 tests
of 21 carbon and alloy steels
showing the effect of operating
temperature on the yield
strength Sy and the ultimate
strength Sut . The ordinate is the
ratio of the strength at the
operating temperature to the
strength at room temperature.
The standard deviations were
0.0442 σ Sy 0.152 for
ˆ
ˆ
Sy and 0.099 σ Sut 0.11
for Sut . (Data source: E. A.
Brandes (ed.), Smithells Metal
Reference Book, 6th ed.,
Butterworth, London, 1983
pp. 22128 to 22131.)
Sy
0.8
0.7
0.6
0.5
0
RT
200
400
Temperature, °C
600
46
40
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Figure 210
Creep deformation
Creep-time curve.
1st stage
2nd stage
3rd stage
Time
Of course, heat treatment, as will be shown, is used to make substantial changes in
the mechanical properties of a material.
Heating due to electric and gas welding also changes the mechanical properties.
Such changes may be due to clamping during the welding process, as well as heating;
the resulting stresses then remain when the parts have cooled and the clamps have been
removed. Hardness tests can be used to learn whether the strength has been changed by
welding, but such tests will not reveal the presence of residual stresses.
27
Numbering Systems
The Society of Automotive Engineers (SAE) was the rst to recognize the need, and to
adopt a system, for the numbering of steels. Later the American Iron and Steel Institute
(AISI) adopted a similar system. In 1975 the SAE published the Unied Numbering
System for Metals and Alloys (UNS); this system also contains cross-reference numbers for other material specications.7 The UNS uses a letter prex to designate the
material, as, for example, G for the carbon and alloy steels, A for the aluminum alloys,
C for the copper-base alloys, and S for the stainless or corrosion-resistant steels. For
some materials, not enough agreement has as yet developed in the industry to warrant
the establishment of a designation.
For the steels, the rst two numbers following the letter prex indicate the composition, excluding the carbon content. The various compositions used are as follows:
G10
Plain carbon
G46
Nickel-molybdenum
G11
Free-cutting carbon steel with
more sulfur or phosphorus
G48
Nickel-molybdenum
G50
Chromium
G13
Manganese
G51
Chromium
G23
Nickel
G52
Chromium
G25
Nickel
G61
Chromium-vanadium
G31
Nickel-chromium
G86
Chromium-nickel-molybdenum
G33
Nickel-chromium
G87
Chromium-nickel-molybdenum
G40
Molybdenum
G92
Manganese-silicon
G41
Chromium-molybdenum
G94
Nickel-chromium-molybdenum
G43
Nickel-chromium-molybdenum
7
Many of the materials discussed in the balance of this chapter are listed in the Appendix tables. Be sure to
review these.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
47
© The McGrawHill
Companies, 2008
2. Materials
Materials
Table 21
Aluminum 99.00% pure and greater
Aluminum Alloy
Designations
41
Ax1xxx
Copper alloys
Ax2xxx
Manganese alloys
Ax3xxx
Silicon alloys
Ax4xxx
Magnesium alloys
Ax5xxx
Magnesium-silicon alloys
Ax6xxx
Zinc alloys
Ax7xxx
The second number pair refers to the approximate carbon content. Thus, G10400 is a
plain carbon steel with a nominal carbon content of 0.40 percent (0.37 to 0.44 percent).
The fth number following the prex is used for special situations. For example, the old
designation AISI 52100 represents a chromium alloy with about 100 points of carbon.
The UNS designation is G52986.
The UNS designations for the stainless steels, prex S, utilize the older AISI designations for the rst three numbers following the prex. The next two numbers are
reserved for special purposes. The rst number of the group indicates the approximate
composition. Thus 2 is a chromium-nickel-manganese steel, 3 is a chromium-nickel
steel, and 4 is a chromium alloy steel. Sometimes stainless steels are referred to by their
alloy content. Thus S30200 is often called an 18-8 stainless steel, meaning 18 percent
chromium and 8 percent nickel.
The prefix for the aluminum group is the letter A. The first number following the
prefix indicates the processing. For example, A9 is a wrought aluminum, while A0 is
a casting alloy. The second number designates the main alloy group as shown in
Table 21. The third number in the group is used to modify the original alloy or to
designate the impurity limits. The last two numbers refer to other alloys used with the
basic group.
The American Society for Testing and Materials (ASTM) numbering system for
cast iron is in widespread use. This system is based on the tensile strength. Thus ASTM
A18 speaks of classes; e.g., 30 cast iron has a minimum tensile strength of 30 kpsi. Note
from Appendix A-24, however, that the typical tensile strength is 31 kpsi. You should
be careful to designate which of the two values is used in design and problem work
because of the signicance of factor of safety.
28
Sand Casting
Sand casting is a basic low-cost process, and it lends itself to economical production
in large quantities with practically no limit to the size, shape, or complexity of the part
produced.
In sand casting, the casting is made by pouring molten metal into sand molds. A
pattern, constructed of metal or wood, is used to form the cavity into which the molten
metal is poured. Recesses or holes in the casting are produced by sand cores introduced
into the mold. The designer should make an effort to visualize the pattern and casting
in the mold. In this way the problems of core setting, pattern removal, draft, and solidication can be studied. Castings to be used as test bars of cast iron are cast separately
and properties may vary.
48
42
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Steel castings are the most difcult of all to produce, because steel has the highest
melting temperature of all materials normally used for casting. This high temperature
aggravates all casting problems.
The following rules will be found quite useful in the design of any sand casting:
1
2
3
4
All sections should be designed with a uniform thickness.
The casting should be designed so as to produce a gradual change from section
to section where this is necessary.
Adjoining sections should be designed with generous llets or radii.
A complicated part should be designed as two or more simple castings to be
assembled by fasteners or by welding.
Steel, gray iron, brass, bronze, and aluminum are most often used in castings. The
minimum wall thickness for any of these materials is about 5 mm, though with particular care, thinner sections can be obtained with some materials.
29
Shell Molding
The shell-molding process employs a heated metal pattern, usually made of cast iron,
aluminum, or brass, which is placed in a shell-molding machine containing a mixture
of dry sand and thermosetting resin. The hot pattern melts the plastic, which, together
with the sand, forms a shell about 5 to 10 mm thick around the pattern. The shell is then
baked at from 400 to 700°F for a short time while still on the pattern. It is then stripped
from the pattern and placed in storage for use in casting.
In the next step the shells are assembled by clamping, bolting, or pasting; they are
placed in a backup material, such as steel shot; and the molten metal is poured into the
cavity. The thin shell permits the heat to be conducted away so that solidication takes
place rapidly. As solidication takes place, the plastic bond is burned and the mold collapses. The permeability of the backup material allows the gases to escape and the casting to air-cool. All this aids in obtaining a ne-grain, stress-free casting.
Shell-mold castings feature a smooth surface, a draft that is quite small, and close
tolerances. In general, the rules governing sand casting also apply to shell-mold casting.
210
Investment Casting
Investment casting uses a pattern that may be made from wax, plastic, or other material.
After the mold is made, the pattern is melted out. Thus a mechanized method of casting
a great many patterns is necessary. The mold material is dependent upon the melting
point of the cast metal. Thus a plaster mold can be used for some materials while
others would require a ceramic mold. After the pattern is melted out, the mold is baked
or red; when ring is completed, the molten metal may be poured into the hot mold
and allowed to cool.
If a number of castings are to be made, then metal or permanent molds may be suitable. Such molds have the advantage that the surfaces are smooth, bright, and accurate,
so that little, if any, machining is required. Metal-mold castings are also known as die
castings and centrifugal castings.
211
Powder-Metallurgy Process
The powder-metallurgy process is a quantity-production process that uses powders
from a single metal, several metals, or a mixture of metals and nonmetals. It consists
essentially of mechanically mixing the powders, compacting them in dies at high pressures,
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
49
© The McGrawHill
Companies, 2008
2. Materials
Materials
43
and heating the compacted part at a temperature less than the melting point of the major
ingredient. The particles are united into a single strong part similar to what would be
obtained by melting the same ingredients together. The advantages are (1) the elimination of scrap or waste material, (2) the elimination of machining operations, (3) the low
unit cost when mass-produced, and (4) the exact control of composition. Some of the
disadvantages are (1) the high cost of dies, (2) the lower physical properties, (3) the
higher cost of materials, (4) the limitations on the design, and (5) the limited range of
materials that can be used. Parts commonly made by this process are oil-impregnated
bearings, incandescent lamp laments, cemented-carbide tips for tools, and permanent
magnets. Some products can be made only by powder metallurgy: surgical implants, for
example. The structure is different from what can be obtained by melting the same
ingredients.
212
Hot-Working Processes
By hot working are meant such processes as rolling, forging, hot extrusion, and hot
pressing, in which the metal is heated above its recrystallation temperature.
Hot rolling is usually used to create a bar of material of a particular shape and
dimension. Figure 211 shows some of the various shapes that are commonly produced
by the hot-rolling process. All of them are available in many different sizes as well as
in different materials. The materials most available in the hot-rolled bar sizes are steel,
aluminum, magnesium, and copper alloys.
Tubing can be manufactured by hot-rolling strip or plate. The edges of the strip are
rolled together, creating seams that are either butt-welded or lap-welded. Seamless tubing is manufactured by roll-piercing a solid heated rod with a piercing mandrel.
Extrusion is the process by which great pressure is applied to a heated metal billet
or blank, causing it to ow through a restricted orice. This process is more common
with materials of low melting point, such as aluminum, copper, magnesium, lead, tin,
and zinc. Stainless steel extrusions are available on a more limited basis.
Forging is the hot working of metal by hammers, presses, or forging machines. In
common with other hot-working processes, forging produces a rened grain structure
that results in increased strength and ductility. Compared with castings, forgings have
greater strength for the same weight. In addition, drop forgings can be made smoother
and more accurate than sand castings, so that less machining is necessary. However, the
initial cost of the forging dies is usually greater than the cost of patterns for castings,
although the greater unit strength rather than the cost is usually the deciding factor
between these two processes.
Figure 211
Common shapes available
through hot rolling.
Round
Square
Half oval
Flat
Hexagon
(a) Bar shapes
Wide flange
Channel
Angle
(b) Structural shapes
Tee
Zee
50
44
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 212
100
Cold-drawn
80
Strength, kpsi
Stress-strain diagram for
hot-rolled and cold-drawn
UNS G10350 steel.
Hot-rolled
Yield point
60
Yield point
40
20
0
0
0.2
0.4
0.6
Elongation, in
213
Cold-Working Processes
By cold working is meant the forming of the metal while at a low temperature (usually
room temperature). In contrast to parts produced by hot working, cold-worked parts
have a bright new nish, are more accurate, and require less machining.
Cold-nished bars and shafts are produced by rolling, drawing, turning, grinding,
and polishing. Of these methods, by far the largest percentage of products are made by
the cold-rolling and cold-drawing processes. Cold rolling is now used mostly for the
production of wide ats and sheets. Practically all cold-nished bars are made by cold
drawing but even so are sometimes mistakenly called cold-rolled bars. In the drawing
process, the hot-rolled bars are rst cleaned of scale and then drawn by pulling them
1
1
through a die that reduces the size about 32 to 16 in. This process does not remove
material from the bar but reduces, or draws down, the size. Many different shapes of
hot-rolled bars may be used for cold drawing.
Cold rolling and cold drawing have the same effect upon the mechanical properties. The cold-working process does not change the grain size but merely distorts it.
Cold working results in a large increase in yield strength, an increase in ultimate
strength and hardness, and a decrease in ductility. In Fig. 212 the properties of a colddrawn bar are compared with those of a hot-rolled bar of the same material.
Heading is a cold-working process in which the metal is gathered, or upset. This
operation is commonly used to make screw and rivet heads and is capable of producing
a wide variety of shapes. Roll threading is the process of rolling threads by squeezing
and rolling a blank between two serrated dies. Spinning is the operation of working sheet
material around a rotating form into a circular shape. Stamping is the term used to
describe punch-press operations such as blanking, coining, forming, and shallow
drawing.
214
The Heat Treatment of Steel
Heat treatment of steel refers to time- and temperature-controlled processes that relieve
residual stresses and/or modies material properties such as hardness (strength), ductility, and toughness. Other mechanical or chemical operations are sometimes grouped
under the heading of heat treatment. The common heat-treating operations are annealing, quenching, tempering, and case hardening.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
51
© The McGrawHill
Companies, 2008
Materials
45
Annealing
When a material is cold- or hot-worked, residual stresses are built in, and, in addition,
the material usually has a higher hardness as a result of these working operations. These
operations change the structure of the material so that it is no longer represented by the
equilibrium diagram. Full annealing and normalizing is a heating operation that permits
the material to transform according to the equilibrium diagram. The material to be
annealed is heated to a temperature that is approximately 100°F above the critical temperature. It is held at this temperature for a time that is sufcient for the carbon to
become dissolved and diffused through the material. The object being treated is then
allowed to cool slowly, usually in the furnace in which it was treated. If the transformation is complete, then it is said to have a full anneal. Annealing is used to soften a
material and make it more ductile, to relieve residual stresses, and to rene the grain
structure.
The term annealing includes the process called normalizing. Parts to be normalized
may be heated to a slightly higher temperature than in full annealing. This produces a
coarser grain structure, which is more easily machined if the material is a low-carbon
steel. In the normalizing process the part is cooled in still air at room temperature. Since
this cooling is more rapid than the slow cooling used in full annealing, less time is available for equilibrium, and the material is harder than fully annealed steel. Normalizing
is often used as the nal treating operation for steel. The cooling in still air amounts to
a slow quench.
Quenching
Eutectoid steel that is fully annealed consists entirely of pearlite, which is obtained
from austenite under conditions of equilibrium. A fully annealed hypoeutectoid steel
would consist of pearlite plus ferrite, while hypereutectoid steel in the fully annealed
condition would consist of pearlite plus cementite. The hardness of steel of a given
carbon content depends upon the structure that replaces the pearlite when full annealing is not carried out.
The absence of full annealing indicates a more rapid rate of cooling. The rate of
cooling is the factor that determines the hardness. A controlled cooling rate is called
quenching. A mild quench is obtained by cooling in still air, which, as we have seen, is
obtained by the normalizing process. The two most widely used media for quenching
are water and oil. The oil quench is quite slow but prevents quenching cracks caused by
rapid expansion of the object being treated. Quenching in water is used for carbon steels
and for medium-carbon, low-alloy steels.
The effectiveness of quenching depends upon the fact that when austenite is cooled
it does not transform into pearlite instantaneously but requires time to initiate and complete the process. Since the transformation ceases at about 800°F, it can be prevented
by rapidly cooling the material to a lower temperature. When the material is cooled
rapidly to 400°F or less, the austenite is transformed into a structure called martensite.
Martensite is a supersaturated solid solution of carbon in ferrite and is the hardest and
strongest form of steel.
If steel is rapidly cooled to a temperature between 400 and 800°F and held there
for a sufcient length of time, the austenite is transformed into a material that is generally called bainite. Bainite is a structure intermediate between pearlite and martensite.
Although there are several structures that can be identied between the temperatures
given, depending upon the temperature used, they are collectively known as bainite. By
the choice of this transformation temperature, almost any variation of structure may be
obtained. These range all the way from coarse pearlite to ne martensite.
52
46
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Tempering
When a steel specimen has been fully hardened, it is very hard and brittle and has high
residual stresses. The steel is unstable and tends to contract on aging. This tendency
is increased when the specimen is subjected to externally applied loads, because the
resultant stresses contribute still more to the instability. These internal stresses can
be relieved by a modest heating process called stress relieving, or a combination of
stress relieving and softening called tempering or drawing. After the specimen has been
fully hardened by being quenched from above the critical temperature, it is reheated to
some temperature below the critical temperature for a certain period of time and then
allowed to cool in still air. The temperature to which it is reheated depends upon the
composition and the degree of hardness or toughness desired.8 This reheating operation
releases the carbon held in the martensite, forming carbide crystals. The structure
obtained is called tempered martensite. It is now essentially a superne dispersion of
iron carbide(s) in ne-grained ferrite.
The effect of heat-treating operations upon the various mechanical properties of a
low alloy steel is shown graphically in Fig. 213.
Figure 213
300
Tensile strength
600
200
150
Brinell hardness
Tensile and yield strength, kpsi
250
Yield strength
80
500
Brinell
60
400
40
Reduction area
100
50
300
20
Elongation
200
400
600
800
1000
1200
Percent elongation and
reduction in area
The effect of thermalmechanical history on the
mechanical properties of AISI
4340 steel. (Prepared by the
International Nickel Company.)
0
1400
Tempering temperature, °F
Condition
Yield
strength,
kpsi
Reduction
in area,
%
Elongation
in 2 in,
%
Brinell
hardness,
Bhn
Normalized
200
147
20
10
410
As rolled
190
144
18
9
380
Annealed
8
Tensile
strength,
kpsi
120
99
43
18
228
For the quantitative aspects of tempering in plain carbon and low-alloy steels, see Charles R. Mischke,
The Strength of Cold-Worked and Heat-Treated Steels, Chap. 33 in Joseph E. Shigley, Charles R.
Mischke, and Thomas H. Brown, Jr. (eds.), Standard Handbook of Machine Design, 3rd ed., McGraw-Hill,
New York, 2004.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
53
© The McGrawHill
Companies, 2008
Materials
47
Case Hardening
The purpose of case hardening is to produce a hard outer surface on a specimen of lowcarbon steel while at the same time retaining the ductility and toughness in the core.
This is done by increasing the carbon content at the surface. Either solid, liquid, or
gaseous carburizing materials may be used. The process consists of introducing the part
to be carburized into the carburizing material for a stated time and at a stated temperature, depending upon the depth of case desired and the composition of the part. The part
may then be quenched directly from the carburization temperature and tempered, or in
some cases it must undergo a double heat treatment in order to ensure that both the core
and the case are in proper condition. Some of the more useful case-hardening processes
are pack carburizing, gas carburizing, nitriding, cyaniding, induction hardening, and
ame hardening. In the last two cases carbon is not added to the steel in question, generally a medium carbon steel, for example SAE/AISI 1144.
Quantitative Estimation of Properties of Heat-Treated Steels
Courses in metallurgy (or material science) for mechanical engineers usually present the
addition method of Crafts and Lamont for the prediction of heat-treated properties from the
Jominy test for plain carbon steels.9 If this has not been in your prerequisite experience,
then refer to the Standard Handbook of Machine Design, where the addition method is covered with examples.10 If this book is a textbook for a machine elements course, it is a good
class project (many hands make light work) to study the method and report to the class.
For low-alloy steels, the multiplication method of Grossman11 and Field12 is
explained in the Standard Handbook of Machine Design (Secs. 29.6 and 33.6).
Modern Steels and Their Properties Handbook explains how to predict the Jominy
curve by the method of Grossman and Field from a ladle analysis and grain size.13
Bethlehem Steel has developed a circular plastic slide rule that is convenient to the purpose.
215
Alloy Steels
Although a plain carbon steel is an alloy of iron and carbon with small amounts of
manganese, silicon, sulfur, and phosphorus, the term alloy steel is applied when one or
more elements other than carbon are introduced in sufcient quantities to modify its
properties substantially. The alloy steels not only possess more desirable physical
properties but also permit a greater latitude in the heat-treating process.
Chromium
The addition of chromium results in the formation of various carbides of chromium that
are very hard, yet the resulting steel is more ductile than a steel of the same hardness produced by a simple increase in carbon content. Chromium also renes the grain structure
so that these two combined effects result in both increased toughness and increased hardness. The addition of chromium increases the critical range of temperatures and moves
the eutectoid point to the left. Chromium is thus a very useful alloying element.
9
W. Crafts and J. L. Lamont, Hardenability and Steel Selection, Pitman and Sons, London, 1949.
10
Charles R. Mischke, Chap. 33 in Joseph E. Shigley, Charles R. Mischke, and Thomas H. Brown, Jr. (eds.),
Standard Handbook of Machine Design, 3rd ed., McGraw-Hill, New York, 2004, p. 33.9.
11
M. A. Grossman, AIME, February 1942.
12
J. Field, Metals Progress, March 1943.
13
Modern Steels and Their Properties, 7th ed., Handbook 2757, Bethlehem Steel, 1972, pp. 4650.
54
48
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Nickel
The addition of nickel to steel also causes the eutectoid point to move to the left and
increases the critical range of temperatures. Nickel is soluble in ferrite and does not
form carbides or oxides. This increases the strength without decreasing the ductility.
Case hardening of nickel steels results in a better core than can be obtained with plain
carbon steels. Chromium is frequently used in combination with nickel to obtain the
toughness and ductility provided by the nickel and the wear resistance and hardness
contributed by the chromium.
Manganese
Manganese is added to all steels as a deoxidizing and desulfurizing agent, but if the sulfur content is low and the manganese content is over 1 percent, the steel is classied as a
manganese alloy. Manganese dissolves in the ferrite and also forms carbides. It causes
the eutectoid point to move to the left and lowers the critical range of temperatures. It
increases the time required for transformation so that oil quenching becomes practicable.
Silicon
Silicon is added to all steels as a deoxidizing agent. When added to very-low-carbon
steels, it produces a brittle material with a low hysteresis loss and a high magnetic
permeability. The principal use of silicon is with other alloying elements, such as
manganese, chromium, and vanadium, to stabilize the carbides.
Molybdenum
While molybdenum is used alone in a few steels, it nds its greatest use when combined
with other alloying elements, such as nickel, chromium, or both. Molybdenum forms
carbides and also dissolves in ferrite to some extent, so that it adds both hardness and
toughness. Molybdenum increases the critical range of temperatures and substantially
lowers the transformation point. Because of this lowering of the transformation point,
molybdenum is most effective in producing desirable oil-hardening and air-hardening
properties. Except for carbon, it has the greatest hardening effect, and because it also
contributes to a ne grain size, this results in the retention of a great deal of toughness.
Vanadium
Vanadium has a very strong tendency to form carbides; hence it is used only in small
amounts. It is a strong deoxidizing agent and promotes a ne grain size. Since some vanadium is dissolved in the ferrite, it also toughens the steel. Vanadium gives a wide hardening range to steel, and the alloy can be hardened from a higher temperature. It is very
difcult to soften vanadium steel by tempering; hence, it is widely used in tool steels.
Tungsten
Tungsten is widely used in tool steels because the tool will maintain its hardness even
at red heat. Tungsten produces a ne, dense structure and adds both toughness and hardness. Its effect is similar to that of molybdenum, except that it must be added in greater
quantities.
216
Corrosion-Resistant Steels
Iron-base alloys containing at least 12 percent chromium are called stainless steels.
The most important characteristic of these steels is their resistance to many, but not all,
corrosive conditions. The four types available are the ferritic chromium steels, the
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
55
© The McGrawHill
Companies, 2008
Materials
49
austenitic chromium-nickel steels, and the martensitic and precipitation-hardenable
stainless steels.
The ferritic chromium steels have a chromium content ranging from 12 to 27 percent. Their corrosion resistance is a function of the chromium content, so that alloys
containing less than 12 percent still exhibit some corrosion resistance, although they
may rust. The quench-hardenability of these steels is a function of both the chromium
and the carbon content. The very high carbon steels have good quench hardenability up
to about 18 percent chromium, while in the lower carbon ranges it ceases at about
13 percent. If a little nickel is added, these steels retain some degree of hardenability up
to 20 percent chromium. If the chromium content exceeds 18 percent, they become difcult to weld, and at the very high chromium levels the hardness becomes so great that
very careful attention must be paid to the service conditions. Since chromium is expensive, the designer will choose the lowest chromium content consistent with the corrosive conditions.
The chromium-nickel stainless steels retain the austenitic structure at room temperature; hence, they are not amenable to heat treatment. The strength of these steels
can be greatly improved by cold working. They are not magnetic unless cold-worked.
Their work hardenability properties also cause them to be difcult to machine. All
the chromium-nickel steels may be welded. They have greater corrosion-resistant properties than the plain chromium steels. When more chromium is added for greater corrosion resistance, more nickel must also be added if the austenitic properties are to be
retained.
217
Casting Materials
Gray Cast Iron
Of all the cast materials, gray cast iron is the most widely used. This is because it has
a very low cost, is easily cast in large quantities, and is easy to machine. The principal
objections to the use of gray cast iron are that it is brittle and that it is weak in tension.
In addition to a high carbon content (over 1.7 percent and usually greater than 2 percent),
cast iron also has a high silicon content, with low percentages of sulfur, manganese,
and phosphorus. The resultant alloy is composed of pearlite, ferrite, and graphite, and
under certain conditions the pearlite may decompose into graphite and ferrite. The
resulting product then contains all ferrite and graphite. The graphite, in the form of
thin akes distributed evenly throughout the structure, darkens it; hence, the name gray
cast iron.
Gray cast iron is not readily welded, because it may crack, but this tendency may
be reduced if the part is carefully preheated. Although the castings are generally used in
the as-cast condition, a mild anneal reduces cooling stresses and improves the machinability. The tensile strength of gray cast iron varies from 100 to 400 MPa (15 to 60 kpsi),
and the compressive strengths are 3 to 4 times the tensile strengths. The modulus of
elasticity varies widely, with values extending all the way from 75 to 150 GPa (11 to
22 Mpsi).
Ductile and Nodular Cast Iron
Because of the lengthy heat treatment required to produce malleable cast iron, engineers
have long desired a cast iron that would combine the ductile properties of malleable
iron with the ease of casting and machining of gray iron and at the same time would
possess these properties in the as-cast conditions. A process for producing such a material
using magnesium-containing material seems to fulll these requirements.
56
50
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Ductile cast iron, or nodular cast iron, as it is sometimes called, is essentially the
same as malleable cast iron, because both contain graphite in the form of spheroids.
However, ductile cast iron in the as-cast condition exhibits properties very close to
those of malleable iron, and if a simple 1-h anneal is given and is followed by a slow
cool, it exhibits even more ductility than the malleable product. Ductile iron is made by
adding MgFeSi to the melt; since magnesium boils at this temperature, it is necessary
to alloy it with other elements before it is introduced.
Ductile iron has a high modulus of elasticity (172 GPa or 25 Mpsi) as compared
with gray cast iron, and it is elastic in the sense that a portion of the stress-strain
curve is a straight line. Gray cast iron, on the other hand, does not obey Hookes law,
because the modulus of elasticity steadily decreases with increase in stress. Like
gray cast iron, however, nodular iron has a compressive strength that is higher than
the tensile strength, although the difference is not as great. In 40 years it has become
extensively used.
White Cast Iron
If all the carbon in cast iron is in the form of cementite and pearlite, with no graphite
present, the resulting structure is white and is known as white cast iron. This may be
produced in two ways. The composition may be adjusted by keeping the carbon and
silicon content low, or the gray-cast-iron composition may be cast against chills in order
to promote rapid cooling. By either method, a casting with large amounts of cementite
is produced, and as a result the product is very brittle and hard to machine but also very
resistant to wear. A chill is usually used in the production of gray-iron castings in order
to provide a very hard surface within a particular area of the casting, while at the same
time retaining the more desirable gray structure within the remaining portion. This produces a relatively tough casting with a wear-resistant area.
Malleable Cast Iron
If white cast iron within a certain composition range is annealed, a product called
malleable cast iron is formed. The annealing process frees the carbon so that it is present as graphite, just as in gray cast iron but in a different form. In gray cast iron the
graphite is present in a thin ake form, while in malleable cast iron it has a nodular
form and is known as temper carbon. A good grade of malleable cast iron may have
a tensile strength of over 350 MPa (50 kpsi), with an elongation of as much as 18 percent. The percentage elongation of a gray cast iron, on the other hand, is seldom over
1 percent. Because of the time required for annealing (up to 6 days for large and
heavy castings), malleable iron is necessarily somewhat more expensive than gray
cast iron.
Alloy Cast Irons
Nickel, chromium, and molybdenum are the most common alloying elements used in
cast iron. Nickel is a general-purpose alloying element, usually added in amounts up to
5 percent. Nickel increases the strength and density, improves the wearing qualities, and
raises the machinability. If the nickel content is raised to 10 to 18 percent, an austenitic
structure with valuable heat- and corrosion-resistant properties results. Chromium
increases the hardness and wear resistance and, when used with a chill, increases the
tendency to form white iron. When chromium and nickel are both added, the hardness
and strength are improved without a reduction in the machinability rating. Molybdenum
added in quantities up to 1.25 percent increases the stiffness, hardness, tensile strength,
and impact resistance. It is a widely used alloying element.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
57
© The McGrawHill
Companies, 2008
Materials
51
Cast Steels
The advantage of the casting process is that parts having complex shapes can be manufactured at costs less than fabrication by other means, such as welding. Thus the
choice of steel castings is logical when the part is complex and when it must also have
a high strength. The higher melting temperatures for steels do aggravate the casting
problems and require closer attention to such details as core design, section thicknesses,
fillets, and the progress of cooling. The same alloying elements used for the wrought
steels can be used for cast steels to improve the strength and other mechanical properties. Cast-steel parts can also be heat-treated to alter the mechanical properties, and,
unlike the cast irons, they can be welded.
218
Nonferrous Metals
Aluminum
The outstanding characteristics of aluminum and its alloys are their strength-weight
ratio, their resistance to corrosion, and their high thermal and electrical conductivity.
The density of aluminum is about 2770 kg/m3 (0.10 lbf/in3), compared with 7750 kg/m3
(0.28 lbf/in3) for steel. Pure aluminum has a tensile strength of about 90 MPa (13 kpsi),
but this can be improved considerably by cold working and also by alloying with other
materials. The modulus of elasticity of aluminum, as well as of its alloys, is 71.7 GPa
(10.4 Mpsi), which means that it has about one-third the stiffness of steel.
Considering the cost and strength of aluminum and its alloys, they are among the
most versatile materials from the standpoint of fabrication. Aluminum can be processed
by sand casting, die casting, hot or cold working, or extruding. Its alloys can be machined,
press-worked, soldered, brazed, or welded. Pure aluminum melts at 660°C (1215°F),
which makes it very desirable for the production of either permanent or sand-mold
castings. It is commercially available in the form of plate, bar, sheet, foil, rod, and tube
and in structural and extruded shapes. Certain precautions must be taken in joining
aluminum by soldering, brazing, or welding; these joining methods are not recommended
for all alloys.
The corrosion resistance of the aluminum alloys depends upon the formation of a
thin oxide coating. This lm forms spontaneously because aluminum is inherently very
reactive. Constant erosion or abrasion removes this lm and allows corrosion to take
place. An extra-heavy oxide lm may be produced by the process called anodizing. In
this process the specimen is made to become the anode in an electrolyte, which may be
chromic acid, oxalic acid, or sulfuric acid. It is possible in this process to control the
color of the resulting lm very accurately.
The most useful alloying elements for aluminum are copper, silicon, manganese,
magnesium, and zinc. Aluminum alloys are classied as casting alloys or wrought
alloys. The casting alloys have greater percentages of alloying elements to facilitate
casting, but this makes cold working difcult. Many of the casting alloys, and some of
the wrought alloys, cannot be hardened by heat treatment. The alloys that are heattreatable use an alloying element that dissolves in the aluminum. The heat treatment
consists of heating the specimen to a temperature that permits the alloying element to
pass into solution, then quenching so rapidly that the alloying element is not precipitated. The aging process may be accelerated by heating slightly, which results in even
greater hardness and strength. One of the better-known heat-treatable alloys is duraluminum, or 2017 (4 percent Cu, 0.5 percent Mg, 0.5 percent Mn). This alloy hardens in
4 days at room temperature. Because of this rapid aging, the alloy must be stored under
58
52
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
refrigeration after quenching and before forming, or it must be formed immediately
after quenching. Other alloys (such as 5053) have been developed that age-harden much
more slowly, so that only mild refrigeration is required before forming. After forming,
they are articially aged in a furnace and possess approximately the same strength and
hardness as the 2024 alloys. Those alloys of aluminum that cannot be heat-treated can
be hardened only by cold working. Both work hardening and the hardening produced
by heat treatment may be removed by an annealing process.
Magnesium
The density of magnesium is about 1800 kg/m3 (0.065 lb/in3), which is two-thirds that
of aluminum and one-fourth that of steel. Since it is the lightest of all commercial metals, its greatest use is in the aircraft and automotive industries, but other uses are now
being found for it. Although the magnesium alloys do not have great strength, because
of their light weight the strength-weight ratio compares favorably with the stronger
aluminum and steel alloys. Even so, magnesium alloys nd their greatest use in applications where strength is not an important consideration. Magnesium will not withstand
elevated temperatures; the yield point is denitely reduced when the temperature is
raised to that of boiling water.
Magnesium and its alloys have a modulus of elasticity of 45 GPa (6.5 Mpsi) in tension and in compression, although some alloys are not as strong in compression as in
tension. Curiously enough, cold working reduces the modulus of elasticity. A range of
cast magnesium alloys are also available.
Titanium
Titanium and its alloys are similar in strength to moderate-strength steel but weigh half
as much as steel. The material exhibits very good resistence to corrosion, has low thermal conductivity, is nonmagnetic, and has high-temperature strength. Its modulus of
elasticity is between those of steel and aluminum at 16.5 Mpsi (114 GPa). Because of
its many advantages over steel and aluminum, applications include: aerospace and military aircraft structures and components, marine hardware, chemical tanks and processing equipment, uid handling systems, and human internal replacement devices. The
disadvantages of titanium are its high cost compared to steel and aluminum and the difculty of machining it.
Copper-Base Alloys
When copper is alloyed with zinc, it is usually called brass. If it is alloyed with another
element, it is often called bronze. Sometimes the other element is specied too, as, for example, tin bronze or phosphor bronze. There are hundreds of variations in each category.
Brass with 5 to 15 Percent Zinc
The low-zinc brasses are easy to cold work, especially those with the higher zinc content. They are ductile but often hard to machine. The corrosion resistance is good. Alloys
included in this group are gilding brass (5 percent Zn), commercial bronze (10 percent Zn),
and red brass (15 percent Zn). Gilding brass is used mostly for jewelry and articles to
be gold-plated; it has the same ductility as copper but greater strength, accompanied by
poor machining characteristics. Commercial bronze is used for jewelry and for forgings
and stampings, because of its ductility. Its machining properties are poor, but it has
excellent cold-working properties. Red brass has good corrosion resistance as well as
high-temperature strength. Because of this it is used a great deal in the form of tubing or
piping to carry hot water in such applications as radiators or condensers.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
2. Materials
59
© The McGrawHill
Companies, 2008
Materials
53
Brass with 20 to 36 Percent Zinc
Included in the intermediate-zinc group are low brass (20 percent Zn), cartridge brass
(30 percent Zn), and yellow brass (35 percent Zn). Since zinc is cheaper than copper,
these alloys cost less than those with more copper and less zinc. They also have better
machinability and slightly greater strength; this is offset, however, by poor corrosion
resistance and the possibility of cracking at points of residual stresses. Low brass is very
similar to red brass and is used for articles requiring deep-drawing operations. Of the
copper-zinc alloys, cartridge brass has the best combination of ductility and strength.
Cartridge cases were originally manufactured entirely by cold working; the process
consisted of a series of deep draws, each draw being followed by an anneal to place the
material in condition for the next draw, hence the name cartridge brass. Although the
hot-working ability of yellow brass is poor, it can be used in practically any other fabricating process and is therefore employed in a large variety of products.
When small amounts of lead are added to the brasses, their machinability is greatly
improved and there is some improvement in their abilities to be hot-worked. The
addition of lead impairs both the cold-working and welding properties. In this group are
low-leaded brass (32 1 percent Zn, 1 percent Pb), high-leaded brass (34 percent Zn,
2
2
2 percent Pb), and free-cutting brass (35 1 percent Zn, 3 percent Pb). The low-leaded
2
brass is not only easy to machine but has good cold-working properties. It is used for
various screw-machine parts. High-leaded brass, sometimes called engravers brass, is
used for instrument, lock, and watch parts. Free-cutting brass is also used for screwmachine parts and has good corrosion resistance with excellent mechanical properties.
Admiralty metal (28 percent Zn) contains 1 percent tin, which imparts excellent
corrosion resistance, especially to saltwater. It has good strength and ductility but only
fair machining and working characteristics. Because of its corrosion resistance it is used
in power-plant and chemical equipment. Aluminum brass (22 percent Zn) contains
2 percent aluminum and is used for the same purposes as admiralty metal, because it
has nearly the same properties and characteristics. In the form of tubing or piping, it is
favored over admiralty metal, because it has better resistance to erosion caused by highvelocity water.
Brass with 36 to 40 Percent Zinc
Brasses with more than 38 percent zinc are less ductile than cartridge brass and cannot
be cold-worked as severely. They are frequently hot-worked and extruded. Muntz metal
(40 percent Zn) is low in cost and mildly corrosion-resistant. Naval brass has the same
composition as Muntz metal except for the addition of 0.75 percent tin, which contributes to the corrosion resistance.
Bronze
Silicon bronze, containing 3 percent silicon and 1 percent manganese in addition to the
copper, has mechanical properties equal to those of mild steel, as well as good corrosion resistance. It can be hot- or cold-worked, machined, or welded. It is useful wherever corrosion resistance combined with strength is required.
Phosphor bronze, made with up to 11 percent tin and containing small amounts of
phosphorus, is especially resistant to fatigue and corrosion. It has a high tensile strength
and a high capacity to absorb energy, and it is also resistant to wear. These properties
make it very useful as a spring material.
Aluminum bronze is a heat-treatable alloy containing up to 12 percent aluminum. This
alloy has strength and corrosion-resistance properties that are better than those of brass, and
in addition, its properties may be varied over a wide range by cold working, heat treating,
60
54
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
or changing the composition. When iron is added in amounts up to 4 percent, the alloy has
a high endurance limit, a high shock resistance, and excellent wear resistance.
Beryllium bronze is another heat-treatable alloy, containing about 2 percent beryllium. This alloy is very corrosion resistant and has high strength, hardness, and resistance to wear. Although it is expensive, it is used for springs and other parts subjected
to fatigue loading where corrosion resistance is required.
With slight modication most copper-based alloys are available in cast form.
219
Plastics
The term thermoplastics is used to mean any plastic that ows or is moldable when heat
is applied to it; the term is sometimes applied to plastics moldable under pressure. Such
plastics can be remolded when heated.
A thermoset is a plastic for which the polymerization process is nished in a hot
molding press where the plastic is liqueed under pressure. Thermoset plastics cannot
be remolded.
Table 22 lists some of the most widely used thermoplastics, together with some
of their characteristics and the range of their properties. Table 23, listing some of the
Table 22
The Thermoplastics
Source: These data have been obtained from the Machine Design Materials Reference Issue,
published by Penton/IPC, Cleveland. These reference issues are published about every 2 years and constitute an
excellent source of data on a great variety of materials.
Name
ABS group
Su,
kpsi
28
E,
Mpsi
Hardness
Rockwell
0.100.37
60110R
350
4060
Acetal group
810
0.410.52
8094M
Acrylic
510
0.200.47
92110M
Fluoroplastic
group
0.507
···
5080D
Elongation Dimensional Heat
Chemical
%
Stability
Resistance Resistance Processing
375
100300
Nylon
814
0.180.45 112120R
10200
Phenylene
oxide
718
0.350.92 115R, 106L
560
Good
*
Fair
EMST
Excellent
Good
High
M
High
*
Fair
EMS
High
Excellent
Excellent
MPR
Poor
Poor
Good
CEM
Excellent
Good
Fair
EFM
Polycarbonate
816
0.340.86
6291M
10125
Excellent
Excellent
Fair
EMS
Polyester
818
0.281.6
6590M
1300
Excellent
Poor
Excellent
CLMR
Polyimide
650
Excellent
Excellent
Excellent
CLMP
Good
Excellent
Excellent
M
Poor
Poor
EM
Excellent
Excellent
EFM
Poor
Poor
EFM
Polyphenylene
sulde
···
88120M
Very low
1419
0.11
122R
1.0
Polystyrene
group
1.512
0.140.60
1090M
0.560
Polysulfone
10
0.36
120R
50100
Excellent
6585D
40450
··
Polyvinyl
chloride
1.57.5 0.350.60
···
*Heat-resistant grades available.
With exceptions.
C Coatings L Laminates R Resins E Extrusions M Moldings S Sheet F Foams P Press and sinter methods T Tubing
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
61
© The McGrawHill
Companies, 2008
2. Materials
Materials
55
Table 23
The Thermosets Source: These data have been obtained from the Machine Design Materials Reference Issue,
published by Penton/IPC, Cleveland. These reference issues are published about every 2 years and constitute
an excellent source of data on a great variety of materials.
Name
Su,
kpsi
E,
Mpsi
Hardness
Rockwell
0.050.30
99M*
Elongation
%
Dimensional
Stability
Heat
Resistance
Chemical
Resistance
Excellent
Good
Fair
M
Excellent
Excellent
Excellent
CM
Good
Excellent*
Excellent*
LR
Excellent
Excellent
CMR
Processing
Alkyd
39
Allylic
410
Amino
group
58
0.130.24
Epoxy
520
0.030.30*
80120M
110
Excellent
Phenolics
59
0.100.25
7095E
Excellent
Good
EMR
56
···
8090M
···
Excellent
Silicones
···
Excellent
Excellent
CL MR
105120M
···
110120M
···
···
0.300.90
···
*With exceptions.
C Coatings L Laminates R Resins E Extrusions M Moldings S Sheet F Foams P Press and sinter methods T Tubing
thermosets, is similar. These tables are presented for information only and should not
be used to make a nal design decision. The range of properties and characteristics that
can be obtained with plastics is very great. The inuence of many factors, such as cost,
moldability, coefcient of friction, weathering, impact strength, and the effect of llers
and reinforcements, must be considered. Manufacturers catalogs will be found quite
helpful in making possible selections.
220
Composite Materials14
Composite materials are formed from two or more dissimilar materials, each of which
contributes to the nal properties. Unlike metallic alloys, the materials in a composite
remain distinct from each other at the macroscopic level.
Most engineering composites consist of two materials: a reinforcement called a
ller and a matrix. The ller provides stiffness and strength; the matrix holds the material together and serves to transfer load among the discontinuous reinforcements. The
most common reinforcements, illustrated in Fig. 214, are continuous bers, either
straight or woven, short chopped bers, and particulates. The most common matrices
are various plastic resins although other materials including metals are used.
Metals and other traditional engineering materials are uniform, or isotropic, in
nature. This means that material properties, such as strength, stiffness, and thermal conductivity, are independent of both position within the material and the choice of coordinate system. The discontinuous nature of composite reinforcements, though, means
that material properties can vary with both position and direction. For example, an
14
For references see I. M. Daniel and O. Ishai, Engineering Mechanics of Composite Materials, Oxford
University Press, 1994, and ASM Engineered Materials Handbook: Composites, ASM International,
Materials Park, OH, 1988.
62
56
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Figure 214
Composites categorized by
type of reinforcement.
Particulate
composite
Randomly oriented
short fiber composite
Unidirectional continuous
fiber composite
Woven fabric
composite
epoxy resin reinforced with continuous graphite bers will have very high strength and
stiffness in the direction of the bers, but very low properties normal or transverse to
the bers. For this reason, structures of composite materials are normally constructed
of multiple plies (laminates) where each ply is oriented to achieve optimal structural
stiffness and strength performance.
High strength-to-weight ratios, up to 5 times greater than those of high-strength
steels, can be achieved. High stiffness-to-weight ratios can also be obtained, as much as
8 times greater than those of structural metals. For this reason, composite materials are
becoming very popular in automotive, aircraft, and spacecraft applications where
weight is a premium.
The directionality of properties of composite materials increases the complexity of
structural analyses. Isotropic materials are fully dened by two engineering constants:
Youngs modulus E and Poissons ratio ν . A single ply of a composite material, however, requires four constants, dened with respect to the ply coordinate system. The
constants are two Youngs moduli (the longitudinal modulus in the direction of the
bers, E 1 , and the transverse modulus normal to the bers, E 2 ), one Poissons ratio
(ν12 , called the major Poissons ratio), and one shear modulus (G 12 ). A fth constant,
the minor Poissons ratio, ν21 , is determined through the reciprocity relation,
ν21 / E 2 = ν12 / E 1 . Combining this with multiple plies oriented at different angles makes
structural analysis of complex structures unapproachable by manual techniques. For
this reason, computer software is available to calculate the properties of a laminated
composite construction.15
221
Materials Selection
As stated earlier, the selection of a material for a machine part or structural member is
one of the most important decisions the designer is called on to make. Up to this point
in this chapter we have discussed many important material physical properties, various
characteristics of typical engineering materials, and various material production
processes. The actual selection of a material for a particular design application can be
an easy one, say, based on previous applications (1020 steel is always a good candidate because of its many positive attributes), or the selection process can be as
involved and daunting as any design problem with the evaluation of the many material
physical, economical, and processing parameters. There are systematic and optimizing
approaches to material selection. Here, for illustration, we will only look at how
to approach some material properties. One basic technique is to list all the important
material properties associated with the design, e.g., strength, stiffness, and cost. This
can be prioritized by using a weighting measure depending on what properties are more
15
About Composite Materials Software listing, http://composite.about.com/cs/software/index.htm.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
63
© The McGrawHill
Companies, 2008
2. Materials
Materials
57
important than others. Next, for each property, list all available materials and rank
them in order beginning with the best material; e.g., for strength, high-strength steel
such as 4340 steel should be near the top of the list. For completeness of available
materials, this might require a large source of material data. Once the lists are formed,
select a manageable amount of materials from the top of each list. From each reduced
list select the materials that are contained within every list for further review. The
materials in the reduced lists can be graded within the list and then weighted according to the importance of each property.
M. F. Ashby has developed a powerful systematic method using materials selection charts.16 This method has also been implemented in a software package called
CES Edupack.17 The charts display data of various properties for the families and
classes of materials listed in Table 24. For example, considering material stiffness
properties, a simple bar chart plotting Youngs modulus E on the y axis is shown
in Fig. 215. Each vertical line represents the range of values of E for a particular
material. Only some of the materials are labeled. Now, more material information
can be displayed if the x axis represents another material property, say density.
Table 24
Family
Short Name
Metals
(the metals and alloys of
engineering)
Material Families and
Classes
Classes
Aluminum alloys
Al alloys
Copper alloys
Cu alloys
Lead alloys
Lead alloys
Magnesium alloys
Mg alloys
Nickel alloys
Ni alloys
Carbon steels
Steels
Stainless steels
Stainless steels
Tin alloys
Tin alloys
Titanium alloys
W alloys
Lead alloys
Pb alloys
Zinc alloys
Ceramics
Technical ceramics (ne
ceramics capable of
load-bearing application)
Ti alloys
Tungsten alloys
Zn alloys
Alumina
AI2 O3
Aluminum nitride
AIN
Boron carbide
B4 C
Silicon carbide
SiC
Silicon nitride
Si3 N4
Tungsten carbide
Nontechnical ceramics
(porous ceramics of
construction)
WC
Brick
Brick
Concrete
Concrete
Stone
Stone
(continued)
16
M. F. Ashby, Materials Selection in Mechanical Design, 3rd ed., Elsevier Butterworth-Heinemann,
Oxford, 2005.
17
Produced by Granta Design Limited. See www.grantadesign.com.
64
58
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Table 24 (continued)
Family
Classes
Short Name
Glasses
Soda-lime glass
Soda-lime glass
Borosilicate glass
Borosilicate glass
Silica glass
Silica glass
Glass ceramic
Glass ceramic
Polymers
(the thermoplastics and
thermosets of engineering)
Acrylonitrile butadiene styrene
ABS
Cellulose polymers
CA
lonomers
lonomers
Epoxies
Epoxy
Phenolics
Phenolics
Polyamides (nylons)
PA
Polycarbonate
PC
Polyesters
Polyester
Polyetheretherkeytone
PEEK
Polyethylene
PE
Polyethylene terephalate
PET or PETE
Polymethylmethacrylate
PMMA
Polyoxymethylene(Acetal)
POM
Polypropylene
PP
Polystyrene
PTFE
Polyvinylchloride
Elastomers
(engineering rubbers,
natural and synthetic)
PS
Polytetrauorethylene
PVC
Butyl rubber
Butyl rubber
EVA
EVA
lsoprene
lsoprene
Natural rubber
Natural rubber
Polychloroprene (Neoprene)
Neoprene
Polyurethane
PU
Silicon elastomers
Silicones
Al-SiC
Flexible polymer foams
Flexible foams
Rigid foams
Cork
Cork
Bamboo
Wood
Natural materials
GFRP
Bamboo
Foams
CFRP
Glass-ber reinforced polymers
Rigid polymer foams
Composites
Carbon-ber reinforced polymers
SiC reinforced aluminum
Hybrids
Wood
From M. F. Ashby, Materials Selection in Mechanical Design, 3rd ed., Elsevier Butterworth-Heinemann, Oxford, 2005. Table 41,
pp. 4950.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
65
© The McGrawHill
Companies, 2008
2. Materials
Materials
59
Figure 215
Youngs modulus E for various materials. (Figure courtesy of Prof. Mike Ashby, Granta Design, Cambridge, U.K.)
1000
Tungsten carbides
Nickel alloys
100
Cast iron, gray
Titanium alloys
Low-alloy steel
GFRP, epoxy matrix (isotropic)
Copper alloys
Soda-lime glass
10
Polyester
Young's modulus, GPa
Wood, typical along grain
Wood, typical across grain
1
Acrylonitrile butadiene styrene (ABS)
Rigid polymer foam (MD)
0.1
Cork
0.01
Polyurethane
1e-3
Butyl rubber
Flexible polymer foam (VLD)
1e-4
Figure 216, called a bubble chart, represents Youngs modulus E plotted against density
ρ . The line ranges for each material property plotted two-dimensionally now form ellipses,
or bubbles. This plot is more useful than the two separate bar charts of each property. Now,
we also see how stiffness/weight for various materials relate. Figure 216 also shows
groups of bubbles outlined according to the material families of Table 24. In addition, dotted lines in the lower right corner of the chart indicate ratios of E β /ρ , which assist in material selection for minimum mass design. Lines drawn parallel to these lines represent
different values for E β /ρ . For example, several parallel dotted lines are shown in Fig. 216
that represent different values of E /ρ ( β = 1). Since ( E /ρ ) 1/2 represents the speed of
sound in a material, each dotted line, E /ρ , represents a different speed as indicated.
To see how β ts into the mix, consider the following. The performance metric P
of a structural element depends on (1) the functional requirements, (2) the geometry,
and (3) the material properties of the structure. That is,
P
functional
geometric
material
[(requirements F), (parameters G), (properties M)]
or, symbolically,
P = f ( F, G, M )
(220)
66
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
60
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Figure 216
Youngs modulus E versus density ρ for various materials. (Figure courtesy of Prof. Mike Ashby, Granta Design, Cambridge, U.K.)
1000
Composites
100
Wood
n grain
Young's modulus E, GPa
Natural
materials
Longitudinal
wave speed
Metals
Polyester
o
Concrete
Zinc alloys
Lead alloys
y
PEEK
PET
T
Epoxies
PC
E1/3
PTFE
Polymers
Leather
1
WC
W alloys
Cu alloys
y
Wood PS
n
grain PP
PE
104 m/s
Rigid polymer
foams
10
B4C
Al
A alloys
C
CFRP
Glass
s
Mg alloys
l
GFRP
PMMA
PA
10
1
Al2O3
Steels Ni alloys
Si3N4 SiC
Ti alloys
Technical
ceramics
E1/2
Foams
10
2
E
EVA
103 m/s
Silicone elastomers
Polyurethane
Cork
Guidelines for
minimum mass
design
Isoprene
10
Neoprene
3
Flexible polymer
foams
10
4
102 m/s
0.01
Elastomers
Butyl
rubber
MFA C4
0.1
1
10
Density , Mg/m3
If the function is separable, which it often is, we can write Eq. (220) as
P = f 1 ( F ) · f 2 (G ) · f 3 ( M )
(221)
For optimum design, we desire to maximize or minimize P. With regards to material
properties alone, this is done by maximizing or minimizing f 3 ( M ), called the material
efciency coefcient.
For illustration, say we want to design a light, stiff, end-loaded cantilever beam with
a circular cross section. For this we will use the mass m of the beam for the performance
metric to minimize. The stiffness of the beam is related to its material and geometry. The
stiffness of a beam is given by k = F /δ , where F and δ are the end load and deection,
respectively (see Chap. 4). The end deection of an end-loaded cantilever beam is given
in Table A9, beam 1, as δ = ymax = ( Fl 3 )/(3 E I ) , where E is Youngs modulus, I the
second moment of the area, and l the length of the beam. Thus, the stiffness is given by
k=
F
3E I
=3
δ
l
(222)
From Table A-18, the second moment of the area of a circular cross section is
I=
π D4
A2
=
64
4π
(223)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
67
© The McGrawHill
Companies, 2008
2. Materials
Materials
61
where D and A are the diameter and area of the cross section, respectively. Substituting
Eq. (223) in (222) and solving for A, we obtain
A=
4π kl 3
3E
1/2
(224)
The mass of the beam is given by
(225)
m = Alρ
Substituting Eq. (224) into (225) and rearranging yields
π 1/2 5/2
ρ
(k )(l )
(226)
3
E 1/2
Equation (226) is of the form of Eq. (221). The term 2 π/3 is simply a constant and
can be associated with any function, say f 1 ( F ). Thus, f 1 ( F ) = 2 π/3(k 1/2 ) is the
functional requirement, stiffness; f 2 (G ) = (l 5/2 ) , the geometric parameter, length; and
the material efciency coefcient
m=2
f3( M ) =
ρ
E 1/2
(227)
is the material property in terms of density and Youngs modulus. To minimize m we
want to minimize f 3 ( M ), or maximize
M=
E 1/2
ρ
(228)
where M is called the material index, and β = 1 . Returning to Fig. 216, draw lines of
2
various values of E 1/2 /ρ as shown in Fig. 217. Lines of increasing M move up and to
the left as shown. Thus, we see that good candidates for a light, stiff, end-loaded cantilever beam with a circular cross section are certain woods, composites, and ceramics.
Other limits/constraints may warrant further investigation. Say, for further illustration, the design requirements indicate that we need a Youngs modulus greater than
50 GPa. Figure 218 shows how this further restricts the search region. This eliminates
woods as a possible material.
Figure 217
Modulusdensity
Search
region
100
Young's modulus E, GPa
A schematic E versus ρ chart
showing a grid of lines for
various values the material
index M = E1/2 /ρ . (From M. F.
Ashby, Materials Selection in
Mechanical Design, 3rd ed.,
Elsevier ButterworthHeinemann, Oxford, 2005.)
1000
Ceramics
3
1
0.3
Metals
Composites
0.1
Increasing values
of index E1/2/
10
E1/2/
(GPa)1/2/(Mg/m)3
Woods
1
Polymers
Foams
0.1
Elastomers
0.01
0.1
MFA 04
1
10
Density, Mg/m3
100
68
62
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Figure 218
1000
Modulusdensity
Ceramics
Metals
Composites
Search
region
100
Young's modulus E, GPa
The search region of Fig. 216
further reduced by restricting
E 50 GPa. (From M. F.
Ashby, Materials Selection in
Mechanical Design, 3rd ed.,
Elsevier Butterworth-Heinemann,
Oxford, 2005.)
Index
E1/2/r 3
10
Woods
E
Modulus
50 GPa
1
Polymers
0.1
Foams
Elastomers
MFA 04
0.01
0.1
1
10
100
Density, Mg/m3
Figure 219
Strength S versus density ρ for various materials. For metals, S is the 0.2 percent offset yield strength. For polymers, S is the 1 percent yield
strength. For ceramics and glasses, S is the compressive crushing strength. For composites, S is the tensile strength. For elastomers, S is the
tear strength. (Figure courtesy of Prof. Mike Ashby, Granta Design, Cambridge, U.K.)
10000
Strengthdensity
1000
Metals and polymers yield strength
Ceramics and glasses MGR
Elastomers tensile tear strength
Composites tensile failure
Strength S, MPa
100
Natural
materials
Ceramics
Si3N4 Ti alloys
y
Metals
Steels
s
SiC Al2O3
N
Ni alloys
A
Al alloys
Tungsten
CFRP
P
alloys
Mg alloys
Polymers and
Tungsten
GFRP
elastomers
carbide
PEEK
Copper
PET
T
PA
alloys
PC
Wood PMMA
to grain
Composites
Rigid polymer
foams
10
Zinc alloys
Lead alloys
Foams
1
Concrete
Butyl
Wood rubber Silicone
elastomers
to grain
Guide lines for
minimum mass
design
Cork
0.1
S
3
S2/3
Flexible polymer
foams
0.01
0.01
0.1
S1/2
MFA D4
1
Density , Mg/m3
10
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
69
© The McGrawHill
Companies, 2008
2. Materials
Materials
63
Certainly, in a given design exercise, there will be other considerations such as
strength, environment, and cost, and other charts may be necessary to investigate. For
example, Fig. 219 represents strength versus density for the material families. Also,
we have not brought in the material process selection part of the picture. If done properly, material selection can result in a good deal of bookkeeping. This is where software
packages such as CES Edupack become very effective.
PROBLEMS
21
Determine the minimum tensile and yield strengths for SAE 1020 cold-drawn steel.
22
Determine the minimum tensile and yield strengths for UNS G10500 hot-rolled steel.
23
For the materials in Probs. 21 and 22, compare the following properties: minimum tensile and
yield strengths, ductility, and stiffness.
24
Assuming you were specifying an AISI 1040 steel for an application where you desired to maximize the yield strength, how would you specify it?
25
Assuming you were specifying an AISI 1040 steel for an application where you desired to maximize the ductility, how would you specify it?
26
Determine the yield strength-to-weight density ratios (called specic strength) in units of inches
for UNS G10350 hot-rolled steel, 2024-T4 aluminum, Ti-6A1-4V titanium alloy, and ASTM
No. 30 gray cast iron.
27
Determine the stiffness-to-weight density ratios (called specic modulus) in units of inches for
UNS G10350 hot-rolled steel, 2024-T4 aluminum, Ti-6A1-4V titanium alloy, and ASTM No. 30
gray cast iron.
28
Poissons ratio ν is a material property and is the ratio of the lateral strain and the longitudinal
strain for a member in tension. For a homogeneous, isotropic material, the modulus of rigidity G
is related to Youngs modulus as
G=
E
2(1 + ν)
Using the tabulated values of G and E, determine Poissons ratio for steel, aluminum, beryllium
copper, and gray cast iron.
29
A specimen of medium-carbon steel having an initial diameter of 0.503 in was tested in tension
using a gauge length of 2 in. The following data were obtained for the elastic and plastic states:
Elastic State
Plastic State
Load P,
lbf
Elongation,
in
Load P,
lbf
Area Ai,
in2
1 000
0.0004
8 800
0.1984
2 000
0.0006
9 200
0.1978
3 000
0.0010
9 100
0.1963
4 000
0.0013
13 200
0.1924
7 000
0.0023
15 200
0.1875
8 400
0.0028
17 000
0.1563
8 800
0.0036
16 400
0.1307
9 200
0.0089
14 800
0.1077
70
64
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
2. Materials
Mechanical Engineering Design
Note that there is some overlap in the data. Plot the engineering or nominal stress-strain diagram
using two scales for the unit strain ǫ, one from zero to about 0.02 in/in and the other from zero
to maximum strain. From this diagram nd the modulus of elasticity, the 0.2 percent offset yield
strength, the ultimate strength, and the percent reduction in area.
210
Compute the true stress and the logarithmic strain using the data of Prob. 29 and plot the results on
log-log paper. Then nd the plastic strength coefcient σ0 and the strain-strengthening exponent m.
Find also the yield strength and the ultimate strength after the specimen has had 20 percent cold work.
211
The stress-strain data from a tensile test on a cast-iron specimen are
Engineering
stress, kpsi
5
16
19
26
32
40
46
49
54
0.20
Engineering strain,
ǫ · 103 in/in
10
0.44
0.80
1.0
1.5
2.0
2.8
3.4
4.0
5.0
Plot the stress-strain locus and nd the 0.1 percent offset yield strength, and the tangent modulus
of elasticity at zero stress and at 20 kpsi.
212
A straight bar of arbitrary cross section and thickness h is cold-formed to an inner radius R about
an anvil as shown in the gure. Some surface at distance N having an original length L A B will
remain unchanged in length after bending. This length is
π( R + N )
2
L AB = L AB =
The lengths of the outer and inner surfaces, after bending, are
Lo =
π
( R + h)
2
Li =
π
R
2
Using Eq. (24), we then nd the true strains to be
εo = ln
R+h
R+N
εi = ln
R
R+N
Tests show that |εo | = |εi |. Show that
N=R
B
B
h
LAB
Problem 212
N
A
R
1+
h
R
1/2
1
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
71
© The McGrawHill
Companies, 2008
2. Materials
Materials
65
and
εo = ln 1 +
h
R
1/2
213
A hot-rolled AISI 1212 steel is given 20 percent cold work. Determine the new values of the yield
and ultimate strengths.
214
A steel member has a Brinell of HB = 250. Estimate the ultimate strength of the steel in MPa.
215
Brinell hardness tests were made on a random sample of 10 steel parts during processing. The
results were HB values of 252 (2), 260, 254, 257 (2), 249 (3), and 251. Estimate the mean and
standard deviation of the ultimate strength in kpsi.
216
Repeat Prob. 215 assuming the material to be cast iron.
217
Toughness is a term that relates to both strength and ductility. The fracture toughness, for examǫ
ple, is dened as the total area under the stress-strain curve to fracture, u T = 0 f σ d ǫ . This area,
called the modulus of toughness, is the strain energy per unit volume required to cause the
material to fracture. A similar term, but dened within the elastic limit of the material, is called
ǫ
the modulus of resilience, u R = 0 y σ d ǫ, where ǫ y is the strain at yield. If the stress-strain is
2
linear to σ = Sy , then it can be shown that u R = Sy /2 E .
For the material in Prob. 29: (a) Determine the modulus of resilience, and (b) Estimate the
modulus of toughness, assuming that the last data point corresponds to fracture.
218
What is the material composition of AISI 4340 steel?
219
Search the website noted in Sec. 220 and report your ndings.
220
Research the material Inconel, briey described in Table A5. Compare it to various carbon and
alloy steels in stiffness, strength, ductility, and toughness. What makes this material so special?
221
Pick a specic material given in the tables (e.g., 2024-T4 aluminum, SAE 1040 steel), and consult a local or regional distributor (consulting either the Yellow Pages or the Thomas Register) to
obtain as much information as you can about cost and availability of the material and in what
form (bar, plate, etc.).
222
Consider a tie rod transmitting a tensile force F. The corresponding tensile stress is given by
σ = F / A , where A is the area of the cross section. The deection of the rod is given by Eq. (43),
which is δ = ( Fl )/( AE ) , where l is the length of the rod. Using the Ashby charts of Figs. 216
and 219, explore what ductile materials are best suited for a light, stiff, and strong tie rod. Hints:
Consider stiffness and strength separately. For use of Fig. 216, prove that β = 1 . For use of Fig.
219, relate the applied tensile stress to the material strength.
72
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
3
Load and Stress Analysis
Chapter Outline
31
Equilibrium and Free-Body Diagrams
32
Shear Force and Bending Moments in Beams
33
Singularity Functions
34
Stress
35
Cartesian Stress Components
75
36
Mohrs Circle for Plane Stress
76
37
General Three-Dimensional Stress
38
Elastic Strain
39
Uniformly Distributed Stresses
68
71
73
75
82
83
84
310
Normal Stresses for Beams in Bending
311
Shear Stresses for Beams in Bending
312
Torsion
313
Stress Concentration
314
Stresses in Pressurized Cylinders
315
Stresses in Rotating Rings
316
Press and Shrink Fits
317
Temperature Effects
318
Curved Beams in Bending
319
Contact Stresses
320
Summary
85
90
95
105
107
110
110
111
112
117
121
67
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
68
I. Basics
73
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
One of the main objectives of this book is to describe how specic machine components
function and how to design or specify them so that they function safely without failing
structurally. Although earlier discussion has described structural strength in terms of
load or stress versus strength, failure of function for structural reasons may arise from
other factors such as excessive deformations or deections.
Here it is assumed that the reader has completed basic courses in statics of rigid
bodies and mechanics of materials and is quite familiar with the analysis of loads, and
the stresses and deformations associated with the basic load states of simple prismatic
elements. In this chapter and Chap. 4 we will review and extend these topics briey.
Complete derivations will not be presented here, and the reader is urged to return to
basic textbooks and notes on these subjects.
This chapter begins with a review of equilibrium and free-body diagrams associated
with load-carrying components. One must understand the nature of forces before
attempting to perform an extensive stress or deection analysis of a mechanical component. An extremely useful tool in handling discontinuous loading of structures
employs Macaulay or singularity functions. Singularity functions are described in
Sec. 33 as applied to the shear forces and bending moments in beams. In Chap. 4, the
use of singularity functions will be expanded to show their real power in handling
deections of complex geometry and statically indeterminate problems.
Machine components transmit forces and motion from one point to another. The
transmission of force can be envisioned as a ow or force distribution that can be further visualized by isolating internal surfaces within the component. Force distributed
over a surface leads to the concept of stress, stress components, and stress transformations (Mohrs circle) for all possible surfaces at a point.
The remainder of the chapter is devoted to the stresses associated with the basic
loading of prismatic elements, such as uniform loading, bending, and torsion, and topics
with major design ramications such as stress concentrations, thin- and thick-walled
pressurized cylinders, rotating rings, press and shrink ts, thermal stresses, curved beams,
and contact stresses.
31
Equilibrium and Free-Body Diagrams
Equilibrium
The word system will be used to denote any isolated part or portion of a machine or
structureincluding all of it if desiredthat we wish to study. A system, under this
denition, may consist of a particle, several particles, a part of a rigid body, an entire
rigid body, or even several rigid bodies.
If we assume that the system to be studied is motionless or, at most, has constant
velocity, then the system has zero acceleration. Under this condition the system is said
to be in equilibrium. The phrase static equilibrium is also used to imply that the system
is at rest. For equilibrium, the forces and moments acting on the system balance such
that
F=0
(31)
M=0
(32)
which states that the sum of all force and the sum of all moment vectors acting upon a
system in equilibrium is zero.
74
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Load and Stress Analysis
69
Free-Body Diagrams
We can greatly simplify the analysis of a very complex structure or machine by successively
isolating each element and studying and analyzing it by the use of free-body diagrams.
When all the members have been treated in this manner, the knowledge can be assembled
to yield information concerning the behavior of the total system. Thus, free-body diagramming is essentially a means of breaking a complicated problem into manageable segments,
analyzing these simple problems, and then, usually, putting the information together again.
Using free-body diagrams for force analysis serves the following important
purposes:
The diagram establishes the directions of reference axes, provides a place to record
the dimensions of the subsystem and the magnitudes and directions of the known
forces, and helps in assuming the directions of unknown forces.
The diagram simplies your thinking because it provides a place to store one thought
while proceeding to the next.
The diagram provides a means of communicating your thoughts clearly and unambiguously to other people.
Careful and complete construction of the diagram claries fuzzy thinking by bringing
out various points that are not always apparent in the statement or in the geometry
of the total problem. Thus, the diagram aids in understanding all facets of the problem.
The diagram helps in the planning of a logical attack on the problem and in setting
up the mathematical relations.
The diagram helps in recording progress in the solution and in illustrating the
methods used.
The diagram allows others to follow your reasoning, showing all forces.
EXAMPLE 31
Solution
Figure 31a shows a simplied rendition of a gear reducer where the input and output
shafts AB and C D are rotating at constant speeds ωi and ωo, respectively. The input and
output torques (torsional moments) are Ti = 240 lbf · in and To, respectively. The shafts
are supported in the housing by bearings at A, B, C, and D. The pitch radii of gears G1
and G2 are r1 = 0.75 in and r2 = 1.5 in, respectively. Draw the free-body diagrams of
each member and determine the net reaction forces and moments at all points.
First, we will list all simplifying assumptions.
1
2
3
4
5
Gears G1 and G2 are simple spur gears with a standard pressure angle φ = 20°
(see Sec. 135).
The bearings are self-aligning and the shafts can be considered to be simply
supported.
The weight of each member is negligible.
Friction is negligible.
The mounting bolts at E, F, H, and I are the same size.
The separate free-body diagrams of the members are shown in Figs. 31bd. Note that
Newtons third law, called the law of action and reaction, is used extensively where
each member mates. The force transmitted between the spur gears is not tangential but
at the pressure angle φ. Thus, N = F tan φ.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
70
I. Basics
Mechanical Engineering Design
F
i,
G1
240 lbf in
Ti
5 in
D
5 in
C
RCz
x
H
C
RI
4 in
RH
I
4 in
(a) Gear reducer
(b) Gear box
1 in
RAz
r1
G1
F
RAz
RCy
I
B
y
RAy
H
1.5 in
RBy
RF
z
A
RDy
D
G2
E
RBz
RDz
A
0
B
RBy
B
F
RE
E
T0
RBz
75
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
RAy
T0
Ti
240 lbf in
N
F
D
r2
C
A
N
RDy
RDz
G2
RCy
(c) Input shaft
RCz
(d ) Output shaft
Figure 31
(a) Gear reducer; (b d ) free-body diagrams. Diagrams are not drawn to scale.
Summing moments about the x axis of shaft A B in Fig. 31d gives
Mx = F (0.75) 240 = 0
F = 320 lbf
The normal force is N = 320 tan 20° = 116.5 lbf.
Using the equilibrium equations for Figs. 31c and d, the reader should verify that:
R Ay = 192 lbf, R Az = 69.9 lbf, R By = 128 lbf, R Bz = 46.6 lbf, RC y = 192 lbf, RC z =
69.9 lbf, R Dy = 128 lbf, R Dz = 46.6 lbf, and To = 480 lbf · in. The direction of the output
torque To is opposite ωo because it is the resistive load on the system opposing the motion ωo.
Note in Fig. 31b the net force from the bearing reactions is zero whereas the net
moment about the x axis is 2.25 (192) + 2.25 (128) = 720 lbf · in. This value is the same
as Ti + To = 240 + 480 = 720 lbf · in, as shown in Fig. 31a. The reaction forces
R E , R F , R H , and R I , from the mounting bolts cannot be determined from the
equilibrium equations as there are too many unknowns. Only three equations are
Fy =
Fz =
Mx = 0. In case you were wondering about assumption
available,
5, here is where we will use it (see Sec. 812). The gear box tends to rotate about the
x axis because of a pure torsional moment of 720 lbf · in. The bolt forces must provide
76
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
71
an equal but opposite torsional moment. The center of rotation relative to the bolts lies at
the centroid of the bolt cross-sectional areas. Thus if the bolt areas are equal: the center
of rotation is at the center of the four bolts, a distance of (4/2)2 + (5/2)2 = 3.202 in
from each bolt; the bolt forces are equal ( R E = R F = R H = R I = R ), and each bolt force
is perpendicular to the line from the bolt to the center of rotation. This gives a net torque
from the four bolts of 4 R (3.202) = 720. Thus, R E = R F = R H = R I = 56.22 lbf.
32
Shear Force and Bending Moments in Beams
Figure 32a shows a beam supported by reactions R1 and R2 and loaded by the concentrated forces F1 , F2 , and F3 . If the beam is cut at some section located at x = x1 and
the left-hand portion is removed as a free body, an internal shear force V and bending
moment M must act on the cut surface to ensure equilibrium (see Fig. 32b). The shear
force is obtained by summing the forces on the isolated section. The bending moment is
the sum of the moments of the forces to the left of the section taken about an axis through
the isolated section. The sign conventions used for bending moment and shear force in this
book are shown in Fig. 33. Shear force and bending moment are related by the equation
V=
dM
dx
(33)
Sometimes the bending is caused by a distributed load q (x ), as shown in Fig. 34;
q (x ) is called the load intensity with units of force per unit length and is positive in the
Figure 32
y
y
Free-body diagram of simplysupported beam with V and M
shown in positive directions.
F2
F1
F3
F1
V
x
x1
x1
R2
R1
R1
(a)
(b)
Figure 33
Sign conventions for bending
and shear.
Positive bending
y
Negative bending
Positive shear
Figure 34
x
M
Negative shear
q (x )
Distributed load on beam.
x
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
72
I. Basics
77
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
positive y direction. It can be shown that differentiating Eq. (33) results in
dV
d2 M
=
=q
dx
dx 2
(34)
Normally the applied distributed load is directed downward and labeled w (e.g., see
Fig. 36). In this case, w = q .
Equations (33) and (34) reveal additional relations if they are integrated. Thus,
if we integrate between, say, x A and x B , we obtain
VB
xB
dV =
VA
(35)
q dx = VB V A
xA
which states that the change in shear force from A to B is equal to the area of the loading diagram between x A and x B .
In a similar manner,
MB
xB
dM =
MA
(36)
V dx = M B M A
xA
which states that the change in moment from A to B is equal to the area of the shearforce diagram between x A and x B .
Table 31
Function
Singularity (Macaulay )
Functions
Concentrated
moment
(unit doublet)
Graph of fn (x)
xa
Meaning
xa
2
xa
x
2
2
xa
=0 x=a
= ± x = a
2
dx = x a
1
a
Concentrated
force
(unit impulse)
xa
xa
1
xa
x
1
1
xa
=0 x=a
= + x = a
1
dx = x a
0
a
Unit step
xa
0
xa
1
x
0
=
xa
0
1
=
0 x<a
1
xa
dx = x a
1
a
Ramp
xa
1
xa
1
1
x
xa
1
0
x<a
xa
xa
dx =
xa
2
a
W. H. Macaulay, Note on the deection of beams, Messenger of Mathematics, vol. 48, pp. 129130, 1919.
2
78
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
33
73
Singularity Functions
The four singularity functions dened in Table 31 constitute a useful and easy means
of integrating across discontinuities. By their use, general expressions for shear force
and bending moment in beams can be written when the beam is loaded by concentrated
moments or forces. As shown in the table, the concentrated moment and force functions
are zero for all values of x not equal to a. The functions are undened for values of
x = a . Note that the unit step and ramp functions are zero only for values of x that are
less than a. The integration properties shown in the table constitute a part of the mathematical denition too. The rst two integrations of q (x ) for V (x ) and M (x ) do not
require constants of integration provided all loads on the beam are accounted for in
q (x ). The examples that follow show how these functions are used.
EXAMPLE 32
Derive expressions for the loading, shear-force, and bending-moment diagrams for the
beam of Fig. 35.
Figure 35
y
l
q
F1
F2
x
O
a1
R1
Solution
Answer
a2
R2
Using Table 31 and q (x ) for the loading function, we nd
q = R1 x
1
F1 x a1
1
F2 x a2
1
+ R2 x l
1
(1)
Next, we use Eq. (35) to get the shear force.
Answer
V=
0
q dx = R1 x
F1 x a1
0
F2 x a2
0
+ R2 x l
0
(2)
+ R2 x l
1
(3)
Note that V = 0 at x = 0 .
A second integration, in accordance with Eq. (36), yields
Answer
M=
V dx = R1 x
1
F1 x a1
1
F2 x a2
1
The reactions R1 and R2 can be found by taking a summation of moments and forces
as usual, or they can be found by noting that the shear force and bending moment must
be zero everywhere except in the region 0 x l . This means that Eq. (2) should give
V = 0 at x slightly larger than l. Thus
R1 F1 F2 + R2 = 0
(4)
Since the bending moment should also be zero in the same region, we have, from Eq. (3),
R1l F1 (l a1 ) F2 (l a2 ) = 0
Equations (4) and (5) can now be solved for the reactions R1 and R2 .
(5)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
74
I. Basics
79
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
EXAMPLE 33
Solution
Figure 36a shows the loading diagram for a beam cantilevered at A with a uniform
load of 20 lbf/in acting on the portion 3 in x 7 in, and a concentrated counterclockwise moment of 240 lbf · in at x = 10 in. Derive the shear-force and bendingmoment relations, and the support reactions M1 and R1 .
Following the procedure of Example 32, we nd the load intensity function to be
q = M1 x
2
+ R1 x
1
20 x 3 0 + 20 x 7 0 240 x 10
2
(1)
Note that the 20 x 7 0 term was necessary to turn off the uniform load at C.
Integrating successively gives
Answers
V = M1 x
M = M1 x
1
0
+ R1 x
+ R1 x
1
0
20 x 3 1 + 20 x 7 1 240 x 10
10 x 3 2 + 10 x 7 2 240 x 10
0
1
(2)
(3)
The reactions are found by making x slightly larger than 10 in, where both V and M are
zero in this region. Equation (2) will then give
M1 (0) + R1 (1) 20(10 3) + 20(10 7) 240(0) = 0
Answer
which yields R1 = 80 lbf.
From Eq. (3) we get
M1 (1) + 80(10) 10(10 3)2 + 10(10 7)2 240(1) = 0
Answer
which yields M1 = 160 lbf · in.
Figures 36b and c show the shear-force and bending-moment diagrams. Note that
the impulse terms in Eq. (2), M1 x 1 and 240 x 10 1 , are physically not forces
Figure 36
y
(a) Loading diagram for a
beam cantilevered at A.
(b) Shear-force diagram.
(c) Bending-moment diagram.
q
10 in
7 in
3 in
20 lbf/in
240 lbf in
D
B
A
(a)
M1
C
x
R1
V (lbf)
80
(b)
Step
Ramp
x
O
M (lbf in)
240
Parabolic
Step
80
x
O
(c)
160
Ramp
Slope = 80 lbf in/in
80
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Load and Stress Analysis
75
and are not shown in the V diagram. Also note that both the M1 and 240 lbf · in
moments are counterclockwise and negative singularity functions; however, by the convention shown in Fig. 32 the M1 and 240 lbf · in are negative and positive bending
moments, respectively, which is reected in Fig. 36c.
34
Stress
When an internal surface is isolated as in Fig. 32b, the net force and moment acting on
the surface manifest themselves as force distributions across the entire area. The force
distribution acting at a point on the surface is unique and will have components in the
normal and tangential directions called normal stress and tangential shear stress,
respectively. Normal and shear stresses are labeled by the Greek symbols σ and τ ,
respectively. If the direction of σ is outward from the surface it is considered to be a tensile stress and is a positive normal stress. If σ is into the surface it is a compressive stress
and commonly considered to be a negative quantity. The units of stress in U.S.
Customary units are pounds per square inch (psi). For SI units, stress is in newtons per
square meter (N/m2 ); 1 N/m2 = 1 pascal (Pa).
35
Cartesian Stress Components
The Cartesian stress components are established by defining three mutually orthogonal surfaces at a point within the body. The normals to each surface will establish the
x, y, z Cartesian axes. In general, each surface will have a normal and shear stress.
The shear stress may have components along two Cartesian axes. For example, Fig.
37 shows an infinitesimal surface area isolation at a point Q within a body where
the surface normal is the x direction. The normal stress is labeled σx . The symbol σ
indicates a normal stress and the subscript x indicates the direction of the surface
normal. The net shear stress acting on the surface is (τx )net which can be resolved into
components in the y and z directions, labeled as τx y and τx z , respectively (see
Fig. 37). Note that double subscripts are necessary for the shear. The first subscript
indicates the direction of the surface normal whereas the second subscript is the
direction of the shear stress.
The state of stress at a point described by three mutually perpendicular surfaces is
shown in Fig. 38a. It can be shown through coordinate transformation that this is sufcient to determine the state of stress on any surface intersecting the point. As the
y
Figure 37
Stress components on surface
normal to x direction.
xy
( x)net
Q
xz
z
x
x
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
76
I. Basics
81
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
y
Figure 38
y
y
(a) General three-dimensional
stress. (b) Plane stress with
cross-shears equal.
y
xy
yx
yz
xy
xy
x
x
zy
x
x
x
xy
zx
xz
xy
y
z
z
(a )
(b)
dimensions of the cube in Fig. 38a approach zero, the stresses on the hidden faces
become equal and opposite to those on the opposing visible faces. Thus, in general, a
complete state of stress is dened by nine stress components, σx , σ y , σz , τx y ,
τx z , τ yx , τ yz , τzx , and τzy .
For equilibrium, in most cases, cross-shears are equal, hence
τ yx = τx y
τzy = τ yz
τx z = τzx
(37)
This reduces the number of stress components for most three-dimensional states of
stress from nine to six quantities, σx , σ y , σz , τx y , τ yz , and τzx .
A very common state of stress occurs when the stresses on one surface are zero.
When this occurs the state of stress is called plane stress. Figure 38b shows a state of
plane stress, arbitrarily assuming that the normal for the stress-free surface is the
z direction such that σz = τzx = τzy = 0. It is important to note that the element in
Fig. 38b is still a three-dimensional cube. Also, here it is assumed that the cross-shears
are equal such that τ yx = τx y , and τ yz = τzy = τx z = τzx = 0.
36
Mohrs Circle for Plane Stress
Suppose the d x dy dz element of Fig. 38b is cut by an oblique plane with a normal n at
an arbitrary angle φ counterclockwise from the x axis as shown in Fig. 39. This section
is concerned with the stresses σ and τ that act upon this oblique plane. By summing the
forces caused by all the stress components to zero, the stresses σ and τ are found to be
σ=
σx + σ y
σx σ y
+
cos 2φ + τx y sin 2φ
2
2
τ =
σx σ y
sin 2φ + τx y cos 2φ
2
(38)
(39)
Equations (38) and (39) are called the plane-stress transformation equations.
Differentiating Eq. (38) with respect to φ and setting the result equal to zero gives
tan 2φ p =
2τx y
σx σ y
(310)
82
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
77
y
Figure 39
n
x
dy
dy
ds
ds
dx
dx
xy
xy
x
y
Equation (310) denes two particular values for the angle 2φ p , one of which denes
the maximum normal stress σ1 and the other, the minimum normal stress σ2 . These two
stresses are called the principal stresses, and their corresponding directions, the principal directions. The angle between the principal directions is 90°. It is important to note
that Eq. (310) can be written in the form
σx σ y
sin 2φ p τx y cos 2φ p = 0
2
(a)
Comparing this with Eq. (39), we see that τ = 0, meaning that the surfaces containing principal stresses have zero shear stresses.
In a similar manner, we differentiate Eq. (39), set the result equal to zero, and
obtain
tan 2φs =
σx σ y
2τx y
(311)
Equation (311) denes the two values of 2φs at which the shear stress τ reaches an
extreme value. The angle between the surfaces containing the maximum shear stresses
is 90°. Equation (311) can also be written as
σx σ y
cos 2φ p + τx y sin 2φ p = 0
2
(b)
Substituting this into Eq. (38) yields
σ=
σx + σ y
2
(312)
Equation (312) tells us that the two surfaces containing the maximum shear stresses
also contain equal normal stresses of (σx + σ y )/2.
Comparing Eqs. (310) and (311), we see that tan 2φs is the negative reciprocal
of tan 2φ p . This means that 2φs and 2φ p are angles 90° apart, and thus the angles
between the surfaces containing the maximum shear stresses and the surfaces containing the principal stresses are ±45 .
Formulas for the two principal stresses can be obtained by substituting the
angle 2φ p from Eq. (310) in Eq. (38). The result is
σ1 , σ2 =
σx + σ y
±
2
σx σ y
2
2
2
+ τx y
(313)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
78
I. Basics
83
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
In a similar manner the two extreme-value shear stresses are found to be
τ1 , τ2 = ±
σx σ y
2
2
2
+ τx y
(314)
Your particular attention is called to the fact that an extreme value of the shear stress
may not be the same as the actual maximum value. See Sec. 37.
It is important to note that the equations given to this point are quite sufcient for
performing any plane stress transformation. However, extreme care must be exercised
when applying them. For example, say you are attempting to determine the principal
state of stress for a problem where σx = 14 MPa, σ y = 10 MPa, and τx y = 16 MPa.
Equation (310) yields φ p = 26.57 and 63.43° to locate the principal stress surfaces,
whereas, Eq. (313) gives σ1 = 22 MPa and σ2 = 18 MPa for the principal stresses.
If all we wanted was the principal stresses, we would be nished. However, what if
we wanted to draw the element containing the principal stresses properly oriented relative to the x, y axes? Well, we have two values of φ p and two values for the principal stresses. How do we know which value of φ p corresponds to which value of the
principal stress? To clear this up we would need to substitute one of the values of φ p
into Eq. (38) to determine the normal stress corresponding to that angle.
A graphical method for expressing the relations developed in this section, called
Mohrs circle diagram, is a very effective means of visualizing the stress state at a point
and keeping track of the directions of the various components associated with plane
stress. Equations (38) and (39) can be shown to be a set of parametric equations for
σ and τ , where the parameter is 2φ. The relationship between σ and τ is that of a circle plotted in the σ, τ plane, where the center of the circle is located at C = (σ, τ ) =
2
[(σx + σ y )/2, 0] and has a radius of R = [(σx σ y )/2]2 + τx y . A problem arises in
the sign of the shear stress. The transformation equations are based on a positive φ
being counterclockwise, as shown in Fig. 39. If a positive τ were plotted above the
σ axis, points would rotate clockwise on the circle 2φ in the opposite direction of
rotation on the element. It would be convenient if the rotations were in the same
direction. One could solve the problem easily by plotting positive τ below the axis.
However, the classical approach to Mohrs circle uses a different convention for the
shear stress.
Mohrs Circle Shear Convention
This convention is followed in drawing Mohrs circle:
Shear stresses tending to rotate the element clockwise (cw) are plotted above the
σ axis.
Shear stresses tending to rotate the element counterclockwise (ccw) are plotted below
the σ axis.
For example, consider the right face of the element in Fig. 38b. By Mohrs circle convention the shear stress shown is plotted below the σ axis because it tends to rotate the
element counterclockwise. The shear stress on the top face of the element is plotted
above the σ axis because it tends to rotate the element clockwise.
In Fig. 310 we create a coordinate system with normal stresses plotted along the
abscissa and shear stresses plotted as the ordinates. On the abscissa, tensile (positive)
normal stresses are plotted to the right of the origin O and compressive (negative) normal stresses to the left. On the ordinate, clockwise (cw) shear stresses are plotted up;
counterclockwise (ccw) shear stresses are plotted down.
84
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
79
cw
Figure 310
x
(
y
Mohrs circle diagram.
x
y)
x
F
y
B
(
2
y
H
cw
y , xy )
xy
2
E
O
2
D
C
y
x
1
xy
x
2
y2
+
2
xy
(
A
x,
2
p
ccw
xy )
x
ccw
x
+
2
y
G
Using the stress state of Fig. 38b, we plot Mohrs circle, Fig. 310, by rst looking at the right surface of the element containing σx to establish the sign of σx and the
cw or ccw direction of the shear stress. The right face is called the x face where
φ = 0 . If σx is positive and the shear stress τx y is ccw as shown in Fig. 38b, we can
ccw
establish point A with coordinates (σx , τx y ) in Fig. 310. Next, we look at the top y
face, where φ = 90 , which contains σ y , and repeat the process to obtain point B with
cw
coordinates (σ y , τx y ) as shown in Fig. 310. The two states of stress for the element
are φ = 90 from each other on the element so they will be 2 φ = 180 from each
other on Mohrs circle. Points A and B are the same vertical distance from the σ axis.
Thus, AB must be on the diameter of the circle, and the center of the circle C is where
AB intersects the σ axis. With points A and B on the circle, and center C, the complete
circle can then be drawn. Note that the extended ends of line A B are labeled x and y
as references to the normals to the surfaces for which points A and B represent the
stresses.
The entire Mohrs circle represents the state of stress at a single point in a structure. Each point on the circle represents the stress state for a specic surface intersecting the point in the structure. Each pair of points on the circle 180° apart represent the
state of stress on an element whose surfaces are 90° apart. Once the circle is drawn, the
states of stress can be visualized for various surfaces intersecting the point being analyzed. For example, the principal stresses σ1 and σ2 are points D and E, respectively,
and their values obviously agree with Eq. (313). We also see that the shear stresses
are zero on the surfaces containing σ1 and σ2 . The two extreme-value shear stresses, one
clockwise and one counterclockwise, occur at F and G with magnitudes equal to the
radius of the circle. The surfaces at F and G each also contain normal stresses of
(σx + σ y )/2 as noted earlier in Eq. (312). Finally, the state of stress on an arbitrary
surface located at an angle φ counterclockwise from the x face is point H.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
80
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
At one time, Mohrs circle was used graphically where it was drawn to scale very
accurately and values were measured by using a scale and protractor. Here, we are strictly
using Mohrs circle as a visualization aid and will use a semigraphical approach, calculating values from the properties of the circle. This is illustrated by the following example.
EXAMPLE 34
A stress element has σx = 80 MPa and τx y = 50 MPa cw, as shown in Fig. 311a.
(a) Using Mohrs circle, nd the principal stresses and directions, and show these
on a stress element correctly aligned with respect to the x y coordinates. Draw another
stress element to show τ1 and τ2 , nd the corresponding normal stresses, and label the
drawing completely.
(b) Repeat part a using the transformation equations only.
Solution
(a) In the semigraphical approach used here, we rst make an approximate freehand
sketch of Mohrs circle and then use the geometry of the gure to obtain the desired
information.
Draw the σ and τ axes rst (Fig. 311b) and from the x face locate σx = 80 MPa
along the σ axis. On the x face of the element, we see that the shear stress is 50 MPa in
the cw direction. Thus, for the x face, this establishes point A (80, 50cw) MPa.
Corresponding to the y face, the stress is σ = 0 and τ = 50 MPa in the ccw direction.
This locates point B (0, 50ccw) MPa. The line A B forms the diameter of the required circle, which can now be drawn. The intersection of the circle with the σ axis denes σ1
and σ2 as shown. Now, noting the triangle AC D , indicate on the sketch the length of the
legs A D and C D as 50 and 40 MPa, respectively. The length of the hypotenuse AC is
Answer
τ1 =
(50)2 + (40)2 = 64.0 MPa
and this should be labeled on the sketch too. Since intersection C is 40 MPa from the
origin, the principal stresses are now found to be
Answer
σ1 = 40 + 64 = 104 MPa
and
σ2 = 40 64 = 24 MPa
The angle 2φ from the x axis cw to σ1 is
Answer
2φ p = tan1
50
40
= 51.3
To draw the principal stress element (Fig. 311c), sketch the x and y axes parallel
to the original axes. The angle φ p on the stress element must be measured in the same
direction as is the angle 2φ p on the Mohr circle. Thus, from x measure 25.7° (half of
51.3°) clockwise to locate the σ1 axis. The σ2 axis is 90° from the σ1 axis and the stress
element can now be completed and labeled as shown. Note that there are no shear
stresses on this element.
The two maximum shear stresses occur at points E and F in Fig. 311b. The two
normal stresses corresponding to these shear stresses are each 40 MPa, as indicated.
Point E is 38.7° ccw from point A on Mohrs circle. Therefore, in Fig. 311d, draw a
stress element oriented 19.3° (half of 38.7°) ccw from x. The element should then be
labeled with magnitudes and directions as shown.
In constructing these stress elements it is important to indicate the x and y directions of the original reference system. This completes the link between the original
machine element and the orientation of its principal stresses.
85
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
81
Load and Stress Analysis
y
Figure 311
x
cw
1
All stresses in MPa.
(80, 50cw)
E
50
A
80
64
86
38.7°
x
2
p
50
50
y=
2
(a )
0
51.3°
C
40
D
40
1
x=
(0, 50ccw)
80
B
F
2
ccw
y
(b)
y
2
y
= 40
2=
24
2=
64
= 40
F
Answer
104
x
1=
25.7°
1=
19.3°
E
x
64
1
(c)
(d )
(b) The transformation equations are programmable. From Eq. (310),
φp =
1
tan1
2
2τx y
σx σ y
=
1
tan1
2
2(50)
80
= 25.7 , 64.3
From Eq. (38), for the rst angle φ p = 25.7 ,
σ=
80 + 0 80 0
+
cos[2(25.7)] + (50) sin[2(25.7)] = 104.03 MPa
2
2
The shear on this surface is obtained from Eq. (39) as
τ =
80 0
sin[2(25.7)] + (50) cos[2(25.7)] = 0 MPa
2
which conrms that 104.03 MPa is a principal stress. From Eq. (38), for φ p = 64.3 ,
σ=
80 + 0 80 0
+
cos[2(64.3)] + (50) sin[2(64.3)] = 24.03 MPa
2
2
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
82
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Answer
Substituting φ p = 64.3 into Eq. (39) again yields τ = 0, indicating that 24.03 MPa
is also a principal stress. Once the principal stresses are calculated they can be ordered
such that σ1 σ2 . Thus, σ1 = 104.03 MPa and σ2 = 24.03 MPa.
Since for σ1 = 104.03 MPa, φ p = 25.7 , and since φ is dened positive ccw in the
transformation equations, we rotate clockwise 25.7° for the surface containing σ1 . We
see in Fig. 311c that this totally agrees with the semigraphical method.
To determine τ1 and τ2 , we rst use Eq. (311) to calculate φs :
φs =
σx σ y
1
tan1
2
2τx y
=
1
80
tan1
2
2(50)
= 19.3 , 109.3
For φs = 19.3 , Eqs. (38) and (39) yield
Answer
σ=
80 + 0 80 0
+
cos[2(19.3)] + (50) sin[2(19.3)] = 40.0 MPa
2
2
τ =
80 0
sin[2(19.3)] + (50) cos[2(19.3)] = 64.0 MPa
2
Remember that Eqs. (38) and (39) are coordinate transformation equations. Imagine
that we are rotating the x, y axes 19.3° counterclockwise and y will now point up and
to the left. So a negative shear stress on the rotated x face will point down and to the
right as shown in Fig. 311d. Thus again, results agree with the semigraphical method.
For φs = 109.3 , Eqs. (38) and (39) give σ = 40.0 MPa and τ = +64.0 MPa.
Using the same logic for the coordinate transformation we nd that results again agree
with Fig. 311d.
37
General Three-Dimensional Stress
As in the case of plane stress, a particular orientation of a stress element occurs in space
for which all shear-stress components are zero. When an element has this particular orientation, the normals to the faces are mutually orthogonal and correspond to the principal directions, and the normal stresses associated with these faces are the principal
stresses. Since there are three faces, there are three principal directions and three principal stresses σ1 , σ2 , and σ3 . For plane stress, the stress-free surface contains the third
principal stress which is zero.
In our studies of plane stress we were able to specify any stress state σx , σ y , and
τx y and find the principal stresses and principal directions. But six components of
stress are required to specify a general state of stress in three dimensions, and the
problem of determining the principal stresses and directions is more difficult. In
design, three-dimensional transformations are rarely performed since most maximum stress states occur under plane stress conditions. One notable exception is contact stress, which is not a case of plane stress, where the three principal stresses are
given in Sec. 319. In fact, all states of stress are truly three-dimensional, where
they might be described one- or two-dimensionally with respect to specific coordinate axes. Here it is most important to understand the relationship amongst the three
principal stresses. The process in finding the three principal stresses from the six
87
88
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
83
Figure 312
Mohrs circles for threedimensional stress.
1/3
1/2
2/3
1/2
3
2
1
1
2
(a)
(b )
stress components σx , σ y , σz , τx y , τ yz , and τzx , involves finding the roots of the cubic
equation1
2
2
2
σ 3 (σx + σ y + σz )σ 2 + σx σ y + σx σz + σ y σz τx y τ yz τzx σ
2
2
2
σx σ y σz + 2τx y τ yz τzx σx τ yz σ y τzx σz τx y = 0
(315)
In plotting Mohrs circles for three-dimensional stress, the principal normal
stresses are ordered so that σ1 σ2 σ3 . Then the result appears as in Fig. 312a. The
stress coordinates σ , τ for any arbitrarily located plane will always lie on the boundaries or within the shaded area.
Figure 312a also shows the three principal shear stresses τ1/2 , τ2/3 , and τ1/3 .2
Each of these occurs on the two planes, one of which is shown in Fig. 312b. The gure shows that the principal shear stresses are given by the equations
τ1/2 =
σ1 σ2
2
τ2/3 =
σ2 σ3
2
τ1/3 =
σ1 σ3
2
(316)
Of course, τmax = τ1/3 when the normal principal stresses are ordered (σ1 > σ2 > σ3 ) ,
so always order your principal stresses. Do this in any computer code you generate and
youll always generate τmax .
38
Elastic Strain
Normal strain ǫ is dened and discussed in Sec. 2-1 for the tensile specimen and is
given by Eq. (22) as ǫ = δ/ l , where δ is the total elongation of the bar within the
length l. Hookes law for the tensile specimen is given by Eq. (23) as
σ = Eǫ
(317)
where the constant E is called Youngs modulus or the modulus of elasticity.
1
For development of this equation and further elaboration of three-dimensional stress transformations see:
Richard G. Budynas, Advanced Strength and Applied Stress Analysis, 2nd ed., McGraw-Hill, New York,
1999, pp. 4678.
2
Note the difference between this notation and that for a shear stress, say, τx y . The use of the shilling mark is
not accepted practice, but it is used here to emphasize the distinction.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
84
I. Basics
89
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
When a material is placed in tension, there exists not only an axial strain, but also
negative strain (contraction) perpendicular to the axial strain. Assuming a linear,
homogeneous, isotropic material, this lateral strain is proportional to the axial strain. If
the axial direction is x, then the lateral strains are ǫ y = ǫz = νǫx . The constant of proportionality v is called Poissons ratio, which is about 0.3 for most structural metals.
See Table A5 for values of v for common materials.
If the axial stress is in the x direction, then from Eq. (317)
ǫx =
σx
E
ǫ y = ǫz = ν
σx
E
(318)
For a stress element undergoing σx , σ y , and σz simultaneously, the normal strains
are given by
1
σ x ν ( σ y + σz )
E
1
σ y ν ( σ x + σz )
ǫy =
E
1
σz ν ( σ x + σ y )
ǫz =
E
ǫx =
(319)
Shear strain γ is the change in a right angle of a stress element when subjected to
pure shear stress, and Hookes law for shear is given by
τ = Gγ
(320)
where the constant G is the shear modulus of elasticity or modulus of rigidity.
It can be shown for a linear, isotropic, homogeneous material, the three elastic constants are related to each other by
E = 2G (1 + ν)
39
(321)
Uniformly Distributed Stresses
The assumption of a uniform distribution of stress is frequently made in design. The
result is then often called pure tension, pure compression, or pure shear, depending
upon how the external load is applied to the body under study. The word simple is sometimes used instead of pure to indicate that there are no other complicating effects.
The tension rod is typical. Here a tension load F is applied through pins at the ends of
the bar. The assumption of uniform stress means that if we cut the bar at a section
remote from the ends and remove one piece, we can replace its effect by applying a uniformly distributed force of magnitude σ A to the cut end. So the stress σ is said to be
uniformly distributed. It is calculated from the equation
σ=
F
A
(322)
This assumption of uniform stress distribution requires that:
The bar be straight and of a homogeneous material
The line of action of the force contains the centroid of the section
The section be taken remote from the ends and from any discontinuity or abrupt
change in cross section
90
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
85
For simple compression, Eq. (322) is applicable with F normally being considered a negative quantity. Also, a slender bar in compression may fail by buckling,
and this possibility must be eliminated from consideration before Eq. (322) is
used.3
Use of the equation
F
τ=
(323)
A
for a body, say, a bolt, in shear assumes a uniform stress distribution too. It is very
difcult in practice to obtain a uniform distribution of shear stress. The equation is
included because occasions do arise in which this assumption is utilized.
310
Normal Stresses for Beams in Bending
The equations for the normal bending stresses in straight beams are based on the following assumptions:
1
2
3
4
5
6
7
The beam is subjected to pure bending. This means that the shear force is zero,
and that no torsion or axial loads are present.
The material is isotropic and homogeneous.
The material obeys Hookes law.
The beam is initially straight with a cross section that is constant throughout the
beam length.
The beam has an axis of symmetry in the plane of bending.
The proportions of the beam are such that it would fail by bending rather than by
crushing, wrinkling, or sidewise buckling.
Plane cross sections of the beam remain plane during bending.
In Fig. 313 we visualize a portion of a straight beam acted upon by a positive
bending moment M shown by the curved arrow showing the physical action of the
moment together with a straight arrow indicating the moment vector. The x axis is
coincident with the neutral axis of the section, and the xz plane, which contains the
neutral axes of all cross sections, is called the neutral plane. Elements of the beam
coincident with this plane have zero stress. The location of the neutral axis with
respect to the cross section is coincident with the centroidal axis of the cross
section.
Figure 313
y
Straight beam in positive
bending.
M
z
M
3
See Sec. 411.
x
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
86
I. Basics
91
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
y
Figure 314
Bending stresses according to
Eq. (324).
Compression
c
y
Neutral axis, Centroidal axis
x
Tension
The bending stress varies linearly with the distance from the neutral axis, y, and is
given by
σx =
My
I
(324)
where I is the second moment of area about the z axis. That is
I=
y2d A
(325)
The stress distribution given by Eq. (324) is shown in Fig. 314. The maximum magnitude of the bending stress will occur where y has the greatest magnitude. Designating σmax
as the maximum magnitude of the bending stress, and c as the maximum magnitude of y
σmax =
Mc
I
(326a)
Equation (324) can still be used to ascertain as to whether σmax is tensile or compressive.
Equation (326a) is often written as
σmax =
M
Z
(326b)
where Z = I/c is called the section modulus.
EXAMPLE 35
A beam having a T section with the dimensions shown in Fig. 315 is subjected to a
bending moment of 1600 N · m that causes tension at the top surface. Locate the neutral axis and nd the maximum tensile and compressive bending stresses.
Solution
The area of the composite section is A = 1956 mm2 . Now divide the T section into two
rectangles, numbered 1 and 2, and sum the moments of these areas about the top edge.
We then have
1956c1 = 12(75)(6) + 12(88)(56)
and hence c1 = 32.99 mm. Therefore c2 = 100 32.99 = 67.01 mm.
Next we calculate the second moment of area of each rectangle about its own centroidal axis. Using Table A-18, we nd for the top rectangle
I1 =
13
1
bh =
(75)123 = 1.080 × 104 mm4
12
12
92
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
Figure 315
87
y
Dimensions in millimeters.
75
12
1
c1
z
100
2
c2
12
For the bottom rectangle, we have
I2 =
1
(12)883 = 6.815 × 105 mm4
12
We now employ the parallel-axis theorem to obtain the second moment of area of the
composite gure about its own centroidal axis. This theorem states
Iz = Icg + Ad 2
where Icg is the second moment of area about its own centroidal axis and Iz is the second moment of area about any parallel axis a distance d removed. For the top rectangle, the distance is
d1 = 32.99 6 = 26.99 mm
and for the bottom rectangle,
d2 = 67.01 44 = 23.01 mm
Using the parallel-axis theorem for both rectangles, we now nd that
I = [1.080 × 104 + 12(75)26.992 ] + [6.815 × 105 + 12(88)23.012 ]
= 1.907 × 106 mm4
Finally, the maximum tensile stress, which occurs at the top surface, is found to be
Answer
σ=
1600(32.99)103
Mc1
=
= 27.68(106 ) Pa = 27.68 MPa
I
1.907(106 )
Similarly, the maximum compressive stress at the lower surface is found to be
Answer
σ =
1600(67.01)103
Mc2
=
= 56.22(106 ) Pa = 56.22 MPa
I
1.907(106 )
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
88
I. Basics
93
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Two-Plane Bending
Quite often, in mechanical design, bending occurs in both xy and xz planes. Considering
cross sections with one or two planes of symmetry only, the bending stresses are given by
σx =
My z
Mz y
+
Iz
Iy
(327)
where the rst term on the right side of the equation is identical to Eq. (324), M y is
the bending moment in the xz plane (moment vector in y direction), z is the distance
from the neutral y axis, and I y is the second area moment about the y axis.
For noncircular cross sections, Eq. (327) is the superposition of stresses caused
by the two bending moment components. The maximum tensile and compressive bending stresses occur where the summation gives the greatest positive and negative stresses, respectively. For solid circular cross sections, all lateral axes are the same and the
plane containing the moment corresponding to the vector sum of Mz and M y contains
the maximum bending stresses. For a beam of diameter d the maximum distance from
the neutral axis is d/2, and from Table A18, I = π d 4/64. The maximum bending stress
for a solid circular cross section is then
2
2
( M y + Mz ) 1/2 ( d /2)
32
Mc
2
=
=
( M 2 + Mz ) 1/2
4 /64
I
πd
π d3 y
σm =
EXAMPLE 36
As shown in Fig. 316a, beam OC is loaded in the xy plane by a uniform load of 50
lbf/in, and in the xz plane by a concentrated force of 100 lbf at end C. The beam is 8 in
long.
Figure 316
(a) Beam loaded in two
planes; (b) loading and
bending-moment diagrams
in xy plane; (c) loading and
bending-moment diagrams
in xz plane.
(328)
y
y
50 lbf/in
x
A
O
1600 lbf-in 400 lbf
50 lbf/in
OB
z
C
Mz
(lbf-in)
C
1.5 in
0
x
x
100 lbf
1600
0.75 in
(b )
(a )
100 lbf
800 lbf-in
x
O
C
z
100 lbf
My
(lbf-in)
800
x
0
(c)
94
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
89
(a) For the cross section shown determine the maximum tensile and compressive
bending stresses and where they act.
(b) If the cross section was a solid circular rod of diameter, d = 1.25 in, determine
the magnitude of the maximum bending stress.
Solution
(a) The reactions at O and the bending-moment diagrams in the xy and xz planes are
shown in Figs. 316b and c, respectively. The maximum moments in both planes occur
at O where
1
( M y ) O = 100(8) = 800 lbf-in
( Mz ) O = (50)82 = 1600 lbf-in
2
The second moments of area in both planes are
Iz =
1
(0.75)1.53 = 0.2109 in4
12
Iy =
1
(1.5)0.753 = 0.05273 in4
12
The maximum tensile stress occurs at point A, shown in Fig. 316a, where the maximum tensile stress is due to both moments. At A, y A = 0.75 in and z A = 0.375 in. Thus,
from Eq. (327)
Answer
(σx ) A =
1600(0.75) 800(0.375)
+
= 11 380 psi = 11.38 kpsi
0.2109
0.05273
The maximum compressive bending stress occurs at point B where, y B = 0.75 in and
z B = 0.375 in. Thus
Answer
(σx ) B =
1600(0.75) 800(0.375)
+
= 11 380 psi = 11.38 kpsi
0.2109
0.05273
(b) For a solid circular cross section of diameter, d = 1.25 in, the maximum bending
stress at end O is given by Eq. (328) as
Answer
σm =
32
8002 + (1600)2
π(1.25)3
1/2
= 9326 psi = 9.329 kpsi
Beams with Asymmetrical Sections
The relations developed earlier in this section can also be applied to beams having
asymmetrical sections, provided that the plane of bending coincides with one of the two
principal axes of the section. We have found that the stress at a distance y from the neutral axis is
σ =
My
I
(a)
Therefore, the force on the element of area d A in Fig. 317 is
dF = σ dA =
My
dA
I
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
90
I. Basics
95
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 317
y
y
z
M
y
x
dA
z
Taking moments of this force about the y axis and integrating across the section gives
My =
z dF =
σz dA =
M
I
yz d A
(b)
We recognize that the last integral in Eq. (b) is the product of inertia I yz . If the bending
moment on the beam is in the plane of one of the principal axes, say the x y plane, then
I yz =
yz d A = 0
(c)
With this restriction, the relations developed in Sec. 310 hold for any cross-sectional
shape. Of course, this means that the designer has a special responsibility to ensure that
the bending loads do, in fact, come onto the beam in a principal plane!
311
Shear Stresses for Beams in Bending
Most beams have both shear forces and bending moments present. It is only occasionally that we encounter beams subjected to pure bending, that is to say, beams having
zero shear force. The exure formula is developed on the assumption of pure bending.
This is done, however, to eliminate the complicating effects of shear force in the development. For engineering purposes, the exure formula is valid no matter whether a
shear force is present or not. For this reason, we shall utilize the same normal bendingstress distribution [Eqs. (324) and (326)] when shear forces are also present.
In Fig. 318a we show a beam segment of constant cross section subjected to a
shear force V and a bending moment M at x. Because of external loading and V, the
shear force and bending moment change with respect to x. At x + dx the shear force
and bending moment are V + d V and M + d M , respectively. Considering forces in the
x direction only, Fig. 318b shows the stress distribution σx due to the bending
moments. If dM is positive, with the bending moment increasing, the stresses on the
right face, for a given value of y, are larger in magnitude than the stresses on the left
face. If we further isolate the element by making a slice at y = y1 (see Fig. 318b), the
net force in the x direction will be directed to the left with a value of
c
y1
(d M ) y
dA
I
as shown in the rotated view of Fig. 318c. For equilibrium, a shear force on the bottom
face, directed to the right, is required. This shear force gives rise to a shear stress τ ,
where, if assumed uniform, the force is τ b d x . Thus
c
τ b dx =
y1
(d M ) y
dA
I
(a)
96
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
w (x)
y
x
My
I
x
M
x
dM
M
x
V
x
dMy
I
c
y1
V
My
I
91
dV
dx
dx
(b)
(a )
A
Figure 318
Beam section isolation. Note:
Only forces shown in x
direction on dx element in (b).
y
F
dx
c dM y
I
y1
b
y1
x
(c)
The term dM/I can be removed from within the integral and b dx placed on the right
side of the equation; then, from Eq. (33) with V = d M /dx , Eq. (a) becomes
τ=
V
Ib
c
yd A
(329)
y1
In this equation, the integral is the rst moment of the area A with respect to the neutral axis (see Fig. 318c). This integral is usually designated as Q. Thus
c
Q=
y1
yd A = y A
¯
(330)
¯
where, for the isolated area y1 to c, y is the distance in the y direction from the neutral
plane to the centroid of the area A . With this, Eq. (329) can be written as
τ=
VQ
Ib
(331)
In using this equation, note that b is the width of the section at y = y1 . Also, I is the
second moment of area of the entire section about the neutral axis.
Because cross shears are equal, and area A is nite, the shear stress τ given by
Eq. (331) and shown on area A in Fig. 318c occurs only at y = y1 . The shear stress
on the lateral area varies with y (normally maximum at the neutral axis where y = 0,
and zero at the outer bers of the beam where Q A 0).
EXAMPLE 37
A beam 12 in long is to support a load of 488 lbf acting 3 in from the left support, as
shown in Fig. 319a. Basing the design only on bending stress, a designer has selected
a 3-in aluminum channel with the cross-sectional dimensions shown. If the direct shear
is neglected, the stress in the beam may be actually higher than the designer thinks.
Determine the principal stresses considering bending and direct shear and compare
them with that considering bending only.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
92
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 319
y
488 lbf
3 in
9 in
0.273 in
x
O
0.170 in
3 in
1.410 in
R1 = 366 lbf
I
I = 1.66 in , c = 1.10 in3
4
R2 = 122 lbf
(a)
y
dy
dA
488 lbf
x
O
a
b
y
1.227 in
366 lbf
122 lbf
366 lbf
O
122 lbf
1098 lbf in
(c )
O
(b )
Solution
The loading, shear-force, and bending-moment diagrams are shown in Fig. 319b. If
the direct shear force is included in the analysis, the maximum stresses at the top and
bottom of the beam will be the same as if only bending were considered. The maximum
bending stresses are
σ =±
Mc
1098(1.5)
=±
= ± 992 psi
I
1.66
However, the maximum stress due to the combined bending and direct shear
stresses may be maximum at the point (3, 1.227) that is just to the left of the applied
load, where the web joins the ange. To simplify the calculations we assume a cross
section with square corners (Fig. 319c). The normal stress at section ab, with x = 3
in, is
σ =
My
1098(1.227)
=
= 812 psi
I
1.66
For the shear stress at section ab, considering the area above ab and using Eq. (330) gives
Q = y A = 1.227 +
¯
0.273
( 1.410)(0.273) = 0.525 in3
2
97
98
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
93
Using Eq. (331) with V = 366 lbf, I = 1.66 in4 , Q = 0.525 in3 , and b = 0.170 in
yields
τx y =
366(0.525)
VQ
=
= 681 psi
Ib
1.66(0.170)
The negative sign comes from recognizing that the shear stress is down on an x face of
a dx dy element at the location being considered.
The principal stresses at the point can now be determined. Using Eq. (313), we
nd that at x = 3 in, y = 1.227 in,
σ1 , σ2 =
=
σx + σ y
±
2
812 + 0
±
2
σx σ y
2
2
2
+ τx y
812 0
2
2
+ ( 681) 2 = 387, 1200 psi
For a point at x = 3 in, y = 1.227 in, the principal stresses are σ1 , σ2 = 1200,
387 psi. Thus we see that the maximum principal stresses are ±1200 psi, 21 percent
higher than thought by the designer.
Shear Stresses in Standard-Section Beams
The shear stress distribution in a beam depends on how Q /b varies as a function of
y1. Here we will show how to determine the shear stress distribution for a beam with
a rectangular cross section and provide results of maximum values of shear stress for
other standard cross sections. Figure 320 shows a portion of a beam with a rectangular cross section, subjected to a shear force V and a bending moment M. As a
result of the bending moment, a normal stress σ is developed on a cross section such
as A-A, which is in compression above the neutral axis and in tension below. To
investigate the shear stress at a distance y1 above the neutral axis, we select an
element of area d A at a distance y above the neutral axis. Then, d A = b dy , and so
Eq. (330) becomes
c
Q=
y1
c
ydA = b
y1
y dy =
by 2
2
c
y1
=
b2
2
c y1
2
(a)
Substituting this value for Q into Eq. (331) gives
V2
2
τ=
c y1
(332)
2I
This is the general equation for shear stress in a rectangular beam. To learn something about it, let us make some substitutions. From Table A18, the second moment
of area for a rectangular section is I = bh 3 /12; substituting h = 2c and A =
bh = 2bc gives
I=
Ac2
3
(b)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
94
I. Basics
99
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 320
y
y
y
A
Shear stresses in a rectangular
beam.
dy
b
dA
max
=
3V
2A
V
M
c
yy
1
x
O
z
x
h
O
A
(a)
(b )
(c)
y
c
y1
x
(d )
If we now use this value of I for Eq. (332) and rearrange, we get
τ=
3V
2A
1
2
y1
c2
(333)
We note that the maximum shear stress exists when y1 = 0, which is at the bending neutral axis. Thus
τmax =
3V
2A
(334)
for a rectangular section. As we move away from the neutral axis, the shear stress
decreases parabolically until it is zero at the outer surfaces where y1 = ±c , as shown
in Fig. 320c. It is particularly interesting and significant here to observe that the
shear stress is maximum at the bending neutral axis, where the normal stress due to
bending is zero, and that the shear stress is zero at the outer surfaces, where the
bending stress is a maximum. Horizontal shear stress is always accompanied by
vertical shear stress of the same magnitude, and so the distribution can be diagrammed as shown in Fig. 320d. Figure 320c shows that the shear τ on the vertical surfaces varies with y. We are almost always interested in the horizontal shear, τ
in Fig. 320d, which is nearly uniform with constant y. The maximum horizontal
shear occurs where the vertical shear is largest. This is usually at the neutral axis but
may not be if the width b is smaller somewhere else. Furthermore, if the section is
such that b can be minimized on a plane not horizontal, then the horizontal shear
stress occurs on an inclined plane. For example, with tubing, the horizontal shear
stress occurs on a radial plane and the corresponding vertical shear is not vertical,
but tangential.
Formulas for the maximum exural shear stress for the most commonly used
shapes are listed in Table 32.
100
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
95
Load and Stress Analysis
Table 32
Beam Shape
Formula
Formulas for Maximum
Shear Stress Due to
Bending
τmax =
Beam Shape
3V
2A
Formula
τmax =
2V
A
τmax =
V
A web
Rectangular
Hollow, thin-walled round
τmax =
4V
3A
Web
Structural I beam (thin-walled)
Circular
312
Torsion
Any moment vector that is collinear with an axis of a mechanical element is called a
torque vector, because the moment causes the element to be twisted about that axis. A
bar subjected to such a moment is also said to be in torsion.
As shown in Fig. 3 21, the torque T applied to a bar can be designated by drawing
arrows on the surface of the bar to indicate direction or by drawing torque-vector arrows
along the axes of twist of the bar. Torque vectors are the hollow arrows shown on the
x axis in Fig. 321. Note that they conform to the right-hand rule for vectors.
The angle of twist, in radians, for a solid round bar is
Tl
θ=
(335)
GJ
where
T = torque
l = length
G = modulus of rigidity
J = polar second moment of area
Figure 321
T
l
A
y
dx
B
T
C
r
B'
O
C'
z
x
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
96
I. Basics
101
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Shear stresses develop throughout the cross section. For a round bar in torsion,
these stresses are proportional to the radius ρ and are given by
τ=
Tρ
J
(336)
Designating r as the radius to the outer surface, we have
τmax =
Tr
J
(337)
The assumptions used in the analysis are:
The bar is acted upon by a pure torque, and the sections under consideration are
remote from the point of application of the load and from a change in diameter.
Adjacent cross sections originally plane and parallel remain plane and parallel after
twisting, and any radial line remains straight.
The material obeys Hookes law.
Equation (337) applies only to circular sections. For a solid round section,
J=
π d4
32
(338)
where d is the diameter of the bar. For a hollow round section,
J=
π4
d di4
32 o
(339)
where the subscripts o and i refer to the outside and inside diameters, respectively.
In using Eq. (337) it is often necessary to obtain the torque T from a consideration of the power and speed of a rotating shaft. For convenience when U. S. Customary
units are used, three forms of this relation are
H=
where
2π T n
Tn
FV
=
=
33 000
33 000(12)
63 025
(340)
H = power, hp
T = torque, lbf · in
n = shaft speed, rev/min
F = force, lbf
V = velocity, ft/min
When SI units are used, the equation is
where
H = power, W
T = torque, N · m
ω = angular velocity, rad/s
H = Tω
(341)
102
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
97
Load and Stress Analysis
The torque T corresponding to the power in watts is given approximately by
T = 9.55
H
n
(342)
where n is in revolutions per minute.
There are some applications in machinery for noncircular-cross-section members
and shafts where a regular polygonal cross section is useful in transmitting torque to a
gear or pulley that can have an axial change in position. Because no key or keyway is
needed, the possibility of a lost key is avoided. Saint Venant (1855) showed that the
maximum shearing stress in a rectangular b × c section bar occurs in the middle of the
longest side b and is of the magnitude
τmax =
T .T
=2
α bc2
bc
3+
1.8
b/c
(343)
where b is the longer side, c the shorter side, and α a factor that is a function of the ratio
b/c as shown in the following table.4 The angle of twist is given by
θ=
Tl
β bc3 G
(344)
where β is a function of b/c , as shown in the table.
b/c
1.00
1.50
1.75
2.00
2.50
3.00
4.00
6.00
8.00
10
α
0.208
0.231
0.239
0.246
0.258
0.267
0.282
0.299
0.307
0.313
0.333
β
0.141
0.196
0.214
0.228
0.249
0.263
0.281
0.299
0.307
0.313
0.333
In Eqs. (343) and (344) b and c are the width (long side) and thickness (short side)
of the bar, respectively. They cannot be interchanged. Equation (343) is also approximately valid for equal-sided angles; these can be considered as two rectangles, each of
which is capable of carrying half the torque.5
4
S. Timoshenko, Strength of Materials, Part I, 3rd ed., D. Van Nostrand Company, New York, 1955, p. 290.
5
For other sections see W. C. Young and R. G. Budynas, Roarks Formulas for Stress and Strain, 7th ed.,
McGraw-Hill, New York, 2002.
EXAMPLE 38
Figure 322 shows a crank loaded by a force F = 300 lbf that causes twisting and
bending of a 3 -in-diameter shaft xed to a support at the origin of the reference system.
4
In actuality, the support may be an inertia that we wish to rotate, but for the purposes
of a stress analysis we can consider this a statics problem.
(a) Draw separate free-body diagrams of the shaft A B and the arm B C , and compute the values of all forces, moments, and torques that act. Label the directions of the
coordinate axes on these diagrams.
(b) Compute the maxima of the torsional stress and the bending stress in the arm
BC and indicate where these act.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
98
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
y
Figure 322
1.5 in
F
C
A
3
4
in dia.
1
4
1
2
in
1
14
B
in dia.
in
z
4 in
5 in
x
(c) Locate a stress element on the top surface of the shaft at A, and calculate all the
stress components that act upon this element.
(d ) Determine the maximum normal and shear stresses at A.
Solution
(a) The two free-body diagrams are shown in Fig. 323. The results are
F = 300j lbf, TC = 450k lbf · in
F = 300j lbf, M1 = 1200i lbf · in, T1 = 450k lbf · in
F = 300j lbf, T2 = 1200i lbf · in, M2 = 450k lbf · in
F = 300j lbf, MA = 1950k lbf · in, TA = 1200i lbf · in
At end C of arm BC:
At end B of arm BC:
At end B of shaft AB:
At end A of shaft AB:
Figure 323
F
y
TC
C
4 in
B
M1
F
T1
x
z
y
MA
TA
A
z
5 in
F
F
M2
B
T2
x
103
104
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
99
(b) For arm BC , the bending moment will reach a maximum near the shaft at
B. If we assume this is 1200 lbf · in, then the bending stress for a rectangular section will be
Answer
σ=
M
6(1200)
6M
= 18 400 psi
= 2=
I /c
bh
0.25(1.25)2
Of course, this is not exactly correct, because at B the moment is actually being transferred into the shaft, probably through a weldment.
For the torsional stress, use Eq. (343). Thus
Answer
τmax =
T
bc2
3+
1.8
b/c
=
450
1.8
3+
1.25(0.252 )
1.25/0.25
= 19 400 psi
This stress occurs at the middle of the 1 1 -in side.
4
(c) For a stress element at A, the bending stress is tensile and is
Answer
σx =
M
32(1950)
32 M
=
= 47 100 psi
=
3
I /c
πd
π(0.75)3
The torsional stress is
Answer
τx z =
T
16(1200)
16T
=
= 14 500 psi
=
J /c
π d3
π(0.75)3
where the reader should verify that the negative sign accounts for the direction of τx z .
(d ) Point A is in a state of plane stress where the stresses are in the x z plane. Thus
the principal stresses are given by Eq. (313) with subscripts corresponding to the
x, z axes.
Answer
The maximum normal stress is then given by
σ1 =
=
Answer
σ x + σz
+
2
47.1 + 0
+
2
σ x σz
2
2
47.1 0
2
2
+ τx z
2
+ (14.5)2 = 51.2 kpsi
The maximum shear stress at A occurs on surfaces different than the surfaces containing the principal stresses or the surfaces containing the bending and torsional shear
stresses. The maximum shear stress is given by Eq. (314), again with modied subscripts, and is given by
τ1 =
σ x σz
2
2
2
+ τx z =
47.1 0
2
2
+ (14.5)2 = 27.7 kpsi
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
100
I. Basics
105
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
EXAMPLE 39
The 1.5-in-diameter solid steel shaft shown in Fig. 324a is simply supported at the ends.
Two pulleys are keyed to the shaft where pulley B is of diameter 4.0 in and pulley C is of
diameter 8.0 in. Considering bending and torsional stresses only, determine the locations
and magnitudes of the greatest tensile, compressive, and shear stresses in the shaft.
Solution
Figure 324b shows the net forces, reactions, and torsional moments on the shaft.
Although this is a three-dimensional problem and vectors might seem appropriate, we
will look at the components of the moment vector by performing a two-plane analysis.
Figure 324c shows the loading in the x y plane, as viewed down the z axis, where bending moments are actually vectors in the z direction. Thus we label the moment diagram
as Mz versus x. For the x z plane, we look down the y axis, and the moment diagram is
M y versus x as shown in Fig. 324d.
The net moment on a section is the vector sum of the components. That is,
M=
2
2
M y + Mz
(1)
At point B,
MB =
At point C,
20002 + 80002 = 8246 lbf · in
MC = 40002 + 40002 = 5657 lbf · in
Thus the maximum bending moment is 8246 lbf · in and the maximum bending stress
at pulley B is
σ=
32(8246)
M d /2
32 M
=
=
= 24 890 psi
π d 4 /64
π d3
π(1.53 )
The maximum torsional shear stress occurs between B and C and is
16T
16(1600)
T d /2
=
= 2414 psi
=
4 /32
3
πd
πd
π(1.53 )
τ=
The maximum bending and torsional shear stresses occur just to the right of pulley
B at points E and F as shown in Fig. 324e. At point E, the maximum tensile stress will
be σ1 given by
Answer
σ1 =
σ
+
2
2
σ
2
+ τ2 =
24 890
+
2
24 890
2
2
+ 24142 = 25 120 psi
At point F, the maximum compressive stress will be σ2 given by
Answer
σ2 =
σ
2
σ
2
2
+ τ2 =
24 890
2
24 890
2
2
+ 24142 = 25 120 psi
The extreme shear stress also occurs at E and F and is
Answer
τ1 =
±σ
2
2
+ τ2 =
±24 890
2
2
+ 24142 = 12 680 psi
106
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
101
Load and Stress Analysis
y
A
10 in
z
B
200 lbf
10 in
C
10 in
1000 lbf
D
x
100 lbf
500 lbf
(a)
y
800 lbf
A
10 in
z
1600 lbf in
B
200 lbf
10 in
1600 lbf in
1200 lbf
10 in
C
D
600 lbf
x
400 lbf
400 lbf
(b)
1200 lbf
y
600 lbf
A
A
C
B
D
C
B
D
400 lbf
800 lbf
200 lbf
Mz
(lbf in)
z
400 lbf
4000
8000
My
(lbf in)
2000
4000
x
O
x
O
(d )
(c)
Location: at B (x = 10+ )
8000 lbf in
8246 lbf in
F
Max. compression
and shear
2000 lbf in
= tan1 8000 = 76
2000
E
Max. tension
and shear
Figure 324
x
x
(e)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
102
I. Basics
107
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 325
ds
The depicted cross section is
elliptical, but the section need
not be symmetrical nor of
constant thickness.
r
t
dAm = 1 rds
2
Closed Thin-Walled Tubes ( t « r) 6
In closed thin-walled tubes, it can be shown that the product of shear stress times thickness
of the wall τ t is constant, meaning that the shear stress τ is inversely proportional to the
wall thickness t. The total torque T on a tube such as depicted in Fig. 325 is given by
T=
τ tr ds = (τ t )
r ds = τ t (2 Am ) = 2 Am t τ
where Am is the area enclosed by the section median line. Solving for τ gives
τ=
T
2 Am t
(345)
For constant wall thickness t, the angular twist (radians) per unit of length of the tube
θ1 is given by
θ1 =
T Lm
4G A2 t
m
(346)
where L m is the perimeter of the section median line. These equations presume the
buckling of the tube is prevented by ribs, stiffeners, bulkheads, and so on, and that the
stresses are below the proportional limit.
6
See Sec. 313, F. P. Beer, E. R. Johnston, and J. T. De Wolf, Mechanics of Materials, 4th ed., McGraw-Hill,
New York, 2006.
EXAMPLE 310
Solution
A welded steel tube is 40 in long, has a 1 -in wall thickness, and a 2.5-in by 3.6-in
8
rectangular cross section as shown in Fig. 326. Assume an allowable shear stress of
11 500 psi and a shear modulus of 11.5(106) psi.
(a) Estimate the allowable torque T.
(b) Estimate the angle of twist due to the torque.
(a) Within the section median line, the area enclosed is
Am = (2.5 0.125)(3.6 0.125) = 8.253 in2
and the length of the median perimeter is
L m = 2[(2.5 0.125) + (3.6 0.125)] = 11.70 in
108
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
103
Figure 326
A rectangular steel tube
produced by welding.
1
8
in
40 in
2.5 in
3.6 in
Answer
From Eq. (345) the torque T is
T = 2 Am t τ = 2(8.253)0.125(11 500) = 23 730 lbf · in
Answer
(b) The angle of twist θ from Eq. (346) is
θ = θ1l =
EXAMPLE 311
Solution
23 730(11.70)
T Lm
l=
(40) = 0.0284 rad = 1.62
4G A2 t
4(11.5 × 106 )(8.2532 )(0.125)
m
Compare the shear stress on a circular cylindrical tube with an outside diameter of 1 in
and an inside diameter of 0.9 in, predicted by Eq. (337), to that estimated by
Eq. (345).
From Eq. (337),
τmax =
T (0.5)
Tr
Tr
=
=
= 14.809T
4
J
(π/32)(14 0.94 )
(π/32) do di4
From Eq. (345),
τ=
T
T
=
= 14.108T
2 Am t
2(π 0.952 /4)0.05
Taking Eq. (337) as correct, the error in the thin-wall estimate is 4.7 percent.
Open Thin-Walled Sections
When the median wall line is not closed, it is said to be open. Figure 327 presents
some examples. Open sections in torsion, where the wall is thin, have relations derived
from the membrane analogy theory7 resulting in:
τ = G θ1 c =
7
3T
Lc2
(347)
See S. P. Timoshenko and J. N. Goodier, Theory of Elasticity, 3rd ed., McGraw-Hill, New York, 1970, Sec.109.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
104
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 327
c
Some open thin-wall sections.
L
where τ is the shear stress, G is the shear modulus, θ1 is the angle of twist per unit
length, T is torque, and L is the length of the median line. The wall thickness is
designated c (rather than t) to remind you that you are in open sections. By studying the table that follows Eq. (3 44) you will discover that membrane theory presumes b/c . Note that open thin-walled sections in torsion should be avoided
in design. As indicated in Eq. (3 47), the shear stress and the angle of twist are
inversely proportional to c2 and c3 , respectively. Thus, for small wall thickness,
stress and twist can become quite large. For example, consider the thin round tube
with a slit in Fig. 327. For a ratio of wall thickness of outside diameter of
c/do = 0.1, the open section has greater magnitudes of stress and angle of twist by
factors of 12.3 and 61.5, respectively, compared to a closed section of the same
dimensions.
EXAMPLE 312
Solution
A 12-in-long strip of steel is 1 in thick and 1 in wide, as shown in Fig. 328. If the
8
allowable shear stress is 11 500 psi and the shear modulus is 11.5(106) psi, nd the
torque corresponding to the allowable shear stress and the angle of twist, in degrees,
(a) using Eq. (347) and (b) using Eqs. (343) and (344).
(a) The length of the median line is 1 in. From Eq. (347),
(1)(1/8)2 11 500
Lc2 τ
=
= 59.90 lbf · in
3
3
11500(12)
τl
=
= 0.0960 rad = 5.5°
θ = θ1l =
Gc
11.5(106 )(1/8)
T=
A torsional spring rate kt can be expressed as T /θ :
T
kt = 59.90/0.0960 = 624 lbf · in/rad
1 in
(b) From Eq. (343),
T=
1
8
in
Figure 328
The cross-section of a thin strip
of steel subjected to a
torsional moment T.
11 500(1)(0.125)2
τmax bc2
=
= 55.72 lbf · in
3 + 1.8/(b/c)
3 + 1.8/(1/0.125)
From Eq. (344), with b/c = 1/0.125 = 8,
θ=
55.72(12)
Tl
=
= 0.0970 rad = 5.6°
β bc3 G
0.307(1)0.1253 (11.5)106
kt = 55.72/0.0970 = 574 lbf · in/rad
109
110
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
105
Load and Stress Analysis
313
Stress Concentration
In the development of the basic stress equations for tension, compression, bending, and
torsion, it was assumed that no geometric irregularities occurred in the member under
consideration. But it is quite difcult to design a machine without permitting some
changes in the cross sections of the members. Rotating shafts must have shoulders
designed on them so that the bearings can be properly seated and so that they will take
thrust loads; and the shafts must have key slots machined into them for securing pulleys and gears. A bolt has a head on one end and screw threads on the other end, both
of which account for abrupt changes in the cross section. Other parts require holes, oil
grooves, and notches of various kinds. Any discontinuity in a machine part alters the
stress distribution in the neighborhood of the discontinuity so that the elementary stress
equations no longer describe the state of stress in the part at these locations. Such discontinuities are called stress raisers, and the regions in which they occur are called
areas of stress concentration.
The distribution of elastic stress across a section of a member may be uniform as
in a bar in tension, linear as a beam in bending, or even rapid and curvaceous as in a
sharply curved beam. Stress concentrations can arise from some irregularity not inherent in the member, such as tool marks, holes, notches, grooves, or threads. The nominal stress is said to exist if the member is free of the stress raiser. This denition is not
always honored, so check the denition on the stress-concentration chart or table you
are using.
A theoretical, or geometric, stress-concentration factor Kt or Kts is used to relate
the actual maximum stress at the discontinuity to the nominal stress. The factors are
dened by the equations
Kt =
σmax
σ0
K ts =
τmax
τ0
(348)
where Kt is used for normal stresses and Kts for shear stresses. The nominal stress σ0 or
τ0 is more difcult to dene. Generally, it is the stress calculated by using the elementary stress equations and the net area, or net cross section. But sometimes the gross
cross section is used instead, and so it is always wise to double check your source of Kt
or Kts before calculating the maximum stress.
The subscript t in Kt means that this stress-concentration factor depends for its
value only on the geometry of the part. That is, the particular material used has no effect
on the value of Kt. This is why it is called a theoretical stress-concentration factor.
The analysis of geometric shapes to determine stress-concentration factors is a difcult problem, and not many solutions can be found. Most stress-concentration factors
are found by using experimental techniques.8 Though the nite-element method has
been used, the fact that the elements are indeed nite prevents nding the true maximum stress. Experimental approaches generally used include photoelasticity, grid
methods, brittle-coating methods, and electrical strain-gauge methods. Of course, the
grid and strain-gauge methods both suffer from the same drawback as the nite-element
method.
Stress-concentration factors for a variety of geometries may be found in
Tables A15 and A16.
8
The best source book is W. D. Pilkey, Petersons Stress Concentration Factors, 2nd ed., John Wiley &
Sons, New York, 1997.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
106
I. Basics
111
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 329
Thin plate in tension or simple
compression with a transverse
central hole. The net tensile
force is F = σ wt, where t is
the thickness of the plate. The
nominal stress is given by
F
w
σ
σ0 =
=
(w d )t
(w d )
3.0
d
2.8
w
2.6
Kt
2.4
2.2
2.0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
d /w
An example is shown in Fig. 329, that of a thin plate loaded in tension where the
plate contains a centrally located hole.
In static loading, stress-concentration factors are applied as follows. In ductile
(ǫ f 0.05) materials, the stress-concentration factor is not usually applied to predict the
critical stress, because plastic strain in the region of the stress is localized and
has a strengthening effect. In brittle materials (ǫ f < 0.05), the geometric stressconcentration factor Kt is applied to the nominal stress before comparing it with strength.
Gray cast iron has so many inherent stress raisers that the stress raisers introduced by the
designer have only a modest (but additive) effect.
EXAMPLE 313
Be Alert to Viewpoint
On a spade rod end (or lug) a load is transferred through a pin to a rectangular-crosssection rod or strap. The theoretical or geometric stress-concentration factor for this
geometry is known as follows, on the basis of the net area A = (w d )t as shown in
Fig. 330.
d/w
Kt
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
7.4
5.4
4.6
3.7
3.2
2.8
2.6
2.45
As presented in the table, Kt is a decreasing monotone. This rod end is similar to the
square-ended lug depicted in Fig. A15-12 of appendix A.
σmax = K t σ0
Kt F
F
= Kt
σmax =
A
(w d )t
(a)
(b)
It is insightful to base the stress concentration factor on the unnotched area, wt . Let
F
σmax = K t
(c)
wt
By equating Eqs. (b) and (c) and solving for K t we obtain
K t =
Kt
wt
F
Kt
=
F
(w d )t
1 d /w
(d )
112
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
107
Load and Stress Analysis
A power regression curve-t for the data in the above table in the form K t = a (d /w)b
gives the result a = exp(0.204 521 2) = 1.227, b = 0.935, and r 2 = 0.9947. Thus
F
d
A
B
K t = 1.227
0.935
d
w
(e)
which is a decreasing monotone (and unexciting). However, from Eq. (d ),
w
t
K t =
Form another table from Eq. ( f ):
F
d/w
Figure 330
K t
A round-ended lug end to a
rectangular cross-section rod.
The maximum tensile stress in
the lug occurs at locations A
and B. The net area
A = ( w d) t is used in the
denition of K t , but there is an
advantage to using the total
area wt.
0.15
0.20
0.25
1.227
1 d /w
0.30
0.35
d
w
0.935
0.40
0.45
(f )
0.50
0.55
0.60
8.507 6.907 5.980 5.403 5.038 4.817 4.707 4.692 4.769 4.946
which shows a stationary-point minimum for K t . This can be found by differentiating
Eq. ( f ) with respect to d /w and setting it equal to zero:
(1 d /w)ab(d /w)b1 + a (d /w)b
d K t
=
=0
d (d /w)
[1 (d /w)]2
where b = 0.935, from which
d
w
=
b
0.935
=
= 0.483
b1
0.935 1
with a corresponding K t of 4.687. Knowing the section w × t lets the designer specify the
strongest lug immediately by specifying a pin diameter of 0.483w (or, as a rule of thumb,
of half the width). The theoretical K t data in the original form, or a plot based on the data
using net area, would not suggest this. The right viewpoint can suggest valuable insights.
314
po
dr
r
pi
ri
ro
Stresses in Pressurized Cylinders
Cylindrical pressure vessels, hydraulic cylinders, gun barrels, and pipes carrying uids
at high pressures develop both radial and tangential stresses with values that depend
upon the radius of the element under consideration. In determining the radial stress σr
and the tangential stress σt , we make use of the assumption that the longitudinal
elongation is constant around the circumference of the cylinder. In other words, a right
section of the cylinder remains plane after stressing.
Referring to Fig. 331, we designate the inside radius of the cylinder by ri, the outside radius by ro, the internal pressure by pi, and the external pressure by po. Then it can
be shown that tangential and radial stresses exist whose magnitudes are9
σt =
Figure 331
2
2
pi ri2 po ro ri2ro ( po pi )/ r 2
2 r2
ro
i
2
2
pi ri2 po ro + ri2ro ( po pi )/ r 2
σr =
2
ro ri2
A cylinder subjected to both
internal and external pressure.
9
(349)
See Richard G. Budynas, Advanced Strength and Applied Stress Analysis, 2nd ed., McGraw-Hill, New
York, 1999, pp. 348352.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
108
I. Basics
113
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 332
po = 0
po = 0
t
Distribution of stresses in a
thick-walled cylinder subjected
to internal pressure.
ro
ri
pi
ri
pi
r
ro
(a) Tangential stress
distribution
(b) Radial stress
distribution
As usual, positive values indicate tension and negative values, compression.
The special case of po = 0 gives
σt =
ri2 pi
2
ro ri2
r 2 pi
σr = 2 i 2
r o ri
1+
2
ro
r2
r2
1 o
r2
(350)
The equations of set (350) are plotted in Fig. 332 to show the distribution of stresses
over the wall thickness. It should be realized that longitudinal stresses exist when the
end reactions to the internal pressure are taken by the pressure vessel itself. This stress
is found to be
σl =
pi ri2
2
ro ri2
(351)
We further note that Eqs. (349), (350), and (351) apply only to sections taken a signicant distance from the ends and away from any areas of stress concentration.
Thin-Walled Vessels
When the wall thickness of a cylindrical pressure vessel is about one-twentieth, or less,
of its radius, the radial stress that results from pressurizing the vessel is quite small
compared with the tangential stress. Under these conditions the tangential stress can be
obtained as follows: Let an internal pressure p be exerted on the wall of a cylinder of
thickness t and inside diameter di. The force tending to separate two halves of a unit
length of the cylinder is pdi . This force is resisted by the tangential stress, also called
the hoop stress, acting uniformly over the stressed area. We then have pdi = 2t σt , or
pdi
(σt )av =
(352)
2t
This equation gives the average tangential stress and is valid regardless of the wall thickness. For a thin-walled vessel an approximation to the maximum tangential stress is
p(di + t )
(σt )max =
(353)
2t
where di + t is the average diameter.
114
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
109
In a closed cylinder, the longitudinal stress σl exists because of the pressure upon
the ends of the vessel. If we assume this stress is also distributed uniformly over the
wall thickness, we can easily nd it to be
pdi
σl =
(354)
4t
EXAMPLE 314
An aluminum-alloy pressure vessel is made of tubing having an outside diameter of 8 in
and a wall thickness of 1 in.
4
(a) What pressure can the cylinder carry if the permissible tangential stress is
12 kpsi and the theory for thin-walled vessels is assumed to apply?
(b) On the basis of the pressure found in part (a), compute all of the stress components using the theory for thick-walled cylinders.
Solution
(a) Here di = 8 2(0.25) = 7.5 in, ri = 7.5/2 = 3.75 in, and ro = 8/2 = 4 in. Then
1
t / ri = 0.25/3.75 = 0.067. Since this ratio is greater than 20 , the theory for thin-walled
vessels may not yield safe results.
We rst solve Eq. (353) to obtain the allowable pressure. This gives
Answer
p=
2(0.25)(12)(10)3
2t (σt )max
=
= 774 psi
di + t
7.5 + 0.25
Then, from Eq. (354), we nd the average longitudinal stress to be
σl =
774(7.5)
pdi
=
= 5810 psi
4t
4(0.25)
(b) The maximum tangential stress will occur at the inside radius, and so we use
r = ri in the rst equation of Eq. (350). This gives
Answer
(σt )max =
ri2 pi
2
ro ri2
2
ro
ri2
1+
= pi
2
ro + ri2
42 + 3.752
= 774 2
= 12 000 psi
2
4 3.752
ro ri2
Similarly, the maximum radial stress is found, from the second equation of Eq. (350)
to be
Answer
σr = pi = 774 psi
Equation (351) gives the longitudinal stress as
Answer
σl =
pi ri2
774(3.75)2
=2
= 5620 psi
2
4 3.752
ri
2
ro
These three stresses, σt , σr , and σl , are principal stresses, since there is no shear on
these surfaces. Note that there is no signicant difference in the tangential stresses in
parts (a) and (b), and so the thin-wall theory can be considered satisfactory.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
110
I. Basics
115
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
315
Stresses in Rotating Rings
Many rotating elements, such as ywheels and blowers, can be simplied to a rotating
ring to determine the stresses. When this is done it is found that the same tangential and
radial stresses exist as in the theory for thick-walled cylinders except that they are
caused by inertial forces acting on all the particles of the ring. The tangential and radial stresses so found are subject to the following restrictions:
The outside radius of the ring, or disk, is large compared with the thickness ro 10t .
The thickness of the ring or disk is constant.
The stresses are constant over the thickness.
The stresses are10
2
ri2ro
1 + 3ν 2
r
2
r
3+ν
σt = ρω2
3+ν
8
2
ri2 + ro +
σr = ρω2
3+ν
8
r 2r 2
2
ri2 + ro i 2o r 2
r
(355)
where r is the radius to the stress element under consideration, ρ is the mass density,
and ω is the angular velocity of the ring in radians per second. For a rotating disk, use
ri = 0 in these equations.
316
Press and Shrink Fits
When two cylindrical parts are assembled by shrinking or press tting one part upon
another, a contact pressure is created between the two parts. The stresses resulting from
this pressure may easily be determined with the equations of the preceding sections.
Figure 333 shows two cylindrical members that have been assembled with a shrink
t. Prior to assembly, the outer radius of the inner member was larger than the inner radius
of the outer member by the radial interference δ . After assembly, an interference contact
pressure p develops between the members at the nominal radius R, causing radial stresses σr = p in each member at the contacting surfaces. This pressure is given by11
p=
δ
1
R
Eo
2
ro
2
ro
2
+R
1
+ νo +
2
R
Ei
R 2 + ri2
νi
R 2 ri2
(356)
where the subscripts o and i on the material properties correspond to the outer and inner
members, respectively. If the two members are of the same material with
E o = E i = E , νo = vi , the relation simplies to
p=
2
E δ (ro R 2 )( R 2 ri2 )
2
2 R3
ro ri2
(357)
For Eqs. (356) or (357), diameters can be used in place of R, ri , and ro , provided δ is
the diametral interference (twice the radial interference).
10
Ibid, pp. 348357.
11
Ibid, pp. 348354.
116
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
111
Figure 333
Notation for press and shrink
ts. (a) Unassembled parts;
(b) after assembly.
ro
R
ri
(a )
(b )
With p, Eq. (349) can be used to determine the radial and tangential stresses in
each member. For the inner member, po = p and pi = 0, For the outer member, po = 0
and pi = p . For example, the magnitudes of the tangential stresses at the transition
radius R are maximum for both members. For the inner member
(σt )i
r=R
= p
R 2 + ri2
R 2 ri2
(358)
and, for the outer member
(σt )o
r=R
=p
2
ro + R 2
2
ro R 2
(359)
Assumptions
It is assumed that both members have the same length. In the case of a hub that has been
press-tted onto a shaft, this assumption would not be true, and there would be an increased
pressure at each end of the hub. It is customary to allow for this condition by employing a
stress-concentration factor. The value of this factor depends upon the contact pressure and
the design of the female member, but its theoretical value is seldom greater than 2.
317
Temperature Effects
When the temperature of an unrestrained body is uniformly increased, the body
expands, and the normal strain is
ǫx = ǫ y = ǫz = α( T )
(360)
where α is the coefcient of thermal expansion and T is the temperature change, in
degrees. In this action the body experiences a simple volume increase with the components of shear strain all zero.
If a straight bar is restrained at the ends so as to prevent lengthwise expansion and
then is subjected to a uniform increase in temperature, a compressive stress will develop
because of the axial constraint. The stress is
σ = ǫ E = α( T ) E
(361)
In a similar manner, if a uniform at plate is restrained at the edges and also subjected
to a uniform temperature rise, the compressive stress developed is given by the equation
σ =
α( T ) E
1ν
(362)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
112
I. Basics
117
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Table 33
Material
Coefcients of Thermal
Expansion (Linear Mean
Coefcients for the
Temperature Range 0 100°C)
Celsius Scale (°C
1
Fahrenheit Scale (°F1)
)
Aluminum
23.9(10)6
13.3(10)6
Brass, cast
18.7(10)6
10.4(10)6
Carbon steel
6
10.8(10)
6.0(10)6
Cast iron
10.6(10)6
5.9(10)6
Magnesium
25.2(10)
6
14.0(10)6
Nickel steel
13.1(10)6
7.3(10)6
6
9.6(10)6
4.3(10)6
2.4(10)6
Stainless steel
17.3(10)
Tungsten
The stresses expressed by Eqs. (361) and (362) are called thermal stresses. They
arise because of a temperature change in a clamped or restrained member. Such stresses, for example, occur during welding, since parts to be welded must be clamped before
welding. Table 33 lists approximate values of the coefcients of thermal expansion.
318
Curved Beams in Bending
The distribution of stress in a curved exural member is determined by using the
following assumptions:
The cross section has an axis of symmetry in a plane along the length of the beam.
Plane cross sections remain plane after bending.
The modulus of elasticity is the same in tension as in compression.
We shall nd that the neutral axis and the centroidal axis of a curved beam, unlike
the axes of a straight beam, are not coincident and also that the stress does not vary linearly from the neutral axis. The notation shown in Fig. 334 is dened as follows:
ro = radius of outer ber
ri = radius of inner ber
Figure 334
a
Note that y is positive in the
direction toward the center of
curvature, point O.
Centroidal
axis
b' b
co
h
e
y
y
ci
M
d
M
c
c'
Neutral axis
ro
rn
r
d
ri
rn
O
O
rc
118
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
113
h = depth of section
co = distance from neutral axis to outer ber
ci = distance from neutral axis to inner ber
rn = radius of neutral axis
rc = radius of centroidal axis
e = distance from centroidal axis to neutral axis
M = bending moment; positive M decreases curvature
Figure 334 shows that the neutral and centroidal axes are not coincident.12 It turns out
that the location of the neutral axis with respect to the center of curvature O is given by
the equation
A
dA
r
rn =
(363)
The stress distribution can be found by balancing the external applied moment against
the internal resisting moment. The result is found to be
σ=
My
Ae(rn y )
(364)
where M is positive in the direction shown in Fig. 334. Equation (363) shows that the
stress distribution is hyperbolic. The critical stresses occur at the inner and outer surfaces where y = ci and y = co , respectively, and are
σi =
Mci
Aeri
σo =
Mco
Aero
(365)
These equations are valid for pure bending. In the usual and more general case, such as
a crane hook, the U frame of a press, or the frame of a clamp, the bending moment is
due to forces acting to one side of the cross section under consideration. In this case the
bending moment is computed about the centroidal axis, not the neutral axis. Also, an
additional axial tensile or compressive stress must be added to the bending stresses
given by Eqs. (364) and (365) to obtain the resultant stresses acting on the section.
12
For a complete development of the relations in this section, see Richard G. Budynas, Advanced Strength
and Applied Stress Analysis, 2nd ed., Mcgraw-Hill, New York, 1999, pp. 309317.
EXAMPLE 315
Solution
Plot the distribution of stresses across section A-A of the crane hook shown in
Fig.335a. The cross section is rectangular, with b = 0.75 in and h = 4 in, and the load
is F = 5000 lbf.
Since A = bh , we have d A = b dr and, from Eq. (363),
bh
A
h
= ro
rn =
=r
o
dA
b
ln
dr
ri
r
r
r
i
(1)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
114
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
From Fig. 335b, we see that ri = 2 in, ro = 6 in, rc = 4 in, and A = 3 in2. Thus, from
Eq. (1),
4
h
rn =
= 6 = 3.641 in
ln(ro / ri )
ln 2
and so the eccentricity is e = rc rn = 4 3.641 = 0.359 in. The moment M is positive and is M = Frc = 5000(4) = 20 000 lbf · in. Adding the axial component of stress
to Eq. (364) gives
F
My
5000 (20 000)(3.641 r )
σ=
+
=
+
(2)
A
Ae(rn y )
3
3(0.359)r
Substituting values of r from 2 to 6 in results in the stress distribution shown in
Fig. 335c. The stresses at the inner and outer radii are found to be 16.9 and 5.63 kpsi,
respectively, as shown.
Figure 335
3/4 in
(a) Plan view of crane hook;
(b) cross section and notation;
(c) resulting stress distribution.
There is no stress concentration.
Section A-A
rc
e
rn
y
r
6-in R.
F
A
2 in
A
0.75 in
2-in R.
4 in
6 in
(a)
(b)
16.9 kpsi
+
4
2
5
3
6
r
5.63 kpsi
(c)
Note in the hook example, the symmetrical rectangular cross section causes the
maximum tensile stress to be 3 times greater than the maximum compressive stress. If
we wanted to design the hook to use material more effectively we would use more
material at the inner radius and less material at the outer radius. For this reason, trapezoidal, T, or unsymmetric I, cross sections are commonly used. Sections most frequently encountered in the stress analysis of curved beams are shown in Table 34.
119
120
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Table 34
Formulas for Sections of
Curved Beams
rc = r i +
h
2
h
ro
rn =
rc
rn
h
ln (ro /r i )
ri
bo
rc = ri +
e
h
ro
bi
ri
rn
rc
bo
c2
ro
rn
bi
A
bo b i + [(b i ro bo r i )/h] ln (ro /r i )
rc = r i +
e
c1
rn =
rc
rn =
b i c 2 + 2bo c1 c2 + bo c 2
1
2
2(bo c2 + b i c1 )
b i c1 + bo c2
b i ln [(r i + c1 )/r i )] + bo ln [ro /(r i + c1 )]
ri
R
rc = r i + R
e
rn =
ri
rn
rc
R2
2 rc
bo
rc = r i +
to
e
h b i + 2bo
3 b i + bo
r2 R2
c
12
ht
2
t i (b i t ) + to (bo t ) + ht
t
h
rn =
ro
ti
rc r
n
1
+ 2 t i2 (b i t ) + to (bo t )(h to /2)
t i (b i t ) + to (bo t ) + hto
ri + t
ro to
ro
b i ln
+ t ln
+ bo ln
ri
ri + ti
ro to
bi
ri
b
rc = r i +
to
e
t
2
t
2
h
ro
rc r
n
ti
rn =
12
ht
2
1
+ 2 t 2 (b t ) + to (b t )(h to /2)
i
ht + (b t )(t i + to )
(b t )(t i + t o ) + ht
ri + ti
ro
ro to
b ln
+ ln
+ t ln
ri
ro to
ri + ti
ri
115
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
116
I. Basics
121
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Alternative Calculations for e
Calculating rn and rc mathematically and subtracting the difference can lead to large
errors if not done carefully, since rn and rc are typically large values compared to e.
Since e is in the denominator of Eqs. (364) and (365), a large error in e can lead to
an inaccurate stress calculation. Furthermore, if you have a complex cross section that
the tables do not handle, alternative methods for determining e are needed. For a quick
and simple approximation of e, it can be shown that13
.I
e=
rc A
(366)
.
This approximation is good for a large curvature where e is small with rn = rc .
Substituting Eq. (366) into Eq. (364), with rn y = r , gives
. M y rc
σ=
Ir
(367)
.
If rn = rc , which it should be to use Eq. (367), then it is only necessary to calculate rc ,
and to measure y from this axis. Determining rc for a complex cross section can be done
easily by most CAD programs or numerically as shown in the before mentioned reference. Observe that as the curvature increases, r rc , and Eq. (367) becomes the
straight-beam formulation, Eq. (324). Note that the negative sign is missing because y
in Fig. 334 is vertically downward, opposite that for the straight-beam equation.
13
Ibid., pp 317321. Also presents a numerical method.
EXAMPLE 316
Solution
Consider the circular section in Table 34 with rc = 3 in and R = 1 in. Determine e by
using the formula from the table and approximately by using Eq. (366). Compare the
results of the two solutions.
Using the formula from Table 34 gives
rn =
R2
2 rc
2
rc R 2
=
12
= 2.91421 in
2 3 32 1
This gives an eccentricity of
Answer
e = rc rn = 3 2.91421 = 0.08579 in
The approximate method, using Eq. (366), yields
Answer
π R 4 /4
R2
12
.I
=
=
= 0.08333 in
e=
=
rc A
rc (π R 2 )
4rc
4(3)
This differs from the exact solution by 2.9 percent.
122
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
117
Load and Stress Analysis
319
Contact Stresses
When two bodies having curved surfaces are pressed together, point or line contact
changes to area contact, and the stresses developed in the two bodies are threedimensional. Contact-stress problems arise in the contact of a wheel and a rail,
in automotive valve cams and tappets, in mating gear teeth, and in the action of
rolling bearings. Typical failures are seen as cracks, pits, or flaking in the surface
material.
The most general case of contact stress occurs when each contacting body has a
double radius of curvature; that is, when the radius in the plane of rolling is different
from the radius in a perpendicular plane, both planes taken through the axis of the contacting force. Here we shall consider only the two special cases of contacting spheres
and contacting cylinders.14 The results presented here are due to Hertz and so are frequently known as Hertzian stresses.
Spherical Contact
When two solid spheres of diameters d1 and d2 are pressed together with a force
F, a circular area of contact of radius a is obtained. Specifying E 1 , ν1 and E 2 , ν2
as the respective elastic constants of the two spheres, the radius a is given by the
equation
a=
3
2
2
3 F 1 ν1 E 1 + 1 ν2
8
1/d1 + 1/d2
(368)
E2
The pressure distribution within the contact area of each sphere is hemispherical, as shown
in Fig. 336b. The maximum pressure occurs at the center of the contact area and is
pmax =
3F
2π a 2
(369)
Equations (368) and (369) are perfectly general and also apply to the contact of
a sphere and a plane surface or of a sphere and an internal spherical surface. For a plane
surface, use d = . For an internal surface, the diameter is expressed as a negative
quantity.
The maximum stresses occur on the z axis, and these are principal stresses. Their
values are
1
z
(1 + ν)
σ1 = σ2 = σx = σ y = pmax 1
tan1
a
|z /a |
σ3 = σz =
14
pmax
z2
1+ 2
a
1
2
2 1+
z
a2
(370)
(371)
A more comprehensive presentation of contact stresses may be found in Arthur P. Boresi and Richard
J. Schmidt, Advanced Mechanics of Materials, 6th ed., Wiley, New York, 2003 pp. 589623.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
118
I. Basics
123
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Figure 336
F
(a) Two spheres held in
contact by force F; (b) contact
stress has a hemispherical
distribution across contact
zone diameter 2a.
F
x
d1
y
y
2a
d2
F
F
z
z
(a )
(b)
These equations are valid for either sphere, but the value used for Poissons ratio
must correspond with the sphere under consideration. The equations are even more complicated when stress states off the z axis are to be determined, because here the x and y
coordinates must also be included. But these are not required for design purposes,
because the maxima occur on the z axis.
Mohrs circles for the stress state described by Eqs. (370) and (371) are a point
and two coincident circles. Since σ1 = σ2 , we have τ1/2 = 0 and
τmax = τ1/3 = τ2/3 =
σ2 σ3
σ1 σ3
=
2
2
(372)
Figure 337 is a plot of Eqs. (370), (371), and (372) for a distance to 3a below the
surface. Note that the shear stress reaches a maximum value slightly below the surface.
It is the opinion of many authorities that this maximum shear stress is responsible for
the surface fatigue failure of contacting elements. The explanation is that a crack originates at the point of maximum shear stress below the surface and progresses to the surface and that the pressure of the lubricant wedges the chip loose.
Cylindrical Contact
Figure 338 illustrates a similar situation in which the contacting elements are two
cylinders of length l and diameters d1 and d2. As shown in Fig. 338b, the area of contact is a narrow rectangle of width 2b and length l , and the pressure distribution is
elliptical. The half-width b is given by the equation
b=
2
2
2 F 1 ν1 E 1 + 1 ν2
πl
1/d1 + 1/d2
E2
(373)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
Magnitude of the stress
components below the surface
as a function of the maximum
pressure of contacting spheres.
Note that the maximum shear
stress is slightly below the
surface at z = 0.48a and is
approximately 0.3pmax. The
chart is based on a Poisson
ratio of 0.30. Note that the
normal stresses are all
compressive stresses.
119
,
Figure 337
1.0
0.8
Ratio of stress to pmax
124
z
0.6
x
,
y
0.4
max
0.2
0
z
0
0.5a
1.5a
a
2a
2.5a
3a
Distance from contact surface
Figure 338
F
(a) Two right circular cylinders
held in contact by forces F
uniformly distributed along
cylinder length l. (b) Contact
stress has an elliptical
distribution across the
contact zone width 2b.
F
x
x
d1
l
y
y
2b
d2
F
F
z
z
(a )
(b )
The maximum pressure is
pmax =
2F
π bl
(374)
Equations (373) and (374) apply to a cylinder and a plane surface, such as a rail, by
making d = for the plane surface. The equations also apply to the contact of a cylinder and an internal cylindrical surface; in this case d is made negative for the internal
surface.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
120
I. Basics
125
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
The stress state along the z axis is given by the equations
σx = 2ν pmax
1+
z2
z
2
b
b
(375)
z2
1 + 2 b2
z
σ y = pmax
2
2
b
z
1+ 2
b
pmax
σ3 = σz =
(376)
(377)
1 + z 2 /b2
These three equations are plotted in Fig. 339 up to a distance of 3b below the surface.
For 0 z 0.436b, σ1 = σx , and τmax = (σ1 σ3 )/2 = (σx σz )/2. For z 0.436b,
σ1 = σ y , and τmax = (σ y σz )/2. A plot of τmax is also included in Fig. 339, where the
greatest value occurs at z /b = 0.786 with a value of 0.300 pmax .
Hertz (1881) provided the preceding mathematical models of the stress eld when
the contact zone is free of shear stress. Another important contact stress case is line of
contact with friction providing the shearing stress on the contact zone. Such shearing
stresses are small with cams and rollers, but in cams with atfaced followers, wheel-rail
contact, and gear teeth, the stresses are elevated above the Hertzian eld. Investigations
of the effect on the stress eld due to normal and shear stresses in the contact zone were
begun theoretically by Lundberg (1939), and continued by Mindlin (1949), Smith-Liu
(1949), and Poritsky (1949) independently. For further detail, see the reference cited in
Footnote 14.
,
Figure 339
1.0
0.8
y
Ratio of stress to pmax
Magnitude of the stress
components below the surface
as a function of the maximum
pressure for contacting
cylinders. The largest value of
τmax occurs at z / b = 0.786.
Its maximum value is
0.30pmax. The chart is based
on a Poisson ratio of 0.30.
Note that all normal stresses
are compressive stresses.
z
0.6
0.4
x
max
0.2
0
z
0
0.5b
b
1.5b
2b
Distance from contact surface
2.5b
3b
126
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Load and Stress Analysis
320
121
Summary
The ability to quantify the stress condition at a critical location in a machine element
is an important skill of the engineer. Why? Whether the member fails or not is
assessed by comparing the (damaging) stress at a critical location with the corresponding material strength at this location. This chapter has addressed the description
of stress.
Stresses can be estimated with great precision where the geometry is sufciently
simple that theory easily provides the necessary quantitative relationships. In other
cases, approximations are used. There are numerical approximations such as nite
element analysis (FEA, see Chap. 19), whose results tend to converge on the true values. There are experimental measurements, strain gauging, for example, allowing inference of stresses from the measured strain conditions. Whatever the method(s), the goal
is a robust description of the stress condition at a critical location.
The nature of research results and understanding in any field is that the longer
we work on it, the more involved things seem to be, and new approaches are sought
to help with the complications. As newer schemes are introduced, engineers, hungry
for the improvement the new approach promises, begin to use the approach.
Optimism usually recedes, as further experience adds concerns. Tasks that promised
to extend the capabilities of the nonexpert eventually show that expertise is not
optional.
In stress analysis, the computer can be helpful if the necessary equations are available. Spreadsheet analysis can quickly reduce complicated calculations for parametric
studies, easily handling what if questions relating trade-offs (e.g., less of a costly
material or more of a cheaper material). It can even give insight into optimization
opportunities.
When the necessary equations are not available, then methods such as FEA are
attractive, but cautions are in order. Even when you have access to a powerful FEA
code, you should be near an expert while you are learning. There are nagging questions
of convergence at discontinuities. Elastic analysis is much easier than elastic-plastic
analysis. The results are no better than the modeling of reality that was used to formulate the problem. Chapter 19 provides an idea of what nite-element analysis is and how
it can be used in design. The chapter is by no means comprehensive in nite-element
theory and the application of nite elements in practice. Both skill sets require much
exposure and experience to be adept.
PROBLEMS
31
The symbol W is used in the various gure parts to specify the weight of an element. If not
given, assume the parts are weightless. For each gure part, sketch a free-body diagram of each
element, including the frame. Try to get the forces in the proper directions, but do not compute
magnitudes.
32
Using the gure part selected by your instructor, sketch a free-body diagram of each element in
the gure. Compute the magnitude and direction of each force using an algebraic or vector
method, as specied.
33
Find the reactions at the supports and plot the shear-force and bending-moment diagrams for each
of the beams shown in the gure on page 123. Label the diagrams properly.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
127
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
1
1
1
1
2
3
2
W
2
1
1
W
W
(b)
(a )
(c )
Problem 31
1
1
3
2
2
1
2
1
W
W
W
(d )
(e)
y
(f)
y
2
0.4 m
0.15m radius
B
A
1
45°
W = 2 kN
F = 800 N
0.6 m
60°
30°
2
O
(a)
x
1
(b )
y
Problem 32
y
F = 400 N
B
30°
F = 1.2 kN
0.9 m
122
I. Basics
C
3
4
2
2
3
B
60°
A
O
1
x
60°
A
1.9 m
D
60°
O
1
5
E
x
9m
(c)
(d )
34
Repeat Prob. 33 using singularity functions exclusively (for reactions as well).
35
Select a beam from Table A9 and nd general expressions for the loading, shear-force, bendingmoment, and support reactions. Use the method specied by your instructor.
128
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
y
y
4 in
O
4 in
6 in
A
4 in
D
O
x
C
B
R1
2 kN
4 kN/m
60 lbf
40 lbf
A
C
B
x
200 mm 150 mm 150 mm
R2
30 lbf
123
(b)
(a )
y
y
1000 lbf
Problem 33
6 ft
O
1000 lbf
4 ft
x
A
O
B
R1
R2
2 ft
2000 lbf
6 ft
2 ft
A
R2
(d )
Hinge
y
40 lbf/in
y
400 lbf
4 ft
3 ft
B
R1
B
O
x
R2
C
D
x
A
R1
C
R3
R2
8 in
(e)
36
320 lbf
800 lbf
3 ft
A
C
R1
(c)
O
x
B
5 in
2 in
5 in
(f)
A beam carrying a uniform load is simply supported with the supports set back a distance a from
the ends as shown in the gure. The bending moment at x can be found from summing moments
to zero at section x :
1
1
M = M + w(a + x )2 wlx = 0
2
2
or
M=
w
[lx (a + x )2 ]
2
where w is the loading intensity in lbf/in. The designer wishes to minimize the necessary weight
of the supporting beam by choosing a setback resulting in the smallest possible maximum bending stress.
(a) If the beam is congured with a = 2.25 in, l = 10 in, and w = 100 lbf/in, nd the magnitude
of the severest bending moment in the beam.
(b) Since the conguration in part (a) is not optimal, nd the optimal setback a that will result in
the lightest-weight beam.
x
w(a + x)
w, lbf/in
M
Problem 36
V
a
x
a
l
wl
2
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
124
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
37
An artist wishes to construct a mobile using pendants, string, and span wire with eyelets as shown
in the gure.
(a) At what positions w, x , y , and z should the suspension strings be attached to the span wires?
(b) Is the mobile stable? If so, justify; if not, suggest a remedy.
z
y
x
Problem 37
w
l
38
For each of the plane stress states listed below, draw a Mohrs circle diagram properly labeled,
nd the principal normal and shear stresses, and determine the angle from the x axis to σ1 . Draw
stress elements as in Fig. 311c and d and label all details.
(a) σx = 12, σ y = 6, τx y = 4 cw
(b) σx = 16, σ y = 9, τx y = 5 ccw
(c) σx = 10, σ y = 24, τx y = 6 ccw
(d ) σx = 9, σ y = 19, τx y = 8 cw
39
Repeat Prob. 38 for:
(a) σx = 4, σ y = 12, τx y = 7 ccw
(b) σx = 6, σ y = 5, τx y = 8 ccw
(c) σx = 8, σ y = 7, τx y = 6 cw
(d ) σx = 9, σ y = 6, τx y = 3 cw
310
Repeat Prob. 38 for:
(a) σx = 20, σ y = 10, τx y
(b) σx = 30, σ y = 10, τx y
(c) σx = 10, σ y = 18, τx y
(d ) σx = 12, σ y = 22, τx y
= 8 cw
= 10 ccw
= 9 cw
= 12 cw
311
For each of the stress states listed below, nd all three principal normal and shear stresses. Draw
a complete Mohrs three-circle diagram and label all points of interest.
(a) σx = 10, σ y = 4
(b) σx = 10, τx y = 4 ccw
(c) σx = 2, σ y = 8, τx y = 4 cw
(d ) σx = 10, σ y = 30, τx y = 10 ccw
312
Repeat Prob. 311 for:
(a) σx = 80, σ y = 30, τx y = 20 cw
(b) σx = 30, σ y = 60, τx y = 30 cw
(c) σx = 40, σz = 30, τx y = 20 ccw
(d ) σx = 50, σz = 20, τx y = 30 cw
129
130
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
125
313
1
A 2 -in-diameter steel tension rod is 72 in long and carries a load of 2000 lbf. Find the tensile
stress, the total deformation, the unit strains, and the change in the rod diameter.
314
Twin diagonal aluminum alloy tension rods 15 mm in diameter are used in a rectangular frame
to prevent collapse. The rods can safely support a tensile stress of 135 MPa. If the rods are initially 3 m in length, how much must they be stretched to develop this stress?
315
Electrical strain gauges were applied to a notched specimen to determine the stresses in the notch.
The results were ǫx = 0.0021 and ǫ y = 0.00067. Find σx and σ y if the material is carbon steel.
316
An engineer wishes to determine the shearing strength of a certain epoxy cement. The problem
is to devise a test specimen such that the joint is subject to pure shear. The joint shown in the gure, in which two bars are offset at an angle θ so as to keep the loading force F centroidal with
the straight shanks, seems to accomplish this purpose. Using the contact area A and designating
Ssu as the ultimate shearing strength, the engineer obtains
F
cos θ
A
Ssu =
The engineers supervisor, in reviewing the test results, says the expression should be
Ssu =
F
A
1+
1
tan2 θ
4
1/2
cos θ
Resolve the discrepancy. What is your position?
Problem 316
317
318
F
F
The state of stress at a point is σx = 2, σ y = 6, σz = 4, τx y = 3, τ y z = 2, and τz x = 5 kpsi.
Determine the principal stresses, draw a complete Mohrs three-circle diagram, labeling all points
of interest, and report the maximum shear stress for this case.
Repeat Prob. 317 with σx = 10, σ y = 0, σz = 10, τx y = 20, τ y z = 10 2, and τz x = 0 MPa.
319
Repeat Prob. 317 with σx = 1, σ y = 4, σz = 4, τx y = 2, τ y z = 4, and τz x = 2 kpsi.
320
The Roman method for addressing uncertainty in design was to build a copy of a design that was
satisfactory and had proven durable. Although the early Romans did not have the intellectual
tools to deal with scaling size up or down, you do. Consider a simply supported, rectangular-crosssection beam with a concentrated load F, as depicted in the gure.
F
c
a
Problem 320
R2
h
b
R1
l
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
126
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
(a) Show that the stress-to-load equation is
F=
σ bh 2 l
6ac
(b) Subscript every parameter with m (for model) and divide into the above equation. Introduce
a scale factor, s = am /a = bm /b = cm /c etc. Since the Roman method was to not lean on
the material any more than the proven design, set σm /σ = 1. Express Fm in terms of the scale
factors and F, and comment on what you have learned.
321
Using our experience with concentrated loading on a simple beam, Prob. 320, consider a uniformly loaded simple beam (Table A97).
(a) Show that the stress-to-load equation for a rectangular-cross-section beam is given by
W=
4 σ bh 2
3l
where W = wl .
(b) Subscript every parameter with m (for model) and divide the model equation into the prototype equation. Introduce the scale factor s as in Prob. 320, setting σm /σ = 1. Express Wm
and wm in terms of the scale factor, and comment on what you have learned.
322
The Chicago North Shore & Milwaukee Railroad was an electric railway running between the
cities in its corporate title. It had passenger cars as shown in the gure, which weighed 104.4 kip,
had 32-ft, 8-in truck centers, 7-ft-wheelbase trucks, and a coupled length of 55 ft, 3 1 in. Consider
4
the case of a single car on a 100-ft-long, simply supported deck plate girder bridge.
(a) What was the largest bending moment in the bridge?
(b) Where on the bridge was the moment located?
(c) What was the position of the car on the bridge?
(d ) Under which axle is the bending moment?
7 ft
32 ft, 8 in
Problem 322
Copyright 1963 by Central Electric Railfans Association, Bull. 107, p. 145, reproduced by permission.
131
132
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
127
Load and Stress Analysis
323
For each section illustrated, nd the second moment of area, the location of the neutral axis, and
the distances from the neutral axis to the top and bottom surfaces. Suppose a positive bending
moment of 10 kip · in is applied; nd the resulting stresses at the top and bottom surfaces and at
every abrupt change in cross section.
D
y
D
7
8
in
1
4
in
1
4
C
in
C
3
8
1
4
B
B
in
60°
A
1
4
1
2
in
1 in
60°
A
in
2 in
(a)
(b)
y
y
C
D
C
B
3 in
Problem 323
1
2
1
2
in
B
A
1
2
4 in
in
30°
30°
in
A
1 in
2 in
4 in
4 in
4 in
(c)
(d )
y
1 1 in
4
6 in
C
y
1 1 in
2
D
1 in
B
3 in
1
4
C
in
B
A
1 1 in
2
1 in
1 in
A
(e)
(f)
324
From basic mechanics of materials, in the derivation of the bending stresses, it is found that the
radius of curvature of the neutral axis, ρ , is given by ρ = E I / M . Find the x and y coordinates of
the center of curvature corresponding to the place where the beam is bent the most, for each beam
shown in the gure. The beams are both made of Douglas r (see Table A5) and have rectangular sections.
325
For each beam illustrated in the gure, nd the locations and magnitudes of the maximum tensile bending stress and the maximum shear stress due to V.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
128
I. Basics
133
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
y
y
50 lbf
50 lbf
50 lbf
1
2
Problem 324
C
O 20 in
in
1
2
x
B
30 in
A
in
5 in
O
B 20 in
A
50 lbf
5 in
C
x
2 in
2 in
20 in
(a)
(b)
y
y
1000 lbf
1000 lbf
3
4
12 in
A
6 in
in
1 in
B
x
x
O
O
8 in
A
B
8 in
1
1 2 in
(a )
2 in
(b)
Problem 325
y
y
3
4
w = 120 lbf/in
in
1 in
w = 100 lbf/in
x
O 5 in
A
15 in
B
x
5 in C
O 6 in
A
12 in
2 in
2 in
(c)
326
B
(d )
The gure illustrates a number of beam sections. Use an allowable bending stress of 1.2 kpsi for
wood and 12 kpsi for steel and nd the maximum safe uniformly distributed load that each beam
can carry if the given lengths are between simple supports.
1
(a) Wood joist 1 2 by 9 1 in and 12 ft long
2
3
(b) Steel tube, 2 in OD by 8 -in wall thickness, 48 in long
3
(c) Hollow steel tube 3 by 2 in, outside dimensions, formed from 16 -in material and welded, 48 in
long
(d ) Steel angles 3 × 3 × 1 in and 72 in long
4
(e) A 5.4-lb, 4-in steel channel, 72 in long
( f ) A 4-in × 1-in steel bar, 72 in long
y
z
Problem 326
y
z
y
z
(a)
(b)
(c)
y
y
y
z
z
z
(d )
(e)
(f)
134
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
327
129
A pin in a knuckle joint carrying a tensile load F deects somewhat on account of this loading,
making the distribution of reaction and load as shown in part b of the gure. The usual designers assumption of loading is shown in part c; others sometimes choose the loading shown in part
d. If a = 0.5 in, b = 0.75 in, d = 0.5 in, and F = 1000 lbf, estimate the maximum bending stress
and the maximum shear stress due to V for each approximation.
F
(b)
b
2
Problem 327
d
a+b
(c)
a
a
b
F
a+b
b
(d )
(a)
328
The gure illustrates a pin tightly tted into a hole of a substantial member. A usual analysis
is one that assumes concentrated reactions R and M at distance l from F. Suppose the reaction
is distributed linearly along distance a. Is the resulting moment reaction larger or smaller than
the concentrated reaction? What is the loading intensity q? What do you think of using the
usual assumption?
F
l
a
Problem 328
329
For the beam shown, determine (a) the maximum tensile and compressive bending stresses,
(b) the maximum shear stress due to V, and (c) the maximum shear stress in the beam.
3000 lbf
2 in
600 lbf/ft
A
B
C
6 in
Problem 329
2 in
5 ft
15 ft
6 in
Cross section (enlarged)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
130
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
330
Consider a simply supported beam of rectangular cross section of constant width b and variable
depth h, so proportioned that the maximum stress σx at the outer surface due to bending is constant, when subjected to a load F at a distance a from the left support and a distance c from the
right support. Show that the depth h at location x is given by
h=
331
0x a
In Prob. 330, h 0 as x 0, which cannot occur. If the maximum shear stress τmax due to
direct shear is to be constant in this region, show that the depth h at location x is given by
h=
332
6 Fcx
lbσmax
3 Fc
2 lbτmax
0x
3 Fcσmax
2
8 lbτmax
Consider a simply supported static beam of circular cross section of diameter d, so proportioned
by varying the diameter such that the maximum stress σx at the surface due to bending is constant, when subjected to a steady load F located at a distance a from the left support and a distance b from the right support. Show that the diameter d at a location x is given by
d=
32 Fbx
π l σmax
1/3
0x a
333
Two steel thin-wall tubes in torsion of equal length are to be compared. The rst is of square cross
section, side length b, and wall thickness t. The second is a round of diameter b and wall thickness t. The largest allowable shear stress is τall and is to be the same in both cases. How does the
angle of twist per unit length compare in each case?
334
Begin with a 1-in-square thin-wall steel tube, wall thickness t = 0.05 in, length 40 in, then introduce corner radii of inside radii ri , with allowable shear stress τall of 11 500 psi, shear modulus
of 11.5(106) psi; now form a table. Use a column of inside corner radii in the range 0 ri 0.45
in. Useful columns include median line radius rm , periphery of the median line L m , area enclosed
by median curve, torque T, and the angular twist θ . The cross section will vary from square to
circular round. A computer program will reduce the calculation effort. Study the table. What have
you learned?
ri
Problem 334
rm
t
1 in
1 in
335
An unequal leg angle shown in the gure carries a torque T. Show that
T=
G θ1
3
τmax = G θ1 cmax
L i ci3
135
136
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Load and Stress Analysis
131
c1
Problem 335
L1
c2
L2
336
1
In Prob. 335 the angle has one leg thickness 16 in and the other 1 in, with both leg lengths
8
The allowable shear stress is τall = 12 000 psi for this steel angle.
(a) Find the torque carried by each leg, and the largest shear stress therein.
(b) Find the angle of twist per unit length of the section.
337
Two 12 in long thin rectangular steel strips are placed together as shown. Using a maximum
allowable shear stress of 12 000 psi, determine the maximum torque and angular twist, and the
1
torsional spring rate. Compare these with a single strip of cross section 1 in by 8 in.
1
8
1
16
5
8
in.
in
in
Problem 337
1 in
T
338
Using a maximum allowable shear stress of 60 MPa, nd the shaft diameter needed to transmit
35 kw when
(a) The shaft speed is 2000 rev/min.
(b) The shaft speed is 200 rev/min.
339
A 15-mm-diameter steel bar is to be used as a torsion spring. If the torsional stress in the bar is
not to exceed 110 MPa when one end is twisted through an angle of 30°, what must be the length
of the bar?
340
A 3-in-diameter solid steel shaft, used as a torque transmitter, is replaced with a 3-in hollow shaft
1
having a 4 -in wall thickness. If both materials have the same strength, what is the percentage
reduction in torque transmission? What is the percentage reduction in shaft weight?
341
A hollow steel shaft is to transmit 5400 N · m of torque and is to be sized so that the torsional
stress does not exceed 150 MPa.
(a) If the inside diameter is three-fourths of the outside diameter, what size shaft should be used?
Use preferred sizes.
(b) What is the stress on the inside of the shaft when full torque is applied?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
132
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
342
The gure shows an endless-belt conveyor drive roll. The roll has a diameter of 6 in and is driven
at 5 rev/min by a geared-motor source rated at 1 hp. Determine a suitable shaft diameter dC for
an allowable torsional stress of 14 kpsi.
(a) What would be the stress in the shaft you have sized if the motor starting torque is twice the
running torque?
(b) Is bending stress likely to be a problem? What is the effect of different roll lengths B on
bending?
(a )
Problem 342
y
dA
dB
dA
dC
x
A
B
A
C
(b )
343
The conveyer drive roll in the gure for Prob. 342 is 150 mm in diameter and is driven at
8 rev/min by a geared-motor source rated at 1 kW. Find a suitable shaft diameter dC based on an
allowable torsional stress of 75 MPa.
344
For the same cross-sectional area A = s 2 = π d 2 /4, for a square cross-sectional area shaft and a
circular cross-sectional area shaft, in torsion which has the higher maximum shear stress, and by
what multiple is it higher?
345
For the same cross-sectional area A = s 2 = π d 2 /4, for a square cross-sectional area shaft and a
circular cross-sectional area shaft, both of length l , in torsion which has the greater angular twist
θ , and by what multiple is it greater?
346
In the figure, shaft AB is rotating at 1000 rev/min and transmits 10 hp to shaft CD through a
set of bevel gears contacting at point E. The contact force at E on the gear of shaft CD is
determined to be (FE)CD
92.8i 362.8j 808.0k lbf. For shaft CD: (a) draw a free-body
diagram and determine the reactions at C and D assuming simple supports (assume also that
bearing C is a thrust bearing), (b) draw the shear-force and bending-moment diagrams, and
(c) assuming that the shaft diameter is 1.25 in, determine the maximum tensile and shear
stresses in the beam.
137
138
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
133
y
6.50 in
3.90 in
3 in
D
3 in
Problem 346
x
1.30 in
A
B
E
4 in
C
347
Repeat the analysis of Prob. 346 for shaft AB. Let the diameter of the shaft be 1.0 in, and assume
that bearing A is a thrust bearing.
348
A torque of T = 1000 lbf · in is applied to the shaft EFG, which is running at constant speed and contains gear F. Gear F transmits torque to shaft ABCD through gear C, which drives the chain sprocket at B, transmitting a force P as shown. Sprocket B, gear C, and gear F have pitch diameters of 6, 10,
and 5 in, respectively. The contact force between the gears is transmitted through the pressure angle
φ = 20°. Assuming no frictional losses and considering the bearings at A, D, E, and G to be simple
supports, locate the point on shaft ABCD that contains the maximum tensile bending and maximum
torsional shear stresses. From this, determine the maximum tensile and shear stresses in the shaft.
y
a
E
G
F
5 in
y
T
T = 1000 lbf in
B
A
D
C
x
Problem 348
z
10 in
1.25-in dia.
a
P
10 in
3 in
P
5 in
6 in
View aa
349
If the tension-loaded plate of Fig. 329 is innitely wide, then the stress state anywhere in the
plate can be described in polar coordinates as15
σr =
d2
1
d2
σ 1 2 + 1 2
2
4r
4r
1
3d 2
4r 2
cos 2θ
15
See R. G. Budynas, Advanced Strength and Applied Stress Analysis, 2nd ed. McGraw-Hill, New York,
1999, pp. 235238.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
134
I. Basics
139
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
σθ =
d2
1
3 d4
σ 1+ 2 1+
2
4r
16 r 4
1 d2
1
τr θ = σ 1
2
4 r2
1+
3 d2
4 r2
cos 2θ
sin 2θ
for the radial, tangential, and shear components, respectively. Here r is the distance from the center to the point of interest and θ is measured positive counterclockwise from the horizontal axis.
(a) Find the stress components at the top and side of the hole for r = d /2.
(b) If d = 10 mm, plot a graph of the tangential stress distribution σθ /σ for θ = 90º from r = 5 mm
to 20 mm.
(c) Repeat part (b) for θ = 0º
350
Considering the stress concentration at point A in the gure, determine the maximum normal and
shear stresses at A if F = 200 lbf.
y
2 in
O
A
12 in
z
Problem 350
1
12
-in dia.
1
8 -in
B
2 in C
R.
1-in dia.
15 in
F
1
x
1 2 -in dia.
D
351
Develop the formulas for the maximum radial and tangential stresses in a thick-walled cylinder
due to internal pressure only.
352
Repeat Prob. 351 where the cylinder is subject to external pressure only. At what radii do the
maximum stresses occur?
353
Develop the stress relations for a thin-walled spherical pressure vessel.
354
A pressure cylinder has a diameter of 150 mm and has a 6-mm wall thickness. What pressure can
this vessel carry if the maximum shear stress is not to exceed 25 Mpa?
355
A cylindrical pressure vessel has an outside diameter of 10 in and a wall thickness of
internal pressure is 350 psi, what is the maximum shear stress in the vessel walls?
3
8
in. If the
140
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
135
356
1
An AISI 1020 cold-drawn steel tube has an ID of 1 4 in and an OD of 1 3 in. What maximum
4
external pressure can this tube take if the largest principal normal stress is not to exceed 80 percent of the minimum yield strength of the material?
357
An AISI 1020 cold-drawn steel tube has an ID of 40 mm and an OD of 50 mm. What maximum
internal pressure can this tube take if the largest principal normal stress is not to exceed 80 percent of the minimum yield strength of the material?
358
Find the maximum shear stress in a 10-in circular saw if it runs idle at 7200 rev/min. The saw is
14 gauge (0.0747 in) and is used on a 3 -in arbor. The thickness is uniform. What is the maximum
4
radial component of stress?
359
The maximum recommended speed for a 300-mm-diameter abrasive grinding wheel is 2069
rev/min. Assume that the material is isotropic; use a bore of 25 mm, ν = 0.24, and a mass density
of 3320 kg/m3; and nd the maximum tensile stress at this speed.
360
1
An abrasive cutoff wheel has a diameter of 6 in, is 16 in thick, and has a 1-in bore. It weighs 6
oz and is designed to run at 10 000 rev/min. If the material is isotropic and ν = 0.20, nd the
maximum shear stress at the design speed.
361
A rotary lawn-mower blade rotates at 3000 rev/min. The steel blade has a uniform cross section
1
1
in thick by 1 4 in wide, and has a 1 -in-diameter hole in the center as shown in the gure.
8
2
Estimate the nominal tensile stress at the central section due to rotation.
12 in
1
8
in
1 1 in
4
Problem 361
24 in
362 to
367
The table lists the maximum and minimum hole and shaft dimensions for a variety of standard
press and shrink ts. The materials are both hot-rolled steel. Find the maximum and minimum
values of the radial interference and the corresponding interface pressure. Use a collar diameter
of 80 mm for the metric sizes and 3 in for those in inch units.
Problem
Number
362
363
364
365
366
367
Fit
Designation*
Basic
Size
40H7/p6
(1.5 in)H7/p6
40H7/s6
(1.5 in)H7/s6
40H7/u6
(1.5 in)H7/u6
40 mm
1.5 in
40 mm
1.5 in
40 mm
1.5 in
Hole
Dmax
Dmin
40.025
1.5010
40.025
1.5010
40.025
1.5010
40.000
1.5000
40.000
1.5000
40.000
1.5000
Shaft
dmax
dmin
40.042
1.5016
40.059
1.5023
40.076
1.5030
40.026
1.5010
40.043
1.5017
40.060
1.5024
*Note: See Table 79 for description of ts.
368 to
371
The table gives data concerning the shrink fit of two cylinders of differing materials and
dimensional specification in inches. Elastic constants for different materials may be found in
Table A5. Identify the radial interference δ , then find the interference pressure p , and the
tangential normal stress on both sides of the fit surface. If dimensional tolerances are given at
fit surfaces, repeat the problem for the highest and lowest stress levels.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
136
I. Basics
141
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Mechanical Engineering Design
Problem
Number
Inner Cylinder
Material
di
d0
Outer Cylinder
Material
Di
D0
368
369
Steel
Steel
0
0
1.002
1.002
Steel
Cast iron
1.000
1.000
2.00
2.00
370
371
Steel
Steel
0
0
1.002/1.003
2.005/2.003
Steel
Aluminum
1.000/1.001
2.000/2.002
2.00
4.00
372
Force ts of a shaft and gear are assembled in an air-operated arbor press. An estimate of assembly
force and torque capacity of the t is needed. Assume the coefcient of friction is f , the t interface
pressure is p, the nominal shaft or hole radius is R , and the axial length of the gear bore is l .
(a) Show that the estimate of the axial force is Fax = 2π f Rlp.
(b) Show the estimate of the torque capacity of the t is T = 2π f R 2 lp.
373
A utility hook was formed from a 1-in-diameter round rod into the geometry shown in the gure.
What are the stresses at the inner and outer surfaces at section A- A if the load F is 1000 lbf?
F
3 in
Problem 373
10 in
1 in
3 in
A
374
A
F
The steel eyebolt shown in the gure is loaded with a force F of 100 lbf. The bolt is formed of
3
1
-in-diameter wire to a 8 -in radius in the eye and at the shank. Estimate the stresses at the inner
4
and outer surfaces at sections A- A and B - B .
F
3
8
Problem 374
3
8
1
4
375
-in R.
B
in
B
A
F
A
in
Shown in the gure is a 12-gauge (0.1094-in) by 3 -in latching spring that supports a load of
4
1
F = 3 lbf. The inside radius of the bend is 8 in. Estimate the stresses at the inner and outer surfaces at the critical section.
142
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
137
376
The cast-iron bell-crank lever depicted in the gure is acted upon by forces F1 of 250 lbf and F2
of 333 lbf. The section A- A at the central pivot has a curved inner surface with a radius of ri = 1
in. Estimate the stresses at the inner and outer surfaces of the curved portion of the lever.
377
The crane hook depicted in Fig. 335 has a 1-in-diameter hole in the center of the critical section.
For a load of 5 kip, estimate the bending stresses at the inner and outer surfaces at the critical section.
378
A 20-kip load is carried by the crane hook shown in the gure. The cross section of the hook uses
two concave anks. The width of the cross section is given by b = 2/ r ,where r is the radius from
the center. The inside radius ri is 2 in, and the outside radius ro = 6 in. Find the stresses at the
inner and outer surfaces at the critical section.
F
4 in
A
1
-in
8
A
3
4
R.
in
Problem 375
Section A-A
No. 12 gauge (0.1094 in)
F1
8 in
Nylon bushing
1
A
1 in
3 2 in
3
8
Problem 376
1-in R.
A
6 in
1
1 8 in
7
F2
1 8 in
Section A-A
Problem 378
4 in
2 in
in
1
1 4 in
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
138
I. Basics
3. Load and Stress Analysis
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
379
An offset tensile link is shaped to clear an obstruction with a geometry as shown in the gure.
The cross section at the critical location is elliptical, with a major axis of 4 in and a minor axis
of 2 in. For a load of 20 kip, estimate the stresses at the inner and outer surfaces of the critical
section.
10-in R.
Problem 379
380
8 in
A cast-steel C frame as shown in the gure has a rectangular cross section of 1 in by 1.6 in, with
a 0.4-in-radius semicircular notch on both sides that forms midank uting as shown. Estimate A,
rc , rn , and e, and for a load of 3000 lbf, estimate the inner and outer surface stresses at the throat
C . Note: Table 34 can be used to determine rn for this section. From the table, the integral
d A / r can be evaluated for a rectangle and a circle by evaluating A / rn for each shape [see
Eq. (364)]. Subtracting A / rn of the circle from that of the rectangle yields d A / r for the C
frame, and rn can then be evaluated.
0.4-in R.
Problem 380
3000 lbf
4 in
1 in
1-in R.
0.4 in
0.4 in
381
Two carbon steel balls, each 25 mm in diameter, are pressed together by a force F . In terms of
the force F , nd the maximum values of the principal stress, and the maximum shear stress, in
MPa.
382
One of the balls in Prob. 381 is replaced by a at carbon steel plate. If F = 18 N, at what depth
does the maximum shear stress occur?
383
An aluminum alloy roller with diameter 1 in and length 2 in rolls on the inside of a cast-iron ring
having an inside radius of 4 in, which is 2 in thick. Find the maximum contact force F that can
be used if the shear stress is not to exceed 4000 psi.
384
The gure shows a hip prosthesis containing a stem that is cemented into a reamed cavity in the
femur. The cup is cemented and fastened to the hip with bone screws. Shown are porous layers
of titanium into which bone tissue will grow to form a longer-lasting bond than that afforded by
cement alone. The bearing surfaces are a plastic cup and a titanium femoral head. The lip shown
in the gures bears against the cutoff end of the femur to transfer the load to the leg from the hip.
Walking will induce several million stress uctuations per year for an average person, so there is
danger that the prosthesis will loosen the cement bonds or that metal cracks may occur because
of the many repetitions of stress. Prostheses like this are made in many different sizes. Typical
143
144
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
3. Load and Stress Analysis
Load and Stress Analysis
139
C
Offset
D
Neck length
Problem 384
Porous hip prosthesis. (Photograph
and drawing courtesy of Zimmer,
Inc., Warsaw, Indiana.)
B
Stem
length
A distal stem diameter
(a )
(b)
dimensions are ball diameter 50 mm, stem diameter 15 mm, stem length 155 mm, offset 38 mm,
and neck length 39 mm. Develop an outline to follow in making a complete stress analysis of this
prosthesis. Describe the material properties needed, the equations required, and how the loading
is to be dened.
385
Simplify Eqs. (370), (371), and (372) by setting z = 0 and nding σx / pmax , σ y / pmax ,
σz / pmax , and τ2/3 / pmax and, for cast iron, check the ordinate intercepts of the four loci in
Fig. 337.
386
A 6-in-diameter cast-iron wheel, 2 in wide, rolls on a at steel surface carrying an 800-lbf load.
(a) Find the Hertzian stresses σx , σ y , σz , and τ2/3 .
(b) What happens to the stresses at a point A that is 0.010 in below the wheel rim surface during
a revolution?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
4
145
Deection and Stiffness
Chapter Outline
41
Spring Rates
42
Tension, Compression, and Torsion
43
Deection Due to Bending
144
44
Beam Deection Methods
146
45
Beam Deections by Superposition
46
Beam Deections by Singularity Functions
47
Strain Energy
48
Castiglianos Theorem
49
Deection of Curved Members
142
143
147
150
156
158
163
410
Statically Indeterminate Problems
411
Compression MembersGeneral
412
Long Columns with Central Loading
413
Intermediate-Length Columns with Central Loading
414
Columns with Eccentric Loading
415
Struts or Short Compression Members
416
Elastic Stability
417
Shock and Impact
418
Suddenly Applied Loading
168
173
173
176
176
180
182
183
184
141
146
142
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
All real bodies deform under load, either elastically or plastically. A body can be sufciently insensitive to deformation that a presumption of rigidity does not affect an analysis enough to warrant a nonrigid treatment. If the body deformation later proves to be not
negligible, then declaring rigidity was a poor decision, not a poor assumption. A wire
rope is exible, but in tension it can be robustly rigid and it distorts enormously under
attempts at compressive loading. The same body can be both rigid and nonrigid.
Deection analysis enters into design situations in many ways. A snap ring, or retaining ring, must be exible enough to be bent without permanent deformation and
assembled with other parts, and then it must be rigid enough to hold the assembled parts
together. In a transmission, the gears must be supported by a rigid shaft. If the shaft bends
too much, that is, if it is too exible, the teeth will not mesh properly, and the result will
be excessive impact, noise, wear, and early failure. In rolling sheet or strip steel to prescribed thicknesses, the rolls must be crowned, that is, curved, so that the nished product
will be of uniform thickness. Thus, to design the rolls it is necessary to know exactly how
much they will bend when a sheet of steel is rolled between them. Sometimes mechanical
elements must be designed to have a particular force-deection characteristic. The
suspension system of an automobile, for example, must be designed within a very narrow
range to achieve an optimum vibration frequency for all conditions of vehicle loading,
because the human body is comfortable only within a limited range of frequencies.
The size of a load-bearing component is often determined on deections, rather
than limits on stress.
This chapter considers distortion of single bodies due to geometry (shape) and
loading, then, briey, the behavior of groups of bodies.
41
Spring Rates
Elasticity is that property of a material that enables it to regain its original conguration
after having been deformed. A spring is a mechanical element that exerts a force when
deformed. Figure 41a shows a straight beam of length l simply supported at the ends
and loaded by the transverse force F. The deection y is linearly related to the force, as
long as the elastic limit of the material is not exceeded, as indicated by the graph. This
beam can be described as a linear spring.
In Fig. 41b a straight beam is supported on two cylinders such that the length
between supports decreases as the beam is deected by the force F. A larger force is
required to deect a short beam than a long one, and hence the more this beam is
deected, the stiffer it becomes. Also, the force is not linearly related to the deection,
and hence this beam can be described as a nonlinear stiffening spring.
Figure 41
l
l
d
F
(a) A linear spring; (b) a
stiffening spring; (c) a
softening spring.
F
F
y
y
y
F
F
F
(a)
y
y
y
(b)
(c)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
147
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
143
Figure 41c is an edge-view of a dish-shaped round disk. The force necessary to
atten the disk increases at rst and then decreases as the disk approaches a at conguration, as shown by the graph. Any mechanical element having such a characteristic
is called a nonlinear softening spring.
If we designate the general relationship between force and deection by the equation
F = F (y)
(a)
then spring rate is dened as
F
dF
=
y
dy
k ( y ) = lim
y 0
(41)
where y must be measured in the direction of F and at the point of application of F. Most
of the force-deection problems encountered in this book are linear, as in Fig. 41a. For
these, k is a constant, also called the spring constant; consequently Eq. (41) is written
k=
F
y
(42)
We might note that Eqs. (41) and (42) are quite general and apply equally well for
torques and moments, provided angular measurements are used for y. For linear displacements, the units of k are often pounds per inch or newtons per meter, and for
angular displacements, pound-inches per radian or newton-meters per radian.
42
Tension, Compression, and Torsion
The total extension or contraction of a uniform bar in pure tension or compression,
respectively, is given by
δ=
Fl
AE
(43)
This equation does not apply to a long bar loaded in compression if there is a possibility of buckling (see Secs. 411 to 415). Using Eqs. (42) and (43), we see that the
spring constant of an axially loaded bar is
k=
AE
l
(44)
The angular deection of a uniform round bar subjected to a twisting moment T
was given in Eq. (335), and is
θ=
Tl
GJ
(45)
where θ is in radians. If we multiply Eq. (45) by 180/π and substitute J = π d 4 /32
for a solid round bar, we obtain
θ=
583.6T l
Gd 4
(46)
where θ is in degrees.
Equation (45) can be rearranged to give the torsional spring rate as
k=
GJ
T
=
θ
l
(47)
148
144
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
43
Deection Due to Bending
The problem of bending of beams probably occurs more often than any other loading
problem in mechanical design. Shafts, axles, cranks, levers, springs, brackets, and wheels,
as well as many other elements, must often be treated as beams in the design and analysis of mechanical structures and systems. The subject of bending, however, is one that
you should have studied as preparation for reading this book. It is for this reason that
we include here only a brief review to establish the nomenclature and conventions to be
used throughout this book.
The curvature of a beam subjected to a bending moment M is given by
1
M
=
ρ
EI
(48)
where ρ is the radius of curvature. From studies in mathematics we also learn that the
curvature of a plane curve is given by the equation
d 2 y /dx 2
1
=
ρ
[1 + (dy /dx )2 ]3/2
(49)
where the interpretation here is that y is the lateral deection of the beam at any point
x along its length. The slope of the beam at any point x is
θ=
dy
dx
(a)
For many problems in bending, the slope is very small, and for these the denominator
of Eq. (49) can be taken as unity. Equation (48) can then be written
d2 y
M
=
EI
dx 2
(b)
Noting Eqs. (33) and (34) and successively differentiating Eq. (b) yields
d3 y
V
=
EI
dx 3
(c)
d4 y
q
=
EI
dx 4
(d )
It is convenient to display these relations in a group as follows:
d4 y
q
=
EI
dx 4
(410)
d3 y
V
=
EI
dx 3
(411)
d2 y
M
=
EI
dx 2
(412)
θ=
dy
dx
y = f (x )
(413)
(414)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Figure 42
149
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
145
y
l = 20 in
w
x
x
x
Moment, M
M0 = Ml = 0
Slope, EI
l/2 = 0
x
R1 = wl
2
Shear, V
V0 = + 800 lbf
Vl = 800 lbf
x
(a)
Loading, w
w = 80 lbf/in
Deflection, EIy
y0 = yl = 0
R2 = wl
2
V
V0
+
Vl
(b)
M
+
(c)
M0
Ml
EI
+
(d)
EI
0
EI
l
EIy
(e)
The nomenclature and conventions are illustrated by the beam of Fig. 42. Here, a beam
of length l = 20 in is loaded by the uniform load w = 80 lbf per inch of beam length.
The x axis is positive to the right, and the y axis positive upward. All quantities
loading, shear, moment, slope, and deectionhave the same sense as y; they are positive if upward, negative if downward.
The reactions R1 = R2 = +800 lbf and the shear forces V0 = +800 lbf and
Vl = 800 lbf are easily computed by using the methods of Chap. 3. The bending
moment is zero at each end because the beam is simply supported. For a simplysupported beam, the deections are also zero at each end.
EXAMPLE 41
For the beam in Fig. 42, the bending moment equation, for 0 x l, is
M=
w
wl
x x2
2
2
Using Eq. (412), determine the equations for the slope and deection of the beam, the
slopes at the ends, and the maximum deection.
Solution
Integrating Eq. (412) as an indenite integral we have
EI
dy
=
dx
M dx =
wl 2 w 3
x x + C1
4
6
(1)
where C1 is a constant of integration that is evaluated from geometric boundary conditions.
We could impose that the slope is zero at the midspan of the beam, since the beam and
150
146
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
loading are symmetric relative to the midspan. However, we will use the given boundary conditions of the problem and verify that the slope is zero at the midspan. Integrating
Eq. (1) gives
EIy =
M dx =
w
wl 3
x x 4 + C1 x + C2
12
24
(2)
The boundary conditions for the simply supported beam are y = 0 at x = 0 and l.
Applying the rst condition, y = 0 at x = 0, to Eq. (2) results in C2 = 0. Applying the
second condition to Eq. (2) with C2 = 0,
E I y (l ) =
wl 3
w
l l 4 + C1l = 0
12
24
Solving for C1 yields C1 = wl 3 /24. Substituting the constants back into Eqs. (1) and
(2) and solving for the deection and slope results in
y=
wx
(2lx 2 x 3 l 3 )
24 E I
(3)
θ=
w
dy
=
(6lx 2 4x 3 l 3 )
dx
24 E I
(4)
Comparing Eq. (3) with that given in Table A9, beam 7, we see complete agreement.
For the slope at the left end, substituting x = 0 into Eq. (4) yields
θ |x =0 =
and at x = l,
θ |x = l =
wl 3
24 E I
wl 3
24 E I
At the midspan, substituting x = l/2 gives d y /dx = 0, as earlier suspected.
The maximum deection occurs where d y /dx = 0. Substituting x = l/2 into
Eq. (3) yields
ymax =
5wl 4
384 E I
which again agrees with Table A97.
The approach used in the example is ne for simple beams with continuous
loading. However, for beams with discontinuous loading and/or geometry such as a step
shaft with multiple gears, ywheels, pulleys, etc., the approach becomes unwieldy. The
following section discusses bending deections in general and the techniques that are
provided in this chapter.
44
Beam Deection Methods
Equations (410) through (414) are the basis for relating the intensity of loading q,
vertical shear V, bending moment M, slope of the neutral surface θ, and the transverse deflection y. Beams have intensities of loading that range from q = constant
I. Basics
© The McGrawHill
Companies, 2008
151
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
147
4. Deflection and Stiffness
(uniform loading), variable intensity q (x ), to Dirac delta functions (concentrated
loads).
The intensity of loading usually consists of piecewise contiguous zones, the
expressions for which are integrated through Eqs. (410) to (414) with varying
degrees of difculty. Another approach is to represent the deection y (x ) as a Fourier
series, which is capable of representing single-valued functions with a nite number of
nite discontinuities, then differentiating through Eqs. (414) to (410), and stopping
at some level where the Fourier coefcients can be evaluated. A complication is the
piecewise continuous nature of some beams (shafts) that are stepped-diameter bodies.
All of the above constitute, in one form or another, formal integration methods,
which, with properly selected problems, result in solutions for q, V, M, θ, and y. These
solutions may be
1
2
3
Closed-form, or
Represented by innite series, which amount to closed form if the series are
rapidly convergent, or
Approximations obtained by evaluating the rst or the rst and second terms.
The series solutions can be made equivalent to the closed-form solution by the use of a
computer. Roarks1 formulas are committed to commercial software and can be used on
a personal computer.
There are many techniques employed to solve the integration problem for beam
deection. Some of the popular methods include:
Superposition (see Sec. 45)
The moment-area method2
Singularity functions (see Sec. 46)
Numerical integration3
The two methods described in this chapter are easy to implement and can handle a large
array of problems.
There are methods that do not deal with Eqs. (410) to (414) directly. An energy
method, based on Castiglianos theorem, is quite powerful for problems not suitable for
the methods mentioned earlier and is discussed in Secs. 47 to 410. Finite element
programs are also quite useful for determining beam deections.
45
Beam Deections by Superposition
The results of many simple load cases and boundary conditions have been solved
and are available. Table A9 provides a limited number of cases. Roarks4 provides
a much more comprehensive listing. Superposition resolves the effect of combined
loading on a structure by determining the effects of each load separately and adding
1
Warren C. Young and Richard G. Budynas, Roarks Formulas for Stress and Strain, 7th ed., McGraw-Hill,
New York, 2002.
2
See Chap. 9, F. P. Beer, E. R. Johnston Jr., and J. T. DeWolf, Mechanics of Materials, 4th ed., McGraw-Hill,
New York, 2006.
3
See Sec. 44, J. E. Shigley and C. R. Mischke, Mechanical Engineering Design, 6th ed., McGraw-Hill,
New York, 2001.
4
Warren C. Young and Richard G. Budynas, Roarks Formulas for Stress and Strain, 7th ed., McGraw-Hill,
New York, 2002.
152
148
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
the results algebraically. Superposition may be applied provided: (1) each effect is
linearly related to the load that produces it, (2) a load does not create a condition that
affects the result of another load, and (3) the deformations resulting from any specific load are not large enough to appreciably alter the geometric relations of the
parts of the structural system.
The following examples are illustrations of the use of superposition.
EXAMPLE 42
Consider the uniformly loaded beam with a concentrated force as shown in Fig. 43.
Using superposition, determine the reactions and the deection as a function of x.
Solution
Considering each load state separately, we can superpose beams 6 and 7 of Table A9.
For the reactions we nd
Answer
R1 =
Fb wl
+
l
2
Answer
R2 =
wl
Fa
+
l
2
The loading of beam 6 is discontinuous and separate deection equations are given
for regions A B and BC. Beam 7 loading is not discontinuous so there is only one equation. Superposition yields
Answer
y AB =
Fbx 2
wx
( x + b2 l 2 ) +
(2lx 2 x 3 l 3 )
6E I l
24 E I
Answer
y BC =
Fa (l x ) 2
wx
(x + a 2 2lx ) +
(2lx 2 x 3 l 3 )
6E I l
24 E I
y
Figure 43
l
F
b
a
w
C
A
B
R1
x
R2
If we wanted to determine the maximum deection in the previous example, we
would set dy /dx = 0 and solve for the value of x where the deection is a maximum.
If a = l /2, the maximum deection would obviously occur at x = l /2 because of
symmetry. However, if a < l /2, where would the maximum deection be? It can be
shown that as the force F moves toward the left support, the maximum deection moves
toward the left support also, but not as much as F (see Prob. 434). Thus, we would set
dy BC /dx = 0 and solve for x.
Sometimes it may not be obvious that we can use superposition with the tables at
hand, as demonstrated in the next example.
I. Basics
EXAMPLE 43
Solution
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
153
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
149
Consider the beam in Fig. 44a and determine the deection equations using
superposition.
For region A B we can superpose beams 7 and 10 of Table A9 to obtain
Answer
y AB =
wx
Fax 2
(2lx 2 x 3 l 3 ) +
(l x 2 )
24 E I
6E I l
For region B C , how do we represent the uniform load? Considering the uniform
load only, the beam deflects as shown in Fig. 44b. Region B C is straight since
there is no bending moment due to w . The slope of the beam at B is θB and is
obtained by taking the derivative of y given in the table with respect to x and setting
x = l . Thus,
d
w
wx
dy
=
(2lx 2 x 3 l 3 ) =
(6lx 2 4x 3 l 3 )
dx
dx 24 E I
24 E I
Substituting x = l gives
θB =
w
wl 3
(6ll 2 4l 3 l 3 ) =
24 E I
24 E I
The deection in region B C due to w is θ B (x l ), and adding this to the deection due
to F, in BC, yields
Answer
y BC =
y
Figure 44
(a) Beam with uniformly
distributed load and overhang
force; (b) deections due to
uniform load only.
y
a
l
w
F
w
B
A
R2
R1
(a)
EXAMPLE 44
F (x l )
wl 3
(x l ) +
[(x l )2 a (3x l )]
24 E I
6E I
C
B
A
x
B
yBC =
B( x
l)
x
C
l
x
(b)
Figure 45a shows a cantilever beam with an end load. Normally we model this problem by considering the left support as rigid. After testing the rigidity of the wall it was
found that the translational stiffness of the wall was kt force per unit vertical deection,
and the rotational stiffness was kr moment per unit angular (radian) deection (see
Fig. 45b). Determine the deection equation for the beam under the load F.
154
150
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Solution
Here we will superpose the modes of deection. They are: (1) translation due to the
compression of spring kt , (2) rotation of the spring kr , and (3) the elastic deformation
of the beam given by Table A91. The force in spring kt is R1 = F , giving a deection from Eq. (42) of
y1 =
F
kt
(1)
The moment in spring kr is M1 = Fl . This gives a clockwise rotation of θ = Fl / kr .
Considering this mode of deection only, the beam rotates rigidly clockwise, leading to
a deection equation of
y2 =
Fl
x
kr
(2)
Finally, the elastic deformation of the beam from Table A91 is
y3 =
F x2
(x 3l )
6E I
(3)
Adding the deections from each mode yields
Answer
y=
F
F x2
Fl
(x 3l )
x
6E I
kt
kr
y
Figure 45
l
F
x
M1
R1
(a)
kr
F
x
kt
R1
(b )
46
Beam Deections by Singularity Functions
Introduced in Sec. 33, singularity functions are excellent for managing discontinuities, and
their application to beam deection is a simple extension of what was presented in the earlier section. They are easy to program, and as will be seen later, they can greatly simplify
the solution of statically indeterminate problems. The following examples illustrate the use
of singularity functions to evaluate deections of statically determinate beam problems.
I. Basics
EXAMPLE 45
Solution
© The McGrawHill
Companies, 2008
155
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
151
4. Deflection and Stiffness
Consider the beam of Table A96, which is a simply supported beam having a concentrated load F not in the center. Develop the deection equations using singularity
functions.
First, write the load intensity equation from the free-body diagram,
q = R1 x
1
F x a
1
+ R2 x l
1
(1)
Integrating Eq. (1) twice results in
V = R1 x
0
F x a
0
+ R2 x l
0
(2)
M = R1 x
1
F x a
1
+ R2 x l
1
(3)
Recall that as long as the q equation is complete, integration constants are unnecessary
for V and M; therefore, they are not included up to this point. From statics, setting
V = M = 0 for x slightly greater than l yields R1 = Fb/ l and R2 = Fa / l . Thus Eq. (3)
becomes
M=
Fb
x
l
1
F x a
1
+
Fa
x l
l
1
Integrating Eqs. (412) and (413) as indenite integrals gives
EI
dy
Fb
x
=
dx
2l
EIy =
Fb
x
6l
2
F
x a
2
3
F
x a
6
2
+
Fa
x l
2l
3
+
Fa
x l
6l
2
+ C1
3
+ C1 x + C2
Note that the rst singularity term in both equations always exists, so x 2 = x 2
and x 3 = x 3 . Also, the last singularity term in both equations does not exist until
x = l, where it is zero, and since there is no beam for x > l we can drop the last term.
Thus
EI
Fb 2 F
dy
x a
=
x
dx
2l
2
EIy =
Fb 3 F
x a
x
6l
6
2
+ C1
(4)
3
+ C1 x + C2
(5)
The constants of integration C1 and C2 are evaluated by using the two boundary conditions y = 0 at x = 0 and y = 0 at x = l. The rst condition, substituted into Eq. (5),
gives C2 = 0 (recall that 0 a 3 = 0). The second condition, substituted into Eq. (5),
yields
0=
Fb3
Fbl 2
Fb 3 F
l (l a )3 + C1l =
+ C1l
6l
6
6
6
Solving for C1 ,
C1 =
Fb 2
(l b2 )
6l
156
152
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Finally, substituting C1 and C2 in Eq. (5) and simplifying produces
F
[bx (x 2 + b2 l 2 ) l x a 3 ]
6E I l
y=
(6)
Comparing Eq. (6) with the two deection equations in Table A96, we note that the
use of singularity functions enables us to express the deection equation with a single
equation.
EXAMPLE 46
Determine the deection equation for the simply supported beam with the load distribution shown in Fig. 46.
Solution
This is a good beam to add to our table for later use with superposition. The load intensity equation for the beam is
q = R1 x
1
w x
0
+w x a
0
+ R2 x l
1
(1)
where the w x a 0 is necessary to turn off the uniform load at x = a.
From statics, the reactions are
R1 =
wa
(2l a )
2l
R2 =
wa 2
2l
(2)
For simplicity, we will retain the form of Eq. (1) for integration and substitute the values
of the reactions in later.
Two integrations of Eq. (1) reveal
V = R1 x
0
w x
M = R1 x
1
w
x
2
1
+w x a
2
+
w
x a
2
1
+ R2 x l
2
+ R2 x l
0
(3)
1
(4)
As in the previous example, singularity functions of order zero or greater starting at
x = 0 can be replaced by normal polynomial functions. Also, once the reactions are
determined, singularity functions starting at the extreme right end of the beam can be
omitted. Thus, Eq. (4) can be rewritten as
M = R1 x
y
Figure 46
l
a
w
B
A
R1
C
R2
x
w2 w
x a
x+
2
2
2
(5)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
157
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
153
Integrating two more times for slope and deection gives
EI
dy
R1 2 w 3 w
x a
=
x x+
dx
2
6
6
EIy =
3
R1 3
w
w
x a
x x4 +
6
24
24
+ C1
(6)
4
(7)
+ C1 x + C2
The boundary conditions are y = 0 at x = 0 and y = 0 at x = l. Substituting the rst
condition in Eq. (7) shows C2 = 0. For the second condition
0=
R1 3
w
w
l l 4 + (l a )4 + C1l
6
24
24
Solving for C1 and substituting into Eq. (7) yields
EIy =
R1
w
w
w
x a
x (x 2 l 2 ) x (x 3 l 3 )
x (l a )4 +
6
24
24l
24
4
Finally, substitution of R1 from Eq. (2) and simplifying results gives
Answer
y=
w
[2ax (2l a )(x 2 l 2 ) xl (x 3 l 3 ) x (l a )4 + l x a 4 ]
24 E I l
As stated earlier, singularity functions are relatively simple to program, as they are
omitted when their arguments are negative, and the
brackets are replaced with ( )
parentheses when the arguments are positive.
EXAMPLE 47
The steel step shaft shown in Fig. 47a is mounted in bearings at A and F. A pulley
is centered at C where a total radial force of 600 lbf is applied. Using singularity
functions evaluate the shaft displacements at 1 - in increments. Assume the shaft is
2
simply supported.
Solution
The reactions are found to be R1 = 360 lbf and R2 = 240 lbf. Ignoring R2 , using
singularity functions, the moment equation is
M = 360x 600 x 8
1
(1)
This is plotted in Fig. 47b.
For simplication, we will consider only the step at D. That is, we will assume section AB has the same diameter as BC and section EF has the same diameter as DE.
Since these sections are short and at the supports, the size reduction will not add much
to the deformation. We will examine this simplication later. The second area moments
for BC and DE are
I BC =
π
1.54 = 0.2485 in4
64
IDE =
π
1.754 = 0.4604 in4
64
158
154
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Figure 47
y
600 lbf
1.000
AB
Dimensions in inches.
1.750
1.500
C
D
1.000
E
F
x
0.5
8
R1
8.5
19.5
(a)
2880 lbf-in
M
R2
20
2760 lbf-in
(b )
M/I
a
b
c
d
(c )
A plot of M / I is shown in Fig. 47c. The values at points b and c, and the step change are
M
I
b
M
I
=
2760
= 11 106.6 lbf/in3
0.2485
M
I
c
=
2760
= 5 994.8 lbf/in3
0.4604
= 5 994.8 11 106.6 = 5 111.8 lbf/in3
The slopes for ab and cd, and the change are
m ab =
360 600
= 965.8 lbf/in4
0.2485
5 994.8
= 521.3 lbf/in4
11.5
m cd =
m = 521.3 ( 965.8) = 444.5 lbf/in4
Dividing Eq. (1) by I BC and, at x
of slope 444.5 lbf/in4 , gives
8.5 in, adding a step of 5 111.8 lbf/in3 and a ramp
M
= 1 448.7x 2 414.5 x 8 1 5 111.8 x 8.5 0 + 444.5 x 8.5
I
1
(2)
Integrating twice gives
E
dy
= 724.35x 2 1207.3 x 8 2 5 111.8 x 8.5
dx
+222.3 x 8.5 2 + C1
1
(3)
and
E y = 241.5x 3 402.4 x 8 3 2 555.9 x 8.5
2
+ 74.08 x 8.5 3 + C1 x + C2
(4)
At x = 0, y = 0. This gives C2 = 0 (remember, singularity functions do not exist until
the argument is positive). At x = 20 in, y = 0, and
0 = 241.5(20) 3 402.4(20 8) 3 2 555.9(20 8.5) 2 + 74.08(20 8.5) 3 + C1 (20)
I. Basics
© The McGrawHill
Companies, 2008
159
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
155
4. Deflection and Stiffness
Solving, gives C1 = 50 565 lbf/in2 . Thus, Eq. (4) becomes, with E = 30(10) 6 psi,
y=
1
(241.5x 3 402.4 x 8 3 2 555.9 x 8.5
30(106 )
2
+74.08 x 8.5 3 50 565x )
(5)
When using a spreadsheet, program the following equations:
y=
1
(241.5x 3 50 565x )
30(106 )
0 x 8 in
y=
1
[241.5x 3 402.4( x 8) 3 50 565x ]
30(106 )
8 x 8.5 in
y=
1
[241.5x 3 402.4 ( x 8) 3 2 555.9 ( x 8.5) 2
30(106 )
+74.08 ( x 8.5) 3 50 565x ]
8.5 x 20 in
The following table results.
x
y
x
y
x
y
x
y
x
y
0
0.000000
4.5
0.006851
9
0.009335
13.5
0.007001
18
0.002377
0.5
0.000842
5
0.007421
9.5
0.009238
14
0.006571
18.5
0.001790
1
0.001677
5.5
0.007931
10
0.009096
14.5
0.006116
19
0.001197
1.5
0.002501
6
0.008374
10.5
0.008909
15
0.005636
19.5
0.000600
20
0.000000
2
0.003307
6.5
0.008745
11
0.008682
15.5
0.005134
2.5
0.004088
7
0.009037
11.5
0.008415
16
0.004613
3
0.004839
7.5
0.009245
12
0.008112
16.5
0.004075
3.5
0.005554
8
0.009362
12.5
0.007773
17
0.003521
4
0.006227
8.5
0.009385
13
0.007403
17.5
0.002954
where x and y are in inches. We see that the greatest deection is at x = 8.5 in, where
y = 0.009385 in.
Substituting C1 into Eq. (3) the slopes at the supports are found to be θ A = 1.686(103 )
rad = 0.09657 deg, and θ F = 1.198(103 ) rad = 0.06864 deg. You might think these to
be insignicant deections, but as you will see in Chap. 7, on shafts, they are not.
A nite-element analysis was performed for the same model and resulted in
y |x
= 8.5 in
= 0.009380 in
θ A = 0.09653
θ F = 0.06868
Virtually the same answer save some round-off error in the equations.
If the steps of the bearings were incorporated into the model, more equations result,
but the process is the same. The solution to this model is
y |x
= 8.5 in
= 0.009387 in
θ A = 0.09763
θ F = 0.06973
The largest difference between the models is of the order of 1.5 percent. Thus the simplication was justied.
160
156
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
In Sec. 49, we will demonstrate the usefulness of singularity functions in solving
statically indeterminate problems.
47
Strain Energy
The external work done on an elastic member in deforming it is transformed into strain,
or potential, energy. If the member is deformed a distance y, and if the force-deection
relationship is linear, this energy is equal to the product of the average force and the
deection, or
U=
F2
F
y=
2
2k
(a)
This equation is general in the sense that the force F can also mean torque, or moment,
provided, of course, that consistent units are used for k. By substituting appropriate
expressions for k, strain-energy formulas for various simple loadings may be obtained.
For tension and compression and for torsion, for example, we employ Eqs. (44) and
(47) and obtain
U=
F 2l
2 AE
tension and compression
(415)
U=
T 2l
2G J
torsion
(416)
To obtain an expression for the strain energy due to direct shear, consider the
element with one side xed in Fig. 48a. The force F places the element in pure shear,
and the work done is U = F δ/2. Since the shear strain is γ = δ/ l = τ/ G = F / AG ,
we have
U=
F 2l
2 AG
direct shear
(417)
The strain energy stored in a beam or lever by bending may be obtained by referring to Fig. 48b. Here AB is a section of the elastic curve of length d s having a radius
of curvature ρ. The strain energy stored in this element of the beam is d U = ( M /2)d θ.
Since ρ d θ = ds , we have
dU =
O
Figure 48
F
d
A
F
l
ds
F
B
dx
(a) Pure shear element
(b) Beam bending element
M ds
2ρ
(b)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
161
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
157
We can eliminate ρ by using Eq. (48). Thus
dU =
M 2 ds
2E I
(c)
.
For small deections, d s = dx . Then, for the entire beam
M 2 dx
2E I
U=
(418)
bending
Equation (418) is exact only when a beam is subject to pure bending. Even when
shear is present, Eq. (418) continues to give quite good results, except for very short
beams. The strain energy due to shear loading of a beam is a complicated problem. An
approximate solution can be obtained by using Eq. (417) with a correction factor
whose value depends upon the shape of the cross section. If we use C for the correction
factor and V for the shear force, then the strain energy due to shear in bending is the
integral of Eq. (417), or
U=
C V 2 dx
2 AG
bending shear
(419)
Values of the factor C are listed in Table 41.
Table 41
Beam Cross-Sectional Shape
Strain-Energy Correction
Factors for Shear
Source: Richard G. Budynas,
Advanced Strength and
Applied Stress Analysis,
2nd ed., McGraw-Hill,
New York, 1999.
Copyright © 1999 The
McGraw-Hill Companies.
EXAMPLE 48
Solution
Factor C
Rectangular
1.2
Circular
1.11
Thin-walled tubular, round
2.00
Box sections
1.00
Structural sections
1.00
Use area of web only.
Find the strain energy due to shear in a rectangular cross-section beam, simply supported, and having a uniformly distributed load.
Using Appendix Table A97, we nd the shear force to be
V=
wl
wx
2
Substituting into Eq. (419), with C = 1.2, gives
Answer
U=
1.2
2 AG
l
0
wl
wx
2
2
dx =
w2l 3
20 AG
162
158
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
EXAMPLE 49
A cantilever has a concentrated load F at the end, as shown in Fig. 49. Find the strain
energy in the beam by neglecting shear.
Figure 49
l
ymax
F
x
Solution
At any point x along the beam, the moment is M = F x . Substituting this value of M
into Eq. (418), we nd
l
Answer
48
U=
0
F 2l 3
F 2 x 2 dx
=
2E I
6E I
Castiglianos Theorem
A most unusual, powerful, and often surprisingly simple approach to deection analysis is afforded by an energy method called Castiglianos theorem. It is a unique way of
analyzing deections and is even useful for nding the reactions of indeterminate structures. Castiglianos theorem states that when forces act on elastic systems subject to
small displacements, the displacement corresponding to any force, in the direction of
the force, is equal to the partial derivative of the total strain energy with respect to that
force. The terms force and displacement in this statement are broadly interpreted to
apply equally to moments and angular displacements. Mathematically, the theorem of
Castigliano is
δi =
U
Fi
(420)
where δi is the displacement of the point of application of the force Fi in the direction
of Fi . For rotational displacement Eq. (420) can be written as
θi =
U
Mi
(421)
where θi is the rotational displacement, in radians, of the beam where the moment
Mi exists and in the direction of Mi .
As an example, apply Castiglianos theorem using Eqs. (415) and (416) to get
the axial and torsional deections. The results are
δ=
F
F 2l
2 AE
=
Fl
AE
(a)
θ=
T
T 2l
2G J
=
Tl
GJ
(b)
I. Basics
© The McGrawHill
Companies, 2008
163
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
159
4. Deflection and Stiffness
Compare Eqs. (a) and (b) with Eqs. (43) and (45). In Example 48, the bending strain
energy for a cantilever having a concentrated end load was found. According to
Castiglianos theorem, the deection at the end of the beam due to bending is
y=
U
=
F
F
F 2l 3
6E I
=
Fl 3
3E I
(c)
which checks with Table A91.
Castiglianos theorem can be used to nd the deection at a point even though no
force or moment acts there. The procedure is:
1
2
3
Set up the equation for the total strain energy U by including the energy due
to a fictitious force or moment Q i acting at the point whose deflection is to be
found.
Find an expression for the desired deection δi , in the direction of Q i , by taking
the derivative of the total strain energy with respect to Q i .
Since Q i is a ctitious force, solve the expression obtained in step 2 by setting
Q i equal to zero. Thus,
δi =
EXAMPLE 410
Solution
U
Qi
(422)
Q i =0
The cantilever of Ex. 49 is a carbon steel bar 10 in long with a 1-in diameter and is
loaded by a force F = 100 lbf.
(a) Find the maximum deection using Castiglianos theorem, including that due to shear.
(b) What error is introduced if shear is neglected?
(a) From Eq. (419) and Example 49 data, the total strain energy is
U=
F 2l 3
+
6E I
l
0
C V 2 dx
2 AG
(1)
For the cantilever, the shear force is constant with repect to x, V = F . Also, C = 1.11,
from Table 41. Performing the integration and substituting these values in Eq. (1)
gives, for the total strain energy,
U=
1.11 F 2l
F 2l 3
+
6E I
2 AG
(2)
Then, according to Castiglianos theorem, the deection of the end is
y=
U
Fl 3
1.11 Fl
=
+
F
3E I
AG
We also nd that
I=
π d4
π(1)4
=
= 0.0491 in4
64
64
A=
π d2
π(1)2
=
= 0.7854 in2
4
4
(3)
164
160
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Substituting these values, together with F = 100 lbf, l = 10 in, E = 30 Mpsi, and
G = 11.5 Mpsi, in Eq. (3) gives
Answer
Answer
y = 0.022 63 + 0.000 12 = 0.022 75 in
Note that the result is positive because it is in the same direction as the force F.
(b) The error in neglecting shear for this problem is found to be about 0.53 percent.
In performing any integrations, it is generally better to take the partial derivative
with respect to the load Fi rst. This is true especially if the force is a ctitious force
Q i , since it can be set to zero as soon as the derivative is taken. This is demonstrated in
the next example. The forms for deection can then be rewritten. Here we will assume,
for axial and torsional loading, that material and cross section properties and loading
can vary along the length of the members. From Eqs. (415), (416), and (418),
δi =
U
=
Fi
1
AE
F
F
Fi
θi =
U
=
Mi
1
GJ
T
T
Mi
δi =
U
=
Fi
1
EI
M
M
Fi
dx
dx
dx
tension and compression
(423)
torsion
(424)
bending
(425)
EXAMPLE 411
Using Castiglianos method, determine the deections of points A and B due to the
force F applied at the end of the step shaft shown in Fig. 410. The second area
moments for sections AB and BC are I1 and 2 I1 , respectively.
Solution
With cantilever beams we normally set up the coordinate system such that x starts at the
wall and is directed towards the free end. Here, for simplicity, we have reversed
that. With the coordinate system of Fig. 410 the bending moment expression is simpler
than with the usual coordinate system, and does not require the support reactions. For
0 x l , the bending moment is
M = F x
(1)
Since F is at A and in the direction of the desired deection, the deection at A from
Eq. (425) is
Figure 410
y
l /2
x
A
F
l /2
I1
B
Qi
2 I1
C
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
U
=
F
δA =
l
0
1
EI
M
F
M
165
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
161
(2)
dx
Substituting Eq. (1) into Eq. (2), noting that I = I1 for 0 x l /2, and I = 2 I1 for
l /2 x l , we get
δA =
=
Answer
1
E
l /2
1
E
0
l
1
( F x ) (x ) dx +
I1
l /2
1
( F x ) (x ) dx
2 I1
3 Fl 3
Fl3
7 Fl 3
=
+
24 I1
48 I1
16 E I1
which is positive, as it is in the direction of F.
For B, a ctitious force Q i is necessary at the point. Assuming Q i acts down at B,
and x is as before, the moment equation is
M = F x
M = F x Qi
0 x l /2
l
x
2
(3)
l /2 x l
For Eq. (425), we need M / Q i . From Eq. (3),
M
=0
Qi
0 x l /2
l
M
= x
Qi
2
l /2 x l
(4)
Once the derivative is taken, Q i can be set to zero, so from Eq. (3), M = F x for
0 x l , and Eq. (425) becomes
l
δB =
0
1
EI
M
M
Qi
=
dx
Q i =0
1
E I1
l /2
0
( F x )(0) dx +
1
E (2 I1 )
l
l /2
( F x ) x
l
2
dx
Evaluating the last integral gives
Answer
F
δB =
2 E I1
x 3 lx 2
3
4
l
l /2
=
5 Fl 3
96 E I1
which again is positive, in the direction of Q i .
EXAMPLE 412
For the wire form of diameter d shown in Fig. 411a, determine the deection of point
B in the direction of the applied force F (neglect the effect of bending shear).
Solution
It is very important to include the loading effects on all parts of the structure. Coordinate
systems are not important, but loads must be consistent with the problem. Thus
166
162
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Figure 411
G
c
z
F
x
B
D
a
b
C
y
(a )
MG 2 = MD 2 = Fb
G
F
MD 2 = MD = Fb
D
MG1 = MD1 = Fa
F
F
B
MD1 = TD = Fa
MD = Fb
C
MC = Fa
TD = TC = Fa
F
D
F
F
C
TC = MC = Fa
(b )
appropriate use of free-body diagrams is essential here. The reader should verify that the
reactions as functions of F in elements B C , C D , and G D are as shown in Fig. 411b.
The deection of B in the direction of F is given by
δB =
U
F
so the partial derivatives in Eqs. (423) to (425) will all be taken with respect to F.
Element BC is in bending only so from Eq. (425),5
1
U BC
=
F
EI
a
0
( F y )( y ) dy =
Fa 3
3E I
(1)
Element C D is in bending and in torsion. The torsion is constant so Eq. (424) can be
written as
U
T
=T
Fi
Fi
5
l
GJ
It is very tempting to mix techniques and try to use superposition also, for example. However, some subtle
things can occur that you may visually miss. It is highly recommended that if you are using Castiglianos
theorem on a problem, you use it for all parts of the problem.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
167
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
163
where l is the length of the member. So for the torsion in member CD, Fi = F , T = Fa ,
and l = b. Thus,
UCD
F
torsion
= ( Fa )(a )
b
Fa 2 b
=
GJ
GJ
(2)
For the bending in CD,
UCD
F
bending
=
b
1
EI
( F x )(x ) dx =
0
Fb3
3E I
(3)
Member DG is axially loaded and is bending in two planes. The axial loading is
constant, so Eq. (423) can be written as
U
=
Fi
F
F
Fi
l
AE
where l is the length of the member. Thus, for the axial loading of D G , F = Fi , l = c,
and
U DG
F
axial
=
Fc
AE
(4)
The bending moments in each plane of D G are constant along the length of M y = Fb
and Mx = Fa . Considering each one separately in the form of Eq. (425) gives
U DG
F
bending
=
1
EI
c
0
( Fb)(b) dz +
1
EI
c
( Fa )(a ) dz
0
Fc(a 2 + b2 )
=
EI
(5)
Adding Eqs. (1) to (5), noting that I = π d 4 /64, J = 2 I , A = π d 2 /4, and G =
E /[2(1 + ν)], we nd that the deection of B in the direction of F is
Answer
(δ B ) F =
4F
[16(a 3 + b3 ) + 48c(a 2 + b2 ) + 48(1 + ν)a 2 b + 3cd 2 ]
3π Ed 4
Now that we have completed the solution, see if you can physically account for each
term in the result.
49
Deection of Curved Members
Machine frames, springs, clips, fasteners, and the like frequently occur as curved
shapes. The determination of stresses in curved members has already been described in
Sec. 318. Castiglianos theorem is particularly useful for the analysis of deections in
curved parts too. Consider, for example, the curved frame of Fig. 412a. We are interested in nding the deection of the frame due to F and in the direction of F . The total
strain energy consists of four terms, and we shall consider each separately. The rst is
due to the bending moment and is6
6
See Richard G. Budynas, Advanced Strength and Applied Stress Analysis, 2nd ed., Sec. 6.7, McGraw-Hill,
New York, 1999.
168
164
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Fr
d
M
F
h
R
F
F
(b)
(a )
Figure 412
(a) Curved bar loaded by force F. R = radius to centroidal axis of section;
h = section thickness. (b) Diagram showing forces acting on section taken at
angle θ. F r = V = shear component of F; F θ is component of F normal to
section; M is moment caused by force F.
U1 =
M2 dθ
2 AeE
(426)
In this equation, the eccentricity e is
(427)
e = R rn
where rn is the radius of the neutral axis as dened in Sec. 318 and shown in Fig. 334.
An approximate result can be obtained by using the equation
.
U1 =
M2 R dθ
2E I
R
> 10
h
(428)
which is obtained directly from Eq. (418). Note the limitation on the use of Eq. (428).
The strain energy component due to the normal force Fθ consists of two parts, one
of which is axial and analogous to Eq. (415). This part is
U2 =
Fθ2 R d θ
2 AE
(429)
The force Fθ also produces a moment, which opposes the moment M in Fig. 412b. The
resulting strain energy will be subtractive and is
U3 =
M Fθ d θ
AE
(430)
The negative sign of Eq. (430) can be appreciated by referring to both parts of
Fig. 412. Note that the moment M tends to decrease the angle d θ . On the other hand,
the moment due to Fθ tends to increase d θ . Thus U3 is negative. If Fθ had been acting
in the opposite direction, then both M and Fθ would tend to decrease the angle d θ .
The fourth and last term is the shear energy due to Fr . Adapting Eq. (419) gives
U4 =
C Fr2 R d θ
2 AG
(431)
where C is the correction factor of Table 41.
Combining the four terms gives the total strain energy
U=
M2 dθ
+
2 AeE
Fθ2 R d θ
2 AE
M Fθ d θ
+
AE
C Fr2 R d θ
2 AG
(432)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
169
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
165
The deection produced by the force F can now be found. It is
δ=
U
=
F
π
0
π
M
F
M
AeE
0
π
dθ +
0
1 ( M Fθ )
dθ +
AE F
Fθ
F
Fθ R
AE
π
0
dθ
Fr
F
C Fr R
AG
dθ
(433)
Using Fig. 412b, we nd
M = F R sin θ
M
= R sin θ
F
Fθ = F sin θ
Fθ
= sin θ
F
MFθ = F 2 R sin2 θ
M Fθ
= 2 F R sin2 θ
F
Fr
= cos θ
F
Fr = F cos θ
Substituting these into Eq. (433) and factoring yields
F R2
AeE
π
π
0
sin2 θ d θ +
FR
AE
0
sin2 θ d θ
2F R
AE
+
δ=
CFR
AG
y
A
R
x
C
+
F
O
z
B
T axis
=
Figure 413
Ring ABC in the xy plane
subject to force F parallel to
the z axis. Corresponding to
a ring segment CB at angle θ
from the point of application
of F, the moment axis is a line
BO and the torque axis is a
line in the xy plane tangent to
the ring at B. Note the positive
directions of the T and M
axes.
sin2 θ d θ
0
π
cos2 θ d θ
0
π F R2
πFR
πFR
πC F R
π F R2
πFR
πC F R
+
+
=
+
2 AeE
2 AE
AE
2 AG
2 AeE
2 AE
2 AG
(434)
Because the rst term contains the square of the radius, the second two terms will be
small if the frame has a large radius. Also, if R / h > 10, Eq. (428) can be used. An
approximate result then turns out to be
. π F R3
δ=
2E I
M axis
+
π
(435)
The determination of the deection of a curved member loaded by forces at right
angles to the plane of the member is more difcult, but the method is the same.7 We
shall include here only one of the more useful solutions to such a problem, though the
methods for all are similar. Figure 413 shows a cantilevered ring segment having a
span angle φ . Assuming R / h > 10, the strain energy neglecting direct shear, is
obtained from the equation
φ
U=
7
0
M2 R dθ
+
2E I
φ
0
T 2 R dθ
2G J
(436)
For more solutions than are included here, see Joseph E. Shigley, Curved Beams and Rings, Chap. 38 in
Joseph E. Shigley, Charles R. Mischke, and Thomas H. Brown, Jr. (eds.), Standard Handbook of Machine
Design, 3rd ed., McGraw-Hill, New York, 2004.
170
166
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
The moments and torques acting on a section at B, due to the force F, are
M = F R sin θ
T = F R (1 cos θ)
The deection δ of the ring segment at C and in the direction of F is then found to be
δ=
F R3
U
=
F
2
β
α
+
EI
GJ
(437)
where the coefcients α and β are dependent on the span angle φ and are dened as
follows:
α = φ sin φ cos φ
(438)
β = 3φ 4 sin φ + sin φ cos φ
(438)
where φ is in radians.
EXAMPLE 413
Deection in a Variable-Cross-Section Punch-Press Frame
The general result expressed in Eq. (434),
δ=
πFR
πC F R
π F R2
+
2 AeE
2 AE
2 AG
is useful in sections that are uniform and in which the centroidal locus is circular. The
bending moment is largest where the material is farthest from the load axis.
Strengthening requires a larger second area moment I. A variable-depth cross section is
attractive, but it makes the integration to a closed form very difcult. However, if you
are seeking results, numerical integration with computer assistance is helpful.
Consider the steel C frame depicted in Fig. 414a in which the centroidal radius is
32 in, the cross section at the ends is 2 in × 2 in, and the depth varies sinusoidally with
an amplitude of 2 in. The load is 1000 lbf. It follows that C = 1.2, G = 11.5(106 ) psi,
E = 30(106 ) psi. The outer and inner radii are
Rout = 33 + 2sin θ
Rin = 31 2sin θ
The remaining geometrical terms are
h = Rout Rin = 2(1 + 2 sin θ)
A = bh = 4(1 + 2 sin θ
rn =
2(1 + 2 sin θ)
h
=
ln[( R + h /2)/( R h /2)]
ln[(33 + 2 sin θ)/(31 2 sin θ)]
e = R rn = 32 rn
Note that
M = F R sin θ
M / F = R sin θ
Fθ = F sin θ
Fθ / F = sin θ
M Fθ = F 2 R sin2 θ
Fr = F cos θ
M Fθ / F = 2 F R sin2 θ
Fr / F = cos θ
1
I. Basics
© The McGrawHill
Companies, 2008
171
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
167
4. Deflection and Stiffness
Figure 414
(a) A steel punch press has a
C frame with a varying-depth
rectangular cross section
depicted. The cross section
varies sinusoidally from
2 in × 2 in at θ = 0 to
2 in × 6 in at θ = 90 , and
back to 2 in × 2 in at
θ = 180 . Of immediate
interest to the designer is the
deection in the load axis
direction under the load.
(b) Finite-element model.
31- in R
1000 lbf
1000 lbf
1000 lbf
(a )
(b)
Substitution of the terms into Eq. (433) yields three inteqrals
(1)
δ = I1 + I2 + I3
where the integrals are
sin2 θ d θ
π
I1 = 8.5333(103 )
0
(1 + 2 sin θ) 32
π
I2 = 2.6667(104 )
0
π
I3 = 8.3478(104 )
0
sin2 θ d θ
1 + 2 sin θ
2(1 + 2 sin θ)
33 + 2 sin θ
ln
31 2 sin θ
(2)
cos2 θ d θ
1 + 2 sin θ
(3)
(4)
The integrals may be evaluated in a number of ways: by a program using Simpsons
rule integration,8 by a program using a spreadsheet, or by mathematics software. Using
MathCad and checking the results with Excel gives the integrals as I1 = 0.076 615,
I2 = 0.000 159, and I3 = 0.000 773. Substituting these into Eq. (1) gives
Answer
δ = 0.077 23 in
Finite-element (FE) programs are also very accessible. Figure 414b shows a
simple half-model, using symmetry, of the press consisting of 216 plane-stress (2-D)
elements. Creating the model and analyzing it to obtain a solution took minutes.
Doubling the results from the FE analysis yielded δ = 0.07790 in, a less than 1 percent
variation from the results of the numerical integration.
8
See Case Study 4, p. 203, J. E. Shigley and C. R. Mischke, Mechanical Engineering Design, 6th ed.,
McGraw-Hill, New York, 2001.
172
168
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
410
Statically Indeterminate Problems
A system in which the laws of statics are not sufcient to determine all the unknown
forces or moments is said to be statically indeterminate. Problems of which this is true
are solved by writing the appropriate equations of static equilibrium and additional
equations pertaining to the deformation of the part. In all, the number of equations must
equal the number of unknowns.
A simple example of a statically indeterminate problem is furnished by the nested
helical springs in Fig. 415a. When this assembly is loaded by the compressive force
F, it deforms through the distance δ . What is the compressive force in each spring?
Only one equation of static equilibrium can be written. It is
F = F F1 F2 = 0
(a)
which simply says that the total force F is resisted by a force F1 in spring 1 plus the
force F2 in spring 2. Since there are two unknowns and only one equation, the system
is statically indeterminate.
To write another equation, note the deformation relation in Fig. 415b. The two
springs have the same deformation. Thus, we obtain the second equation as
δ1 = δ2 = δ
(b)
If we now substitute Eq. (42) in Eq. (b), we have
F2
F1
=
(c)
k1
k2
Now we solve Eq. (c) for F1 and substitute the result in Eq. (a). This gives
k2 F
k1
F F2 F2 = 0 or F2 =
(d )
k2
k1 + k2
This completes the solution, because with F2 known, F1 can be found from Eq. (c).
F
Figure 415
k1
k2
(a)
F1
F2
k1
k2
(b)
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
173
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
169
In the spring example, obtaining the necessary deformation equation was very
straightforward. However, for other situations, the deformation relations may not be as
easy. A more structured approach may be necessary. Here we will show two basic procedures for general statically indeterminate problems.
Procedure 1
1 Choose the redundant reaction(s). There may be alternative choices (See Example
414).
2 Write the equations of static equilibrium for the remaining reactions in terms of
the applied loads and the redundant reaction(s) of step 1.
3 Write the deection equation(s) for the point(s) at the locations of the redundant
reaction(s) of step 1 in terms of the applied loads and the redundant reaction(s)
of step 1. Normally the deection(s) is (are) zero. If a redundant reaction is a
moment, the corresponding deection equation is a rotational deection equation.
4 The equations from steps 2 and 3 can now be solved to determine the reactions.
In step 3 the deection equations can be solved in any of the standard ways. Here we will
demonstrate the use of superposition and Castiglianos theorem on a beam problem.
EXAMPLE 414
The indeterminate beam of Appendix Table A911 is reproduced in Fig. 416.
Determine the reactions using procedure 1.
Solution
The reactions are shown in Fig. 416b. Without R2 the beam is a statically determinate
cantilever beam. Without M1 the beam is a statically determinate simply supported
beam. In either case, the beam has only one redundant support. We will rst solve this
problem using superposition, choosing R2 as the redundant reaction. For the second
solution, we will use Castiglianos theorem with M1 as the redundant reaction.
Solution 1
1
2
Choose R2 at B to be the redundant reaction.
Using static equilibrium equations solve for R1 and M1 in terms of F and R2 .
This results in
R1 = F R2
3
Fl
R2 l
2
(1)
Write the deection equation for point B in terms of F and R2 . Using
superposition of Table A91 with F = R2 , and Table A92 with a = l /2,
the deection of B, at x = l , is
δB =
Figure 416
M1 =
F (l /2)2
R2 l 2
(l 3l ) +
6E I
6E I
y
l
3l
2
=
R2 l 3
5 Fl 3
=0
3E I
48 E I
(2)
y
F
l
l
2
F
A
O
(a)
B
A
B
x
x
O
M1
R1
x
ˆ
(b)
R2
174
170
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
4
Equation (2) can be solved for R2 directly. This yields
Answer
5F
16
R2 =
(3)
Next, substituting R2 into Eqs. (1) completes the solution, giving
Answer
R1 =
11 F
16
M1 =
3 Fl
16
(4)
Note that the solution agrees with what is given in Table A911.
Solution 2
1
2
Choose M1 at O to be the redundant reaction.
Using static equilibrium equations solve for R1 and R2 in terms of F and M1 .
This results in
R1 =
3
M1
F
+
2
l
R2 =
M1
F
2
l
(5)
Since M1 is the redundant reaction at O, write the equation for the angular
deection at point O. From Castiglianos theorem this is
θO =
U
M1
(6)
We can apply Eq. (425), using the variable x as shown in Fig. 416b. However, simˆ
pler terms can be found by using a variable x that starts at B and is positive to the left.
With this and the expression for R2 from Eq. (5) the moment equations are
M=
F
M1
2
l
x
ˆ
M=
M1
F
2
l
x F x
ˆ
ˆ
l
2
(7)
l
x l
ˆ
2
(8)
0x
ˆ
l
2
For both equations
M
x
ˆ
=
M1
l
(9)
Substituting Eqs. (7) to (9) in Eq. (6), using the form of Eq. (425) where Fi = M1 , gives
θO =
U
1
=
M1
EI
l /2
0
F
M1
2
l
F x
ˆ
l
2
x
ˆ
x
ˆ
l
x
ˆ
l
l
dx +
ˆ
F
M1
2
l
l /2
x
ˆ
dx = 0
ˆ
Canceling 1/ E I l , and combining the rst two integrals, simplies this quite readily to
M1
F
2
l
l
0
l
x2 dx F
ˆˆ
l /2
x
ˆ
l
x dx = 0
ˆˆ
2
Integrating gives
F
M1
2
l
l3
F3
l
3
3
l
2
3
+
Fl 2
l
4
l
2
2
=0
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
175
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
171
which reduces to
M1 =
4
3 Fl
16
(10)
Substituting Eq. (10) into (5) results in
R1 =
11 F
16
R2 =
5F
16
(11)
which again agrees with Table A911.
For some problems even procedure 1 can be a task. Procedure 2 eliminates some
tricky geometric problems that would complicate procedure 1. We will describe the procedure for a beam problem.
Procedure 2
1 Write the equations of static equilibrium for the beam in terms of the applied
loads and unknown restraint reactions.
2 Write the deection equation for the beam in terms of the applied loads and
unknown restraint reactions.
3 Apply boundary conditions consistent with the restraints.
4 Solve the equations from steps 1 and 3.
EXAMPLE 415
The rods A D and C E shown in Fig. 417a each have a diameter of 10 mm. The secondarea moment of beam A BC is I = 62.5(103 ) mm4 . The modulus of elasticity of the
material used for the rods and beam is E = 200 GPa. The threads at the ends of the rods
are single-threaded with a pitch of 1.5 mm. The nuts are rst snugly t with bar A BC
horizontal. Next the nut at A is tightened one full turn. Determine the resulting tension
in each rod and the deections of points A and C.
Solution
There is a lot going on in this problem; a rod shortens, the rods stretch in tension, and
the beam bends. Lets try the procedure!
1
The free-body diagram of the beam is shown in Fig. 417b. Summing forces,
and moments about B, gives
FB FA FC = 0
4 FA 3 FC = 0
Figure 417
200
A
Dimensions in mm.
(1)
(2)
150
FA
B
C
200
A
150
B
C
x
FB
600
800
D
E
(a)
FC
(b) Free-body diagram of beam ABC
176
172
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
2
Using singularity functions, we nd the moment equation for the beam is
M = FA x + FB x 0.2
1
where x is in meters. Integration yields
dy
FA
= x2 +
dx
2
FA
E I y = x3 +
6
EI
3
FB
x 0.2 2 + C1
2
FB
x 0.2 3 + C1 x + C2
6
(3)
The term E I = 200(109 ) 62.5(109 ) = 1.25(104 ) N · m2 .
The upward deection of point A is ( Fl / AE ) AD N p, where the rst term
is the elastic stretch of A D , N is the number of turns of the nut, and p is the
pitch of the thread. Thus, the deection of A is
FA (0.6)
(1)(0.0015)
yA = π
(0.010)2 (200)(109 )
4
(4)
= 3.8197(108 ) FA 1.5(103 )
The upward deection of point C is ( Fl / AE )C E , or
FC (0.8)
yC = π
= 5.093(108 ) FC
(0.010)2 (200)(109 )
4
(5)
Equations (4) and (5) will now serve as the boundary conditions for Eq. (3). At
x = 0, y = y A . Substituting Eq. (4) into (3) with x = 0 and E I = 1.25(104 ), noting
that the singularity function is zero for x = 0, gives
4.7746(104 ) FA + C2 = 18.75
(6)
At x = 0.2 m, y = 0, and Eq. (3) yields
1.3333(103 ) FA + 0.2C1 + C2 = 0
(7)
At x = 0.35 m, y = yC . Substituting Eq. (5) into (3) with x = 0.35 m and E I =
1.25(104 ) gives
7.1458(103 ) FA + 5.625(104 ) FB 6.3662(104 ) FC + 0.35C1 + C2 = 0
Equations (1), (2), (6), (7), and (8) are ve equations in
Written in matrix form, they are
1
1
1
0
4
0
3
0
0
0
0
4.7746(104 )
1.3333(103 )
0
0
0.2
7.1458(103 ) 5.625(104 ) 6.3662(104 ) 0.35
Solving these equations yields
Answer
FA = 2988 N
2
C1 = 106.54 N · m
FB = 6971 N
3
C2 = 17.324 N · m
(8)
FA , FB , FC , C1 , and C2 .
0 FA 0
0 FB 0
1 FC = 18.75
1 C1 0
1
C2
0
FC = 3983 N
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
177
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
173
Equation (3) can be reduced to
y = (39.84x 3 92.95 x 0.2 3 8.523x + 1.386)(103 )
At x = 0, y = y A = 1.386(103 ) m = 1.386 mm.
Answer
Answer
At x = 0.35 m, y = yC = [39.84(0.35)3 92.95(0.35 0.2)3 8.523(0.35)
+ 1.386](103 ) = 0.203(103 ) m = 0.203 mm
Note that we could have easily incorporated the stiffness of the support at B if we
were given a spring constant.
411
Compression MembersGeneral
The analysis and design of compression members can differ signicantly from that of
members loaded in tension or in torsion. If you were to take a long rod or pole, such as
a meterstick, and apply gradually increasing compressive forces at each end, nothing
would happen at rst, but then the stick would bend (buckle), and nally bend so much
as to fracture. Try it. The other extreme would occur if you were to saw off, say, a 5-mm
length of the meterstick and perform the same experiment on the short piece. You would
then observe that the failure exhibits itself as a mashing of the specimen, that is, a
simple compressive failure. For these reasons it is convenient to classify compression
members according to their length and according to whether the loading is central or
eccentric. The term column is applied to all such members except those in which failure would be by simple or pure compression. Columns can be categorized then as:
1
2
3
4
Long columns with central loading
Intermediate-length columns with central loading
Columns with eccentric loading
Struts or short columns with eccentric loading
Classifying columns as above makes it possible to develop methods of analysis and
design specic to each category. Furthermore, these methods will also reveal whether or
not you have selected the category appropriate to your particular problem. The four
sections that follow correspond, respectively, to the four categories of columns listed above.
412
Long Columns with Central Loading
Figure 418 shows long columns with differing end (boundary) conditions. If the axial
force P shown acts along the centroidal axis of the column, simple compression of the
member occurs for low values of the force. However, under certain conditions, when P
reaches a specic value, the column becomes unstable and bending as shown in Fig.
418 develops rapidly. This force is determined by writing the bending deection equation for the column, resulting in a differential equation where when the boundary conditions are applied, results in the critical load for unstable bending.9 The critical force
for the pin-ended column of Fig. 418a is given by
Pcr =
π2E I
l2
9
See F. P. Beer, E. R. Johnston, Jr., and J. T. DeWolf, Mechanics of Materials, 4th ed., McGraw-Hill,
New York, 2006, pp. 610613.
(439)
174
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
(a) Both ends rounded or
pivoted; (b) both ends xed;
(c) one end free and one end
xed; (d ) one end rounded
and pivoted, and one end
xed.
P
P
Figure 418
P
P
y
l
4
A
l
2
l
0.707l
178
l
l
4
l
A
B
x
(a) C
1
(b) C
4
(c) C
1
4
(d ) C
2
which is called the Euler column formula. Equation (439) can be extended to apply to
other end-conditions by writing
Pcr =
Cπ2 E I
l2
(440)
where the constant C depends on the end conditions as shown in Fig. 418.
Using the relation I = Ak 2 , where A is the area and k the radius of gyration,
enables us to rearrange Eq. (440) into the more convenient form
Cπ2 E
Pcr
=
A
(l / k )2
(441)
where l / k is called the slenderness ratio. This ratio, rather than the actual column
length, will be used in classifying columns according to length categories.
The quantity Pcr / A in Eq. (441) is the critical unit load. It is the load per unit area
necessary to place the column in a condition of unstable equilibrium. In this state any
small crookedness of the member, or slight movement of the support or load, will cause
the column to begin to collapse. The unit load has the same units as strength, but this is
the strength of a specic column, not of the column material. Doubling the length of a
member, for example, will have a drastic effect on the value of Pcr / A but no effect at
all on, say, the yield strength Sy of the column material itself.
Equation (441) shows that the critical unit load depends only upon the modulus
of elasticity and the slenderness ratio. Thus a column obeying the Euler formula made
of high-strength alloy steel is no stronger than one made of low-carbon steel, since E is
the same for both.
The factor C is called the end-condition constant, and it may have any one of the
theoretical values 1 , 1, 2, and 4, depending upon the manner in which the load is
4
applied. In practice it is difcult, if not impossible, to x the column ends so that the
factor C = 2 or C = 4 would apply. Even if the ends are welded, some deection will
occur. Because of this, some designers never use a value of C greater than unity.
However, if liberal factors of safety are employed, and if the column load is accurately
known, then a value of C not exceeding 1.2 for both ends xed, or for one end rounded
and one end xed, is not unreasonable, since it supposes only partial xation. Of course,
the value C = 1 must always be used for a column having one end xed and one end
4
free. These recommendations are summarized in Table 42.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Table 42
179
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
175
End-Condition Constant C
Theoretical
Value
Conservative
Value
Recommended
Value*
Fixed-free
1
4
1
4
1
4
Rounded-rounded
1
1
1
Fixed-rounded
2
1
1.2
Fixed-xed
End-Condition Constants
for Euler Columns [to Be
Used with Eq. (440)]
Column End
Conditions
4
1
1.2
*To be used only with liberal factors of safety when the column load is accurately known.
Figure 419
P
Euler curve plotted using
Eq. (440) with C = 1.
Q
Unit load
Pcr
A
Sy
Parabolic
curve
T
Euler curve
R
l
kQ
l
k
1
l
Slenderness ratio k
When Eq. (441) is solved for various values of the unit load Pcr / A in terms of the
slenderness ratio l / k , we obtain the curve PQR shown in Fig. 419. Since the yield
strength of the material has the same units as the unit load, the horizontal line through
Sy and Q has been added to the gure. This would appear to make the gure cover the
entire range of compression problems from the shortest to the longest compression
member. Thus it would appear that any compression member having an l / k value less
than (l / k ) Q should be treated as a pure compression member while all others are to be
treated as Euler columns. Unfortunately, this is not true.
In the actual design of a member that functions as a column, the designer will be
aware of the end conditions shown in Fig. 418, and will endeavor to congure the ends,
using bolts, welds, or pins, for example, so as to achieve the required ideal end conditions. In spite of these precautions, the result, following manufacture, is likely to contain
defects such as initial crookedness or load eccentricities. The existence of such defects
and the methods of accounting for them will usually involve a factor-of-safety approach
or a stochastic analysis. These methods work well for long columns and for simple
compression members. However, tests show numerous failures for columns with
slenderness ratios below and in the vicinity of point Q, as shown in the shaded area in
Fig. 419. These have been reported as occurring even when near-perfect geometric
specimens were used in the testing procedure.
A column failure is always sudden, total, unexpected, and hence dangerous. There
is no advance warning. A beam will bend and give visual warning that it is overloaded,
but not so for a column. For this reason neither simple compression methods nor the
180
176
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Euler column equation should be used when the slenderness ratio is near (l / k ) Q . Then
what should we do? The usual approach is to choose some point T on the Euler curve
of Fig. 419. If the slenderness ratio is specied as (l / k )1 corresponding to point T,
then use the Euler equation only when the actual slenderness ratio is greater than
(l / k )1 . Otherwise, use one of the methods in the sections that follow. See Examples
417 and 418.
Most designers select point T such that Pcr / A = Sy /2. Using Eq. (440), we nd
the corresponding value of (l / k )1 to be
l
k
413
1
2π 2 C E
Sy
=
1/2
(442)
Intermediate-Length Columns with Central Loading
Over the years there have been a number of column formulas proposed and used for the
range of l / k values for which the Euler formula is not suitable. Many of these are based
on the use of a single material; others, on a so-called safe unit load rather than the critical value. Most of these formulas are based on the use of a linear relationship between
the slenderness ratio and the unit load. The parabolic or J. B. Johnson formula now
seems to be the preferred one among designers in the machine, automotive, aircraft, and
structural-steel construction elds.
The general form of the parabolic formula is
l
Pcr
=ab
A
k
2
(a)
where a and b are constants that are evaluated by tting a parabola to the Euler curve
of Fig. 419 as shown by the dashed line ending at T . If the parabola is begun at Sy ,
then a = Sy . If point T is selected as previously noted, then Eq. (a) gives the value of
(l / k )1 and the constant b is found to be
b=
Sy
2π
2
1
CE
(b)
Upon substituting the known values of a and b into Eq. (a), we obtain, for the parabolic
equation,
Pcr
= Sy
A
414
Sy l
2π k
2
1
CE
l
k
l
k
(443)
1
Columns with Eccentric Loading
We have noted before that deviations from an ideal column, such as load eccentricities
or crookedness, are likely to occur during manufacture and assembly. Though these
deviations are often quite small, it is still convenient to have a method of dealing with
them. Frequently, too, problems occur in which load eccentricities are unavoidable.
Figure 420a shows a column in which the line of action of the column forces is
separated from the centroidal axis of the column by the eccentricity e. This problem is
developed by using Eq. (412) and the free-body diagram of Fig. 420b.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
181
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
177
x
Figure 420
P
Notation for an eccentrically
loaded column.
A
x
P
l
M
y
x
y
y
O
P
Pe
e
P
(a )
(b)
This results in the differential equation
d2 y
Pe
P
y=
+
2
dx
EI
EI
(a)
The solution of Eq. (a), for the boundary conditions that y
y
[(
e tan
lP
2 EI
)(
sin
P
x
EI
)
(
cos
0 at x
P
x
EI
0, l is
)]
1
(b)
By substituting x = l /2 in Eq. (b) and using a trigonometric identity, we obtain
[(
e sec
Pl
EI 2
)]
(444)
1
The maximum bending moment also occurs at midspan and is
Mmax = P (e + δ) = Pe sec
l
2
P
EI
(445)
The magnitude of the maximum compressive stress at midspan is found by superposing
the axial component and the bending component. This gives
σc =
Mc
P
Mc
P
=
A
I
A
Ak 2
(c)
Substituting Mmax from Eq. (445) yields
σc =
ec
P
l
1 + 2 sec
A
k
2k
P
EA
(446)
By imposing the compressive yield strength Syc as the maximum value of σc , we can
write Eq. (446) in the form
Syc
P
=
2 ) sec[(l /2k ) P / AE ]
A
1 + (ec/ k
(447)
This is called the secant column formula. The term ec/ k 2 is called the eccentricity
ratio. Figure 421 is a plot of Eq. (447) for a steel having a compressive (and tensile)
182
178
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Figure 421
ec/k 2 = 0.1
Unit load P/A
Comparison of secant and
Euler equations for steel with
Sy = 40 kpsi.
Sy
0.3
0.6
Euler's curve
1.0
0
50
100
150
200
250
Slenderness ratio l / k
yield strength of 40 kpsi. Note how the P / A contours asymptotically approach the
Euler curve as l / k increases.
Equation (447) cannot be solved explicitly for the load P . Design charts, in the
fashion of Fig. 421, can be prepared for a single material if much column design
is to be done. Otherwise, a root-nding technique using numerical methods must
be used.
EXAMPLE 416
Solution
Develop specic Euler equations for the sizes of columns having
(a) Round cross sections
(b) Rectangular cross sections
(a) Using A = π d 2 /4 and k =
gives
Answer
I / A = [(π d 4 /64)/(π d 2 /4)]1/2 = d /4 with Eq. (441)
d=
1/4
64 Pcrl 2
π 3C E
(448)
(b) For the rectangular column, we specify a cross section h × b with the
restriction that h b . If the end conditions are the same for buckling in both directions,
then buckling will occur in the direction of the least thickness. Therefore
I=
bh 3
12
A = bh
k2 = I / A =
h2
12
Substituting these in Eq. (441) gives
Answer
b=
12 Pcrl 2
π 2 C Eh 3
(449)
Note, however, that rectangular columns do not generally have the same end conditions
in both directions.
I. Basics
EXAMPLE 417
Solution
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
183
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
179
Specify the diameter of a round column 1.5 m long that is to carry a maximum load
estimated to be 22 kN. Use a design factor n d = 4 and consider the ends as pinned
(rounded). The column material selected has a minimum yield strength of 500 MPa and
a modulus of elasticity of 207 GPa.
We shall design the column for a critical load of
Pcr = n d P = 4(22) = 88 kN
Then, using Eq. (448) with C = 1 (see Table 42) gives
64 Pcrl 2
π 3C E
d=
1/4
=
64(88)(1.5)2
π 3 (1)(207)
1/4
1/4
103
109
(103 ) = 37.48 mm
Table A17 shows that the preferred size is 40 mm. The slenderness ratio for this size is
l
1.5(103 )
l
=
=
= 150
k
d /4
40/4
To be sure that this is an Euler column, we use Eq. (548) and obtain
l
k
1
=
2π 2 C E
Sy
1/2
=
2π 2 (1)(207)
500
1/2
109
106
1/2
= 90.4
which indicates that it is indeed an Euler column. So select
Answer
EXAMPLE 418
Solution
d = 40 mm
Repeat Ex. 416 for J. B. Johnson columns.
(a) For round columns, Eq. (443) yields
Sy l 2
Pcr
d=2
+2
π Sy
π CE
Answer
1/2
(450)
(b) For a rectangular section with dimensions h b , we nd
Answer
EXAMPLE 419
b=
Pcr
h Sy
3l 2 Sy
1 2
π C Eh 2
hb
(451)
Choose a set of dimensions for a rectangular link that is to carry a maximum compressive load of 5000 lbf. The material selected has a minimum yield strength of 75 kpsi
and a modulus of elasticity E = 30 Mpsi. Use a design factor of 4 and an end condition constant C = 1 for buckling in the weakest direction, and design for (a) a length
of 15 in, and (b) a length of 8 in with a minimum thickness of 1 in.
2
184
180
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
Solution
(a) Using Eq. (441), we nd the limiting slenderness ratio to be
l
k
1
=
2π 2 C E
Sy
1/2
=
2π 2 (1)(30)(106 )
75(10)3
1/2
= 88.9
By using Pcr = n d P = 4(5000) = 20 000 lbf, Eqs. (449) and (451) are solved, using
various values of h, to form Table 43. The table shows that a cross section of 5 by 3
8
4
in, which is marginally suitable, gives the least area.
(b) An approach similar to that in part (a) is used with l = 8 in. All trial computations are found to be in the J. B. Johnson region of l / k values. A minimum area occurs
when the section is a near square. Thus a cross section of 1 by 3 in is found to be suit2
4
able and safe.
Table 43
h
Table Generated to
Solve Ex. 419, part (a)
b
A
l/k
Type
Eq. No.
0.375
3.46
1.298
139
Euler
(449)
0.500
1.46
0.730
104
Euler
(449)
0.625
415
0.475
83
Johnson
(451)
1.03
0.579
92
Euler
(449)
Struts or Short Compression Members
A short bar loaded in pure compression by a force P acting along the centroidal axis
will shorten in accordance with Hookes law, until the stress reaches the elastic limit of
the material. At this point, permanent set is introduced and usefulness as a machine
member may be at an end. If the force P is increased still more, the material either
becomes barrel-like or fractures. When there is eccentricity in the loading, the elastic
limit is encountered at smaller loads.
A strut is a short compression member such as the one shown in Fig. 422. The
magnitude of the maximum compressive stress in the x direction at point B in an intermediate section is the sum of a simple component P / A and a exural component
Mc/ I ; that is,
P
x
e
B
0.76
0.5625
l
c
y
P
Figure 422
Eccentrically loaded strut.
σc =
ec
Mc
P
Pec A
P
P
1+ 2
+
=
+
=
A
I
A
IA
A
k
(452)
where k = ( I / A)1/2 and is the radius of gyration, c is the coordinate of point B, and e
is the eccentricity of loading.
Note that the length of the strut does not appear in Eq. (452). In order to use the
equation for design or analysis, we ought, therefore, to know the range of lengths for
which the equation is valid. In other words, how long is a short member?
The difference between the secant formula Eq. (447) and Eq. (452) is that the
secant equation, unlike Eq. (452), accounts for an increased bending moment due to
bending deection. Thus the secant equation shows the eccentricity to be magnied by
the bending deection. This difference between the two formulas suggests that one way
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
185
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
181
of differentiating between a secant column and a strut, or short compression member,
is to say that in a strut, the effect of bending deection must be limited to a certain small
percentage of the eccentricity. If we decide that the limiting percentage is to be 1 percent of e, then, from Eq. (444), the limiting slenderness ratio turns out to be
l
k
= 0.282
2
AE
P
1/2
(453)
This equation then gives the limiting slenderness ratio for using Eq. (452). If the actual
slenderness ratio is greater than (l / k )2 , then use the secant formula; otherwise, use
Eq. (452).
EXAMPLE 420
Figure 423a shows a workpiece clamped to a milling machine table by a bolt tightened to a tension of 2000 lbf. The clamp contact is offset from the centroidal axis of the
strut by a distance e = 0.10 in, as shown in part b of the gure. The strut, or block, is
steel, 1 in square and 4 in long, as shown. Determine the maximum compressive stress
in the block.
Solution
First we nd A = bh = 1(1) = 1 in2 , I = bh 3 /12 = 1(1)3 /12 = 0.0833 in4 , k 2 =
I / A = 0.0833/1 = 0.0833 in2, and l / k = 4/(0.0833)1/2 = 13.9. Equation (453)
gives the limiting slenderness ratio as
l
k
2
= 0.282
1/2
AE
P
= 0.282
1(30)(106 )
1000
1/2
= 48.8
Thus the block could be as long as
l = 48.8k = 48.8(0.0833)1/2 = 14.1 in
before it need be treated by using the secant formula. So Eq. (452) applies and the
maximum compressive stress is
Answer
σc =
ec
P
1+ 2
A
k
=
0.1(0.5)
1000
1+
= 1600 psi
1
0.0833
Figure 423
P = 1000 lbf
A strut that is part of a
workpiece clamping assembly.
1-in square
4 in
0.10 in
P
(a )
(b)
186
182
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
416
Elastic Stability
Section 412 presented the conditions for the unstable behavior of long, slender
columns. Elastic instability can also occur in structural members other than columns.
Compressive loads/stresses within any long, thin structure can cause structural instabilities (buckling). The compressive stress may be elastic or inelastic and the instability
may be global or local. Global instabilities can cause catastrophic failure, whereas local
instabilities may cause permanent deformation and function failure but not a catastrophic failure. The buckling discussed in Sec. 412 was global instability. However,
consider a wide ange beam in bending. One ange will be in compression, and if thin
enough, can develop localized buckling in a region where the bending moment is a
maximum. Localized buckling can also occur in the web of the beam, where transverse
shear stresses are present at the beam centroid. Recall, for the case of pure shear stress
τ , a stress transformation will show that at 45 , a compressive stress of σ = τ exists.
If the web is sufciently thin where the shear force V is a maximum, localized buckling
of the web can occur. For this reason, additional support in the form of bracing is typically applied at locations of high shear forces.10
Thin-walled beams in bending can buckle in a torsional mode as illustrated in
Fig. 424. Here a cantilever beam is loaded with a lateral force, F. As F is increases
from zero, the end of the beam will deect in the negative y direction normally according to the bending equation, y = F L 3 /(3 E I ) . However, if the beam is long enough
and the ratio of b/h is sufciently small, there is a critical value of F for which the beam
will collapse in a twisting mode as shown. This is due to the compression in the bottom
bers of the beam which cause the bers to buckle sideways (z direction).
There are a great many other examples of unstable structural behavior, such as thinwalled pressure vessels in compression or with outer pressure or inner vacuum, thin-walled
open or closed members in torsion, thin arches in compression, frames in compression,
and shear panels. Because of the vast array of applications and the complexity of their
analyses, further elaboration is beyond the scope of this book. The intent of this section
is to make the reader aware of the possibilities and potential safety issues. The key issue
is that the designer should be aware that if any unbraced part of a structural member is
thin, and/or long, and in compression (directly or indirectly), the possibility of buckling
should be investigated.11
Figure 424
y
Torsional buckling of a
thin-walled beam in bending.
z
h
z
y
x
b
F
Figure 425
10
Finite-element representation of
ange buckling of a channel
in compression.
11
See S. P. Timoshenko and J. M. Gere, Theory of Elastic Stability, 2nd ed., McGraw-Hill, New York, 1961.
See also, Z. P. Bazant and L. Cedolin, Stability of Structures, Oxford University Press, New York, 1991.
See C. G. Salmon and J. E. Johnson, Steel Structures: Design and Behavior, 4th ed., Harper, Collins,
New York, 1996.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
187
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
183
For unique applications, the designer may need to revert to a numerical solution
such as using nite elements. Depending on the application and the nite-element code
available, an analysis can be performed to determine the critical loading (see Fig. 425).
417
Shock and Impact
Impact refers to the collision of two masses with initial relative velocity. In some cases
it is desirable to achieve a known impact in design; for example, this is the case in the
design of coining, stamping, and forming presses. In other cases, impact occurs because
of excessive deections, or because of clearances between parts, and in these cases it is
desirable to minimize the effects. The rattling of mating gear teeth in their tooth spaces
is an impact problem caused by shaft deection and the clearance between the teeth.
This impact causes gear noise and fatigue failure of the tooth surfaces. The clearance
space between a cam and follower or between a journal and its bearing may result in
crossover impact and also cause excessive noise and rapid fatigue failure.
Shock is a more general term that is used to describe any suddenly applied force or
disturbance. Thus the study of shock includes impact as a special case.
Figure 426 represents a highly simplied mathematical model of an automobile
in collision with a rigid obstruction. Here m 1 is the lumped mass of the engine. The
displacement, velocity, and acceleration are described by the coordinate x1 and its
time derivatives. The lumped mass of the vehicle less the engine is denoted by m 2 , and
its motion by the coordinate x2 and its derivatives. Springs k1 , k2 , and k3 represent the
linear and nonlinear stiffnesses of the various structural elements that compose
the vehicle. Friction and damping can and should be included, but is not shown in this
model. The determination of the spring rates for such a complex structure will almost
certainly have to be performed experimentally. Once these valuesthe ks, ms, damping
and frictional coefcientsare obtained, a set of nonlinear differential equations can be
written and a computer solution obtained for any impact velocity.
Figure 427 is another impact model. Here mass m 1 has an initial velocity v and is
just coming into contact with spring k1 . The part or structure to be analyzed is represented by mass m 2 and spring k2 . The problem facing the designer is to nd the
maximum deection of m 2 and the maximum force exerted by k2 against m 2 . In the
analysis it doesnt matter whether k1 is fastened to m 1 or to m 2 , since we are interested
Figure 426
x2
x1
Two-degree-of-freedom
mathematical model of an
automobile in collision with a
rigid obstruction.
k1
k2
m1
m2
k3
Figure 427
x1
x2
k1
m1
k2
m2
188
184
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
only in a solution up to the point in time for which x2 reaches a maximum. That is, the
solution for the rebound isnt needed. The differential equations are not difcult to
derive. They are
m 1 x 1 + k1 ( x 1 x 2 ) = 0
¨
m 2 x 2 + k2 x 2 k1 ( x 1 x 2 ) = 0
¨
(454)
The analytical solution of Eq. pair (454) is harmonic and is studied in a course on
mechanical vibrations.12 If the values of the ms and ks are known, the solution can be
obtained easily using a program such as MATLAB.
418
Suddenly Applied Loading
A simple case of impact is illustrated in Fig. 428a. Here a weight W falls a distance h
and impacts a cantilever of stiffness EI and length l. We want to nd the maximum
deection and the maximum force exerted on the beam due to the impact.
Figure 428b shows an abstract model of the system. Using Table A91, we nd
the spring rate to be k = F / y = 3 E I / l 3 . The beam mass and damping can be accounted
for, but for this example will be considered negligible. The origin of the coordinate y
corresponds to the point where the weight is released. Two free-body diagrams, shown
in Fig. 428c and d are necessary. The rst corresponds to y h , and the second when
y > h to account for the spring force.
For each of these free-body diagrams we can write Newtons law by stating that the
¨
inertia force (W /g ) y is equal to the sum of the external forces acting on the weight. We
then have
W
y=W
¨
g
yh
W
y = k ( y h ) + W
¨
g
y>h
(a)
We must also include in the mathematical statement of the problem the knowledge that the
weight is released with zero initial velocity. Equation pair (a) constitutes a set of piecewise
differential equations. Each equation is linear, but each applies only for a certain range of y .
Figure 428
(a) A weight free to fall a
distance h to free end of a
beam. (b) Equivalent spring
model. (c) Free body of
weight during fall. (d ) Free
body of weight during arrest.
W
W
y
h
y
EI, l
y
W
h
y
W
W
k
(a)
12
(b )
W
k ( y h)
(c) y
h
(d ) y
h
See William T. Thomson and Marie Dillon Dahleh, Theory of Vibrations with Applications, Prentice Hall,
5th ed., 1998.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
© The McGrawHill
Companies, 2008
189
Deection and Stiffness
I. Basics
185
4. Deflection and Stiffness
The solution to the set is valid for all values of t, but we are interested in values of y only
up until the time that the spring or structure reaches its maximum deection.
The solution to the rst equation in the set is
y=
gt 2
2
(455)
yh
and you can verify this by direct substitution. Equation (455) is no longer valid after
y = h ; call this time t1 . Then
t1 =
(b)
2h / g
Differentiating Eq. (455) to get the velocity gives
y = gt
˙
(c)
yh
and so the velocity of the weight at t = t1 is
y1 = gt1 = g 2h /g =
˙
(d)
2gh
Having moved from y = 0 to y = h , we then need to solve the second equation of
the set (a). It is convenient to dene a new time t = t t1 . Thus t = 0 at the instant
the weight strikes the spring. Applying your knowledge of differential equations, you
should nd the solution to be
y = A cos ωt + B sin ωt + h +
W
k
(e)
y>h
where
ω=
kg
W
(456)
is the circular frequency of vibration. The initial conditions for the beam motion at
t = 0, are y = h and y = y1 = 2gh (neglecting the mass of the beam, the velocity is
˙
˙
the same as the weight at t = 0). Substituting the initial conditions into Eq. (e) yields
A and B, and Eq. (e) becomes
W
W
2W h
cos ωt +
sin ωt + h +
y>h
(f)
k
k
k
2W h / k = C sin φ , where it can be shown that
Let W / k = C cos φ and
C = [(W / k )2 + 2W h / k ]1/2 . Substituting this into Eq. ( f ) and using a trigonometric
identity gives
y=
y=
W
k
2
+
2W h
k
1/2
cos[ωt φ ] + h +
W
k
y>h
(457)
The maximum deection of the spring (beam) occurs when the cosine term in
Eq. (457) is unity. We designate this as δ and, after rearranging, nd it to be
δ = ymax h =
W
W
1+
+
k
k
1/2
2hk
W
(458)
The maximum force acting on the beam is now found to be
F = kδ = W + W 1 +
2hk
W
1/2
(459)
190
186
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Note, in this equation, that if h = 0, then F = 2W . This says that when the weight is
released while in contact with the spring but is not exerting any force on the spring, the
largest force is double the weight.
Most systems are not as ideal as those explored here, so be wary about using these
relations for nonideal systems.
PROBLEMS
41
Structures can often be considered to be composed of a combination of tension and torsion
members and beams. Each of these members can be analyzed separately to determine its
force-deflection relationship and its spring rate. It is possible, then, to obtain the deflection of
a structure by considering it as an assembly of springs having various series and parallel relationships.
(a) What is the overall spring rate of three springs in series?
(b) What is the overall spring rate of three springs in parallel?
(c) What is the overall spring rate of a single spring in series with a pair of parallel springs?
42
The gure shows a torsion bar O A xed at O , simply supported at A, and connected to a cantilever A B . The spring rate of the torsion bar is k T , in newton-meters per radian, and that of the
cantilever is kC , in newtons per meter. What is the overall spring rate based on the deection y at
point B ?
F
O
B
L
Problem 42
l
A
y
R
43
A torsion-bar spring consists of a prismatic bar, usually of round cross section, that is twisted
at one end and held fast at the other to form a stiff spring. An engineer needs a stiffer one than
usual and so considers building in both ends and applying the torque somewhere in the central portion of the span, as shown in the figure. If the bar is uniform in diameter, that is, if
d = d1 = d2 , investigate how the allowable angle of twist, the largest torque, and the spring
rate depend on the location x at which the torque is applied. Hint: Consider two springs in
parallel.
d2
T
Problem 43
d1
l
x
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
191
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
187
44
An engineer is forced by geometric considerations to apply the torque on the spring of Prob. 43
at the location x = 0.2l . For a uniform-diameter spring, this would cause the long leg of the span
to be underutilized when both legs have the same diameter. If the diameter of the long leg is
reduced sufciently, the shear stress in the two legs can be made equal. How would this change
affect the allowable angle of twist, the largest torque, and the spring rate?
45
A bar in tension has a circular cross section and includes a conical portion of length l , as
shown. The task is to find the spring rate of the entire bar. Equation (44) is useful for the
outer portions of diameters d1 and d2 , but a new relation must be derived for the tapered section. If α is the apex half-angle, as shown, show that the spring rate of the tapered portion of
the shaft is
k=
Problem 45
d2
E A1
l
1+
2l
tan α
d1
dl
l
46
When a hoisting cable is long, the weight of the cable itself contributes to the elongation. If a
cable has a weight per unit length of w, a length of l , and a load P attached to the free end, show
that the cable elongation is
δ=
Pl
wl 2
+
AE
2 AE
47
Use integration to verify the deection equation given for the uniformly loaded cantilever beam
of appendix Table A93.
48
Use integration to verify the deection equation given for the end moment loaded cantilever beam
of appendix Table A94.
49
When an initially straight beam sags under transverse loading, the ends contract because the
neutral surface of zero strain neither extends nor contracts. The length of the deflected neutral surface is the same as the original beam length l . Consider a segment of the initially
straight beam s . After bending, the x-direction component is shorter than s , namely, x .
The contraction is s x , and these summed for the entire beam gives the end contraction λ.
Show that
.1
λ=
2
l
0
dy
dx
2
dx
410
Using the results of Prob. 49, determine the end contraction of the uniformly loaded cantilever
beam of appendix Table A93.
411
Using the results of Prob. 49, determine the end contraction of the uniformly loaded simplysupported beam of appendix Table A97. Assume the left support cannot deect in the x direction,
whereas the right support can.
412
1
The gure shows a cantilever consisting of steel angles size 4 × 4 × 2 in mounted back to back.
Using superposition, nd the deection at B and the maximum stress in the beam.
192
188
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
y
10 ft
600 lbf
Problem 412
7 ft
50 lbf/ft
x
O
B
A
413
A simply supported beam loaded by two forces is shown in the gure. Select a pair of structural steel channels mounted back to back to support the loads in such a way that the deec1
tion at midspan will not exceed 16 in and the maximum stress will not exceed 6 kpsi. Use
superposition.
y
800 lbf 600 lbf
Problem 413
3 ft
5 ft
2 ft
C
O
A
414
x
B
Using superposition, nd the deection of the steel shaft at A in the gure. Find the deection at
midspan. By what percentage do these two values differ?
y
400 mm
600 mm
1500 N
Problem 414
2 kN/m
B
O
x
A
40 mm-dia. shaft
415
A rectangular steel bar supports the two overhanging loads shown in the gure. Using superposition,
nd the deection at the ends and at the center.
y
250
250
500
500 N
500 N
Problem 415
Dimensions in millimeters.
A
B
x
C
O
Bar, b = 9, h = 35
416
Using the formulas in Appendix Table A9 and superposition, nd the deection of the cantilever
at B if I = 13 in4 and E = 30 Mpsi.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
193
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
189
y
400 lbf
400 lbf
Problem 416
3 ft
3 ft
O
x
A
B
417
The cantilever shown in the gure consists of two structural-steel channels size 3 in, 5.0 lbf/ft.
Using superposition, nd the deection at A.
y
48 in
220 lbf
Problem 417
10 lbf/in
x
A
O
418
Using superposition, determine the maximum deection of the beam shown in the gure. The
material is carbon steel.
y
10 in
10 in
10 in
10 in
120 lbf
85 lbf
85 lbf
Problem 418
D
O
A
B
x
C
2-in-dia. shaft
419
Illustrated is a rectangular steel bar with simple supports at the ends and loaded by a force F at
the middle; the bar is to act as a spring. The ratio of the width to the thickness is to be about
b = 16h , and the desired spring scale is 2400 lbf/in.
(a) Find a set of cross-section dimensions, using preferred sizes.
(b) What deection would cause a permanent set in the spring if this is estimated to occur at a
normal stress of 90 kpsi?
F
A
b
Problem 419
A
4 ft
420
h
Section AA
1
Illustrated in the gure is a 1 2 -in-diameter steel countershaft that supports two pulleys. Pulley A
delivers power to a machine causing a tension of 600 lbf in the tight side of the belt and 80 lbf in
194
190
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
the loose side, as indicated. Pulley B receives power from a motor. The belt tensions on pulley B
have the relation T1 = 0.125T2 . Find the deection of the shaft in the z direction at pulleys A and
B . Assume that the bearings constitute simple supports.
y
12 in
21 in
O
A
15 in
Problem 420
T2
z
600 lbf
T1
80 lbf
C
9-in dia.
B
1 1 -in dia.
2
x
12-in dia.
421
The gure shows a steel countershaft that supports two pulleys. Pulley C receives power from a
motor producing the belt tensions shown. Pulley A transmits this power to another machine
through the belt tensions T1 and T2 such that T1 = 8T2 .
y
9 in
O
T2
z
T1
Problem 421
A
11 in
1 1 -in dia.
4
12 in
B
10-in dia.
C
16-in dia.
50 lbf
x
400 lbf
(a) Find the deection of the overhanging end of the shaft, assuming simple supports at the
bearings.
(b) If roller bearings are used, the slope of the shaft at the bearings should not exceed 0.06 for
1
good bearing life. What shaft diameter is needed to conform to this requirement? Use 8 -in
increments in any iteration you may make. What is the deection at pulley C now?
422
The structure of a diesel-electric locomotive is essentially a composite beam supporting a
deck. Above the deck are mounted the diesel prime mover, generator or alternator, radiators,
switch gear, and auxiliaries. Beneath the deck are found fuel and lubricant tanks, air reservoirs, and small auxiliaries. This assembly is supported at bolsters by the trucks that house the
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
195
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
191
traction motors and brakes. This equipment is distributed as uniformly as possible in the span
between the bolsters. In an approximate way, the loading can be viewed as uniform between
the bolsters and simply supported. Because the hoods that shield the equipment from the
weather have many rectangular access doors, which are mass-produced, it is important that
the hood structure be level and plumb and sit on a flat deck. Aesthetics plays a role too. The
center sill beam has a second moment of area of I = 5450 in4 , the bolsters are 36 ft apart, and
the deck loading is 5000 lbf/ft.
(a) What is the camber of the curve to which the deck will be built in order that the service-ready
locomotive will have a at deck?
(b) What equation would you give to locate points on the curve of part (a)?
423
The designer of a shaft usually has a slope constraint imposed by the bearings used. This limit
will be denoted as ξ . If the shaft shown in the gure is to have a uniform diameter d except in
the locality of the bearing mounting, it can be approximated as a uniform beam with simple supports. Show that the minimum diameters to meet the slope constraints at the left and right bearings are, respectively,
dL =
32 Fb(l 2 b2 )
3π El ξ
1/4
dR =
32 Fa (l 2 a 2 )
3π El ξ
1/4
F
a
b
l
Problem 423
y
F
x
424
A shaft is to be designed so that it is supported by roller bearings. The basic geometry is shown
in the gure. The allowable slope at the bearings is 0.001 mm/mm without bearing life penalty.
For a design factor of 1.28, what uniform-diameter shaft will support the 3.5-kN load 100 mm
from the left bearing without penalty? Use E = 207 GPa.
F = 3.5 kN
100
150
Problem 424
Dimensions in millimeters.
d
250
425
Determine the maximum deection of the shaft of Prob. 424.
426
For the shaft shown in the gure, let a1 = 4 in, b1 = 12 in, a2 = 10 in, F1 = 100 lbf, F2 = 300 lbf,
and E = 30 Mpsi. The shaft is to be sized so that the maximum slope at either bearing A or bearing B does not exceed 0.001 rad. Determine a suitable diameter d.
196
192
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
y
F1
a1
A
Problem 426
b1
z
a2
B
b2
F2
x
427
If the diameter of the beam for Prob. 426 is 1.375 in, determine the deection of the beam at
x = 8 in.
428
See Prob. 426 and the accompanying gure. The loads and dimensions are F1 = 3.5 kN,
F2 = 2.7 kN, a1 = 100 mm, b1 = 150 mm, and a2 = 175 mm. Find the uniform shaft diameter
necessary to limit the slope at the bearings to 0.001 rad. Use a design factor of n d = 1.5 and
E = 207 Gpa.
429
Shown in the gure is a uniform-diameter shaft with bearing shoulders at the ends; the shaft is
subjected to a concentrated moment M = 1200 lbf · in. The shaft is of carbon steel and has a = 5
in and l = 9 in. The slope at the ends must be limited to 0.002 rad. Find a suitable diameter d .
a
b
MB
Problem 429
B
l
430
The rectangular member O AB , shown in the gure, is held horizontal by the round hooked bar
AC. The modulus of elasticity of both parts is 10 Mpsi. Use superposition to nd the deection
at B due to a force F = 80 lbf.
1
-in
2
dia.
C
y
Problem 430
12 in
2 in
1
-in
4
thick
F
x
A
B
O
6 in
431
12 in
The gure illustrates a torsion-bar spring O A having a diameter d = 12 mm. The actuating
cantilever A B also has d = 12 mm. Both parts are of carbon steel. Use superposition and nd
the spring rate k corresponding to a force F acting at B.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
197
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
193
y
O
d
x
Problem 431
1.5 m
A
F
d
z
0.1 m
B
432
Consider the simply supported beam with an intermediate load in Appendix A96. Determine
the deection equation if the stiffness of the left and right supports are k1 and k2 , respectively.
433
Consider the simply supported beam with a uniform load in Appendix A97. Determine the
deection equation if the stiffness of the left and right supports are k1 and k2 , respectively.
434
Prove that for a uniform-cross-section beam with simple supports at the ends loaded by a single
concentrated load, the location of the maximum deection will never be outside the range of
0.423l x 0.577l regardless of the location of the load along the beam. The importance of this
is that you can always get a quick estimate of ymax by using x = l /2.
435
Solve Prob. 412 using singularity functions. Use statics to determine the reactions.
436
Solve Prob. 413 using singularity functions. Use statics to determine the reactions.
437
Solve Prob. 414 using singularity functions. Use statics to determine the reactions.
438
Consider the uniformly loaded simply supported beam with an overhang as shown. Use singularity
functions to determine the deection equation of the beam. Use statics to determine the reactions.
w
Problem 438
l
a
439
Solve Prob. 415 using singularity functions. Since the beam is symmetric, only write the equation for half the beam and use the slope at the beam center as a boundary condition. Use statics
to determine the reactions.
440
Solve Prob. 430 using singularity functions. Use statics to determine the reactions.
441
Determine the deection equation for the steel beam shown using singularity functions. Since the
beam is symmetric, write the equation for only half the beam and use the slope at the beam center as a boundary condition. Use statics to determine the reactions.
w = 200 lbf/in
1.5-in diameter
1.5-in diameter
Problem 441
2-in diameter
4 in
12 in
4 in
198
194
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
442
Determine the deection equation for the cantilever beam shown using singularity functions.
Evaluate the deections at B and C and compare your results with Example 411.
y
l /2
Problem 442
A
2 I1
l /2
B
I1
x
C
F
443
Examine the expression for the deection of the cantilever beam, end-loaded, shown in
Appendix Table A91 for some intermediate point, x = a , as
F1 a 2
(a 3l )
6E I
y |x = a =
In Table A92, for a cantilever with intermediate load, the deection at the end is
y |x = l =
F2 a 2
(a 3l )
6E I
These expressions are remarkably similar and become identical when F1 = F2 = 1. In other
words, the deection at x = a (station 1) due to a unit load at x = l (station 2) is the same as the
deection at station 2 due to a unit load at station 1. Prove that this is true generally for an elastic body even when the lines of action of the loads are not parallel. This is known as a special
case of Maxwells reciprocal theorem. (Hint: Consider the potential energy of strain when the
body is loaded by two forces in either order of application.)
444
A steel shaft of uniform 2-in diameter has a bearing span l of 23 in and an overhang of 7 in on
which a coupling is to be mounted. A gear is to be attached 9 in to the right of the left bearing
and will carry a radial load of 400 lbf. We require an estimate of the bending deection at the
coupling. Appendix Table A96 is available, but we cant be sure of how to expand the equation
to predict the deection at the coupling.
(a) Show how Appendix Table A910 and Maxwells theorem (see Prob. 443) can be used to
obtain the needed estimate.
(b) Check your work by nding the slope at the right bearing and extending it to the coupling
location.
445
Use Castiglianos theorem to verify the maximum deection for the uniformly loaded beam of
Appendix Table A97. Neglect shear.
446
Solve Prob. 417 using Castiglianos theorem. Hint: Write the moment equation using a position
variable positive to the left starting at the right end of the beam.
447
Solve Prob. 430 using Castiglianos theorem.
448
Solve Prob. 431 using Castiglianos theorem.
449
Determine the deection at midspan for the beam of Prob. 441 using Castiglianos theorem.
I. Basics
450
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
199
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
195
Using Castiglianos theorem, determine the deection of point B in the direction of the force F
for the bar shown. The solid bar has a uniform diameter, d. Neglect bending shear.
l
O
A
Problem 450
a
B
4
3
F
451
A cable is made using a 16-gauge (0.0625-in) steel wire and three strands of 12-gauge (0.0801-in)
copper wire. Find the stress in each wire if the cable is subjected to a tension of 250 lbf.
452
The gure shows a steel pressure cylinder of diameter 4 in which uses six SAE grade 5 steel bolts
having a grip of 12 in. These bolts have a proof strength (see Chap. 8) of 85 kpsi for this size of
bolt. Suppose the bolts are tightened to 90 percent of this strength in accordance with some
recommendations.
(a) Find the tensile stress in the bolts and the compressive stress in the cylinder walls.
(b) Repeat part (a), but assume now that a uid under a pressure of 600 psi is introduced into the
cylinder.
Six
3
8
-in grade 5 bolts
t=
Problem 452
lc = 11 in
1
4
in
D = 4 in
lb = 12 in
453
A torsion bar of length L consists of a round core of stiffness (G J )c and a shell of stiffness
(G J )s . If a torque T is applied to this composite bar, what percentage of the total torque is carried by the shell?
454
A rectangular aluminum bar 12 mm thick and 50 mm wide is welded to xed supports at the ends,
and the bar supports a load W = 3.5 kN, acting through a pin as shown. Find the reactions at the
supports.
200
196
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
y
B
750 mm
Problem 454
50 mm
W
12 mm thick
A
500 mm
x
O
455
The steel shaft shown in the gure is subjected to a torque of 50 lbf-in applied at point A. Find
the torque reactions at O and B .
y
1
50 lbf-in 1 2 -in dia.
Problem 455
x
O
A
B
4 in
6 in
456
Repeat Prob. 455 with the diameters of section OA being 1.5 in and section AB being 1.75 in.
457
In testing the wear life of gear teeth, the gears are assembled by using a pretorsion. In this way,
a large torque can exist even though the power input to the tester is small. The arrangement shown
in the gure uses this principle. Note the symbol used to indicate the location of the shaft bearings used in the gure. Gears A, B , and C are assembled rst, and then gear C is held xed. Gear
D is assembled and meshed with gear C by twisting it through an angle of 4 to provide the pretorsion. Find the maximum shear stress in each shaft resulting from this preload.
4 ft
C, 6-in dia.
Problem 457
B, 6-in dia.
1 1 -in dia.
4
7
8
2
-in dia.
1
D, 2 1 -in dia.
2
A, 2 1 -in dia.
2
458
3
The gure shows a 8 - by 1 1 -in rectangular steel bar welded to xed supports at each end. The
2
bar is axially loaded by the forces FA = 10 kip and FB = 5 kip acting on pins at A and B .
Assuming that the bar will not buckle laterally, nd the reactions at the xed supports. Use procedure 1 from Sec. 410.
459
For the beam shown, determine the support reactions using superposition and procedure 1 from
Sec. 410.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
201
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
197
y
10 in
20 in
Problem 458
A
1
15 in
B
FA
1 2 in
C
FB
x
O
3
8
in thick
w
Problem 459
B
A
C
a
l
460
Solve Prob. 459 using Castiglianos theorem and procedure 1 from Sec. 410.
461
The steel beam A BC D shown is simply supported at A and supported at B and D by steel cables,
each having an effective diameter of 12 mm. The second area moment of the beam is I =
8(105 ) mm4 . A force of 20 kN is applied at point C. Using procedure 2 of Sec. 410 determine
the stresses in the cables and the deections of B, C, and D. For steel, let E = 209 GPa.
E
F
1m
A
Problem 461
B
C
D
20 kN
500 mm
462
500 mm
500 mm
The steel beam A BC D shown is supported at C as shown and supported at B and D by steel bolts
5
each having a diameter of 16 in. The lengths of B E and D F are 2 and 2.5 in, respectively. The
beam has a second area moment of 0.050 in4 . Prior to loading, the nuts are just in contact with
the horizontal beam. A force of 500 lbf is then applied at point A. Using procedure 2 of Sec. 410,
determine the stresses in the bolts and the deections of points A, B, and D. For steel, let
E = 30 Mpsi.
E
500 lbf
A
B
D
C
Problem 462
F
3 in
463
3 in
3 in
The horizontal deection of the right end of the curved bar of Fig. 412 is given by Eq. (435)
for R / h > 10. For the same conditions, determine the vertical deection.
202
198
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
464
A cast-iron piston ring has a mean diameter of 81 mm, a radial height h = 6 mm, and a thickness
b = 4 mm. The ring is assembled using an expansion tool that separates the split ends a distance
δ by applying a force F as shown. Use Castiglianos theorem and determine the deection δ as a
function of F . Use E = 131 GPa and assume Eq. (428) applies.
h = 6 mm
F
Problem 464
+
F
465
For the wire form shown use Castiglianos method to determine the vertical deection of point A.
Consider bending only and assume Eq. (428) applies for the curved part.
C
Problem 465
P
R
A
B
l
466
For the wire form shown determine the vertical deections of points A and B. Consider bending
only and assume Eq. (428) applies.
A
C
R
P
Problem 466
B
467
For the wire form shown, determine the deection of point A in the y direction. Assume
R / h > 10 and consider the effects of bending and torsion only. The wire is steel with E =
200 GPa, ν = 0.29, and has a diameter of 5 mm. Before application of the 200-N force the wire
form is in the x z plane where the radius R is 100 mm.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
203
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
199
y
x
Problem 467
R
z
90°
A
200 N
468
For the wire form shown, determine (a) the reactions at points A and B, (b) how the bending
moment varies along the wire, and (c) the deection of the load F. Assume that the entire energy
is described by Eq. (428).
F
Problem 468
R
A
469
B
For the curved beam shown, F = 30 kN. The material is steel with E = 207 GPa and G =
79 GPa. Determine the relative deection of the applied forces.
80
10
50
F
F
A
A
20
40
Problem 469
10
Section AA
100
(All dimensions in millimeters.)
470
Solve Prob. 463 using Eq. (432).
471
A thin ring is loaded by two equal and opposite forces F in part a of the gure. A free-body diagram of one quadrant is shown in part b. This is a statically indeterminate problem, because the
moment M A cannot be found by statics. We wish to nd the maximum bending moment in the ring
due to the forces F. Assume that the radius of the ring is large so that Eq. (428) can be used.
204
200
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
Mechanical Engineering Design
y
y
F
B
B
ds
d
R
Problem 471
A
C
A
x
O
O
x
MA
F
2
D
F
(b )
(a)
472
Find the increase in the diameter of the ring of Prob. 471 due to the forces F and along the y axis.
473
A round tubular column has outside and inside diameters of D and d, respectively, and a diametral ratio of K = d / D . Show that buckling will occur when the outside diameter is
64 Pcr l 2
3 C E (1 K 4 )
π
D=
474
For the conditions of Prob. 473, show that buckling according to the parabolic formula will
occur when the outside diameter is
D=2
475
1/4
Sy l 2
Pcr
+2
2)
π Sy (1 K
π C E (1 + K 2 )
1/2
Link 2, shown in the gure, is 1 in wide, has 1 -in-diameter bearings at the ends, and is cut from
2
low-carbon steel bar stock having a minimum yield strength of 24 kpsi. The end-condition constants are C = 1 and C = 1.2 for buckling in and out of the plane of the drawing, respectively.
(a) Using a design factor n d = 5, nd a suitable thickness for the link.
(b) Are the bearing stresses at O and B of any signicance?
y
1
Problem 475
x
2
O
A
3
3
180 lbf
1 4 ft
B
3 ft
476
C
1
2 2 ft
Link 3, shown schematically in the gure, acts as a brace to support the 1.2-kN load. For buckling in the plane of the gure, the link may be regarded as pinned at both ends. For out-of-plane
buckling, the ends are xed. Select a suitable material and a method of manufacture, such as forging, casting, stamping, or machining, for casual applications of the brace in oil-eld machinery.
Specify the dimensions of the cross section as well as the ends so as to obtain a strong, safe, wellmade, and economical brace.
I. Basics
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
205
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
201
y
B
F = 1.2 kN
3
Problem 476
0.9 m
2
O
60°
1
477
A
x
The hydraulic cylinder shown in the gure has a 3-in bore and is to operate at a pressure of 800 psi.
With the clevis mount shown, the piston rod should be sized as a column with both ends rounded for
any plane of buckling. The rod is to be made of forged AISI 1030 steel without further heat treatment.
d
Problem 477
3 in
(a) Use a design factor n d = 3 and select a preferred size for the rod diameter if the column
length is 60 in.
(b) Repeat part (a) but for a column length of 18 in.
(c) What factor of safety actually results for each of the cases above?
478
The gure shows a schematic drawing of a vehicular jack that is to be designed to support a
maximum mass of 400 kg based on the use of a design factor n d = 2.50. The opposite-handed
threads on the two ends of the screw are cut to allow the link angle θ to vary from 15 to 70 . The
links are to be machined from AISI 1020 hot-rolled steel bars with a minimum yield strength of
380 MPa. Each of the four links is to consist of two bars, one on each side of the central bearings. The bars are to be 300 mm long and have a bar width of 25 mm. The pinned ends are to be
designed to secure an end-condition constant of at least C = 1.4 for out-of-plane buckling. Find
a suitable preferred thickness and the resulting factor of safety for this thickness.
W
l
Problem 478
w
206
202
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
I. Basics
4. Deflection and Stiffness
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
479
If drawn, a gure for this problem would resemble that for Prob. 452. A strut that is a standard
3
hollow right circular cylinder has an outside diameter of 4 in and a wall thickness of 8 in and is
compressed between two circular end plates held by four bolts equally spaced on a bolt circle of
5.68-in diameter. All four bolts are hand-tightened, and then bolt A is tightened to a tension
of 2000 lbf and bolt C, diagonally opposite, is tightened to a tension of 10 000 lbf. The strut axis
of symmetry is coincident with the center of the bolt circles. Find the maximum compressive
load, the eccentricity of loading, and the largest compressive stress in the strut.
480
Design link C D of the hand-operated toggle press shown in the gure. Specify the cross-section
dimensions, the bearing size and rod-end dimensions, the material, and the method of processing.
F
A
B
L
l
Problem 480
C
L = 12 in, l = 4 in, θmin = 0°.
l
D
481
Find expressions for the maximum values of the spring force and deection y of the impact system shown in the gure. Can you think of a realistic application for this model?
W
y
k
Problem 481
h
482
As shown in the gure, the weight W1 strikes W2 from a height h. Find the maximum values of the
spring force and the deection of W2 . Name an actual system for which this model might be used.
h
W1
W2
Problem 482
y
k
I. Basics
483
© The McGrawHill
Companies, 2008
4. Deflection and Stiffness
207
Deection and Stiffness
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
203
Part a of the gure shows a weight W mounted between two springs. If the free end of spring k1
is suddenly displaced through the distance x = a , as shown in part b, what would be the maximum displacement y of the weight?
x
y
k1
k2
W
Problem 483
a
t
x
(a )
( b)
208
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
PART
II. Failure Prevention
Introduction
2
Failure Prevention
© The McGrawHill
Companies, 2008
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
5
209
Failures Resulting
from Static Loading
Chapter Outline
51
Static Strength
52
Stress Concentration
53
Failure Theories
54
Maximum-Shear-Stress Theory for Ductile Materials
55
Distortion-Energy Theory for Ductile Materials
56
Coulomb-Mohr Theory for Ductile Materials
57
Failure of Ductile Materials Summary
58
Maximum-Normal-Stress Theory for Brittle Materials
59
Modications of the Mohr Theory for Brittle Materials
208
209
211
510
Failure of Brittle Materials Summary
511
Selection of Failure Criteria
512
Introduction to Fracture Mechanics
513
Stochastic Analysis
514
Important Design Equations
211
213
219
222
226
227
229
230
231
240
246
205
210
206
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
In Chap. 1 we learned that strength is a property or characteristic of a mechanical
element. This property results from the material identity, the treatment and processing
incidental to creating its geometry, and the loading, and it is at the controlling or critical
location.
In addition to considering the strength of a single part, we must be cognizant
that the strengths of the mass-produced parts will all be somewhat different from the
others in the collection or ensemble because of variations in dimensions, machining,
forming, and composition. Descriptors of strength are necessarily statistical in
nature, involving parameters such as mean, standard deviations, and distributional
identification.
A static load is a stationary force or couple applied to a member. To be stationary,
the force or couple must be unchanging in magnitude, point or points of application,
and direction. A static load can produce axial tension or compression, a shear load, a
bending load, a torsional load, or any combination of these. To be considered static, the
load cannot change in any manner.
In this chapter we consider the relations between strength and static loading in order
to make the decisions concerning material and its treatment, fabrication, and geometry
for satisfying the requirements of functionality, safety, reliability, competitiveness,
usability, manufacturability, and marketability. How far we go down this list is related
to the scope of the examples.
Failure is the rst word in the chapter title. Failure can mean a part has separated into two or more pieces; has become permanently distorted, thus ruining its
geometry; has had its reliability downgraded; or has had its function compromised,
whatever the reason. A designer speaking of failure can mean any or all of these possibilities. In this chapter our attention is focused on the predictability of permanent
distortion or separation. In strength-sensitive situations the designer must separate
mean stress and mean strength at the critical location sufciently to accomplish his
or her purposes.
Figures 51 to 55 are photographs of several failed parts. The photographs exemplify the need of the designer to be well-versed in failure prevention. Toward this end
we shall consider one-, two-, and three-dimensional stress states, with and without
stress concentrations, for both ductile and brittle materials.
Figure 51
(a) Failure of a truck drive-shaft
spline due to corrosion
fatigue. Note that it was
necessary to use clear tape
to hold the pieces in place.
(b) Direct end view of failure.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Failures Resulting from Static Loading
Figure 52
Impact failure of a lawnmower blade driver hub. The
blade impacted a surveying
pipe marker.
Figure 53
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.
Figure 54
Chain test xture that failed in one cycle. To alleviate complaints of excessive wear, the manufacturer decided to
case-harden the material. (a) Two halves showing fracture; this is an excellent example of brittle fracture initiated
by stress concentration. (b) Enlarged view of one portion to show cracks induced by stress concentration at the
support-pin holes.
211
207
212
208
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 55
Valve-spring failure caused by
spring surge in an oversped
engine. The fractures exhibit
the classic 45 shear failure.
51
Static Strength
Ideally, in designing any machine element, the engineer should have available the results
of a great many strength tests of the particular material chosen. These tests should be
made on specimens having the same heat treatment, surface nish, and size as the element the engineer proposes to design; and the tests should be made under exactly the
same loading conditions as the part will experience in service. This means that if the part
is to experience a bending load, it should be tested with a bending load. If it is to be
subjected to combined bending and torsion, it should be tested under combined bending
and torsion. If it is made of heat-treated AISI 1040 steel drawn at 500 C with a ground
nish, the specimens tested should be of the same material prepared in the same manner.
Such tests will provide very useful and precise information. Whenever such data are
available for design purposes, the engineer can be assured of doing the best possible job
of engineering.
The cost of gathering such extensive data prior to design is justied if failure of the
part may endanger human life or if the part is manufactured in sufciently large quantities. Refrigerators and other appliances, for example, have very good reliabilities
because the parts are made in such large quantities that they can be thoroughly tested
in advance of manufacture. The cost of making these tests is very low when it is divided by the total number of parts manufactured.
You can now appreciate the following four design categories:
1
2
3
Failure of the part would endanger human life, or the part is made in extremely
large quantities; consequently, an elaborate testing program is justied during
design.
The part is made in large enough quantities that a moderate series of tests is
feasible.
The part is made in such small quantities that testing is not justied at all; or the
design must be completed so rapidly that there is not enough time for testing.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Failures Resulting from Static Loading
213
209
The part has already been designed, manufactured, and tested and found to be
unsatisfactory. Analysis is required to understand why the part is unsatisfactory
and what to do to improve it.
4
More often than not it is necessary to design using only published values of yield
strength, ultimate strength, percentage reduction in area, and percentage elongation,
such as those listed in Appendix A. How can one use such meager data to design against
both static and dynamic loads, two- and three-dimensional stress states, high and low
temperatures, and very large and very small parts? These and similar questions will be
addressed in this chapter and those to follow, but think how much better it would be to
have data available that duplicate the actual design situation.
52
Stress Concentration
Stress concentration (see Sec. 313) is a highly localized effect. In some instances it
may be due to a surface scratch. If the material is ductile and the load static, the design
load may cause yielding in the critical location in the notch. This yielding can involve
strain strengthening of the material and an increase in yield strength at the small critical notch location. Since the loads are static and the material is ductile, that part can
carry the loads satisfactorily with no general yielding. In these cases the designer sets
the geometric (theoretical) stress concentration factor K t to unity.
The rationale can be expressed as follows. The worst-case scenario is that of an
idealized nonstrain-strengthening material shown in Fig. 56. The stress-strain curve
rises linearly to the yield strength Sy , then proceeds at constant stress, which is equal to
Sy . Consider a lleted rectangular bar as depicted in Fig. A155, where the crosssection area of the small shank is 1 in2. If the material is ductile, with a yield point of
40 kpsi, and the theoretical stress-concentration factor (SCF) K t is 2,
A load of 20 kip induces a tensile stress of 20 kpsi in the shank as depicted at point A
in Fig. 56. At the critical location in the llet the stress is 40 kpsi, and the SCF is
K = σmax /σnom = 40/20 = 2.
Figure 56
An idealized stress-strain
curve. The dashed line depicts
a strain-strengthening material.
50
Tensile stress , kpsi
C
Sy
D
E
B
A
0
Tensile strain,
214
210
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
A load of 30 kip induces a tensile stress of 30 kpsi in the shank at point B. The llet
stress is still 40 kpsi (point D), and the SCF K = σmax /σnom = Sy /σ = 40/30 = 1.33.
At a load of 40 kip the induced tensile stress (point C) is 40 kpsi in the shank.
At the critical location in the llet, the stress (at point E) is 40 kpsi. The SCF
K = σmax /σnom = Sy /σ = 40/40 = 1.
For materials that strain-strengthen, the critical location in the notch has a higher Sy .
The shank area is at a stress level a little below 40 kpsi, is carrying load, and is very
near its failure-by-general-yielding condition. This is the reason designers do not apply
K t in static loading of a ductile material loaded elastically, instead setting K t = 1.
When using this rule for ductile materials with static loads, be careful to assure
yourself that the material is not susceptible to brittle fracture (see Sec. 512) in the
environment of use. The usual denition of geometric (theoretical) stress-concentration
factor for normal stress K t and shear stress K ts is
σmax = K t σnom
(a)
τmax = K ts τnom
(b)
Since your attention is on the stress-concentration factor, and the denition of σnom or
τnom is given in the graph caption or from a computer program, be sure the value of
nominal stress is appropriate for the section carrying the load.
Brittle materials do not exhibit a plastic range. A brittle material feels the stress
concentration factor K t or K ts , which is applied by using Eq. (a) or (b).
An exception to this rule is a brittle material that inherently contains microdiscontinuity stress concentration, worse than the macrodiscontinuity that the designer has in
mind. Sand molding introduces sand particles, air, and water vapor bubbles. The grain
structure of cast iron contains graphite akes (with little strength), which are literally
cracks introduced during the solidication process. When a tensile test on a cast iron is
performed, the strength reported in the literature includes this stress concentration. In
such cases K t or K ts need not be applied.
An important source of stress-concentration factors is R. E. Peterson, who compiled them from his own work and that of others.1 Peterson developed the style of presentation in which the stress-concentration factor K t is multiplied by the nominal stress
σnom to estimate the magnitude of the largest stress in the locality. His approximations
were based on photoelastic studies of two-dimensional strips (Hartman and Levan,
1951; Wilson and White, 1973), with some limited data from three-dimensional
photoelastic tests of Hartman and Levan. A contoured graph was included in the presentation of each case. Filleted shafts in tension were based on two-dimensional strips.
Table A15 provides many charts for the theoretical stress-concentration factors for
several fundamental load conditions and geometry. Additional charts are also available
from Peterson.2
Finite element analysis (FEA) can also be applied to obtain stress-concentration
factors. Improvements on K t and K ts for lleted shafts were reported by Tipton, Sorem,
and Rolovic.3
1
R. E. Peterson, Design Factors for Stress Concentration, Machine Design, vol. 23, no. 2, February 1951;
no. 3, March 1951; no. 5, May 1951; no. 6, June 1951; no. 7, July 1951.
2
Walter D. Pilkey, Petersons Stress Concentration Factors, 2nd ed, John Wiley & Sons, New York, 1997.
3
S. M. Tipton, J. R. Sorem Jr., and R. D. Rolovic, Updated Stress-Concentration Factors for Filleted Shafts in
Bending and Tension, Trans. ASME, Journal of Mechanical Design, vol. 118, September 1996, pp. 321327.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
53
215
211
Failure Theories
Section 51 illustrated some ways that loss of function is manifested. Events such as
distortion, permanent set, cracking, and rupturing are among the ways that a machine
element fails. Testing machines appeared in the 1700s, and specimens were pulled, bent,
and twisted in simple loading processes.
If the failure mechanism is simple, then simple tests can give clues. Just what is
simple? The tension test is uniaxial (thats simple) and elongations are largest in the axial
direction, so strains can be measured and stresses inferred up to failure. Just what is
important: a critical stress, a critical strain, a critical energy? In the next several sections,
we shall show failure theories that have helped answer some of these questions.
Unfortunately, there is no universal theory of failure for the general case of material properties and stress state. Instead, over the years several hypotheses have been
formulated and tested, leading to todays accepted practices. Being accepted, we will
characterize these practices as theories as most designers do.
Structural metal behavior is typically classied as being ductile or brittle, although
under special situations, a material normally considered ductile can fail in a brittle
manner (see Sec. 512). Ductile materials are normally classied such that ε f 0.05
and have an identiable yield strength that is often the same in compression as in tension ( Syt = Syc = Sy ). Brittle materials, ε f < 0.05, do not exhibit an identiable yield
strength, and are typically classied by ultimate tensile and compressive strengths, Sut
and Suc , respectively (where Suc is given as a positive quantity). The generally accepted
theories are:
Ductile materials (yield criteria)
Maximum shear stress (MSS), Sec. 54
Distortion energy (DE), Sec. 55
Ductile Coulomb-Mohr (DCM), Sec. 56
Brittle materials (fracture criteria)
Maximum normal stress (MNS), Sec. 58
Brittle Coulomb-Mohr (BCM), Sec. 59
Modied Mohr (MM), Sec. 59
It would be inviting if we had one universally accepted theory for each material
type, but for one reason or another, they are all used. Later, we will provide rationales
for selecting a particular theory. First, we will describe the bases of these theories and
apply them to some examples.
54
Maximum-Shear-Stress Theory
for Ductile Materials
The maximum-shear-stress theory predicts that yielding begins whenever the maximum
shear stress in any element equals or exceeds the maximum shear stress in a tensiontest specimen of the same material when that specimen begins to yield. The MSS theory
is also referred to as the Tresca or Guest theory.
Many theories are postulated on the basis of the consequences seen from tensile
tests. As a strip of a ductile material is subjected to tension, slip lines (called Lüder
lines) form at approximately 45° with the axis of the strip. These slip lines are the
216
212
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
beginning of yield, and when loaded to fracture, fracture lines are also seen at angles
approximately 45° with the axis of tension. Since the shear stress is maximum at 45°
from the axis of tension, it makes sense to think that this is the mechanism of failure. It
will be shown in the next section, that there is a little more going on than this. However,
it turns out the MSS theory is an acceptable but conservative predictor of failure; and
since engineers are conservative by nature, it is quite often used.
Recall that for simple tensile stress, σ = P / A , and the maximum shear stress
occurs on a surface 45° from the tensile surface with a magnitude of τmax = σ/2. So the
maximum shear stress at yield is τmax = Sy /2. For a general state of stress, three principal stresses can be determined and ordered such that σ1 σ2 σ3 . The maximum
shear stress is then τmax = (σ1 σ3 )/2 (see Fig. 312). Thus, for a general state of
stress, the maximum-shear-stress theory predicts yielding when
τmax =
Sy
σ1 σ3
2
2
or
σ1 σ3 Sy
(51)
Note that this implies that the yield strength in shear is given by
(52)
Ssy = 0.5 Sy
which, as we will see later is about 15 percent low (conservative).
For design purposes, Eq. (51) can be modied to incorporate a factor of safety, n.
Thus,
τmax =
Sy
2n
or
σ1 σ3 =
Sy
n
(53)
Plane stress problems are very common where one of the principal stresses is zero,
and the other two, σ A and σ B , are determined from Eq. (313). Assuming that σ A σ B ,
there are three cases to consider in using Eq. (51) for plane stress:
Case 1: σ A σ B 0. For this case, σ1 = σ A and σ3 = 0. Equation (51)
reduces to a yield condition of
σ A Sy
(54)
Case 2: σ A 0 σ B . Here, σ1 = σ A and σ3 = σ B , and Eq. (51) becomes
σ A σ B Sy
(55)
Case 3: 0 σ A σ B . For this case, σ1 = 0 and σ3 = σ B , and Eq. (51) gives
σ B Sy
(56)
Equations (54) to (56) are represented in Fig. 57 by the three lines indicated in the
σ A , σ B plane. The remaining unmarked lines are cases for σ B σ A , which are not normally used. Equations (54) to (56) can also be converted to design equations by substituting equality for the equal to or greater sign and dividing Sy by n.
Note that the rst part of Eq. (5-3), τmax = Sy /2n , is sufcient for design purposes
provided the designer is careful in determining τmax . For plane stress, Eq. (314) does
not always predict τmax . However, consider the special case when one normal stress is
zero in the plane, say σx and τx y have values and σ y = 0. It can be easily shown that this
is a Case 2 problem, and the shear stress determined by Eq. (314) is τmax . Shaft design
problems typically fall into this category where a normal stress exists from bending
and/or axial loading, and a shear stress arises from torsion.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 57
217
213
B
Sy
The maximum-shear-stress
(MSS) theory for plane stress,
where σ A and σ B are the two
nonzero principal stresses.
Case 1
Sy
A
Sy
Case 2
Sy
Case 3
55
Distortion-Energy Theory for Ductile Materials
The distortion-energy theory predicts that yielding occurs when the distortion strain
energy per unit volume reaches or exceeds the distortion strain energy per unit volume
for yield in simple tension or compression of the same material.
The distortion-energy (DE) theory originated from the observation that ductile
materials stressed hydrostatically exhibited yield strengths greatly in excess of the values given by the simple tension test. Therefore it was postulated that yielding was not
a simple tensile or compressive phenomenon at all, but, rather, that it was related somehow to the angular distortion of the stressed element. To develop the theory, note, in Fig.
58a, the unit volume subjected to any three-dimensional stress state designated by the
stresses σ1 , σ2 , and σ3 . The stress state shown in Fig. 58b is one of hydrostatic tension
due to the stresses σav acting in each of the same principal directions as in Fig. 58a.
The formula for σav is simply
σ1 + σ2 + σ3
3
σav =
(a)
Thus the element in Fig. 58b undergoes pure volume change, that is, no angular distortion. If we regard σav as a component of σ1 , σ2 , and σ3 , then this component can be
2
2
av
=
+
1
3
1
>
2
(a) Triaxial stresses
>
av
1
av
av
3
av
3
(b) Hydrostatic component
(c) Distortional component
Figure 58
(a) Element with triaxial stresses; this element undergoes both volume
change and angular distortion. (b) Element under hydrostatic tension
undergoes only volume change. (c) Element has angular distortion
without volume change.
av
218
214
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
subtracted from them, resulting in the stress state shown in Fig. 58c. This element is
subjected to pure angular distortion, that is, no volume change.
The strain energy per unit volume for simple tension is u = 1 ǫσ . For the element
2
of Fig. 58a the strain energy per unit volume is u = 1 [ǫ1 σ1 + ǫ2 σ2 + ǫ3 σ3 ].
2
Substituting Eq. (319) for the principal strains gives
u=
1
2
2
σ 2 + σ2 + σ3 2ν(σ1 σ2 + σ2 σ3 + σ3 σ1 )
2E 1
(b)
The strain energy for producing only volume change u v can be obtained by substituting σav for σ1 , σ2 , and σ3 in Eq. (b). The result is
uv =
2
3σav
(1 2ν)
2E
(c)
If we now substitute the square of Eq. (a) in Eq. (c) and simplify the expression, we get
uv =
1 2ν 2
2
2
σ1 + σ2 + σ3 + 2σ1 σ2 + 2σ2 σ3 + 2σ3 σ1
6E
(57)
Then the distortion energy is obtained by subtracting Eq. (57) from Eq. (b). This
gives
ud = u uv =
1+ν
3E
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
2
(58)
Note that the distortion energy is zero if σ1 = σ2 = σ3 .
For the simple tensile test, at yield, σ1 = Sy and σ2 = σ3 = 0, and from Eq. (58)
the distortion energy is
ud =
1+ν 2
S
3E y
(59)
So for the general state of stress given by Eq. (58), yield is predicted if Eq. (58)
equals or exceeds Eq. (59). This gives
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
2
1/2
Sy
(510)
If we had a simple case of tension σ , then yield would occur when σ Sy . Thus, the
left of Eq. (510) can be thought of as a single, equivalent, or effective stress for the
entire general state of stress given by σ1 , σ2 , and σ3 . This effective stress is usually
called the von Mises stress, σ , named after Dr. R. von Mises, who contributed to the
theory. Thus Eq. (510), for yield, can be written as
σ Sy
(511)
where the von Mises stress is
σ =
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
2
1/2
(512)
For plane stress, let σ A and σ B be the two nonzero principal stresses. Then from
Eq. (512), we get
2
2
σ = σ A σ AσB + σB
1/2
(513)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
219
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 59
215
B
The distortion-energy (DE)
theory for plane stress states.
This is a plot of points
obtained from Eq. (513)
with σ = Sy .
Sy
Sy
Sy
A
Pure shear load line (
Sy
A
B
)
DE
MSS
Equation (513) is a rotated ellipse in the σ A , σ B plane, as shown in Fig. 59 with
σ = Sy . The dotted lines in the gure represent the MSS theory, which can be seen to
be more restrictive, hence, more conservative.4
Using xyz components of three-dimensional stress, the von Mises stress can be
written as
1
2
2
2
σ = (σx σ y )2 + (σ y σz )2 + (σz σx )2 + 6 τx y + τ yz + τzx
2
1/2
(514)
and for plane stress,
2
2
2
σ = σx σx σ y + σ y + 3τx y
1/2
(515)
The distortion-energy theory is also called:
The von Mises or von MisesHencky theory
The shear-energy theory
The octahedral-shear-stress theory
Understanding octahedral shear stress will shed some light on why the MSS is conservative. Consider an isolated element in which the normal stresses on each surface are
equal to the hydrostatic stress σav . There are eight surfaces symmetric to the principal
directions that contain this stress. This forms an octahedron as shown in Fig. 510. The
shear stresses on these surfaces are equal and are called the octahedral shear stresses
(Fig. 510 has only one of the octahedral surfaces labeled). Through coordinate transformations the octahedral shear stress is given by5
τoct =
1
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
3
1/2
(516)
4
The three-dimensional equations for DE and MSS can be plotted relative to three-dimensional σ1 , σ2 , σ3 ,
coordinate axes. The failure surface for DE is a circular cylinder with an axis inclined at 45° from each
principal stress axis, whereas the surface for MSS is a hexagon inscribed within the cylinder. See Arthur P.
Boresi and Richard J. Schmidt, Advanced Mechanics of Materials, 6th ed., John Wiley & Sons, New York,
2003, Sec. 4.4.
5
For a derivation, see Arthur P. Boresi, op. cit., pp. 3637.
220
216
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
Figure 510
2
Octahedral surfaces.
av
oct
1
3
Under the name of the octahedral-shear-stress theory, failure is assumed to occur whenever the octahedral shear stress for any stress state equals or exceeds the octahedral
shear stress for the simple tension-test specimen at failure.
As before, on the basis of the tensile test results, yield occurs when σ1 = Sy and
σ2 = σ3 = 0. From Eq. (516) the octahedral shear stress under this condition is
2
Sy
τoct =
(517)
3
When, for the general stress case, Eq. (516) is equal or greater than Eq. (517), yield
is predicted. This reduces to
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
2
1/2
Sy
(518)
which is identical to Eq. (510), verifying that the maximum-octahedral-shear-stress
theory is equivalent to the distortion-energy theory.
The model for the MSS theory ignores the contribution of the normal stresses on
the 45° surfaces of the tensile specimen. However, these stresses are P /2 A , and not the
hydrostatic stresses which are P /3 A . Herein lies the difference between the MSS and
DE theories.
The mathematical manipulation involved in describing the DE theory might tend
to obscure the real value and usefulness of the result. The equations given allow the
most complicated stress situation to be represented by a single quantity, the von Mises
stress, which then can be compared against the yield strength of the material through
Eq. (511). This equation can be expressed as a design equation by
σ =
Sy
n
(519)
The distortion-energy theory predicts no failure under hydrostatic stress and
agrees well with all data for ductile behavior. Hence, it is the most widely used theory for ductile materials and is recommended for design problems unless otherwise
specified.
One nal note concerns the shear yield strength. Consider a case of pure shear τx y ,
where for plane stress σx = σ y = 0. For yield, Eq. (511) with Eq. (515) gives
2
3τx y
1/2
= Sy
or
Sy
τx y = = 0.577 Sy
3
(520)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
221
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
217
Failures Resulting from Static Loading
Thus, the shear yield strength predicted by the distortion-energy theory is
Ssy = 0.577 Sy
(521)
which as stated earlier, is about 15 percent greater than the 0.5 Sy predicted by the MSS
theory. For pure shear, τx y the principal stresses from Eq. (313) are σ A = σ B = τx y .
The load line for this case is in the third quadrant at an angle of 45o from the σ A , σ B
axes shown in Fig. 59.
EXAMPLE 51
A hot-rolled steel has a yield strength of Syt = Syc = 100 kpsi and a true strain at
fracture of ε f = 0.55. Estimate the factor of safety for the following principal stress
states:
(a) 70, 70, 0 kpsi.
(b) 30, 70, 0 kpsi.
(c) 0, 70, 30 kpsi.
(d ) 0, 30, 70 kpsi.
(e) 30, 30, 30 kpsi.
Solution
Since ε f > 0.05 and Syc and Syt are equal, the material is ductile and the distortionenergy (DE) theory applies. The maximum-shear-stress (MSS) theory will also be
applied and compared to the DE results. Note that cases a to d are plane stress
states.
(a) The ordered principal stresses are σ A = σ1 = 70, σ B = σ2 = 70, σ3 = 0 kpsi.
DE From Eq. (513),
σ = [702 70(70) + 702 ]1/2 = 70 kpsi
Answer
n=
Sy
100
=
= 1.43
σ
70
MSS Case 1, using Eq. (54) with a factor of safety,
Answer
n=
Sy
100
= 1.43
=
σA
70
(b) The ordered principal stresses are σ A = σ1 = 70, σ B = σ2 = 30, σ3 = 0 kpsi.
DE
Answer
σ = [702 70(30) + 302 ]1/2 = 60.8 kpsi
n=
Sy
100
=
= 1.64
σ
60.8
MSS Case 1, using Eq. (54),
Answer
n=
Sy
100
=
= 1.43
σA
70
222
218
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
(c) The ordered principal stresses are σ A = σ1 = 70, σ2 = 0, σ B = σ3 = 30 kpsi.
σ = [702 70(30) + (30)2 ]1/2 = 88.9 kpsi
DE
Answer
Sy
100
= 1.13
=
σ
88.9
n=
MSS Case 2, using Eq. (55),
Answer
n=
Sy
100
= 1.00
=
σ A σB
70 (30)
(d ) The ordered principal stresses are σ1 = 0, σ A = σ2 = 30, σ B = σ3 = 70 kpsi.
DE
σ = [(70)2 (70)(30) + (30)2 ]1/2 = 60.8 kpsi
Answer
Sy
100
=
= 1.64
σ
60.8
n=
MSS Case 3, using Eq. (56),
Answer
n=
Sy
100
=
= 1.43
σB
70
(e) The ordered principal stresses are σ1 = 30, σ2 = 30, σ3 = 30 kpsi
DE From Eq. (512),
σ =
(30 30)2 + (30 30)2 + (30 30)2
2
Answer
n=
1/2
= 0 kpsi
Sy
100
=
σ
0
MSS From Eq. (53),
Answer
n=
Sy
100
=
σ1 σ3
30 30
A tabular summary of the factors of safety is included for comparisons.
(a)
(b)
(c)
(d )
(e)
DE
1.43
1.64
1.13
1.64
MSS
1.43
1.43
1.00
1.43
Since the MSS theory is on or within the boundary of the DE theory, it will always predict a factor of safety equal to or less than the DE theory, as can be seen in the table.
For each case, except case (e), the coordinates and load lines in the σ A , σ B plane are
shown in Fig. 511. Case (e) is not plane stress. Note that the load line for case (a) is
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 511
223
219
B
(a)
Load lines for Example 51.
Sy
(b)
B
A
Sy
Sy
A
(c)
Sy
DE
MSS
Load lines
(d )
the only plane stress case given in which the two theories agree, thus giving the same
factor of safety.
56
Coulomb-Mohr Theory for Ductile Materials
Not all materials have compressive strengths equal to their corresponding tensile
values. For example, the yield strength of magnesium alloys in compression may be as
little as 50 percent of their yield strength in tension. The ultimate strength of gray cast
irons in compression varies from 3 to 4 times greater than the ultimate tensile strength.
So, in this section, we are primarily interested in those theories that can be used to predict failure for materials whose strengths in tension and compression are not equal.
Historically, the Mohr theory of failure dates to 1900, a date that is relevant to its
presentation. There were no computers, just slide rules, compasses, and French curves.
Graphical procedures, common then, are still useful today for visualization. The idea of Mohr
is based on three simple tests: tension, compression, and shear, to yielding if the material
can yield, or to rupture. It is easier to dene shear yield strength as Ssy than it is to test for it.
The practical difculties aside, Mohrs hypothesis was to use the results of tensile,
compressive, and torsional shear tests to construct the three circles of Fig. 512 dening
a failure envelope, depicted as line ABCDE in the gure, above the σ axis. The failure
envelope need not be straight. The argument amounted to the three Mohr circles
describing the stress state in a body (see Fig. 312) growing during loading until one of
them became tangent to the failure envelope, thereby dening failure. Was the form of
the failure envelope straight, circular, or quadratic? A compass or a French curve
dened the failure envelope.
A variation of Mohrs theory, called the Coulomb-Mohr theory or the internal-friction
theory, assumes that the boundary BCD in Fig. 512 is straight. With this assumption only
the tensile and compressive strengths are necessary. Consider the conventional ordering of
the principal stresses such that σ1 σ2 σ3 . The largest circle connects σ1 and σ3 , as
shown in Fig. 513. The centers of the circles in Fig. 513 are C1, C2, and C3. Triangles
OBiCi are similar, therefore
B3 C3 B1 C1
B2 C2 B1 C1
=
OC2 OC1
OC3 OC1
224
220
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 512
Three Mohr circles, one for the
uniaxial compression test, one
for the test in pure shear, and
one for the uniaxial tension test,
are used to dene failure by the
Mohr hypothesis. The strengths
Sc and S t are the compressive
and tensile strengths,
respectively; they can be used
for yield or ultimate strength.
A
B
C
D
E
Sc
Figure 513
St
Coulomb-Mohr
failure line
Mohrs largest circle for a
general state of stress.
B3
B2
B1
O
Sc
3
C3 C2
1 C1
St
or
Sc
σ1 σ3
St
St
2
2=2
2
σ1 + σ3
St
St
Sc
+
2
2
2
2
Cross-multiplying and simplifying reduces this equation to
σ3
σ1
=1
St
Sc
(522)
where either yield strength or ultimate strength can be used.
For plane stress, when the two nonzero principal stresses are σ A σ B , we have
a situation similar to the three cases given for the MSS theory, Eqs. (54) to (56).
That is,
Case 1: σ A σ B 0. For this case, σ1 = σ A and σ3 = 0. Equation (522)
reduces to a failure condition of
σ A St
(523)
Case 2: σ A 0 σ B . Here, σ1 = σ A and σ3 = σ B , and Eq. (522) becomes
σB
σA
1
St
Sc
(524)
Case 3: 0 σ A σ B . For this case, σ1 = 0 and σ3 = σ B , and Eq. (522) gives
σ B Sc
(525)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 514
225
221
B
Plot of the Coulomb-Mohr
theory of failure for plane
stress states.
St
Sc
St
A
Sc
A plot of these cases, together with the normally unused cases corresponding to
σ B σ A , is shown in Fig. 514.
For design equations, incorporating the factor of safety n, divide all strengths by n.
For example, Eq. (522) as a design equation can be written as
σ3
1
σ1
=
St
Sc
n
(526)
Since for the Coulomb-Mohr theory we do not need the torsional shear strength
circle we can deduce it from Eq. (522). For pure shear τ, σ1 = σ3 = τ . The torsional
yield strength occurs when τmax = Ssy . Substituting σ1 = σ3 = Ssy into Eq. (522)
and simplifying gives
Ssy =
EXAMPLE 52
Solution
Syt Syc
Syt + Syc
(527)
A 25-mm-diameter shaft is statically torqued to 230 N · m. It is made of cast 195-T6
aluminum, with a yield strength in tension of 160 MPa and a yield strength in compression of 170 MPa. It is machined to nal diameter. Estimate the factor of safety of
the shaft.
The maximum shear stress is given by
τ=
16T
16(230)
=
3
πd
π 25 103
3
= 75 106 N/m2 = 75 MPa
The two nonzero principal stresses are 75 and 75 MPa, making the ordered principal
stresses σ1 = 75, σ2 = 0, and σ3 = 75 MPa. From Eq. (526), for yield,
Answer
n=
1
1
= 1.10
=
σ1 / Syt σ3 / Syc
75/160 (75)/170
Alternatively, from Eq. (527),
Ssy =
Syt Syc
160(170)
=
= 82.4 MPa
Syt + Syc
160 + 170
226
222
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
and τmax = 75 MPa. Thus,
Answer
57
n=
Ssy
82.4
= 1.10
=
τmax
75
Failure of Ductile Materials Summary
Having studied some of the various theories of failure, we shall now evaluate them and
show how they are applied in design and analysis. In this section we limit our studies to
materials and parts that are known to fail in a ductile manner. Materials that fail in a brittle manner will be considered separately because these require different failure theories.
To help decide on appropriate and workable theories of failure, Marin6 collected
data from many sources. Some of the data points used to select failure theories for ductile materials are shown in Fig. 515.7 Mann also collected many data for copper and
nickel alloys; if shown, the data points for these would be mingled with those already
diagrammed. Figure 515 shows that either the maximum-shear-stress theory or the
distortion-energy theory is acceptable for design and analysis of materials that would
Figure 515
2
Experimental data superposed
on failure theories. (From Fig.
7.11, p. 257, Mechanical
Behavior of Materials, 2nd
ed., N. E. Dowling, Prentice
Hall, Englewood Cliffs, N.J.,
1999. Modied to show only
ductile failures.)
/Sc
Oct. shear
Yielding ( Sc = Sy )
1.0
Ni-Cr-Mo steel
AISI 1023 steel
2024-T4 Al
3S-H Al
Max. shear
1.0
0
1.0
1
/Sc
1.0
6
Joseph Marin was one of the pioneers in the collection, development, and dissemination of material on the
failure of engineering elements. He has published many books and papers on the subject. Here the
reference used is Joseph Marin, Engineering Materials, Prentice-Hall, Englewood Cliffs, N.J., 1952.
(See pp. 156 and 157 for some data points used here.)
7
Note that some data in Fig. 515 are displayed along the top horizontal boundary where σ B σ A . This is
often done with failure data to thin out congested data points by plotting on the mirror image of the line
σB = σ A .
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
227
223
fail in a ductile manner. You may wish to plot other theories using a red or blue pencil
on Fig. 515 to show why they are not acceptable or are not used.
The selection of one or the other of these two theories is something that you, the
engineer, must decide. For design purposes the maximum-shear-stress theory is easy,
quick to use, and conservative. If the problem is to learn why a part failed, then the
distortion-energy theory may be the best to use; Fig. 515 shows that the plot of the
distortion-energy theory passes closer to the central area of the data points, and thus is
generally a better predictor of failure.
For ductile materials with unequal yield strengths, Syt in tension and Syc in compression, the Mohr theory is the best available. However, the theory requires the results
from three separate modes of tests, graphical construction of the failure locus, and tting the largest Mohrs circle to the failure locus. The alternative to this is to use the
Coulomb-Mohr theory, which requires only the tensile and compressive yield strengths
and is easily dealt with in equation form.
EXAMPLE 53
This example illustrates the use of a failure theory to determine the strength of a mechanical element or component. The example may also clear up any confusion existing
between the phrases strength of a machine part, strength of a material, and strength of
a part at a point.
A certain force F applied at D near the end of the 15-in lever shown in Fig. 516,
which is quite similar to a socket wrench, results in certain stresses in the cantilevered
bar OABC. This bar (OABC) is of AISI 1035 steel, forged and heat-treated so that it has
a minimum (ASTM) yield strength of 81 kpsi. We presume that this component would
be of no value after yielding. Thus the force F required to initiate yielding can be
regarded as the strength of the component part. Find this force.
y
Figure 516
2 in
O
A
12 in
z
1 1 -in D.
2
B
1
8
-in R.
2 in C
1-in D.
15 in
F
D
x
1 1 -in D.
2
228
224
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
Solution
We will assume that lever DC is strong enough and hence not a part of the problem. A 1035
steel, heat-treated, will have a reduction in area of 50 percent or more and hence is a ductile material at normal temperatures. This also means that stress concentration at shoulder
A need not be considered. A stress element at A on the top surface will be subjected to a
tensile bending stress and a torsional stress. This point, on the 1-in-diameter section, is the
weakest section, and governs the strength of the assembly. The two stresses are
σx =
32 M
M
32(14 F )
=
= 142.6 F
=
3
I /c
πd
π(13 )
τzx =
16T
Tr
16(15 F )
=
= 76.4 F
=
3
J
πd
π(13 )
Employing the distortion-energy theory, we nd, from Eq. (515), that
2
2
σ = σx + 3τzx
1/2
1/2
= [(142.6 F )2 + 3(76.4 F )2 ]
= 194.5 F
Equating the von Mises stress to Sy , we solve for F and get
Answer
F=
Sy
81 000
=
= 416 lbf
194.5
194.5
In this example the strength of the material at point A is Sy = 81 kpsi. The strength of
the assembly or component is F = 416 lbf.
Let us see how to apply the MSS theory. For a point undergoing plane stress with
only one non-zero normal stress and one shear stress, the two nonzero principal stresses
σ A and σ B will have opposite signs and hence t case 2 for the MSS theory. From
Eq. (313),
σ A σB = 2
σx
2
2
1/2
2
+ τzx
2
2
= σx + 4τzx
1/2
For case 2 of the MSS theory, Eq. (55) applies and hence
2
2
σx + 4τzx
1/2
2
= Sy
[(142.6 F ) + 4(76.4 F )2 ]1/2 = 209.0 F = 81 000
F = 388 lbf
which is about 7 percent less than found for the DE theory. As stated earlier, the MSS
theory is more conservative than the DE theory.
EXAMPLE 54
The cantilevered tube shown in Fig. 517 is to be made of 2014 aluminum alloy treated
to obtain a specied minimum yield strength of 276 MPa. We wish to select a stock-size
tube from Table A8 using a design factor n d = 4. The bending load is F = 1.75 kN,
the axial tension is P = 9.0 kN, and the torsion is T = 72 N · m. What is the realized
factor of safety?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 517
229
225
y
12
0m
m
F
z
P
T
x
Solution
Since the maximum bending moment is M = 120 F , the normal stress, for an element
on the top surface of the tube at the origin, is
σx =
Mc
9
120(1.75)(do /2)
9
105do
P
+
=+
=+
A
I
A
I
A
I
(1)
where, if millimeters are used for the area properties, the stress is in gigapascals.
The torsional stress at the same point is
72(do /2)
36do
Tr
=
=
J
J
J
τzx =
(2)
For accuracy, we choose the distortion-energy theory as the design basis. The von Mises
stress, as in the previous example, is
2
2
σ = σx + 3τzx
1/2
(3)
On the basis of the given design factor, the goal for σ is
σ
Sy
0.276
= 0.0690 GPa
=
nd
4
(4)
where we have used gigapascals in this relation to agree with Eqs. (1) and (2).
Programming Eqs. (1) to (3) on a spreadsheet and entering metric sizes from
Table A8 reveals that a 42- × 5-mm tube is satisfactory. The von Mises stress is found
to be σ = 0.06043 GPa for this size. Thus the realized factor of safety is
Answer
n=
Sy
0.276
= 4.57
=
σ
0.06043
For the next size smaller, a 42- × 4-mm tube, σ = 0.07105 GPa giving a factor of
safety of
n=
Sy
0.276
= 3.88
=
σ
0.07105
230
226
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
58
Maximum-Normal-Stress Theory
for Brittle Materials
The maximum-normal-stress (MNS) theory states that failure occurs whenever one of
the three principal stresses equals or exceeds the strength. Again we arrange the principal stresses for a general stress state in the ordered form σ1 σ2 σ3 . This theory
then predicts that failure occurs whenever
σ1 Sut
or
σ3 Suc
(528)
where Sut and Suc are the ultimate tensile and compressive strengths, respectively, given
as positive quantities.
For plane stress, with the principal stresses given by Eq. (313), with σ A σ B ,
Eq. (528) can be written as
σ A Sut
or
σ B Suc
(529)
which is plotted in Fig. 518a. As before, the failure criteria equations can be converted to
design equations. We can consider two sets of equations for load lines where σ A σ B as
Figure 518
(a) Graph of maximum-normalstress (MNS) theory of failure
for plane stress states. Stress
states that plot inside the
failure locus are safe.
(b) Load line plot.
B
Sut
Suc
A
Sut
Suc
(a)
B
Load line 1
O
A
Sut
Load line 2
Suc
Load line 4
(b)
Load line 3
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
σA =
Sut
n
σ A σB 0
Suc
n
227
Load line 1
σ A 0 σB
σB =
231
and
σ A 0 σB
and
Suc
σB
σA
Sut
Suc
σB
>
σA
Sut
0 σ A σB
Load line 2
(530a)
Load line 3
Load line 4
(530b)
where the load lines are shown in Fig. 518b.
Before we comment any further on the MNS theory we will explore some modications to the Mohr theory for brittle materials.
59
Modications of the Mohr Theory
for Brittle Materials
We will discuss two modications of the Mohr theory for brittle materials: the BrittleCoulomb-Mohr (BCM) theory and the modied Mohr (MM) theory. The equations
provided for the theories will be restricted to plane stress and be of the design type
incorporating the factor of safety.
The Coulomb-Mohr theory was discussed earlier in Sec. 56 with Eqs. (523) to
(525). Written as design equations for a brittle material, they are:
Brittle-Coulomb-Mohr
σA =
Sut
n
σA
σB
1
=
Sut
Suc
n
σB =
Suc
n
(531a)
σ A σB 0
(531b)
σ A 0 σB
(531c)
0 σ A σB
On the basis of observed data for the fourth quadrant, the modied Mohr theory
expands the fourth quadrant as shown in Fig. 519.
Modied Mohr
σA =
Sut
n
σ A σB 0
σ A 0 σB
σB
1
( Suc Sut ) σ A
=
Suc Sut
Suc
n
σB =
Suc
n
and
σ A 0 σB
0 σ A σB
(532a)
σB
1
σA
and
σB
>1
σA
(532b)
(532c)
Data are still outside this extended region. The straight line introduced by the modied
Mohr theory, for σ A 0 σ B and |σ B /σ A | > 1, can be replaced by a parabolic relation
232
228
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
Figure 519
B,
Biaxial fracture data of gray
cast iron compared with
various failure criteria.
(Dowling, N. E., Mechanical
Behavior of Materials, 2/e,
1999, p. 261. Reprinted by
permission of Pearson
Education, Inc., Upper Saddle
River, New Jersey.)
MPa
300
Sut
max. normal
ohr
d. M
mo
Suc
700
Cou
lomb
- Mo
hr
Sut
300
300
0
A,
MPa
Sut
r
To
sio
n
300
Gray cast-iron data
Suc
700
which can more closely represent some of the data.8 However, this introduces a nonlinear equation for the sake of a minor correction, and will not be presented here.
8
See J. E. Shigley, C. R. Mischke, R. G. Budynas, Mechanical Engineering Design, 7th ed., McGraw-Hill,
New York, 2004, p. 275.
EXAMPLE 55
Consider the wrench in Ex. 53, Fig. 516, as made of cast iron, machined to dimension. The force F required to fracture this part can be regarded as the strength of the
component part. If the material is ASTM grade 30 cast iron, nd the force F with
(a) Coulomb-Mohr failure model.
(b) Modied Mohr failure model.
Solution
We assume that the lever DC is strong enough, and not part of the problem. Since grade
30 cast iron is a brittle material and cast iron, the stress-concentration factors K t and K ts
are set to unity. From Table A24, the tensile ultimate strength is 31 kpsi and the compressive ultimate strength is 109 kpsi. The stress element at A on the top surface will be
subjected to a tensile bending stress and a torsional stress. This location, on the 1-indiameter section llet, is the weakest location, and it governs the strength of the assembly. The normal stress σx and the shear stress at A are given by
σx = K t
32(14 F )
M
32 M
= Kt
= (1)
= 142.6 F
I /c
π d3
π(1)3
τx y = K ts
16(15 F )
Tr
16T
= K ts
= (1)
= 76.4 F
3
J
πd
π(1)3
From Eq. (313) the nonzero principal stresses σ A and σ B are
σ A, σB =
142.6 F + 0
±
2
142.6 F 0
2
2
+ (76.4 F )2 = 175.8 F , 33.2 F
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
233
229
This puts us in the fourth-quadrant of the σ A , σ B plane.
(a) For BCM, Eq. (531b) applies with n = 1 for failure.
(33.2 F )
σB
175.8 F
σA
=1
=
Sut
Suc
31(103 )
109(103 )
Solving for F yields
Answer
F = 167 lbf
(b) For MM, the slope of the load line is |σ B /σ A | = 33.2/175.8 = 0.189 < 1.
Obviously, Eq. (532a) applies.
175.8 F
σA
=1
=
Sut
31(103 )
Answer
F = 176 lbf
As one would expect from inspection of Fig. 519, Coulomb-Mohr is more conservative.
510
Failure of Brittle Materials Summary
We have identied failure or strength of brittle materials that conform to the usual
meaning of the word brittle, relating to those materials whose true strain at fracture
is 0.05 or less. We also have to be aware of normally ductile materials that for some
reason may develop a brittle fracture or crack if used below the transition temperature. Figure 520 shows data for a nominal grade 30 cast iron taken under biaxial
Figure 520
A plot of experimental data
points obtained from tests on
cast iron. Shown also are the
graphs of three failure theories
of possible usefulness for brittle
materials. Note points A, B,
C, and D. To avoid
congestion in the rst
quadrant, points have been
plotted for σ A > σ B as well as
for the opposite sense. (Source
of data: Charles F. Walton
(ed.), Iron Castings
Handbook, Iron Founders
Society, 1971, pp. 215,
216, Cleveland, Ohio.)
B
Modified Mohr
Sut
30 Sut
120
Suc 90
60
30
30
ASTM No. 30 C.I.
Sut = 31 kpsi, Suc = 109 kpsi
30
Sut
A
Sut
B
Coulomb-Mohr
60
A
Maximum-normal-stress
90
B
120
A
C
D
A
B
150
Suc
= 1
234
230
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
stress conditions, with several brittle failure hypotheses shown, superposed. We note
the following:
In the rst quadrant the data appear on both sides and along the failure curves of
maximum-normal-stress, Coulomb-Mohr, and modied Mohr. All failure curves are
the same, and data t well.
In the fourth quadrant the modied Mohr theory represents the data best.
In the third quadrant the points A, B, C, and D are too few to make any suggestion
concerning a fracture locus.
511
Selection of Failure Criteria
For ductile behavior the preferred criterion is the distortion-energy theory, although
some designers also apply the maximum-shear-stress theory because of its simplicity
and conservative nature. In the rare case when Syt = Syc , the ductile Coulomb-Mohr
method is employed.
For brittle behavior, the original Mohr hypothesis, constructed with tensile, compression, and torsion tests, with a curved failure locus is the best hypothesis we have.
However, the difculty of applying it without a computer leads engineers to choose
modications, namely, Coulomb Mohr, or modied Mohr. Figure 521 provides a summary owchart for the selection of an effective procedure for analyzing or predicting
failures from static loading for brittle or ductile behavior.
Figure 521
Brittle behavior
Failure theory selection
owchart.
Ductile behavior
< 0.05
0.05
f
No
Mod. Mohr
(MM)
Eq. (5-32)
Conservative?
Yes
No
Yes
·
Syt = Syc?
Brittle Coulomb-Mohr Ductile Coulomb-Mohr
(BCM)
(DCM)
Eq. (5-31)
Eq. (5-26)
No
Conservative?
Distortion-energy
(DE)
Eqs. (5-15)
and (5-19)
Yes
Maximum shear stress
(MSS)
Eq. (5-3)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
512
235
231
Introduction to Fracture Mechanics
The idea that cracks exist in parts even before service begins, and that cracks can grow
during service, has led to the descriptive phrase damage-tolerant design. The focus of
this philosophy is on crack growth until it becomes critical, and the part is removed
from service. The analysis tool is linear elastic fracture mechanics (LEFM). Inspection
and maintenance are essential in the decision to retire parts before cracks reach catastrophic size. Where human safety is concerned, periodic inspections for cracks are
mandated by codes and government ordinance.
We shall now briey examine some of the basic ideas and vocabulary needed for
the potential of the approach to be appreciated. The intent here is to make the reader
aware of the dangers associated with the sudden brittle fracture of so-called ductile
materials. The topic is much too extensive to include in detail here and the reader is
urged to read further on this complex subject.9
The use of elastic stress-concentration factors provides an indication of the average
load required on a part for the onset of plastic deformation, or yielding; these factors
are also useful for analysis of the loads on a part that will cause fatigue fracture.
However, stress-concentration factors are limited to structures for which all dimensions
are precisely known, particularly the radius of curvature in regions of high stress concentration. When there exists a crack, aw, inclusion, or defect of unknown small radius
in a part, the elastic stress-concentration factor approaches innity as the root radius
approaches zero, thus rendering the stress-concentration factor approach useless.
Furthermore, even if the radius of curvature of the aw tip is known, the high local
stresses there will lead to local plastic deformation surrounded by a region of elastic
deformation. Elastic stress-concentration factors are no longer valid for this situation,
so analysis from the point of view of stress-concentration factors does not lead to criteria useful for design when very sharp cracks are present.
By combining analysis of the gross elastic changes in a structure or part that occur
as a sharp brittle crack grows with measurements of the energy required to produce new
fracture surfaces, it is possible to calculate the average stress (if no crack were present)
that will cause crack growth in a part. Such calculation is possible only for parts with
cracks for which the elastic analysis has been completed, and for materials that crack in a
relatively brittle manner and for which the fracture energy has been carefully measured.
The term relatively brittle is rigorously dened in the test procedures,10 but it means,
roughly, fracture without yielding occurring throughout the fractured cross section.
Thus glass, hard steels, strong aluminum alloys, and even low-carbon steel below
the ductile-to-brittle transition temperature can be analyzed in this way. Fortunately,
ductile materials blunt sharp cracks, as we have previously discovered, so that fracture
occurs at average stresses of the order of the yield strength, and the designer is prepared
9
References on brittle fracture include:
H. Tada and P. C. Paris, The Stress Analysis of Cracks Handbook, 2nd ed., Paris Productions,
St. Louis, 1985.
D. Broek, Elementary Engineering Fracture Mechanics, 4th ed., Martinus Nijhoff, London, 1985.
D. Broek, The Practical Use of Fracture Mechanics, Kluwar Academic Pub., London, 1988.
David K. Felbeck and Anthony G. Atkins, Strength and Fracture of Engineering Solids, Prentice-Hall,
Englewood Cliffs, N.J., 1984.
Kåre Hellan, Introduction to Fracture Mechanics, McGraw-Hill, New York, 1984.
10
BS 5447:1977 and ASTM E399-78.
236
232
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
for this condition. The middle ground of materials that lie between relatively brittle
and ductile is now being actively analyzed, but exact design criteria for these materials are not yet available.
Quasi-Static Fracture
Many of us have had the experience of observing brittle fracture, whether it is the breaking of a cast-iron specimen in a tensile test or the twist fracture of a piece of blackboard
chalk. It happens so rapidly that we think of it as instantaneous, that is, the cross section
simply parting. Fewer of us have skated on a frozen pond in the spring, with no one near
us, heard a cracking noise, and stopped to observe. The noise is due to cracking. The
cracks move slowly enough for us to see them run. The phenomenon is not instantaneous,
since some time is necessary to feed the crack energy from the stress eld to the crack for
propagation. Quantifying these things is important to understanding the phenomenon in
the small. In the large, a static crack may be stable and will not propagate. Some level of
loading can render the crack unstable, and the crack propagates to fracture.
The foundation of fracture mechanics was rst established by Grifth in 1921
using the stress eld calculations for an elliptical aw in a plate developed by Inglis in
1913. For the innite plate loaded by an applied uniaxial stress σ in Fig. 522, the maximum stress occurs at (±a, 0) and is given by
(σ y )max = 1 + 2
a
σ
b
(533)
Note that when a = b , the ellipse becomes a circle and Eq. (533) gives a stress concentration factor of 3. This agrees with the well-known result for an innite plate with
a circular hole (see Table A151). For a ne crack, b/a 0, and Eq. (534) predicts
that (σ y )max . However, on a microscopic level, an innitely sharp crack is a
hypothetical abstraction that is physically impossible, and when plastic deformation
occurs, the stress will be nite at the crack tip.
Grifth showed that the crack growth occurs when the energy release rate from
applied loading is greater than the rate of energy for crack growth. Crack growth can be
stable or unstable. Unstable crack growth occurs when the rate of change of the energy
release rate relative to the crack length is equal to or greater than the rate of change of
the crack growth rate of energy. Grifths experimental work was restricted to brittle
materials, namely glass, which pretty much conrmed his surface energy hypothesis.
However, for ductile materials, the energy needed to perform plastic work at the crack
tip is found to be much more crucial than surface energy.
Figure 522
y
b
x
a
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
237
233
Figure 523
Crack propagation modes.
Mode I
Mode II
Mode III
Crack Modes and the Stress Intensity Factor
Three distinct modes of crack propagation exist, as shown in Fig. 523. A tensile stress
eld gives rise to mode I, the opening crack propagation mode, as shown in Fig. 523a.
This mode is the most common in practice. Mode II is the sliding mode, is due to
in-plane shear, and can be seen in Fig. 523b. Mode III is the tearing mode, which
arises from out-of-plane shear, as shown in Fig. 523c. Combinations of these modes
can also occur. Since mode I is the most common and important mode, the remainder
of this section will consider only this mode.
Consider a mode I crack of length 2a in the innite plate of Fig. 524. By using
complex stress functions, it has been shown that the stress eld on a d x dy element in
the vicinity of the crack tip is given by
σx = σ
θ
θ
3θ
a
1 sin sin
cos
2r
2
2
2
(534a)
σy = σ
θ
3θ
θ
a
cos
1 + sin sin
2r
2
2
2
(534b)
τx y = σ
θ
θ
3θ
a
sin cos cos
2r
2
2
2
(534c)
0
ν(σx + σ y )
σz =
Figure 524
y
Mode I crack model.
dx
dy
r
a
x
(for plane stress)
(for plane strain)
(534d)
238
234
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
The stress σ y near the tip, with θ = 0, is
σ y |θ =0 = σ
a
2r
(a)
As with the elliptical crack, we see that σ y |θ =0 as r 0, and again the concept
of an
innite stress concentration at the crack tip is inappropriate. The quantity
σ y |θ =0 2r = σ a , however, does remain constant as r 0. It is common practice to
dene a factor K called the stress intensity factor given by
K = σ πa
(b)
where the units are MPa m or kpsi in. Since we are dealing with a mode I crack, Eq.
(b) is written as
KI = σ πa
(535)
The stress intensity factor is not to be confused with the static stress concentration
factors K t and K ts dened in Secs. 313 and 52.
Thus Eqs. (534) can be rewritten as
3θ
KI
θ
θ
1 sin sin
cos
σx =
2
2
2
2π r
(536a)
θ
θ
3θ
KI
cos
1 + sin sin
σy =
2
2
2
2π r
(536b)
θ
θ
3θ
KI
sin cos cos
τx y =
2
2
2
2π r
(536c)
σz =
0
ν(σx + σ y )
(for plane stress)
(for plane strain)
(536d)
The stress intensity factor is a function of geometry, size and shape of the crack,
and the type of loading. For various load and geometric congurations, Eq. (535) can
be written as
K I = βσ π a
(537)
where β is the stress intensity modication factor. Tables for β are available in the literature for basic congurations.11 Figures 525 to 530 present a few examples of β for
mode I crack propagation.
Fracture Toughness
When the magnitude of the mode I stress intensity factor reaches a critical value,
K I c crack propagation initiates. The critical stress intensity factor K I c is a material property that depends on the material, crack mode, processing of the material, temperature,
11
See, for example:
H. Tada and P. C. Paris, The Stress Analysis of Cracks Handbook, 2nd ed., Paris Productions, St. Louis, 1985.
G. C. Sib, Handbook of Stress Intensity Factors for Researchers and Engineers, Institute of Fracture and
Solid Mechanics, Lehigh University, Bethlehem, Pa., 1973.
Y. Murakami, ed., Stress Intensity Factors Handbook, Pergamon Press, Oxford, U.K., 1987.
W. D. Pilkey, Formulas for Stress, Strain, and Structural Matrices, 2nd ed. John Wiley& Sons,
New York, 2005.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
Figure 525
Off-center crack in a plate in
longitudinal tension; solid
curves are for the crack tip at
A; dashed curves are for the
tip at B.
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
2.2
239
AA
2.0
2a
A
A
B
d
1.8
2b
1.6
0.4
1.4
d b = 1.0
B
0.2
B
0.4
1.2
0.2
1.0
Figure 526
Plate loaded in longitudinal
tension with a crack at the
edge; for the solid curve there
are no constraints to bending;
the dashed curve was
obtained with bending
constraints added.
0
0.2
0.4
a d ratio
0.6
0.8
0.6
0.8
7.0
6.0
h
a
b
h
5.0
4.0
3.0
h b = 0.5
1.0
2.0
1.0
0
0.2
0.4
a b ratio
235
240
236
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
Figure 527
Beams of rectangular cross
section having an edge crack.
2.0
h
a
M
M
F
1.8
h
a
F
2
F
2
l
l
1.6
Pure bending
1.4
l =4
h
1.2
l =2
h
1.0
0
Figure 528
0.2
0.4
a h ratio
0.6
0.8
3
Plate in tension containing a
circular hole with two cracks.
2a
r = 0.5
b
2
r
2b
r = 0.25
b
1r
=0
b
0
0
0.2
0.4
a b ratio
0.6
0.8
loading rate, and the state of stress at the crack site (such as plane stress versus plane
strain). The critical stress intensity factor K I c is also called the fracture toughness of
the material. The fracture toughness for plane strain is normally lower than that for
plane stress. For this reason, the term K I c is typically dened as the mode I, plane strain
fracture toughness. Fracture toughness K I c for engineering metals lies in the range
20 K I 200 MPa · m; for engineering polymers and ceramics, 1 K I c
c
5 MPa · m. For a 4340 steel, where the yield strength due heat treatment ranges
to
from 800 to 1600 MPa, K I c decreases from 190 to 40 MPa · m.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 529
A cylinder loading in axial
tension having a radial crack
of depth a extending
completely around the
circumference of the cylinder.
241
237
4.0
ri ro = 0
a
a
3.0
0.1
0.4
2.0
1.0
Figure 530
Cylinder subjected to internal
pressure p, having a radial
crack in the longitudinal
direction of depth a. Use
Eq. (451) for the tangential
stress at r = r 0 .
0
0.8
ro
ri
0.2
0.4
a (ro ri ) ratio
0.6
0.8
0.6
0.8
3.4
a
3.0
pi
ri
ro
2.6
2.2
1.8
ri ro = 0.9
0.75
0.35
1.4
1.0
0
0.2
0.4
a (ro ri ) ratio
Table 51 gives some approximate typical room-temperature values of K I c for
several materials. As previously noted, the fracture toughness depends on many factors
and the table is meant only to convey some typical magnitudes of K I c . For an actual
application, it is recommended that the material specied for the application be certied using standard test procedures [see the American Society for Testing and Materials
(ASTM) standard E399].
242
238
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
Table 51
K Ic, MPa m
Material
Values of KIc for Some
Engineering Materials
at Room Temperature
Aluminum
2024
7075
7178
Sy, MPa
26
24
33
115
55
52100
860
1515
14
Steel
4340
4340
910
1035
99
60
Titanium
Ti-6AL-4V
Ti-6AL-4V
455
495
490
2070
One of the rst problems facing the designer is that of deciding whether the conditions exist, or not, for a brittle fracture. Low-temperature operation, that is, operation
below room temperature, is a key indicator that brittle fracture is a possible failure
mode. Tables of transition temperatures for various materials have not been published,
possibly because of the wide variation in values, even for a single material. Thus, in
many situations, laboratory testing may give the only clue to the possibility of a brittle
fracture. Another key indicator of the possibility of fracture is the ratio of the yield
strength to the ultimate strength. A high ratio of Sy / Su indicates there is only a small
ability to absorb energy in the plastic region and hence there is a likelihood of brittle
fracture.
The strength-to-stress ratio K I c / K I can be used as a factor of safety as
n=
KIc
KI
(538)
EXAMPLE 56
A steel ship deck plate is 30 mm thick and 12 m wide. It is loaded with a nominal uniaxial tensile stress of 50 MPa. It is operated below its ductile-to-brittle transition temperature with K I c equal to 28.3 MPa. If a 65-mm-long central transverse crack is
present, estimate the tensile stress at which catastrophic failure will occur. Compare this
stress with the yield strength of 240 MPa for this steel.
Solution
For Fig. 525, with d = b, 2a = 65 mm and 2b = 12 m, so that d /b = 1 and a /d =
65/12(103 ) = 0.00542. Since a/d is so small, β = 1, so that
K I = σ π a = 50 π (32.5 × 103 ) = 16.0 MPa m
From Eq. (538),
n=
KIc
28.3
= 1.77
=
KI
16.0
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
243
239
The stress at which catastrophic failure occurs is
Answer
σc =
KIc
28.3
(50) = 88.4 MPa
σ=
KI
16.0
The yield strength is 240 MPa, and catastrophic failure occurs at 88.4/240 = 0.37, or
at 37 percent of yield. The factor of safety in this circumstance is K I c / K I =
28.3/16 = 1.77 and not 240/50 = 4.8.
EXAMPLE 57
A plate of width 1.4 m and length 2.8 m is required to support a tensile force in the
2.8-m direction of 4.0 MN. Inspection procedures will detect only through-thickness
edge cracks larger than 2.7 mm. The two Ti-6AL-4V alloys in Table 51 are being considered for this application, for which the safety factor must be 1.3 and minimum
weight is important. Which alloy should be used?
Solution
(a) We elect rst to estimate the thickness required to resist yielding. Since σ = P /wt ,
we have t = P /wσ. For the weaker alloy, we have, from Table 51, Sy = 910 MPa.
Thus,
σall =
Sy
910
=
= 700 MPa
n
1.3
Thus
t=
4.0(10)3
P
=
= 4.08 mm or greater
wσall
1.4(700)
For the stronger alloy, we have, from Table 51,
σall =
1035
= 796 MPa
1.3
and so the thickness is
Answer
t=
P
4.0(10)3
= 3.59 mm or greater
=
wσall
1.4(796)
(b) Now let us nd the thickness required to prevent crack growth. Using Fig. 526, we
have
h
2.8/2
=
=1
b
1.4
a
2.7
=
= 0.001 93
b
1.4(103 )
.
Corresponding to these ratios we nd from Fig. 526 that β = 1.1, and K I = 1.1σ π a .
KIc
115 103
KIc
n=
=
σ=
,
KI
1.1σ π a
1.1n π a
244
240
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
From Table 51, K I c = 115 MPa m for the weaker of the two alloys. Solving for σ
with n = 1 gives the fracture stress
σ=
115
1.1 π (2.7 × 103 )
= 1135 MPa
which is greater than the yield strength of 910 MPa, and so yield strength is the basis
for the geometry decision. For the stronger alloy Sy = 1035 MPa, with n = 1 the fracture stress is
σ=
55
KIc
= 542.9 MPa
=
nKI
1(1.1) π (2.7 × 103 )
which is less than the yield strength of 1035 MPa. The thickness t is
t=
P
4.0(103 )
= 6.84 mm or greater
=
wσall
1.4(542.9/1.3)
This example shows that the fracture toughness K I c limits the geometry when the
stronger alloy is used, and so a thickness of 6.84 mm or larger is required. When the
weaker alloy is used the geometry is limited by the yield strength, giving a thickness of
only 4.08 mm or greater. Thus the weaker alloy leads to a thinner and lighter weight
choice since the failure modes differ.
513
Stochastic Analysis12
Reliability is the probability that machine systems and components will perform their
intended function satisfactorily without failure. Up to this point, discussion in this chapter has been restricted to deterministic relations between static stress, strength, and the
design factor. Stress and strength, however, are statistical in nature and very much tied
to the reliability of the stressed component. Consider the probability density functions
for stress and strength, and S, shown in Fig. 531a. The mean values of stress and
strength are µσ and µ S , respectively. Here, the average factor of safety is
n=
¯
µS
µσ
(a)
The margin of safety for any value of stress σ and strength S is dened as
m = Sσ
(b)
¯
The average part will have a margin of safety of m = µ S µσ . However, for the overlap of the distributions shown by the shaded area in Fig. 531a, the stress exceeds the
strength, the margin of safety is negative, and these parts are expected to fail. This
shaded area is called the interference of and S.
Figure 531b shows the distribution of m, which obviously depends on the distributions of stress and strength. The reliability that a part will perform without failure, R,
is the area of the margin of safety distribution for m > 0. The interference is the area
12
Review Chap. 20 before reading this section.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
245
241
S
s
Stress
(a)
m
f (m)
Plot of density functions
showing how the interference
of S and is used to obtain
the stress margin m. (a) Stress
and strength distributions.
(b) Distribution of interference;
the reliability R is the area of
the density function for m
greater than zero; the
interference is the area
(1 R).
f (s), f ( )
Figure 531
(1 R)
R
+
0
m
Stress margin
(b)
1 R where parts are expected to fail. We next consider some typical cases involving
stress-strength interference.
Normal-Normal Case
ˆ
ˆ
Consider the normal distributions, S = N(µ S , σ S ) and = N(µσ , σσ ) . The stress
margin is m = S , and will be normally distributed because the addition or subˆ
traction of normals is normal. Thus m = N(µm , σm ) . Reliability is the probability p that
m > 0. That is,
(539)
R = p( S > σ ) = p( S σ > 0) = p(m > 0)
To nd the chance that m > 0 we form the z variable of m and substitute m = 0 [See
ˆ2 ˆ2
ˆ
Eq. (2016)]. Noting that µm = µ S µσ and σm = (σ S + σσ )1/2 , we write
z=
0 µm
µm
µ S µσ
m µm
=
=
=
1/2
σm
ˆ
σm
ˆ
σm
ˆ
σ S + σσ
ˆ2 ˆ2
(540)
Equation (540) is called the normal coupling equation. The reliability associated with
z is given by
R=
x
u2
1
du = 1 F = 1
exp
2
2π
(z )
(541)
The body of Table A10 gives R when z > 0 and (1 R = F ) when z 0. Noting that
n = µ S /µσ , square both sides of Eq. (540), and introduce C S and Cσ where
¯
¯
Cs = σs /µs and Cσ = σσ /µσ . Solve the resulting quadratic for n to obtain
ˆ
ˆ
n=
¯
1±
2
1 1 z2C S
2
1 z2C S
2
1 z 2 Cσ
The plus sign is associated with R > 0.5, and the minus sign with R < 0.5.
(542)
246
242
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
LognormalLognormal Case
ˆ
ˆ
Consider the lognormal distributions S = LN(µ S , σ S ) and = LN(µσ , σσ ). If we
interfere their companion normals using Eqs. (2018) and (2019), we obtain
2
µln S = ln µ S ln 1 + C S
σln S =
ˆ
(strength)
2
ln 1 + C S
and
2
µln σ = ln µσ ln 1 + Cσ
σln σ =
ˆ
(stress)
2
ln 1 + Cσ
Using Eq. (540) for interfering normal distributions gives
z=
µln S µln σ
ˆ2
σln S + σln σ
ˆ2
1/2
=
2
1 + Cσ
2
1 + CS
µS
µσ
ln
2
ln 1 + C S
(543)
2
1 + Cσ
The reliability R is expressed by Eq. (541). The design factor n is the random variable
that is the quotient of S/ . The quotient of lognormals is lognormal, so pursuing the
z variable of the lognormal n, we note
µn =
µS
µσ
Cn =
2
2
C S + Cσ
2
1 + Cσ
σn = Cn µn
ˆ
ˆ
The companion normal to n = LN(µn , σn ) , from Eqs. (2018) and (2019), has a mean
and standard deviation of
2
µ y = ln µn ln 1 + Cn
σy =
ˆ
2
ln 1 + Cn
The z variable for the companion normal y distribution is
z=
y µy
σy
ˆ
¯
Failure will occur when the stress is greater than the strength, when n < 1, or when
y < 0.
z=
2
2
ln µn / 1 + Cn
ln µn ln 1 + Cn
µy
0 µy
=
=
=
˙
σy
ˆ
σy
2
2
ln 1 + Cn
ln 1 + Cn
(544)
Solving for µn gives
Cn
.
2
2
µn = n = exp z ln 1 + Cn + ln 1 + Cn = exp Cn z +
¯
2
(545)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
247
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
243
Equations (542) and (545) are remarkable for several reasons:
¯
They relate design factor n to the reliability goal (through z) and the coefcients of
variation of strength and stress.
They are not functions of the means of stress and strength.
They estimate the design factor necessary to achieve the reliability goal before decisions involving means are made. The C S depends slightly on the particular material.
The Cσ has the coefcient of variation (COV) of the load, and that is generally given.
EXAMPLE 58
Solution
A round cold-drawn 1018 steel rod has an 0.2 percent yield strength S y = N(78.4, 5.90)
kpsi and is to be subjected to a static axial load of P = N(50, 4.1) kip. What value of
¯
the design factor n corresponds to a reliability of 0.999 against yielding (z = 3.09)?
Determine the corresponding diameter of the rod.
C S = 5.90/78.4 = 0.0753 , and
P
4P
=
A
π d2
Since the COV of the diameter is an order of magnitude less than the COV of the load
or strength, the diameter is treated deterministically:
4.1
Cσ = C P =
= 0.082
50
From Eq. (542),
=
1
1
[1
( 3.09) (0.0753 )][1
2
( 3.09) (0.082 )]
2
2
2
n
1
2
2
( 3.09) (0.0753 )
The diameter is found deterministically:
Answer
d=
Check
¯
4P
=
¯¯
π Sy /n
4(50 000)
= 1.072 in
π(78 400)/1.416
S y = N(78.4, 5.90) kpsi, P = N(50, 4.1) kip, and d = 1.072 in. Then
π d2
π(1.0722 )
=
= 0.9026 in2
4
4
¯
(50 000)
P
=
= 55 400 psi
σ=
¯
A
0.9026
4.1
= 0.082
C P = Cσ =
50
A=
σσ = Cσ σ = 0.082(55 400) = 4540 psi
ˆ
¯
From Eq. (540)
σ S = 5.90 kpsi
ˆ
78.4 55.4
= 3.09
(5.902 + 4.542 )1/2
From Appendix Table A10, R = (3.09) = 0.999.
z=
1.416
248
244
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
EXAMPLE 59
Solution
Rework Ex. 58 with lognormally distributed stress and strength.
C S = 5.90/78.4 = 0.0753, and Cσ = C P = 4.1/50 = 0.082. Then
P
4P
=
A
π d2
=
2
2
C S + Cσ
=
2
1 + Cσ
Cn =
0.07532 + 0.0822
= 0.1110
1 + 0.0822
From Table A10, z = 3.09. From Eq. (545),
n = exp (3.09) ln(1 + 0.1112 ) + ln 1 + 0.1112 = 1.416
¯
d=
Check
¯
4P
=
¯¯
π Sy /n
4(50 000)
= 1.0723 in
π(78 400)/1.416
S y = LN(78.4, 5.90), P = LN (50, 4.1) kip. Then
π d2
π(1.07232 )
=
= 0.9031
4
4
¯
50 000
P
=
= 55 365 psi
σ=
¯
A
0.9031
4.1
= 0.082
Cσ = C P =
50
A=
σσ = Cσ µσ = 0.082(55 367) = 4540 psi
ˆ
From Eq. (543),
z=
ln
78.4
55.365
2
1 + 0.082
1 + 0.07532
ln[(1 + 0.07532 )(1 + 0.0822 )]
= 3.1343
Appendix Table A10 gives R = 0.99950.
InterferenceGeneral
In the previous segments, we employed interference theory to estimate reliability when the
distributions are both normal and when they are both lognormal. Sometimes, however, it
turns out that the strength has, say, a Weibull distribution while the stress is distributed
lognormally. In fact, stresses are quite likely to have a lognormal distribution, because the
multiplication of variates that are normally distributed produces a result that approaches
lognormal. What all this means is that we must expect to encounter interference problems
involving mixed distributions and we need a general method to handle the problem.
It is quite likely that we will use interference theory for problems involving distributions other than strength and stress. For this reason we employ the subscript 1 to
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
Figure 532
249
245
f1(S )
(a) PDF of the strength
distribution; (b) PDF of the
load-induced stress
distribution.
dF1(x) = f1(x) dx
S
dx
(a)
x
Cursor
f2( )
F2(x)
R2(x)
(b)
designate the strength distribution and the subscript 2 to designate the stress distribution. Figure 532 shows these two distributions aligned so that a single cursor x can be
used to identify points on both distributions. We can now write
Probability that
stress is less
than strength
= dp(σ < x ) = d R = F2 (x ) d F1 (x )
By substituting 1 R2 for F2 and d R1 for d F1 , we have
d R = [1 R2 (x )] d R1 (x )
The reliability for all possible locations of the cursor is obtained by integrating x
from to ; but this corresponds to an integration from 1 to 0 on the reliability R1 .
Therefore
0
R=
1
[1 R2 (x )] d R1 (x )
which can be written
1
R2 d R1
R =1
(546)
f1( S) d S
(547)
f 2 (σ ) d σ
(548)
0
where
R1 ( x ) =
R2 ( x ) =
x
x
250
246
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
1
1
R2
R2
R1
R1
1
(a)
1
(b)
Figure 533
Curve shapes of the R 1 R 2 plot. In each case the shaded area
is equal to 1 R and is obtained by numerical integration.
(a) Typical curve for asymptotic distributions; (b) curve shape
obtained from lower truncated distributions such as the Weibull.
For the usual distributions encountered, plots of R1 versus R2 appear as shown in
Fig. 533. Both of the cases shown are amenable to numerical integration and computer solution. When the reliability is high, the bulk of the integration area is under the
right-hand spike of Fig. 533a.
514
Important Design Equations
The following equations and their locations are provided as a summary.
Maximum Shear Theory
τmax =
p. 212
Sy
σ1 σ3
=
2
2n
(53)
Distortion-Energy Theory
Von Mises stress, p. 214
σ =
(σ1 σ2 )2 + (σ2 σ3 )2 + (σ3 σ1 )2
2
1/2
1
2
2
2
p. 215 σ = (σx σ y )2 + (σ y σz )2 + (σz σx )2 + 6(τx y + τ yz + τzx )
2
(512)
1/2
(514)
Plane stress, p. 214
2
2
σ = (σ A σ A σ B + σ B )1/2
p. 215
2
2
2
σ = (σx σx σ y + σ y + 3τx y )1/2
(513)
(515)
Yield design equation, p. 216
σ =
Sy
n
(519)
Shear yield strength, p. 217
Ssy = 0.577 Sy
(521)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
251
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
247
Coulomb-Mohr Theory
σ1
σ3
1
=
St
Sc
n
p. 221
(526)
where St is tensile yield (ductile) or ultimate tensile (brittle), and St is compressive
yield (ductile) or ultimate compressive (brittle) strengths.
Maximum-Normal-Stress Theory
σ1 =
p. 226
Sut
n
σ3 =
or
Suc
n
(530)
Modied Mohr (Plane Stress)
Use maximum-normal-stress equations, or
p. 227
σB
1
( Suc Sut )σ A
=
Suc Sut
Suc
n
σ A 0 σB
and
σB
>1
σA
(532b)
Failure Theory Flowchart
Fig. 521, p. 230
Brittle behavior
Ductile behavior
< 0.05
0.05
f
No
Mod. Mohr
(MM)
Eq. (5-32)
Conservative?
Yes
No
Yes
·
Syt = Syc?
Brittle Coulomb-Mohr Ductile Coulomb-Mohr
(BCM)
(DCM)
Eq. (5-31)
Eq. (5-26)
No
Conservative?
Distortion-energy
(DE)
Eqs. (5-15)
and (5-19)
Yes
Maximum shear stress
(MSS)
Eq. (5-3)
Fracture Mechanics
p. 234
K I = βσ π a
where β is found in Figs. 525 to 530 (pp. 235 to 237)
(537)
252
248
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
n=
p. 238
KIc
KI
(538)
where K I c is found in Table 51 (p. 238)
Stochastic Analysis
¯
Mean factor of safety dened as n = µ S /µσ (µ S and µσ are mean strength and stress,
respectively)
Normal-Normal Case
n=
p. 241
1±
2
2
1 (1 z 2 Cs )(1 z 2 Cσ )
2C 2
1z s
(542)
ˆ
ˆ
where z can be found in Table A10, C S = σ S /µ S , and Cσ = σσ /µσ .
Lognormal-Lognormal Case
p. 242
Cn
.
2
2
n = exp z ln(1 + Cn ) + ln 1 + Cn = exp Cn z +
2
(545)
where
Cn =
2
2
C S + Cσ
2
1 + Cσ
(See other denitions in normal-normal case.)
PROBLEMS
51
A ductile hot-rolled steel bar has a minimum yield strength in tension and compression of 50 kpsi.
Using the distortion-energy and maximum-shear-stress theories determine the factors of safety
for the following plane stress states:
(a) σx = 12 kpsi, σ y = 6 kpsi
(b) σx = 12 kpsi, τx y = 8 kpsi
(c) σx = 6 kpsi, σ y = 10 kpsi, τx y = 5 kpsi
(d ) σx = 12 kpsi, σ y = 4 kpsi, τx y = 1 kpsi
52
Repeat Prob. 51 for:
(a) σ A = 12 kpsi, σ B = 12 kpsi
(b) σ A = 12 kpsi, σ B = 6 kpsi
(c) σ A = 12 kpsi, σ B = 12 kpsi
(d ) σ A = 6 kpsi, σ B = 12 kpsi
53
Repeat Prob. 51 for a bar of AISI 1020 cold-drawn steel and:
(a) σx = 180 MPa, σ y = 100 MPa
(b) σx = 180 MPa, τx y = 100 MPa
(c) σx = 160 MPa, τx y = 100 MPa
(d ) τx y = 150 MPa
54
Repeat Prob. 51 for a bar of AISI 1018 hot-rolled steel and:
(a) σ A = 100 MPa, σ B = 80 MPa
(b) σ A = 100 MPa, σ B = 10 MPa
(c) σ A = 100 MPa, σ B = 80 MPa
(d ) σ A = 80 MPa, σ B = 100 MPa
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Failures Resulting from Static Loading
253
249
55
Repeat Prob. 53 by rst plotting the failure loci in the σ A , σ B plane to scale; then, for each stress
state, plot the load line and by graphical measurement estimate the factors of safety.
56
Repeat Prob. 54 by rst plotting the failure loci in the σ A , σ B plane to scale; then, for each stress
state, plot the load line and by graphical measurement estimate the factors of safety.
57
An ASTM cast iron has minimum ultimate strengths of 30 kpsi in tension and 100 kpsi in compression. Find the factors of safety using the MNS, BCM, and MM theories for each of the following stress states. Plot the failure diagrams in the σ A , σ B plane to scale and locate the
coordinates of each stress state.
(a) σx = 20 kpsi, σ y = 6 kpsi
(b) σx = 12 kpsi, τx y = 8 kpsi
(c) σx = 6 kpsi, σ y = 10 kpsi, τx y = 5 kpsi
(d ) σx = 12 kpsi, τx y = 8 kpsi
58
For Prob. 57, case (d ), estimate the factors of safety from the three theories by graphical measurements of the load line.
59
Among the decisions a designer must make is selection of the failure criteria that is applicable to
the material and its static loading. A 1020 hot-rolled steel has the following properties:
Sy = 42 kpsi, Sut = 66.2 kpsi, and true strain at fracture ε f = 0.90. Plot the failure locus and, for
the static stress states at the critical locations listed below, plot the load line and estimate the factor of safety analytically and graphically.
(a) σx = 9 kpsi, σ y = 5 kpsi.
(b) σx = 12 kpsi, τx y = 3 kpsi ccw.
(c) σx = 4 kpsi, σ y = 9 kpsi, τx y = 5 kpsi cw.
(d) σx = 11 kpsi, σ y = 4 kpsi, τx y = 1 kpsi cw.
510
A 4142 steel Q&T at 80 F exhibits Syt = 235 kpsi, Syc = 275 kpsi, and ε f = 0.06. Choose and
plot the failure locus and, for the static stresses at the critical locations, which are 10 times those
in Prob. 59, plot the load lines and estimate the factors of safety analytically and graphically.
511
For grade 20 cast iron, Table A24 gives Sut = 22 kpsi, Suc = 83 kpsi. Choose and plot the failure locus and, for the static loadings inducing the stresses at the critical locations of Prob. 59,
plot the load lines and estimate the factors of safety analytically and graphically.
512
A cast aluminum 195-T6 has an ultimate strength in tension of Sut = 36 kpsi and ultimate
strength in compression of Suc = 35 kpsi, and it exhibits a true strain at fracture ε f = 0.045.
Choose and plot the failure locus and, for the static loading inducing the stresses at the critical
locations of Prob. 59, plot the load lines and estimate the factors of safety analytically and graphically.
513
An ASTM cast iron, grade 30 (see Table A24), carries static loading resulting in the stress state
listed below at the critical locations. Choose the appropriate failure locus, plot it and the load
lines, and estimate the factors of safety analytically and graphically.
(a) σ A = 20 kpsi, σ B = 20 kpsi.
(b) τx y = 15 kpsi.
(c) σ A = σ B = 80 kpsi.
(d ) σ A = 15 kpsi, σ B = 25 kpsi.
514
This problem illustrates that the factor of safety for a machine element depends on the particular point
selected for analysis. Here you are to compute factors of safety, based upon the distortion-energy
theory, for stress elements at A and B of the member shown in the gure. This bar is made of AISI
1006 cold-drawn steel and is loaded by the forces F = 0.55 kN, P = 8.0 kN, and T = 30 N · m.
254
250
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
y
10
0m
m
A
B
Problem 514
F
z
20-mm D.
P
T
x
515
The gure shows a crank loaded by a force F = 190 lbf which causes twisting and bending of
the 3 -in-diameter shaft xed to a support at the origin of the reference system. In actuality, the
4
support may be an inertia which we wish to rotate, but for the purposes of a strength analysis we
can consider this to be a statics problem. The material of the shaft AB is hot-rolled AISI 1018
steel (Table A20). Using the maximum-shear-stress theory, nd the factor of safety based on the
stress at point A.
y
1 in
F
C
A
3
-in
4
1
-in
2
dia.
1
4
Problem 515
B
in
1
14
dia.
in
z
4 in
5 in
x
516
517*
518
Solve Prob. 515 using the distortion energy theory. If you have solved Prob. 515, compare the
results and discuss the difference.
Design the lever arm CD of Fig. 516 by specifying a suitable size and material.
A spherical pressure vessel is formed of 18-gauge (0.05-in) cold-drawn AISI 1018 sheet steel. If the
vessel has a diameter of 8 in, estimate the pressure necessary to initiate yielding. What is the estimated bursting pressure?
*The asterisk indicates a problem that may not have a unique result or may be a particularly challenging
problem.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
255
251
519
This problem illustrates that the strength of a machine part can sometimes be measured in units
other than those of force or moment. For example, the maximum speed that a ywheel can reach
without yielding or fracturing is a measure of its strength. In this problem you have a rotating ring
made of hot-forged AISI 1020 steel; the ring has a 6-in inside diameter and a 10-in outside diameter
and is 1.5 in thick. What speed in revolutions per minute would cause the ring to yield? At what
radius would yielding begin? [Note: The maximum radial stress occurs at r = (ro ri )1/2 ; see Eq.
(355).]
520
A light pressure vessel is made of 2024-T3 aluminum alloy tubing with suitable end closures.
1
This cylinder has a 3 2 -in OD, a 0.065-in wall thickness, and ν = 0.334. The purchase order species a minimum yield strength of 46 kpsi. What is the factor of safety if the pressure-release valve
is set at 500 psi?
521
A cold-drawn AISI 1015 steel tube is 300 mm OD by 200 mm ID and is to be subjected to an
external pressure caused by a shrink t. What maximum pressure would cause the material of the
tube to yield?
522
What speed would cause fracture of the ring of Prob. 519 if it were made of grade 30 cast iron?
523
The gure shows a shaft mounted in bearings at A and D and having pulleys at B and C. The
forces shown acting on the pulley surfaces represent the belt tensions. The shaft is to be made of
ASTM grade 25 cast iron using a design factor n d = 2.8. What diameter should be used for the
shaft?
x
6-in D.
300 lbf
50 lbf
y
27 lbf
Problem 523
8-in D.
z
A
B
360 lbf D
C 6 in
8 in
8 in
524
By modern standards, the shaft design of Prob. 523 is poor because it is so long. Suppose it is
redesigned by halving the length dimensions. Using the same material and design factor as in
Prob. 523, nd the new shaft diameter.
525
The gear forces shown act in planes parallel to the yz plane. The force on gear A is 300 lbf.
Consider the bearings at O and B to be simple supports. For a static analysis and a factor of safety of 3.5, use distortion energy to determine the minimum safe diameter of the shaft. Consider
the material to have a yield strength of 60 kpsi.
526
Repeat Prob. 525 using maximum-shear-stress.
527
The gure is a schematic drawing of a countershaft that supports two V-belt pulleys. For each
pulley, the belt tensions are parallel. For pulley A consider the loose belt tension is 15 percent of
the tension on the tight side. A cold-drawn UNS G10180 steel shaft of uniform diameter is to be
selected for this application. For a static analysis with a factor of safety of 3.0, determine the
minimum preferred size diameter. Use the distortion-energy theory.
256
252
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
y
20 in
O
16 in
FC
10 in
Problem 525
20°
z
Gear A
24-in D.
B
A
C
FA
Gear C
10-in D.
x
20°
y
300
45°
O
400
T2
Problem 527
T1
z
150
Dimensions in millimeters
250 Dia.
300 Dia.
A
50 N
B
C
x
270 N
528
Repeat Prob. 527 using maximum shear stress.
529
The clevis pin shown in the gure is 12 mm in diameter and has the dimensions a = 12 mm and
b = 18 mm. The pin is machined from AISI 1018 hot-rolled steel (Table A20) and is to be
loaded to no more than 4.4 kN. Determine whether or not the assumed loading of gure c yields
a factor of safety any different from that of gure d. Use the maximum-shear-stress theory.
530
Repeat Prob. 529, but this time use the distortion-energy theory.
531
1
A split-ring clamp-type shaft collar is shown in the gure. The collar is 2 in OD by 1 in ID by 2
1
in wide. The screw is designated as 4 -28 UNF. The relation between the screw tightening torque
T, the nominal screw diameter d, and the tension in the screw Fi is approximately T = 0.2 Fi d .
The shaft is sized to obtain a close running t. Find the axial holding force Fx of the collar as a
function of the coefcient of friction and the screw torque.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Failures Resulting from Static Loading
257
253
F
(b)
Problem 529
b
2
d
a+b
a
a
(c)
F
b
b
(a)
a+b
(d )
A
Problem 531
532
Suppose the collar of Prob. 531 is tightened by using a screw torque of 190 lbf · in. The collar
material is AISI 1040 steel heat-treated to a minimum tensile yield strength of 63 kpsi.
(a) Estimate the tension in the screw.
(b) By relating the tangential stress to the hoop tension, nd the internal pressure of the shaft on
the ring.
(c) Find the tangential and radial stresses in the ring at the inner surface.
(d ) Determine the maximum shear stress and the von Mises stress.
(e) What are the factors of safety based on the maximum-shear-stress hypothesis and the distortionenergy theory?
533
In Prob. 531, the role of the screw was to induce the hoop tension that produces the clamping.
The screw should be placed so that no moment is induced in the ring. Just where should the screw
be located?
534
A tube has another tube shrunk over it. The specications are:
Inner Member
ID
OD
Outer Member
1.000 ± 0.002 in
2.000 ± 0.0004 in
1.999 ± 0.0004 in
3.000 ± 0.004 in
Both tubes are made of a plain carbon steel.
(a) Find the nominal shrink-t pressure and the von Mises stresses at the t surface.
(b) If the inner tube is changed to solid shafting with the same outside dimensions, nd the
nominal shrink-t pressure and the von Mises stresses at the t surface.
258
254
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
5. Failures Resulting from
Static Loading
Mechanical Engineering Design
535
Steel tubes with a Youngs modulus of 207 GPa have the specications:
Inner Tube
Outer Tube
ID
25 ± 0.050 mm
49.98 ± 0.010 mm
OD
50 ± 0.010 mm
75 ± 0.10 mm
These are shrink-tted together. Find the nominal shrink-t pressure and the von Mises stress in
each body at the t surface.
536
Repeat Prob. 535 for maximum shrink-t conditions.
537
A 2-in-diameter solid steel shaft has a gear with ASTM grade 20 cast-iron hub ( E = 14.5 Mpsi)
shrink-tted to it. The specications for the shaft are
2.000
+ 0.0000
0.0004
in
1
The hole in the hub is sized at 1.999 ± 0.0004 in with an OD of 4.00 ± 32 in. Using the midrange
values and the modied Mohr theory, estimate the factor of safety guarding against fracture in the
gear hub due to the shrink t.
538
Two steel tubes are shrink-tted together where the nominal diameters are 1.50, 1.75, and 2.00
in. Careful measurement before tting revealed that the diametral interference between the tubes
to be 0.00246 in. After the t, the assembly is subjected to a torque of 8000 lbf · in and a bending-moment of 6000 lbf · in. Assuming no slipping between the cylinders, analyze the outer
cylinder at the inner and outer radius. Determine the factor of safety using distortion energy with
Sy = 60 kpsi.
539
Repeat Prob. 538 for the inner tube.
540
For Eqs. (536) show that the principal stresses are given by
θ
KI
θ
1 + sin
σ1 =
cos
2
2
2π r
σ3 =
θ
θ
KI
1 sin
cos
σ2 =
2
2
2π r
(plane stress)
0
θ
2
ν K I cos
πr
2
(plane strain)
541
Use the results of Prob. 540 for plane strain near the tip with θ = 0 and ν = 1 . If the yield
3
strength of the plate is Sy , what is σ1 when yield occurs?
(a) Use the distortion-energy theory.
(b) Use the maximum-shear-stress theory. Using Mohrs circles, explain your answer.
542
A plate 4 in wide, 8 in long, and 0.5 in thick is loaded in tension in the direction of the length.
The plate contains a crack as
shown in Fig. 526 with the crack length of 0.625 in. The material
is steel with K I c = 70 kpsi · in, and Sy = 160 kpsi. Determine the maximum possible load that
can be applied before the plate (a) yields, and (b) has uncontrollable crack growth.
543
A cylinder subjected to internal pressure pi has an outer diameter of 350 mm and a 25-mm wall
thickness. For the cylinder material, K I c = 80 MPa · m, Sy = 1200 MPa, and Sut = 1350 MPa.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
5. Failures Resulting from
Static Loading
© The McGrawHill
Companies, 2008
Failures Resulting from Static Loading
259
255
If the cylinder contains a radial crack in the longitudinal direction of depth 12.5 mm determine
the pressure that will cause uncontrollable crack growth.
544
A carbon steel collar of length 1 in is to be machined to inside and outside diameters, respectively, of
Di = 0.750 ± 0.0004 in
Do = 1.125 ± 0.002 in
This collar is to be shrink-tted to a hollow steel shaft having inside and outside diameters,
respectively, of
di = 0.375 ± 0.002 in
do = 0.752 ± 0.0004 in
These tolerances are assumed to have a normal distribution, to be centered in the spread interval,
and to have a total spread of ±4 standard deviations. Determine the means and the standard deviations of the tangential stress components for both cylinders at the interface.
545
Suppose the collar of Prob. 544 has a yield strength of S y = N(95.5, 6.59) kpsi. What is the
probability that the material will not yield?
546
A carbon steel tube has an outside diameter of 1 in and a wall thickness of 1 in. The tube is to
8
carry an internal hydraulic pressure given as p = N(6000, 500) psi. The material of the tube has
a yield strength of S y = N(50, 4.1) kpsi. Find the reliability using thin-wall theory.
260
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
6
Fatigue Failure Resulting
from Variable Loading
Chapter Outline
61
Introduction to Fatigue in Metals
62
Approach to Fatigue Failure in Analysis and Design
63
Fatigue-Life Methods
64
The Stress-Life Method
265
65
The Strain-Life Method
268
66
The Linear-Elastic Fracture Mechanics Method
67
The Endurance Limit
68
Fatigue Strength
69
Endurance Limit Modifying Factors
258
264
265
270
274
275
278
610
Stress Concentration and Notch Sensitivity
611
Characterizing Fluctuating Stresses
612
Fatigue Failure Criteria for Fluctuating Stress
613
Torsional Fatigue Strength under Fluctuating Stresses
614
Combinations of Loading Modes
615
Varying, Fluctuating Stresses; Cumulative Fatigue Damage
616
Surface Fatigue Strength
617
Stochastic Analysis
618
Road Maps and Important Design Equations for the Stress-Life Method
287
292
295
309
309
313
319
322
336
257
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
258
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
In Chap. 5 we considered the analysis and design of parts subjected to static loading.
The behavior of machine parts is entirely different when they are subjected to timevarying loading. In this chapter we shall examine how parts fail under variable loading
and how to proportion them to successfully resist such conditions.
61
Introduction to Fatigue in Metals
In most testing of those properties of materials that relate to the stress-strain diagram,
the load is applied gradually, to give sufcient time for the strain to fully develop.
Furthermore, the specimen is tested to destruction, and so the stresses are applied only
once. Testing of this kind is applicable, to what are known as static conditions; such
conditions closely approximate the actual conditions to which many structural and
machine members are subjected.
The condition frequently arises, however, in which the stresses vary with time or
they uctuate between different levels. For example, a particular ber on the surface of
a rotating shaft subjected to the action of bending loads undergoes both tension and compression for each revolution of the shaft. If the shaft is part of an electric motor rotating
at 1725 rev/min, the ber is stressed in tension and compression 1725 times each minute.
If, in addition, the shaft is also axially loaded (as it would be, for example, by a helical
or worm gear), an axial component of stress is superposed upon the bending component.
In this case, some stress is always present in any one ber, but now the level of stress is
uctuating. These and other kinds of loading occurring in machine members produce
stresses that are called variable, repeated, alternating, or uctuating stresses.
Often, machine members are found to have failed under the action of repeated or
uctuating stresses; yet the most careful analysis reveals that the actual maximum
stresses were well below the ultimate strength of the material, and quite frequently even
below the yield strength. The most distinguishing characteristic of these failures is that
the stresses have been repeated a very large number of times. Hence the failure is called
a fatigue failure.
When machine parts fail statically, they usually develop a very large deection,
because the stress has exceeded the yield strength, and the part is replaced before fracture
actually occurs. Thus many static failures give visible warning in advance. But a fatigue
failure gives no warning! It is sudden and total, and hence dangerous. It is relatively simple to design against a static failure, because our knowledge is comprehensive. Fatigue is
a much more complicated phenomenon, only partially understood, and the engineer seeking competence must acquire as much knowledge of the subject as possible.
A fatigue failure has an appearance similar to a brittle fracture, as the fracture surfaces are at and perpendicular to the stress axis with the absence of necking. The fracture features of a fatigue failure, however, are quite different from a static brittle fracture
arising from three stages of development. Stage I is the initiation of one or more microcracks due to cyclic plastic deformation followed by crystallographic propagation
extending from two to ve grains about the origin. Stage I cracks are not normally discernible to the naked eye. Stage II progresses from microcracks to macrocracks forming
parallel plateau-like fracture surfaces separated by longitudinal ridges. The plateaus are
generally smooth and normal to the direction of maximum tensile stress. These surfaces
can be wavy dark and light bands referred to as beach marks or clamshell marks, as seen
in Fig. 61. During cyclic loading, these cracked surfaces open and close, rubbing
together, and the beach mark appearance depends on the changes in the level or frequency of loading and the corrosive nature of the environment. Stage III occurs during
the nal stress cycle when the remaining material cannot support the loads, resulting in
261
262
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Fatigue Failure Resulting from Variable Loading
259
Figure 61
Fatigue failure of a bolt due
to repeated unidirectional
bending. The failure started
at the thread root at A,
propagated across most of
the cross section shown by
the beach marks at B, before
final fast fracture at C. (From
ASM Handbook, Vol. 12:
Fractography, ASM International, Materials Park, OH
44073-0002, g 50, p. 120.
Reprinted by permission of
ASM International ®,
www.asminternational.org.)
a sudden, fast fracture. A stage III fracture can be brittle, ductile, or a combination of
both. Quite often the beach marks, if they exist, and possible patterns in the stage III fracture called chevron lines, point toward the origins of the initial cracks.
There is a good deal to be learned from the fracture patterns of a fatigue failure.1
Figure 62 shows representations of failure surfaces of various part geometries under
differing load conditions and levels of stress concentration. Note that, in the case of
rotational bending, even the direction of rotation inuences the failure pattern.
Fatigue failure is due to crack formation and propagation. A fatigue crack will typically initiate at a discontinuity in the material where the cyclic stress is a maximum.
Discontinuities can arise because of:
Design of rapid changes in cross section, keyways, holes, etc. where stress concentrations occur as discussed in Secs. 313 and 52.
Elements that roll and/or slide against each other (bearings, gears, cams, etc.) under
high contact pressure, developing concentrated subsurface contact stresses (Sec. 319)
that can cause surface pitting or spalling after many cycles of the load.
Carelessness in locations of stamp marks, tool marks, scratches, and burrs; poor joint
design; improper assembly; and other fabrication faults.
Composition of the material itself as processed by rolling, forging, casting, extrusion,
drawing, heat treatment, etc. Microscopic and submicroscopic surface and subsurface
discontinuities arise, such as inclusions of foreign material, alloy segregation, voids,
hard precipitated particles, and crystal discontinuities.
Various conditions that can accelerate crack initiation include residual tensile stresses,
elevated temperatures, temperature cycling, a corrosive environment, and high-frequency
cycling.
The rate and direction of fatigue crack propagation is primarily controlled by localized stresses and by the structure of the material at the crack. However, as with crack
formation, other factors may exert a signicant inuence, such as environment, temperature, and frequency. As stated earlier, cracks will grow along planes normal to the
1
See the ASM Handbook, Fractography, ASM International, Metals Park, Ohio, vol. 12, 9th ed., 1987.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
260
II. Failure Prevention
Mechanical Engineering Design
Figure 62
Schematics of fatigue fracture
surfaces produced in smooth
and notched components with
round and rectangular cross
sections under various loading
conditions and nominal stress
levels. (From ASM Handbook,
Vol. 11: Failure Analysis and
Prevention, ASM International,
Materials Park, OH
44073-0002, g 18, p. 111.
Reprinted by permission
of ASM International ®,
www.asminternational.org.)
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
263
264
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Fatigue Failure Resulting from Variable Loading
261
maximum tensile stresses. The crack growth process can be explained by fracture
mechanics (see Sec. 66).
A major reference source in the study of fatigue failure is the 21-volume
ASM Metals Handbook. Figures 61 to 68, reproduced with permission from ASM
International, are but a minuscule sample of examples of fatigue failures for a great
variety of conditions included in the handbook. Comparing Fig. 63 with Fig. 62, we
see that failure occurred by rotating bending stresses, with the direction of rotation
being clockwise with respect to the view and with a mild stress concentration and low
nominal stress.
Figure 63
Fatigue fracture of an AISI
4320 drive shaft. The fatigue
failure initiated at the end of
the keyway at points B and
progressed to nal rupture at
C. The nal rupture zone is
small, indicating that loads
were low. (From ASM
Handbook, Vol. 11: Failure
Analysis and Prevention, ASM
International, Materials Park,
OH 44073-0002, g 18,
p. 111. Reprinted by
permission of ASM
International ®,
www.asminternational.org.)
Figure 64
Fatigue fracture surface of an
AISI 8640 pin. Sharp corners
of the mismatched grease
holes provided stress
concentrations that initiated
two fatigue cracks indicated
by the arrows. (From ASM
Handbook, Vol. 12:
Fractography, ASM
International, Materials Park,
OH 44073-0002, g 520,
p. 331. Reprinted by
permission of ASM
International ®,
www.asminternational.org.)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
262
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 65
Fatigue fracture surface of a
forged connecting rod of AISI
8640 steel. The fatigue crack
origin is at the left edge, at the
ash line of the forging, but no
unusual roughness of the ash
trim was indicated. The
fatigue crack progressed
halfway around the oil hole
at the left, indicated by the
beach marks, before nal fast
fracture occurred. Note the
pronounced shear lip in the
nal fracture at the right edge.
(From ASM Handbook,
Vol. 12: Fractography, ASM
International, Materials Park,
OH 44073-0002, g 523,
p. 332. Reprinted by
permission of ASM
International ®,
www.asminternational.org.)
Figure 66
Fatigue fracture surface of a 200-mm (8-in) diameter piston rod of an alloy
steel steam hammer used for forging. This is an example of a fatigue fracture
caused by pure tension where surface stress concentrations are absent and
a crack may initiate anywhere in the cross section. In this instance, the initial
crack formed at a forging ake slightly below center, grew outward
symmetrically, and ultimately produced a brittle fracture without warning.
(From ASM Handbook, Vol. 12: Fractography, ASM International, Materials
Park, OH 44073-0002, g 570, p. 342. Reprinted by permission of ASM
International ®, www.asminternational.org.)
© The McGrawHill
Companies, 2008
265
266
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
263
Medium-carbon steel
(ASTM A186)
30 dia
Web
Fracture
Fracture Tread
Flange
(1 of 2)
(a) Coke-oven-car wheel
Figure 67
Fatigue failure of an ASTM A186 steel double-ange trailer wheel caused by stamp marks. (a) Coke-oven car wheel showing position of
stamp marks and fractures in the rib and web. (b) Stamp mark showing heavy impression and fracture extending along the base of the lower
row of numbers. (c) Notches, indicated by arrows, created from the heavily indented stamp marks from which cracks initiated along the top
at the fracture surface. (From ASM Handbook, Vol. 11: Failure Analysis and Prevention, ASM International, Materials Park, OH 440730002, g 51, p. 130. Reprinted by permission of ASM International ®, www.asminternational.org.)
Figure 68
Aluminum alloy 7075-T73
landing-gear torque-arm
assembly redesign to eliminate
fatigue fracture at a lubrication
hole. (a) Arm conguration,
original and improved design
(dimensions given in inches).
(b) Fracture surface where
arrows indicate multiple crack
origins. (From ASM
Handbook, Vol. 11: Failure
Analysis and Prevention, ASM
International, Materials Park,
OH 44073-0002, g 23,
p. 114. Reprinted
by permission of ASM
International ®,
www.asminternational.org.)
4.94
Aluminum alloy 7075-T73
Rockwell B 85.5
25.5
10.200
Lug
(1 of 2)
Fracture
A
Primary-fracture
surface
Lubrication hole
1.750-in.-dia
bushing,
0.090-in. wall
Lubrication hole
1 in
3.62 dia
Secondary
fracture
Improved design
Original design
Detail A
(a)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
264
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
62
Approach to Fatigue Failure in Analysis and Design
As noted in the previous section, there are a great many factors to be considered, even
for very simple load cases. The methods of fatigue failure analysis represent a combination of engineering and science. Often science fails to provide the complete answers
that are needed. But the airplane must still be made to ysafely. And the automobile
must be manufactured with a reliability that will ensure a long and troublefree life and
at the same time produce prots for the stockholders of the industry. Thus, while science has not yet completely explained the complete mechanism of fatigue, the engineer
must still design things that will not fail. In a sense this is a classic example of the true
meaning of engineering as contrasted with science. Engineers use science to solve their
problems if the science is available. But available or not, the problem must be solved,
and whatever form the solution takes under these conditions is called engineering.
In this chapter, we will take a structured approach in the design against fatigue
failure. As with static failure, we will attempt to relate to test results performed on simply loaded specimens. However, because of the complex nature of fatigue, there is
much more to account for. From this point, we will proceed methodically, and in stages.
In an attempt to provide some insight as to what follows in this chapter, a brief description of the remaining sections will be given here.
Fatigue-Life Methods (Secs. 63 to 66)
Three major approaches used in design and analysis to predict when, if ever, a cyclically
loaded machine component will fail in fatigue over a period of time are presented. The
premises of each approach are quite different but each adds to our understanding of the
mechanisms associated with fatigue. The application, advantages, and disadvantages of
each method are indicated. Beyond Sec. 66, only one of the methods, the stress-life
method, will be pursued for further design applications.
Fatigue Strength and the Endurance Limit (Secs. 67 and 68)
The strength-life (S-N ) diagram provides the fatigue strength S f versus cycle life N of a
material. The results are generated from tests using a simple loading of standard laboratorycontrolled specimens. The loading often is that of sinusoidally reversing pure bending.
The laboratory-controlled specimens are polished without geometric stress concentration at the region of minimum area.
For steel and iron, the S-N diagram becomes horizontal at some point. The strength
at this point is called the endurance limit Se and occurs somewhere between 106 and 107
cycles. The prime mark on Se refers to the endurance limit of the controlled laboratory
specimen. For nonferrous materials that do not exhibit an endurance limit, a fatigue
strength at a specic number of cycles, S f , may be given, where again, the prime denotes
the fatigue strength of the laboratory-controlled specimen.
The strength data are based on many controlled conditions that will not be the same
as that for an actual machine part. What follows are practices used to account for the
differences between the loading and physical conditions of the specimen and the actual
machine part.
Endurance Limit Modifying Factors (Sec. 69)
Modifying factors are dened and used to account for differences between the specimen and the actual machine part with regard to surface conditions, size, loading, temperature, reliability, and miscellaneous factors. Loading is still considered to be simple
and reversing.
267
268
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Fatigue Failure Resulting from Variable Loading
265
Stress Concentration and Notch Sensitivity (Sec. 610)
The actual part may have a geometric stress concentration by which the fatigue behavior depends on the static stress concentration factor and the component materials sensitivity to fatigue damage.
Fluctuating Stresses (Secs. 611 to 613)
These sections account for simple stress states from uctuating load conditions that are
not purely sinusoidally reversing axial, bending, or torsional stresses.
Combinations of Loading Modes (Sec. 614)
Here a procedure based on the distortion-energy theory is presented for analyzing combined uctuating stress states, such as combined bending and torsion. Here it is
assumed that the levels of the uctuating stresses are in phase and not time varying.
Varying, Fluctuating Stresses; Cumulative
Fatigue Damage (Sec. 615)
The uctuating stress levels on a machine part may be time varying. Methods are provided to assess the fatigue damage on a cumulative basis.
Remaining Sections
The remaining three sections of the chapter pertain to the special topics of surface
fatigue strength, stochastic analysis, and roadmaps with important equations.
63
Fatigue-Life Methods
The three major fatigue life methods used in design and analysis are the stress-life
method, the strain-life method, and the linear-elastic fracture mechanics method. These
methods attempt to predict the life in number of cycles to failure, N, for a specic level
of loading. Life of 1 N 103 cycles is generally classied as low-cycle fatigue,
whereas high-cycle fatigue is considered to be N > 103 cycles. The stress-life method,
based on stress levels only, is the least accurate approach, especially for low-cycle
applications. However, it is the most traditional method, since it is the easiest to implement for a wide range of design applications, has ample supporting data, and represents
high-cycle applications adequately.
The strain-life method involves more detailed analysis of the plastic deformation at
localized regions where the stresses and strains are considered for life estimates. This
method is especially good for low-cycle fatigue applications. In applying this method,
several idealizations must be compounded, and so some uncertainties will exist in the
results. For this reason, it will be discussed only because of its value in adding to the
understanding of the nature of fatigue.
The fracture mechanics method assumes a crack is already present and detected. It
is then employed to predict crack growth with respect to stress intensity. It is most practical when applied to large structures in conjunction with computer codes and a periodic inspection program.
64
The Stress-Life Method
To determine the strength of materials under the action of fatigue loads, specimens are
subjected to repeated or varying forces of specied magnitudes while the cycles or
stress reversals are counted to destruction. The most widely used fatigue-testing device
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
266
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
is the R. R. Moore high-speed rotating-beam machine. This machine subjects the specimen
to pure bending (no transverse shear) by means of weights. The specimen, shown in
Fig. 69, is very carefully machined and polished, with a nal polishing in an axial
direction to avoid circumferential scratches. Other fatigue-testing machines are available for applying uctuating or reversed axial stresses, torsional stresses, or combined
stresses to the test specimens.
To establish the fatigue strength of a material, quite a number of tests are necessary
because of the statistical nature of fatigue. For the rotating-beam test, a constant bending load is applied, and the number of revolutions (stress reversals) of the beam required
for failure is recorded. The rst test is made at a stress that is somewhat under the ultimate strength of the material. The second test is made at a stress that is less than that
used in the rst. This process is continued, and the results are plotted as an S-N diagram
(Fig. 610). This chart may be plotted on semilog paper or on log-log paper. In the case
of ferrous metals and alloys, the graph becomes horizontal after the material has been
stressed for a certain number of cycles. Plotting on log paper emphasizes the bend in
the curve, which might not be apparent if the results were plotted by using Cartesian
coordinates.
7
3 16 in
0.30 in
9 7 in R.
8
Figure 69
Test-specimen geometry for the R. R. Moore rotatingbeam machine. The bending moment is uniform over the
curved at the highest-stressed portion, a valid test of
material, whereas a fracture elsewhere (not at the higheststress level) is grounds for suspicion of material aw.
Figure 610
High cycle
Finite life
Infinite
life
Sut
100
Fatigue strength Sf , kpsi
An S-N diagram plotted from
the results of completely
reversed axial fatigue tests.
Material: UNS G41300
steel, normalized;
Sut = 116 kpsi; maximum
Sut = 125 kpsi. (Data from
NACA Tech. Note 3866,
December 1966.)
Low cycle
50
100
Se
101
102
103
10 4
10 5
Number of stress cycles, N
106
107
108
269
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Figure 611
S-N bands for representative
aluminum alloys, excluding
wrought alloys with
Sut < 38 kpsi. (From R. C.
Juvinall, Engineering
Considerations of Stress,
Strain and Strength. Copyright
© 1967 by The McGraw-Hill
Companies, Inc. Reprinted by
permission.)
267
80
70
60
Peak alternating bending stress S, kpsi (log)
270
50
40
35
30
25
Wrought
20
18
16
14
12
Permanent mold cast
10
Sand cast
8
7
6
5
103
104
105
106
Life N, cycles (log)
107
108
109
The ordinate of the S-N diagram is called the fatigue strength S f ; a statement of
this strength value must always be accompanied by a statement of the number of cycles
N to which it corresponds.
Soon we shall learn that S-N diagrams can be determined either for a test specimen
or for an actual mechanical element. Even when the material of the test specimen and
that of the mechanical element are identical, there will be signicant differences
between the diagrams for the two.
In the case of the steels, a knee occurs in the graph, and beyond this knee failure
will not occur, no matter how great the number of cycles. The strength corresponding
to the knee is called the endurance limit Se , or the fatigue limit. The graph of Fig. 610
never does become horizontal for nonferrous metals and alloys, and hence these materials do not have an endurance limit. Figure 611 shows scatter bands indicating the S-N
curves for most common aluminum alloys excluding wrought alloys having a tensile
strength below 38 kpsi. Since aluminum does not have an endurance limit, normally the
fatigue strength S f is reported at a specic number of cycles, normally N = 5(108 )
cycles of reversed stress (see Table A24).
We note that a stress cycle ( N = 1) constitutes a single application and removal of
a load and then another application and removal of the load in the opposite direction.
Thus N = 1 means the load is applied once and then removed, which is the case with
2
the simple tension test.
The body of knowledge available on fatigue failure from N = 1 to N = 1000
cycles is generally classied as low-cycle fatigue, as indicated in Fig. 610. High-cycle
fatigue, then, is concerned with failure corresponding to stress cycles greater than 103
cycles.
We also distinguish a nite-life region and an innite-life region in Fig. 610. The
boundary between these regions cannot be clearly dened except for a specic material;
but it lies somewhere between 106 and 107 cycles for steels, as shown in Fig. 610.
As noted previously, it is always good engineering practice to conduct a testing
program on the materials to be employed in design and manufacture. This, in fact, is a
requirement, not an option, in guarding against the possibility of a fatigue failure.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
268
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Because of this necessity for testing, it would really be unnecessary for us to proceed
any further in the study of fatigue failure except for one important reason: the desire to
know why fatigue failures occur so that the most effective method or methods can be
used to improve fatigue strength. Thus our primary purpose in studying fatigue is to
understand why failures occur so that we can guard against them in an optimum manner. For this reason, the analytical design approaches presented in this book, or in any
other book, for that matter, do not yield absolutely precise results. The results should be
taken as a guide, as something that indicates what is important and what is not important in designing against fatigue failure.
As stated earlier, the stress-life method is the least accurate approach especially
for low-cycle applications. However, it is the most traditional method, with much
published data available. It is the easiest to implement for a wide range of design
applications and represents high-cycle applications adequately. For these reasons the
stress-life method will be emphasized in subsequent sections of this chapter.
However, care should be exercised when applying the method for low-cycle applications,
as the method does not account for the true stress-strain behavior when localized
yielding occurs.
65
The Strain-Life Method
The best approach yet advanced to explain the nature of fatigue failure is called by some
the strain-life method. The approach can be used to estimate fatigue strengths, but when
it is so used it is necessary to compound several idealizations, and so some uncertainties will exist in the results. For this reason, the method is presented here only because
of its value in explaining the nature of fatigue.
A fatigue failure almost always begins at a local discontinuity such as a notch,
crack, or other area of stress concentration. When the stress at the discontinuity exceeds
the elastic limit, plastic strain occurs. If a fatigue fracture is to occur, there must exist
cyclic plastic strains. Thus we shall need to investigate the behavior of materials subject to cyclic deformation.
In 1910, Bairstow veried by experiment Bauschingers theory that the elastic limits of iron and steel can be changed, either up or down, by the cyclic variations of stress.2
In general, the elastic limits of annealed steels are likely to increase when subjected to
cycles of stress reversals, while cold-drawn steels exhibit a decreasing elastic limit.
R. W. Landgraf has investigated the low-cycle fatigue behavior of a large number
of very high-strength steels, and during his research he made many cyclic stress-strain
plots.3 Figure 612 has been constructed to show the general appearance of these plots
for the rst few cycles of controlled cyclic strain. In this case the strength decreases
with stress repetitions, as evidenced by the fact that the reversals occur at ever-smaller
stress levels. As previously noted, other materials may be strengthened, instead, by
cyclic stress reversals.
The SAE Fatigue Design and Evaluation Steering Committee released a report in
1975 in which the life in reversals to failure is related to the strain amplitude ε/2.4
2
L. Bairstow, The Elastic Limits of Iron and Steel under Cyclic Variations of Stress, Philosophical
Transactions, Series A, vol. 210, Royal Society of London, 1910, pp. 3555.
3
R. W. Landgraf, Cyclic Deformation and Fatigue Behavior of Hardened Steels, Report no. 320, Department
of Theoretical and Applied Mechanics, University of Illinois, Urbana, 1968, pp. 8490.
4
Technical Report on Fatigue Properties, SAE J1099, 1975.
271
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Figure 612
True stresstrue strain hysteresis
loops showing the rst ve
stress reversals of a cyclicsoftening material. The graph
is slightly exaggerated for
clarity. Note that the slope of
the line AB is the modulus of
elasticity E. The stress range is
σ , ε p is the plastic-strain
range, and εe is the
elastic strain range. The
total-strain range is
ε = ε p + εe .
269
1st reversal
A
3d
5th
4th
2d
B
p
e
Figure 613
A log-log plot showing how
the fatigue life is related to
the true-strain amplitude for
hot-rolled SAE 1020 steel.
(Reprinted with permission
from SAE J1099_200208
© 2002 SAE International.)
10 0
'
F
101
Strain amplitude, /2
272
c
1.0
102
'
F
E
Total strain
Plastic strain
b
1.0
103
Elastic strain
10 4
100
101
10 2
10 3
10 4
10 5
106
Reversals to failure, 2 N
The report contains a plot of this relationship for SAE 1020 hot-rolled steel; the graph
has been reproduced as Fig. 613. To explain the graph, we rst dene the following
terms:
Fatigue ductility coefcient εF is the true strain corresponding to fracture in one reversal (point A in Fig. 612). The plastic-strain line begins at this point in Fig. 613.
Fatigue strength coefcient σ F is the true stress corresponding to fracture in one
reversal (point A in Fig. 612). Note in Fig. 613 that the elastic-strain line begins at
σF / E .
Fatigue ductility exponent c is the slope of the plastic-strain line in Fig. 613 and is
the power to which the life 2N must be raised to be proportional to the true plasticstrain amplitude. If the number of stress reversals is 2N, then N is the number of
cycles.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
270
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
273
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Fatigue strength exponent b is the slope of the elastic-strain line, and is the power to
which the life 2N must be raised to be proportional to the true-stress amplitude.
Now, from Fig. 612, we see that the total strain is the sum of the elastic and plastic
components. Therefore the total strain amplitude is half the total strain range
εp
ε
εe
=
+
2
2
2
(a)
The equation of the plastic-strain line in Fig. 613 is
εp
= εF (2 N )c
2
(61)
The equation of the elastic strain line is
σ
εe
= F (2 N )b
2
E
(62)
Therefore, from Eq. (a), we have for the total-strain amplitude
ε
σ
= F (2 N )b + εF (2 N )c
2
E
(63)
which is the Manson-Cofn relationship between fatigue life and total strain.5 Some
values of the coefcients and exponents are listed in Table A23. Many more are
included in the SAE J1099 report.6
Though Eq. (63) is a perfectly legitimate equation for obtaining the fatigue life of
a part when the strain and other cyclic characteristics are given, it appears to be of little use to the designer. The question of how to determine the total strain at the bottom
of a notch or discontinuity has not been answered. There are no tables or charts of strain
concentration factors in the literature. It is possible that strain concentration factors will
become available in research literature very soon because of the increase in the use of
nite-element analysis. Moreover, nite element analysis can of itself approximate the
strains that will occur at all points in the subject structure.7
66
The Linear-Elastic Fracture Mechanics Method
The rst phase of fatigue cracking is designated as stage I fatigue. Crystal slip that
extends through several contiguous grains, inclusions, and surface imperfections is presumed to play a role. Since most of this is invisible to the observer, we just say that stage
I involves several grains. The second phase, that of crack extension, is called stage II
fatigue. The advance of the crack (that is, new crack area is created) does produce evidence that can be observed on micrographs from an electron microscope. The growth of
5
J. F. Tavernelli and L. F. Cofn, Jr., Experimental Support for Generalized Equation Predicting Low Cycle
Fatigue, and S. S. Manson, discussion, Trans. ASME, J. Basic Eng., vol. 84, no. 4, pp. 533537.
6
See also, Landgraf, Ibid.
7
For further discussion of the strain-life method see N. E. Dowling, Mechanical Behavior of Materials,
2nd ed., Prentice-Hall, Englewood Cliffs, N.J., 1999, Chap. 14.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
271
the crack is orderly. Final fracture occurs during stage III fatigue, although fatigue is not
involved. When the crack is sufciently long that K I = K Ic for the stress amplitude
involved, then K Ic is the critical stress intensity for the undamaged metal, and there is
sudden, catastrophic failure of the remaining cross section in tensile overload (see
Sec. 512). Stage III fatigue is associated with rapid acceleration of crack growth then
fracture.
Crack Growth
Fatigue cracks nucleate and grow when stresses vary and there is some tension in
each stress cycle. Consider the stress to be uctuating between the limits of σmin and
σmax , where the stress range is dened as σ = σmax σmin . From Eq. (537) the
stress intensity is given by K I = βσ π a . Thus, for σ, the stress intensity range per
cycle is
K I = β(σmax σmin ) π a = β σ π a
(64)
To develop fatigue strength data, a number of specimens of the same material are tested
at various levels of σ. Cracks nucleate at or very near a free surface or large discontinuity. Assuming an initial crack length of ai , crack growth as a function of the number of stress cycles N will depend on σ, that is, K I . For K I below some threshold
value ( K I )th a crack will not grow. Figure 614 represents the crack length a as a
function of N for three stress levels ( σ )3 > ( σ )2 > ( σ )1 , where ( K I )3 >
( K I )2 > ( K I )1 . Notice the effect of the higher stress range in Fig. 614 in the production of longer cracks at a particular cycle count.
When the rate of crack growth per cycle, d a /d N in Fig. 614, is plotted as shown
in Fig. 615, the data from all three stress range levels superpose to give a sigmoidal
curve. The three stages of crack development are observable, and the stage II data are
linear on log-log coordinates, within the domain of linear elastic fracture mechanics
(LEFM) validity. A group of similar curves can be generated by changing the stress
ratio R = σmin /σmax of the experiment.
Here we present a simplied procedure for estimating the remaining life of a cyclically stressed part after discovery of a crack. This requires the assumption that plane strain
Figure 614
The increase in crack length a
from an initial length of ai as
a function of cycle count for
three stress ranges, ( σ ) 3 >
( σ) 2 > ( σ) 1.
(KI )3
Crack length a
274
(KI )2
(KI )1
da
a
dN
ai
Log N
Stress cycles N
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
272
II. Failure Prevention
275
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 615
Log da
dN
When da/dN is measured in
Fig. 614 and plotted on
loglog coordinates, the data
for different stress ranges
superpose, giving rise to a
sigmoid curve as shown.
( K I ) th is the threshold value
of K I , below which a crack
does not grow. From threshold
to rupture an aluminum alloy
will spend 85--90 percent of
life in region I, 5--8 percent in
region II, and 1--2 percent in
region III.
Region I
Region II
Crack
initiation
Crack
propagation
Region III
Crack
unstable
Increasing
stress ratio
R
Kc
(K )th
Log K
Table 61
Conservative Values of
Factor C and Exponent
m in Eq. (65) for
Various Forms of Steel
.
(R = 0)
C,
Material
m/cycle
m
MPa m
C,
6.89(1012 )
Ferritic-pearlitic steels
10
Martensitic steels
1.36(10
Austenitic stainless steels
in/cycle
m
kpsi in
3.60(1010 )
9
5.61(1012 )
)
6.60(10
)
3.00(1010 )
m
3.00
2.25
3.25
From J.M. Barsom and S.T. Rolfe, Fatigue and Fracture Control in Structures, 2nd ed., Prentice Hall, Upper Saddle River, NJ, 1987,
pp. 288291, Copyright ASTM International. Reprinted with permission.
conditions prevail.8 Assuming a crack is discovered early in stage II, the crack growth in
region II of Fig. 615 can be approximated by the Paris equation, which is of the form
da
= C ( K I )m
dN
(65)
where C and m are empirical material constants and K I is given by Eq. (64).
Representative, but conservative, values of C and m for various classes of steels are
listed in Table 61. Substituting Eq. (64) and integrating gives
Nf
0
d N = Nf =
1
C
af
ai
da
(β σ π a )m
(66)
Here ai is the initial crack length, a f is the nal crack length corresponding to failure,
and N f is the estimated number of cycles to produce a failure after the initial crack is
formed. Note that β may vary in the integration variable (e.g., see Figs. 525 to 530).
8
Recommended references are: Dowling, op. cit.; J. A. Collins, Failure of Materials in Mechanical Design,
John Wiley & Sons, New York, 1981; H. O. Fuchs and R. I. Stephens, Metal Fatigue in Engineering, John
Wiley & Sons, New York, 1980; and Harold S. Reemsnyder, Constant Amplitude Fatigue Life Assessment
Models, SAE Trans. 820688, vol. 91, Nov. 1983.
276
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
273
If this should happen, then Reemsnyder9 suggests the use of numerical integration
employing the algorithm
δ a j = C ( K I )m (δ N ) j
j
a j +1 = a j + δ a j
(67)
N j +1 = N j + δ N j
Nf =
δ Nj
Here δ a j and δ N j are increments of the crack length and the number of cycles. The procedure is to select a value of δ N j , using ai determine β and compute K I , determine
δ a j , and then nd the next value of a. Repeat the procedure until a = a f .
The following example is highly simplied with β constant in order to give some
understanding of the procedure. Normally, one uses fatigue crack growth computer programs such as NASA/FLAGRO 2.0 with more comprehensive theoretical models to
solve these problems.
9
Op. cit.
EXAMPLE 61
Solution
The bar shown in Fig. 616 is subjected to a repeated moment 0 M 1200 lbf in.
·
The bar is AISI 4430 steel with Sut = 185 kpsi, Sy = 170 kpsi, and K Ic = 73 kpsi in.
Material tests on various specimens of this material with identical heat treatment
indicate worst-case constants of C = 3.8(1011 ) (in/cycle) (kpsi in)m and m = 3.0. As
shown, a nick of size 0.004 in has been discovered on the bottom of the bar. Estimate
the number of cycles of life remaining.
The stress range
σ is always computed by using the nominal (uncracked) area. Thus
I
bh 2
0.25(0.5)2
=
=
= 0.010 42 in3
c
6
6
Therefore, before the crack initiates, the stress range is
σ=
M
1200
=
= 115.2(103 ) psi = 115.2 kpsi
I /c
0.010 42
which is below the yield strength. As the crack grows, it will eventually become long
enough such that the bar will completely yield or undergo a brittle fracture. For the ratio
of Sy / Sut it is highly unlikely that the bar will reach complete yield. For brittle fracture,
designate the crack length as a f . If β = 1, then from Eq. (537) with K I = K Ic , we
approximate a f as
af =
1
π
K Ic
βσmax
Figure 616
2
1
4
M
M
Nick
.1
=
π
73
115.2
in
1
2
in
2
= 0.1278 in
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
274
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
From Fig. 527, we compute the ratio a f / h as
af
0.1278
=
= 0.256
h
0.5
Thus a f / h varies from near zero to approximately 0.256. From Fig. 527, for this range
β is nearly constant at approximately 1.07. We will assume it to be so, and re-evaluate
a f as
af =
1
π
73
1.07(115.2)
2
= 0.112 in
Thus, from Eq. (66), the estimated remaining life is
Nf =
1
C
=
67
af
ai
1
da
=
m
3.8(1011 )
(β σ π a )
5.047(103 )
a
0.112
0.004
da
[1.07(115.2) π a ]3
0.112
0.004
= 64.7 (103 ) cycles
The Endurance Limit
The determination of endurance limits by fatigue testing is now routine, though a lengthy
procedure. Generally, stress testing is preferred to strain testing for endurance limits.
For preliminary and prototype design and for some failure analysis as well, a quick
method of estimating endurance limits is needed. There are great quantities of data in
the literature on the results of rotating-beam tests and simple tension tests of specimens
taken from the same bar or ingot. By plotting these as in Fig. 617, it is possible to see
whether there is any correlation between the two sets of results. The graph appears to
suggest that the endurance limit ranges from about 40 to 60 percent of the tensile
strength for steels up to about 210 kpsi (1450 MPa). Beginning at about Sut = 210 kpsi
(1450 MPa), the scatter appears to increase, but the trend seems to level off, as sug
gested by the dashed horizontal line at Se = 105 kpsi.
We wish now to present a method for estimating endurance limits. Note that estimates obtained from quantities of data obtained from many sources probably have a
large spread and might deviate signicantly from the results of actual laboratory tests of
the mechanical properties of specimens obtained through strict purchase-order specications. Since the area of uncertainty is greater, compensation must be made by employing larger design factors than would be used for static design.
For steels, simplifying our observation of Fig. 617, we will estimate the endurance
limit as
Sut 200 kpsi (1400 MPa)
0.5 Sut
Se = 100 kpsi
Sut > 200 kpsi
(68)
700 MPa
Sut > 1400 MPa
where Sut is the minimum tensile strength. The prime mark on Se in this equation refers
to the rotating-beam specimen itself. We wish to reserve the unprimed symbol Se for the
endurance limit of any particular machine element subjected to any kind of loading.
Soon we shall learn that the two strengths may be quite different.
277
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
275
Fatigue Failure Resulting from Variable Loading
140
0
S 'e =
u
S
Carbon steels
Alloy steels
Wrought irons
120
Endurance limit S 'e , kpsi
278
0.5
.6
0.4
105 kpsi
100
80
60
40
20
0
0
20
40
60
80
100
120
140
160
180
200
220
240
260
280
300
Tensile strength Su t , kpsi
Figure 617
Graph of endurance limits versus tensile strengths from actual test results for a large number of wrought
irons and steels. Ratios of Se/ Sut of 0.60, 0.50, and 0.40 are shown by the solid and dashed lines.
Note also the horizontal dashed line for Se = 105 kpsi. Points shown having a tensile strength greater
than 210 kpsi have a mean endurance limit of Se = 105 kpsi and a standard deviation of 13.5 kpsi.
(Collated from data compiled by H. J. Grover, S. A. Gordon, and L. R. Jackson in Fatigue of Metals
and Structures, Bureau of Naval Weapons Document NAVWEPS 00-25-534, 1960; and from Fatigue
Design Handbook, SAE, 1968, p. 42.)
Steels treated to give different microstructures have different Se / Sut ratios. It
appears that the more ductile microstructures have a higher ratio. Martensite has a very
brittle nature and is highly susceptible to fatigue-induced cracking; thus the ratio is low.
When designs include detailed heat-treating specifications to obtain specific
microstructures, it is possible to use an estimate of the endurance limit based on test
data for the particular microstructure; such estimates are much more reliable and indeed
should be used.
The endurance limits for various classes of cast irons, polished or machined, are
given in Table A24. Aluminum alloys do not have an endurance limit. The fatigue
strengths of some aluminum alloys at 5(108) cycles of reversed stress are given in
Table A24.
68
Fatigue Strength
As shown in Fig. 610, a region of low-cycle fatigue extends from N = 1 to about
103 cycles. In this region the fatigue strength S f is only slightly smaller than the tensile strength Sut . An analytical approach has been given by Mischke10 for both
10
J. E. Shigley, C. R. Mischke, and T. H. Brown, Jr., Standard Handbook of Machine Design, 3rd ed.,
McGraw-Hill, New York, 2004, pp. 29.2529.27.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
276
II. Failure Prevention
279
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
high-cycle and low-cycle regions, requiring the parameters of the Manson-Coffin
equation plus the strain-strengthening exponent m. Engineers often have to work
with less information.
Figure 610 indicates that the high-cycle fatigue domain extends from 103 cycles
for steels to the endurance limit life Ne , which is about 106 to 107 cycles. The purpose
of this section is to develop methods of approximation of the S-N diagram in the highcycle region, when information may be as sparse as the results of a simple tension test.
Experience has shown high-cycle fatigue data are rectied by a logarithmic transform
to both stress and cycles-to-failure. Equation (62) can be used to determine the fatigue
strength at 103 cycles. Dening the specimen fatigue strength at a specic number of
cycles as ( S f ) N = E εe /2, write Eq. (62) as
( S f ) N = σ F (2 N )b
(69)
At 103 cycles,
( S f )103 = σ F (2.103 )b = f Sut
where f is the fraction of Sut represented by ( S f )103 cycles . Solving for f gives
f=
σF
(2 · 103 )b
Sut
(610)
Now, from Eq. (211), σ F = σ0 εm , with ε = εF . If this true-stresstrue-strain equation
is not known, the SAE approximation11 for steels with HB 500 may be used:
σ F = Sut + 50 kpsi
or
σ F = Sut + 345 MPa
(611)
To nd b, substitute the endurance strength and corresponding cycles, Se and Ne ,
respectively into Eq. (69) and solving for b
b=
log σ F / Se
log (2 N e )
(612)
Thus, the equation S f = σ F (2 N )b is known. For example, if Sut = 105 kpsi and
Se = 52.5 kpsi at failure,
Eq. (611)
σ F = 105 + 50 = 155 kpsi
Eq. (612)
b=
Eq. (610)
f=
log(155/52.5)
= 0.0746
log 2 · 106
155
2 · 103
105
0.0746
= 0.837
and for Eq. (69), with S f = ( S f ) N ,
S f = 155(2 N )0.0746 = 147 N 0.0746
11
Fatigue Design Handbook, vol. 4, Society of Automotive Engineers, New York, 1958, p. 27.
(a)
280
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Figure 618
Fatigue strength fraction, f, of
Sut at 103 cycles for
Se = Se = 0.5 Sut .
f
277
0.9
0.88
0.86
0.84
0.82
0.8
0.78
0.76
70
80
90
100 110 120 130 140 150 160 170 180 190 200
Su t , kpsi
The process given for nding f can be repeated for various ultimate strengths.
Figure 618 is a plot of f for 70 Sut 200 kpsi. To be conservative, for Sut < 70 kpsi,
let f 0.9.
For an actual mechanical component, Se is reduced to S e (see Sec. 69) which is
less than 0.5 Sut . However, unless actual data is available, we recommend using the
value of f found from Fig. 618. Equation (a), for the actual mechanical component, can
be written in the form
Sf = a N b
(613)
where N is cycles to failure and the constants a and b are dened by the points
103 , S f 103 and 106 , Se with S f 103 = f Sut . Substituting these two points in Eq.
(613) gives
a=
( f Sut )2
Se
1
b = log
3
(614)
f Sut
Se
(615)
If a completely reversed stress σa is given, setting S f = σa in Eq. (613), the number
of cycles-to-failure can be expressed as
N=
σa
a
1/b
(616)
Low-cycle fatigue is often dened (see Fig. 610) as failure that occurs in a range
of 1 N 103 cycles. On a loglog plot such as Fig. 610 the failure locus in this range
is nearly linear below 103 cycles. A straight line between 103 , f Sut and 1, Sut (transformed) is conservative, and it is given by
S f Sut N (log f )/3
1 N 103
(617)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
278
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
EXAMPLE 62
Solution
Given a 1050 HR steel, estimate
(a) the rotating-beam endurance limit at 106 cycles.
(b) the endurance strength of a polished rotating-beam specimen corresponding to 104
cycles to failure
(c) the expected life of a polished rotating-beam specimen under a completely reversed
stress of 55 kpsi.
(a) From Table A20, Sut = 90 kpsi. From Eq. (68),
Se = 0.5(90) = 45 kpsi
Answer
.
(b) From Fig. 618, for Sut = 90 kpsi, f = 0.86. From Eq. (614),
a=
[0.86(90)2 ]
= 133.1 kpsi
45
From Eq. (615),
1
0.86(90)
b = log
= 0.0785
3
45
Thus, Eq. (613) is
S f = 133.1 N 0.0785
Answer
For 104 cycles to failure, S f = 133.1(104 ) 0.0785 = 64.6 kpsi
(c) From Eq. (616), with σa = 55 kpsi,
Answer
N=
55
133.1
1/0.0785
= 77 500 = 7.75(104 )cycles
Keep in mind that these are only estimates. So expressing the answers using three-place
accuracy is a little misleading.
69
Endurance Limit Modifying Factors
We have seen that the rotating-beam specimen used in the laboratory to determine
endurance limits is prepared very carefully and tested under closely controlled conditions. It is unrealistic to expect the endurance limit of a mechanical or structural member to match the values obtained in the laboratory. Some differences include
Material: composition, basis of failure, variability
Manufacturing: method, heat treatment, fretting corrosion, surface condition, stress
concentration
Environment: corrosion, temperature, stress state, relaxation times
Design: size, shape, life, stress state, stress concentration, speed, fretting, galling
281
282
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
279
Fatigue Failure Resulting from Variable Loading
Marin12 identied factors that quantied the effects of surface condition, size, loading,
temperature, and miscellaneous items. The question of whether to adjust the endurance
limit by subtractive corrections or multiplicative corrections was resolved by an extensive statistical analysis of a 4340 (electric furnace, aircraft quality) steel, in which a
correlation coefcient of 0.85 was found for the multiplicative form and 0.40 for the
additive form. A Marin equation is therefore written as
Se = ka kb kc kd ke k f Se
where
(618)
ka = surface condition modication factor
kb = size modication factor
kc = load modication factor
kd = temperature modication factor
ke = reliability factor13
kf = miscellaneous-effects modication factor
Se = rotary-beam test specimen endurance limit
Se = endurance limit at the critical location of a machine part in the geometry and condition of use
When endurance tests of parts are not available, estimations are made by applying
Marin factors to the endurance limit.
Surface Factor ka
The surface of a rotating-beam specimen is highly polished, with a nal polishing in the
axial direction to smooth out any circumferential scratches. The surface modication
factor depends on the quality of the nish of the actual part surface and on the tensile
strength of the part material. To nd quantitative expressions for common nishes of
machine parts (ground, machined, or cold-drawn, hot-rolled, and as-forged), the coordinates of data points were recaptured from a plot of endurance limit versus ultimate
tensile strength of data gathered by Lipson and Noll and reproduced by Horger.14 The
data can be represented by
b
ka = aSut
(619)
where Sut is the minimum tensile strength and a and b are to be found in Table 62.
12
Joseph Marin, Mechanical Behavior of Engineering Materials, Prentice-Hall, Englewood Cliffs, N.J.,
1962, p. 224.
13
Complete stochastic analysis is presented in Sec. 617. Until that point the presentation here is one of a
deterministic nature. However, we must take care of the known scatter in the fatigue data. This means that
we will not carry out a true reliability analysis at this time but will attempt to answer the question: What is
the probability that a known (assumed) stress will exceed the strength of a randomly selected component
made from this material population?
14
C. J. Noll and C. Lipson, Allowable Working Stresses, Society for Experimental Stress Analysis, vol. 3,
no. 2, 1946, p. 29. Reproduced by O. J. Horger (ed.), Metals Engineering Design ASME Handbook,
McGraw-Hill, New York, 1953, p. 102.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
280
II. Failure Prevention
283
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Table 62
Parameters for Marin
Surface Modication
Factor, Eq. (619)
Factor a
Surface
Finish
Sut, kpsi
Sut, MPa
Ground
1.34
1.58
Machined or cold-drawn
2.70
4.51
Hot-rolled
14.4
As-forged
39.9
57.7
Exponent
b
0.085
0.265
0.718
272.
0.995
From C.J. Noll and C. Lipson, Allowable Working Stresses, Society for Experimental Stress Analysis, vol. 3,
no. 2, 1946 p. 29. Reproduced by O.J. Horger (ed.) Metals Engineering Design ASME Handbook, McGraw-Hill,
New York. Copyright © 1953 by The McGraw-Hill Companies, Inc. Reprinted by permission.
EXAMPLE 63
Solution
Answer
A steel has a minimum ultimate strength of 520 MPa and a machined surface.
Estimate ka.
From Table 62, a = 4.51 and b = 0.265. Then, from Eq. (619)
ka = 4.51(520)0.265 = 0.860
Again, it is important to note that this is an approximation as the data is typically
quite scattered. Furthermore, this is not a correction to take lightly. For example, if in
the previous example the steel was forged, the correction factor would be 0.540, a signicant reduction of strength.
Size Factor kb
The size factor has been evaluated using 133 sets of data points.15 The results for bending and torsion may be expressed as
(d /0.3)0.107 = 0.879d 0.107
0.91d 0.157
kb =
(d /7.62)0.107 = 1.24d 0.107
1.51d 0.157
0.11 d 2 in
2 < d 10 in
2.79 d 51 mm
51 < d 254 mm
( 620)
For axial loading there is no size effect, so
kb = 1
(621)
but see kc .
One of the problems that arises in using Eq. (620) is what to do when a round bar
in bending is not rotating, or when a noncircular cross section is used. For example,
what is the size factor for a bar 6 mm thick and 40 mm wide? The approach to be used
15
Charles R. Mischke, Prediction of Stochastic Endurance Strength, Trans. of ASME, Journal of Vibration,
Acoustics, Stress, and Reliability in Design, vol. 109, no. 1, January 1987, Table 3.
284
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
281
here employs an effective dimension de 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.16 It turns out that when these two volumes are equated,
the lengths cancel, and so we need only consider the areas. For a rotating round section,
the 95 percent stress area is the area in a ring having an outside diameter d and an inside
diameter of 0.95d. So, designating the 95 percent stress area A0.95σ , we have
π
A0.95σ = [d 2 (0.95d )2 ] = 0.0766d 2
(622)
4
This equation is also valid for a rotating hollow round. For nonrotating solid or hollow
rounds, the 95 percent stress area is twice the area outside of two parallel chords having a spacing of 0.95d, where d is the diameter. Using an exact computation, this is
A0.95σ = 0.01046d 2
(623)
with de in Eq. (622), setting Eqs. (622) and (623) equal to each other enables us to
solve for the effective diameter. This gives
(624)
de = 0.370d
as the effective size of a round corresponding to a nonrotating solid or hollow round.
A rectangular section of dimensions h × b has A0.95σ = 0.05hb . Using the same
approach as before,
de = 0.808(hb)1/2
(625)
Table 63 provides A0.95σ areas of common structural shapes undergoing nonrotating bending.
16
See R. Kuguel, A Relation between Theoretical Stress Concentration Factor and Fatigue Notch Factor
Deduced from the Concept of Highly Stressed Volume, Proc. ASTM, vol. 61, 1961, pp. 732748.
EXAMPLE 64
Solution
A steel shaft loaded in bending is 32 mm in diameter, abutting a lleted shoulder 38 mm
in diameter. The shaft material has a mean ultimate tensile strength of 690 MPa.
Estimate the Marin size factor kb if the shaft is used in
(a) A rotating mode.
(b) A nonrotating mode.
(a) From Eq. (620)
Answer
kb =
d
7.62
0.107
=
32
7.62
0.107
= 0.858
(b) From Table 63,
de = 0.37d = 0.37(32) = 11.84 mm
From Eq. (620),
Answer
kb =
11.84
7.62
0.107
= 0.954
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
282
II. Failure Prevention
285
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Table 63
A0.95σ Areas of
Common Nonrotating
Structural Shapes
A 0.95σ = 0.01046d 2
d
de = 0.370d
b
2
h
1
A 0 .95 σ = 0.05hb
de = 0.808 hb
1
2
a
1
b
2
A 0 .95 σ =
2
0.10at f
0.05ba
axis 1-1
t f > 0.025a
axis 2-2
tf
1
a
1
x
2
b
A 0 .95 σ =
2
tf
0.05ab
axis 1-1
0.052xa + 0.1t f ( b x )
axis 2-2
1
Loading Factor kc
When fatigue tests are carried out with rotating bending, axial (push-pull), and torsional loading, the endurance limits differ with Sut. This is discussed further in Sec. 617.
Here, we will specify average values of the load factor as
kc =
1
0.85
0.59
bending
axial
torsion17
(626)
Temperature Factor kd
When operating temperatures are below room temperature, brittle fracture is a strong
possibility and should be investigated rst. When the operating temperatures are higher than room temperature, yielding should be investigated rst because the yield
strength drops off so rapidly with temperature; see Fig. 29. Any stress will induce
creep in a material operating at high temperatures; so this factor must be considered too.
17
Use this only for pure torsional fatigue loading. When torsion is combined with other stresses, such
as bending, kc = 1 and the combined loading is managed by using the effective von Mises stress as in
Sec. 55. Note: For pure torsion, the distortion energy predicts that (kc)torsion = 0.577.
286
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Table 64
Effect of Operating
Temperature on the
Tensile Strength of
Steel.* (ST = tensile
strength at operating
temperature;
SRT = tensile strength
at room temperature;
ˆ
0.099 σ 0.110)
Temperature, °C
20
ST/SRT
1.000
Temperature, °F
70
283
ST/SRT
1.000
50
1.010
100
1.008
100
1.020
200
1.020
150
1.025
300
1.024
200
1.020
400
1.018
250
1.000
500
0.995
300
0.975
600
0.963
350
0.943
700
0.927
400
0.900
800
0.872
450
0.843
900
0.797
500
0.768
1000
0.698
550
0.672
1100
0.567
600
0.549
*Data source: Fig. 29.
Finally, it may be true that there is no fatigue limit for materials operating at high temperatures. Because of the reduced fatigue resistance, the failure process is, to some
extent, dependent on time.
The limited amount of data available show that the endurance limit for steels
increases slightly as the temperature rises and then begins to fall off in the 400 to 700°F
range, not unlike the behavior of the tensile strength shown in Fig. 29. For this reason
it is probably true that the endurance limit is related to tensile strength at elevated temperatures in the same manner as at room temperature.18 It seems quite logical, therefore,
to employ the same relations to predict endurance limit at elevated temperatures as are
used at room temperature, at least until more comprehensive data become available. At
the very least, this practice will provide a useful standard against which the performance of various materials can be compared.
Table 64 has been obtained from Fig. 29 by using only the tensile-strength data.
Note that the table represents 145 tests of 21 different carbon and alloy steels. A fourthorder polynomial curve t to the data underlying Fig. 29 gives
2
kd = 0.975 + 0.432(103 )TF 0.115(105 )TF
3
4
+ 0.104(108 )TF 0.595(1012 )TF
( 627)
where 70 TF 1000 F.
Two types of problems arise when temperature is a consideration. If the rotatingbeam endurance limit is known at room temperature, then use
ST
kd =
(628)
S RT
18
For more, see Table 2 of ANSI/ASME B106. 1M-1985 shaft standard, and E. A. Brandes (ed.), Smithells
Metals Reference Book, 6th ed., Butterworth, London, 1983, pp. 22134 to 22136, where endurance limits
from 100 to 650°C are tabulated.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
284
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
287
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
from Table 64 or Eq. (627) and proceed as usual. If the rotating-beam endurance limit
is not given, then compute it using Eq. (68) and the temperature-corrected tensile
strength obtained by using the factor from Table 64. Then use kd = 1.
EXAMPLE 65
Solution
A 1035 steel has a tensile strength of 70 kpsi and is to be used for a part that sees 450°F
in service. Estimate the Marin temperature modication factor and ( Se )450 if
(a) The room-temperature endurance limit by test is ( Se )70 = 39.0 kpsi.
(b) Only the tensile strength at room temperature is known.
(a) First, from Eq. (627),
kd = 0.975 + 0.432(103 )(450) 0.115(105 )(4502 )
+ 0.104(108 )(4503 ) 0.595(1012 )(4504 ) = 1.007
Thus,
( Se )450 = kd ( Se )70 = 1.007(39.0) = 39.3 kpsi
Answer
(b) Interpolating from Table 64 gives
( ST / S RT )450 = 1.018 + (0.995 1.018)
450 400
= 1.007
500 400
Thus, the tensile strength at 450°F is estimated as
( Sut )450 = ( ST / S RT )450 ( Sut )70 = 1.007(70) = 70.5 kpsi
From Eq. (68) then,
Answer
( S e )450 = 0.5 ( Sut )450 = 0.5(70.5) = 35.2 kpsi
Part a gives the better estimate due to actual testing of the particular material.
Reliability Factor ke
The discussion presented here accounts for the scatter of data such as shown in
.
Fig. 617 where the mean endurance limit is shown to be Se / Sut = 0.5, or as given by
Eq. (68). Most endurance strength data are reported as mean values. Data presented
by Haugen and Wirching19 show standard deviations of endurance strengths of less than
8 percent. Thus the reliability modication factor to account for this can be written as
ke = 1 0.08 z a
(629)
where za is dened by Eq. (2016) and values for any desired reliability can be determined from Table A10. Table 65 gives reliability factors for some standard specied
reliabilities.
For a more comprehensive approach to reliability, see Sec. 617.
19
E. B. Haugen and P. H. Wirsching, Probabilistic Design, Machine Design, vol. 47, no. 12, 1975,
pp. 1014.
288
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Table 65
Reliability Factors ke
Corresponding to
8 Percent Standard
Deviation of the
Endurance Limit
Reliability, %
Transformation Variate za
Reliability Factor ke
50
0
1.000
90
1.288
0.897
95
1.645
0.868
99
2.326
0.814
99.9
3.091
0.753
99.99
3.719
0.702
99.999
4.265
0.659
99.9999
4.753
0.620
Figure 619
The failure of a case-hardened
part in bending or torsion. In
this example, failure occurs in
the core.
285
Se (case)
or
Case
Core
Se (core)
Miscellaneous-Effects Factor kf
Though the factor k f is intended to account for the reduction in endurance limit due to
all other effects, it is really intended as a reminder that these must be accounted for,
because actual values of k f are not always available.
Residual stresses may either improve the endurance limit or affect it adversely.
Generally, if the residual stress in the surface of the part is compression, the endurance
limit is improved. Fatigue failures appear to be tensile failures, or at least to be caused
by tensile stress, and so anything that reduces tensile stress will also reduce the possibility of a fatigue failure. Operations such as shot peening, hammering, and cold rolling
build compressive stresses into the surface of the part and improve the endurance limit
signicantly. Of course, the material must not be worked to exhaustion.
The endurance limits of parts that are made from rolled or drawn sheets or bars,
as well as parts that are forged, may be affected by the so-called directional characteristics of the operation. Rolled or drawn parts, for example, have an endurance limit
in the transverse direction that may be 10 to 20 percent less than the endurance limit in
the longitudinal direction.
Parts that are case-hardened may fail at the surface or at the maximum core radius,
depending upon the stress gradient. Figure 619 shows the typical triangular stress distribution of a bar under bending or torsion. Also plotted as a heavy line in this gure are
the endurance limits Se for the case and core. For this example the endurance limit of the
core rules the design because the gure shows that the stress σ or τ, whichever applies,
at the outer core radius, is appreciably larger than the core endurance limit.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
286
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Of course, if stress concentration is also present, the stress gradient is much
steeper, and hence failure in the core is unlikely.
Corrosion
It is to be expected that parts that operate in a corrosive atmosphere will have a lowered
fatigue resistance. This is, of course, true, and it is due to the roughening or pitting of
the surface by the corrosive material. But the problem is not so simple as the one of
nding the endurance limit of a specimen that has been corroded. The reason for this is
that the corrosion and the stressing occur at the same time. Basically, this means that in
time any part will fail when subjected to repeated stressing in a corrosive atmosphere.
There is no fatigue limit. Thus the designers problem is to attempt to minimize the factors that affect the fatigue life; these are:
Mean or static stress
Alternating stress
Electrolyte concentration
Dissolved oxygen in electrolyte
Material properties and composition
Temperature
Cyclic frequency
Fluid ow rate around specimen
Local crevices
Electrolytic Plating
Metallic coatings, such as chromium plating, nickel plating, or cadmium plating, reduce
the endurance limit by as much as 50 percent. In some cases the reduction by coatings
has been so severe that it has been necessary to eliminate the plating process. Zinc
plating does not affect the fatigue strength. Anodic oxidation of light alloys reduces
bending endurance limits by as much as 39 percent but has no effect on the torsional
endurance limit.
Metal Spraying
Metal spraying results in surface imperfections that can initiate cracks. Limited tests
show reductions of 14 percent in the fatigue strength.
Cyclic Frequency
If, for any reason, the fatigue process becomes time-dependent, then it also becomes
frequency-dependent. Under normal conditions, fatigue failure is independent of frequency. But when corrosion or high temperatures, or both, are encountered, the cyclic
rate becomes important. The slower the frequency and the higher the temperature, the
higher the crack propagation rate and the shorter the life at a given stress level.
Frettage Corrosion
The phenomenon of frettage corrosion is the result of microscopic motions of tightly
tting parts or structures. Bolted joints, bearing-race ts, wheel hubs, and any set of
tightly tted parts are examples. The process involves surface discoloration, pitting, and
eventual fatigue. The frettage factor k f depends upon the material of the mating pairs
and ranges from 0.24 to 0.90.
289
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
610
287
Stress Concentration and Notch Sensitivity
In Sec. 313 it was pointed out that the existence of irregularities or discontinuities,
such as holes, grooves, or notches, in a part increases the theoretical stresses signicantly in the immediate vicinity of the discontinuity. Equation (348) dened a stress
concentration factor K t (or K ts ), which is used with the nominal stress to obtain the
maximum resulting stress due to the irregularity or defect. It turns out that some materials are not fully sensitive to the presence of notches and hence, for these, a reduced
value of Kt can be used. For these materials, the maximum stress is, in fact,
σmax = K f σ0
or
(630)
τmax = K f s τ0
where K f is a reduced value of K t and σ0 is the nominal stress. The factor K f is commonly called a fatigue stress-concentration factor, and hence the subscript f. So it is
convenient to think of Kf as a stress-concentration factor reduced from Kt because of
lessened sensitivity to notches. The resulting factor is dened by the equation
Kf =
maximum stress in notched specimen
stress in notch-free specimen
(a)
Notch sensitivity q is dened by the equation
q=
Kf 1
Kt 1
qshear =
or
Kfs 1
K ts 1
(631)
where q is usually between zero and unity. Equation (631) shows that if q = 0, then
K f = 1, and the material has no sensitivity to notches at all. On the other hand, if
q = 1, then K f = K t , and the material has full notch sensitivity. In analysis or design
work, nd Kt rst, from the geometry of the part. Then specify the material, nd q, and
solve for Kf from the equation
K f = 1 + q ( K t 1)
K f s = 1 + qshear ( K ts 1)
or
(632)
For steels and 2024 aluminum alloys, use Fig. 620 to nd q for bending and axial
loading. For shear loading, use Fig. 621. In using these charts it is well to know that
the actual test results from which the curves were derived exhibit a large amount of
Figure 620
Notch-sensitivity charts for
steels and UNS A92024-T
wrought aluminum alloys
subjected to reversed bending
or reversed axial loads. For
larger notch radii, use the
values of q corresponding
to the r = 0.16-in (4-mm)
ordinate. (From George Sines
and J. L. Waisman (eds.),
Metal Fatigue, McGraw-Hill,
New York. Copyright ©
1969 by The McGraw-Hill
Companies, Inc. Reprinted by
permission.)
Notch radius r, mm
1.0
0
0.5
S ut
=
1.0
kpsi
200
15
0.8
2.0
2.5
3.0
3.5
4.0
(0.7)
0
(0.4)
10
0.6
1.5
(1.4 GPa)
(1.0)
0
Notch sensitivity q
290
60
0.4
Steels
Alum. alloy
0.2
0
0
0.02
0.04
0.06
0.08
0.10
Notch radius r, in
0.12
0.14
0.16
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
288
II. Failure Prevention
291
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 621
1.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
0.14
0.16
0.8
Notch sensitivity qshear
Notch-sensitivity curves for
materials in reversed torsion.
For larger notch radii, use
the values of qshear
corresponding to r = 0.16 in
(4 mm).
Notch radius r, mm
0
Quenched and drawn steels (Bhn > 200)
Annealed steels (Bhn < 200)
0.6
0.4
Aluminum alloys
0.2
0
0
0.02
0.04
0.06
0.08
0.10
0.12
Notch radius r, in
scatter. Because of this scatter it is always safe to use K f = K t if there is any doubt
about the true value of q. Also, note that q is not far from unity for large notch radii.
The notch sensitivity of the cast irons is very low, varying from 0 to about 0.20,
depending upon the tensile strength. To be on the conservative side, it is recommended
that the value q = 0.20 be used for all grades of cast iron.
Figure 620 has as its basis the Neuber equation, which is given by
Kf = 1 +
Kt 1
1 + a /r
(633)
where a is dened as the Neuber constant and is a material constant. Equating
Eqs. (631) and (633) yields the notch sensitivity equation
q=
1
a
1+
r
(634)
For steel, with Sut in kpsi, the Neuber constant can be approximated by a third-order
polynomial t of data as
a = 0.245 799 0.307 794(102 ) Sut
2
3
+ 0.150 874(104 ) Sut 0.266 978(107 ) Sut
(635)
To use Eq. (633) or (634) for torsion for low-alloy
steels, increase the ultimate
strength by 20 kpsi in Eq. (635) and apply this value of a .
EXAMPLE 66
A steel shaft in bending has an ultimate strength of 690 MPa and a shoulder with a llet radius of 3 mm connecting a 32-mm diameter with a 38-mm diameter. Estimate Kf
using:
(a) Figure 620.
(b) Equations (633) and (635).
292
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Solution
Answer
Answer
289
From Fig. A159, using D /d = 38/32 = 1.1875, r /d = 3/32 = 0.093 75, we read
.
the graph to nd K t = 1.65.
.
(a) From Fig. 620, for Sut = 690 MPa and r = 3 mm, q = 0.84. Thus, from Eq. (632)
.
K f = 1 + q ( K t 1) = 1 + 0.84(1.65 1) = 1.55
(b) From Eq. (635) with Sut = 690 MPa = 100 kpsi, a = 0.0622 in = 0.313 mm.
Substituting this into Eq. (633) with r = 3 mm gives
Kf = 1 +
Kt 1 .
1.65 1
=1+
= 1.55
0.313
1 + a /r
1+
3
For simple loading, it is acceptable to reduce the endurance limit by either dividing
the unnotched specimen endurance limit by K f or multiplying the reversing stress by
K f . However, in dealing with combined stress problems that may involve more than one
value of fatigue-concentration factor, the stresses are multiplied by K f .
EXAMPLE 67
Solution
Consider an unnotched specimen with an endurance limit of 55 kpsi. If the specimen
was notched such that K f = 1.6, what would be the factor of safety against failure for
N > 106 cycles at a reversing stress of 30 kpsi?
(a) Solve by reducing Se .
(b) Solve by increasing the applied stress.
(a) The endurance limit of the notched specimen is given by
Se =
Se
55
= 34.4 kpsi
=
Kf
1.6
and the factor of safety is
Answer
n=
Se
34.4
= 1.15
=
σa
30
(b) The maximum stress can be written as
(σa )max = K f σa = 1.6(30) = 48.0 kpsi
and the factor of safety is
Answer
n=
Se
55
= 1.15
=
K f σa
48
Up to this point, examples illustrated each factor in Marins equation and stress
concentrations alone. Let us consider a number of factors occurring simultaneously.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
290
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
EXAMPLE 68
A 1015 hot-rolled steel bar has been machined to a diameter of 1 in. It is to be placed
in reversed axial loading for 70 000 cycles to failure in an operating environment of
550°F. Using ASTM minimum properties, and a reliability of 99 percent, estimate the
endurance limit and fatigue strength at 70 000 cycles.
Solution
From Table A20, Sut = 50 kpsi at 70°F. Since the rotating-beam specimen endurance
limit is not known at room temperature, we determine the ultimate strength at the elevated temperature rst, using Table 64. From Table 64,
ST
S RT
550
=
0.995 + 0.963
= 0.979
2
The ultimate strength at 550°F is then
( Sut )550 = ( ST / S RT )550 ( Sut )70 = 0.979(50) = 49.0 kpsi
The rotating-beam specimen endurance limit at 550°F is then estimated from Eq. (68)
as
Se = 0.5(49) = 24.5 kpsi
Next, we determine the Marin factors. For the machined surface, Eq. (619) with
Table 62 gives
b
ka = aSut = 2.70(490.265 ) = 0.963
For axial loading, from Eq. (621), the size factor kb = 1, and from Eq. (626) the loading factor is kc = 0.85. The temperature factor kd = 1, since we accounted for the temperature in modifying the ultimate strength and consequently the endurance limit. For
99 percent reliability, from Table 65, ke = 0.814. Finally, since no other conditions
were given, the miscellaneous factor is kf = 1. The endurance limit for the part is estimated by Eq. (618) as
Se = ka kb kc kd ke k f Se
Answer
= 0.963(1)(0.85)(1)(0.814)(1)24.5 = 16.3 kpsi
For the fatigue strength at 70 000 cycles we need to construct the S-N equation. From
p. 277, since Sut = 49 < 70 kpsi, then f 0.9. From Eq. (614)
a=
( f Sut )2
[0.9(49)]2
= 119.3 kpsi
=
Se
16.3
and Eq. (615)
1
b = log
3
f Sut
Se
1
0.9(49)
= log
= 0.1441
3
16.3
Finally, for the fatigue strength at 70 000 cycles, Eq. (613) gives
Answer
S f = a N b = 119.3(70 000)0.1441 = 23.9 kpsi
293
294
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
291
EXAMPLE 69
Figure 622a shows a rotating shaft simply supported in ball bearings at A and D and
loaded by a nonrotating force F of 6.8 kN. Using ASTM minimum strengths, estimate
the life of the part.
Solution
From Fig. 622b we learn that failure will probably occur at B rather than at C or at the
point of maximum moment. Point B has a smaller cross section, a higher bending
moment, and a higher stress-concentration factor than C, and the location of maximum
moment has a larger size and no stress-concentration factor.
We shall solve the problem by rst estimating the strength at point B, since the strength
will be different elsewhere, and comparing this strength with the stress at the same point.
From Table A20 we nd Sut = 690 MPa and Sy = 580 MPa. The endurance limit
Se is estimated as
Se = 0.5(690) = 345 MPa
From Eq. (619) and Table 62,
ka = 4.51(690)0.265 = 0.798
From Eq. (620),
kb = (32/7.62)0.107 = 0.858
Since kc = kd = ke = k f = 1,
Se = 0.798(0.858)345 = 236 MPa
To nd the geometric stress-concentration factor K t we enter Fig. A159 with D /d =
.
1.
38/32 = 1.1875 and r /d = 3/32 = 0.093 75 and read K t = 65. Substituting
Sut = 690/6.89 = 100 kpsi into Eq. (635) yields a = 0.0622 in = 0.313 mm.
Substituting this into Eq. (633) gives
Kf = 1 +
Figure 622
(a) Shaft drawing showing all
dimensions in millimeters; all
llets 3-mm radius. The shaft
rotates and the load is
stationary; material is
machined from AISI 1050
cold-drawn steel. (b) Bendingmoment diagram.
A
1.65 1
Kt 1
=1+
= 1.55
1 + a /r
1 + 0.313/ 3
6.8 kN
B
250
75
C
100
125
10
10
32
30
D
35
38
30
R2
R1
(a )
Mmax
MB
MC
A
B
C
(b)
D
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
292
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
The next step is to estimate the bending stress at point B. The bending moment
is
M B = R1 x =
225(6.8)
225 F
250 =
250 = 695.5 N · m
550
550
Just to the left of B the section modulus is I /c = π d 3 /32 = π 323 /32 = 3.217 (103 )mm3 .
The reversing bending stress is, assuming innite life,
σ = Kf
695.5
MB
= 1.55
(10)6 = 335.1(106 ) Pa = 335.1 MPa
I /c
3.217
This stress is greater than Se and less than Sy. This means we have both nite life and
no yielding on the rst cycle.
For nite life, we will need to use Eq. (616). The ultimate strength, Sut = 690
MPa = 100 kpsi. From Fig. 618, f = 0.844. From Eq. (614)
a=
[0.844(690)]2
( f Sut )2
= 1437 MPa
=
Se
236
and from Eq. (615)
1
b = log
3
f Sut
Se
0.844(690)
1
= 0.1308
= log
3
236
From Eq. (616),
Answer
611
N=
σa
a
1/b
=
335.1
1437
1/0.1308
= 68(103 ) cycles
Characterizing Fluctuating Stresses
Fluctuating stresses in machinery often take the form of a sinusoidal pattern because
of the nature of some rotating machinery. However, other patterns, some quite irregular, do occur. It has been found that in periodic patterns exhibiting a single maximum and a single minimum of force, the shape of the wave is not important, but the
peaks on both the high side (maximum) and the low side (minimum) are important.
Thus Fmax and Fmin in a cycle of force can be used to characterize the force pattern.
It is also true that ranging above and below some baseline can be equally effective
in characterizing the force pattern. If the largest force is Fmax and the smallest force
is Fmin , then a steady component and an alternating component can be constructed
as follows:
Fm =
Fmax + Fmin
2
Fa =
Fmax Fmin
2
where Fm is the midrange steady component of force, and Fa is the amplitude of the
alternating component of force.
295
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
293
Fatigue Failure Resulting from Variable Loading
Figure 623
Stress
a
Stress
r
Time
a
max
m
min
(a )
O
Time
Stress
(d )
Stress
Time
a
r
max
a
O
(b )
min
=0
m
Time
(e)
+
a
Time
Stress
Some stress-time relations:
(a) uctuating stress with highfrequency ripple; (b and c)
nonsinusoidal uctuating
stress; (d) sinusoidal uctuating
stress; (e) repeated stress;
(f ) completely reversed
sinusoidal stress.
Stress
296
Time
O
r
a
m
(c)
=0
(f)
Figure 623 illustrates some of the various stress-time traces that occur. The components of stress, some of which are shown in Fig. 623d, are
σmin = minimum stress
σmax = maximum stress
σa = amplitude component
σm = midrange component
σr = range of stress
σs = static or steady stress
The steady, or static, stress is not the same as the midrange stress; in fact, it may have
any value between σmin and σmax . The steady stress exists because of a xed load or preload applied to the part, and it is usually independent of the varying portion of the load.
A helical compression spring, for example, is always loaded into a space shorter than
the free length of the spring. The stress created by this initial compression is called the
steady, or static, component of the stress. It is not the same as the midrange stress.
We shall have occasion to apply the subscripts of these components to shear stresses as well as normal stresses.
The following relations are evident from Fig. 623:
σmax + σmin
2
σmax σmin
σa =
2
σm =
(636)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
294
II. Failure Prevention
297
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
In addition to Eq. (636), the stress ratio
σmin
σmax
R=
(637)
σa
σm
(638)
and the amplitude ratio
A=
are also dened and used in connection with uctuating stresses.
Equations (636) utilize symbols σa and σm as the stress components at the location under scrutiny. This means, in the absence of a notch, σa and σm are equal to the
nominal stresses σao and σmo induced by loads Fa and Fm , respectively; in the presence
of a notch they are K f σao and K f σmo , respectively, as long as the material remains
without plastic strain. In other words, the fatigue stress concentration factor K f is
applied to both components.
When the steady stress component is high enough to induce localized notch yielding, the designer has a problem. The rst-cycle local yielding produces plastic strain
and strain-strengthening. This is occurring at the location where fatigue crack nucleation and growth are most likely. The material properties ( Sy and Sut ) are new and difcult to quantify. The prudent engineer controls the concept, material and condition of
use, and geometry so that no plastic strain occurs. There are discussions concerning
possible ways of quantifying what is occurring under localized and general yielding in
the presence of a notch, referred to as the nominal mean stress method, residual stress
method, and the like.20 The nominal mean stress method (set σa = K f σao and
σm = σmo ) gives roughly comparable results to the residual stress method, but both are
approximations.
There is the method of Dowling21 for ductile materials, which, for materials with a
pronounced yield point and approximated by an elasticperfectly plastic behavior
model, quantitatively expresses the steady stress component stress-concentration factor
K f m as
Kfm = Kf
Kfm =
Sy K f σao
|σmo |
Kfm = 0
K f |σmax,o | < Sy
K f |σmax,o | > Sy
(639)
K f |σmax,o σmin,o | > 2 Sy
For the purposes of this book, for ductile materials in fatigue,
Avoid localized plastic strain at a notch. Set σa = K f σa ,o and σm = K f σmo .
When plastic strain at a notch cannot be avoided, use Eqs. (639); or conservatively,
set σa = K f σao and use K f m = 1, that is, σm = σmo .
20
R. C. Juvinall, Stress, Strain, and Strength, McGraw-Hill, New York, 1967, articles 14.914.12; R. C.
Juvinall and K. M. Marshek, Fundamentals of Machine Component Design, 4th ed., Wiley, New York, 2006,
Sec. 8.11; M. E. Dowling, Mechanical Behavior of Materials, 2nd ed., Prentice Hall, Englewood Cliffs,
N.J., 1999, Secs. 10.310.5.
21
Dowling, op. cit., p. 437438.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
612
295
Fatigue Failure Criteria for Fluctuating Stress
Now that we have dened the various components of stress associated with a part subjected to uctuating stress, we want to vary both the midrange stress and the stress
amplitude, or alternating component, to learn something about the fatigue resistance of
parts when subjected to such situations. Three methods of plotting the results of such
tests are in general use and are shown in Figs. 624, 625, and 626.
The modied Goodman diagram of Fig. 624 has the midrange stress plotted along
the abscissa and all other components of stress plotted on the ordinate, with tension in
the positive direction. The endurance limit, fatigue strength, or nite-life strength,
whichever applies, is plotted on the ordinate above and below the origin. The midrangestress line is a 45 line from the origin to the tensile strength of the part. The modied
Goodman diagram consists of the lines constructed to Se (or S f ) above and below the
origin. Note that the yield strength is also plotted on both axes, because yielding would
be the criterion of failure if σmax exceeded Sy .
Another way to display test results is shown in Fig. 625. Here the abscissa represents the ratio of the midrange strength Sm to the ultimate strength, with tension plotted to the right and compression to the left. The ordinate is the ratio of the alternating
strength to the endurance limit. The line B C then represents the modied Goodman
criterion of failure. Note that the existence of midrange stress in the compressive region
has little effect on the endurance limit.
The very clever diagram of Fig. 626 is unique in that it displays four of the stress
components as well as the two stress ratios. A curve representing the endurance limit
for values of R beginning at R = 1 and ending with R = 1 begins at Se on the σa axis
and ends at Sut on the σm axis. Constant-life curves for N = 105 and N = 104 cycles
Figure 624
+
Modied Goodman diagram
showing all the strengths and
the limiting values of all the
stress components for a
particular midrange stress.
Su
Stress
Sy
max
ss
tre
x. s
Ma
Se
a
M
id
str ran
es ge
s
r
a
min
45°
0
Sy
m
M
in.
str
ess
Parallel
298
Se
Midrange stress
Su
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
296
II. Failure Prevention
299
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
1.2
Amplitude ratio Sa / Se
1.0
A
B
0.8
0.6
0.4
0.2
C
1.2
1.0
0.8
0.6
0.4
0.2
0
0.2
Compression Sm / Suc
0.4
0.6
0.8
1.0
Tension Sm / Sut
Midrange ratio
Figure 625
Plot of fatigue failures for midrange stresses in both tensile and compressive regions. Normalizing
the data by using the ratio of steady strength component to tensile strength Sm / Sut , steady strength
component to compressive strength Sm / Suc and strength amplitude component to endurance limit
Sa / Se enables a plot of experimental results for a variety of steels. [Data source: Thomas J. Dolan,
Stress Range, Sec. 6.2 in O. J. Horger (ed.), ASME HandbookMetals Engineering Design,
McGraw-Hill, New York, 1953.]
Figure 626
1.5
0.2
A=1
R=0
0.67
0.2
RA
0.43
0.4
0.25
0.6
0.11
0.8
0
1.0
16
0
18
0
2.33
0.4
0
180
,k
m
0
10
tre
es
ng
80
ra
id
60
40
40
i
ps
,k
60
a
s
es
str
g
tin
Se
M
a
rn
40
80
lte
A
60
20
20
20
kpsi
max ,
10
0
Maximum stress
A
6
10
80
160
14
5
10
0
100
Sut
ss
120
4c
10
s
y cle
12
0
ps
i
A=
R = 1.0
12
Master fatigue diagram
created for AISI 4340 steel
having Sut = 158 and
Sy = 147 kpsi. The stress
components at A are
σmin = 20, σmax = 120,
σm = 70, and σa = 50, all in
kpsi. (Source: H. J. Grover,
Fatigue of Aircraft Structures,
U.S. Government Printing
Ofce, Washington, D.C.,
1966, pp. 317, 322. See
also J. A. Collins, Failure of
Materials in Mechanical
Design, Wiley, New York,
1981, p. 216.)
4.0
0.6
120 100 80
60
40
20
0
20
40
Minimum stress
60
min ,
80
100
120
140
kpsi
have been drawn too. Any stress state, such as the one at A, can be described by the minimum and maximum components, or by the midrange and alternating components. And
safety is indicated whenever the point described by the stress components lies below the
constant-life line.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Figure 627
Yield (Langer) line
a
Fatigue diagram showing
various criteria of failure. For
each criterion, points on or
above the respective line
indicate failure. Some point A
on the Goodman line, for
example, gives the strength Sm
as the limiting value of σm
corresponding to the strength
Sa , which, paired with σm , is
the limiting value of σa .
297
Sy
Alternating stress
300
Se
Gerber line
Load line, slope r = Sa /Sm
Modified Goodman line
Sa
A
ASME-elliptic line
Soderberg line
0
0
Sm
Midrange stress
Sy
Sut
m
When the midrange stress is compression, failure occurs whenever σa = Se or
whenever σmax = Syc , as indicated by the left-hand side of Fig. 625. Neither a fatigue
diagram nor any other failure criteria need be developed.
In Fig. 627, the tensile side of Fig. 625 has been redrawn in terms of strengths,
instead of strength ratios, with the same modied Goodman criterion together with four
additional criteria of failure. Such diagrams are often constructed for analysis and
design purposes; they are easy to use and the results can be scaled off directly.
The early viewpoint expressed on a σa σm diagram was that there existed a locus which
divided safe from unsafe combinations of σa and σm . Ensuing proposals included the
parabola of Gerber (1874), the Goodman (1890)22 (straight) line, and the Soderberg (1930)
(straight) line. As more data were generated it became clear that a fatigue criterion, rather
than being a fence, was more like a zone or band wherein the probability of failure could
be estimated. We include the failure criterion of Goodman because
It is a straight line and the algebra is linear and easy.
It is easily graphed, every time for every problem.
It reveals subtleties of insight into fatigue problems.
Answers can be scaled from the diagrams as a check on the algebra.
We also caution that it is deterministic and the phenomenon is not. It is biased and we
cannot quantify the bias. It is not conservative. It is a stepping-stone to understanding; it
is history; and to read the work of other engineers and to have meaningful oral exchanges
with them, it is necessary that you understand the Goodman approach should it arise.
Either the fatigue limit Se or the nite-life strength S f is plotted on the ordinate of
Fig. 627. These values will have already been corrected using the Marin factors of
Eq. (618). Note that the yield strength Sy is plotted on the ordinate too. This serves as
a reminder that rst-cycle yielding rather than fatigue might be the criterion of failure.
The midrange-stress axis of Fig. 627 has the yield strength Sy and the tensile
strength Sut plotted along it.
22
It is difcult to date Goodmans work because it went through several modications and was never
published.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
298
II. Failure Prevention
301
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Five criteria of failure are diagrammed in Fig. 627: the Soderberg, the modied
Goodman, the Gerber, the ASME-elliptic, and yielding. The diagram shows that only
the Soderberg criterion guards against any yielding, but is biased low.
Considering the modied Goodman line as a criterion, point A represents a limiting point with an alternating strength Sa and midrange strength Sm. The slope of the load
line shown is dened as r = Sa / Sm .
The criterion equation for the Soderberg line is
Sm
Sa
+
=1
Se
Sy
(640)
Similarly, we nd the modied Goodman relation to be
Sa
Sm
+
=1
Se
Sut
(641)
Examination of Fig. 625 shows that both a parabola and an ellipse have a better
opportunity to pass among the midrange tension data and to permit quantication of the
probability of failure. The Gerber failure criterion is written as
Sa
+
Se
2
Sm
Sut
(642)
=1
and the ASME-elliptic is written as
2
Sa
Se
2
Sm
Sy
+
(643)
=1
The Langer first-cycle-yielding criterion is used in connection with the fatigue
curve:
(644)
Sa + Sm = Sy
The stresses n σa and n σm can replace Sa and Sm , where n is the design factor or factor
of safety. Then, Eq. (640), the Soderberg line, becomes
σa
σm
1
+
=
Se
Sy
n
Soderberg
(645)
Equation (641), the modied Goodman line, becomes
mod-Goodman
σa
σm
1
+
=
Se
Sut
n
(646)
Equation (642), the Gerber line, becomes
Gerber
n σa
+
Se
n σm
Sut
2
(647)
=1
Equation (643), the ASME-elliptic line, becomes
ASME-elliptic
n σa
Se
2
+
n σm
Sy
2
=1
(648)
We will emphasize the Gerber and ASME-elliptic for fatigue failure criterion and the
Langer for rst-cycle yielding. However, conservative designers often use the modied
Goodman criterion, so we will continue to include it in our discussions. The design
equation for the Langer rst-cycle-yielding is
Langer static yield σa + σm =
Sy
n
(649)
302
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
299
The failure criteria are used in conjunction with a load line, r = Sa / Sm = σa /σm .
Principal intersections are tabulated in Tables 66 to 68. Formal expressions for
fatigue factor of safety are given in the lower panel of Tables 66 to 68. The rst row
of each table corresponds to the fatigue criterion, the second row is the static Langer
criterion, and the third row corresponds to the intersection of the static and fatigue
Table 66
Intersecting Equations
Amplitude and Steady
Coordinates of Strength
and Important
Intersections in First
Quadrant for Modied
Goodman and Langer
Failure Criteria
Sa
Sm
+
=1
Se
Sut
Intersection Coordinates
Sa =
Sa
Sm
r Se Sut
r Sut + Se
Sm =
Sa
Sm
+
=1
Sy
Sy
Sa
Sm
Load line r =
Sa
r
Sa =
Load line r =
r Sy
1+r
Sm =
Sy
1+r
Sy Se Sut
Sut Se
Sa
Sm
+
=1
Se
Sut
Sm =
Sa
Sm
+
=1
Sy
Sy
Sa = Sy Sm , r crit = Sa / Sm
Fatigue factor of safety
1
n f = σa
σm
+
Se
Sut
Table 67
Amplitude and Steady
Coordinates of Strength
and Important
Intersections in First
Quadrant for Gerber
and Langer Failure
Criteria
Intersecting Equations
Sa
+
Se
Sm
Sut
Intersection Coordinates
2
2 Se 2
r 2 Sut
1 + 1 +
Sa =
2 Se
r Sut
2
=1
Load line r =
Sm =
Sa
Sm
+
=1
Sy
Sy
Load line r =
Sa
+
Se
Sm
Sut
Sa
r
Sa =
Sa
Sm
r Sy
1+r
Sy
1+r
2
Sut
Sm =
1
2 Se
Sa
Sm
Sm =
2
=1
Sa
Sm
+
=1
Sy
Sy
1+
2 Se
Sut
2
Sy
1
Se
Sa = Sy Sm , r crit = Sa / Sm
Fatigue factor of safety
1
nf =
2
Sut
σm
2
σa
1 +
Se
1+
2σm Se
Sut σa
2
σm > 0
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
300
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Table 68
Intersecting Equations
Amplitude and Steady
Coordinates of Strength
and Important
Intersections in First
Quadrant for ASMEElliptic and Langer
Failure Criteria
Sa
Se
2
+
Sm
Sy
Intersection Coordinates
2
22
r 2 Se Sy
Sa =
=1
2
2
Se + r 2 Sy
Load line r = Sa / Sm
Sm =
Sa
r
Sa
Sm
+
=1
Sy
Sy
Sa =
r Sy
1+r
Load line r = Sa / Sm
Sm =
Sy
1+r
Sa
Se
2
+
Sm
Sy
2
Sa = 0,
=1
Sa
Sm
+
=1
Sy
Sy
2
2 Sy Se
2 + S2
Se
y
Sm = Sy Sa , r crit = Sa / Sm
Fatigue factor of safety
nf =
1
2
( σa / Se ) + σm / Sy
2
criteria. The rst column gives the intersecting equations and the second column the
intersection coordinates.
There are two ways to proceed with a typical analysis. One method is to assume
that fatigue occurs first and use one of Eqs. (645) to (648) to determine n or size,
depending on the task. Most often fatigue is the governing failure mode. Then
follow with a static check. If static failure governs then the analysis is repeated using
Eq. (649).
Alternatively, one could use the tables. Determine the load line and establish which
criterion the load line intersects rst and use the corresponding equations in the tables.
Some examples will help solidify the ideas just discussed.
EXAMPLE 610
A 1.5-in-diameter bar has been machined from an AISI 1050 cold-drawn bar. This part
is to withstand a uctuating tensile load varying from 0 to 16 kip. Because of the ends,
and the llet radius, a fatigue stress-concentration factor K f is 1.85 for 106 or larger
life. Find Sa and Sm and the factor of safety guarding against fatigue and rst-cycle
yielding, using (a) the Gerber fatigue line and (b) the ASME-elliptic fatigue line.
Solution
We begin with some preliminaries. From Table A20, Sut = 100 kpsi and Sy = 84 kpsi.
Note that Fa = Fm = 8 kip. The Marin factors are, deterministically,
ka = 2.70(100)0.265 = 0.797: Eq. (619), Table 62, p. 279
kb = 1 (axial loading, see kc )
303
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
301
kc = 0.85: Eq. (626), p. 282
kd = ke = k f = 1
Se = 0.797(1)0.850(1)(1)(1)0.5(100) = 33.9 kpsi: Eqs. (68), (618), p. 274, p. 279
The nominal axial stress components σao and σmo are
σao =
4 Fa
4(8)
=
= 4.53 kpsi
2
πd
π 1.52
σmo =
4 Fm
4(8)
=
= 4.53 kpsi
2
πd
π 1.52
Applying K f to both components σao and σmo constitutes a prescription of no notch
yielding:
σa = K f σao = 1.85(4.53) = 8.38 kpsi = σm
Answer
(a) Let us calculate the factors of safety rst. From the bottom panel from Table 67 the
factor of safety for fatigue is
1 100 2 8.38
2(8.38)33.9 2
nf =
= 3.66
1 + 1 +
2 8.38
33.9
100(8.38)
From Eq. (649) the factor of safety guarding against rst-cycle yield is
Answer
Answer
Figure 628
ny =
Sy
84
= 5.01
=
σa + σm
8.38 + 8.38
Thus, we see that fatigue will occur rst and the factor of safety is 3.68. This can be
seen in Fig. 628 where the load line intersects the Gerber fatigue curve rst at point B.
If the plots are created to true scale it would be seen that n f = O B / O A.
From the rst panel of Table 67, r = σa /σm = 1,
2
2
2
(1) 100
2(33.9)
= 30.7 kpsi
Sa =
1 + 1 +
2(33.9)
(1)100
100
Principal points A, B, C, and
D on the designers diagram
drawn for Gerber, Langer, and
load line.
a,
kpsi
84
Stress amplitude
304
50
Load line
C
42
Langer line
33.9
30.7
B
D
20
rcrit
Gerber
fatigue curve
A
8.38
0
0
8.38
30.7
42 50
Midrange stress
m,
64
kpsi
84
100
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
302
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Answer
Sm =
30.7
Sa
=
= 30.7 kpsi
r
1
As a check on the previous result, n f = O B / O A = Sa /σa = Sm /σm = 30.7/8.38 =
3.66 and we see total agreement.
We could have detected that fatigue failure would occur rst without drawing Fig.
628 by calculating rcrit . From the third row third column panel of Table 67, the intersection point between fatigue and rst-cycle yield is
1002
2(33.9) 2
84
1 1+
= 64.0 kpsi
Sm =
1
2(33.9)
100
33.9
Sa = Sy Sm = 84 64 = 20 kpsi
The critical slope is thus
Sa
20
= 0.312
=
Sm
64
rcrit =
which is less than the actual load line of r = 1. This indicates that fatigue occurs before
rst-cycle-yield.
(b) Repeating the same procedure for the ASME-elliptic line, for fatigue
Answer
nf =
(8.38/33.9) 2
1
= 3.75
+ (8.38/84) 2
Again, this is less than n y = 5.01 and fatigue is predicted to occur rst. From the rst
row second column panel of Table 68, with r = 1, we obtain the coordinates Sa and
Sm of point B in Fig. 629 as
Figure 629
100
Principal points A, B, C, and
D on the designers diagram
drawn for ASME-elliptic,
Langer, and load lines.
Stress amplitude
a,
kpsi
84
50
Load line
C
42
Langer line
B
31.4
D
23.5
ASME-elliptic line
A
8.38
0
0
8.38
31.4
42 50
Midrange stress
60.5
kpsi
m,
84
100
305
306
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Answer
Sa =
( 1) 2 33.92 (84) 2
= 31.4 kpsi,
33.92 + (1) 2 842
Sm =
303
Sa
31.4
=
= 31.4 kpsi
r
1
To verify the fatigue factor of safety, n f = Sa /σa = 31.4/8.38 = 3.75.
As before, let us calculate rcrit . From the third row second column panel of
Table 68,
2(84)33.92
= 23.5 kpsi,
33.92 + 842
Sa =
rcrit =
Sm = Sy Sa = 84 23.5 = 60.5 kpsi
Sa
23.5
=
= 0.388
Sm
60.5
which again is less than r = 1, verifying that fatigue occurs rst with n f = 3.75.
The Gerber and the ASME-elliptic fatigue failure criteria are very close to each
other and are used interchangeably. The ANSI/ASME Standard B106.1M1985 uses
ASME-elliptic for shafting.
EXAMPLE 611
A at-leaf spring is used to retain an oscillating at-faced follower in contact with a
plate cam. The follower range of motion is 2 in and xed, so the alternating component
of force, bending moment, and stress is xed, too. The spring is preloaded to adjust to
various cam speeds. The preload must be increased to prevent follower oat or jump.
For lower speeds the preload should be decreased to obtain longer life of cam and
follower surfaces. The spring is a steel cantilever 32 in long, 2 in wide, and 1 in thick,
4
as seen in Fig. 630a. The spring strengths are Sut = 150 kpsi, Sy = 127 kpsi, and Se =
28 kpsi fully corrected. The total cam motion is 2 in. The designer wishes to preload
the spring by deecting it 2 in for low speed and 5 in for high speed.
(a) Plot the Gerber-Langer failure lines with the load line.
(b) What are the strength factors of safety corresponding to 2 in and 5 in preload?
Solution
We begin with preliminaries. The second area moment of the cantilever cross section is
I=
bh 3
2(0.25)3
=
= 0.00260 in4
12
12
Since, from Table A9, beam 1, force F and deection y in a cantilever are related by
F = 3 E I y / l 3, then stress σ and deection y are related by
σ=
where K =
Mc
32 Fc
32(3 E I y ) c
96 Ecy
=
=
=
= Ky
I
I
l3
I
l3
96(30 · 106 )0.125
96 Ec
=
= 10.99(103 ) psi/in = 10.99 kpsi/in
3
l
323
Now the minimums and maximums of y and σ can be dened by
ymin = δ
ymax = 2 + δ
σmin = K δ
σmax = K (2 + δ)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
304
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 630
Cam follower retaining spring.
(a) Geometry; (b) designers
fatigue diagram for Ex. 611.
1
4
2 in
+
in
32 in
= 2 in
+
= 2 in preload
= 5 in
= 5 in preload
+
(a )
Amplitude stress component
a,
kpsi
150
100
Langer line
50
Gerber line
A
0
A'
11
33
A"
50
65.9
Steady stress component
m,
100
kpsi
115.6 127
150
(b)
The stress components are thus
σa =
K (2 + δ) K δ
= K = 10.99 kpsi
2
σm =
K (2 + δ) + K δ
= K (1 + δ) = 10.99(1 + δ)
2
For δ = 0,
σa = σm = 10.99 = 11 kpsi
307
308
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
For δ = 2 in,
σa = 11 kpsi, σm = 10.99(1 + 2) = 33 kpsi
For δ = 5 in,
305
σa = 11 kpsi, σm = 10.99(1 + 5) = 65.9 kpsi
(a) A plot of the Gerber and Langer criteria is shown in Fig. 630b. The three preload
deections of 0, 2, and 5 in are shown as points A, A , and A . Note that since σa is
constant at 11 kpsi, the load line is horizontal and does not contain the origin. The
intersection between the Gerber line and the load line is found from solving Eq. (642)
for Sm and substituting 11 kpsi for Sa :
Sm = Sut 1
Sa
11
= 116.9 kpsi
= 150 1
Se
28
The intersection of the Langer line and the load line is found from solving Eq. (644)
for Sm and substituting 11 kpsi for Sa :
Sm = Sy Sa = 127 11 = 116 kpsi
The threats from fatigue and rst-cycle yielding are approximately equal.
(b) For δ = 2 in,
Answer
nf =
Sm
116.9
= 3.54
=
σm
33
ny =
116
= 3.52
33
and for δ = 5 in,
Answer
EXAMPLE 612
Solution
nf =
116.9
= 1.77
65.9
ny =
116
= 1.76
65.9
A steel bar undergoes cyclic loading such that σmax = 60 kpsi and σmin = 20 kpsi. For
the material, Sut = 80 kpsi, Sy = 65 kpsi, a fully corrected endurance limit of Se =
40 kpsi, and f = 0.9. Estimate the number of cycles to a fatigue failure using:
(a) Modied Goodman criterion.
(b) Gerber criterion.
From the given stresses,
σa =
60 (20)
= 40 kpsi
2
σm =
60 + (20)
= 20 kpsi
2
From the material properties, Eqs. (614) to (616), p. 277, give
a=
( f Sut )2
[0.9(80)]2
= 129.6 kpsi
=
Se
40
1
b = log
3
N=
Sf
a
1
0.9(80)
= log
= 0.0851
3
40
f Sut
Se
1/b
=
Sf
129.6
where S f replaced σa in Eq. (616).
1/0.0851
(1)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
306
II. Failure Prevention
309
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
(a) The modied Goodman line is given by Eq. (646), p. 298, where the endurance
limit Se is used for innite life. For nite life at S f > Se , replace Se with S f in Eq.
(646) and rearrange giving
Sf =
σa
40
= 53.3 kpsi
σm =
20
1
1
Sut
80
Substituting this into Eq. (1) yields
Answer
N=
53.3
129.6
1/0.0851
.
= 3.4(104 ) cycles
(b) For Gerber, similar to part (a), from Eq. (647),
Sf =
σa
1
σm
Sut
2
=
40
1
20
80
2
= 42.7 kpsi
Again, from Eq. (1),
Answer
N=
42.7
129.6
1/0.0851
.
= 4.6(105 ) cycles
Comparing the answers, we see a large difference in the results. Again, the modied
Goodman criterion is conservative as compared to Gerber for which the moderate difference in S f is then magnied by a logarithmic S, N relationship.
For many brittle materials, the rst quadrant fatigue failure criteria follows a concave upward Smith-Dolan locus represented by
Sa
1 Sm / Sut
=
Se
1 + Sm / Sut
(650)
n σa
1 n σm / Sut
=
Se
1 + n σm / Sut
(651)
or as a design equation,
For a radial load line of slope r, we substitute Sa / r for Sm in Eq. (650) and solve for
Sa , obtaining
Sa =
r Sut + Se
1 +
2
1+
4r Sut Se
(r Sut + Se )2
(652)
The fatigue diagram for a brittle material differs markedly from that of a ductile material
because:
Yielding is not involved since the material may not have a yield strength.
Characteristically, the compressive ultimate strength exceeds the ultimate tensile
strength severalfold.
310
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
307
First-quadrant fatigue failure locus is concave-upward (Smith-Dolan), for example,
and as at as Goodman. Brittle materials are more sensitive to midrange stress, being
lowered, but compressive midrange stresses are benecial.
Not enough work has been done on brittle fatigue to discover insightful generalities,
so we stay in the rst and a bit of the second quadrant.
The most likely domain of designer use is in the range from Sut σm Sut . The
locus in the rst quadrant is Goodman, Smith-Dolan, or something in between. The portion of the second quadrant that is used is represented by a straight line between the
points Sut , Sut and 0, Se , which has the equation
Sa = Se +
Se
1 Sm
Sut
Sut Sm 0 (for cast iron)
(653)
Table A24 gives properties of gray cast iron. The endurance limit stated is really
ka kb Se and only corrections kc , kd , ke , and k f need be made. The average kc for axial
and torsional loading is 0.9.
EXAMPLE 613
A grade 30 gray cast iron is subjected to a load F applied to a 1 by 3 -in cross-section
8
link with a 1 -in-diameter hole drilled in the center as depicted in Fig. 631a. The sur4
faces are machined. In the neighborhood of the hole, what is the factor of safety guarding against failure under the following conditions:
(a) The load F = 1000 lbf tensile, steady.
(b) The load is 1000 lbf repeatedly applied.
(c) The load uctuates between 1000 lbf and 300 lbf without column action.
Use the Smith-Dolan fatigue locus.
Alternating stress,
a
F
Sut
1 in
1
4
r = 1.86
in D. drill
Sa = 18.5 kpsi
Se
3
8
r=1
in
Sa = 7.63
Sm
F
Sut
9.95
0
Midrange stress
(a)
7.63 10
m,
20
30 Sut
kpsi
(b)
Figure 631
The grade 30 cast-iron part in axial fatigue with (a) its geometry displayed and (b) its designers fatigue diagram for the
circumstances of Ex. 613.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
308
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Solution
Some preparatory work is needed. From Table A24, Sut = 31 kpsi, Suc = 109 kpsi,
ka kb Se = 14 kpsi. Since kc for axial loading is 0.9, then Se = (ka kb Se )kc = 14(0.9) =
12.6 kpsi. From Table A151, A = t (w d ) = 0.375(1 0.25) = 0.281 in2 , d /w =
0.25/1 = 0.25, and K t = 2.45. The notch sensitivity for cast iron is 0.20 (see p. 288),
so
K f = 1 + q ( K t 1) = 1 + 0.20(2.45 1) = 1.29
(a) σa =
K f Fa
1.29(0)
=
=0
A
0.281
σm =
K f Fm
1.29(1000) 3
=
(10 ) = 4.59 kpsi
A
0.281
and
Answer
n=
Sut
31.0
= 6.75
=
σm
4.59
Fa = Fm =
1000
F
=
= 500 lbf
2
2
σa = σm =
(b)
K f Fa
1.29(500) 3
=
(10 ) = 2.30 kpsi
A
0.281
r=
σa
=1
σm
From Eq. (652),
Sa =
(1)31 + 12.6
1 +
2
Answer
n=
(c)
Fa =
Fm =
1+
= 7.63 kpsi
Sa
7.63
= 3.32
=
σa
2.30
1
|300 (1000)| = 650 lbf
2
1
[300 + (1000)] = 350 lbf
2
r=
4(1)31(12.6)
[(1)31 + 12.6]2
σa =
σm =
1.29(650) 3
(10 ) = 2.98 kpsi
0.281
1.29(350) 3
(10 ) = 1.61 kpsi
0.281
σa
3.0
= 1.86
=
σm
1.61
From Eq. (653), Sa = Se + ( Se / Sut 1) Sm and Sm = Sa / r . It follows that
Sa =
Answer
1
1
r
Se
Se
1
Sut
n=
=
12.6
12.6
1
1
1
1.86 31
= 18.5 kpsi
Sa
18.5
= 6.20
=
σa
2.98
Figure 631b shows the portion of the designers fatigue diagram that was constructed.
311
312
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
613
309
Torsional Fatigue Strength under
Fluctuating Stresses
Extensive tests by Smith23 provide some very interesting results on pulsating torsional
fatigue. Smiths rst result, based on 72 tests, shows that the existence of a torsional
steady-stress component not more than the torsional yield strength has no effect on
the torsional endurance limit, provided the material is ductile, polished, notch-free, and
cylindrical.
Smiths second result applies to materials with stress concentration, notches, or
surface imperfections. In this case, he nds that the torsional fatigue limit decreases
monotonically with torsional steady stress. Since the great majority of parts will have
surfaces that are less than perfect, this result indicates Gerber, ASME-elliptic, and other
approximations are useful. Joerres of Associated Spring-Barnes Group, conrms
Smiths results and recommends the use of the modied Goodman relation for pulsating torsion. In constructing the Goodman diagram, Joerres uses
Ssu = 0.67 Sut
(654)
Also, from Chap. 5, Ssy = 0.577 Syt from distortion-energy theory, and the mean load
factor kc is given by Eq. (626), or 0.577. This is discussed further in Chap. 10.
614
Combinations of Loading Modes
It may be helpful to think of fatigue problems as being in three categories:
Completely reversing simple loads
Fluctuating simple loads
Combinations of loading modes
The simplest category is that of a completely reversed single stress which is handled with the S-N diagram, relating the alternating stress to a life. Only one type of
loading is allowed here, and the midrange stress must be zero. The next category incorporates general uctuating loads, using a criterion to relate midrange and alternating
stresses (modied Goodman, Gerber, ASME-elliptic, or Soderberg). Again, only one
type of loading is allowed at a time. The third category, which we will develop in this
section, involves cases where there are combinations of different types of loading, such
as combined bending, torsion, and axial.
In Sec. 69 we learned that a load factor kc is used to obtain the endurance limit,
and hence the result is dependent on whether the loading is axial, bending, or torsion.
In this section we want to answer the question, How do we proceed when the loading
is a mixture of, say, axial, bending, and torsional loads? This type of loading introduces
a few complications in that there may now exist combined normal and shear stresses,
each with alternating and midrange values, and several of the factors used in determining the endurance limit depend on the type of loading. There may also be multiple
stress-concentration factors, one for each mode of loading. The problem of how to deal
with combined stresses was encountered when developing static failure theories. The
distortion energy failure theory proved to be a satisfactory method of combining the
23
James O. Smith, The Effect of Range of Stress on the Fatigue Strength of Metals, Univ. of Ill. Eng. Exp.
Sta. Bull. 334, 1942.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
310
II. Failure Prevention
313
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
multiple stresses on a stress element into a single equivalent von Mises stress. The same
approach will be used here.
The rst step is to generate two stress elementsone for the alternating stresses and
one for the midrange stresses. Apply the appropriate fatigue stress concentration factors
to each of the stresses; i.e., apply ( K f ) bending for the bending stresses, ( K f s ) torsion for the
torsional stresses, and ( K f ) axial for the axial stresses. Next, calculate an equivalent von
Mises stress for each of these two stress elements, σa and σm . Finally, select a fatigue
failure criterion (modied Goodman, Gerber, ASME-elliptic, or Soderberg) to complete
the fatigue analysis. For the endurance limit, Se , use the endurance limit modiers,
ka , kb , and kc , for bending. The torsional load factor, kc = 0.59 should not be applied as
it is already accounted for in the von Mises stress calculation (see footnote 17 on page
282). The load factor for the axial load can be accounted for by dividing the alternating
axial stress by the axial load factor of 0.85. For example, consider the common case of
a shaft with bending stresses, torsional shear stresses, and axial stresses. For this case,
1/2
the von Mises stress is of the form σ = σx 2 + 3τx y 2
. 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
σa =
( K f ) bending ( σa ) bending + ( K f ) axial
( σa ) axial
0.85
1/2
2
+ 3 ( K f s ) torsion ( τa ) torsion
2
(655)
σm =
( K f ) bending ( σm ) bending + ( K f ) axial ( σm ) axial
2
+ 3 ( K f s ) torsion ( τm ) torsion
2 1/2
(656)
For rst-cycle localized yielding, the maximum von Mises stress is calculated. This
would be done by rst adding the axial and bending alternating and midrange stresses to
obtain σmax and adding the alternating and midrange shear stresses to obtain τmax . Then
substitute σmax and τmax into the equation for the von Mises stress. A simpler and more con.
servative method is to add Eq. (655) and Eq. (656). That is, let σmax = σa + σm
If the stress components are not in phase but have the same frequency, the maxima
can be found by expressing each component in trigonometric terms, using phase angles,
and then nding the sum. If two or more stress components have differing frequencies,
the problem is difcult; one solution is to assume that the two (or more) components
often reach an in-phase condition, so that their magnitudes are additive.
EXAMPLE 614
A rotating shaft is made of 42- × 4-mm AISI 1018 cold-drawn steel tubing and has a
6-mm-diameter hole drilled transversely through it. Estimate the factor of safety guarding against fatigue and static failures using the Gerber and Langer failure criteria for the
following loading conditions:
(a) The shaft is subjected to a completely reversed torque of 120 N · m in phase with a
completely reversed bending moment of 150 N · m.
(b) The shaft is subjected to a pulsating torque uctuating from 20 to 160 N · m and a
steady bending moment of 150 N · m.
Solution
Here we follow the procedure of estimating the strengths and then the stresses, followed
by relating the two.
314
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
311
From Table A20 we nd the minimum strengths to be Sut = 440 MPa and Sy =
370 MPa. The endurance limit of the rotating-beam specimen is 0.5(440) = 220 MPa.
The surface factor, obtained from Eq. (619) and Table 62, p. 279 is
ka = 4.51 Sut0.265 = 4.51(440)0.265 = 0.899
From Eq. (620) the size factor is
kb =
d
7.62
0.107
=
42
7.62
0.107
= 0.833
The remaining Marin factors are all unity, so the modied endurance strength Se is
Se = 0.899(0.833)220 = 165 MPa
(a) Theoretical stress-concentration factors are found from Table A16. Using a / D =
6/42 = 0.143 and d / D = 34/42 = 0.810, and using linear interpolation, we obtain
A = 0.798 and K t = 2.366 for bending; and A = 0.89 and K ts = 1.75 for torsion.
Thus, for bending,
Z net =
πA
π (0.798)
( D4 d 4) =
[(42) 4 (34) 4 ] = 3.31 (103 )mm3
32 D
32(42)
and for torsion
Jnet =
πA 4
π (0.89)
( D d 4) =
[(42) 4 (34) 4 ] = 155 (103 )mm4
32
32
Next, using Figs. 620 and 621, pp. 287288, with a notch radius of 3 mm we nd the
notch sensitivities to be 0.78 for bending and 0.96 for torsion. The two corresponding
fatigue stress-concentration factors are obtained from Eq. (632) as
K f = 1 + q ( K t 1) = 1 + 0.78(2.366 1) = 2.07
K f s = 1 + 0.96(1.75 1) = 1.72
The alternating bending stress is now found to be
σxa = K f
M
150
= 93.8(106 )Pa = 93.8 MPa
= 2.07
Z net
3.31(106 )
and the alternating torsional stress is
τx ya = K f s
120(42)(103 )
TD
= 28.0(106 )Pa = 28.0 MPa
= 1.72
2 Jnet
2(155)(109 )
The midrange von Mises component σm is zero. The alternating component σa is given
by
2
2
σa = σxa + 3τx ya
1/2
= [93.82 + 3(282 )]1/2 = 105.6 MPa
Since Se = Sa , the fatigue factor of safety n f is
Answer
nf =
Sa
165
= 1.56
=
σa
105.6
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
312
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 632
Designers fatigue diagram for
Ex. 614.
Von Mises amplitude stress component a , MPa
'
400
300
200
Gerber
165
r = 0.28
100
105.6
85.5
0
305
Von Mises steady stress component m, MPa
'
440
500
The rst-cycle yield factor of safety is
Answer
ny =
Sy
370
= 3.50
=
σa
105.6
There is no localized yielding; the threat is from fatigue. See Fig. 632.
(b) This part asks us to nd the factors of safety when the alternating component is due
to pulsating torsion, and a steady component is due to both torsion and bending. We
have Ta = (160 20)/2 = 70 N · m and Tm = (160 + 20) /2 = 90 N · m. The corresponding amplitude and steady-stress components are
τx ya = K f s
70(42)(103 )
Ta D
= 16.3(106 )Pa = 16.3 MPa
= 1.72
2 Jnet
2(155)(109 )
τx ym = K f s
Tm D
90(42)(103 )
= 21.0(106 )Pa = 21.0 MPa
= 1.72
2 Jnet
2(155)(109 )
The steady bending stress component σxm is
σxm = K f
150
Mm
= 93.8(106 )Pa = 93.8 MPa
= 2.07
Z net
3.31(106 )
The von Mises components σa and σm are
σa = [3(16.3)2 ]1/2 = 28.2 MPa
σm = [93.82 + 3(21)2 ]1/2 = 100.6 MPa
From Table 67, p. 299, the fatigue factor of safety is
Answer
nf =
1
2
440
100.6
2
28.2
1 +
165
1+
2(100.6)165
440(28.2)
2
= 3.03
315
316
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
313
From the same table, with r = σa /σm = 28.2/100.6 = 0.280, the strengths can be
shown to be Sa = 85.5 MPa and Sm = 305 MPa. See the plot in Fig. 632.
The rst-cycle yield factor of safety n y is
Answer
ny =
σa
Sy
370
= 2.87
=
+ σm
28.2 + 100.6
There is no notch yielding. The likelihood of failure may rst come from rst-cycle
yielding at the notch. See the plot in Fig. 632.
615
Varying, Fluctuating Stresses;
Cumulative Fatigue Damage
Instead of a single fully reversed stress history block composed of n cycles, suppose a
machine part, at a critical location, is subjected to
A fully reversed stress σ1 for n 1 cycles, σ2 for n 2 cycles, . . . , or
A wiggly time line of stress exhibiting many and different peaks and valleys.
What stresses are signicant, what counts as a cycle, and what is the measure of
damage incurred? Consider a fully reversed cycle with stresses varying 60, 80, 40, and
60 kpsi and a second fully reversed cycle 40, 60, 20, and 40 kpsi as depicted in
Fig. 633a. First, it is clear that to impose the pattern of stress in Fig. 633a on a part
it is necessary that the time trace look like the solid line plus the dashed line in Fig.
633a. Figure 633b moves the snapshot to exist beginning with 80 kpsi and ending
with 80 kpsi. Acknowledging the existence of a single stress-time trace is to discover a
hidden cycle shown as the dashed line in Fig. 633b. If there are 100 applications of
the all-positive stress cycle, then 100 applications of the all-negative stress cycle, the
Figure 633
100
100
50
50
0
0
50
50
Variable stress diagram
prepared for assessing
cumulative damage.
(a)
(b)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
314
II. Failure Prevention
317
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
hidden cycle is applied but once. If the all-positive stress cycle is applied alternately
with the all-negative stress cycle, the hidden cycle is applied 100 times.
To ensure that the hidden cycle is not lost, begin on the snapshot with the largest
(or smallest) stress and add previous history to the right side, as was done in Fig. 633b.
Characterization of a cycle takes on a maxminsame max (or minmaxsame min)
form. We identify the hidden cycle rst by moving along the dashed-line trace in
Fig. 633b identifying a cycle with an 80-kpsi max, a 60-kpsi min, and returning to
80 kpsi. Mentally deleting the used part of the trace (the dashed line) leaves a 40, 60,
40 cycle and a 40, 20, 40 cycle. Since failure loci are expressed in terms of stress
amplitude component σa and steady component σm , we use Eq. (636) to construct the
table below:
Cycle Number
max
1
80
60
70
10
2
60
40
10
50
3
20
40
10
30
min
a
m
The most damaging cycle is number 1. It could have been lost.
Methods for counting cycles include:
Number of tensile peaks to failure.
All maxima above the waveform mean, all minima below.
The global maxima between crossings above the mean and the global minima
between crossings below the mean.
All positive slope crossings of levels above the mean, and all negative slope crossings of levels below the mean.
A modication of the preceding method with only one count made between successive crossings of a level associated with each counting level.
Each local maxi-min excursion is counted as a half-cycle, and the associated amplitude is half-range.
The preceding method plus consideration of the local mean.
Rain-ow counting technique.
The method used here amounts to a variation of the rain-ow counting technique.
The Palmgren-Miner24 cycle-ratio summation rule, also called Miners rule, is
written
ni
=c
Ni
(657)
where n i is the number of cycles at stress level σi and Ni is the number of cycles to failure at stress level σi . The parameter c has been determined by experiment; it is usually
found in the range 0.7 < c < 2.2 with an average value near unity.
24
A. Palmgren, Die Lebensdauer von Kugellagern, ZVDI, vol. 68, pp. 339341, 1924; M. A. Miner,
Cumulative Damage in Fatigue, J. Appl. Mech., vol. 12, Trans. ASME, vol. 67, pp. A159A164, 1945.
318
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
315
Using the deterministic formulation as a linear damage rule we write
ni
Ni
D=
(658)
where D is the accumulated damage. When D = c = 1, failure ensues.
EXAMPLE 615
Solution
Given a part with Sut = 151 kpsi and at the critical location of the part, Se = 67.5 kpsi.
For the loading of Fig. 633, estimate the number of repetitions of the stress-time block
in Fig. 633 that can be made before failure.
From Fig. 618, p. 277, for Sut = 151 kpsi, f = 0.795. From Eq. (614),
a=
( f Sut )2
[0.795(151)]2
= 213.5 kpsi
=
Se
67.5
From Eq. (615),
1
b = log
3
f Sut
Se
1
0.795(151)
= log
= 0.0833
3
67.5
So,
S f = 213.5 N 0.0833
N=
Sf
213.5
1/0.0833
(1), (2)
We prepare to add two columns to the previous table. Using the Gerber fatigue criterion,
Eq. (647), p. 298, with Se = S f , and n = 1, we can write
σa
1 (σm / Sut )2
Se
Sf =
σm > 0
σm 0
(3)
Cycle 1: r = σa /σm = 70/10 = 7, and the strength amplitude from Table 67, p. 299, is
72 1512
2(67.5) 2
= 67.2 kpsi
1 + 1 +
Sa =
2(67.5)
7(151)
Since σa > Sa , that is, 70 > 67.2, life is reduced. From Eq. (3),
Sf =
70
= 70.3 kpsi
1 (10/151)2
and from Eq. (2)
N=
70.3
213.5
1/0.0833
= 619(103 ) cycles
Cycle 2: r = 10/50 = 0.2, and the strength amplitude is
2
2
2
0.2 151
2(67.5)
Sa =
1 + 1 +
= 24.2 kpsi
2(67.5)
0.2(151)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
316
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Since σa < Sa , that is 10 < 24.2, then S f = Se and indenite life follows. Thus,
.
N
Cycle 3: r = 10/30 = 0.333, and since σm < 0, S f = Se , indenite life follows and
N
Cycle Number
Sf , kpsi
N , cycles
1
70.3
619(103)
2
67.5
3
67.5
From Eq. (658) the damage per block is
N
1
1
1
ni
=
+
+
=N
3)
Ni
619(10
619(103 )
D=
Answer
Setting D = 1 yields N = 619(103 ) cycles.
To further illustrate the use of the Miner rule, let us choose a steel having the prop
erties Sut = 80 kpsi, Se,0 = 40 kpsi, and f = 0.9, where we have used the designation
Se,0 instead of the more usual Se to indicate the endurance limit of the virgin, or undamaged, material. The log Slog N diagram for this material is shown in Fig. 634 by the
heavy solid line. Now apply, say, a reversed stress σ1 = 60 kpsi for n 1 = 3000 cycles.
Since σ1 > Se,0 , the endurance limit will be damaged, and we wish to nd the new
endurance limit Se,1 of the damaged material using the Miner rule. The equation of the
virgin material failure line in Fig. 634 in the 103 to 106 cycle range is
S f = a N b = 129.6 N 0.085 091
The cycles to failure at stress level σ1 = 60 kpsi are
N1 =
Figure 634
σ1
129.6
1/0.085 091
60
129.6
=
1/0.085 091
= 8520 cycles
4.9
Use of the Miner rule to
predict the endurance limit of
a material that has been
overstressed for a nite
number of cycles.
0.9Sut
72
4.8
Sf, 0
1
60
Sf, 1
4.7
So kpsi
Log S
n1 = 3 (10 3)
N1 = 8.52(10 3)
N1 n1 = 5.52(10 3)
4.6
Se ,0
40
38.6
Sf,2
Se ,1
n 2 = 0.648(106)
4.5
10 3
10 4
10 5
10 6
5
6
N
3
4
Log N
319
320
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
317
Figure 634 shows that the material has a life N1 = 8520 cycles at 60 kpsi, and consequently, after the application of σ1 for 3000 cycles, there are N1 n 1 = 5520
cycles of life remaining at σ1 . This locates the nite-life strength S f ,1 of the damaged
material, as shown in Fig. 634. To get a second point, we ask the question: With n 1
and N1 given, how many cycles of stress σ2 = Se,0 can be applied before the damaged
material fails?
This corresponds to n 2 cycles of stress reversal, and hence, from Eq. (658), we
have
n2
n1
+
=1
N1
N2
(a)
or
n2 = 1
n1
N1
(b)
N2
Then
n2 = 1
3(10)3
(106 ) = 0.648(106 ) cycles
8.52(10)3
This corresponds to the nite-life strength S f ,2 in Fig. 634. A line through S f ,1 and S f ,2
is the log Slog N diagram of the damaged material according to the Miner rule. The
new endurance limit is Se,1 = 38.6 kpsi.
We could leave it at this, but a little more investigation can be helpful. We have
two points on the new fatigue locus, N1 n 1 , σ1 and n 2 , σ2 . It is useful to prove that
the slope of the new line is still b. For the equation S f = a N b , where the values of a
and b are established by two points α and β . The equation for b is
b =
log σα /σβ
log Nα / Nβ
(c)
Examine the denominator of Eq. (c):
log
N1 n 1
N1 n 1
N1
Nα
= log
= log
= log
Nβ
n2
(1 n 1 / N1 ) N2
N2
= log
(σ1 /a )1/b
σ1
= log
1/b
(σ2 /a )
σ2
1/b
=
σ1
1
log
b
σ2
Substituting this into Eq. (c) with σα /σβ = σ1 /σ2 gives
b =
log(σ1 /σ2 )
=b
(1/b) log(σ1 /σ2 )
which means the damaged material line has the same slope as the virgin material line;
therefore, the lines are parallel. This information can be helpful in writing a computer
program for the Palmgren-Miner hypothesis.
Though the Miner rule is quite generally used, it fails in two ways to agree with
experiment. First, note that this theory states that the static strength Sut is damaged, that
is, decreased, because of the application of σ1 ; see Fig. 634 at N = 103 cycles.
Experiments fail to verify this prediction.
The Miner rule, as given by Eq. (658), does not account for the order in which the
stresses are applied, and hence ignores any stresses less than Se,0 . But it can be seen in
Fig. 634 that a stress σ3 in the range Se,1 < σ3 < Se,0 would cause damage if applied
after the endurance limit had been damaged by the application of σ1 .
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 635
4.9
Use of the Manson method to
predict the endurance limit of
a material that has been
overstressed for a nite
number of cycles.
0.9Sut
72
4.8
Sf, 0
1
60
Sf, 1
Log S
n1 = 3 (10 3)
4.7
So kpsi
318
II. Failure Prevention
N1 = 8.52(10 3)
N1 n1 = 5.52(10 3)
4.6
S'e ,0
40
S'e ,1
34.4
4.5
10 3
10 4
10 5
10 6
5
6
N
3
4
Log N
Mansons25 approach overcomes both of the deciencies noted for the PalmgrenMiner method; historically it is a much more recent approach, and it is just as easy to
use. Except for a slight change, we shall use and recommend the Manson method in this
book. Manson plotted the Slog N diagram instead of a log Slog N plot as is recommended here. Manson also resorted to experiment to nd the point of convergence of
the Slog N lines corresponding to the static strength, instead of arbitrarily selecting the
intersection of N = 103 cycles with S = 0.9 Sut as is done here. Of course, it is always
better to use experiment, but our purpose in this book has been to use the simple test
data to learn as much as possible about fatigue failure.
The method of Manson, as presented here, consists in having all log Slog N lines,
that is, lines for both the damaged and the virgin material, converge to the same point,
0.9 Sut at 103 cycles. In addition, the log Slog N lines must be constructed in the same
historical order in which the stresses occur.
The data from the preceding example are used for illustrative purposes. The results
are shown in Fig. 635. Note that the strength S f ,1 corresponding to N1 n 1 =
5.52(103 ) cycles is found in the same manner as before. Through this point and through
0.9 Sut at 103 cycles, draw the heavy dashed line to meet N = 106 cycles and dene the
endurance limit Se,1 of the damaged material. In this case the new endurance limit is
34.4 kpsi, somewhat less than that found by the Miner method.
It is now easy to see from Fig. 635 that a reversed stress σ = 36 kpsi, say, would
not harm the endurance limit of the virgin material, no matter how many cycles it might
be applied. However, if σ = 36 kpsi should be applied after the material was damaged
by σ1 = 60 kpsi, then additional damage would be done.
Both these rules involve a number of computations, which are repeated every time
damage is estimated. For complicated stress-time traces, this might be every cycle.
Clearly a computer program is useful to perform the tasks, including scanning the trace
and identifying the cycles.
25
S. S. Manson, A. J. Nachtigall, C. R. Ensign, and J. C. Fresche, Further Investigation of a Relation for
Cumulative Fatigue Damage in Bending, Trans. ASME, J. Eng. Ind., ser. B, vol. 87, No. 1, pp. 2535,
February 1965.
321
322
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
319
Collins said it well: In spite of all the problems cited, the Palmgren linear damage
rule is frequently used because of its simplicity and the experimental fact that other
more complex damage theories do not always yield a signicant improvement in failure prediction reliability.26
616
Surface Fatigue Strength
The surface fatigue mechanism is not denitively understood. The contact-affected
zone, in the absence of surface shearing tractions, entertains compressive principal
stresses. Rotary fatigue has its cracks grown at or near the surface in the presence of
tensile stresses that are associated with crack propagation, to catastrophic failure. There
are shear stresses in the zone, which are largest just below the surface. Cracks seem to
grow from this stratum until small pieces of material are expelled, leaving pits on the surface. Because engineers had to design durable machinery before the surface fatigue phenomenon was understood in detail, they had taken the posture of conducting tests,
observing pits on the surface, and declaring failure at an arbitrary projected area of hole,
and they related this to the Hertzian contact pressure. This compressive stress did
not produce the failure directly, but whatever the failure mechanism, whatever the
stress type that was instrumental in the failure, the contact stress was an index to its
magnitude.
Buckingham27 conducted a number of tests relating the fatigue at 108 cycles to
endurance strength (Hertzian contact pressure). While there is evidence of an endurance
limit at about 3(107 ) cycles for cast materials, hardened steel rollers showed no endurance
limit up to 4(108 ) cycles. Subsequent testing on hard steel shows no endurance limit.
Hardened steel exhibits such high fatigue strengths that its use in resisting surface fatigue
is widespread.
Our studies thus far have dealt with the failure of a machine element by yielding,
by fracture, and by fatigue. The endurance limit obtained by the rotating-beam test is
frequently called the exural endurance limit, because it is a test of a rotating beam. In
this section we shall study a property of mating materials called the surface endurance
shear. The design engineer must frequently solve problems in which two machine elements mate with one another by rolling, sliding, or a combination of rolling and sliding
contact. Obvious examples of such combinations are the mating teeth of a pair of gears,
a cam and follower, a wheel and rail, and a chain and sprocket. A knowledge of the surface strength of materials is necessary if the designer is to create machines having a
long and satisfactory life.
When two surfaces roll or roll and slide against one another with sufcient force,
a pitting failure will occur after a certain number of cycles of operation. Authorities are
not in complete agreement on the exact mechanism of the pitting; although the subject
is quite complicated, they do agree that the Hertz stresses, the number of cycles, the surface nish, the hardness, the degree of lubrication, and the temperature all inuence the
strength. In Sec. 319 it was learned that, when two surfaces are pressed together, a
maximum shear stress is developed slightly below the contacting surface. It is postulated
by some authorities that a surface fatigue failure is initiated by this maximum shear
stress and then is propagated rapidly to the surface. The lubricant then enters the crack
that is formed and, under pressure, eventually wedges the chip loose.
26
J. A. Collins, Failure of Materials in Mechanical Design, John Wiley & Sons, New York, 1981, p. 243.
27
Earle Buckingham, Analytical Mechanics of Gears, McGraw-Hill, New York, 1949.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
320
II. Failure Prevention
323
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
To determine the surface fatigue strength of mating materials, Buckingham designed
a simple machine for testing a pair of contacting rolling surfaces in connection with his
investigation of the wear of gear teeth. Buckingham and, later, Talbourdet gathered large
numbers of data from many tests so that considerable design information is now
available. To make the results useful for designers, Buckingham dened a load-stress
factor, also called a wear factor, which is derived from the Hertz equations. Equations
(373) and (374), pp. 118119, for contacting cylinders are found to be
2
2
2 F 1 ν1 / E 1 + 1 ν2 / E 2
πl
(1/d1 ) + (1/d2 )
b=
pmax =
2F
π bl
(659)
(660)
where b = half width of rectangular contact area
F = contact force
l = length of cylinders
ν = Poissons ratio
E = modulus of elasticity
d = cylinder diameter
It is more convenient to use the cylinder radius, so let 2r = d . If we then designate
the length of the cylinders as w (for width of gear, bearing, cam, etc.) instead of l and
remove the square root sign, Eq. (659) becomes
b2 =
2
2
4 F 1 ν1 / E 1 + 1 ν2 / E 2
πw
1/ r1 + 1/ r2
(661)
We can dene a surface endurance strength SC using
pmax =
2F
π bw
(662)
as
SC =
2F
π bw
(663)
which may also be called contact strength, the contact fatigue strength, or the Hertzian
endurance strength. The strength is the contacting pressure which, after a specied
number of cycles, will cause failure of the surface. Such failures are often called wear
because they occur over a very long time. They should not be confused with abrasive
wear, however. By squaring Eq. (663), substituting b2 from Eq. (661), and rearranging, we obtain
F
w
1
1
+
r1 r2
2
= π SC
2
2
1 ν2
1 ν1
+
= K1
E1
E2
(664)
The left expression consists of parameters a designer may seek to control independently.
The central expression consists of material properties that come with the material and
condition specication. The third expression is the parameter K 1 , Buckinghams loadstress factor, determined by a test xture with values F, w , r1 , r2 and the number of
324
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
321
cycles associated with the rst tangible evidence of fatigue. In gear studies a similar
K factor is used:
K1
sin φ
Kg =
(665)
4
2
2
where φ is the tooth pressure angle, and the term [(1 ν1 )/ E 1 + (1 ν2 )/ E 2 ] is
dened as 1/(π C 2 ), so that
P
SC = C P
F
w
1
1
+
r1 r2
(666)
Buckingham and others reported K 1 for 108 cycles and nothing else. This gives only one
point on the SC N curve. For cast metals this may be sufcient, but for wrought steels, heattreated, some idea of the slope is useful in meeting design goals of other than 108 cycles.
Experiments show that K 1 versus N, K g versus N, and SC versus N data are rectied by loglog transformation. This suggests that
Kg = a N b
K 1 = α1 N β1
SC = α N β
The three exponents are given by
β1 =
log( K 1 / K 2 )
log( N1 / N2 )
b=
log( K g1 / K g2 )
log( N1 / N2 )
β=
log( SC 1 / SC 2 )
log( N1 / N2 )
(667)
Data on induction-hardened steel on steel give ( SC )107 = 271 kpsi and ( SC )108 =
239 kpsi, so β , from Eq. (667), is
β=
log(271/239)
= 0.055
log(107 /108 )
It may be of interest that the American Gear Manufacturers Association (AGMA) uses
β
0.056 between 104 < N < 1010 if the designer has no data to the contrary
beyond 107 cycles.
A longstanding correlation in steels between SC and HB at 108 cycles is
( SC )108 =
0.4 HB 10 kpsi
2.76 HB 70 MPa
(668)
0.99 ( SC )107
= 0.327 HB + 26 kpsi
(669)
AGMA uses
Equation (666) can be used in design to nd an allowable surface stress by using
a design factor. Since this equation is nonlinear in its stress-load transformation, the
designer must decide if loss of function denotes inability to carry the load. If so, then
to nd the allowable stress, one divides the load F by the design factor n d :
σC = C P
F
wn d
1
1
+
r1 r2
CP
=
nd
F
w
1
1
+
r1 r2
SC
=
nd
and n d = ( SC /σC )2 . If the loss of function is focused on stress, then n d = SC /σC . It is
recommended that an engineer
Decide whether loss of function is failure to carry load or stress.
Dene the design factor and factor of safety accordingly.
Announce what he or she is using and why.
Be prepared to defend his or her position.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
322
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
In this way everyone who is party to the communication knows what a design factor
(or factor of safety) of 2 means and adjusts, if necessary, the judgmental perspective.
617
Stochastic Analysis28
As already demonstrated in this chapter, there are a great many factors to consider in
a fatigue analysis, much more so than in a static analysis. So far, each factor has been
treated in a deterministic manner, and if not obvious, these factors are subject to variability and control the overall reliability of the results. When reliability is important,
then fatigue testing must certainly be undertaken. There is no other way. Consequently,
the methods of stochastic analysis presented here and in other sections of this book
constitute guidelines that enable the designer to obtain a good understanding of the
various issues involved and help in the development of a safe and reliable design.
In this section, key stochastic modications to the deterministic features and equations described in earlier sections are provided in the same order of presentation.
Endurance Limit
To begin, a method for estimating endurance limits, the tensile strength correlation
¯
method, is presented. The ratio = Se / Sut is called the fatigue ratio.29 For ferrous
metals, most of which exhibit an endurance limit, the endurance limit is used as a
numerator. For materials that do not show an endurance limit, an endurance strength at
a specied number of cycles to failure is used and noted. Gough30 reported the stochastic nature of the fatigue ratio for several classes of metals, and this is shown in
Fig. 636. The rst item to note is that the coefcient of variation is of the order 0.10
to 0.15, and the distribution varies for classes of metals. The second item to note is that
Goughs data include materials of no interest to engineers. In the absence of testing,
engineers use the correlation that represents to estimate the endurance limit Se from
¯
the mean ultimate strength Sut .
Goughs data are for ensembles of metals, some chosen for metallurgical interest,
and include materials that are not commonly selected for machine parts. Mischke31
analyzed data for 133 common steels and treatments in varying diameters in rotating
bending,32 and the result was
= 0.445d 0.107 LN(1, 0.138)
where d is the specimen diameter in inches and LN(1, 0.138) is a unit lognormal variate with a mean of 1 and a standard deviation (and coefcient of variation) of 0.138. For
the standard R. R. Moore specimen,
0.30
= 0.445(0.30)0.107 LN(1, 0.138) = 0.506LN(1, 0.138)
28
Review Chap. 20 before reading this section.
29
From this point, since we will be dealing with statistical distributions in terms of means, standard
¯
deviations, etc. A key quantity, the ultimate strength, will here be presented by its mean value, Sut . This
means that certain terms that were dened earlier in terms of the minimum value of Sut will change slightly.
30
In J. A. Pope, Metal Fatigue, Chapman and Hall, London, 1959.
31
Charles R. Mischke, Prediction of Stochastic Endurance Strength, Trans. ASME, Journal of Vibration,
Acoustics, Stress, and Reliability in Design, vol. 109, no. 1, January 1987, pp. 113122.
32
Data from H. J. Grover, S. A. Gordon, and L. R. Jackson, Fatigue of Metals and Structures, Bureau of
Naval Weapons, Document NAVWEPS 00-2500435, 1960.
325
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Figure 636
The lognormal probability
density PDF of the fatigue ratio
φb of Gough.
4
5
Class
All metals
Nonferrous
Iron and carbon steels
Low alloy steels
Special alloy steels
1
2
3
4
5
3
Probability density
326
No.
380
152
111
78
39
2
1
0
323
0.3
5
0.4
0.5
0.6
Rotary bending fatigue ratio
0.7
b
Also, 25 plain carbon and low-alloy steels with Sut > 212 kpsi are described by
Se = 107LN(1, 0.139) kpsi
In summary, for the rotating-beam specimen,
¯
0.506 Sut LN(1, 0.138) kpsi or MPa
Se = 107LN(1, 0.139) kpsi
740LN(1, 0.139) MPa
¯
Sut 212 kpsi (1460 MPa)
¯
Sut > 212 kpsi
(670)
¯ut > 1460 MPa
S
¯
where Sut is the mean ultimate tensile strength.
Equations (670) represent the state of information before an engineer has chosen
a material. In choosing, the designer has made a random choice from the ensemble of
possibilities, and the statistics can give the odds of disappointment. If the testing is lim¯
ited to nding an estimate of the ultimate tensile strength mean Sut with the chosen
material, Eqs. (670) are directly helpful. If there is to be rotary-beam fatigue testing,
then statistical information on the endurance limit is gathered and there is no need for
the correlation above.
¯
Table 69 compares approximate mean values of the fatigue ratio φ0.30 for several
classes of ferrous materials.
Endurance Limit Modifying Factors
A Marin equation can be written as
Se = ka kb kc kd kf Se
(671)
where the size factor kb is deterministic and remains unchanged from that given in
Sec. 69. Also, since we are performing a stochastic analysis, the reliability factor ke
is unnecessary here.
The surface factor ka cited earlier in deterministic form as Eq. (620), p. 280, is
now given in stochastic form by
¯b
ka = a Sut LN(1, C )
¯
( Sut in kpsi or MPa)
where Table 610 gives values of a, b, and C for various surface conditions.
(672)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
324
II. Failure Prevention
327
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Table 69
Material Class
Wrought steels
0.50
Cast steels
0.40
Powdered steels
0.38
Gray cast iron
0.35
Malleable cast iron
0.40
Normalized nodular cast iron
Comparison of
Approximate Values of
Mean Fatigue Ratio for
Some Classes of Metals
φ 0.30
0.33
Table 610
ka
Parameters in Marin
Surface Condition
Factor
Surface
Finish
b
aSut LN(1, C )
a
kpsi
Ground
1.34
Machined or Cold-rolled
2.67
Hot-rolled
14.5
As-forged
39.8
MPa
Solution
0.086
1.58
0.265
4.45
0.719
58.1
0.995
271
*Due to the wide scatter in ground surface data, an alternate function is ka
EXAMPLE 616
b
Coefcient of
Variation, C
0.120
0.058
0.110
0.145
0.878LN(1, 0.120). Note: Sut in kpsi or MPa.
A steel has a mean ultimate strength of 520 MPa and a machined surface. Estimate ka .
From Table 610,
ka = 4.45(520)0.265 LN(1, 0.058)
¯
ka = 4.45(520)0.265 (1) = 0.848
Answer
¯
σka = C ka = (0.058)4.45(520)0.265 = 0.049
ˆ
so ka = LN(0.848, 0.049) .
The load factor kc for axial and torsional loading is given by
¯
(kc )axial = 1.23 Sut0.0778 LN(1, 0.125)
¯0
(kc )torsion = 0.328 Sut.125 LN(1, 0.125)
(673)
(674)
¯
where Sut is in kpsi. There are fewer data to study for axial fatigue. Equation (673) was
deduced from the data of Landgraf and of Grover, Gordon, and Jackson (as cited earlier).
Torsional data are sparser, and Eq. (674) is deduced from data in Grover et al.
Notice the mild sensitivity to strength in the axial and torsional load factor, so kc in
these cases is not constant. Average values are shown in the last column of Table 611,
and as footnotes to Tables 612 and 613. Table 614 shows the inuence of material
classes on the load factor kc . Distortion energy theory predicts (kc )torsion = 0.577 for
materials to which the distortion-energy theory applies. For bending, kc = LN(1, 0) .
328
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
Table 611
Parameters in Marin
Loading Factor
kc
Mode of
Loading
MPa
Bending
1
1
Axial
1.23
1.43
Torsion
Table 612
0.328
0.258
¯
Sut ,
kpsi
C
Average
kc
0
1
0.125
0.85
0.125
0.59
50
0.125
0.860
0.832
200
0
0.0778
0.907
150
β
k*
c
100
Average Marin Loading
Factor for Axial Load
β
αSut LN(1, C)
α
kpsi
0.814
*Average entry 0.85.
Table 613
Average Marin Loading
Factor for Torsional Load
¯
Sut ,
kpsi
k*
c
50
0.535
100
0.583
150
0.614
200
0.636
*Average entry 0.59.
Table 614
Average Marin Torsional
Loading Factor kc for
Several Materials
Range
n
¯c
k
ˆkc
σ
Wrought steels
0.520.69
31
0.60
0.03
Wrought Al
0.430.74
13
0.55
0.09
Wrought Cu and alloy
0.410.67
7
0.56
0.10
Wrought Mg and alloy
0.490.60
2
0.54
0.08
Material
325
Titanium
0.370.57
3
0.48
0.12
Cast iron
0.791.01
9
0.90
0.07
Cast Al, Mg, and alloy
0.710.91
5
0.85
0.09
Source: The table is an extension of P. G. Forrest, Fatigue of Metals, Pergamon Press, London, 1962, Table 17, p. 110,
with standard deviations estimated from range and sample size using Table A1 in J. B. Kennedy and A. M. Neville,
Basic Statistical Methods for Engineers and Scientists, 3rd ed., Harper & Row, New York, 1986, pp. 5455.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
326
II. Failure Prevention
329
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
EXAMPLE 617
Solution
Estimate the Marin loading factor kc for a 1in-diameter bar that is used as follows.
(a) In bending. It is made of steel with Sut = 100LN(1, 0.035) kpsi, and the designer
¯
intends to use the correlation Se = 0.30 Sut to predict Se .
(b) In bending, but endurance testing gave Se = 55LN(1, 0.081) kpsi.
(c) In push-pull (axial) fatigue, Sut = LN(86.2, 3.92) kpsi, and the designer intended to
¯
use the correlation Se = 0.30 Sut .
(d) In torsional fatigue. The material is cast iron, and Se is known by test.
(a) Since the bar is in bending,
Answer
kc = (1, 0)
(b) Since the test is in bending and use is in bending,
Answer
kc = (1, 0)
(c) From Eq. (673),
Answer
(kc )ax = 1.23(86.2)0.0778 LN(1, 0.125)
¯
kc = 1.23(86.2)0.0778 (1) = 0.870
¯
σkc = C kc = 0.125(0.870) = 0.109
ˆ
¯
ˆ
(d) From Table 615, kc = 0.90, σkc = 0.07, and
Answer
Ckc =
0.07
= 0.08
0.90
The temperature factor kd is
¯
kd = kd LN(1, 0.11)
(675)
¯
where kd = kd , given by Eq. (627), p. 283.
Finally, kf is, as before, the miscellaneous factor that can come about from a great
many considerations, as discussed in Sec. 69, where now statistical distributions, possibly from testing, are considered.
Stress Concentration and Notch Sensitivity
Notch sensitivity q was dened by Eq. (631), p. 287. The stochastic equivalent is
q=
Kf 1
Kt 1
(676)
where K t is the theoretical (or geometric) stress-concentration factor, a deterministic
quantity. A study of lines 3 and 4 of Table 206, will reveal that adding a scalar to (or
subtracting one from) a variate x will affect only the mean. Also, multiplying (or dividing) by a scalar affects both the mean and standard deviation. With this in mind, we can
330
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
a( in) ,
a( mm) ,
Notch Type
Sut in kpsi
Sut in MPa
Coefcient of
Variation CKf
Transverse hole
5/Sut
174/Sut
0.10
Shoulder
4/Sut
139/Sut
0.11
Groove
3/Sut
104/Sut
0.15
Table 615
Heywoods Parameter
a and coefcients of
variation CKf for steels
327
relate the statistical parameters of the fatigue stress-concentration factor K f to those of
notch sensitivity q. It follows that
q = LN
¯
¯
Kf 1 CKf
,
Kt 1 Kt 1
where C = C K f and
q=
¯
σq =
ˆ
Cq =
¯
Kf 1
Kt 1
¯
CKf
Kt 1
(677)
¯
CKf
¯
Kf 1
The fatigue stress-concentration factor K f has been investigated more in England than in
¯
the United States. For K f , consider a modied Neuber equation (after Heywood33 ),
where the fatigue stress-concentration factor is given by
Kt
(678)
2( K t 1) a
1+
Kt
r
where Table 615 gives values of a and C K f for steels with transverse holes,
shoulders, or grooves. Once K f is described, q can also be quantified using the set
Eqs. (677).
The modied Neuber equation gives the fatigue stress concentration factor as
¯
Kf =
¯
K f = K f LN 1, C K f
(679)
33
R. B. Heywood, Designing Against Fatigue, Chapman & Hall, London, 1962.
EXAMPLE 618
Solution
Estimate K f and q for the steel shaft given in Ex. 66, p. 288.
From Ex. 66, a steel shaft with Sut = 690 Mpa and a shoulder with a llet of 3 mm
.
was found to have a theoretical stress-concentration-factor of K t = 1.65. From Table
615,
139
139
= 0.2014 mm
a=
=
Sut
690
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
328
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
From Eq. (678),
Kf =
Answer
1.65
Kt
= 1.51
=
2(1.65 1) 0.2014
2( K t 1) a
1+
1+
1.65
Kt
3
r
which is 2.5 percent lower than what was found in Ex. 66.
From Table 615, C K f = 0.11. Thus from Eq. (679),
K f = 1.51 LN(1, 0.11)
From Eq. (677), with K t = 1.65
q=
¯
Cq =
1.51 1
= 0.785
1.65 1
¯
CK f K f
0.11(1.51)
=
= 0.326
¯
1.51 1
Kf 1
σq = Cq q = 0.326(0.785) = 0.256
ˆ
¯
So,
Answer
EXAMPLE 619
Solution
q = LN(0.785, 0.256)
The bar shown in Fig. 637 is machined from a cold-rolled at having an ultimate
strength of Sut = LN(87.6, 5.74) kpsi. The axial load shown is completely reversed.
The load amplitude is Fa = LN(1000, 120) lbf.
(a) Estimate the reliability.
(b) Reestimate the reliability when a rotating bending endurance test shows that Se =
LN(40, 2) kpsi.
¯
(a) From Eq. (670), Se = 0.506 Sut LN(1, 0.138) = 0.506(87.6)LN(1, 0.138)
= 44.3LN(1, 0.138) kpsi
From Eq. (672) and Table 610,
¯
ka = 2.67 Sut0.265 LN(1, 0.058) = 2.67(87.6)0.265 LN(1, 0.058)
= 0.816LN(1, 0.058)
kb = 1
(axial loading)
3
16
Figure 637
1000 lbf
2 1 in
4
in R.
1000 lbf
1 1 in
2
1
4
in
3
4
in D.
331
332
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
329
From Eq. (673),
¯
kc = 1.23 Sut0.0778 LN(1, 0.125) = 1.23(87.6)0.0778 LN(1, 0.125)
= 0.869LN(1, 0.125)
kd = k f = (1, 0)
The endurance strength, from Eq. (671), is
Se = ka kb kc kd k f Se
Se = 0.816LN(1, 0.058)(1)0.869LN(1, 0.125)(1)(1)44.3LN(1, 0.138)
The parameters of Se are
¯
Se = 0.816(0.869)44.3 = 31.4 kpsi
C Se = (0.0582 + 0.1252 + 0.1382 )1/2 = 0.195
so Se = 31.4LN(1, 0.195) kpsi.
In computing the stress, the section at the hole governs. Using the terminology
.
of Table A151 we nd d /w = 0.50, therefore K t = 2.18. From Table 615,
a = 5/ Sut = 5/87.6 = 0.0571 and Ck f = 0.10. From Eqs. (678) and (679) with
r = 0.375 in,
2.18
Kt
LN(1, 0.10)
LN 1, C K f =
2(2.18 1) 0.0571
2( K t 1) a
1+
1+
2.18
0.375
Kt
r
= 1.98LN(1, 0.10)
Kf =
The stress at the hole is
= Kf
1000LN(1, 0.12)
F
= 1.98LN(1, 0.10)
A
0.25(0.75)
σ = 1.98
¯
1000
103 = 10.56 kpsi
0.25(0.75)
Cσ = (0.102 + 0.122 )1/2 = 0.156
so stress can be expressed as = 10.56LN(1, 0.156) kpsi.34
The endurance limit is considerably greater than the load-induced stress, indicating that nite life is not a problem. For interfering lognormal-lognormal distributions,
Eq. (543), p. 242, gives
ln
z=
¯
Se
σ
¯
2
1 + Cσ
2
1 + C Se
2
ln 1 + C Se
2
1 + Cσ
=
ln
31.4
10.56
ln[(1 + 0.1952 )(1 + 0.1562 )]
From Table A10 the probability of failure p f =
reliability is
Answer
1 + 0.1562
1 + 0.1952
= 4.37
(4.37) = .000 006 35, and the
R = 1 0.000 006 35 = 0.999 993 65
34
Note that there is a simplication here. The area is not a deterministic quantity. It will have a statistical
distribution also. However no information was given here, and so it was treated as being deterministic.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
330
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
(b) The rotary endurance tests are described by Se = 40LN(1, 0.05) kpsi whose mean
¯
is less than the predicted mean in part a. The mean endurance strength Se is
¯
Se = 0.816(0.869)40 = 28.4 kpsi
C Se = (0.0582 + 0.1252 + 0.052 )1/2 = 0.147
so the endurance strength can be expressed as Se = 28.3LN(1, 0.147) kpsi. From
Eq. (543),
28.4 1 + 0.1562
ln
10.56 1 + 0.1472
z=
= 4.65
ln[(1 + 0.1472 )(1 + 0.1562 )]
Using Table A10, we see the probability of failure p f =
and
(4.65) = 0.000 001 71,
R = 1 0.000 001 71 = 0.999 998 29
an increase! The reduction in the probability of failure is (0.000 001 71 0.000
006 35)/0.000 006 35 = 0.73, a reduction of 73 percent. We are analyzing an existing
¯¯
¯
design, so in part (a) the factor of safety was n = S /σ = 31.4/10.56 = 2.97. In part (b)
n = 28.4/ 10.56 = 2.69, a decrease. This example gives you the opportunity to see the role
¯
¯
¯
of the design factor. Given knowledge of S, C S, σ , Cσ , and reliability (through z), the mean
¯
¯
factor of safety (as a design factor) separates S and σ so that the reliability goal is achieved.
¯
¯
Knowing n alone says nothing about the probability of failure. Looking at n = 2.97 and
n = 2.69 says nothing about the respective probabilities of failure. The tests did not reduce
¯
¯
Se signicantly, but reduced the variation C S such that the reliability was increased.
¯¯
When a mean design factor (or mean factor of safety) defined as Se /σ is said to
be silent on matters of frequency of failures, it means that a scalar factor of safety
by itself does not offer any information about probability of failure. Nevertheless,
some engineers let the factor of safety speak up, and they can be wrong in their
conclusions.
As revealing as Ex. 619 is concerning the meaning (and lack of meaning) of a
design factor or factor of safety, let us remember that the rotary testing associated with
part (b) changed nothing about the part, but only our knowledge about the part. The
mean endurance limit was 40 kpsi all the time, and our adequacy assessment had to
move with what was known.
Fluctuating Stresses
Deterministic failure curves that lie among the data are candidates for regression models. Included among these are the Gerber and ASME-elliptic for ductile materials, and,
for brittle materials, Smith-Dolan models, which use mean values in their presentation.
Just as the deterministic failure curves are located by endurance strength and ultimate
tensile (or yield) strength, so too are stochastic failure curves located by Se and by Sut
or S y . Figure 632, p. 312, shows a parabolic Gerber mean curve. We also need to
establish a contour located one standard deviation from the mean. Since stochastic
333
334
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
331
curves are most likely to be used with a radial load line we will use the equation given
in Table 67, p. 299, expressed in terms of the strength means as
¯2
¯2
r 2 Sut
2 Se
¯
1 + 1 +
Sa =
(680)
¯
¯
2 Se
r Sut
¯¯
¯
Because of the positive correlation between Se and Sut , we increment Se by C Se Se , Sut
¯
¯
¯
by C Sut Sut , and Sa by C Sa Sa , substitute into Eq. (680), and solve for C Sa to obtain
2
¯
2 Se (1 + C Se )
1 + 1 +
¯
r Sut (1 + C Sut )
(1 + C Sut )2
1
C Sa =
(681)
1 + C Se
¯2
2 Se
1 + 1 +
¯
r Sut
Equation (681) can be viewed as an interpolation formula for C Sa , which falls between
¯
C Se and C Sut depending on load line slope r. Note that Sa = Sa LN(1, C Sa ) .
Similarly, the ASME-elliptic criterion of Table 68, p. 300, expressed in terms of
its means is
¯
Sa =
¯¯
r Sy Se
(682)
¯2 ¯2
r 2 Sy + Se
¯¯
¯
¯
¯
¯
Similarly, we increment Se by C Se Se , Sy by C Sy Sy , and Sa by C Sa Sa , substitute into
Eq. (682), and solve for C Sa :
C Sa = (1 + C Sy )(1 + C Se )
¯2 ¯2
r 2 Sy + Se
1
¯2
¯2
r 2 Sy (1 + C Sy )2 + Se (1 + C Se )2
(683)
Many brittle materials follow a Smith-Dolan failure criterion, written deterministically as
n σa
1 n σm / Sut
=
Se
1 + n σm / Sut
(684)
¯¯
¯
1 Sm / Sut
Sa
=
¯
¯¯
Se
1 + Sm / Sut
(685)
Expressed in terms of its means,
¯
¯
¯
For a radial load line slope of r, we substitute Sa / r for Sm and solve for Sa , obtaining
¯
¯¯
¯
r Sut + Se
4r Sut Se
¯
1 + 1 +
Sa =
(686)
¯
¯
2
(r Sut + Se )2
and the expression for C Sa is
¯
¯
r Sut (1 + C Sut ) + Se (1 + C Se )
C Sa =
¯
2 Sa
· 1 +
1+
¯¯
4r Sut Se (1 + C Se )(1 + C Sut )
¯
¯
[r Sut (1 + C Sut ) + Se (1 + C Se )]2
(687)
1
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
332
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
EXAMPLE 620
Solution
A rotating shaft experiences a steady torque T = 1360LN(1, 0.05) lbf · in, and at a
shoulder with a 1.1-in small diameter, a fatigue stress-concentration factor K f =
1.50LN(1, 0.11) , K f s = 1.28LN(1, 0.11) , and at that location a bending moment of
M = 1260LN(1, 0.05) lbf · in. The material of which the shaft is machined is hot-rolled
1035 with Sut = 86.2LN(1, 0.045) kpsi and S y = 56.0LN(1, 0.077) kpsi. Estimate the
reliability using a stochastic Gerber failure zone.
Establish the endurance strength. From Eqs. (670) to (672) and Eq. (620), p. 280,
Se = 0.506(86.2)LN(1, 0.138) = 43.6LN(1, 0.138) kpsi
ka = 2.67(86.2)0.265 LN(1, 0.058) = 0.820LN(1, 0.058)
kb = (1.1/0.30)0.107 = 0.870
kc = kd = k f = LN(1, 0)
Se = 0.820LN(1, 0.058)0.870(43.6)LN(1, 0.138)
¯
Se = 0.820(0.870)43.6 = 31.1 kpsi
C Se = (0.0582 + 0.1382 )1/2 = 0.150
and so Se = 31.1LN(1, 0.150) kpsi.
Stress (in kpsi):
σa =
32K f Ma
32(1.50)LN(1, 0.11)1.26LN(1, 0.05)
=
π d3
π(1.1)3
σa =
¯
32(1.50)1.26
= 14.5 kpsi
π(1.1)3
Cσ a = (0.112 + 0.052 )1/2 = 0.121
m
=
τm =
¯
16K f s Tm
16(1.28)LN(1, 0.11)1.36LN(1, 0.05)
=
3
πd
π(1.1)3
16(1.28)1.36
= 6.66 kpsi
π(1.1)3
Cτ m = (0.112 + 0.052 )1/2 = 0.121
σa = σa + 3τa
¯
¯2
¯2
1/2
= [14.52 + 3(0)2 ]1/2 = 14.5 kpsi
σm = σm + 3τm
¯
¯2
¯2
1/2
= [0 + 3(6.66)2 ]1/2 = 11.54 kpsi
r=
σa
¯
14.5
=
= 1.26
σm
¯
11.54
Strength: From Eqs. (680) and (681),
1.26 86.2
¯
1 +
Sa =
2(31.1)
2
2
1+
2(31.1)
1.26(86.2)
2
= 28.9 kpsi
335
336
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
C Sa
1 +
(1 + 0.045)2
=
1 + 0.150
1+
1 +
2(31.1)(1 + 0.15)
1.26(86.2)(1 + 0.045)
2(31.1)
1+
1.26(86.2)
2
333
2
1 = 0.134
Reliability: Since Sa = 28.9LN(1, 0.134) kpsi and a = 14.5LN(1, 0.121) kpsi,
Eq. (544), p. 242, gives
2
28.9 1 + 0.1212
¯a 1 + Cσa
S
ln
ln
2
14.5 1 + 0.1342
σa 1 + C Sa
¯
z=
=
= 3.83
ln[(1 + 0.1342 )(1 + 0.1212 )]
ln 1 + C 2 1 + C 2
Sa
σa
From Table A10 the probability of failure is p f = 0.000 065, and the reliability is,
against fatigue,
Answer
R = 1 p f = 1 0.000 065 = 0.999 935
The chance of rst-cycle yielding is estimated by interfering S y with max . The
¯ ¯
quantity max is formed from a + m . The mean of max is σa + σm = 14.5 +
11.54 = 26.04 kpsi. The coefcient of variation of the sum is 0.121, since both
COVs are 0.121, thus Cσ max = 0.121. We interfere S y = 56LN(1, 0.077) kpsi with
max = 26.04LN (1, 0.121) kpsi. The corresponding z variable is
1 + 0.1212
56
ln
26.04 1 + 0.0772
= 5.39
z=
ln[(1 + 0.0772 )(1 + 0.1212 )]
which represents, from Table A10, a probability of failure of approximately 0.07 358
[which represents 3.58(108 )] of rst-cycle yield in the llet.
The probability of observing a fatigue failure exceeds the probability of a yield
failure, something a deterministic analysis does not foresee and in fact could lead one
to expect a yield failure should a failure occur. Look at the a Sa interference and the
max S y interference and examine the z expressions. These control the relative probabilities. A deterministic analysis is oblivious to this and can mislead. Check your statistics text for events that are not mutually exclusive, but are independent, to quantify
the probability of failure:
p f = p(yield) + p(fatigue) p(yield and fatigue)
= p(yield) + p(fatigue) p(yield) p(fatigue)
= 0.358(107 ) + 0.65(104 ) 0.358(107 )0.65(104 ) = 0.650(104 )
R = 1 0.650(104 ) = 0.999 935
against either or both modes of failure.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
334
II. Failure Prevention
337
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
Figure 638
ea
M
Designers fatigue diagram
for Ex. 620.
50
n
La
ng
cu
rv
e
Amplitude stress component
a,
kpsi
er
40
Load line
30
1 S ig ma c
urve
Sa
Mea
nG
+1
erbe
r cu
_
Sa
Sig
ma
rve
cur
ve
20
a
_
a
10
0
0
10
20
30
40
50
Steady stress component
60
m,
70
80
90
kpsi
Examine Fig. 638, which depicts the results of Ex. 620. The problem distribution
of Se was compounded of historical experience with Se and the uncertainty manifestations
due to features requiring Marin considerations. The Gerber failure zone displays this.
The interference with load-induced stress predicts the risk of failure. If additional information is known (R. R. Moore testing, with or without Marin features), the stochastic
Gerber can accommodate to the information. Usually, the accommodation to additional
test information is movement and contraction of the failure zone. In its own way the stochastic failure model accomplishes more precisely what the deterministic models and
conservative postures intend. Additionally, stochastic models can estimate the probability
of failure, something a deterministic approach cannot address.
The Design Factor in Fatigue
The designer, in envisioning how to execute the geometry of a part subject to the imposed
constraints, can begin making a priori decisions without realizing the impact on the
design task. Now is the time to note how these things are related to the reliability goal.
The mean value of the design factor is given by Eq. (545), repeated here as
2
2.
n = exp z ln 1 + Cn + ln 1 + Cn = exp[Cn (z + Cn /2)]
¯
(688)
in which, from Table 206 for the quotient n = S/ ,
Cn =
2
2
C S + Cσ
2
1 + Cσ
where C S is the COV of the signicant strength and Cσ is the COV of the signicant
¯
stress at the critical location. Note that n is a function of the reliability goal (through
z) and the COVs of the strength and stress. There are no means present, just measures
of variability. The nature of C S in a fatigue situation may be C Se for fully reversed
loading, or C Sa otherwise. Also, experience shows C Se > C Sa > C Sut , so C Se can be
used as a conservative estimate of C Sa . If the loading is bending or axial, the form of
338
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
a
335
might be
a
= Kf
Ma c
I
respectively. This makes the COV of
a
or
a,
= Kf
F
A
namely Cσa , expressible as
2
2
Cσa = C K f + C F
1/2
again a function of variabilities. The COV of Se , namely C Se , is
2
2
2
2
2
C Se = Cka + Ckc + Ckd + Ck f + C Se
1/2
again, a function of variabilities. An example will be useful.
EXAMPLE 621
Solution
A strap to be made from a cold-drawn steel strip workpiece is to carry a fully reversed
axial load F = LN(1000, 120) lbf as shown in Fig. 639. Consideration of adjacent
parts established the geometry as shown in the gure, except for the thickness t. Make a
decision as to the magnitude of the design factor if the reliability goal is to be 0.999 95,
then make a decision as to the workpiece thickness t.
Let us take each a priori decision and note the consequence:
A Priori Decision
Use 1018 CD steel
Consequence
¯
Sut
87.6kpsi, CSut
0.0655
Function:
Carry axial load
Fa = 1000 lbf
R 0.999 95
CF
z
in D. drill
Cka
Hole critical
3
8
Machined surfaces
CKf
Ambient temperature Ckd
CS e
Hole drilled
3
4
Correlation method
CSe
Cn
in
n
¯
Fa = 1000 lbf
0.12, Ckc
3.891
0.058
0.10, C a
0
0.138
0.125
(0.102
0.122)1/2 = 0.156
(0.0582 + 0.1252 + 0.1382 ) 1/2 = 0.195
2
2
CSe + Cσ
a
1+
2
Cσ
a
=
0.1952 + 0.1562
= 0.2467
1 + 0.1562
exp ( 3.891) ln(1 + 0.24672 ) + ln 1 + 0.24672
= 2.65
Figure 639
A strap with a thickness t is
subjected to a fully reversed
axial load of 1000 lbf.
Example 621 considers the
thickness necessary to attain a
reliability of 0.999 95 against
a fatigue failure.
¯
These eight a priori decisions have quantied the mean design factor as n = 2.65.
Proceeding deterministically hereafter we write
¯
¯
Se
F
¯
= Kf
σa =
n
¯
(w d )t
from which
¯ ¯¯
Kf nF
t=
(1)
¯
(w d ) Se
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
336
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
339
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
¯
¯
To evaluate the preceding equation we need Se and K f . The Marin factors are
¯
ka = 2.67 Sut0.265 LN(1, 0.058) = 2.67(87.6)0.265 LN(1, 0.058)
¯
ka = 0.816
kb = 1
¯
kc = 1.23 Sut0.078 LN(1, 0.125) = 0.868LN(1, 0.125)
¯
kc = 0.868
¯
¯
kd = k f = 1
and the endurance strength is
¯
Se = 0.816(1)(0.868)(1)(1)0.506(87.6) = 31.4 kpsi
The hole governs. From Table A151 we nd d /w = 0.50, therefore K t = 2.18. From
¯
Table 615 a = 5/ Sut = 5/87.6 = 0.0571, r = 0.1875 in. From Eq. (678) the
fatigue stress concentration factor is
2.18
¯
= 1.91
Kf =
2(2.18 1) 0.0571
1+
2.18
0.1875
The thickness t can now be determined from Eq. (1)
¯ ¯¯
Kf nF
1.91(2.65)1000
t
= 0.430 in
=
(w d ) Se
(0.75 0.375)31 400
Use 1 -in-thick strap for the workpiece. The 1 -in thickness attains and, in the rounding
2
2
to available nominal size, exceeds the reliability goal.
The example demonstrates that, for a given reliability goal, the fatigue design factor
that facilitates its attainment is decided by the variabilities of the situation. Furthermore,
the necessary design factor is not a constant independent of the way the concept unfolds.
Rather, it is a function of a number of seemingly unrelated a priori decisions that are made
in giving denition to the concept. The involvement of stochastic methodology can be
limited to dening the necessary design factor. In particular, in the example, the design
factor is not a function of the design variable t; rather, t follows from the design factor.
618
Road Maps and Important Design Equations
for the Stress-Life Method
As stated in Sec. 615, there are three categories of fatigue problems. The important
procedures and equations for deterministic stress-life problems are presented here.
Completely Reversing Simple Loading
1 Determine Se either from test data or
p. 274
0.5 Sut
Se = 100 kpsi
700 MPa
Sut 200 kpsi (1400 MPa)
Sut > 200 kpsi
Sut > 1400 MPa
(68)
340
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
337
2 Modify Se to determine Se .
Se = ka kb kc kd ke k f Se
p. 279
ka =
Table 62
Parameters for Marin
Surface Modication
Factor, Eq. (619)
(619)
Factor a
Surface
Finish
Sut, kpsi
Sut, MPa
Ground
1.34
1.58
Machined or cold-drawn
2.70
4.51
Hot-rolled
14.4
As-forged
(618)
b
aSut
39.9
57.7
0.085
0.265
0.718
272.
Rotating shaft. For bending or torsion,
( d /0.3) 0.107 = 0.879d 0.107
0.91d 0.157
kb =
p. 280
( d /7.62) 0.107 = 1.24d 0.107
1.51d 0.157
Exponent
b
0.995
0.11 d 2 in
2 < d 10 in
2.79 d 51 mm
51 < 254 mm
(620)
For axial,
(621)
kb = 1
Nonrotating member. Use Table 63, p. 282, for de and substitute into Eq. (620)
for d.
bending
1
kc = 0.85 axial
p. 282
(626)
0.59 torsion
p. 283 Use Table 64 for kd, or
2
kd = 0.975 + 0.432(103 ) TF 0.115(105 ) TF
3
4
+ 0.104(108 ) TF 0.595(1012 ) TF
(627)
pp. 284285, ke
Table 65
Reliability Factors ke
Corresponding to
8 Percent Standard
Deviation of the
Endurance Limit
Reliability, %
Transformation Variate za
Reliability Factor ke
50
0
1.000
90
1.288
0.897
95
1.645
0.868
99
2.326
0.814
99.9
3.091
0.753
99.99
3.719
0.702
99.999
4.265
0.659
99.9999
4.753
0.620
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
338
II. Failure Prevention
341
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
pp. 285286, k f
3 Determine fatigue stress-concentration factor, K f or K f s . First, nd K t or K ts from
Table A15.
p. 287
K f = 1 + q ( K t 1)
K f s = 1 + q ( K ts 1)
or
(632)
Obtain q from either Fig. 620 or 621, pp. 287288.
Alternatively, for reversed bending or axial loads,
Kf = 1 +
p. 288
Kt 1
1 + a /r
(633)
For Sut in kpsi,
a = 0.245 799 0.307 794(102 ) Sut
2
3
+0.150 874(104 ) Sut 0.266 978(107 ) Sut
(635)
For torsion for low-alloy steels, increase Sut by 20 kpsi and apply to Eq. (635).
4 Apply K f or K f s by either dividing Se by it or multiplying it with the purely
reversing stress not both.
5 Determine fatigue life constants a and b. If Sut 70 kpsi, determine f from
Fig. 618, p. 277. If Sut < 70 kpsi, let f = 0.9.
a = ( f Sut ) 2 / Se
(614)
b = [log( f Sut / Se )]/3
p. 277
(615)
6 Determine fatigue strength S f at N cycles, or, N cycles to failure at a reversing
stress σa
(Note: this only applies to purely reversing stresses where σm = 0).
Sf = a N b
p. 277
(613)
N = ( σa /a )
1/b
(616)
Fluctuating Simple Loading
For Se , K f or K f s , see previous subsection.
1 Calculate σm and σa . Apply K f to both stresses.
p. 293
σm = ( σmax + σmin ) /2
σa = |σmax σmin |/2
(636)
2 Apply to a fatigue failure criterion, p. 298
σm 0
Soderburg
σa / Se + σm / Sy = 1/ n
(645)
mod-Goodman
σa / Se + σm / Sut = 1/ n
(646)
n σa / Se + ( n σm / Sut ) = 1
(647)
Gerber
ASME-elliptic
2
( σa / Se ) 2 + ( σm / Sut ) 2 = 1/ n 2
σm < 0
p. 297
σa = Se / n
(648)
342
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
339
Fatigue Failure Resulting from Variable Loading
Torsion. Use the same equations as apply for σm 0, except replace σm and σa with
τm and τa , use kc = 0.59 for Se , replace Sut with Ssu = 0.67 Sut [Eq. (654), p. 309],
and replace Sy with Ssy = 0.577 Sy [Eq. (521), p. 217]
3 Check for localized yielding.
(649)
p. 298
σa + σm = Sy / n
or, for torsion,
τa + τm = 0.577 Sy / n
4 For nite-life fatigue strength (see Ex. 612, pp. 305306),
mod-Goodman
Sf =
σa
1 ( σm / Sut )
Gerber
Sf =
σa
1 ( σm / Sut ) 2
If determining the nite life N with a factor of safety n, substitute S f / n for σa in
Eq. (616). That is,
N=
Sf /n
a
1/b
Combination of Loading Modes
See previous subsections for earlier denitions.
1 Calculate von Mises stresses for alternating and midrange stress states, σa and σm .
When determining Se , do not use kc nor divide by K f or K f s . Apply K f and/or K f s
directly to each specic alternating and midrange stress. If axial stress is present
divide the alternating axial stress by kc = 0.85. For the special case of combined
bending, torsional shear, and axial stresses
p. 310
σa
=
( σa ) axial
( K f ) bending ( σa ) bending + ( K f ) axial
0.85
1/2
2
+ 3 ( K f s ) torsion ( τa ) torsion
2
(655)
σm =
( K f ) bending ( σm ) bending + ( K f ) axial ( σm ) axial
2
+ 3 ( K f s ) torsion ( τm ) torsion
2 1/2
(656)
2 Apply stresses to fatigue criterion [see Eq. (645) to (648), p. 338 in previous
subsection].
3 Conservative check for localized yielding using von Mises stresses.
p. 298
σa + σm = Sy / n
(649)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
340
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
PROBLEMS
Problems 61 to 631 are to be solved by deterministic methods. Problems 632 to 638 are to
be solved by stochastic methods. Problems 639 to 646 are computer problems.
Deterministic Problems
61
A 1 -in drill rod was heat-treated and ground. The measured hardness was found to be 490 Brinell.
4
Estimate the endurance strength if the rod is used in rotating bending.
62
Estimate Se for the following materials:
(a) AISI 1020 CD steel.
(b) AISI 1080 HR steel.
(c) 2024 T3 aluminum.
(d) AISI 4340 steel heat-treated to a tensile strength of 250 kpsi.
63
Estimate the fatigue strength of a rotating-beam specimen made of AISI 1020 hot-rolled steel corresponding to a life of 12.5 kilocycles of stress reversal. Also, estimate the life of the specimen
corresponding to a stress amplitude of 36 kpsi. The known properties are Sut = 66.2 kpsi, σ0 =
115 kpsi, m = 0.22, and ε f = 0.90.
64
65
66
Derive Eq. (617). For the specimen of Prob. 63, estimate the strength corresponding to
500 cycles.
For the interval 103 N 106 cycles, develop an expression for the axial fatigue strength
( S f )ax for the polished specimens of 4130 used to obtain Fig. 610. The ultimate strength is
Sut = 125 kpsi and the endurance limit is ( Se )ax = 50 kpsi.
Estimate the endurance strength of a 32-mm-diameter rod of AISI 1035 steel having a machined
nish and heat-treated to a tensile strength of 710 MPa.
67
Two steels are being considered for manufacture of as-forged connecting rods. One is AISI 4340
Cr-Mo-Ni steel capable of being heat-treated to a tensile strength of 260 kpsi. The other is a plain carbon steel AISI 1040 with an attainable Sut of 113 kpsi. If each rod is to have a size giving an equivalent diameter de of 0.75 in, is there any advantage to using the alloy steel for this fatigue application?
68
A solid round bar, 25 mm in diameter, has a groove 2.5-mm deep with a 2.5-mm radius machined
into it. The bar is made of AISI 1018 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 C, estimate the number
of cycles to failure.
69
A solid square rod is cantilevered at one end. The rod is 0.8 m long and supports a completely
reversing transverse load at the other end of ±1 kN. The material is AISI 1045 hot-rolled steel.
If the rod must support this load for 104 cycles with a factor of safety of 1.5, what dimension
should the square cross section have? Neglect any stress concentrations at the support end and
assume that f = 0.9.
610
A rectangular bar is cut from an AISI 1018 cold-drawn steel at. The bar is 60 mm wide by
10 mm thick and has a 12-mm hole drilled through the center as depicted in Table A151. The
bar is concentrically loaded in push-pull fatigue by axial forces Fa , uniformly distributed across
the width. Using a design factor of n d = 1.8, estimate the largest force Fa that can be applied
ignoring column action.
611
Bearing reactions R1 and R2 are exerted on the shaft shown in the gure, which rotates at
1150 rev/min and supports a 10-kip bending force. Use a 1095 HR steel. Specify a diameter d
using a design factor of n d = 1.6 for a life of 3 min. The surfaces are machined.
343
344
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
341
F = 10 kip
12 in
6 in
6 in
d /5 R.
Problem 611
d
R1
1.5 d
d /10 R.
R2 d
1 in
612
A bar of steel has the minimum properties Se = 276 MPa, Sy = 413 MPa, and Sut = 551 MPa.
The bar is subjected to a steady torsional stress of 103 MPa and an alternating bending stress of
172 MPa. Find the factor of safety guarding against a static failure, and either the factor of safety guarding against a fatigue failure or the expected life of the part. For the fatigue analysis use:
(a) Modied Goodman criterion.
(b) Gerber criterion.
(c) ASME-elliptic criterion.
613
Repeat Prob. 612 but with a steady torsional stress of 138 MPa and an alternating bending stress
of 69 MPa.
614
Repeat Prob. 612 but with a steady torsional stress of 103 MPa, an alternating torsional stress
of 69 MPa, and an alternating bending stress of 83 MPa.
615
Repeat Prob. 612 but with an alternating torsional stress of 207 MPa.
616
Repeat Prob. 612 but with an alternating torsional stress of 103 MPa and a steady bending stress
of 103 MPa.
617
The cold-drawn AISI 1018 steel bar shown in the gure is subjected to an axial load uctuating
between 800 and 3000 lbf. Estimate the factors of safety n y and n f using (a) a Gerber fatigue
failure criterion as part of the designers fatigue diagram, and (b) an ASME-elliptic fatigue failure criterion as part of the designers fatigue diagram.
1
4
in D.
1 in
Problem 617
3
8
in
618
Repeat Prob. 617, with the load uctuating between 800 and 3000 lbf. Assume no buckling.
619
Repeat Prob. 617, with the load uctuating between 800 and 3000 lbf. Assume no buckling.
620
The gure shows a formed round-wire cantilever spring subjected to a varying force. The hardness tests made on 25 springs gave a minimum hardness of 380 Brinell. It is apparent from the
mounting details that there is no stress concentration. A visual inspection of the springs indicates
16 in
Problem 620
3
8
in D.
Fmax = 30 lbf
Fmin = 15 lbf
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
342
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
that the surface nish corresponds closely to a hot-rolled nish. What number of applications is
likely to cause failure? Solve using:
(a) Modied Goodman criterion.
(b) Gerber criterion.
621
The figure is a drawing of a 3- by 18-mm latching spring. A preload is obtained during assembly by shimming under the bolts to obtain an estimated initial deflection of 2 mm. The latching operation itself requires an additional deflection of exactly 4 mm. The material is ground
high-carbon steel, bent then hardened and tempered to a minimum hardness of 490 Bhn. The
radius of the bend is 3 mm. Estimate the yield strength to be 90 percent of the ultimate
strength.
(a) Find the maximum and minimum latching forces.
(b) Is it likely the spring will fail in fatigue? Use the Gerber criterion.
F
100
A
A
Problem 621
Dimensions in millimeters
18
3
Section
AA
622
Repeat Prob. 621, part b, using the modied Goodman criterion.
623
The gure shows the free-body diagram of a connecting-link portion having stress concentration
at three sections. The dimensions are r = 0.25 in, d = 0.75 in, h = 0.50 in, w1 = 3.75 in, and
w2 = 2.5 in. The forces F uctuate between a tension of 4 kip and a compression of 16 kip.
Neglect column action and nd the least factor of safety if the material is cold-drawn AISI 1018
steel.
A
Problem 623
F
F
w1
w2
A
624
h
r
d
Section AA
The torsional coupling in the gure is composed of a curved beam of square cross section that is
welded to an input shaft and output plate. A torque is applied to the shaft and cycles from zero to
T. The cross section of the beam has dimensions of 5 by 5 mm, and the centroidal axis of the
beam describes a curve of the form r = 20 + 10 θ/π , where r and θ are in mm and radians,
respectively (0 θ 4π ). The curved beam has a machined surface with yield and ultimate
strength values of 420 and 770 MPa, respectively.
(a) Determine the maximum allowable value of T such that the coupling will have an innite life
with a factor of safety, n = 3, using the modied Goodman criterion.
(b) Repeat part (a) using the Gerber criterion.
(c) Using T found in part (b), determine the factor of safety guarding against yield.
345
346
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Fatigue Failure Resulting from Variable Loading
343
T
5
T
20
Problem 624
60
(Dimensions in mm)
625
Repeat Prob. 624 ignoring curvature effects on the bending stress.
626
In the gure shown, shaft A, made of AISI 1010 hot-rolled steel, is welded to a xed support and is
subjected to loading by equal and opposite forces F via shaft B. A theoretical stress concentration
K t s of 1.6 is induced by the 3-mm llet. The length of shaft A from the xed support to the connection at shaft B is 1 m. The load F cycles from 0.5 to 2 kN.
(a) For shaft A, nd the factor of safety for innite life using the modied Goodman fatigue failure criterion.
(b) Repeat part (a) using the Gerber fatigue failure criterion.
F
20 mm
25
Problem 626
mm
m
mm m
2510
3 mm
fillet
Shaft B
Shaft A
F
627
A schematic of a clutch-testing machine is shown. The steel shaft rotates at a constant speed ω.
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 is given by
f P(D + d)
T=
4
where D and d are dened in the gure and f is the coefcient of friction of the clutch face. The
shaft is machined with Sy = 800 MPa and Sut = 1000 MPa. The theoretical stress concentration
factors for the llet are 3.0 and 1.8 for the axial and torsional loading, respectively.
(a) Assume the load variation P is synchronous with shaft rotation. With f = 0.3, nd the maximum allowable load P such that the shaft will survive a minimum of 106 cycles with a factor
of safety of 3. Use the modied Goodman criterion. Determine the corresponding factor of
safety guarding against yielding.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
344
II. Failure Prevention
© The McGrawHill
Companies, 2008
6. Fatigue Failure Resulting
from Variable Loading
Mechanical Engineering Design
(b) Suppose the shaft is not rotating, but the load P is cycled as shown. With f = 0.3, nd the
maximum allowable load P so that the shaft will survive a minimum of 106 cycles with a factor of safety of 3. Use the modied Goodman criterion. Determine the corresponding factor
of safety guarding against yielding.
R=3
d = 30 mm
Problem 627
P
Friction pad
D = 150 mm
628
For the clutch of Prob. 627, the external load P is cycled between 20 kN and 80 kN. Assuming
that the shaft is rotating synchronous with the external load cycle, estimate the number of cycles
to failure. Use the modied Goodman fatigue failure criteria.
629
A at leaf spring has uctuating stress of σmax = 420 MPa and σmin = 140 MPa applied for
5 (104) cycles. If the load changes to σmax = 350 MPa and σmin = 200 MPa, how many cycles
should the spring survive? The material is AISI 1040 CD and has a fully corrected endurance
strength of Se = 200 MPa. Assume that f = 0.9.
(a) Use Miners method.
(b) Use Mansons method.
630
A machine part will be cycled at ±48 kpsi for 4 (103) cycles. Then the loading will be changed
to ±38 kpsi for 6 (104) cycles. Finally, the load will be changed to ±32 kpsi. How many cycles
of operation can be expected at this stress level? For the part, Sut = 76 kpsi, f = 0.9, and has a
fully corrected endurance strength of Se = 30 kpsi.
(a) Use Miners method.
(b) Use Mansons method.
631
A rotating-beam specimen with an endurance limit of 50 kpsi and an ultimate strength of 100 kpsi
is cycled 20 percent of the time at 70 kpsi, 50 percent at 55 kpsi, and 30 percent at 40 kpsi. Let
f = 0.9 and estimate the number of cycles to failure.
Stochastic Problems
632
Solve Prob. 61 if the ultimate strength of production pieces is found to be Sut = 245LN
(1, 0.0508)kpsi.
633
The situation is similar to that of Prob. 610 wherein the imposed completely reversed axial load
Fa = 15LN(1, 0.20) kN is to be carried by the link with a thickness to be specied by you, the
designer. Use the 1018 cold-drawn steel of Prob. 610 with Sut = 440LN(1, 0.30) MPa and
S yt = 370LN(1, 0.061) . The reliability goal must exceed 0.999. Using the correlation method,
specify the thickness t.
634
A solid round steel bar is machined to a diameter of 1.25 in. A groove 1 in deep with a radius of
8
1
in is cut into the bar. The material has a mean tensile strength of 110 kpsi. A completely
8
reversed bending moment M = 1400 lbf · in is applied. Estimate the reliability. The size factor
should be based on the gross diameter. The bar rotates.
347
348
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
II. Failure Prevention
6. Fatigue Failure Resulting
from Variable Loading
© The McGrawHill
Companies, 2008
Fatigue Failure Resulting from Variable Loading
345
635
Repeat Prob. 634, with a completely reversed torsional moment of T = 1400 lbf · in applied.
636
1
A 1 1 -in-diameter hot-rolled steel bar has a 8 -in diameter hole drilled transversely through it. The
4
bar is nonrotating and is subject to a completely reversed bending moment of M = 1600 lbf · in in
the same plane as the axis of the transverse hole. The material has a mean tensile strength of 58 kpsi.
Estimate the reliability. The size factor should be based on the gross size. Use Table A16 for K t .
637
Repeat Prob. 636, with the bar subject to a completely reversed torsional moment of 2400 lbf · in.
638
The plan view of a link is the same as in Prob. 623; however, the forces F are completely
reversed, the reliability goal is 0.998, and the material properties are Sut = 64LN(1, 0.045) kpsi
and S y = 54LN(1, 0.077) kpsi. Treat Fa as deterministic, and specify the thickness h.
Computer Problems
639
1
1
A 4 by 1 2 -in steel bar has a 3 -in drilled hole located in the center, much as is shown in
4
Table A151. The bar is subjected to a completely reversed axial load with a deterministic load
¯
of 1200 lbf. The material has a mean ultimate tensile strength of Sut = 80 kpsi.
(a) Estimate the reliability.
(b) Conduct a computer simulation to conrm your answer to part a.
640
From your experience with Prob. 639 and Ex. 619, you observed that for completely reversed
axial and bending fatigue, it is possible to
Observe the COVs associated with a priori design considerations.
Note the reliability goal.
¯
Find the mean design factor n d which will permit making a geometric design decision that
¯
will attain the goal using deterministic methods in conjunction with n d .
¯
Formulate an interactive computer program that will enable the user to nd n d . While the material properties Sut , S y , and the load COV must be input by the user, all of the COVs associated with
0 .30 , ka , kc , kd , and K f can be internal, and answers to questions will allow C σ and C S , as well
¯
as Cn and n d , to be calculated. Later you can add improvements. Test your program with problems you have already solved.
641
When using the Gerber fatigue failure criterion in a stochastic problem, Eqs. (680) and (681)
are useful. They are also computationally complicated. It is helpful to have a computer subroutine
or procedure that performs these calculations. When writing an executive program, and it is
appropriate to nd Sa and C Sa , a simple call to the subroutine does this with a minimum of effort.
Also, once the subroutine is tested, it is always ready to perform. Write and test such a program.
642
Repeat Problem. 641 for the ASME-elliptic fatigue failure locus, implementing Eqs. (682) and
(683).
643
Repeat Prob. 641 for the Smith-Dolan fatigue failure locus, implementing Eqs. (686) and (687).
644
Write and test computer subroutines or procedures that will implement
¯
(a) Table 62, returning a, b, C, and ka .
(b) Equation (620) using Table 64, returning kb .
¯
(c) Table 611, returning α , β , C, and kc .
¯
(d) Equations (627) and (675), returning kd and Ckd .
645
Write and test a computer subroutine or procedure that implements Eqs. (676) and (677),
¯ˆ
returning q , σq , and Cq .
646
Write and test a computer subroutine or procedure that implements Eq. (678) and Table 615,
¯
returning a , C K f , and K f .
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
PART
III. Design of Mechanical
Elements
Introduction
3
© The McGrawHill
Companies, 2008
Design of Mechanical
Elements
349
350
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
7
Shafts and Shaft
Components
Chapter Outline
71
Introduction
72
Shaft Materials
73
Shaft Layout
74
Shaft Design for Stress
75
Deection Considerations
367
76
Critical Speeds for Shafts
371
77
Miscellaneous Shaft Components
78
Limits and Fits
348
348
349
354
376
383
347
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
348
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
71
Introduction
A shaft is a rotating member, usually of circular cross section, used to transmit power
or motion. It provides the axis of rotation, or oscillation, of elements such as gears,
pulleys, ywheels, cranks, sprockets, and the like and controls the geometry of their
motion. An axle is a nonrotating member that carries no torque and is used to support rotating wheels, pulleys, and the like. The automotive axle is not a true axle; the
term is a carry-over from the horse-and-buggy era, when the wheels rotated on nonrotating members. A non-rotating axle can readily be designed and analyzed as a static
beam, and will not warrant the special attention given in this chapter to the rotating
shafts which are subject to fatigue loading.
There is really nothing unique about a shaft that requires any special treatment
beyond the basic methods already developed in previous chapters. However, because of
the ubiquity of the shaft in so many machine design applications, there is some advantage in giving the shaft and its design a closer inspection. A complete shaft design has
much interdependence on the design of the components. The design of the machine itself
will dictate that certain gears, pulleys, bearings, and other elements will have at least been
partially analyzed and their size and spacing tentatively determined. Chapter 18 provides
a complete case study of a power transmission, focusing on the overall design process.
In this chapter, details of the shaft itself will be examined, including the following:
Material selection
Geometric layout
Stress and strength
Static strength
Fatigue strength
Deection and rigidity
Bending deection
Torsional deection
Slope at bearings and shaft-supported elements
Shear deection due to transverse loading of short shafts
Vibration due to natural frequency
In deciding on an approach to shaft sizing, it is necessary to realize that a stress analysis at a specic point on a shaft can be made using only the shaft geometry in the vicinity of that point. Thus the geometry of the entire shaft is not needed. In design it is usually
possible to locate the critical areas, size these to meet the strength requirements, and then
size the rest of the shaft to meet the requirements of the shaft-supported elements.
The deection and slope analyses cannot be made until the geometry of the entire
shaft has been dened. Thus deection is a function of the geometry everywhere,
whereas the stress at a section of interest is a function of local geometry. For this reason, shaft design allows a consideration of stress rst. Then, after tentative values for
the shaft dimensions have been established, the determination of the deections and
slopes can be made.
72
Shaft Materials
Deection is not affected by strength, but rather by stiffness as represented by the
modulus of elasticity, which is essentially constant for all steels. For that reason, rigidity cannot be controlled by material decisions, but only by geometric decisions.
351
352
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Shafts and Shaft Components
349
Necessary strength to resist loading stresses affects the choice of materials and
their treatments. Many shafts are made from low carbon, cold-drawn or hot-rolled
steel, such as ANSI 1020-1050 steels.
Signicant strengthening from heat treatment and high alloy content are often not
warranted. Fatigue failure is reduced moderately by increase in strength, and then only
to a certain level before adverse effects in endurance limit and notch sensitivity begin
to counteract the benets of higher strength. A good practice is to start with an inexpensive, low or medium carbon steel for the rst time through the design calculations.
If strength considerations turn out to dominate over deection, then a higher strength
material should be tried, allowing the shaft sizes to be reduced until excess deection
becomes an issue. The cost of the material and its processing must be weighed against
the need for smaller shaft diameters. When warranted, typical alloy steels for heat
treatment include ANSI 1340-50, 3140-50, 4140, 4340, 5140, and 8650.
Shafts usually dont need to be surface hardened unless they serve as the actual
journal of a bearing surface. Typical material choices for surface hardening include
carburizing grades of ANSI 1020, 4320, 4820, and 8620.
Cold drawn steel is usually used for diameters under about 3 inches. The nominal diameter of the bar can be left unmachined in areas that do not require fitting
of components. Hot rolled steel should be machined all over. For large shafts
requiring much material removal, the residual stresses may tend to cause warping.
If concentricity is important, it may be necessary to rough machine, then heat treat
to remove residual stresses and increase the strength, then finish machine to the
final dimensions.
In approaching material selection, the amount to be produced is a salient factor.
For low production, turning is the usual primary shaping process. An economic viewpoint may require removing the least material. High production may permit a volumeconservative shaping method (hot or cold forming, casting), and minimum material in
the shaft can become a design goal. Cast iron may be specied if the production quantity is high, and the gears are to be integrally cast with the shaft.
Properties of the shaft locally depend on its historycold work, cold forming,
rolling of llet features, heat treatment, including quenching medium, agitation, and
tempering regimen.1
Stainless steel may be appropriate for some environments.
73
Shaft Layout
The general layout of a shaft to accommodate shaft elements, e.g. gears, bearings, and
pulleys, must be specied early in the design process in order to perform a free body
force analysis and to obtain shear-moment diagrams. The geometry of a shaft is generally that of a stepped cylinder. The use of shaft shoulders is an excellent means of
axially locating the shaft elements and to carry any thrust loads. Figure 71 shows an
example of a stepped shaft supporting the gear of a worm-gear speed reducer. Each
shoulder in the shaft serves a specic purpose, which you should attempt to determine by observation.
1
See Joseph E. Shigley, Charles R. Mischke, and Thomas H. Brown, Jr. (eds-in-chief), Standard Handbook
of Machine Design, 3rd ed., McGraw-Hill, New York, 2004. For cold-worked property prediction see
Chap. 29, and for heat-treated property prediction see Chaps. 29 and 33.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
350
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
Figure 71
A vertical worm-gear speed
reducer. (Courtesy of the
Cleveland Gear Company.)
Figure 72
(a) Choose a shaft
conguration to support and
locate the two gears and two
bearings. (b) Solution uses an
integral pinion, three shaft
shoulders, key and keyway,
and sleeve. The housing
locates the bearings on their
outer rings and receives the
thrust loads. (c) Choose fanshaft conguration. (d) Solution
uses sleeve bearings, a
straight-through shaft, locating
collars, and setscrews for
collars, fan pulley, and fan
itself. The fan housing supports
the sleeve bearings.
(a)
(b)
Fan
(c)
(d)
The geometric conguration of a shaft to be designed is often simply a revision
of existing models in which a limited number of changes must be made. If there is
no existing design to use as a starter, then the determination of the shaft layout may
have many solutions. This problem is illustrated by the two examples of Fig. 72. In
Fig. 72a a geared countershaft is to be supported by two bearings. In Fig. 72c a
fanshaft is to be congured. The solutions shown in Fig. 72b and 72d are not necessarily the best ones, but they do illustrate how the shaft-mounted devices are xed
and located in the axial direction, and how provision is made for torque transfer from
one element to another. There are no absolute rules for specifying the general layout,
but the following guidelines may be helpful.
353
354
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Shafts and Shaft Components
351
Axial Layout of Components
The axial positioning of components is often dictated by the layout of the housing
and other meshing components. In general, it is best to support load-carrying components between bearings, such as in Fig. 72a, rather than cantilevered outboard of
the bearings, such as in Fig. 72c. Pulleys and sprockets often need to be mounted
outboard for ease of installation of the belt or chain. The length of the cantilever
should be kept short to minimize the deection.
Only two bearings should be used in most cases. For extremely long shafts carrying
several load-bearing components, it may be necessary to provide more than two bearing
supports. In this case, particular care must be given to the alignment of the bearings.
Shafts should be kept short to minimize bending moments and deections. Some
axial space between components is desirable to allow for lubricant ow and to provide access space for disassembly of components with a puller. Load bearing components should be placed near the bearings, again to minimize the bending moment
at the locations that will likely have stress concentrations, and to minimize the deection at the load-carrying components.
The components must be accurately located on the shaft to line up with other
mating components, and provision must be made to securely hold the components in
position. The primary means of locating the components is to position them against
a shoulder of the shaft. A shoulder also provides a solid support to minimize deection and vibration of the component. Sometimes when the magnitudes of the forces
are reasonably low, shoulders can be constructed with retaining rings in grooves,
sleeves between components, or clamp-on collars. In cases where axial loads are very
small, it may be feasible to do without the shoulders entirely, and rely on press ts,
pins, or collars with setscrews to maintain an axial location. See Fig. 72b and 72d
for examples of some of these means of axial location.
Supporting Axial Loads
In cases where axial loads are not trivial, it is necessary to provide a means to transfer the axial loads into the shaft, then through a bearing to the ground. This will be
particularly necessary with helical or bevel gears, or tapered roller bearings, as each
of these produces axial force components. Often, the same means of providing axial
location, e.g., shoulders, retaining rings, and pins, will be used to also transmit the
axial load into the shaft.
It is generally best to have only one bearing carry the axial load, to allow
greater tolerances on shaft length dimensions, and to prevent binding if the shaft
expands due to temperature changes. This is particularly important for long shafts.
Figures 73 and 74 show examples of shafts with only one bearing carrying the
axial load against a shoulder, while the other bearing is simply press-fit onto the
shaft with no shoulder.
Providing for Torque Transmission
Most shafts serve to transmit torque from an input gear or pulley, through the shaft, to
an output gear or pulley. Of course, the shaft itself must be sized to support the torsional
stress and torsional deection. It is also necessary to provide a means of transmitting the
torque between the shaft and the gears. Common torque-transfer elements are:
Keys
Splines
Setscrews
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
352
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 73
Tapered roller bearings used
in a mowing machine spindle.
This design represents good
practice for the situation in
which one or more torquetransfer elements must be
mounted outboard. (Source:
Redrawn from material
furnished by The Timken
Company.)
Figure 74
A bevel-gear drive in which
both pinion and gear are
straddle-mounted. (Source:
Redrawn from material
furnished by Gleason
Machine Division.)
Pins
Press or shrink ts
Tapered ts
In addition to transmitting the torque, many of these devices are designed to fail if
the torque exceeds acceptable operating limits, protecting more expensive components.
Details regarding hardware components such as keys, pins, and setscrews are
addressed in detail in Sec. 77. One of the most effective and economical means
of transmitting moderate to high levels of torque is through a key that fits in a
groove in the shaft and gear. Keyed components generally have a slip fit onto the
shaft, so assembly and disassembly is easy. The key provides for positive angular
orientation of the component, which is useful in cases where phase angle timing
is important.
355
356
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Shafts and Shaft Components
353
Splines are essentially stubby gear teeth formed on the outside of the shaft and on
the inside of the hub of the load-transmitting component. Splines are generally much
more expensive to manufacture than keys, and are usually not necessary for simple
torque transmission. They are typically used to transfer high torques. One feature of a
spline is that it can be made with a reasonably loose slip t to allow for large axial
motion between the shaft and component while still transmitting torque. This is useful for connecting two shafts where relative motion between them is common, such as
in connecting a power takeoff (PTO) shaft of a tractor to an implement. SAE and ANSI
publish standards for splines. Stress concentration factors are greatest where the spline
ends and blends into the shaft, but are generally quite moderate.
For cases of low torque transmission, various means of transmitting torque are
available. These include pins, setscrews in hubs, tapered ts, and press ts.
Press and shrink ts for securing hubs to shafts are used both for torque transfer and for preserving axial location. The resulting stress-concentration factor is usually quite small. See Sec. 78 for guidelines regarding appropriate sizing and tolerancing to transmit torque with press and shrink ts. A similar method is to use a split
hub with screws to clamp the hub to the shaft. This method allows for disassembly
and lateral adjustments. Another similar method uses a two-part hub consisting of a
split inner member that ts into a tapered hole. The assembly is then tightened to the
shaft with screws, which forces the inner part into the wheel and clamps the whole
assembly against the shaft.
Tapered ts between the shaft and the shaft-mounted device, such as a wheel, are
often used on the overhanging end of a shaft. Screw threads at the shaft end then permit
the use of a nut to lock the wheel tightly to the shaft. This approach is useful because it
can be disassembled, but it does not provide good axial location of the wheel on the shaft.
At the early stages of the shaft layout, the important thing is to select an appropriate means of transmitting torque, and to determine how it affects the overall shaft
layout. It is necessary to know where the shaft discontinuities, such as keyways, holes,
and splines, will be in order to determine critical locations for analysis.
Assembly and Disassembly
Consideration should be given to the method of assembling the components onto the
shaft, and the shaft assembly into the frame. This generally requires the largest diameter in the center of the shaft, with progressively smaller diameters towards the ends
to allow components to be slid on from the ends. If a shoulder is needed on both sides
of a component, one of them must be created by such means as a retaining ring or
by a sleeve between two components. The gearbox itself will need means to physically position the shaft into its bearings, and the bearings into the frame. This is typically accomplished by providing access through the housing to the bearing at one
end of the shaft. See Figs. 75 through 78 for examples.
Figure 75
Arrangement showing bearing
inner rings press-tted to shaft
while outer rings oat in the
housing. The axial clearance
should be sufcient only to
allow for machinery vibrations.
Note the labyrinth seal on the
right.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
354
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 76
Similar to the arrangement of
Fig. 7--5 except that the outer
bearing rings are preloaded.
Figure 77
In this arrangement the inner ring of the left-hand bearing is locked to the shaft between a
nut and a shaft shoulder. The locknut and washer are AFBMA standard. The snap ring in
the outer race is used to positively locate the shaft assembly in the axial direction. Note
the oating right-hand bearing and the grinding runout grooves in the shaft.
Figure 78
This arrangement is similar to
Fig. 7--7 in that the left-hand
bearing positions the entire
shaft assembly. In this case
the inner ring is secured to
the shaft using a snap ring.
Note the use of a shield to
prevent dirt generated
from within the machine from
entering the bearing.
74
When components are to be press-t to the shaft, the shaft should be designed
so that it is not necessary to press the component down a long length of shaft. This
may require an extra change in diameter, but it will reduce manufacturing and assembly cost by only requiring the close tolerance for a short length.
Consideration should also be given to the necessity of disassembling the components from the shaft. This requires consideration of issues such as accessibility of
retaining rings, space for pullers to access bearings, openings in the housing to allow
pressing the shaft or bearings out, etc.
Shaft Design for Stress
Critical Locations
It is not necessary to evaluate the stresses in a shaft at every point; a few potentially
critical locations will sufce. Critical locations will usually be on the outer surface,
at axial locations where the bending moment is large, where the torque is present, and
where stress concentrations exist. By direct comparison of various points along the
shaft, a few critical locations can be identied upon which to base the design. An
assessment of typical stress situations will help.
357
358
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
355
Most shafts will transmit torque through a portion of the shaft. Typically the
torque comes into the shaft at one gear and leaves the shaft at another gear. A free
body diagram of the shaft will allow the torque at any section to be determined. The
torque is often relatively constant at steady state operation. The shear stress due to
the torsion will be greatest on outer surfaces.
The bending moments on a shaft can be determined by shear and bending
moment diagrams. Since most shaft problems incorporate gears or pulleys that introduce forces in two planes, the shear and bending moment diagrams will generally
be needed in two planes. Resultant moments are obtained by summing moments as
vectors at points of interest along the shaft. The phase angle of the moments is not
important since the shaft rotates. A steady bending moment will produce a completely reversed moment on a rotating shaft, as a specic stress element will alternate from compression to tension in every revolution of the shaft. The normal stress
due to bending moments will be greatest on the outer surfaces. In situations where
a bearing is located at the end of the shaft, stresses near the bearing are often not
critical since the bending moment is small.
Axial stresses on shafts due to the axial components transmitted through helical gears or tapered roller bearings will almost always be negligibly small compared
to the bending moment stress. They are often also constant, so they contribute little to fatigue. Consequently, it is usually acceptable to neglect the axial stresses
induced by the gears and bearings when bending is present in a shaft. If an axial
load is applied to the shaft in some other way, it is not safe to assume it is negligible without checking magnitudes.
Shaft Stresses
Bending, torsion, and axial stresses may be present in both midrange and alternating
components. For analysis, it is simple enough to combine the different types of
stresses into alternating and midrange von Mises stresses, as shown in Sec. 614,
p. 309. It is sometimes convenient to customize the equations specically for shaft
applications. Axial loads are usually comparatively very small at critical locations
where bending and torsion dominate, so they will be left out of the following equations. The uctuating stresses due to bending and torsion are given by
σa = K f
Ma c
I
σm = K f
Mm c
I
(71)
Ta c
J
τm = K f s
Tm c
J
(72)
τa = K f s
where Mm and Ma are the midrange and alternating bending moments, Tm and Ta are
the midrange and alternating torques, and K f and K f s are the fatigue stress concentration factors for bending and torsion, respectively.
Assuming a solid shaft with round cross section, appropriate geometry terms can
be introduced for c, I, and J resulting in
σa = K f
32 Ma
π d3
σm = K f
16Ta
π d3
τm = K f s
τa = K f s
32 Mm
π d3
(73)
16Tm
π d3
(74)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
356
III. Design of Mechanical
Elements
359
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
Combining these stresses in accordance with the distortion energy failure theory,
the von Mises stresses for rotating round, solid shafts, neglecting axial loads, are given
by
σa
σm
=
2
(σa
=
2
(σm
+
2
3τa )1/2
+
2
3τm )1/2
2
=
32 K f Ma
π d3
2
=
32 K f Mm
π d3
16 K f s Ta
+3
π d3
16 K f s Tm
+3
π d3
2 1/2
(75)
2 1/2
(76)
Note that the stress concentration factors are sometimes considered optional for the
midrange components with ductile materials, because of the capacity of the ductile
material to yield locally at the discontinuity.
These equivalent alternating and midrange stresses can be evaluated using an
appropriate failure curve on the modied Goodman diagram (See Sec. 612, p. 295,
and Fig. 627). For example, the fatigue failure criteria for the modied Goodman
line as expressed previously in Eq. (646) is
σ
1
σ
= a+ m
n
Se
Sut
Substitution of σa and σm from Eqs. (75) and (76) results in
1
16
=
n
π d3
1
4( K f Ma )2 + 3( K f s Ta )2
Se
1/2
+
1
4( K f Mm )2 + 3( K f s Tm )2
Sut
1/2
For design purposes, it is also desirable to solve the equation for the diameter.
This results in
d=
16n
π
1
1/2
4( K f Ma )2 + 3( K f s Ta )2
Se
1
1/2
4( K f Mm )2 + 3( K f s Tm )2
+
Sut
1/3
Similar expressions can be obtained for any of the common failure criteria by substituting the von Mises stresses from Eqs. (75) and (76) into any of the failure
criteria expressed by Eqs. (645) through (648), p. 298. The resulting equations
for several of the commonly used failure curves are summarized below. The names
given to each set of equations identifies the significant failure theory, followed by
a fatigue failure locus name. For example, DE-Gerber indicates the stresses are
combined using the distortion energy (DE) theory, and the Gerber criteria is used
for the fatigue failure.
DE-Goodman
16
1
=
n
π d3
1
4( K f Ma )2 + 3( K f s Ta )2
Se
1/2
+
1
4( K f Mm )2 + 3( K f s Tm )2
Sut
1/2
(77)
d=
16n
π
1
1/2
4( K f Ma )2 + 3( K f s Ta )2
Se
1
1/2
4( K f Mm )2 + 3( K f s Tm )2
+
Sut
1/3
(78)
360
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
357
DE-Gerber
1
8A
=
n
π d 3 Se
8n A
d=
π Se
where
1+ 1+
2 B Se
ASut
2 B Se
ASut
1+ 1+
2 1/2
2 1/2
(79)
1/3
(710)
A=
4( K f Ma ) 2 + 3( K f s Ta ) 2
B=
4( K f Mm ) 2 + 3( K f s Tm ) 2
DE-ASME Elliptic
1
16
=
4
n
π d3
2
K f Ma
Se
+3
2
K f s Ta
Se
+4
2
K f Mm
Sy
+3
K f s Tm
Sy
2 1/2
(711)
d=
16n
π
4
K f Ma
Se
2
+3
K f s Ta
Se
2
+4
2
K f Mm
Sy
+3
K f s Tm
Sy
1/3
2 1/2
(712)
DE-Soderberg
1
16
=
n
π d3
1
4( K f Ma )2 + 3( K f s Ta )2
Se
1/2
+
1
4( K f Mm )2 + 3( K f s Tm )2
Syt
1/2
(713)
d=
16n
π
+
1
4( K f Ma )2 + 3( K f s Ta )2
Se
1
4( K f Mm )2 + 3( K f s Tm )2
Syt
1/2
1/2
1/3
(714)
For a rotating shaft with constant bending and torsion, the bending stress is completely reversed and the torsion is steady. Equations (77) through (714) can be simplied by setting Mm and Ta equal to 0, which simply drops out some of the terms.
Note that in an analysis situation in which the diameter is known and the factor
of safety is desired, as an alternative to using the specialized equations above, it is
always still valid to calculate the alternating and mid-range stresses using Eqs. (75)
and (76), and substitute them into one of the equations for the failure criteria, Eqs.
(645) through (648), and solve directly for n. In a design situation, however, having the equations pre-solved for diameter is quite helpful.
It is always necessary to consider the possibility of static failure in the rst load cycle.
The Soderberg criteria inherently guards against yielding, as can be seen by noting that
its failure curve is conservatively within the yield (Langer) line on Fig. 627, p. 297. The
ASME Elliptic also takes yielding into account, but is not entirely conservative
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
358
III. Design of Mechanical
Elements
361
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
throughout its entire range. This is evident by noting that it crosses the yield line in
Fig. 627. The Gerber and modied Goodman criteria do not guard against yielding,
requiring a separate check for yielding. A von Mises maximum stress is calculated for
this purpose.
σmax = ( σm + σa ) 2 + 3 ( τm + τa ) 2
=
32 K f ( Mm + Ma )
π d3
2
1/2
16 K f s ( Tm + Ta )
+3
π d3
2 1/2
(715)
To check for yielding, this von Mises maximum stress is compared to the yield
strength, as usual.
ny =
Sy
σmax
(716)
For a quick, conservative check, an estimate for σmax can be obtained by simply
adding σa and σm . (σa + σm ) will always be greater than or equal to σmax , and will
therefore be conservative.
EXAMPLE 71
At a machined shaft shoulder the small diameter d is 1.100 in, the large diameter D
is 1.65 in, and the llet radius is 0.11 in. The bending moment is 1260 lbf · in and
the steady torsion moment is 1100 lbf · in. The heat-treated steel shaft has an ultimate
strength of Sut = 105 kpsi and a yield strength of Sy = 82 kpsi. The reliability goal
is 0.99.
(a) Determine the fatigue factor of safety of the design using each of the fatigue failure
criteria described in this section.
(b) Determine the yielding factor of safety.
Solution
(a) D /d = 1.65/1.100 = 1.50, r /d = 0.11/1.100 = 0.10, K t = 1.68 (Fig. A159),
K ts = 1.42 (Fig. A158), q = 0.85 (Fig. 620), qshear = 0.92 (Fig. 621).
From Eq. (632),
K f = 1 + 0.85(1.68 1) = 1.58
K f s = 1 + 0.92(1.42 1) = 1.39
Eq. (68):
Se = 0.5(105) = 52.5 kpsi
Eq. (619):
ka = 2.70(105) 0.265 = 0.787
Eq. (620):
kb =
1.100
0.30
0.107
kc = kd = k f = 1
= 0.870
362
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
359
ke = 0.814
Table 66:
Se = 0.787(0.870)0.814(52.5) = 29.3 kpsi
For a rotating shaft, the constant bending moment will create a completely reversed
bending stress.
Ma = 1260 lbf · in
Tm = 1100 lbf · in
Mm = Ta = 0
Applying Eq. (77) for the DE-Goodman criteria gives
16
1
=
n
π (1.1) 3
4 (1.58 · 1260) 2
29 300
Answer
n = 1.62
1/2
+
3 (1.39 · 1100) 2
105 000
1/2
= 0.615
DE-Goodman
Similarly, applying Eqs. (79), (711), and (713) for the other failure criteria,
Answer
n = 1.87
DE-Gerber
Answer
n = 1.88
DE-ASME Elliptic
Answer
n = 1.56
DE-Soderberg
For comparison, consider an equivalent approach of calculating the stresses and applying the fatigue failure criteria directly. From Eqs. (75) and (76),
σa
32 · 1.58 · 1260
π (1.1) 3
=
σm = 3
16 · 1.39 · 1100
π (1.1) 3
2 1/2
= 15 235 psi
2 1/2
= 10 134 psi
Taking, for example, the Goodman failure critera, application of Eq. (646)
gives
1
σ
10 134
σ
15 235
= a+ m=
+
= 0.616
n
Se
Sut
29 300 105 000
n = 1.62
which is identical with the previous result. The same process could be used for the
other failure criteria.
(b) For the yielding factor of safety, determine an equivalent von Mises maximum
stress using Eq. (715).
σmax
Answer
=
32(1.58) (1260)
π (1.1) 3
ny =
2
16(1.39) (1100)
+3
π (1.1) 3
2 1/2
= 18 300 psi
Sy
82 000
= 4.48
=
σmax
18 300
For comparison, a quick and very conservative check on yielding can be obtained
by replacing σmax with σa + σm . This just saves the extra time of calculating σmax
if σa and σm have already been determined. For this example,
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
360
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
ny =
σa
Sy
82 000
= 3.23
=
+ σm
15 235 + 10 134
which is quite conservative compared with ny
4.48.
Estimating Stress Concentrations
The stress analysis process for fatigue is highly dependent on stress concentrations.
Stress concentrations for shoulders and keyways are dependent on size specications
that are not known the rst time through the process. Fortunately, since these elements
are usually of standard proportions, it is possible to estimate the stress concentration
factors for initial design of the shaft. These stress concentrations will be ne-tuned in
successive iterations, once the details are known.
Shoulders for bearing and gear support should match the catalog recommendation for the specic bearing or gear. A look through bearing catalogs shows that a
typical bearing calls for the ratio of D /d to be between 1.2 and 1.5. For a rst approximation, the worst case of 1.5 can be assumed. Similarly, the llet radius at the shoulder needs to be sized to avoid interference with the llet radius of the mating component. There is a signicant variation in typical bearings in the ratio of llet radius
versus bore diameter, with r /d typically ranging from around 0.02 to 0.06. A quick
look at the stress concentration charts (Figures A158 and A159) shows that the
stress concentrations for bending and torsion increase signicantly in this range. For
example, with D /d = 1.5 for bending, K t = 2.7 at r /d = 0.02, and reduces to
K t = 2.1 at r /d = 0.05, and further down to K t = 1.7 at r /d = 0.1. This indicates
that this is an area where some attention to detail could make a signicant difference.
Fortunately, in most cases the shear and bending moment diagrams show that bending moments are quite low near the bearings, since the bending moments from the
ground reaction forces are small.
In cases where the shoulder at the bearing is found to be critical, the designer
should plan to select a bearing with generous llet radius, or consider providing for
a larger llet radius on the shaft by relieving it into the base of the shoulder as shown
in Fig. 79a. This effectively creates a dead zone in the shoulder area that does not
Sharp radius
Large radius undercut
Stress flow
Large-radius
relief groove
Shoulder
relief groove
Bearing
Shaft
(a)
(b )
(c)
Figure 79
Techniques for reducing stress concentration at a shoulder supporting a bearing with a sharp radius. (a) Large radius undercut
into the shoulder. (b) Large radius relief groove into the back of the shoulder. (c) Large radius relief groove into the small diameter
363
364
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
361
carry the bending stresses, as shown by the stress ow lines. A shoulder relief groove
as shown in Fig. 79b can accomplish a similar purpose. Another option is to cut a
large-radius relief groove into the small diameter of the shaft, as shown in Fig. 79c.
This has the disadvantage of reducing the cross-sectional area, but is often used
in cases where it is useful to provide a relief groove before the shoulder to
prevent the grinding or turning operation from having to go all the way to the
shoulder.
For the standard shoulder fillet, for estimating K t values for the first iteration,
an r /d ratio should be selected so K t values can be obtained. For the worst end
of the spectrum, with r /d = 0.02 and D /d = 1.5, K t values from the stress concentration charts for shoulders indicate 2.7 for bending, 2.2 for torsion, and 3.0 for
axial.
A keyway will produce a stress concentration near a critical point where the loadtransmitting component is located. The stress concentration in an end-milled keyseat
is a function of the ratio of the radius r at the bottom of the groove and the shaft
diameter d. For early stages of the design process, it is possible to estimate the stress
concentration for keyways regardless of the actual shaft dimensions by assuming a
typical ratio of r /d = 0.02. This gives K t = 2.2 for bending and K ts = 3.0 for torsion, assuming the key is in place.
Figures A1516 and A1517 give values for stress concentrations for atbottomed grooves such as used for retaining rings. By examining typical retaining
ring specications in vendor catalogs, it can be seen that the groove width is typically
slightly greater than the groove depth, and the radius at the bottom of the groove is
around 1/10 of the groove width. From Figs. A1516 and A1517, stress concentration factors for typical retaining ring dimensions are around 5 for bending and axial,
and 3 for torsion. Fortunately, the small radius will often lead to a smaller notch sensitivity, reducing K f .
Table 71 summarizes some typical stress concentration factors for the rst iteration in the design of a shaft. Similar estimates can be made for other features. The
point is to notice that stress concentrations are essentially normalized so that they are
dependent on ratios of geometry features, not on the specic dimensions. Consequently, by estimating the appropriate ratios, the rst iteration values for stress concentrations can be obtained. These values can be used for initial design, then actual
values inserted once diameters have been determined.
Table 71
First Iteration Estimates
for Stress Concentration
Factors Kt.
Warning: These factors are
only estimates for use when
actual dimensions are not
yet determined. Do not
use these once actual
dimensions are available.
Bending
Shoulder lletsharp (r/d
Shoulder lletwell rounded (r/d
End-mill keyseat (r/d
0.02)
0.1)
Axial
2.7
0.02)
Torsional
2.2
3.0
1.7
1.5
1.9
2.2
3.0
Sled runner keyseat
1.7
Retaining ring groove
5.0
3.0
5.0
Missing values in the table are not readily available.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
EXAMPLE 72
This example problem is part of a larger case study. See Chap. 18 for the full
context.
A double reduction gearbox design has developed to the point that the general layout and axial dimensions of the countershaft carrying two spur gears
has been proposed, as shown in Fig. 710. The gears and bearings are located
and supported by shoulders, and held in place by retaining rings. The gears
transmit torque through keys. Gears have been specified as shown, allowing the
tangential and radial forces transmitted through the gears to the shaft to be
determined as follows.
t
W23 = 540 lbf
t
W54 = 2431 lbf
r
W23 = 197 lbf
r
W54 = 885 lbf
where the superscripts t and r represent tangential and radial directions,
respectively; and, the subscripts 23 and 54 represent the forces exerted
by gears 2 and 5 (not shown) on gears 3 and 4, respectively.
Proceed with the next phase of the design, in which a suitable material
is selected, and appropriate diameters for each section of the shaft are
estimated, based on providing sufficient fatigue and static stress capacity
for infinite life of the shaft, with minimum safety factors of 1.5.
Bearing A
Bearing B
Gear 3
d3 12
Gear 4
d4 2.67
D5
Figure 710
Shaft layout for Example 72. Dimensions in inches.
KL
MBN
11.50
D7
11.25
J
10.25
I
9.50
9.75
3.50
H
8.50
G
D6
7.50
C A D EF
2.75
1.75
2.0
1.25
0.75
D4
D2
10.75
D3
D1
Datum
0.25
362
III. Design of Mechanical
Elements
365
366
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Solution
r
W23
Perform free body diagram
analysis to get reaction
forces at the bearings.
t
W23
y
r
W54
RBy
RAy
R Az = 115.0 lbf
R Ay = 356.7 lbf
t
W54
A
x
R Bz = 1776.0 lbf
R By = 725.3 lbf
G
I
RAz
J
B
K
RBz
z
From Mx , find the torque in
the shaft between the gears,
T
3240
t
T = W23 ( d3 /2) = 540 (12/2) =
3240 lbf · in
Generate shear-moment
diagrams for two planes.
V
655
115
1776
x-z Plane
3341
M
3996
2220
230
V
357
160
725
1472
x-y Plane
1632
M
713
907
3651
Combine orthogonal planes as
vectors to get total moments,
e.g. at J, 39962 + 16322 =
4316 lbf · in.
4316
MTOT
2398
749
Start with Point I, where the bending moment is high, there is a stress concentration at the shoulder, and the torque is present.
At I , Ma = 3651 lbf-in, Tm = 3240 lbf-in, Mm = Ta = 0
363
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
364
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
Assume generous fillet radius for gear at I.
From Table 71, estimate K t = 1.7, K ts = 1.5 . For quick, conservative first
pass, assume K f = K t , K f s = K ts .
Choose inexpensive steel, 1020 CD, with Sut = 68 kpsi. For Se ,
b
ka = aSut = 2.7(68) 0.265 = 0.883
Eq. (619)
Guess kb = 0.9. Check later when d is known.
kc = kd = ke = 1
Se = (0.883)(0.9)(0.5)(68) = 27.0 kpsi.
Eq. (618)
For first estimate of the small diameter at the shoulder at point I, use the
DE-Goodman criterion of Eq. (78). This criterion is good for the initial design,
since it is simple and conservative. With Mm = Ta = 0, Eq. (78) reduces to
16n 2 K M
3 K f s Tm
f
a
d=
+
π
Se
Sut
d=
16(1.5)
π
2 1/2
1/3
3 [(1.5) (3240)]2
2 (1.7) (3651)
+
27 000
68 000
1/2
1/3
d = 1.65 in.
All estimates have probably been conservative, so select the next standard size
below 1.65 in. and check, d
1.625 in.
A typical D/d ratio for support at a shoulder is D/d
1.2, thus, D
1.2(1.625)
1.95 in. Increase to D
2.0 in. A nominal 2 in. cold-drawn shaft diameter can
be used. Check if estimates were acceptable.
D /d = 2/1.625 = 1.23
=
Assume fillet radius r = d /10 0.16 in. r /d = 0.1
K t = 1.6 (Fig. A159), q = 0.82 (Fig. 620)
Eq. (632)
K f = 1 + 0.82(1.6 1) = 1.49
K ts = 1.35 (Fig. A158), qs = 0.95 (Fig. 621)
K f s = 1 + 0.95(1.35 1) = 1.33
ka = 0.883 (no change)
Eq. (620)
kb =
1.625
0.3
0.107
= 0.835
Se = (0.883)(0.835)(0.5)(68) = 25.1 kpsi
Eq. (75)
Eq. (76)
32 K f Ma
32(1.49)(3651)
=
= 12 910 psi 1
3
πd
π (1.625) 3
1/2
16 K f s Tm 2
3(16)(1.33)(3240)
σm = 3
=
= 8859 psi
π d3
π (1.625) 3
σa =
Using Goodman criterion
σ
σ
129 10
8859
1
= a+ m=
+
= 0.645
nf
Se
Sut
25 100 68 000
n f = 1.55
367
368
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Note that we could have used Eq. (77) directly.
Check yielding.
ny =
Sy
Sy
57 000
= 2.62
>
=
σmax
σa + σm
12 910 + 8859
Also check this diameter at the end of the keyway, just to the right of point I,
and at the groove at point K. From moment diagram, estimate M at end of
keyway to be M
3750 lbf-in.
Assume the radius at the bottom of the keyway will be the standard
rd
0.02, r
0.02 d
0.02 (1.625)
0.0325 in.
K t = 2.14 (Fig. A1518), q
0.65 (Fig. 620)
K f = 1 + 0.65(2.14 1) = 1.74
K ts = 3.0 (Fig. A1519), qs = 0.9 (Fig. 621)
K f s = 1 + 0.9(3 1) = 2.8
32 K f Ma
32(1.74)(3750)
σa =
=
= 15 490 psi
3
πd
π (1.625) 3
K f s Tm
3(16)(2.8)(3240)
σm = 3(16)
=
= 18 650 psi
π d3
π (1.625) 3
σ
σ
15 490 18 650
1
+
= 0.891
= a+ m=
nf
Se
Sut
25 100 68 000
n f = 1.12
The keyway turns out to be more critical than the shoulder. We can either
increase the diameter, or use a higher strength material. Unless the deflection
analysis shows a need for larger diameters, let us choose to increase the
strength. We started with a very low strength, and can afford to increase it
some to avoid larger sizes. Try 1050 CD, with Sut = 100 kpsi.
Recalculate factors affected by Sut , i.e. ka Se ; q K f σa
ka = 2.7(100) 0.265 = 0.797,
Se = 0.797(0.835)(0.5)(100) = 33.3 kpsi
q = 0.72, K f = 1 + 0.72(2.14 1) = 1.82
32(1.82)(3750)
= 16 200 psi
π (1.625) 3
1
18 650
16 200
+
= 0.673
=
nf
33 300 100 000
σa =
n f = 1.49
Since the Goodman criterion is conservative, we will accept this as close enough
to the requested 1.5.
Check at the groove at K, since K t for flat-bottomed grooves are often very
high. From the torque diagram, note that no torque is present at the groove.
From the moment diagram, Ma = 2398 lbf in, Mm = Ta = Tm = 0 . To quickly
check if this location is potentially critical just use K f = K t = 5.0 as an
estimate, from Table 71.
σa =
32 K f Ma
32(5)(2398)
=
= 28 460 psi
3
πd
π (1.625) 3
nf =
Se
33 300
=
= 1.17
σa
28 460
365
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
366
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
This is low. We will look up data for a specific retaining ring to obtain K f more
accurately. With a quick on-line search of a retaining ring specification using the
website www.globalspec.com, appropriate groove specifications for a retaining ring
for a shaft diameter of 1.625 in are obtained as follows: width, a = 0.068 in;
depth, t = 0.048 in; and corner radius at bottom of groove, r = 0.01in. From
Fig. A1516, with r / t = 0.01/0.048 = 0.208 , and a / t = 0.068/0.048 = 1.42
K t = 4.3 , q = 0.65 (Fig. 620)
K f = 1 + 0.65(4.3 1) = 3.15
32 K f Ma
32(3.15)(2398)
=
= 17 930 psi
π d3
π (1.625) 3
Se
33 300
= 1.86
=
nf =
σa
17 930
σa =
Quickly check if point M might be critical. Only bending is present, and the
moment is small, but the diameter is small and the stress concentration is high
for a sharp fillet required for a bearing. From the moment diagram,
Ma = 959 lbf · in, and Mm = Tm = Ta = 0 .
Estimate K t = 2.7 from Table 71, d = 1.0 in, and fillet radius r to fit a
typical bearing.
r /d = 0.02, r = 0.02(1) = 0.02
q = 0.7 (Fig. 620)
K f = 1 + (0.7)(2.7 1) = 2.19
32 K f Ma
32(2.19)(959)
=
= 21 390 psi
σa =
π d3
π (1) 3
nf =
Se
33 300
= 1.56
=
σa
21 390
Should be OK. Close enough to recheck after bearing is selected.
With the diameters specified for the critical locations, fill in trial values for
the rest of the diameters, taking into account typical shoulder heights for
bearing and gear support.
D1 = D7 = 1.0 in
D2 = D6 = 1.4 in
D3 = D5 = 1.625 in
D4 = 2.0 in
The bending moments are much less on the left end of shaft, so D1 , D2 , and D3
could be smaller. However, unless weight is an issue, there is little advantage to
requiring more material removal. Also, the extra rigidity may be needed to keep
deflections small.
369
370
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Table 72
367
Slopes
Typical Maximum
Ranges for Slopes and
Transverse Deections
0.00050.0012 rad
Tapered roller
Cylindrical roller
0.00080.0012 rad
Deep-groove ball
0.0010.003 rad
Spherical ball
0.0260.052 rad
Self-align ball
0.0260.052 rad
Uncrowned spur gear
0.0005 rad
Transverse deections
Spur gears with P < 10 teeth/in
0.010 in
Spur gears with 11 < P < 19
75
0.005 in
Spur gears with 20 < P < 50
0.003 in
Deection Considerations
Deection analysis at even a single point of interest requires complete geometry information for the entire shaft. For this reason, it is desirable to design the dimensions at
critical locations to handle the stresses, and ll in reasonable estimates for all other
dimensions, before performing a deection analysis. Deection of the shaft, both linear and angular, should be checked at gears and bearings. Allowable deections will
depend on many factors, and bearing and gear catalogs should be used for guidance
on allowable misalignment for specic bearings and gears. As a rough guideline, typical ranges for maximum slopes and transverse deections of the shaft centerline are
given in Table 72. The allowable transverse deections for spur gears are dependent
on the size of the teeth, as represented by the diametral pitch P
number of
teeth/pitch diameter.
In Sec. 44 several beam deection methods are described. For shafts, where the
deections may be sought at a number of different points, integration using either
singularity functions or numerical integration is practical. In a stepped shaft, the crosssectional properties change along the shaft at each step, increasing the complexity of
integration, since both M and I vary. Fortunately, only the gross geometric dimensions
need to be included, as the local factors such as llets, grooves, and keyways do not
have much impact on deection. Example 47 demonstrates the use of singularity
functions for a stepped shaft. Many shafts will include forces in multiple planes,
requiring either a three dimensional analysis, or the use of superposition to obtain
deections in two planes which can then be summed as vectors.
A deection analysis is straightforward, but it is lengthy and tedious to carry out
manually, particularly for multiple points of interest. Consequently, practically all
shaft deflection analysis will be evaluated with the assistance of software. Any
general-purpose finite-element software can readily handle a shaft problem (see
Chap. 19). This is practical if the designer is already familiar with using the software
and with how to properly model the shaft. Special-purpose software solutions for
3-D shaft analysis are available, but somewhat expensive if only used occasionally.
Software requiring very little training is readily available for planar beam analysis,
and can be downloaded from the internet. Example 73 demonstrates how to incorporate such a program for a shaft with forces in multiple planes.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
368
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
EXAMPLE 73
This example problem is part of a larger case study. See Chap. 18 for the full
context.
In Example 72 a preliminary shaft geometry was obtained on the basis of
design for stress. The resulting shaft is shown in Fig. 710, with proposed
diameters of
D1 = D7 = 1 in
D2 = D6 = 1.4 in
D3 = D5 = 1.625 in
D4 = 2.0 in
Check that the deflections and slopes at the gears and bearings are acceptable.
If necessary, propose changes in the geometry to resolve any problems.
Solution
A simple planar beam analysis program will be used. By modeling the shaft
twice, with loads in two orthogonal planes, and combining the results, the shaft
deflections can readily be obtained. For both planes, the material is selected
(steel with E = 30 Mpsi), the shaft lengths and diameters are entered, and the
bearing locations are specified. Local details like grooves and keyways are
ignored, as they will have insignificant effect on the deflections. Then the tangential gear forces are entered in the horizontal xz plane model, and the radial
gear forces are entered in the vertical xy plane model. The software can calculate the bearing reaction forces, and numerically integrate to generate plots
for shear, moment, slope, and deflection, as shown in Fig. 711.
xy plane
xz plane
Beam length: 11.5 in
Beam length: 11.5 in
in
Deflection
in
Deflection
deg
Slope
deg
Slope
lbf-in
Moment
lbf-in
Moment
lbf
Shear
lbf
Shear
Figure 711
Shear, moment, slope, and deection plots from two planes. (Source: Beam 2D Stress Analysis, Orand
Systems, Inc.)
371
372
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Point of interest
xz plane
xy plane
Total
Left bearing slope
0.02263 deg
0.01770 deg
0.02872 deg
Right bearing slope
0.05711 deg
0.02599 deg
0.06274 deg
Left gear slope
0.02067 deg
0.01162 deg
0.02371 deg
Right gear slope
0.02155 deg
0.01149 deg
0.02442 deg
Left gear deection
0.0007568 in
0.0005153 in
0.0009155 in
Right gear deection
0.0015870 in
0.0007535 in
369
0.0017567 in
0.000501 rad
0.001095 rad
0.000414 rad
0.000426 rad
Table 73
Slope and Deection Values at Key Locations
The deflections and slopes at points of interest are obtained from the plots,
2
2
and combined with orthogonal vector addition, that is, δ = δx z + δx y . Results
are shown in Table 73.
Whether these values are acceptable will depend on the specific bearings
and gears selected, as well as the level of performance expected. According
to the guidelines in Table 72, all of the bearing slopes are well below typical
limits for ball bearings. The right bearing slope is within the typical range for
cylindrical bearings. Since the load on the right bearing is relatively high, a
cylindrical bearing might be used. This constraint should be checked against
the specific bearing specifications once the bearing is selected.
The gear slopes and deflections more than satisfy the limits recommended
in Table 72. It is recommended to proceed with the design, with an
awareness that changes that reduce rigidity should warrant another
deflection check.
Once deections at various points have been determined, if any value is larger
than the allowable deection at that point, a new diameter can be found from
n d yold 1/4
dnew = dold
(717)
yall
where yall is the allowable deection at that station and n d is the design factor. Similarly,
if any slope is larger than the allowable slope θall , a new diameter can be found from
n d ( dy /dx ) old 1/4
dnew = dold
(718)
(slope) all
where (slope)all is the allowable slope. As a result of these calculations, determine the
largest dnew /dold ratio, then multiply all diameters by this ratio. The tight constraint
will be just tight, and all others will be loose. Dont be too concerned about end journal sizes, as their inuence is usually negligible. The beauty of the method is that the
deections need to be completed just once and constraints can be rendered loose but
for one, with diameters all identied without reworking every deection.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
370
III. Design of Mechanical
Elements
373
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
EXAMPLE 74
Solution
For the shaft in Example 73, it was noted that the slope at the right bearing is near
the limit for a cylindrical roller bearing. Determine an appropriate increase in diameters to bring this slope down to 0.0005 rad.
Applying Eq. (717) to the deection at the right bearing gives
dnew = dold
n d slopeold
slopeall
1/4
= 1.0
( 1)(0.001095)
(0.0005)
1/4
= 1.216 in
Multiplying all diameters by the ratio
dnew
1.216
= 1.216
=
dold
1.0
gives a new set of diameters,
D1 = D7 = 1.216 in
D2 = D6 = 1.702 in
D3 = D5 = 1.976 in
D4 = 2.432 in
Repeating the beam deection analysis of Example 73 with these new diameters produces a slope at the right bearing of 0.0005 in, with all other deections less than
their previous values.
The transverse shear V at a section of a beam in exure imposes a shearing deection, which is superposed on the bending deection. Usually such shearing deection
is less than 1 percent of the transverse bending deection, and it is seldom evaluated.
However, when the shaft length-to-diameter ratio is less than 10, the shear component of transverse deection merits attention. There are many short shafts. A tabular
method is explained in detail elsewhere2, including examples.
For right-circular cylindrical shafts in torsion the angular deection θ is given in
Eq. (45). For a stepped shaft with individual cylinder length li and torque Ti , the
angular deection can be estimated from
θ=
θi =
Ti li
G i Ji
(719)
or, for a constant torque throughout homogeneous material, from
θ=
T
G
li
Ji
(720)
This should be treated only as an estimate, since experimental evidence shows that
the actual θ is larger than given by Eqs. (719) and (720).3
2
C.R. Mischke, Tabular Method for Transverse Shear Deection, Sec. 17.3 in Joseph E. Shigley, Charles
R. Mischke, and Thomas H. Brown, Jr. (eds.), Standard Handbook of Machine Design, 3rd ed., McGrawHill, New York, 2004.
3
R. Bruce Hopkins, Design Analysis of Shafts and Beams, McGraw-Hill, New York, 1970, pp. 9399.
374
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
371
Shafts and Shaft Components
If torsional stiffness is defined as ki = Ti /θi and, since θi = Ti / ki and
θ = θi = ( Ti / ki ), for constant torque θ = T ( 1/ ki ), it follows that the torsional stiffness of the shaft k in terms of segment stiffnesses is
1
1
=
(721)
k
ki
76
Critical Speeds for Shafts
When a shaft is turning, eccentricity causes a centrifugal force deection, which is
resisted by the shafts exural rigidity E I . As long as deections are small, no harm
is done. Another potential problem, however, is called critical speeds: at certain speeds
the shaft is unstable, with deections increasing without upper bound. It is fortunate
that although the dynamic deection shape is unknown, using a static deection curve
gives an excellent estimate of the lowest critical speed. Such a curve meets the boundary condition of the differential equation (zero moment and deection at both bearings) and the shaft energy is not particularly sensitive to the exact shape of the deection curve. Designers seek rst critical speeds at least twice the operating speed.
The shaft, because of its own mass, has a critical speed. The ensemble of attachments to a shaft likewise has a critical speed that is much lower than the shafts intrinsic critical speed. Estimating these critical speeds (and harmonics) is a task of the
designer. When geometry is simple, as in a shaft of uniform diameter, simply
supported, the task is easy. It can be expressed4 as
ω1 =
π
l
2
EI
=
m
π
l
2
gE I
Aγ
(722)
where m is the mass per unit length, A the cross-sectional area, and γ the specic
weight. For an ensemble of attachments, Rayleighs method for lumped masses gives5
ω1 =
g
wi yi
wi yi2
(723)
where wi is the weight of the ith location and yi is the deection at the ith body location. It is possible to use Eq. (723) for the case of Eq. (722) by partitioning the shaft
into segments and placing its weight force at the segment centroid as seen in Fig. 712.
Figure 712
y
(a) A uniform-diameter shaft for
Eq. (722). (b) A segmented
uniform-diameter shaft for
Eq. (723).
x
(a)
y
x
(b)
4
William T. Thomson and Marie Dillon Dahleh, Theory of Vibration with Applications, Prentice Hall,
5th ed., 1998, p. 273.
5
Thomson, op. cit., p. 357.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
372
III. Design of Mechanical
Elements
375
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
y
Figure 713
Unit load
aj
The inuence coefcient δi j is
the deection at i due to a unit
load at j.
bj
xi
x
l
Computer assistance is often used to lessen the difculty in nding transverse deections of a stepped shaft. Rayleighs equation overestimates the critical speed.
To counter the increasing complexity of detail, we adopt a useful viewpoint. Inasmuch as the shaft is an elastic body, we can use inuence coefcients. An inuence
coefcient is the transverse deection at location i on a shaft due to a unit load at
location j on the shaft. From Table A96 we obtain, for a simply supported beam
with a single unit load as shown in Fig. 713,
b j xi l 2 b2 x 2
j
i
6E I l
δi j = a (l x )
i
j
2
2
6 E I l 2lxi a j xi
xi ai
(724)
xi > ai
For three loads the inuence coefcients may be displayed as
j
i
1
2
3
1
δ11
δ12
δ13
2
δ21
δ22
δ23
3
δ31
δ32
δ33
Maxwells reciprocity theorem6 states that there is a symmetry about the main diagonal, composed of δ11 , δ22 , and δ33 , of the form δi j = δ ji . This relation reduces the
work of nding the inuence coefcients. From the inuence coefcients above, one
can nd the deections y1 , y2 , and y3 of Eq. (723) as follows:
y1 = F1 δ11 + F2 δ12 + F3 δ13
y2 = F1 δ21 + F2 δ22 + F3 δ23
y3 = F1 δ31 + F2 δ32 + F3 δ33
(725)
The forces Fi can arise from weight attached wi or centrifugal forces m i ω2 yi . The
equation set (725) written with inertial forces can be displayed as
y1 = m 1 ω2 y1 δ11 + m 2 ω2 y2 δ12 + m 3 ω2 y3 δ13
y2 = m 1 ω2 y1 δ21 + m 2 ω2 y2 δ22 + m 3 ω2 y3 δ23
y3 = m 1 ω2 y1 δ31 + m 2 ω2 y2 δ32 + m 3 ω2 y3 δ33
6
Thomson, op. cit., p. 167.
376
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
373
which can be rewritten as
(m 1 δ11 1/ω2 ) y1 + (m 2 δ12 ) y2 + (m 3 δ13 ) y3 = 0
(m 1 δ21 ) y1 + (m 2 δ22 1/ω2 ) y2 + (m 3 δ23 ) y3 = 0
(a )
(m 1 δ31 ) y1 + (m 2 δ32 ) y2 + (m 3 δ33 1/ω2 ) y3 = 0
Equation set (a) is three simultaneous equations in terms of y1 , y2 , and y3 . To avoid
the trivial solution y1 = y2 = y3 = 0, the determinant of the coefcients of y1 , y2 , and
y3 must be zero (eigenvalue problem). Thus,
m 2 δ12
m 3 δ13
(m 1 δ11 1/2 )
2
=0
m 1 δ21
(m 2 δ22 1/ω )
m 3 δ23
2
m 1 δ31
m 2 δ32
(m 3 δ33 1/ω )
(726)
which says that a deection other than zero exists only at three distinct values of ω,
the critical speeds. Expanding the determinant, we obtain
1
ω2
3
(m 1 δ11 + m 2 δ22 + m 3 δ33 )
1
ω2
2
+ ··· = 0
(727)
2
2
2
The three roots of Eq. (727) can be expressed as 1/ω1 , 1/ω2 , and 1/ω3 . Thus
Eq. (727) can be written in the form
1
1
2
ω2
ω1
1
1
2
ω2
ω2
1
1
2
ω2
ω3
=0
or
1
ω2
3
1
ω2
1
1
1
+ 2+ 2
2
ω1
ω2
ω3
2
+ ··· = 0
(728)
Comparing Eqs. (727) and (728) we see that
1
1
1
+ 2 + 2 = m 1 δ11 + m 2 δ22 + m 3 δ33
2
ω1
ω2
ω3
(729)
If we had only a single mass m 1 alone, the critical speed would be given by 1/ω2 =
m 1 δ11 . Denote this critical speed as ω11 (which considers only m 1 acting alone). Like2
wise for m 2 or m 3 acting alone, we similarly dene the terms 1/ω22 = m 2 δ22 or
2
1/ω33 = m 3 δ33 , respectively. Thus, Eq. (729) can be rewritten as
1
1
1
1
1
1
+ 2+ 2= 2 + 2 + 2
2
ω1
ω2
ω3
ω11
ω22
ω33
(730)
2
2
2
If we order the critical speeds such that ω1 < ω2 < ω3 , then 1/ω1 1/ω2 , and 1/ω3 .
So the rst, or fundamental, critical speed ω1 can be approximated by
1
1
1.1
= 2+ 2+ 2
2
ω1
ω11
ω22
ω33
(731)
This idea can be extended to an n-body shaft:
1.
=
2
ω1
n
1=1
1
2
ωii
(732)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
374
III. Design of Mechanical
Elements
377
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
This is called Dunkerleys equation. By ignoring the higher mode term(s), the rst
critical speed estimate is lower than actually is the case.
Since Eq. (732) has no loads appearing in the equation, it follows that if each
load could be placed at some convenient location transformed into an equivalent load,
then the critical speed of an array of loads could be found by summing the equivalent loads, all placed at a single convenient location. For the load at station 1, placed
at the center of span, denoted with the subscript c, the equivalent load is found from
2
ω11 =
g
g
=
w1 δ11
w1c δcc
or
w1c = w1
EXAMPLE 75
Solution
δ11
δcc
Consider a simply supported steel shaft as depicted in Fig. 714, with 1 in diameter
and a 31-in span between bearings, carrying two gears weighing 35 and 55 lbf.
(a) Find the inuence coefcients.
w y and
w y 2 and the rst critical speed using Rayleighs equation,
(b) Find
Eq. (723).
(c) From the inuence coefcients, nd ω11 and ω22 .
(d) Using Dunkerleys equation, Eq. (732), estimate the rst critical speed.
(e) Use superposition to estimate the rst critical speed.
( f ) Estimate the shafts intrinsic critical speed. Suggest a modication to Dunkerleys
equation to include the effect of the shafts mass on the rst critical speed of the
attachments.
I=
(a)
π(1)4
π d4
=
= 0.049 09 in4
64
64
6 E I l = 6(30)106 (0.049 09)31 = 0.2739(109 ) lbf · in3
Figure 714
(733)
y
w1 = 35 lbf
(a) A 1-in uniform-diameter
shaft for Ex. 75.
(b) Superposing of equivalent
loads at the center of the shaft
for the purpose of nding the
rst critical speed.
7 in
w2 = 55 lbf
13 in
11 in
x
31 in
(a )
y
w1c
17.1 lbf
w2c
46.1 lbf
15.5 in
15.5 in
x
(b )
378
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
375
From Eq. set (724),
δ11 =
24(7)(312 242 72 )
= 2.061(104 ) in/lbf
0.2739(109 )
δ22 =
11(20)(312 112 202 )
= 3.534(104 ) in/lbf
0.2739(109 )
δ12 = δ21 =
Answer
11(7)(312 112 72 )
= 2.224(104 ) in/lbf
0.2739(109 )
j
i
1
1
2.061(10 4)
2
2
2.224(10 4)
4
3.534(10 4)
2.224(10 )
y1 = w1 δ11 + w2 δ12 = 35(2.061)104 + 55(2.224)104 = 0.019 45 in
y2 = w1 δ21 + w2 δ22 = 35(2.224)104 + 55(3.534)104 = 0.027 22 in
wi yi = 35(0.019 45) + 55(0.027 22) = 2.178 lbf · in
(b)
wi yi2 = 35(0.019 45)2 + 55(0.027 22)2 = 0.053 99 lbf · in2
Answer
Answer
ω=
386.1(2.178)
= 124.8 rad/s , or 1192 rev/min
0.053 99
(c)
w1
1
δ11
=
2
g
ω11
Answer
ω11 =
386.1
= 231.4 rad/s, or 2210 rev/min
35(2.061)104
ω22 =
Answer
g
=
w1 δ11
g
=
w2 δ22
386.1
= 140.9 rad/s, or 1346 rev/min
55(3.534)104
1.
=
2
ω1
(d )
.
ω1 =
Answer
1
1
1
=
+
= 6.905(105 )
2
2
231.4
140.92
ωii
1
= 120.3 rad/s, or 1149 rev/min
6.905(105 )
which is less than part b, as expected.
(e) From Eq. (724),
2
2
bcc xcc l 2 bcc xcc
15.5(15.5)(312 15.52 15.52 )
=
6E I l
0.2739(109 )
4
= 4.215(10 ) in/lbf
δcc =
(1)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
376
III. Design of Mechanical
Elements
379
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
From Eq. (733),
w1c = w1
w2c = w2
Answer
ω=
g
δcc
wic
2.061(104 )
δ11
= 17.11 lbf
= 35
δcc
4.215(104 )
3.534(104 )
δ22
= 46.11 lbf
= 55
δcc
4.215(104 )
386.1
= 120.4 rad/s, or 1150 rev/min
4.215(104 )(17.11 + 46.11)
=
which, except for rounding, agrees with part d, as expected.
( f ) For the shaft, E = 30(106 ) psi, γ = 0.282 lbf/in3, and A = π(12 )/4 = 0.7854 in2.
Considering the shaft alone, the critical speed, from Eq. (722), is
Answer
ωs =
π
l
2
gE I
=
Aγ
π
31
2
386.1(30)106 (0.049 09)
0.7854(0.282)
= 520.4 rad/s, or 4970 rev/min
2
We can simply add 1/ωs to the right side of Dunkerleys equation, Eq. (1), to include
the shafts contribution,
Answer
1
1.
=
+ 6.905(105 ) = 7.274(105 )
2
520.42
ω1
.
ω1 = 117.3 rad/s, or 1120 rev/min
which is slightly less than part d, as expected.
The shafts rst critical speed ωs is just one more single effect to add to Dunkerleys equation. Since it does not t into the summation, it is usually written up front.
Answer
1.1
= 2+
2
ωs
ω1
n
i =1
1
2
ωii
(734)
Common shafts are complicated by the stepped-cylinder geometry, which makes the
inuence-coefcient determination part of a numerical solution.
77
Miscellaneous Shaft Components
Setscrews
Unlike bolts and cap screws, which depend on tension to develop a clamping force, the
setscrew depends on compression to develop the clamping force. The resistance to axial
motion of the collar or hub relative to the shaft is called holding power. This holding
power, which is really a force resistance, is due to frictional resistance of the contacting
portions of the collar and shaft as well as any slight penetration of the setscrew into the
shaft.
380
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
377
Figure 715 shows the point types available with socket setscrews. These are also
manufactured with screwdriver slots and with square heads.
Table 74 lists values of the seating torque and the corresponding holding power
for inch-series setscrews. The values listed apply to both axial holding power, for
Figure 715
Socket setscrews: (a) at point;
(b) cup point; (c) oval point;
(d ) cone point; (e) half-dog
point.
L
L
L
T
T
D
T
D
(a)
D
(b)
(c)
L
L
T
T
D
(d)
Table 74
Typical Holding Power
(Force) for Socket
Setscrews*
Source: Unbrako Division, SPS
Technologies, Jenkintown, Pa.
Size,
in
P
D
Seating
Torque,
lbf . in
(e)
Holding
Power,
lbf
#0
1.0
50
#1
1.8
65
#2
1.8
85
#3
5
120
#4
5
160
#5
10
200
#6
10
250
#8
20
385
#10
36
540
1
4
5
16
87
1000
165
1500
3
8
290
2000
7
16
430
2500
1
2
620
3000
9
16
620
3500
5
8
1325
4000
3
4
7
8
2400
5000
5200
6000
1
7200
7000
*Based on alloy-steel screw against steel shaft, class 3A coarse or
ne threads in class 2B holes, and cup-point socket setscrews.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
378
III. Design of Mechanical
Elements
381
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
resisting thrust, and the tangential holding power, for resisting torsion. Typical factors
of safety are 1.5 to 2.0 for static loads and 4 to 8 for various dynamic loads.
Setscrews should have a length of about half of the shaft diameter. Note that this
practice also provides a rough rule for the radial thickness of a hub or collar.
Keys and Pins
Keys and pins are used on shafts to secure rotating elements, such as gears, pulleys,
or other wheels. Keys are used to enable the transmission of torque from the shaft to
the shaft-supported element. Pins are used for axial positioning and for the transfer
of torque or thrust or both.
Figure 716 shows a variety of keys and pins. Pins are useful when the principal loading is shear and when both torsion and thrust are present. Taper pins are sized
according to the diameter at the large end. Some of the most useful sizes of these are
listed in Table 75. The diameter at the small end is
(735)
d = D 0.0208 L
where d
D
L
diameter at small end, in
diameter at large end, in
length, in
Figure 716
(a) Square key; (b) round key;
(c and d ) round pins; (e) taper
pin; (f ) split tubular spring pin.
The pins in parts (e) and
(f ) are shown longer than
necessary, to illustrate the
chamfer on the ends, but their
lengths should be kept smaller
than the hub diameters to
prevent injuries due to
projections on rotating parts.
(a)
Dimensions at Large End
of Some Standard Taper
PinsInch Series
(c)
(d )
Table 75
(b )
(e)
( f)
Commercial
Precision
Size
Maximum
Minimum
Maximum
Minimum
4/0
0.1103
0.1083
0.1100
0.1090
2/0
0.1423
0.1403
0.1420
0.1410
0
0.1573
0.1553
0.1570
0.1560
2
0.1943
0.1923
0.1940
0.1930
4
0.2513
0.2493
0.2510
0.2500
6
0.3423
0.3403
0.3420
0.3410
8
0.4933
0.4913
0.4930
0.4920
382
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Table 76
Shaft Diameter
Key Size
Over
w
h
Keyway Depth
7
16
3
32
3
32
7
16
9
16
1
8
3
32
1
8
1
8
3
64
3
64
1
16
3
16
1
8
1
16
3
16
Source: Joseph E. Shigley,
Unthreaded Fasteners,
Chap. 24 in Joseph E.
Shigley, Charles R. Mischke,
and Thomas H. Brown, Jr.
(eds.), Standard Handbook of
Machine Design, 3rd ed.,
McGraw-Hill, New York,
2004.
To (Incl.)
5
16
Inch Dimensions for
Some Standard Squareand Rectangular-Key
Applications
379
3
16
3
32
1
4
1
4
5
16
3
16
3
32
1
4
1
4
5
16
1
8
9
16
7
8
11
4
7
8
1
14
3
18
5
16
5
32
3
8
21
4
23
4
1
24
3
24
1
34
3
16
1
2
3
8
3
16
1
2
5
8
7
16
1
4
7
32
5
8
13
4
1
4
3
8
1
2
13
8
3
14
1
8
5
8
5
16
3
4
3
4
1
2
1
4
3
8
3
8
3
4
1
8
For less important applications, a dowel pin or a drive pin can be used. A large
variety of these are listed in manufacturers catalogs.7
The square key, shown in Fig. 716a, is also available in rectangular sizes. Standard sizes of these, together with the range of applicable shaft diameters, are listed in
Table 76. The shaft diameter determines standard sizes for width, height, and key depth.
The designer chooses an appropriate key length to carry the torsional load. Failure of
the key can be by direct shear, or by bearing stress. Example 76 demonstrates the
process to size the length of a key. The maximum length of a key is limited by the hub
length of the attached element, and should generally not exceed about 1.5 times the
shaft diameter to avoid excessive twisting with the angular deection of the shaft. Mulo
tiple keys may be used as necessary to carry greater loads, typically oriented at 90 from
one another. Excessive safety factors should be avoided in key design, since it is desirable in an overload situation for the key to fail, rather than more costly components.
Stock key material is typically made from low carbon cold-rolled steel, and is
manufactured such that its dimensions never exceed the nominal dimension. This
allows standard cutter sizes to be used for the keyseats. A setscrew is sometimes used
along with a key to hold the hub axially, and to minimize rotational backlash when
the shaft rotates in both directions.
7
See also Joseph E. Shigley, Unthreaded Fasteners, Chap. 24. In Joseph E. Shigley, Charles R. Mischke, and
Thomas H. Brown, Jr. (eds.), Standard Handbook of Machine Design, 3rd ed., McGraw-Hill, New York, 2004.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
380
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
Figure 717
(a) Gib-head key;
(b) Woodruff key.
1
Taper 8 " in 12"
w
w
h
(a )
D
w
(b )
The gib-head key, in Fig. 717a, is tapered so that, when rmly driven, it acts to
prevent relative axial motion. This also gives the advantage that the hub position can
be adjusted for the best axial location. The head makes removal possible without
access to the other end, but the projection may be hazardous.
The Woodruff key, shown in Fig. 717b, is of general usefulness, especially when a
wheel is to be positioned against a shaft shoulder, since the keyslot need not be machined
into the shoulder stress-concentration region. The use of the Woodruff key also yields
better concentricity after assembly of the wheel and shaft. This is especially important at
high speeds, as, for example, with a turbine wheel and shaft. Woodruff keys are particularly useful in smaller shafts where their deeper penetration helps prevent key rolling.
Dimensions for some standard Woodruff key sizes can be found in Table 77, and Table
78 gives the shaft diameters for which the different keyseat widths are suitable.
Pilkey8 gives values for stress concentrations in an end-milled keyseat, as a function of the ratio of the radius r at the bottom of the groove and the shaft diameter d.
For fillets cut by standard milling-machine cutters, with a ratio of r /d = 0.02,
Petersons charts give K t = 2.14 for bending and K ts = 2.62 for torsion without the
key in place, or K ts = 3.0 for torsion with the key in place. The stress concentration
at the end of the keyseat can be reduced somewhat by using a sled-runner keyseat,
eliminating the abrupt end to the keyseat, as shown in Fig. 717. It does, however,
still have the sharp radius in the bottom of the groove on the sides. The sled-runner
keyseat can only be used when denite longitudinal key positioning is not necessary.
It is also not as suitable near a shoulder. Keeping the end of a keyseat at least a distance
8
W. D. Pilkey, Petersons Stress Concentration Factors, 2nd ed., John Wiley & Sons, New York, 1997,
pp. 408409.
383
384
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
Table 77
Dimensions of Woodruff
KeysInch Series
Table 78
Sizes of Woodruff Keys
Suitable for Various
Shaft Diameters
Key Size
Height
b
Offset
e
1
4
3
8
3
8
1
2
5
8
1
2
5
8
3
4
5
8
3
4
7
8
3
4
7
8
0.109
1
0.438
7
8
0.375
1
64
1
64
1
64
3
64
1
16
3
64
1
16
1
16
1
16
1
16
1
16
1
16
1
16
1
16
1
16
1
16
5
64
1
16
5
64
7
64
5
64
7
64
w
D
1
16
1
16
3
32
3
32
3
32
1
8
1
8
1
8
5
32
5
32
5
32
3
16
3
16
3
16
1
4
1
4
1
4
5
16
5
16
5
16
3
8
3
8
0.172
0.172
0.203
0.250
0.203
0.250
0.313
0.250
0.313
0.375
0.313
0.375
1
0.438
11
4
0.547
1
0.438
11
4
0.547
11
2
0.641
11
4
0.547
11
2
0.641
Keyseat
Width, in
1
16
3
32
1
8
5
32
3
16
1
4
5
16
3
8
Shaft Diameter, in
From
To (inclusive)
5
16
3
8
3
8
1
2
9
16
11
16
3
4
1
2
7
8
11
2
15
8
1
25
8
2
21
4
23
8
Keyseat Depth
Shaft
Hub
0.0728
0.0372
0.1358
0.0372
0.1202
0.0529
0.1511
0.0529
0.1981
0.0529
0.1355
0.0685
0.1825
0.0685
0.2455
0.0685
0.1669
0.0841
0.2299
0.0841
0.2919
0.0841
0.2143
0.0997
0.2763
0.0997
0.3393
0.0997
0.2450
0.1310
0.3080
0.1310
0.4170
0.1310
0.2768
0.1622
0.3858
0.1622
0.4798
0.1622
0.3545
0.1935
0.4485
0.1935
381
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
382
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
Figure 718
Typical uses for retaining rings.
(a) External ring and (b) its
application; (c) internal ring
and (d ) its application.
Retaining ring
Retaining ring
(a )
(b )
(c)
(d)
of d/10 from the start of the shoulder llet will prevent the two stress concentrations
from combining with each other.9
Retaining Rings
A retaining ring is frequently used instead of a shaft shoulder or a sleeve to axially
position a component on a shaft or in a housing bore. As shown in Fig. 718, a groove
is cut in the shaft or bore to receive the spring retainer. For sizes, dimensions, and
axial load ratings, the manufacturers catalogs should be consulted.
Appendix Tables A1516 and A1517 give values for stress concentration factors for at-bottomed grooves in shafts, suitable for retaining rings. For the rings to
seat nicely in the bottom of the groove, and support axial loads against the sides of
the groove, the radius in the bottom of the groove must be reasonably sharp, typically
about one-tenth of the groove width. This causes comparatively high values for stress
concentration factors, around 5 for bending and axial, and 3 for torsion. Care should
be taken in using retaining rings, particularly in locations with high bending stresses.
9
Ibid, p. 381.
EXAMPLE 76
A UNS G10350 steel shaft, heat-treated to a minimum yield strength of 75 kpsi, has
7
a diameter of 1 16 in. The shaft rotates at 600 rev/min and transmits 40 hp through a
gear. Select an appropriate key for the gear.
Solution
A 3 -in square key is selected, UNS G10200 cold-drawn steel being used. The design
8
will be based on a yield strength of 65 kpsi. A factor of safety of 2.80 will be
employed in the absence of exact information about the nature of the load.
The torque is obtained from the horsepower equation
t
a
F
F
b
r
T=
63 025 H
(63 025)(40)
=
= 4200 lbf · in
n
600
From Fig. 719, the force F at the surface of the shaft is
F=
4200
T
=
= 5850 lbf
r
1.4375/2
By the distortion-energy theory, the shear strength is
Figure 719
Ssy = 0.577 Sy = (0.577)(65) = 37.5 kpsi
385
386
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
383
Failure by shear across the area ab will create a stress of τ = F / tl . Substituting the
strength divided by the factor of safety for τ gives
Ssy
F
=
n
tl
or
37.5(10) 3
5850
=
2.80
0.375l
or l = 1.16 in. To resist crushing, the area of one-half the face of the key is used:
Sy
F
=
n
tl /2
or
65(10) 3
5850
=
2.80
0.375l /2
and l = 1.34 in. The hub length of a gear is usually greater than the shaft diameter,
for stability. If the key, in this example, is made equal in length to the hub, it would
7
therefore have ample strength, since it would probably be 1 16 in or longer.
78
Limits and Fits
The designer is free to adopt any geometry of t for shafts and holes that will ensure
the intended function. There is sufcient accumulated experience with commonly recurring situations to make standards useful. There are two standards for limits and ts in
the United States, one based on inch units and the other based on metric units.10 These
differ in nomenclature, denitions, and organization. No point would be served by separately studying each of the two systems. The metric version is the newer of the two
and is well organized, and so here we present only the metric version but include a set
of inch conversions to enable the same system to be used with either system of units.
In using the standard, capital letters always refer to the hole; lowercase letters
are used for the shaft.
The denitions illustrated in Fig. 720 are explained as follows:
Basic size is the size to which limits or deviations are assigned and is the same for
both members of the t.
Deviation is the algebraic difference between a size and the corresponding basic size.
Upper deviation is the algebraic difference between the maximum limit and the
corresponding basic size.
Lower deviation is the algebraic difference between the minimum limit and the
corresponding basic size.
Fundamental deviation is either the upper or the lower deviation, depending on
which is closer to the basic size.
Tolerance is the difference between the maximum and minimum size limits of a part.
International tolerance grade numbers (IT) designate groups of tolerances such that
the tolerances for a particular IT number have the same relative level of accuracy
but vary depending on the basic size.
Hole basis represents a system of ts corresponding to a basic hole size. The fundamental deviation is H.
10
Preferred Limits and Fits for Cylindrical Parts, ANSI B4.1-1967. Preferred Metric Limits and Fits, ANSI
B4.2-1978.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
384
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Figure 720
Denitions applied to a
cylindrical t.
Upper deviation,
Max. size, dmax
u
Lower deviation,
l
Min. size, dmin
International tolerance
grade, d (IT number)
Fundamental deviation,
F (letter)
Basic size, D(d)
Lower deviation,
Upper deviation, u
International tolerance
grade, D (IT number)
l
Fundamental deviation,
F (letter)
Min. size, Dmin
Max. size, Dmax
Shaft basis represents a system of ts corresponding to a basic shaft size. The
fundamental deviation is h. The shaft-basis system is not included here.
The magnitude of the tolerance zone is the variation in part size and is the same
for both the internal and the external dimensions. The tolerance zones are specied
in international tolerance grade numbers, called IT numbers. The smaller grade numbers specify a smaller tolerance zone. These range from IT0 to IT16, but only grades
IT6 to IT11 are needed for the preferred ts. These are listed in Tables A11 to A13
for basic sizes up to 16 in or 400 mm.
The standard uses tolerance position letters, with capital letters for internal dimensions (holes) and lowercase letters for external dimensions (shafts). As shown in Fig. 720,
the fundamental deviation locates the tolerance zone relative to the basic size.
Table 79 shows how the letters are combined with the tolerance grades to establish a preferred t. The ISO symbol for the hole for a sliding t with a basic size of
32 mm is 32H7. Inch units are not a part of the standard. However, the designation
(1 3 in) H7 includes the same information and is recommended for use here. In both
8
cases, the capital letter H establishes the fundamental deviation and the number 7
denes a tolerance grade of IT7.
For the sliding t, the corresponding shaft dimensions are dened by the symbol
32g6 [(1 3 in)g6].
8
The fundamental deviations for shafts are given in Tables A11 and A13. For
letter codes c, d, f, g, and h,
Upper deviation = fundamental deviation
Lower deviation = upper deviation tolerance grade
For letter codes k, n, p, s, and u, the deviations for shafts are
Lower deviation = fundamental deviation
Upper deviation = lower deviation + tolerance grade
387
388
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
385
Shafts and Shaft Components
Table 79
Type of Fit
Description
Symbol
Descriptions of Preferred
Fits Using the Basic
Hole System
Clearance
Loose running t: for wide commercial tolerances or
allowances on external members
H11/c11
Free running t: not for use where accuracy is
essential, but good for large temperature variations,
high running speeds, or heavy journal pressures
H9/d9
Close running t: for running on accurate machines
and for accurate location at moderate speeds and
journal pressures
H8/f7
Sliding t: where parts are not intended to run freely,
but must move and turn freely and locate accurately
H7/g6
Locational clearance t: provides snug t for location
of stationary parts, but can be freely assembled and
disassembled
H7/h6
Locational transition t for accurate location, a
compromise between clearance and interference
H7/k6
Locational transition t for more accurate location
where greater interference is permissible
H7/n6
Locational interference t: for parts requiring rigidity
and alignment with prime accuracy of location but
without special bore pressure requirements
H7/p6
Medium drive t: for ordinary steel parts or shrink ts on
light sections, the tightest t usable with cast iron
H7/s6
Source: Preferred Metric Limits
and Fits, ANSI B4.2-1978.
See also BS 4500.
Transition
Interference
Force t: suitable for parts that can be highly stressed
H7/u6
or for shrink ts where the heavy pressing forces required
are impractical
The lower deviation H (for holes) is zero. For these, the upper deviation equals the
tolerance grade.
As shown in Fig. 720, we use the following notation:
D = basic size of hole
d = basic size of shaft
δu = upper deviation
δl = lower deviation
δ F = fundamental deviation
D = tolerance grade for hole
d = tolerance grade for shaft
Note that these quantities are all deterministic. Thus, for the hole,
Dmax = D +
D
(736)
Dmin = D
For shafts with clearance ts c, d, f, g, and h,
dmax = d + δ F
dmin = d + δ F
d
(737)
d
(738)
For shafts with interference ts k, n, p, s, and u,
dmin = d + δ F
dmax = d + δ F +
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
386
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
EXAMPLE 77
Find the shaft and hole dimensions for a loose running t with a 34-mm basic size.
Solution
From Table 79, the ISO symbol is 34H11/c11. From Table A11, we nd that tolerance grade IT11 is 0.160 mm. The symbol 34H11/c11 therefore says that
D = d = 0.160 mm. Using Eq. (736) for the hole, we get
Answer
Dmax = D +
Answer
D = 34 + 0.160 = 34.160 mm
Dmin = D = 34.000 mm
The shaft is designated as a 34c11 shaft. From Table A12, the fundamental deviation is δ F = 0.120 mm. Using Eq. (737), we get for the shaft dimensions
Answer
Answer
dmax = d + δ F = 34 + (0.120) = 33.880 mm
dmin = d + δ F
d = 34 + (0.120) 0.160 = 33.720 mm
EXAMPLE 78
Find the hole and shaft limits for a medium drive t using a basic hole size of 2 in.
Solution
The symbol for the t, from Table 78, in inch units is (2 in)H7/s6. For the hole, we
use Table A13 and nd the IT7 grade to be D = 0.0010 in. Thus, from Eq. (736),
Answer
Dmax = D +
Answer
D = 2 + 0.0010 = 2.0010 in
Dmin = D = 2.0000 in
The IT6 tolerance for the shaft is d = 0.0006 in. Also, from Table A14, the
fundamental deviation is δ F = 0.0017 in. Using Eq. (738), we get for the shaft that
Answer
Answer
dmin = d + δ F = 2 + 0.0017 = 2.0017 in
dmax = d + δ F +
d = 2 + 0.0017 + 0.0006 = 2.0023 in
Stress and Torque Capacity in Interference Fits
Interference ts between a shaft and its components can sometimes be used effectively to minimize the need for shoulders and keyways. The stresses due to an interference t can be obtained by treating the shaft as a cylinder with a uniform external
pressure, and the hub as a hollow cylinder with a uniform internal pressure. Stress
equations for these situations were developed in Sec. 316, and will be converted here
from radius terms into diameter terms to match the terminology of this section.
389
390
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
387
The pressure p generated at the interface of the interference t, from Eq. (356)
converted into terms of diameters, is given by
δ
p=
(739)
2
d d 2 + di2
d do + d 2
+ νo +
νi
2
E o do d 2
E i d 2 di2
or, in the case where both members are of the same material,
p=
2
E δ (do d 2 )(d 2 di2 )
2
2d 3
do di2
(740)
where d is the nominal shaft diameter, di is the inside diameter (if any) of the shaft,
do is the outside diameter of the hub, E is Youngs modulus, and v is Poissons ratio,
with subscripts o and i for the outer member (hub) and inner member (shaft), respectively. δ is the diametral interference between the shaft and hub, that is, the difference between the shaft outside diameter and the hub inside diameter.
δ = dshaft dhub
(741)
Since there will be tolerances on both diameters, the maximum and minimum
pressures can be found by applying the maximum and minimum interferences. Adopting the notation from Fig. 720, we write
δmin = dmin Dmax
(742)
δmax = dmax Dmin
(743)
where the diameter terms are dened in Eqs. (736) and (738). The maximum interference should be used in Eq. (739) or (740) to determine the maximum pressure
to check for excessive stress.
From Eqs. (358) and (359), with radii converted to diameters, the tangential
stresses at the interface of the shaft and hub are
σt , shaft = p
σt , hub = p
d 2 + di 2
d 2 di 2
do 2 + d 2
do 2 d 2
(744)
(745)
The radial stresses at the interface are simply
σr, shaft = p
(746)
σr, hub = p
(747)
The tangential and radial stresses are orthogonal, and should be combined using a
failure theory to compare with the yield strength. If either the shaft or hub yields during assembly, the full pressure will not be achieved, diminishing the torque that can be
transmitted. The interaction of the stresses due to the interference t with the other
stresses in the shaft due to shaft loading is not trivial. Finite-element analysis of the
interface would be appropriate when warranted. A stress element on the surface of a
rotating shaft will experience a completely reversed bending stress in the longitudinal
direction, as well as the steady compressive stresses in the tangential and radial directions. This is a three-dimensional stress element. Shear stress due to torsion in shaft
may also be present. Since the stresses due to the press t are compressive, the fatigue
situation is usually actually improved. For this reason, it may be acceptable to simplify
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
388
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
391
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
the shaft analysis by ignoring the steady compressive stresses due to the press t. There
is, however, a stress concentration effect in the shaft bending stress near the ends of the
hub, due to the sudden change from compressed to uncompressed material. The design
of the hub geometry, and therefore its uniformity and rigidity, can have a signicant
effect on the specic value of the stress concentration factor, making it difcult to report
generalized values. For rst estimates, values are typically not greater than 2.
The amount of torque that can be transmitted through an interference t can be
estimated with a simple friction analysis at the interface. The friction force is the product of the coefcient of friction f and the normal force acting at the interface. The
normal force can be represented by the product of the pressure p and the surface area
A of interface. Therefore, the friction force Ff is
Ff = f N = f ( p A) = f [ p2π(d /2)l ] = f pπ dl
(748)
where l is the length of the hub. This friction force is acting with a moment arm of
d/2 to provide the torque capacity of the joint, so
T = Ff d /2 = f pπ dl (d /2)
T = (π/2) f pld 2
(749)
The minimum interference, from Eq. (742), should be used to determine the
minimum pressure to check for the maximum amount of torque that the joint should
be designed to transmit without slipping.
PROBLEMS
71
A shaft is loaded in bending and torsion such that Ma = 600 lbf · in, Ta = 400 lbf · in, Mm =
500 lbf · in, and Tm = 300 lbf · in. For the shaft, Su = 100 kpsi and Sy = 80 kpsi, and a fully
corrected endurance limit of Se = 30 kpsi is assumed. Let K f = 2.2 and K f s = 1.8. With a
design factor of 2.0 determine the minimum acceptable diameter of the shaft using the
(a) DE-Gerber criterion.
(b) DE-elliptic criterion.
(c) DE-Soderberg criterion.
(d ) DE-Goodman criterion.
Discuss and compare the results.
72
The section of shaft shown in the gure is to be designed to approximate relative sizes of
d = 0.75 D and r = D /20 with diameter d conforming to that of standard metric rolling-bearing
bore sizes. The shaft is to be made of SAE 2340 steel, heat-treated to obtain minimum strengths
in the shoulder area of 1226-MPa ultimate tensile strength and 1130-MPa yield strength with a
Brinell hardness not less than 368. At the shoulder the shaft is subjected to a completely reversed
bending moment of 70 N · m, accompanied by a steady torsion of 45 N · m. Use a design factor
of 2.5 and size the shaft for an innite life.
Problem 72
Section of a shaft containing a
grinding-relief groove. Unless
otherwise specied, the diameter at
the root of the groove dr = d 2r,
and though the section of diameter d
is ground, the root of the groove is
still a machined surface.
r
D
d
392
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
73
389
The rotating solid steel shaft is simply supported by bearings at points B and C and is driven
by a gear (not shown) which meshes with the spur gear at D, which has a 6-in pitch diameter.
o
The force F from the drive gear acts at a pressure angle of 20 . The shaft transmits a torque
to point A of T A = 3000 lbf · in. The shaft is machined from steel with Sy = 60 kpsi and
Sut = 80 kpsi. Using a factor of safety of 2.5, determine the minimum allowable diameter of
the 10 in section of the shaft based on (a) a static yield analysis using the distortion energy
theory and (b) a fatigue-failure analysis. Assume sharp llet radii at the bearing shoulders for
estimating stress concentration factors.
TA
10 in
A
F
Problem 73
4 in
B
20
C
D
74
A geared industrial roll shown in the gure is driven at 300 rev/min by a force F acting on a
3-in-diameter pitch circle as shown. The roll exerts a normal force of 30 lbf/in of roll length
on the material being pulled through. The material passes under the roll. The coefcient of friction is 0.40. Develop the moment and shear diagrams for the shaft modeling the roll force as
(a) a concentrated force at the center of the roll, and (b) a uniformly distributed force along
the roll. These diagrams will appear on two orthogonal planes.
y
O
4 dia.
F
Problem 74
Material moves under the roll.
Dimensions in inches.
A
z
20°
3
14
3
B
8
3
14
3
24
2
x
Gear 4
3 dia.
75
Design a shaft for the situation of the industrial roll of Prob. 74 with a design factor of 2 and
a reliability goal of 0.999 against fatigue failure. Plan for a ball bearing on the left and a cylindrical roller on the right. For deformation use a factor of safety of 2.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
390
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
76
The gure shows a proposed design for the industrial roll shaft of Prob. 74. Hydrodynamic lm
bearings are to be used. All surfaces are machined except the journals, which are ground and polished. The material is 1035 HR steel. Perform a design assessment. Is the design satisfactory?
11
4
Problem 76
Bearing shoulder llets 0.030 in,
1
others 16 in. Sled-runner keyway is
31 in long. Dimensions in inches.
2
keyway
A
1
1
1
10
12
77
1
4
1
O
7
8
4
12
In the double-reduction gear train shown, shaft a is driven by a motor attached by a exible
coupling attached to the overhang. The motor provides a torque of 2500 lbf · in at a speed of
1200 rpm. The gears have 20o pressure angles, with diameters shown on the gure. Use an
AISI 1020 cold-drawn steel. Design one of the shafts (as specied by the instructor) with a
design factor of 1.5 by performing the following tasks.
(a) Sketch a general shaft layout, including means to locate the gears and bearings, and to transmit the torque.
(b) Perform a force analysis to nd the bearing reaction forces, and generate shear and bending moment diagrams.
(c) Determine potential critical locations for stress design.
(d) Determine critical diameters of the shaft based on fatigue and static stresses at the critical
locations.
(e) Make any other dimensional decisions necessary to specify all diameters and axial dimensions. Sketch the shaft to scale, showing all proposed dimensions.
(f) Check the deection at the gear, and the slopes at the gear and the bearings for satisfaction
of the recommended limits in Table 72.
(g) If any of the deections exceed the recommended limits, make appropriate changes to bring
them all within the limits.
3
8
24
F
E
c
Problem 77
16
Dimensions in inches.
20
4
D
C
b
8
A
B
a
12
78
9
2
6
In the gure is a proposed shaft design to be used for the input shaft a in Prob. 77. A ball
bearing is planned for the left bearing, and a cylindrical roller bearing for the right.
(a) Determine the minimum fatigue factor of safety by evaluating at any critical locations. Use a
fatigue failure criteria that is considered to be typical of the failure data, rather than one that
is considered conservative. Also ensure that the shaft does not yield in the rst load cycle.
393
394
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
391
Shafts and Shaft Components
(b) Check the design for adequacy with respect to deformation, according to the recommendations in Table 72.
8
3
74
Problem 78
Shoulder llets at bearing seat
0.030-in radius, others 1 -in radius,
8
except right-hand bearing seat
1
transition, 4 in. The material is
1030 HR. Keyways 3 in wide by
8
3
in deep. Dimensions in inches.
16
0.354
0.453
1.875
1.875
1.500
1.574
1.574
9
6
11
79
The shaft shown in the gure is driven by a gear at the right keyway, drives a fan at the left
keyway, and is supported by two deep-groove ball bearings. The shaft is made from AISI 1020
cold-drawn steel. At steady-state speed, the gear transmits a radial load of 230 lbf and a tangential load of 633 lbf at a pitch diameter of 8 in.
(a) Determine fatigue factors of safety at any potentially critical locations.
(b) Check that deections satisfy the suggested minimums for bearings and gears.
12.87
8.50
1.181
2.0
2.20
0.20
0.75
0.485
1.750
2.75
1.70
1.40
1.181
1.000
Problem 79
Dimensions in inches.
2.0
1
16
1
4
710
×
1
8
R.
keyway
1
32
1
8
0.15
R.
3
8
R.
×
3
16
0.1 R.
1
8
keyway
R.
1
32
R.
An AISI 1020 cold-drawn steel shaft with the geometry shown in the gure carries a transverse
load of 7 kN and a torque of 107 N · m. Examine the shaft for strength and deection. If the
largest allowable slope at the bearings is 0.001 rad and at the gear mesh is 0.0005 rad, what
7 kN
155
40
35
30
55
45
40
35
30
20
Problem 710
Dimensions in millimeters.
30
30
60
55
115
85
10
150
375
All fillets 2 mm
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
392
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Mechanical Engineering Design
is the factor of safety guarding against damaging distortion? What is the factor of safety guarding against a fatigue failure? If the shaft turns out to be unsatisfactory, what would you recommend to correct the problem?
711
A shaft is to be designed to support the spur pinion and helical gear shown in the gure on
two bearings spaced 28 in center-to-center. Bearing A is a cylindrical roller and is to take only
radial load; bearing B is to take the thrust load of 220 lbf produced by the helical gear and its
share of the radial load. The bearing at B can be a ball bearing. The radial loads of both gears
are in the same plane, and are 660 lbf for the pinion and 220 lbf for the gear. The shaft speed
is 1150 rev/min. Design the shaft. Make a sketch to scale of the shaft showing all llet sizes,
keyways, shoulders, and diameters. Specify the material and its heat treatment.
C brg
L
C brg
L
2
4
Problem 711
Dimensions in inches.
A
B
7
712
16
5
A heat-treated steel shaft is to be designed to support the spur gear and the overhanging worm
shown in the gure. A bearing at A takes pure radial load. The bearing at B takes the wormthrust load for either direction of rotation. The dimensions and the loading are shown in the
gure; note that the radial loads are in the same plane. Make a complete design of the shaft,
including a sketch of the shaft showing all dimensions. Identify the material and its heat treatment (if necessary). Provide an assessment of your nal design. The shaft speed is 310 rev/min.
4
4
A
B
Problem 712
Dimensions in inches.
4
3
14
950 lbf
600 lbf
RB
5600 lbf
T = 4800 lbf-in
T
RA
713
RB
A bevel-gear shaft mounted on two 40-mm 02-series ball bearings is driven at 1720 rev/min
by a motor connected through a exible coupling. The gure shows the shaft, the gear, and the
bearings. The shaft has been giving troublein fact, two of them have already failedand the
down time on the machine is so expensive that you have decided to redesign the shaft yourself rather than order replacements. A hardness check of the two shafts in the vicinity of the fracture of the two shafts showed an average of 198 Bhn for one and 204 Bhn of the other. As closely
as you can estimate the two shafts failed at a life measure between 600 000 and 1 200 000 cycles
395
396
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
7. Shafts and Shaft
Components
Shafts and Shaft Components
393
of operation. The surfaces of the shaft were machined, but not ground. The llet sizes were not
measured, but they correspond with the recommendations for the ball bearings used. You know
that the load is a pulsating or shock-type load, but you have no idea of the magnitude, because
the shaft drives an indexing mechanism, and the forces are inertial. The keyways are 3 in wide
8
3
by 16 in deep. The straight-toothed bevel pinion drives a 48-tooth bevel gear. Specify a new
shaft in sufcient detail to ensure a long and trouble-free life.
2
Shaft failed here
1
3
Problem 713
1 2 dia.
1 8 dia.
Dimensions in inches.
4
6
1
2
2
4P, 16T
714
A 1-in-diameter uniform steel shaft is 24 in long between bearings.
(a) Find the lowest critical speed of the shaft.
(b) If the goal is to double the critical speed, nd the new diameter.
(c) A half-size model of the original shaft has what critical speed?
715
Demonstrate how rapidly Rayleighs method converges for the uniform-diameter solid shaft of
Prob. 714, by partitioning the shaft into rst one, then two, and nally three elements.
716
Compare Eq. (727) for the angular frequency of a two-disk shaft with Eq. (728), and note
that the constants in the two equations are equal.
(a) Develop an expression for the second critical speed.
(b) Estimate the second critical speed of the shaft addressed in Ex. 75, parts a and b.
717
For a uniform-diameter shaft, does hollowing the shaft increase or decrease the critical speed?
718
The shaft shown in the gure carries a 20-lbf gear on the left and a 35-lbf gear on the right.
Estimate the rst critical speed due to the loads, the shafts critical speed without the loads,
and the critical speed of the combination.
35 lbf
20 lbf
2.000
2.763
2.472
2.000
Problem 718
Dimensions in inches.
1
2
9
14
15
16
719
A transverse drilled and reamed hole can be used in a solid shaft to hold a pin that locates and
holds a mechanical element, such as the hub of a gear, in axial position, and allows for the
transmission of torque. Since a small-diameter hole introduces high stress concentration, and a
larger diameter hole erodes the area resisting bending and torsion, investigate the existence of
a pin diameter with minimum adverse affect on the shaft. Then formulate a design rule. (Hint:
Use Table A16.)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
394
III. Design of Mechanical
Elements
7. Shafts and Shaft
Components
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
720
A guide pin is required to align the assembly of a two-part xture. The nominal size of the pin
is 15 mm. Make the dimensional decisions for a 15-mm basic size locational clearance t.
721
An interference t of a cast-iron hub of a gear on a steel shaft is required. Make the dimensional decisions for a 45-mm basic size medium drive t.
722
A pin is required for forming a linkage pivot. Find the dimensions required for a 50-mm basic
size pin and clevis with a sliding t.
723
A journal bearing and bushing need to be described. The nominal size is 1 in. What dimensions are needed for a 1-in basic size with a close running t if this is a lightly loaded journal
and bushing assembly?
724
A gear and shaft with nominal diameter of 1.5 in are to be assembled with a medium drive t,
as specied in Table 79. The gear has a hub, with an outside diameter of 2.5 in, and an overall length of 2 in. The shaft is made from AISI 1020 CD steel, and the gear is made from steel
that has been through hardened to provide Su
100 kpsi and Sy
85 kpsi.
(a) Specify dimensions with tolerances for the shaft and gear bore to achieve the desired t.
(b) Determine the minimum and maximum pressures that could be experienced at the interface
with the specied tolerances.
(c) Determine the worst-case static factors of safety guarding against yielding at assembly for
the shaft and the gear based on the distortion energy failure theory.
(d ) Determine the maximum torque that the joint should be expected to transmit without slipping, i.e., when the interference pressure is at a minimum for the specied tolerances.
397
398
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
8
Screws, Fasteners, and the
Design of Nonpermanent
Joints
Chapter Outline
81
Thread Standards and Denitions
396
82
The Mechanics of Power Screws
400
83
Threaded Fasteners
84
JointsFastener Stiffness
410
85
JointsMember Stiffness
413
86
Bolt Strength
87
Tension JointsThe External Load
88
Relating Bolt Torque to Bolt Tension
89
Statically Loaded Tension Joint with Preload
408
417
421
422
810
Gasketed Joints
811
Fatigue Loading of Tension Joints
812
Bolted and Riveted Joints Loaded in Shear
425
429
429
435
395
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
396
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
The helical-thread screw was undoubtably an extremely important mechanical invention. It is the basis of power screws, which change angular motion to linear motion
to transmit power or to develop large forces (presses, jacks, etc.), and threaded fasteners, an important element in nonpermanent joints.
This book presupposes a knowledge of the elementary methods of fastening. Typical methods of fastening or joining parts use such devices as bolts, nuts, cap screws,
setscrews, rivets, spring retainers, locking devices, pins, keys, welds, and adhesives.
Studies in engineering graphics and in metal processes often include instruction on various joining methods, and the curiosity of any person interested in mechanical engineering naturally results in the acquisition of a good background knowledge of fastening methods. Contrary to rst impressions, the subject is one of the most interesting in
the entire eld of mechanical design.
One of the key targets of current design for manufacture is to reduce the number
of fasteners. However, there will always be a need for fasteners to facilitate disassembly for whatever purposes. For example, jumbo jets such as Boeings 747 require
as many as 2.5 million fasteners, some of which cost several dollars apiece. To keep
costs down, aircraft manufacturers, and their subcontractors, constantly review new
fastener designs, installation techniques, and tooling.
The number of innovations in the fastener eld over any period you might care
to mention has been tremendous. An overwhelming variety of fasteners are available
for the designers selection. Serious designers generally keep specic notebooks on
fasteners alone. Methods of joining parts are extremely important in the engineering
of a quality design, and it is necessary to have a thorough understanding of the performance of fasteners and joints under all conditions of use and design.
81
Thread Standards and Denitions
The terminology of screw threads, illustrated in Fig. 81, is explained as follows:
The pitch is the distance between adjacent thread forms measured parallel to
the thread axis. The pitch in U.S. units is the reciprocal of the number of thread forms
per inch N.
The major diameter d is the largest diameter of a screw thread.
The minor (or root) diameter dr is the smallest diameter of a screw thread.
The pitch diameter d p is a theoretical diameter between the major and minor
diameters.
The lead l, not shown, is the distance the nut moves parallel to the screw axis when
the nut is given one turn. For a single thread, as in Fig. 81, the lead is the same as
the pitch.
A multiple-threaded product is one having two or more threads cut beside each
other (imagine two or more strings wound side by side around a pencil). Standardized products such as screws, bolts, and nuts all have single threads; a double-threaded
screw has a lead equal to twice the pitch, a triple-threaded screw has a lead equal to
3 times the pitch, and so on.
All threads are made according to the right-hand rule unless otherwise noted.
The American National (Unied) thread standard has been approved in this country and in Great Britain for use on all standard threaded products. The thread angle
is 60 and the crests of the thread may be either at or rounded.
Figure 82 shows the thread geometry of the metric M and MJ proles. The M
prole replaces the inch class and is the basic ISO 68 prole with 60 symmetric
threads. The MJ prole has a rounded llet at the root of the external thread and a
399
400
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
Figure 81
397
Major diameter
Pitch diameter
Terminology of screw threads.
Sharp vee threads shown for
clarity; the crests and roots are
actually attened or rounded
during the forming operation.
Minor diameter
Pitch p
45° chamfer
Root
Thread angle 2α
Crest
Figure 82
Basic prole for metric M
and M J threads.
d major diameter
dr minor diameter
dp pitch diameter
p itch
p
H
3
2
H
8
H
p
8
5H
8
p
2
p
4
p
Internal threads
p
2
3H
8
60°
H
4
60°
H
4
d
30°
dp
p
External threads
dr
larger minor diameter of both the internal and external threads. This prole is especially useful where high fatigue strength is required.
Tables 81 and 82 will be useful in specifying and designing threaded parts.
Note that the thread size is specied by giving the pitch p for metric sizes and by
giving the number of threads per inch N for the Unied sizes. The screw sizes in
Table 82 with diameter under 1 in are numbered or gauge sizes. The second column
4
in Table 82 shows that a No. 8 screw has a nominal major diameter of 0.1640 in.
A great many tensile tests of threaded rods have shown that an unthreaded rod
having a diameter equal to the mean of the pitch diameter and minor diameter will have
the same tensile strength as the threaded rod. The area of this unthreaded rod is called
the tensile-stress area At of the threaded rod; values of At are listed in both tables.
Two major Unied thread series are in common use: UN and UNR. The difference between these is simply that a root radius must be used in the UNR series.
Because of reduced thread stress-concentration factors, UNR series threads have
improved fatigue strengths. Unied threads are specied by stating the nominal major
diameter, the number of threads per inch, and the thread series, for example, 5 in-18
8
UNRF or 0.625 in-18 UNRF.
Metric threads are specied by writing the diameter and pitch in millimeters, in
that order. Thus, M12 × 1.75 is a thread having a nominal major diameter of 12 mm
and a pitch of 1.75 mm. Note that the letter M, which precedes the diameter, is the
clue to the metric designation.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
398
III. Design of Mechanical
Elements
401
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 81
Diameters and Areas of
Coarse-Pitch and FinePitch Metric Threads.*
Nominal
Major
Diameter
d
mm
Coarse-Pitch Series
Pitch
p
mm
TensileStress
Area At
mm2
MinorDiameter
Area Ar
mm2
1.6
0.35
1.27
0.40
2.07
0.45
3.39
0.5
5.03
4.47
3.5
0.6
6.78
6.00
4
0.7
8.78
7.75
5
0.8
14.2
12.7
6
1
20.1
MinorDiameter
Area Ar
mm2
2.98
3
TensileStress
Area At
mm2
1.79
2.5
Pitch
p
mm
1.07
2
Fine-Pitch Series
17.9
1.25
36.6
32.8
1
39.2
36.0
10
8
1.5
58.0
52.3
1.25
61.2
56.3
12
1.75
76.3
1.25
14
2
16
2
157
144
1.5
167
157
20
2.5
245
225
1.5
272
259
24
3
353
324
2
384
365
30
3.5
561
519
2
621
596
36
4
42
4.5
48
5
1470
1380
2
1670
1630
56
5.5
2030
1910
2
2300
2250
64
6
2680
2520
2
3030
2980
72
6
3460
3280
2
3860
3800
80
6
4340
4140
1.5
4850
4800
90
6
5590
5360
2
6100
6020
100
6
6990
6740
110
84.3
115
104
1.5
92.1
125
86.0
116
817
759
2
915
884
1120
1050
2
1260
1230
2
7560
7470
2
9180
9080
*The equations and data used to develop this table have been obtained from ANSI B1.1-1974 and B18.3.1-1978. The minor
diameter was found from the equation dr d 1.226 869p, and the pitch diameter from dp d 0.649 519p. The mean of
the pitch diameter and the minor diameter was used to compute the tensile-stress area.
Square and Acme threads, shown in Fig. 83a and b, respectively, are used on
screws when power is to be transmitted. Table 83 lists the preferred pitches for inchseries Acme threads. However, other pitches can be and often are used, since the need
for a standard for such threads is not great.
Modications are frequently made to both Acme and square threads. For instance,
the square thread is sometimes modied by cutting the space between the teeth so as
to have an included thread angle of 10 to 15 . This is not difcult, since these threads
are usually cut with a single-point tool anyhow; the modication retains most of the
high efciency inherent in square threads and makes the cutting simpler. Acme threads
402
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
399
Table 82
Diameters and Area of Unied Screw Threads UNC and UNF*
Coarse SeriesUNC
Size
Designation
Nominal
Major
Diameter
in
Fine SeriesUNF
Threads
per Inch
N
MinorDiameter
Area Ar
in2
TensileStress
Area At
in2
MinorDiameter
Area Ar
in2
80
Threads
per Inch
N
TensileStress
Area At
in2
0.001 80
0.001 51
0
0.0600
1
0.0730
64
0.002 63
0.002 18
72
0.002 78
0.002 37
2
0.0860
56
0.003 70
0.003 10
64
0.003 94
0.003 39
3
0.0990
48
0.004 87
0.004 06
56
0.005 23
0.004 51
4
0.1120
40
0.006 04
0.004 96
48
0.006 61
0.005 66
5
0.1250
40
0.007 96
0.006 72
44
0.008 80
0.007 16
6
0.1380
32
0.009 09
0.007 45
40
0.010 15
0.008 74
8
0.1640
32
0.014 0
0.011 96
36
0.014 74
0.012 85
10
0.1900
24
0.017 5
0.014 50
32
0.020 0
0.017 5
12
0.2160
24
0.024 2
0.020 6
28
0.025 8
0.022 6
1
4
5
16
0.2500
20
0.031 8
0.026 9
28
0.036 4
0.032 6
0.3125
18
0.052 4
0.045 4
24
0.058 0
0.052 4
0.3750
16
0.077 5
0.067 8
24
0.087 8
0.080 9
0.4375
14
0.106 3
0.093 3
20
0.118 7
0.109 0
0.5000
13
0.141 9
0.125 7
20
0.159 9
0.148 6
3
8
7
16
1
2
9
16
0.5625
12
0.182
0.162
18
0.203
0.189
5
8
3
4
7
8
0.6250
11
0.226
0.202
18
0.256
0.240
0.7500
10
0.334
0.302
16
0.373
0.351
0.8750
9
0.462
0.419
14
0.509
0.480
1
1.0000
8
0.606
0.551
12
0.663
0.625
1
14
1
12
1.2500
7
0.969
0.890
12
1.073
1.024
1.5000
6
1.405
1.294
12
1.581
1.521
*This table was compiled from ANSI B1.1-1974. The minor diameter was found from the equation dr
mean of the pitch diameter and the minor diameter was used to compute the tensile-stress area.
Figure 83
p
1.299 038p, and the pitch diameter from dp
p
p
2
(a) Square thread; (b) Acme
thread.
d
29°
p
2
d
p
2
d
dr
dr
(a )
(b)
p
2
d
0.649 519p. The
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
400
III. Design of Mechanical
Elements
403
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 83
d, in
Preferred Pitches for
Acme Threads
1
4
5
16
3
8
1
2
5
8
3
4
7
8
1
11
4
11
2
13
4
2
21
2
3
p, in
1
16
1
14
1
12
1
10
1
8
1
6
1
6
1
5
1
5
1
4
1
4
1
4
1
3
1
2
are sometimes modied to a stub form by making the teeth shorter. This results in a
larger minor diameter and a somewhat stronger screw.
82
The Mechanics of Power Screws
A power screw is a device used in machinery to change angular motion into linear
motion, and, usually, to transmit power. Familiar applications include the lead screws
of lathes, and the screws for vises, presses, and jacks.
An application of power screws to a power-driven jack is shown in Fig. 84. You
should be able to identify the worm, the worm gear, the screw, and the nut. Is the
worm gear supported by one bearing or two?
Figure 84
The Joyce worm-gear screw
jack. (Courtesy Joyce-Dayton
Corp., Dayton, Ohio.)
404
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
401
Figure 85
Portion of a power screw.
dm
F
p
Nut
F 2
F 2
Figure 86
Force diagrams: (a) lifting the
load; (b) lowering the load.
F
F
fN
PR
l
fN
PL
N
N
dm
(a)
l
dm
(b)
In Fig. 85 a square-threaded power screw with single thread having a mean
diameter dm , a pitch p, a lead angle λ, and a helix angle ψ is loaded by the axial
compressive force F. We wish to nd an expression for the torque required to raise
this load, and another expression for the torque required to lower the load.
First, imagine that a single thread of the screw is unrolled or developed (Fig. 86)
for exactly a single turn. Then one edge of the thread will form the hypotenuse of a right
triangle whose base is the circumference of the mean-thread-diameter circle and whose
height is the lead. The angle λ, in Figs. 85 and 86, is the lead angle of the thread. We
represent the summation of all the unit axial forces acting upon the normal thread area
by F. To raise the load, a force PR acts to the right (Fig. 86a), and to lower the load,
PL acts to the left (Fig. 86b). The friction force is the product of the coefcient of friction f with the normal force N, and acts to oppose the motion. The system is in equilibrium under the action of these forces, and hence, for raising the load, we have
FH = PR N sin λ f N cos λ = 0
(a)
FV = F + f N sin λ N cos λ = 0
In a similar manner, for lowering the load, we have
FH = PL N sin λ + f N cos λ = 0
FV = F f N sin λ N cos λ = 0
(b)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
402
III. Design of Mechanical
Elements
405
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Since we are not interested in the normal force N, we eliminate it from each of these
sets of equations and solve the result for P. For raising the load, this gives
PR =
F (sin λ + f cos λ)
cos λ f sin λ
(c)
PL =
F ( f cos λ sin λ)
cos λ + f sin λ
(d )
and for lowering the load,
Next, divide the numerator and the denominator of these equations by cos λ and use
the relation tan λ = l /π dm (Fig. 86). We then have, respectively,
PR =
F [(l /π dm ) + f ]
1 ( f l /π dm )
(e)
PL =
F [ f (l /π dm )]
1 + ( f l /π dm )
(f )
Finally, noting that the torque is the product of the force P and the mean radius dm /2,
for raising the load we can write
TR =
Fdm
2
l + π f dm
π dm f l
(81)
where TR is the torque required for two purposes: to overcome thread friction and to
raise the load.
The torque required to lower the load, from Eq. ( f ), is found to be
TL =
Fdm
2
π f dm l
π dm + f l
(82)
This is the torque required to overcome a part of the friction in lowering the load. It may
turn out, in specic instances where the lead is large or the friction is low, that the load
will lower itself by causing the screw to spin without any external effort. In such cases,
the torque TL from Eq. (82) will be negative or zero. When a positive torque is
obtained from this equation, the screw is said to be self-locking. Thus the condition
for self-locking is
π f dm > l
Now divide both sides of this inequality by π dm . Recognizing that l /π dm = tan λ, we
get
f > tan λ
(83)
This relation states that self-locking is obtained whenever the coefcient of thread
friction is equal to or greater than the tangent of the thread lead angle.
An expression for efciency is also useful in the evaluation of power screws. If
we let f = 0 in Eq. (81), we obtain
T0 =
Fl
2π
(g)
406
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
403
which, since thread friction has been eliminated, is the torque required only to raise
the load. The efciency is therefore
e=
T0
Fl
=
TR
2π TR
(84)
The preceding equations have been developed for square threads where the normal thread loads are parallel to the axis of the screw. In the case of Acme or other
threads, the normal thread load is inclined to the axis because of the thread angle 2α
and the lead angle λ. Since lead angles are small, this inclination can be neglected
and only the effect of the thread angle (Fig. 87a) considered. The effect of the angle
α is to increase the frictional force by the wedging action of the threads. Therefore
the frictional terms in Eq. (81) must be divided by cos α . For raising the load, or for
tightening a screw or bolt, this yields
TR =
Fdm
2
l + π f dm sec α
π dm f l sec α
(85)
In using Eq. (85), remember that it is an approximation because the effect of the
lead angle has been neglected.
For power screws, the Acme thread is not as efcient as the square thread, because
of the additional friction due to the wedging action, but it is often preferred because
it is easier to machine and permits the use of a split nut, which can be adjusted to
take up for wear.
Usually a third component of torque must be applied in power-screw applications.
When the screw is loaded axially, a thrust or collar bearing must be employed between
the rotating and stationary members in order to carry the axial component. Figure
87b shows a typical thrust collar in which the load is assumed to be concentrated at
the mean collar diameter dc . If f c is the coefcient of collar friction, the torque
required is
Tc =
F f c dc
2
(86)
For large collars, the torque should probably be computed in a manner similar to that
employed for disk clutches.
Figure 87
(a) Normal thread force is
increased because of angle α ;
(b) thrust collar has frictional
diameter dc.
dc
F 2
F
cos
F
F 2
Collar
Nut
2=
Thread
angle
F 2
(a)
F 2
(b)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
404
III. Design of Mechanical
Elements
407
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Nominal body stresses in power screws can be related to thread parameters as
follows. The maximum nominal shear stress τ in torsion of the screw body can be
expressed as
τ=
16T
π dr3
(87)
The axial stress σ in the body of the screw due to load F is
σ=
F
4F
=
A
π dr2
(88)
in the absence of column action. For a short column the J. B. Johnson buckling
formula is given by Eq. (443), which is
F
A
crit
= Sy
Sy l
2π k
2
1
CE
(89)
Nominal thread stresses in power screws can be related to thread parameters as
follows. The bearing stress in Fig. 88, σ B , is
σB =
F
2F
=
π dm n t p/2
π dm n t p
(810)
where n t is the number of engaged threads. The bending stress at the root of the thread
σb is found from
I
π
(π dr n t ) ( p/2)2
=
=
dr n t p2
c
6
24
M=
Fp
4
so
M
6F
Fp
24
=
=
2
I /c
4 π dr n t p
π dr n t p
σb =
(811)
The transverse shear stress τ at the center of the root of the thread due to load F is
3V
3
3F
F
=
=
2A
2 π dr n t p/2
π dr n t p
τ=
(812)
and at the top of the root it is zero. The von Mises stress σ at the top of the root plane
is found by rst identifying the orthogonal normal stresses and the shear stresses. From
dm
Figure 88
Geometry of square thread
useful in nding bending and
transverse shear stresses at the
thread root.
F
z
p2
x
p 2
dr
408
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
405
the coordinate system of Fig. 88, we note
σx =
6F
π dr n t p
τ yz =
σy = 0
σz =
τx y = 0
4F
π dr2
16T
π dr3
τzx = 0
then use Eq. (514) of Sec. 55.
The screw-thread form is complicated from an analysis viewpoint. Remember the
origin of the tensile-stress area At , which comes from experiment. A power screw lifting a load is in compression and its thread pitch is shortened by elastic deformation.
Its engaging nut is in tension and its thread pitch is lengthened. The engaged threads
cannot share the load equally. Some experiments show that the rst engaged thread
carries 0.38 of the load, the second 0.25, the third 0.18, and the seventh is free of load.
In estimating thread stresses by the equations above, substituting 0.38 F for F and setting n t to 1 will give the largest level of stresses in the thread-nut combination.
EXAMPLE 81
A square-thread power screw has a major diameter of 32 mm and a pitch of 4 mm
with double threads, and it is to be used in an application similar to that in Fig. 84.
The given data include f = f c = 0.08, dc = 40 mm, and F = 6.4 kN per screw.
(a) Find the thread depth, thread width, pitch diameter, minor diameter, and lead.
(b) Find the torque required to raise and lower the load.
(c) Find the efciency during lifting the load.
(d ) Find the body stresses, torsional and compressive.
(e) Find the bearing stress.
( f ) Find the thread stresses bending at the root, shear at the root, and von Mises stress
and maximum shear stress at the same location.
Solution
(a) From Fig. 83a the thread depth and width are the same and equal to half the
pitch, or 2 mm. Also
dm = d p/2 = 32 4/2 = 30 mm
Answer
dr = d p = 32 4 = 28 mm
l = np = 2(4) = 8 mm
(b) Using Eqs. (81) and (86), the torque required to turn the screw against the load is
TR =
=
Answer
Fdm
2
l + π f dm
π dm f l
+
F f c dc
2
6.4(30) 8 + π(0.08)(30)
6.4(0.08)40
+
2
π(30) 0.08(8)
2
= 15.94 + 10.24 = 26.18 N · m
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
406
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Using Eqs. (82) and (86), we nd the load-lowering torque is
TL =
=
Answer
π f dm l
π dm + f l
Fdm
2
+
F f c dc
2
6.4(0.08)(40)
6.4(30) π (0.08)30 8
+
2
π(30) + 0.08(8)
2
= 0.466 + 10.24 = 9.77 N · m
The minus sign in the rst term indicates that the screw alone is not self-locking and
would rotate under the action of the load except for the fact that the collar friction is
present and must be overcome, too. Thus the torque required to rotate the screw with
the load is less than is necessary to overcome collar friction alone.
(c) The overall efciency in raising the load is
Answer
e=
Fl
6.4(8)
=
= 0.311
2π TR
2π(26.18)
(d ) The body shear stress τ due to torsional moment TR at the outside of the screw
body is
Answer
τ=
16TR
16(26.18)(103 )
=
= 6.07 MPa
π dr3
π(283 )
The axial nominal normal stress σ is
Answer
σ =
4(6.4)103
4F
=
= 10.39 MPa
π dr2
π(282 )
(e) The bearing stress σ B is, with one thread carrying 0.38 F ,
Answer
σB =
2(0.38 F )
2(0.38)(6.4)103
=
= 12.9 MPa
π dm (1) p
π(30)(1)(4)
( f ) The thread-root bending stress σb with one thread carrying 0.38 F is
σb =
6(0.38 F )
6(0.38)(6.4)103
=
= 41.5 MPa
π dr (1) p
π(28)(1)4
The transverse shear at the extreme of the root cross section due to bending is zero.
However, there is a circumferential shear stress at the extreme of the root cross section of the thread as shown in part (d ) of 6.07 MPa. The three-dimensional stresses,
after Fig. 88, noting the y coordinate is into the page, are
σx = 41.5 MPa
τx y = 0
σy = 0
τ yz = 6.07 MPa
σz = 10.39 MPa
τzx = 0
Equation (514) of Sec. 55 can be written as
Answer
1
σ = {(41.5 0) 2 + [0 ( 10.39)]2 + ( 10.39 41.5) 2 + 6(6.07) 2 }1/2
2
= 48.7 MPa
409
410
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
407
Alternatively, you can determine the principal stresses and then use Eq. (512) to nd
the von Mises stress. This would prove helpful in evaluating τmax as well. The principal stresses can be found from Eq. (315); however, sketch the stress element and
note that there are no shear stresses on the x face. This means that σx is a principal
stress. The remaining stresses can be transformed by using the plane stress equation,
Eq. (313). Thus, the remaining principal stresses are
10.39
±
2
10.39
2
2
+ 6.072 = 2.79, 13.18 MPa
Ordering the principal stresses gives σ1 , σ2 , σ3 = 41.5, 2.79, 13.18 MPa. Substituting these into Eq. (512) yields
σ =
Answer
[41.5 2.79]2 + [2.79 (13.18)]2 + [13.18 41.5]2
2
1/2
= 48.7 MPa
The maximum shear stress is given by Eq. (316), where τmax = τ1/3 , giving
Answer
Table 84
Screw Bearing
Pressure pb
Source: H. A. Rothbart,
Mechanical Design and
Systems Handbook, 2nd ed.,
McGraw-Hill, New York,
1985.
τmax =
σ1 σ3
41.5 (13.18)
=
= 27.3 MPa
2
2
Screw
Material
Nut
Material
Safe pb, psi
Notes
Steel
Bronze
25003500
Low speed
Steel
Bronze
16002500
10 fpm
Cast iron
18002500
Steel
8 fpm
8001400
2040 fpm
Cast iron
Steel
Bronze
6001000
2040 fpm
Bronze
150240
50 fpm
Ham and Ryan1 showed that the coefcient of friction in screw threads is independent of axial load, practically independent of speed, decreases with heavier lubricants, shows little variation with combinations of materials, and is best for steel on
bronze. Sliding coefcients of friction in power screws are about 0.100.15.
Table 84 shows safe bearing pressures on threads, to protect the moving surfaces from abnormal wear. Table 85 shows the coefcients of sliding friction for
common material pairs. Table 86 shows coefcients of starting and running friction
for common material pairs.
1
Ham and Ryan, An Experimental Investigation of the Friction of Screw-threads, Bulletin 247, University of
Illinois Experiment Station, Champaign-Urbana, Ill., June 7, 1932.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
408
III. Design of Mechanical
Elements
411
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 85
Coefcients of Friction f
for Threaded Pairs
Source: H. A. Rothbart,
Mechanical Design and
Systems Handbook, 2nd ed.,
McGraw-Hill, New York,
1985.
Table 86
Nut Material
Screw
Material
Steel
Bronze
Brass
Cast Iron
Steel, dry
0.150.25
0.150.23
0.150.19
0.150.25
Steel, machine oil
0.110.17
0.100.16
0.100.15
0.110.17
Bronze
0.080.12
0.040.06
0.060.09
Combination
Running
Starting
Thrust-Collar Friction
Coefcients
Soft steel on cast iron
0.12
0.17
Hard steel on cast iron
0.09
0.15
Source: H. A. Rothbart,
Mechanical Design and
Systems Handbook, 2nd ed.,
McGraw-Hill, New York,
1985.
Soft steel on bronze
0.08
0.10
Hard steel on bronze
0.06
0.08
83
Threaded Fasteners
As you study the sections on threaded fasteners and their use, be alert to the stochastic
and deterministic viewpoints. In most cases the threat is from overproof loading of
fasteners, and this is best addressed by statistical methods. The threat from fatigue is
lower, and deterministic methods can be adequate.
Figure 89 is a drawing of a standard hexagon-head bolt. Points of stress concentration are at the llet, at the start of the threads (runout), and at the thread-root
llet in the plane of the nut when it is present. See Table A29 for dimensions. The
diameter of the washer face is the same as the width across the ats of the hexagon.
The thread length of inch-series bolts, where d is the nominal diameter, is
LT =
2d +
2d +
1
4
1
2
in
in
L 6 in
(813)
L > 6 in
and for metric bolts is
LT =
2d + 6
2d + 12
2d + 25
L 125
125 < L 200
d 48
(814)
L > 200
where the dimensions are in millimeters. The ideal bolt length is one in which only
one or two threads project from the nut after it is tightened. Bolt holes may have burrs
or sharp edges after drilling. These could bite into the llet and increase stress concentration. Therefore, washers must always be used under the bolt head to prevent
this. They should be of hardened steel and loaded onto the bolt so that the rounded
edge of the stamped hole faces the washer face of the bolt. Sometimes it is necessary
to use washers under the nut too.
The purpose of a bolt is to clamp two or more parts together. The clamping load
stretches or elongates the bolt; the load is obtained by twisting the nut until the bolt
has elongated almost to the elastic limit. If the nut does not loosen, this bolt tension
412
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
Figure 89
H
Approx.
Hexagon-head bolt; note the
washer face, the llet under
the head, the start of threads,
and the chamfer on both ends.
Bolt lengths are always
measured from below the
head.
1
64
409
W
in
R
30°
Figure 810
Typical cap-screw heads:
(a) llister head; (b) at head;
(c) hexagonal socket head.
Cap screws are also
manufactured with hexagonal
heads similar to the one
shown in Fig. 89, as well as
a variety of other head styles.
This illustration uses one of the
conventional methods of
representing threads.
A
A
A
80 to 82°
H
H
H
D
D
D
L
L
l
L
l
l
(a)
(b )
(c)
remains as the preload or clamping force. When tightening, the mechanic should, if
possible, hold the bolt head stationary and twist the nut; in this way the bolt shank
will not feel the thread-friction torque.
The head of a hexagon-head cap screw is slightly thinner than that of a hexagon-head bolt. Dimensions of hexagon-head cap screws are listed in Table A30.
Hexagon-head cap screws are used in the same applications as bolts and also in applications in which one of the clamped members is threaded. Three other common capscrew head styles are shown in Fig. 810.
A variety of machine-screw head styles are shown in Fig. 811. Inch-series
machine screws are generally available in sizes from No. 0 to about 3 in.
8
Several styles of hexagonal nuts are illustrated in Fig. 812; their dimensions are
given in Table A31. The material of the nut must be selected carefully to match that
of the bolt. During tightening, the rst thread of the nut tends to take the entire load;
but yielding occurs, with some strengthening due to the cold work that takes place,
and the load is eventually divided over about three nut threads. For this reason you
should never reuse nuts; in fact, it can be dangerous to do so.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
410
III. Design of Mechanical
Elements
413
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Types of heads used on
machine screws.
A
A
D
H
80 to 82°
Figure 811
D
H
L
(a) Round head
80 to 82°
(b) Flat head
A
A
D
H
L
D
H
L
(c) Fillister head
L
(d) Oval head
5° ±3°
A
A
D
D
R
H
L
L
(e) Truss head
( f ) Binding head
D
D
W
W
H
L
H
(g) Hex head (trimmed)
Figure 812
W
Hexagonal nuts: (a) end view,
general; (b) washer-faced
regular nut; (c) regular nut
chamfered on both sides;
(d) jam nut with washer face;
(e) jam nut chamfered on both
sides.
84
(h) Hex head (upset)
H
1
Approx. 64 in
30
(a)
L
H
H
30
( b)
( c)
Approx.
1
64
in
H
30
30
( d)
(e)
JointsFastener Stiffness
When a connection is desired that can be disassembled without destructive methods
and that is strong enough to resist external tensile loads, moment loads, and shear
loads, or a combination of these, then the simple bolted joint using hardened-steel
washers is a good solution. Such a joint can also be dangerous unless it is properly
designed and assembled by a trained mechanic.
414
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
Figure 813
P
411
P
A bolted connection loaded in
tension by the forces P. Note
the use of two washers. Note
how the threads extend into
the body of the connection.
This is usual and is desired. l
is the grip of the connection.
l
P
P
Figure 814
Section of cylindrical pressure
vessel. Hexagon-head cap
screws are used to fasten the
cylinder head to the body.
Note the use of an O-ring seal.
l is the effective grip of the
connection (see Table 87).
l'
A section through a tension-loaded bolted joint is illustrated in Fig. 813. Notice
the clearance space provided by the bolt holes. Notice, too, how the bolt threads
extend into the body of the connection.
As noted previously, the purpose of the bolt is to clamp the two, or more, parts
together. Twisting the nut stretches the bolt to produce the clamping force. This clamping
force is called the pretension or bolt preload. It exists in the connection after the nut has
been properly tightened no matter whether the external tensile load P is exerted or not.
Of course, since the members are being clamped together, the clamping force that
produces tension in the bolt induces compression in the members.
Figure 814 shows another tension-loaded connection. This joint uses cap screws
threaded into one of the members. An alternative approach to this problem (of not using
a nut) would be to use studs. A stud is a rod threaded on both ends. The stud is screwed
into the lower member rst; then the top member is positioned and fastened down
with hardened washers and nuts. The studs are regarded as permanent, and so the joint
can be disassembled merely by removing the nut and washer. Thus the threaded part
of the lower member is not damaged by reusing the threads.
The spring rate is a limit as expressed in Eq. (41). For an elastic member such
as a bolt, as we learned in Eq. (42), it is the ratio between the force applied to the
member and the deection produced by that force. We can use Eq. (44) and the results
of Prob. 41 to nd the stiffness constant of a fastener in any bolted connection.
The grip l of a connection is the total thickness of the clamped material. In
Fig. 813 the grip is the sum of the thicknesses of both members and both washers.
In Fig. 814 the effective grip is given in Table 87.
The stiffness of the portion of a bolt or screw within the clamped zone will generally consist of two parts, that of the unthreaded shank portion and that of the
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
412
III. Design of Mechanical
Elements
415
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 87
Suggested Procedure for Finding Fastener Stiffness
ld
h
t1
t
lt
H
t
t2
d
d
lt
LT
l'
l
LT
ld
L
L
(a)
(b)
Given fastener diameter d
and pitch p or number of threads
l =
Grip is thickness l
Washer thickness from
Table A32 or A33
Threaded length LT
Inch series:
LT =
Fastener length: L > l
H
2d +
2d +
1
4
1
2
in,
in,
Effective grip
h + t2 /2,
h + d /2,
t2 < d
t2 d
L 6 in
L > 6 in
Metric series:
2d + 6 mm, L 125, d 48 mm
LT = 2d + 12 mm, 125 < L 200 mm
2d + 25 mm, L > 200 mm
Fastener length: L > h
1.5d
Round up using Table A17
Length of useful unthreaded
portion: ld L LT
Length of threaded portion:
lt l ld
Length of useful
portion: ld L
Length of useful
portion: lt l
unthreaded
LT
threaded
ld
Area of unthreaded portion:
Ad π d 2 4
Area of threaded portion:
At, Table 81 or 82
Fastener stiffness:
AdAtE
kb =
A d l t + A t ld
*Bolts and cap screws may not be available in all the preferred lengths listed in Table A17. Large fasteners may not be available in fractional inches or in millimeter lengths ending in
a nonzero digit. Check with your bolt supplier for availability.
416
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
413
threaded portion. Thus the stiffness constant of the bolt is equivalent to the stiffnesses
of two springs in series. Using the results of Prob. 41, we nd
1
1
1
+
=
k
k1
k2
k=
or
k1 k2
k1 + k2
(815)
for two springs in series. From Eq. (44), the spring rates of the threaded and
unthreaded portions of the bolt in the clamped zone are, respectively,
kt =
where
At E
lt
kd =
Ad E
ld
(816)
At = tensile-stress area (Tables 81, 82)
lt = length of threaded portion of grip
Ad = major-diameter area of fastener
ld = length of unthreaded portion in grip
Substituting these stiffnesses in Eq. (815) gives
kb =
Ad At E
Ad lt + At ld
(817)
where kb is the estimated effective stiffness of the bolt or cap screw in the clamped
zone. For short fasteners, the one in Fig. 814, for example, the unthreaded area is
small and so the rst of the expressions in Eq. (816) can be used to nd kb . For long
fasteners, the threaded area is relatively small, and so the second expression in Eq.
(816) can be used. Table 87 is useful.
85
JointsMember Stiffness
In the previous section, we determined the stiffness of the fastener in the clamped zone.
In this section, we wish to study the stiffnesses of the members in the clamped zone.
Both of these stiffnesses must be known in order to learn what happens when the
assembled connection is subjected to an external tensile loading.
There may be more than two members included in the grip of the fastener. All
together these act like compressive springs in series, and hence the total spring rate
of the members is
1
1
1
1
1
=
+
+
+ ··· +
km
k1
k2
k3
ki
(818)
If one of the members is a soft gasket, its stiffness relative to the other members is
usually so small that for all practical purposes the others can be neglected and only
the gasket stiffness used.
If there is no gasket, the stiffness of the members is rather difcult to obtain,
except by experimentation, because the compression spreads out between the bolt head
and the nut and hence the area is not uniform. There are, however, some cases in
which this area can be determined.
Ito2 has used ultrasonic techniques to determine the pressure distribution at the member interface. The results show that the pressure stays high out to about 1.5 bolt radii.
2
Y. Ito, J. Toyoda, and S. Nagata, Interface Pressure Distribution in a Bolt-Flange Assembly, ASME paper
no. 77-WA/DE-11, 1977.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
414
III. Design of Mechanical
Elements
417
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Figure 815
D
x
Compression of a member
with the equivalent elastic
properties represented by a
frustum of a hollow cone.
Here, l represents the grip
length.
y
dw
t
y
x
l
2
d
t
d
dx
x
(a)
(b )
The pressure, however, falls off farther away from the bolt. Thus Ito suggests the use
of Rotschers pressure-cone method for stiffness calculations with a variable cone
angle. This method is quite complicated, and so here we choose to use a simpler
approach using a xed cone angle.
Figure 815 illustrates the general cone geometry using a half-apex angle α . An
angle α = 45 has been used, but Little3 reports that this overestimates the clamping
stiffness. When loading is restricted to a washer-face annulus (hardened steel, cast
iron, or aluminum), the proper apex angle is smaller. Osgood4 reports a range of
25 α 33 for most combinations. In this book we shall use α = 30 except in
cases in which the material is insufcient to allow the frusta to exist.
Referring now to Fig. 815b, the contraction of an element of the cone of thickness dx subjected to a compressive force P is, from Eq. (43),
dδ =
P dx
EA
(a)
The area of the element is
2
A = π ro ri2 = π
= π x tan α +
x tan α +
D+d
2
D
2
2
x tan α +
d
2
2
Dd
2
(b)
Substituting this in Eq. (a) and integrating gives a total contraction of
δ=
t
P
πE
0
dx
[x tan α + ( D + d )/2][x tan α + ( D d )/2]
(c)
Using a table of integrals, we nd the result to be
δ=
P
(2t tan α + D d )( D + d )
ln
π Ed tan α (2t tan α + D + d )( D d )
(d )
Thus the spring rate or stiffness of this frustum is
k=
P
=
δ
π Ed tan α
(2t tan α + D d )( D + d )
ln
(2t tan α + D + d )( D d )
3
R. E. Little, Bolted Joints: How Much Give? Machine Design, Nov. 9, 1967.
4
C. C. Osgood, Saving Weight on Bolted Joints, Machine Design, Oct. 25, 1979.
(819)
418
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
415
With α = 30 , this becomes
k=
0.5774π Ed
(1.155t + D d )( D + d )
ln
(1.155t + D + d )( D d )
(820)
Equation (820), or (819), must be solved separately for each frustum in the
joint. Then individual stiffnesses are assembled to obtain km using Eq. (818).
If the members of the joint have the same Youngs modulus E with symmetrical
frusta back to back, then they act as two identical springs in series. From Eq. (818)
we learn that km = k /2. Using the grip as l = 2t and dw as the diameter of the washer
face, we nd the spring rate of the members to be
km =
π Ed tan α
(l tan α + dw d ) (dw + d )
2 ln
(l tan α + dw + d ) (dw d )
(821)
The diameter of the washer face is about 50 percent greater than the fastener diameter for standard hexagon-head bolts and cap screws. Thus we can simplify Eq. (821)
by letting dw = 1.5d . If we also use α = 30 , then Eq. (821) can be written as
km =
0.5774π Ed
0.5774l + 0.5d
2 ln 5
0.5774l + 2.5d
(822)
It is easy to program the numbered equations in this section, and you should do so.
The time spent in programming will save many hours of formula plugging.
To see how good Eq. (821) is, solve it for km / Ed :
km
=
Ed
π tan α
(l tan α + dw d ) (dw + d )
2 ln
(l tan α + dw + d ) (dw d )
Earlier in the section use of α = 30 was recommended for hardened steel, cast iron,
or aluminum members. Wileman, Choudury, and Green5 conducted a nite element
study of this problem. The results, which are depicted in Fig. 816, agree with the
α = 30 recommendation, coinciding exactly at the aspect ratio d / l = 0.4. Additionally, they offered an exponential curve-t of the form
km
= A exp( Bd / l )
Ed
(823)
with constants A and B dened in Table 88. For standard washer faces and members of the same material, Eq. (823) offers a simple calculation for member stiffness km . For departure from these conditions, Eq. (820) remains the basis for
approaching the problem.
5
J.Wileman, M. Choudury, and I. Green, Computation of Member Stiffness in Bolted Connections, Trans.
ASME, J. Mech. Design, vol. 113, December 1991, pp. 432437.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
416
III. Design of Mechanical
Elements
419
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Figure 816
3.4
3.2
3.0
2.8
2.6
Dimensionless stiffness, k m Ed
The dimensionless plot of
stiffness versus aspect ratio of
the members of a bolted joint,
showing the relative accuracy
of methods of Rotscher,
Mischke, and Motosh,
compared to a nite-element
analysis (FEA) conducted by
Wileman, Choudury, and
Green.
2.4
2.2
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.1
0.3
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
Aspect ratio, d l
FEA
Table 88
Stiffness Parameters
of Various Member
Materials
Source: J. Wileman, M.
Choudury, and I. Green,
Computation of Member
Stiffness in Bolted
Connections, Trans. ASME,
J. Mech. Design, vol. 113,
December 1991,
pp. 432437.
Rotscher
Material
Used
Mischke 45°
Mischke 30°
Modulus
Mpsi
Motosh
Poisson
Ratio
Elastic
GPa
Steel
0.291
207
30.0
0.787 15
0.628 73
Aluminum
0.334
71
10.3
0.796 70
0.638 16
A
B
Copper
0.326
119
17.3
0.795 68
0.635 53
Gray cast iron
0.211
100
14.5
0.778 71
0.616 16
0.789 52
0.629 14
General expression
EXAMPLE 82
Two 1 -in-thick steel plates with a modulus of elasticity of 30(106 ) psi are clamped
2
by washer-faced 1 -in-diameter UNC SAE grade 5 bolts with a 0.095-in-thick washer
2
under the nut. Find the member spring rate km using the method of conical frusta,
and compare the result with the nite element analysis (FEA) curve-t method of
Wileman et al.
Solution
The grip is 0.5 + 0.5 + 0.095 = 1.095 in. Using Eq. (822) with l = 1.095 and
d = 0.5 in, we write
km =
0.5774π 30(106 )0.5
0.5774(1.095) + 0.5(0.5)
2 ln 5
0.5774(1.095) + 2.5(0.5)
= 15.97(106 ) lbf/in
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
417
From Table 88, A = 0.787 15, B = 0.628 73. Equation (823) gives
km = 30(106 )(0.5)(0.787 15) exp[0.628 73(0.5)/1.095]
= 15.73(106 ) lbf/in
For this case, the difference between the results for Eqs. (822) and (823) is less
than 2 percent.
86
Bolt Strength
In the specication standards for bolts, the strength is specied by stating ASTM minimum quantities, the minimum proof strength, or minimum proof load, and the minimum tensile strength.
The proof load is the maximum load (force) that a bolt can withstand without
acquiring a permanent set. The proof strength is the quotient of the proof load and
the tensile-stress area. The proof strength thus corresponds roughly to the proportional
limit and corresponds to 0.0001 in permanent set in the fastener (rst measurable deviation from elastic behavior). The value of the mean proof strength, the mean tensile
strength, and the corresponding standard deviations are not part of the specication
codes, so it is the designers responsibility to obtain these values, perhaps by laboratory testing, before designing to a reliability specication. Figure 817 shows the distribution of ultimate tensile strength from a bolt production run. If the ASTM minimum strength equals or exceeds 120 kpsi, the bolts can be offered as SAE grade 5.
The designer does not see this histogram. Instead, in Table 89, the designer sees the
entry Sut = 120 kpsi under the 1 1-in size in grade 5 bolts. Similarly, minimum
4
strengths are shown in Tables 810 and 811.
The SAE specications are found in Table 89. The bolt grades are numbered
according to the tensile strengths, with decimals used for variations at the same
strength level. Bolts and screws are available in all grades listed. Studs are available
in grades 1, 2, 4, 5, 8, and 8.1. Grade 8.1 is not listed.
120
Figure 817
Histogram of bolt ultimate
tensile strength based on
539 tests displaying a mean
ultimate tensile strength
¯
Sut = 145.1 kpsi and a
standard deviation of
σ Sut = 10.3 kpsi.
ˆ
100
Number of specimens
420
80
60
40
20
0
0
120
130
140
150
160
Tensile strength, Sut , kpsi
170
180
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
418
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 89
SAE Specications for Steel Bolts
Minimum
Proof
Strength,*
kpsi
Minimum
Tensile
Strength,*
kpsi
Minimum
Yield
Strength,*
kpsi
1
1
1 1
4
2
33
60
36
Low or medium carbon
2
13
44
55
74
57
Low or medium carbon
7
1 1
8
2
33
60
36
4
1
1 1
4
2
65
115
100
5
1
1
4
85
120
92
1 1 1 1
8
2
74
105
81
5.2
1
1
4
85
120
92
7
1
1 1
4
2
105
133
115
Medium-carbon alloy, Q&T
8
1
1 1
4
2
120
150
130
Medium-carbon alloy, Q&T
8.2
1
1
4
120
150
130
Low-carbon martensite, Q&T
SAE
Grade
No.
Size
Range
Inclusive,
in
Material
Head Marking
Medium carbon, cold-drawn
Medium carbon, Q&T
Low-carbon martensite, Q&T
*Minimum strengths are strengths exceeded by 99 percent of fasteners.
ASTM specications are listed in Table 810. ASTM threads are shorter because
ASTM deals mostly with structures; structural connections are generally loaded in
shear, and the decreased thread length provides more shank area.
Specications for metric fasteners are given in Table 811.
It is worth noting that all specication-grade bolts made in this country bear a manufacturers mark or logo, in addition to the grade marking, on the bolt head. Such marks
conrm that the bolt meets or exceeds specications. If such marks are missing, the bolt
may be imported; for imported bolts there is no obligation to meet specications.
Bolts in fatigue axial loading fail at the llet under the head, at the thread runout,
and at the rst thread engaged in the nut. If the bolt has a standard shoulder under
421
422
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
419
Table 810
ASTM Specications for Steel Bolts
ASTM
Size
DesigRange,
nation Inclusive,
No.
in
Minimum
Proof
Strength,*
kpsi
Minimum
Tensile
Strength,*
kpsi
A307
1
1 1
4
2
33
60
36
Low carbon
A325,
1
1
2
85
120
92
Medium carbon, Q&T
type 1
1 1 1 1
8
2
74
105
81
A325,
1
1
2
85
120
92
Low-carbon, martensite,
type 2
1 1 1 1
8
2
74
105
81
Q&T
A325,
1
1
2
85
120
92
Weathering steel,
type 3
1 1 1 1
8
2
74
105
81
Q&T
A354,
1
2 1
4
2
105
125
109
2 3 4
4
95
115
99
1
4
4
120
150
130
1
1
4
85
120
92
1 1 1 1
8
2
74
105
81
1 3 3
4
55
90
58
1
1 1
2
2
120
150
130
1
1 1
2
2
120
150
130
grade BC
A354,
Minimum
Yield
Strength,*
kpsi
Material
Head Marking
A325
A325
A325
Alloy steel, Q&T
BC
Alloy steel, Q&T
grade BD
A449
A490,
Medium-carbon, Q&T
Alloy steel, Q&T
type 1
A490,
type 3
*Minimum strengths are strengths exceeded by 99 percent of fasteners.
A490
Weathering steel,
Q&T
A490
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
420
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 811
Metric Mechanical-Property Classes for Steel Bolts, Screws, and Studs*
Property
Class
4.6
Size
Range,
Inclusive
Minimum
Proof
Strength,
MPa
Minimum
Tensile
Strength,
MPa
Minimum
Yield
Strength,
MPa
Material
M5M36
225
400
240
Low or medium carbon
Head Marking
4.6
4.8
M1.6M16
310
420
340
Low or medium carbon
4.8
5.8
M5M24
380
520
420
Low or medium carbon
5.8
8.8
M16M36
600
830
660
Medium carbon, Q&T
8.8
9.8
M1.6M16
650
900
720
Medium carbon, Q&T
9.8
10.9
M5M36
830
1040
940
Low-carbon martensite,
Q&T
12.9
M1.6M36
970
1220
1100
10.9
Alloy, Q&T
12.9
*The thread length for bolts and cap screws is
2d + 6
L T = 2d + 12
2d + 25
L 125
125 < L 200
L > 200
where L is the bolt length. The thread length for structural bolts is slightly shorter than given above.
strengths are strength exceeded by 99 percent of fasteners.
Minimum
the head, it has a value of K f from 2.1 to 2.3, and this shoulder llet is protected
from scratching or scoring by a washer. If the thread runout has a 15 or less halfcone angle, the stress is higher at the rst engaged thread in the nut. Bolts are sized
by examining the loading at the plane of the washer face of the nut. This is the weakest part of the bolt if and only if the conditions above are satised (washer protection
of the shoulder llet and thread runout 15 ). Inattention to this requirement has led
to a record of 15 percent fastener fatigue failure under the head, 20 percent at thread
runout, and 65 percent where the designer is focusing attention. It does little good to
concentrate on the plane of the nut washer face if it is not the weakest location.
423
424
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
421
Nuts are graded so that they can be mated with their corresponding grade of bolt.
The purpose of the nut is to have its threads deect to distribute the load of the bolt
more evenly to the nut. The nuts properties are controlled in order to accomplish this.
The grade of the nut should be the grade of the bolt.
87
Tension JointsThe External Load
Let us now consider what happens when an external tensile load P, as in Fig. 813,
is applied to a bolted connection. It is to be assumed, of course, that the clamping
force, which we will call the preload Fi , has been correctly applied by tightening the
nut before P is applied. The nomenclature used is:
Fi = preload
P = external tensile load
Pb = portion of P taken by bolt
Pm = portion of P taken by members
Fb = Pb + Fi = resultant bolt load
Fm = Pm Fi = resultant load on members
C = fraction of external load P carried by bolt
1 C = fraction of external load P carried by members
The load P is tension, and it causes the connection to stretch, or elongate, through
some distance δ . We can relate this elongation to the stiffnesses by recalling that k is
the force divided by the deection. Thus
δ=
Pb
kb
δ=
and
Pm
km
(a)
or
Pm = Pb
km
kb
(b)
Since P = Pb + Pm , we have
Pb =
kb P
= CP
kb + km
(c)
and
Pm = P Pb = (1 C ) P
(d )
where
C=
kb
kb + km
(e)
is called the stiffness constant of the joint. The resultant bolt load is
Fb = Pb + Fi = C P + Fi
Fm < 0
(824)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
422
III. Design of Mechanical
Elements
425
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 812
Stiffnesses, M lbf/in
Computation of Bolt
and Member Stiffnesses.
Steel members clamped
using a 1 in-13 NC
2
kb
steel bolt. C =
Bolt Grip, in
kb
km
C
1
C
2
2.57
12.69
0.168
0.832
3
1.79
11.33
0.136
0.864
4
1.37
10.63
0.114
0.886
kb + km
and the resultant load on the connected members is
Fm = Pm Fi = (1 C ) P Fi
Fm < 0
(825)
Of course, these results are valid only as long as some clamping load remains in the
members; this is indicated by the qualier in the equations.
Table 812 is included to provide some information on the relative values of the
stiffnesses encountered. The grip contains only two members, both of steel, and no
washers. The ratios C and 1 C are the coefcients of P in Eqs. (824) and (825),
respectively. They describe the proportion of the external load taken by the bolt and
by the members, respectively. In all cases, the members take over 80 percent of the
external load. Think how important this is when fatigue loading is present. Note also
that making the grip longer causes the members to take an even greater percentage
of the external load.
88
Relating Bolt Torque to Bolt Tension
Having learned that a high preload is very desirable in important bolted connections,
we must next consider means of ensuring that the preload is actually developed when
the parts are assembled.
If the overall length of the bolt can actually be measured with a micrometer when
it is assembled, the bolt elongation due to the preload Fi can be computed using the
formula δ = Fi l /( AE ). Then the nut is simply tightened until the bolt elongates
through the distance δ. This ensures that the desired preload has been attained.
The elongation of a screw cannot usually be measured, because the threaded end is
often in a blind hole. It is also impractical in many cases to measure bolt elongation. In
such cases the wrench torque required to develop the specied preload must be estimated.
Then torque wrenching, pneumatic-impact wrenching, or the turn-of-the-nut method may
be used.
The torque wrench has a built-in dial that indicates the proper torque.
With impact wrenching, the air pressure is adjusted so that the wrench stalls when
the proper torque is obtained, or in some wrenches, the air automatically shuts off at
the desired torque.
The turn-of-the-nut method requires that we rst dene the meaning of snug-tight.
The snug-tight condition is the tightness attained by a few impacts of an impact
wrench, or the full effort of a person using an ordinary wrench. When the snug-tight
condition is attained, all additional turning develops useful tension in the bolt. The
turn-of-the-nut method requires that you compute the fractional number of turns necessary to develop the required preload from the snug-tight condition. For example, for
heavy hexagonal structural bolts, the turn-of-the-nut specication states that the nut
should be turned a minimum of 180 from the snug-tight condition under optimum
426
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
Table 813
Distribution of Preload Fi
for 20 Tests of
Unlubricated Bolts
Torqued to 90 N · m
423
23.6,
27.6,
28.0,
29.4,
30.3,
30.7,
32.9,
33.8,
33.8,
33.8,
34.7,
35.6,
35.6,
37.4,
37.8,
37.8,
39.2,
40.0,
40.5,
42.7
ˆ
*Mean value Fi = 34.3 kN. Standard deviation, σ = 4.91 kN.
conditions. Note that this is also about the correct rotation for the wheel nuts of a passenger car. Problems 815 to 817 illustrate the method further.
Although the coefcients of friction may vary widely, we can obtain a good estimate
of the torque required to produce a given preload by combining Eqs. (85) and (86):
T=
l + π f dm sec α
π dm f l sec α
Fi dm
2
+
Fi f c dc
2
(a)
where dm is the average of the major and minor diameters. Since tan λ = l /π dm , we
divide the numerator and denominator of the rst term by π dm and get
T=
tan λ + f sec α
l f tan λ sec α
Fi dm
2
+
Fi f c dc
2
(b)
The diameter of the washer face of a hexagonal nut is the same as the width across
ats and equal to 1 1 times the nominal size. Therefore the mean collar diameter is
2
dc = (d + 1.5d )/2 = 1.25d . Equation (b) can now be arranged to give
T=
dm
2d
tan λ + f sec α
1 f tan λ sec α
+ 0.625 f c Fi d
(c)
We now dene a torque coefcient K as the term in brackets, and so
K=
dm
2d
tan λ + f sec α
1 f tan λ sec α
+ 0.625 f c
(826)
Equation (c) can now be written
T = K Fi d
(827)
The coefcient of friction depends upon the surface smoothness, accuracy, and
degree of lubrication. On the average, both f and f c are about 0.15. The interesting
.
fact about Eq. (826) is that K = 0.20 for f = f c = 0.15 no matter what size bolts
are employed and no matter whether the threads are coarse or ne.
Blake and Kurtz have published results of numerous tests of the torquing of bolts.6
By subjecting their data to a statistical analysis, we can learn something about the
distribution of the torque coefcients and the resulting preload. Blake and Kurtz determined the preload in quantities of unlubricated and lubricated bolts of size 1 in-20
2
UNF when torqued to 800 lbf · in. This corresponds roughly to an M12 × 1.25 bolt
torqued to 90 N · m. The statistical analyses of these two groups of bolts, converted
to SI units, are displayed in Tables 813 and 814.
6
J. C. Blake and H. J. Kurtz, The Uncertainties of Measuring Fastener Preload, Machine Design, vol. 37,
Sept. 30, 1965, pp. 128131.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
424
III. Design of Mechanical
Elements
427
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 814
30.3,
Distribution of Preload Fi
for 10 Tests of
Lubricated Bolts Torqued
to 90 N · m
Table 815
32.5,
32.5,
32.9,
33.8,
34.3,
34.7,
37.4,
40.5
ˆ
*Mean value, Fi = 34.18 kN. Standard deviation, σ = 2.88 kN.
Bolt Condition
Torque Factors K for Use
with Eq. (827)
32.9,
K
Nonplated, black nish
0.30
Zinc-plated
0.20
Lubricated
0.18
Cadmium-plated
0.16
With Bowman Anti-Seize
0.12
With Bowman-Grip nuts
0.09
We rst note that both groups have about the same mean preload, 34 kN. The
unlubricated bolts have a standard deviation of 4.9 kN and a COV of about 0.15. The
lubricated bolts have a standard deviation of 3 kN and a COV of about 0.9.
The means obtained from the two samples are nearly identical, approximately
34 kN; using Eq. (827), we nd, for both samples, K = 0.208.
Bowman Distribution, a large manufacturer of fasteners, recommends the values
shown in Table 815. In this book we shall use these values and use K = 0.2 when
the bolt condition is not stated.
EXAMPLE 83
Solution
A 3 in-16 UNF × 2 1 in SAE grade 5 bolt is subjected to a load P of 6 kip in a ten4
2
sion joint. The initial bolt tension is Fi = 25 kip. The bolt and joint stiffnesses are
kb = 6.50 and km = 13.8 Mlbf/in, respectively.
(a) Determine the preload and service load stresses in the bolt. Compare these to the
SAE minimum proof strength of the bolt.
(b) Specify the torque necessary to develop the preload, using Eq. (827).
(c) Specify the torque necessary to develop the preload, using Eq. (826) with f =
f c = 0.15.
From Table 82, At = 0.373 in2.
(a) The preload stress is
Answer
σi =
Fi
25
=
= 67.02 kpsi
At
0.373
The stiffness constant is
C=
6.5
kb
=
= 0.320
kb + km
6.5 + 13.8
428
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
425
From Eq. (824), the stress under the service load is
σb =
Answer
Fb
C P + Fi
P
=
=C
+ σi
At
At
At
= 0.320
6
+ 67.02 = 72.17 kpsi
0.373
From Table 89, the SAE minimum proof strength of the bolt is Sp = 85 kpsi. The
preload and service load stresses are respectively 21 and 15 percent less than the proof
strength.
(b) From Eq. (827), the torque necessary to achieve the preload is
T = K Fi d = 0.2(25)(103 )(0.75) = 3750 lbf · in
Answer
(c
) The minor diameter can be determined from the minor area in Table 82. Thus dr =
4 Ar /π = 4(0.351)/π = 0.6685 in. Thus, the mean diameter is dm = (0.75 +
0.6685)/2 = 0.7093 in. The lead angle is
λ = tan1
l
1
1
= tan1
= tan1
= 1.6066
π dm
π dm N
π(0.7093)(16)
For α = 30 , Eq. (826) gives
T=
0.7093
2(0.75)
tan 1.6066 + 0.15(sec 30 )
+ 0.625(0.15) 25(103 )(0.75)
1 0.15(tan 1.6066 )(sec 30 )
= 3551 lbf · in
which is 5.3 percent less than the value found in part (b).
89
Statically Loaded Tension Joint with Preload
Equations (824) and (825) represent the forces in a bolted joint with preload. The
tensile stress in the bolt can be found as in Ex. 83 as
σb =
CP
Fi
+
At
At
(a)
The limiting value of σb is the proof strength Sp . Thus, with the introduction of a
load factor n, Eq. (a) becomes
Cn P
Fi
+
= Sp
At
At
(b)
or
n=
Sp At Fi
CP
(828)
Here we have called n a load factor rather than a factor of safety, though the two
ideas are somewhat related. Any value of n > 1 in Eq. (828) ensures that the bolt
stress is less than the proof strength.
Another means of ensuring a safe joint is to require that the external load be
smaller than that needed to cause the joint to separate. If separation does occur, then
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
426
III. Design of Mechanical
Elements
429
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
the entire external load will be imposed on the bolt. Let P0 be the value of the external load that would cause joint separation. At separation, Fm = 0 in Eq. (825), and so
(1 C ) P0 Fi = 0
(c)
Let the factor of safety against joint separation be
n0 =
P0
P
(d )
Substituting P0 = n 0 P in Eq. (c), we nd
n0 =
Fi
P (1 C )
(829)
as a load factor guarding against joint separation.
Figure 818 is the stress-strain diagram of a good-quality bolt material. Notice that
there is no clearly dened yield point and that the diagram progresses smoothly up to
fracture, which corresponds to the tensile strength. This means that no matter how much
preload is given the bolt, it will retain its load-carrying capacity. This is what keeps the
bolt tight and determines the joint strength. The pre-tension is the muscle of the joint,
and its magnitude is determined by the bolt strength. If the full bolt strength is not used
in developing the pre-tension, then money is wasted and the joint is weaker.
Good-quality bolts can be preloaded into the plastic range to develop more
strength. Some of the bolt torque used in tightening produces torsion, which increases
the principal tensile stress. However, this torsion is held only by the friction of the
bolt head and nut; in time it relaxes and lowers the bolt tension slightly. Thus, as a
rule, a bolt will either fracture during tightening, or not at all.
Above all, do not rely too much on wrench torque; it is not a good indicator of
preload. Actual bolt elongation should be used whenever possibleespecially with
fatigue loading. In fact, if high reliability is a requirement of the design, then preload
should always be determined by bolt elongation.
Russell, Burdsall & Ward Inc. (RB&W) recommendations for preload are 60 kpsi
for SAE grade 5 bolts for nonpermanent connections, and that A325 bolts (equivalent
to SAE grade 5) used in structural applications be tightened to proof load or beyond
Sut
Figure 818
Typical stress-strain diagram
for bolt materials showing
proof strength Sp, yield
strength Sy, and ultimate
tensile strength Sut.
Sy
Stress
Sp
Strain
430
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
427
(85 kpsi up to a diameter of 1 in).7 Bowman8 recommends a preload of 75 percent of
proof load, which is about the same as the RB&W recommendations for reused bolts.
In view of these guidelines, it is recommended for both static and fatigue loading that
the following be used for preload:
Fi =
0.75 Fp
0.90 Fp
for nonpermanent connections, reused fasteners
for permanent connections
(830)
where Fp is the proof load, obtained from the equation
(831)
Fp = At Sp
Here Sp is the proof strength obtained from Tables 89 to 811. For other materials,
an approximate value is Sp = 0.85 Sy . Be very careful not to use a soft material in a
threaded fastener. For high-strength steel bolts used as structural steel connectors, if
advanced tightening methods are used, tighten to yield.
You can see that the RB&W recommendations on preload are in line with what
we have encountered in this chapter. The purposes of development were to give the
reader the perspective to appreciate Eqs. (830) and a methodology with which to
handle cases more specically than the recommendations.
7
Russell, Burdsall & Ward Inc., Helpful Hints for Fastener Design and Application, Mentor, Ohio, 1965, p. 42.
8
Bowman DistributionBarnes Group, Fastener Facts, Cleveland, 1985, p. 90.
EXAMPLE 84
Solution
Figure 819 is a cross section of a grade 25 cast-iron pressure vessel. A total of N
bolts are to be used to resist a separating force of 36 kip.
(a) Determine kb , km , and C.
(b) Find the number of bolts required for a load factor of 2 where the bolts may be
reused when the joint is taken apart.
(a) The grip is l = 1.50 in. From Table A31, the nut thickness is
2
threads beyond the nut of 11 in gives a bolt length of
L=
Figure 819
5
8
35
2
+ 1.50 +
= 2.229 in
64
11
in-11 UNC × 2 1 in grade 5
4
finished hex head bolt
No. 25 CI
3
4
in
3
4
in
35
64
in. Adding two
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
428
III. Design of Mechanical
Elements
431
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
From Table A17 the next fraction size bolt is L = 2 1 in. From Eq. (813), the thread
4
length is L T = 2(0.625) + 0.25 = 1.50 in. Thus the length of the unthreaded portion
in the grip is ld = 2.25 1.50 = 0.75 in. The threaded length in the grip is lt = l
ld = 0.75 in. From Table 82, At = 0.226 in2. The major-diameter area is Ad =
π(0.625)2 /4 = 0.3068 in2. The bolt stiffness is then
kb =
Answer
Ad At E
0.3068(0.226)(30)
=
Ad lt + At ld
0.3068(0.75) + 0.226(0.75)
= 5.21 Mlbf/in
From Table A24, for no. 25 cast iron we will use E = 14 Mpsi. The stiffness of the
members, from Eq. (822), is
km =
Answer
0.5774π Ed
0.5774l + 0.5d
2 ln 5
0.5774l + 2.5d
=
0.5774π(14)(0.625)
0.5774 (1.5) + 0.5 (0.625)
2 ln 5
0.5774 (1.5) + 2.5 (0.625)
= 8.95 Mlbf/in
If you are using Eq. (823), from Table 88, A = 0.778 71 and B = 0.616 16, and
km = Ed A exp( Bd / l )
= 14(0.625)(0.778 71) exp[0.616 16(0.625)/1.5]
= 8.81 Mlbf/in
which is only 1.6 percent lower than the previous result.
From the rst calculation for km , the stiffness constant C is
Answer
C=
kb
5.21
=
= 0.368
kb + km
5.21 + 8.95
(b) From Table 89, Sp = 85 kpsi. Then, using Eqs. (830) and (831), we nd the
recommended preload to be
Fi = 0.75 At Sp = 0.75(0.226)(85) = 14.4 kip
For N bolts, Eq. (828) can be written
n=
Sp At Fi
C( P/N )
(1)
or
N=
0.368(2)(36)
Cn P
=
= 5.52
Sp At Fi
85(0.226) 14.4
With six bolts, Eq. (1) gives
n=
85(0.226) 14.4
= 2.18
0.368(36/6)
which is greater than the required value. Therefore we choose six bolts and use the
recommended tightening preload.
432
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
810
429
Gasketed Joints
If a full gasket is present in the joint, the gasket pressure p is found by dividing the
force in the member by the gasket area per bolt. Thus, for N bolts,
p=
Fm
Ag / N
(a)
With a load factor n, Eq. (825) can be written as
Fm = (1 C )n P Fi
(b)
Substituting this into Eq. (a) gives the gasket pressure as
p = [ Fi n P (1 C )]
N
Ag
(832)
In full-gasketed joints uniformity of pressure on the gasket is important. To maintain
adequate uniformity of pressure adjacent bolts should not be placed more than six nominal diameters apart on the bolt circle. To maintain wrench clearance, bolts should be
placed at least three diameters apart. A rough rule for bolt spacing around a bolt circle is
3
π Db
6
Nd
(833)
where Db is the diameter of the bolt circle and N is the number of bolts.
811
Fatigue Loading of Tension Joints
Tension-loaded bolted joints subjected to fatigue action can be analyzed directly by the
methods of Chap. 6. Table 816 lists average fatigue stress-concentration factors for the
llet under the bolt head and also at the beginning of the threads on the bolt shank.
These are already corrected for notch sensitivity and for surface nish. Designers should
be aware that situations may arise in which it would be advisable to investigate these
factors more closely, since they are only average values. In fact, Peterson9 observes that
the distribution of typical bolt failures is about 15 percent under the head, 20 percent
at the end of the thread, and 65 percent in the thread at the nut face.
Use of rolled threads is the predominant method of thread-forming in screw fasteners, where Table 816 applies. In thread-rolling, the amount of cold work and strainstrengthening is unknown to the designer; therefore, fully corrected (including K f )
axial endurance strength is reported in Table 817. For cut threads, the methods of
Chap. 6 are useful. Anticipate that the endurance strengths will be considerably lower.
Most of the time, the type of fatigue loading encountered in the analysis of bolted
joints is one in which the externally applied load uctuates between zero and some
Table 816
SAE
Grade
Cut
Threads
3.6 to 5.8
2.2
2.8
2.1
4 to 8
9
Rolled
Threads
0 to 2
Fatigue StressConcentration Factors Kf
for Threaded Elements
Metric
Grade
6.6 to 10.9
3.0
3.8
2.3
Fillet
W. D. Pilkey, Petersons Stress Concentration Factors, 2nd ed., John Wiley & Sons, New York, 1997, p. 387.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
430
III. Design of Mechanical
Elements
433
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Table 817
Fully Corrected
Endurance Strengths for
Bolts and Screws with
Rolled Threads*
Grade or Class
Size Range
Endurance Strength
1
1 in
4
1
1
1 8 1 2 in
1
1
1 2 in
4
1
1
1 2 in
4
SAE 5
SAE 7
SAE 8
18.6 kpsi
16.3 kpsi
20.6 kpsi
23.2 kpsi
ISO 8.8
M16M36
129 MPa
ISO 9.8
M1.6M16
140 MPa
ISO 10.9
M5M36
162 MPa
ISO 12.9
M1.6M36
190 MPa
*Repeatedly-applied, axial loading, fully corrected.
Figure 820
Se
a
Load line
Alternating stress
Designers fatigue diagram
showing a Goodman failure
line and how a load line is
used to dene failure and
safety in preloaded bolted
joints in fatigue. Point B
represents nonfailure; point C,
failure.
1
1
C
Sa
B
a
A
Fi
i=A
t
m
D
Sm
Sut
Sa
Steady stress
m
maximum force P. This would be the situation in a pressure cylinder, for example, where
a pressure either exists or does not exist. For such cases, Fmax = Fb and Fmin = Fi and
the alternating component of the force is Fa = ( Fmax Fmin )/2 = ( Fb Fi )/2. Dividing this by At yields the alternating component of the bolt stress. Employing the notation from Sec. 87 with Eq. (824), we obtain
σa =
Fb Fi
(C P + Fi ) Fi
CP
=
=
2 At
2 At
2 At
(834)
The mean stress is equal to the alternating component plus the minimum stress, σi =
Fi / At , which results in
σm =
CP
Fi
+
2 At
At
(835)
On the designers fatigue diagram, shown in Fig. 820, the load line is
σm = σa + σi
(836)
434
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
431
The next problem is to nd the strength components Sa and Sm of the fatigue
failure line. These depend on the failure criteria:
Goodman:
Sa
Sm
+
=1
Se
Sut
(837)
Gerber:
2
Sm
Sut
Sa
+
Se
=1
(838)
ASME-elliptic:
Sa
Se
2
+
Sm
Sp
2
=1
(839)
For simultaneous solution between Eq. (836), as Sm = Sa + σi , and each of Eqs. (837)
to (839) gives
Goodman:
Sa =
Se ( Sut σi )
Sut + Se
Sm = Sa + σi
(840)
(841)
Gerber:
Sa =
1
2
2
Sut Sut + 4 Se ( Se + σi ) Sut 2σi Se
2 Se
(842)
Sm = Sa + σi
ASME-elliptic:
Sa =
2
Sp
Se
2
2
Sp Sp + Se σi2 σi Se
2
+ Se
(843)
Sm = Sa + σi
When using relations of this section, be sure to use Kf for both σa and σm . Otherwise, the slope of the load line will not remain 1 to 1.
Examination of Eqs. (837) to (843) shows parametric equations that relate the
coordinates of interest to the form of the criteria. The factor of safety guarding against
fatigue is given by
nf =
Sa
σa
(844)
Applying this to the Goodman criterion, for example, with Eqs. (834) and (840)
and σi = Fi / At gives
nf =
2 Se ( Sut At Fi )
C P ( Sut + Se )
(845)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
432
III. Design of Mechanical
Elements
435
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
when preload Fi is present. With no preload, C = 1, Fi = 0, and Eq. (845) becomes
nf0 =
2 Se Sut At
P ( Sut + Se )
(846)
Preload is benecial for resisting fatigue when n f / n f 0 is greater than unity. For Goodman, Eqs. (845) and (846) with n f / n f 0 1 puts an upper bound on the preload Fi
of
Fi (1 C ) Sut At
(847)
If this cannot be achieved, and nf is unsatisfactory, use the Gerber or ASME-elliptic criterion
to obtain a less conservative assessment. If the design is still not satisfactory, additional
bolts and/or a different size bolt may be called for. Bolts loosen, as they are friction devices,
and cyclic loading and vibration as well as other effects allow the fasteners to lose tension
with time. How does one ght loosening? Within strength limitations, the higher the preload the better. A rule of thumb is that preloads of 60 percent of proof load rarely loosen.
If more is better, how much more? Well, not enough to create reused fasteners as a future
threat. Alternatively, fastener-locking schemes can be employed.
After solving Eq. (844), you should also check the possibility of yielding, using
the proof strength
np =
Sp
σm + σa
(848)
EXAMPLE 85
Figure 821 shows a connection using cap screws. The joint is subjected to a uctuating force whose maximum value is 5 kip per screw. The required data are: cap screw,
1
5/8 in-11 NC, SAE 5; hardened-steel washer, tw = 16 in thick; steel cover plate, t1 =
5
5
E = 30 Mpsi; and cast-iron base, t2 = 8 in, E ci = 16 Mpsi.
8 in, s
(a) Find kb , km , and C using the assumptions given in the caption of Fig. 821.
(b) Find all factors of safety and explain what they mean.
Solution
(a) For the symbols of Figs. 815 and 821, h = t1 + tw = 0.6875 in, l = h + d /2 =
1 in, and D2 = 1.5d = 0.9375 in. The joint is composed of three frusta; the upper
two frusta are steel and the lower one is cast iron.
For the upper frustum: t = l /2 = 0.5 in, D = 0.9375 in, and E = 30 Mpsi. Using
these values in Eq. (820) gives k1 = 46.46 Mlbf/in.
Figure 821
Pressure-cone frustum member
model for a cap screw. For this
model the signicant sizes are
t2 < d
h + t 2 /2
l=
h + d /2
t2 d
D1 = dw + l tan α =
1.5d + 0.577l
D2 = dw = 1.5d
where l = effective grip. The
solutions are for α = 30 and
dw = 1.5d .
D1
l
l
2
t1
t2
d
D2
h
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
433
For the middle frustum: t = h l /2 = 0.1875 in and D = 0.9375 + 2(l h )
tan 30 = 1.298 in. With these and E s = 30 Mpsi, Eq. (820) gives k2 = 197.43 Mlbf/in.
The lower frustum has D = 0.9375 in, t = l h = 0.3125 in, and E ci = 16
Mpsi. The same equation yields k3 = 32.39 Mlbf/in.
Substituting these three stiffnesses into Eq. (818) gives km = 17.40 Mlbf/in. The
cap screw is short and threaded all the way. Using l = 1 in for the grip and
At = 0.226 in2 from Table 82, we find the stiffness to be kb = At E / l = 6.78
Mlbf/in. Thus the joint constant is
Answer
kb
6.78
=
= 0.280
kb + km
6.78 + 17.40
C=
(b) Equation (830) gives the preload as
Fi = 0.75 Fp = 0.75 At Sp = 0.75(0.226)(85) = 14.4 kip
where from Table 89, Sp = 85 kpsi for an SAE grade 5 cap screw. Using Eq. (828),
we obtain the load factor as
Answer
n=
Sp At Fi
85(0.226) 14.4
=
= 3.44
CP
0.280(5)
This factor prevents the bolt stress from becoming equal to the proof strength.
Next, using Eq. (829), we have
Answer
Fi
14.4
=
= 4.00
P (1 C )
5(1 0.280)
n0 =
If the force P gets too large, the joint will separate and the bolt will take the entire
load. This factor guards against that event.
For the remaining factors, refer to Fig. 822. This diagram contains the modied
Goodman line, the Gerber line, the proof-strength line, and the load line. The intersection
Figure 822
E
Sa
D
Sa
Sa
C
Sp
a
B
A
a
Designers fatigue diagram for
preloaded bolts, drawn to
scale, showing the modied
Goodman line, the Gerber
line, and the Langer proofstrength line, with an exploded
view of the area of interest.
The strengths used are
Sp = 85 kpsi, Se = 18.6 kpsi,
and Sut = 120 kpsi. The
coordinates are A, σi =
63.72 kpsi; B, σa = 3.10
kpsi, σm = 66.82 kpsi; C,
Sa = 7.55 kpsi, Sm = 71.29
kpsi; D, Sa = 10.64 kpsi,
Sm = 74.36 kpsi; E, Sa =
11.32 kpsi, Sm = 75.04 kpsi.
L
60
Stress amplitude
436
i
m
Sm
Sm
70
Sm
80
Sp
Proof
strength
line
Gerber line
L
Se
Modified Goodman line
i
Sp
Steady stress component
Sut
m
90
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
434
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
437
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
of the load line L with the respective failure lines at points C, D, and E denes a set of
strengths Sa and Sm at each intersection. Point B represents the stress state σa , σm . Point A
is the preload stress σi . Therefore the load line begins at A and makes an angle having a
unit slope. This angle is 45° only when both stress axes have the same scale.
The factors of safety are found by dividing the distances AC , AD , and A E by
the distance AB. Note that this is the same as dividing Sa for each theory by σa .
The quantities shown in the caption of Fig. 822 are obtained as follows:
Point A
σi =
Fi
14.4
=
= 63.72 kpsi
At
0.226
Point B
σa =
CP
0.280(5)
=
= 3.10 kpsi
2 At
2(0.226)
σm = σa + σi = 3.10 + 63.72 = 66.82 kpsi
Point C
This is the modied Goodman criteria. From Table 817, we nd Se = 18.6 kpsi.
Then, using Eq. (840), we get
Sa =
Se ( Sut σi )
18.6(120 63.72)
=
= 7.55 kpsi
Sut + Se
120 + 18.6
The factor of safety is found to be
Answer
nf =
Sa
7.55
=
= 2.44
σa
3.10
Point D
This is on the proof-strength line where
Sm + Sa = Sp
(1)
In addition, the horizontal projection of the load line AD is
Sm = σi + Sa
(2)
Solving Eqs. (1) and (2) simultaneously results in
Sa =
Sp σi
85 63.72
=
= 10.64 kpsi
2
2
The factor of safety resulting from this is
Answer
np =
Sa
10.64
=
= 3.43
σa
3.10
which, of course, is identical to the result previously obtained by using Eq. (828).
A similar analysis of a fatigue diagram could have been done using yield strength
instead of proof strength. Though the two strengths are somewhat related, proof
strength is a much better and more positive indicator of a fully loaded bolt than is the
yield strength. It is also worth remembering that proof-strength values are specied
in design codes; yield strengths are not.
438
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
435
Screws, Fasteners, and the Design of Nonpermanent Joints
We found n f = 2.44 on the basis of fatigue and the modied Goodman line, and
n p = 3.43 on the basis of proof strength. Thus the danger of failure is by fatigue, not
by overproof loading. These two factors should always be compared to determine
where the greatest danger lies.
Point E
For the Gerber criterion, from Eq. (842),
Sa =
=
1
2
2
Sut Sut + 4 Se ( Se + σi ) Sut 2σi Se
2 Se
1
120 1202 + 4(18.6)(18.6 + 63.72) 1202 2(63.72)(18.6)
2(18.6)
= 11.33 kpsi
Thus for the Gerber criterion the safety factor is
Answer
nf =
Sa
11.33
=
= 3.65
σa
3.10
which is greater than n p = 3.43 and contradicts the conclusion earlier that the danger of failure is fatigue. Figure 822 clearly shows the conict where point D lies
between points C and E. Again, the conservative nature of the Goodman criterion
explains the discrepancy and the designer must form his or her own conclusion.
812
Bolted and Riveted Joints Loaded in Shear10
Riveted and bolted joints loaded in shear are treated exactly alike in design and
analysis.
Figure 823a shows a riveted connection loaded in shear. Let us now study the
various means by which this connection might fail.
Figure 823b shows a failure by bending of the rivet or of the riveted members.
The bending moment is approximately M = Ft /2, where F is the shearing force and
t is the grip of the rivet, that is, the total thickness of the connected parts. The bending stress in the members or in the rivet is, neglecting stress concentration,
σ=
M
I /c
(849)
where I /c is the section modulus for the weakest member or for the rivet or rivets,
depending upon which stress is to be found. The calculation of the bending stress in
10
The design of bolted and riveted connections for boilers, bridges, buildings, and other structures in which
danger to human life is involved is strictly governed by various construction codes. When designing these
structures, the engineer should refer to the American Institute of Steel Construction Handbook, the American
Railway Engineering Association specications, or the Boiler Construction Code of the American Society of
Mechanical Engineers.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
436
III. Design of Mechanical
Elements
439
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Figure 823
Modes of failure in shear
loading of a bolted or riveted
connection: (a) shear loading;
(b) bending of rivet; (c) shear
of rivet; (d) tensile failure of
members; (e) bearing of rivet
on members or bearing of
members on rivet; (f ) shear
tear-out; (g) tensile tear-out.
(a )
(b )
(e )
( c)
(d )
(f)
( g)
this manner is an assumption, because we do not know exactly how the load is distributed to the rivet or the relative deformations of the rivet and the members.
Although this equation can be used to determine the bending stress, it is seldom used
in design; instead its effect is compensated for by an increase in the factor of safety.
In Fig. 823c failure of the rivet by pure shear is shown; the stress in the rivet is
τ=
F
A
(850)
where A is the cross-sectional area of all the rivets in the group. It may be noted that
it is standard practice in structural design to use the nominal diameter of the rivet
rather than the diameter of the hole, even though a hot-driven rivet expands and nearly
lls up the hole.
Rupture of one of the connected membes or plates by pure tension is illustrated
in Fig. 823d. The tensile stress is
σ=
F
A
(851)
where A is the net area of the plate, that is, the area reduced by an amount equal to
the area of all the rivet holes. For brittle materials and static loads and for either ductile or brittle materials loaded in fatigue, the stress-concentration effects must be
included. It is true that the use of a bolt with an initial preload and, sometimes, a rivet
will place the area around the hole in compression and thus tend to nullify the effects
of stress concentration, but unless denite steps are taken to ensure that the preload
does not relax, it is on the conservative side to design as if the full stress-concentration
effect were present. The stress-concentration effects are not considered in structural
design, because the loads are static and the materials ductile.
440
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
437
In calculating the area for Eq. (851), the designer should, of course, use the
combination of rivet or bolt holes that gives the smallest area.
Figure 823e illustrates a failure by crushing of the rivet or plate. Calculation of
this stress, which is usually called a bearing stress, is complicated by the distribution
of the load on the cylindrical surface of the rivet. The exact values of the forces acting upon the rivet are unknown, and so it is customary to assume that the components
of these forces are uniformly distributed over the projected contact area of the rivet.
This gives for the stress
σ =
F
A
(852)
where the projected area for a single rivet is A = td . Here, t is the thickness of the
thinnest plate and d is the rivet or bolt diameter.
Edge shearing, or tearing, of the margin is shown in Fig. 823f and g, respectively. In structural practice this failure is avoided by spacing the rivets at least 1 1
2
diameters away from the edge. Bolted connections usually are spaced an even greater
distance than this for satisfactory appearance, and hence this type of failure may usually be neglected.
In a rivet joint, the rivets all share the load in shear, bearing in the rivet, bearing
in the member, and shear in the rivet. Other failures are participated in by only some
of the joint. In a bolted joint, shear is taken by clamping friction, and bearing does
not exist. When bolt preload is lost, one bolt begins to carry the shear and bearing
until yielding slowly brings other fasteners in to share the shear and bearing. Finally,
all participate, and this is the basis of most bolted-joint analysis if loss of bolt preload is complete. The usual analysis involves
Bearing in the bolt (all bolts participate)
Bearing in members (all holes participate)
Shear of bolt (all bolts participate eventually)
Distinguishing between thread and shank shear
Edge shearing and tearing of member (edge bolts participate)
Tensile yielding of member across bolt holes
Checking member capacity
EXAMPLE 86
Two 1- by 4-in 1018 cold-rolled steel bars are butt-spliced with two 1 - by 4-in 1018
2
cold-rolled splice plates using four 3 in-16 UNF grade 5 bolts as depicted in Fig.
4
824. For a design factor of n d = 1.5 estimate the static load F that can be carried if
the bolts lose preload.
Solution
From Table A20, minimum strengths of Sy = 54 kpsi and Sut = 64 kpsi are found for
the members, and from Table 89 minimum strengths of Sp = 85 kpsi and
Sut = 120 kpsi for the bolts are found.
F /2 is transmitted by each of the splice plates, but since the areas of the splice
plates are half those of the center bars, the stresses associated with the plates are the
same. So for stresses associated with the plates, the force and areas used will be those
of the center plates.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
438
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Figure 824
1
1 2 in
1
1
1 2 in
1 2 in
1
1 2 in
1
1 4 in
F
1
1 2 in
w
F
1
1 4 in
(a )
1
2
3
4
in
in - 16 UNF SAE grade 5
1in
F
1
2
F
in
(b )
Bearing in bolts, all bolts loaded:
Sp
F
=
σ=
2td
nd
F=
2(1) 3 85
2td Sp
4
=
= 85 kip
nd
1.5
Bearing in members, all bolts active:
σ=
( Sy )mem
F
=
2td
nd
F=
2(1) 3 54
2td ( Sy )mem
4
=
= 54 kip
nd
1.5
Shear of bolt, all bolts active: If the bolt threads do not extend into the shear planes
for four shanks:
τ=
Sp
F
= 0.577
4π d 2 /4
nd
F = 0.577π d 2
Sp
85
= 0.577π(0.75)2
= 57.8 kip
nd
1.5
If the bolt threads extend into a shear plane:
τ=
Sp
F
= 0.577
4 Ar
nd
F=
0.577(4) Ar Sp
0.577(4)0.351(85)
=
= 45.9 kip
nd
1.5
441
442
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
439
Edge shearing of member at two margin bolts: From Fig. 825,
τ=
0.577( Sy )mem
F
=
4at
nd
F=
4at 0.577( Sy )mem
4(1.125)(1)0.577(54)
= 93.5 kip
=
nd
1.5
Tensile yielding of members across bolt holes:
σ=
F=
F
42
42
=
( Sy )mem
nd
3
4
t
3
4
t ( Sy )mem
nd
=
42
3
4
(1)54
1.5
= 90 kip
Member yield:
F=
wt ( Sy )mem
4(1)54
= 144 kip
=
nd
1.5
On the basis of bolt shear, the limiting value of the force is 45.9 kip, assuming the
threads extend into a shear plane. However, it would be poor design to allow the
threads to extend into a shear plane. So, assuming a good design based on bolt shear,
the limiting value of the force is 57.8 kip. For the members, the bearing stress limits
the load to 54 kip.
Figure 825
Edge shearing of member.
Bolt
d
a
Shear Joints with Eccentric Loading
Integral to the analysis of a shear joint is locating the center of relative motion between
the two members. In Fig. 826 let A1 to A5 be the respective cross-sectional areas of
a group of ve pins, or hot-driven rivets, or tight-tting shoulder bolts. Under this
assumption the rotational pivot point lies at the centroid of the cross-sectional area pattern of the pins, rivets, or bolts. Using statics, we learn that the centroid G is located
¯
¯
by the coordinates x and y , where x1 and yi are the distances to the ith area center:
x=
¯
A1 x 1 + A2 x 2 + A3 x 3 + A4 x 4 + A5 x 5
=
A1 + A2 + A3 + A4 + A5
A1 y1 + A2 y2 + A3 y3 + A4 y4 + A5 y5
y=
¯
=
A1 + A2 + A3 + A4 + A5
n
1
Ai x i
Ai
n
1
n
1
Ai yi
n
1 Ai
(853)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
440
III. Design of Mechanical
Elements
443
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
Figure 826
y
Centroid of pins, rivets, or
bolts.
A3
A2
A4
G
A1
_
y
A5
O
x
_
x
Figure 827
w lbf in
M1
(a) Beam bolted at both ends
with distributed load; (b) freebody diagram of beam;
(c) enlarged view of bolt group
centered at O showing
primary and secondary
resultant shear forces.
O
M2
V2
V1
(b )
FA
'
FB
'
FB
"
w lbf in
A
O
FA
"
+
B
rB
rA
O
Beam
FC
'
rC
rD
FD
'
FD
"
(a )
C
D
FC
"
(c)
In many instances the centroid can be located by symmetry.
An example of eccentric loading of fasteners is shown in Fig. 827. This is a
portion of a machine frame containing a beam subjected to the action of a bending
load. In this case, the beam is fastened to vertical members at the ends with specially
prepared load-sharing bolts. You will recognize the schematic representation in Fig.
827b as a statically indeterminate beam with both ends xed and with moment and
shear reactions at each end.
For convenience, the centers of the bolts at the left end of the beam are drawn to a
larger scale in Fig. 827c. Point O represents the centroid of the group, and it is assumed
in this example that all the bolts are of the same diameter. Note that the forces shown
in Fig. 827c are the resultant forces acting on the pins with a net force and moment
equal and opposite to the reaction loads V1 and M1 acting at O. The total load taken by
each bolt will be calculated in three steps. In the rst step the shear V1 is divided equally
among the bolts so that each bolt takes F = V1 / n , where n refers to the number of bolts
in the group and the force F is called the direct load, or primary shear.
It is noted that an equal distribution of the direct load to the bolts assumes an
absolutely rigid member. The arrangement of the bolts or the shape and size of the
444
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
441
members sometimes justies the use of another assumption as to the division of the
load. The direct loads F are shown as vectors on the loading diagram (Fig. 827c).
The moment load, or secondary shear, is the additional load on each bolt due to
the moment M1 . If r A , r B , rC , etc., are the radial distances from the centroid to the
center of each bolt, the moment and moment loads are related as follows:
M1 = FA r A + FB r B + FC rC + · · ·
(a)
where the F are the moment loads. The force taken by each bolt depends upon its
radial distance from the centroid; that is, the bolt farthest from the centroid takes the
greatest load, while the nearest bolt takes the smallest. We can therefore write
F
F
FA
= B= C
rA
rB
rC
(b)
where again, the diameters of the bolts are assumed equal. If not, then one replaces
F in Eq. (b) with the shear stresses τ = 4 F /π d 2 for each bolt. Solving Eqs. (a)
and (b) simultaneously, we obtain
M1 r n
2
2
2
r A + r B + rC + · · ·
Fn =
(854)
where the subscript n refers to the particular bolt whose load is to be found. These
moment loads are also shown as vectors on the loading diagram.
In the third step the direct and moment loads are added vectorially to obtain the
resultant load on each bolt. Since all the bolts or rivets are usually the same size, only
that bolt having the maximum load need be considered. When the maximum load is
found, the strength may be determined by using the various methods already described.
EXAMPLE 87
Shown in Fig. 828 is a 15- by 200-mm rectangular steel bar cantilevered to a 250-mm
steel channel using four tightly tted bolts located at A, B, C, and D.
Figure 828
250
Dimensions in millimeters.
10
15
M16
C
2 bolts
F = 16 kN
B
60
200
O
D
60
A
75
75
50
300
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
442
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
For a F = 16 kN load nd
(a) The resultant load on each bolt
(b) The maximum shear stress in each bolt
(c) The maximum bearing stress
(d) The critical bending stress in the bar
Solution
(a) Point O, the centroid of the bolt group in Fig. 828, is found by symmetry. If a
free-body diagram of the beam were constructed, the shear reaction V would pass
through O and the moment reactions M would be about O. These reactions are
V = 16 kN
M = 16(425) = 6800 N · m
In Fig. 829, the bolt group has been drawn to a larger scale and the reactions
are shown. The distance from the centroid to the center of each bolt is
(60)2 + (75)2 = 96.0 mm
r=
The primary shear load per bolt is
F =
V
16
=
= 4 kN
n
4
Since the secondary shear forces are equal, Eq. (854) becomes
F =
Mr
M
6800
=
=
= 17.7 kN
2
4r
4r
4(96.0)
The primary and secondary shear forces are plotted to scale in Fig. 829 and the resultants obtained by using the parallelogram rule. The magnitudes are found by measurement
Figure 829
y
FC
"
FC
B
C
'
FC
FB
'
rB
rC
FB
"
FB
O
"
FD
M
V
rA
rD
FD
D
A
'
FA
'
FD
"
FA
FA
x
445
446
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Screws, Fasteners, and the Design of Nonpermanent Joints
443
(or analysis) to be
Answer
FA = FB = 21.0 kN
Answer
FC = FD = 14.8 kN
(b) Bolts A and B are critical because they carry the largest shear load. Does this shear
act on the threaded portion of the bolt, or on the unthreaded portion? The bolt length
will be 25 mm plus the height of the nut plus about 2 mm for a washer. Table A31
gives the nut height as 14.8 mm. Including two threads beyond the nut, this adds up
to a length of 43.8 mm, and so a bolt 46 mm long will be needed. From Eq. (814)
we compute the thread length as L T = 38 mm. Thus the unthreaded portion of the bolt
is 46 38 = 8 mm long. This is less than the 15 mm for the plate in Fig. 828, and
so the bolt will tend to shear across its minor diameter. Therefore the shear-stress area
is As = 144 mm2, and so the shear stress is
Answer
τ=
F
21.0(10)3
=
= 146 MPa
As
144
(c) The channel is thinner than the bar, and so the largest bearing stress is due to the
pressing of the bolt against the channel web. The bearing area is Ab = td = 10(16) =
160 mm2. Thus the bearing stress is
Answer
σ =
F
21.0(10)3
=
= 131 MPa
Ab
160
(d ) The critical bending stress in the bar is assumed to occur in a section parallel to
the y axis and through bolts A and B. At this section the bending moment is
M = 16(300 + 50) = 5600 N · m
The second moment of area through this section is obtained by the use of the transfer formula, as follows:
¯
I = Ibar 2( Iholes + d 2 A)
=
15(16)3
15(200)3
2
+ (60)2 (15)(16) = 8.26(10)6 mm4
12
12
Then
Answer
σ=
Mc
5600(100)
(10)3 = 67.8 MPa
=
I
8.26(10)6
PROBLEMS
81
A power screw is 25 mm in diameter and has a thread pitch of 5 mm.
(a) Find the thread depth, the thread width, the mean and root diameters, and the lead, provided square threads are used.
(b) Repeat part (a) for Acme threads.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
444
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
82
Using the information in the footnote of Table 81, show that the tensile-stress area is
At =
83
π
(d 0.938 194 p)2
4
Show that for zero collar friction the efciency of a square-thread screw is given by the equation
e = tan λ
1 f tan λ
tan λ + f
Plot a curve of the efciency for lead angles up to 45 . Use f = 0.08.
84
A single-threaded 25-mm power screw is 25 mm in diameter with a pitch of 5 mm. A vertical
load on the screw reaches a maximum of 6 kN. The coefcients of friction are 0.05 for the
collar and 0.08 for the threads. The frictional diameter of the collar is 40 mm. Find the overall efciency and the torque to raise and lower the load.
85
The machine shown in the gure can be used for a tension test but not for a compression test.
Why? Can both screws have the same hand?
Motor
Bearings
Worm
Spur gears
[
Problem 85
Bronze
bushings
2 's
C.I.
Collar
bearing
B
C
2 [ 's
Foot
A
86
The press shown for Prob. 85 has a rated load of 5000 lbf. The twin screws have Acme threads,
1
a diameter of 3 in, and a pitch of 2 in. Coefcients of friction are 0.05 for the threads and 0.06
for the collar bearings. Collar diameters are 5 in. The gears have an efciency of 95 percent
and a speed ratio of 75:1. A slip clutch, on the motor shaft, prevents overloading. The full-load
motor speed is 1720 rev/min.
(a) When the motor is turned on, how fast will the press head move?
(b) What should be the horsepower rating of the motor?
87
3
A screw clamp similar to the one shown in the gure has a handle with diameter 16 in made
7
of cold-drawn AISI 1006 steel. The overall length is 3 in. The screw is 16 in-14 UNC and is
3
3
5 4 in long, overall. Distance A is 2 in. The clamp will accommodate parts up to 4 16 in high.
(a) What screw torque will cause the handle to bend permanently?
(b) What clamping force will the answer to part (a) cause if the collar friction is neglected and
if the thread friction is 0.075?
447
448
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Screws, Fasteners, and the Design of Nonpermanent Joints
445
(c) What clamping force will cause the screw to buckle?
(d ) Are there any other stresses or possible failures to be checked?
Problem 87
A
B
88
The C clamp shown in the gure for Prob. 87 uses a 5 in-6 Acme thread. The frictional coef8
cients are 0.15 for the threads and for the collar. The collar, which in this case is the anvil
7
strikers swivel joint, has a friction diameter of 16 in. Calculations are to be based on a max3
imum force of 6 lbf applied to the handle at a radius of 2 4 in from the screw centerline. Find
the clamping force.
89
Find the power required to drive a 40-mm power screw having double square threads with a
pitch of 6 mm. The nut is to move at a velocity of 48 mm/s and move a load of F = 10 kN.
The frictional coefcients are 0.10 for the threads and 0.15 for the collar. The frictional diameter of the collar is 60 mm.
810
A single square-thread power screw has an input power of 3 kW at a speed of 1 rev/s. The
screw has a diameter of 36 mm and a pitch of 6 mm. The frictional coefcients are 0.14 for
the threads and 0.09 for the collar, with a collar friction radius of 45 mm. Find the axial resisting load F and the combined efciency of the screw and collar.
811
A bolted joint is to have a grip consisting of two 1 -in steel plates and one wide 1 -in American
2
2
Standard plain washer to t under the head of the 1 in-13 × 1.75 in UNC hex-head bolt.
2
(a) What is the length of the thread L T for this diameter inch-series bolt?
(b) What is the length of the grip l?
(c) What is the height H of the nut?
(d ) Is the bolt long enough? If not, round to the next larger preferred length (Table A17).
(e) What is the length of the shank and threaded portions of the bolt within the grip? These
lengths are needed in order to estimate the bolt spring rate kb .
812
A bolted joint is to have a grip consisting of two 14-mm steel plates and one 14R metric plain
washer to t under the head of the M14 × 2 hex-head bolt, 50 mm long.
(a) What is the length of the thread L T for this diameter metric coarse-pitch series bolt?
(b) What is the length of the grip l?
(c) What is the height H of the nut?
(d ) Is the bolt long enough? If not, round to the next larger preferred length (Table A17).
(e) What is the length of the shank and the threaded portions of the bolt within the grip? These
lengths are needed in order to estimate bolt spring rate kb .
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
446
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
813
A blanking disk 0.875 in thick is to be fastened to a spool whose ange is 1 in thick, using
eight 1 in-13 × 1.75 in hex-head cap screws.
2
(a) What is the length of threads L T for this cap screw?
(b) What is the effective length of the grip l ?
(c) Is the length of this cap screw sufcient? If not, round up.
(d ) Find the shank length ld and the useful thread length lt within the grip. These lengths are
needed for the estimate of the fastener spring rate kb .
814
A blanking disk is 20 mm thick and is to be fastened to a spool whose ange is 25 mm thick,
using eight M12 × 40 hex-head metric cap screws.
(a) What is the length of the threads L T for this fastener?
(b) What is the effective grip length l ?
(c) Is the length of this fastener sufcient? If not, round to the next preferred length.
(d ) Find the shank length ld and the useful threaded length in the grip lt . These lengths are
needed in order to estimate the fastener spring rate kb .
815
A 3 in-16 UNF series SAE grade 5 bolt has a 3 -in ID tube 13 in long, clamped between washer
4
4
faces of bolt and nut by turning the nut snug and adding one-third of a turn. The tube OD is
the washer-face diameter dw = 1.5d = 1.5(0.75) = 1.125 in = OD.
3
4
in-16 UNF grade
1.125 in
Problem 815
13 in
(a) What is the spring rate of the bolt and the tube, if the tube is made of steel? What is the
joint constant C?
(b) When the one-third turn-of-nut is applied, what is the initial tension Fi in the bolt?
(c) What is the bolt tension at opening if additional tension is applied to the bolt external to
the joint?
816
From your experience with Prob. 815, generalize your solution to develop a turn-of-nut equation
Nt =
θ
=
360
kb + km
kb km
Fi N
where Nt = turn of the nut from snug tight
θ = turn of the nut in degrees
N = number of thread/in (1/ p where p is pitch)
Fi = initial preload
kb , km = spring rates of the bolt and members, respectively
Use this equation to nd the relation between torque-wrench setting T and turn-of-nut Nt .
(Snug tight means the joint has been tightened to perhaps half the intended preload to atten asperities on the washer faces and the members. Then the nut is loosened and retightened
449
450
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
447
nger tight, and the nut is rotated the number of degrees indicated by the equation. Properly
done, the result is competitive with torque wrenching.)
817
RB&W11 recommends turn-of-nut from snug t to preload as follows: 1/3 turn for bolt grips
of 14 diameters, 1/2 turn for bolt grips 48 diameters, and 2/3 turn for grips of 812 diameters. These recommendations are for structural steel fabrication (permanent joints), producing
preloads of 100 percent of proof strength and beyond. Machinery fabricators with fatigue loadings and possible joint disassembly have much smaller turns-of-nut. The RB&W recommendation enters the nonlinear plastic deformation zone.
Position mark
on work surface
Position
mark on nut
Problem 817
Turn-of-nut method
Position
mark on nut
Tighten nut
to snug fit
Addition turn
(a) For Ex. 84, use Eq. (827) with K = 0.2 to estimate the torque necessary to establish the
desired preload. Then, using the results from Prob. 816, determine the turn of the nut in
degrees. How does this compare with the RB&W recommendations?
(b) Repeat part (a) for Ex. 85.
818
Take Eq. (822) and express km /( Ed ) as a function of l /d , then compare with Eq. (823) for
d / l = 0.5.
819
A joint has the same geometry as Ex. 84, but the lower member is steel. Use Eq. (823) to
nd the spring rate of the members in the grip. Hint: Equation (823) applies to the stiffness
of two sections of a joint of one material. If each section has the same thickness, then what is
the stiffness of one of the sections?
820
The gure illustrates the connection of a cylinder head to a pressure vessel using 10 bolts and
a confined-gasket seal. The effective sealing diameter is 150 mm. Other dimensions are:
A = 100, B = 200, C = 300, D = 20, and E = 20, all in millimeters. The cylinder is used to
store gas at a static pressure of 6 MPa. ISO class 8.8 bolts with a diameter of 12 mm have
been selected. This provides an acceptable bolt spacing. What load factor n results from this
selection?
C
B
D
Problem 820
E
Cylinder head is steel; cylinder is
grade 30 cast iron.
A
11
Russell, Burdsall & Ward, Inc., Metal Forming Spets, Mentor, Ohio.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
448
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
821
The computer can be very helpful to the engineer. In matters of analysis it can take the drudgery
out of calculations and improve accuracy. In synthesis, good programming is a matter of organizing decisions that must be made, soliciting them while displaying enough information, accepting
them, and doing the number crunching. In either case, one cannot program what one does not
understand. Understanding comes from experience with problems executed manually. It is useful
to program the protocol of Table 87 because it is so easy to make a mistake in longhand. Focusing on the fastener, recognize two situations: (1) the fastener has been chosen, its diameter and
length are known, and the designer needs to know all the pertinent dimensions, including the
effective grip of a cap-screw joint and whether the length is adequate; and (2) the fastener diameter, nut, and washers are chosen, and the designer has to make the length decision, after which
documentation of pertinent dimensions is in order. Code the protocol of Table 87, bearing in mind
that you may wish to embed some of it in a larger program.
822
Figure P820 illustrates the connection of a cylinder head to a pressure vessel using 10 bolts and
a conned-gasket seal. The effective sealing diameter is 150 mm. Other dimensions are: A = 100,
B = 200, C = 300, D = 20, and E = 25, all in millimeters. The cylinder is used to store gas at
a static pressure of 6 MPa. ISO class 8.8 bolts with a diameter of 12 mm have been selected.
This provides an acceptable bolt spacing. What load factor n results from this selection?
823
We wish to alter the gure for Prob. 822 by decreasing the inside diameter of the seal to
the diameter A = 100 mm. This makes an effective sealing diameter of 120 mm. Then, by
using cap screws instead of bolts, the bolt circle diameter B can be reduced as well as the
outside diameter C. If the same bolt spacing and the same edge distance are used, then eight
12-mm cap screws can be used on a bolt circle with B = 160 mm and an outside diameter
of 260 mm, a substantial savings. With these dimensions and all other data the same as in
Prob. 822, nd the load factor.
824
In the gure for Prob. 820, the bolts have a diameter of 1 in and the cover plate is steel, with
2
1
D = 2 in. The cylinder is cast iron, with E = 5 in and a modulus of elasticity of 18 Mpsi.
8
The 1 -in SAE washer to be used under the nut has OD = 1.062 in and is 0.095 in thick. Find
2
the stiffnesses of the bolt and the members and the joint constant C.
825
1
The same as Prob. 824, except that 2 -in cap screws are used with washers (see Fig. 821).
826
In addition to the data of Prob. 824, the dimensions of the cylinder are A = 3.5 in and an
effective seal diameter of 4.25 in. The internal static pressure is 1500 psi. The outside diameter of the head is C = 8 in. The diameter of the bolt circle is 6 in, and so a bolt spacing in the
range of 3 to 5 bolt diameters would require from 8 to 13 bolts. Select 10 SAE grade 5 bolts
and nd the resulting load factor n.
827
A 3 -in class 5 cap screw and steel washer are used to secure a cap to a cast-iron frame of a
8
machine having a blind threaded hole. The washer is 0.065 in thick. The frame has a modulus
1
of elasticity of 14 Mpsi and is 4 in thick. The screw is 1 in long. The material in the frame also
has a modulus of elasticity of 14 Mpsi. Find the stiffnesses kb and km of the bolt and members.
828
Bolts distributed about a bolt circle are often called upon to resist an external bending moment
as shown in the gure. The external moment is 12 kip · in and the bolt circle has a diameter
of 8 in. The neutral axis for bending is a diameter of the bolt circle. What needs to be determined is the most severe external load seen by a bolt in the assembly.
(a) View the effect of the bolts as placing a line load around the bolt circle whose intensity
Fb , in pounds per inch, varies linearly with the distance from the neutral axis according to
the relation Fb = Fb ,max R sin θ . The load on any particular bolt can be viewed as the effect
of the line load over the arc associated with the bolt. For example, there are 12 bolts shown
451
452
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
449
in the gure. Thus each bolt load is assumed to be distributed on a 30° arc of the bolt circle. Under these conditions, what is the largest bolt load?
(b) View the largest load as the intensity Fb ,max multiplied by the arc length associated with
each bolt and nd the largest bolt load.
(c) Express the load on any bolt as F = Fmax sin θ , sum the moments due to all the bolts, and
estimate the largest bolt load. Compare the results of these three approaches to decide how
to attack such problems in the future.
R
Problem 828
Bolted connection subjected to
bending.
M
M
Neutral
axis
829
The gure shows a cast-iron bearing block that is to be bolted to a steel ceiling joist and is
to support a gravity load. Bolts used are M20 ISO 8.8 with coarse threads and with 3.4-mmthick steel washers under the bolt head and nut. The joist anges are 20 mm in thickness,
and the dimension A, shown in the gure, is 20 mm. The modulus of elasticity of the bearing block is 135 GPa.
A
Problem 829
B
d
C
(a) Find the wrench torque required if the fasteners are lubricated during assembly and the joint
is to be permanent.
(b) Determine the load factor for the design if the gravity load is 15 kN.
830
The upside-down steel A frame shown in the gure is to be bolted to steel beams on the ceiling of a machine room using ISO grade 8.8 bolts. This frame is to support the 40-kN radial
load as illustrated. The total bolt grip is 48 mm, which includes the thickness of the steel beam,
the A-frame feet, and the steel washers used. The bolts are size M20 × 2.5.
(a) What tightening torque should be used if the connection is permanent and the fasteners are
lubricated?
(b) What portion of the external load is taken by the bolts? By the members?
831
If the pressure in Prob. 820 is cycling between 0 and 6 MPa, determine the fatigue factor of
safety using the:
(a) Goodman criterion.
(b) Gerber criterion.
(c) ASME-elliptic criterion.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
450
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
Drill 2 holes for
M20 × 2.5 bolts
Problem 830
W = 40 kN
832
In the gure for Prob. 820, let A = 0.9 m, B = 1 m, C = 1.10 m, D = 20 mm, and E = 25 mm.
The cylinder is made of ASTM No. 35 cast iron ( E = 96 GPa), and the head, of low-carbon
steel. There are thirty-six M10 × 1.5 ISO 10.9 bolts tightened to 75 percent of proof load.
During use, the cylinder pressure uctuates between 0 and 550 kPa. Find the factor of safety
guarding against a fatigue failure of a bolt using the:
(a) Goodman criterion.
(b) Gerber criterion.
(c) ASME-elliptic criterion.
833
A 1-in-diameter hot-rolled AISI 1144 steel rod is hot-formed into an eyebolt similar to that shown
in the gure for Prob. 374, with an inner 2-in-diameter eye. The threads are 1 in-12 UNF and
are die-cut.
(a) For a repeatedly applied load collinear with the thread axis, using the Gerber criterion is
fatigue failure more likely in the thread or in the eye?
(b) What can be done to strengthen the bolt at the weaker location?
(c) If the factor of safety guarding against a fatigue failure is n f = 2, what repeatedly applied
load can be applied to the eye?
834
The section of the sealed joint shown in the gure is loaded by a repeated force P = 6 kip.
The members have E = 16 Mpsi. All bolts have been carefully preloaded to Fi = 25 kip each.
3
4
in-16 UNF
SAE grade 5
Problem 834
1
1 2 in
No. 40 CI
(a) If hardened-steel washers 0.134 in thick are to be used under the head and nut, what length
of bolts should be used?
(b) Find kb , km , and C.
(c) Using the Goodman criterion, nd the factor of safety guarding against a fatigue failure.
(d ) Using the Gerber criterion, nd the factor of safety guarding against a fatigue failure.
(e) Find the load factor guarding against overproof loading.
453
454
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
835
451
Suppose the welded steel bracket shown in the gure is bolted underneath a structural-steel
ceiling beam to support a uctuating vertical load imposed on it by a pin and yoke. The bolts
1
are 2 in coarse-thread SAE grade 5, tightened to recommended preload. The stiffnesses have
already been computed and are kb = 4.94 Mlb/in and km = 15.97 Mlb/in.
A
C
Problem 835
d
B
(a) Assuming that the bolts, rather than the welds, govern the strength of this design, determine the safe repeated load P that can be imposed on this assembly using the Goodman
criterion and a fatigue design factor of 2.
(b) Repeat part (a) using the Gerber criterion.
(c) Compute the load factors based on the load found in part (b).
836
Using the Gerber fatigue criterion and a fatigue-design factor of 2, determine the external
repeated load P that a 1 1 -in SAE grade 5 coarse-thread bolt can take compared with that for
4
a ne-thread bolt. The joint constants are C = 0.30 for coarse- and 0.32 for ne-thread bolts.
837
An M30 × 3.5 ISO 8.8 bolt is used in a joint at recommended preload, and the joint is subject to a repeated tensile fatigue load of P = 80 kN per bolt. The joint constant is C = 0.33.
Find the load factors and the factor of safety guarding against a fatigue failure based on the
Gerber fatigue criterion.
838
The gure shows a uid-pressure linear actuator (hydraulic cylinder) in which D = 4 in, t = 3
8
in, L = 12 in, and w = 3 in. Both brackets as well as the cylinder are of steel. The actuator
4
3
has been designed for a working pressure of 2000 psi. Six 8 -in SAE grade 5 coarse-thread
bolts are used, tightened to 75 percent of proof load.
w
Problem 838
t
L
w
D
(a) Find the stiffnesses of the bolts and members, assuming that the entire cylinder is compressed uniformly and that the end brackets are perfectly rigid.
(b) Using the Goodman fatigue criterion, nd the factor of safety guarding against a fatigue
failure.
(c) Repeat part (b) using the Gerber fatigue criterion.
(d ) What pressure would be required to cause total joint separation?
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
452
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
839
The gure shows a bolted lap joint that uses SAE grade 8 bolts. The members are made of
cold-drawn AISI 1040 steel. Find the safe tensile shear load F that can be applied to this connection if the following factors of safety are specied: shear of bolts 3, bearing on bolts 2,
bearing on members 2.5, and tension of members 3.
5
8
Problem 839
3
8
in
5
in
16
in-16 UNC
1 1 in
8
5
8
in
1
4
1
1 4 in
840
in
The bolted connection shown in the gure uses SAE grade 5 bolts. The members are hot-rolled
AISI 1018 steel. A tensile shear load F = 4000 lbf is applied to the connection. Find the factor of safety for all possible modes of failure.
5
8
1
5
8
in
3
8
in
5
8
5
8
1 8 in
in
1
4
in
in-16 UNC
in
Problem 840
1
4
841
in
A bolted lap joint using SAE grade 5 bolts and members made of cold-drawn SAE 1040 steel
is shown in the gure. Find the tensile shear load F that can be applied to this connection if
the following factors of safety are specied: shear of bolts 1.8, bearing on bolts 2.2, bearing
on members 2.4, and tension of members 2.6.
7
8
1
3
4
in
in-9 UNC
1 2 in
3
Problem 841
2 4 in
1
1 2 in
3 in
842
3
4
in
The bolted connection shown in the gure is subjected to a tensile shear load of 20 kip. The
bolts are SAE grade 5 and the material is cold-drawn AISI 1015 steel. Find the factor of safety
of the connection for all possible modes of failure.
455
456
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
3
3
1 8 in
Problem 842
3
3
1 8 in
2 8 in
2 8 in
453
3
4
1 3 in
8
5
8
in
in-10 UNC
1 3 in
8
3
4
843
The gure shows a connection that employs three SAE grade 5 bolts. The tensile shear load
on the joint is 5400 lbf. The members are cold-drawn bars of AISI 1020 steel. Find the factor
of safety for each possible mode of failure.
5
8
5
8
Problem 843
1
1 8 in
in
3
8
in
5
in
16
in-16 UNC
1 in
5
8
in
3
116 in
5
in
16
2 3 in
8
844
in
A beam is made up by bolting together two cold-drawn bars of AISI 1018 steel as a lap joint,
as shown in the gure. The bolts used are ISO 5.8. Ignoring any twisting, determine the factor of safety of the connection.
y
A
2.8 kN
Problem 844
200
50
100
350
10
Dimensions in millimeters.
x
50
10
A
845
M10
1.5
Section AA
Standard design practice, as exhibited by the solutions to Probs. 839 to 843, is to assume
that the bolts, or rivets, share the shear equally. For many situations, such an assumption may
lead to an unsafe design. Consider the yoke bracket of Prob. 835, for example. Suppose this
bracket is bolted to a wide-ange column with the centerline through the two bolts in the vertical direction. A vertical load through the yoke-pin hole at distance B from the column ange
would place a shear load on the bolts as well as a tensile load. The tensile load comes about
because the bracket tends to pry itself about the bottom corner, much like a claw hammer, exerting a large tensile load on the upper bolt. In addition, it is almost certain that both the spacing of the bolt holes and their diameters will be slightly different on the column ange from
what they are on the yoke bracket. Thus, unless yielding occurs, only one of the bolts will take
the shear load. The designer has no way of knowing which bolt this will be.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
454
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Mechanical Engineering Design
In this problem the bracket is 8 in long, A = 1 in, B = 3 in, C = 6 in, and the column
2
1
ange is 2 in thick. The bolts are 1 in UNC SAE 5. Steel washers 0.095 in thick are used
2
under the nuts. The nuts are tightened to 75 percent of proof load. The vertical yoke-pin load
is 3000 lbf. If the upper bolt takes all the shear load as well as the tensile load, how closely
does the bolt stress approach the proof strength?
846
The bearing of Prob. 829 is bolted to a vertical surface and supports a horizontal shaft. The
bolts used have coarse threads and are M20 ISO 5.8. The joint constant is C = 0.30, and the
dimensions are A = 20 mm, B = 50 mm, and C = 160 mm. The bearing base is 240 mm long.
The bearing load is 12 kN. If the bolts are tightened to 75 percent of proof load, will the bolt
stress exceed the proof strength? Use worst-case loading, as discussed in Prob. 845.
847
A split-ring clamp-type shaft collar such as is described in Prob. 531 must resist an axial load of
1000 lbf. Using a design factor of n = 3 and a coefcient of friction of 0.12, specify an SAE Grade
5 cap screw using ne threads. What wrench torque should be used if a lubricated screw is used?
848
A vertical channel 152 × 76 (see Table A7) has a cantilever beam bolted to it as shown. The
channel is hot-rolled AISI 1006 steel. The bar is of hot-rolled AISI 1015 steel. The shoulder
bolts are M12 × 1.75 ISO 5.8. For a design factor of 2.8, nd the safe force F that can be
applied to the cantilever.
12
F
Problem 848
Dimensions in millimeters.
A
50
849
O
50
B
50
125
Find the total shear load on each of the three bolts for the connection shown in the gure and
compute the signicant bolt shear stress and bearing stress. Find the second moment of area
of the 8-mm plate on a section through the three bolt holes, and nd the maximum bending
stress in the plate.
Holes for M12
1.75 bolts
8 mm thick
36
Problem 849
Dimensions in millimeters.
12 kN
32
64
36
200
Column
457
458
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
Screws, Fasteners, and the Design of Nonpermanent Joints
850
A 3 - × 2-in AISI 1018 cold-drawn steel bar is cantilevered to support a static load of 300 lbf
8
as illustrated. The bar is secured to the support using two 1 in-13 UNC SAE 5 bolts. Find the
2
factor of safety for the following modes of failure: shear of bolt, bearing on bolt, bearing on
member, and strength of member.
3
8
Problem 850
455
1 in
in
14 in
3 in 1 in
300 lbf
851
The gure shows a welded tting which has been tentatively designed to be bolted to a channel so as to transfer the 2500-lbf load into the channel. The channel is made of hot-rolled lowcarbon steel having a minimum yield strength of 46 kpsi; the two tting plates are of hot-rolled
stock having a minimum Sy of 45.5 kpsi. The tting is to be bolted using six SAE grade
2 shoulder bolts. Check the strength of the design by computing the factor of safety for all possible modes of failure.
6 holes for
5
8
in-11 NC bolts
F = 2500 lbf
1
4
in
4 in 1 in
Problem 851
2
5 in
1
4
in
8 in [ 11.5
8 in
3
16
in
7 1 in
2
852
A cantilever is to be attached to the at side of a 6-in, 13.0-lbf/in channel used as a column.
The cantilever is to carry a load as shown in the gure. To a designer the choice of a bolt array
is usually an a priori decision. Such decisions are made from a background of knowledge of
the effectiveness of various patterns.
1
2
Problem 852
in steel plate
6 in
6 in
6 in
2000 lbf
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
456
III. Design of Mechanical
Elements
8. Screws, Fasteners, and
the Design of
Nonpermanent Joints
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
(a) If two fasteners are used, should the array be arranged vertically, horizontally, or diagonally? How would you decide?
(b) If three fasteners are used, should a linear or triangular array be used? For a triangular array,
what should be the orientation of the triangle? How would you decide?
853
Using your experience with Prob. 852, specify a bolt pattern for Prob. 852, and size the bolts.
854
Determining the joint stiffness of nonsymmetric joints of two or more different materials using
a frustum of a hollow cone can be time-consuming and prone to error. Develop a computer
program to determine km for a joint composed of two different materials of differing thickness.
Test the program to determine km for problems such as Ex. 85 and Probs. 819, 820, 822,
824, and 827.
459
460
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
9
Welding, Bonding, and the
Design of Permanent Joints
Chapter Outline
91
Welding Symbols
92
Butt and Fillet Welds
93
Stresses in Welded Joints in Torsion
94
Stresses in Welded Joints in Bending
95
The Strength of Welded Joints
96
Static Loading
97
Fatigue Loading
98
Resistance Welding
99
Adhesive Bonding
458
460
464
469
471
474
478
480
480
457
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
458
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Mechanical Engineering Design
Form can more readily pursue function with the help of joining processes such as welding, brazing, soldering, cementing, and gluingprocesses that are used extensively in
manufacturing today. Whenever parts have to be assembled or fabricated, there is usually good cause for considering one of these processes in preliminary design work.
Particularly when sections to be joined are thin, one of these methods may lead to signicant savings. The elimination of individual fasteners, with their holes and assembly
costs, is an important factor. Also, some of the methods allow rapid machine assembly,
furthering their attractiveness.
Riveted permanent joints were common as the means of fastening rolled steel
shapes to one another to form a permanent joint. The childhood fascination of seeing a
cherry-red hot rivet thrown with tongs across a building skeleton to be unerringly
caught by a person with a conical bucket, to be hammered pneumatically into its nal
shape, is all but gone. Two developments relegated riveting to lesser prominence.
The rst was the development of high-strength steel bolts whose preload could be
controlled. The second was the improvement of welding, competing both in cost and in
latitude of possible form.
91
Welding Symbols
A weldment is fabricated by welding together a collection of metal shapes, cut to particular congurations. During welding, the several parts are held securely together,
often by clamping or jigging. The welds must be precisely specied on working
drawings, and this is done by using the welding symbol, shown in Fig. 91, as standardized by the American Welding Society (AWS). The arrow of this symbol points to
the joint to be welded. The body of the symbol contains as many of the following elements as are deemed necessary:
Reference line
Arrow
Groove angle; included
angle of countersink
for plug welds
Length of weld
Size; size or strength
for resistance welds
Pitch (center-to-center
spacing) of welds
F
A
R
Arrow connecting reference
line to arrow side of joint,
to grooved member, or both
Other
side
Reference line
sides)
S
LP
T
Specification; process;
or other reference
Tail (may be omitted
when reference
is not used)
Basic weld symbol
or detail reference
Arrow
side
The AWS standard welding
symbol showing the location
of the symbol elements.
Finish symbol
Contour symbol
Root opening; depth of filling
for plug and slot welds
(Both
Figure 91
(N)
Field weld symbol
Weld all around symbol
Number of spot or
projection welds
461
462
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Welding, Bonding, and the Design of Permanent Joints
459
Basic weld symbols as in Fig. 92
Dimensions and other data
Supplementary symbols
Finish symbols
Tail
Specication or process
The arrow side of a joint is the line, side, area, or near member to which the arrow
points. The side opposite the arrow side is the other side.
Figures 93 to 96 illustrate the types of welds used most frequently by designers.
For general machine elements most welds are llet welds, though butt welds are used a
great deal in designing pressure vessels. Of course, the parts to be joined must be
arranged so that there is sufcient clearance for the welding operation. If unusual joints
are required because of insufcient clearance or because of the section shape, the
design may be a poor one and the designer should begin again and endeavor to synthesize another solution.
Since heat is used in the welding operation, there are metallurgical changes in the
parent metal in the vicinity of the weld. Also, residual stresses may be introduced because
of clamping or holding or, sometimes, because of the order of welding. Usually these
Figure 92
Arc- and gas-weld symbols.
Type of weld
Bead
Fillet
Figure 93
Fillet welds. (a) The number
indicates the leg size; the
arrow should point only to one
weld when both sides are the
same. (b) The symbol indicates
that the welds are intermittent
and staggered 60 mm along
on 200-mm centers.
Plug
or
slot
Groove
Square
V
Bevel
60
5
(b)
Figure 94
The circle on the weld symbol
indicates that the welding is to
go all around.
5
J
200
60200
(a)
U
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
460
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Mechanical Engineering Design
Figure 95
60°
Butt or groove welds:
(a) square butt-welded on both
sides; (b) single V with 60°
bevel and root opening of
2 mm; (c) double V; (d ) single
bevel.
2
2
60°
(a)
(b)
60°
45°
(d )
(c)
Figure 96
Special groove welds: (a) T
joint for thick plates; (b) U and
J welds for thick plates;
(c) corner weld (may also
have a bead weld on inside
for greater strength but should
not be used for heavy loads);
(d) edge weld for sheet metal
and light loads.
(a)
(b)
(c)
(d )
residual stresses are not severe enough to cause concern; in some cases a light heat treatment after welding has been found helpful in relieving them. When the parts to be welded
are thick, a preheating will also be of benet. If the reliability of the component is to be
quite high, a testing program should be established to learn what changes or additions to
the operations are necessary to ensure the best quality.
92
Butt and Fillet Welds
Figure 97a shows a single V-groove weld loaded by the tensile force F. For either
tension or compression loading, the average normal stress is
F
σ=
(91)
hl
where h is the weld throat and l is the length of the weld, as shown in the gure. Note that
the value of h does not include the reinforcement. The reinforcement can be desirable,
but it varies somewhat and does produce stress concentration at point A in the gure. If
fatigue loads exist, it is good practice to grind or machine off the reinforcement.
463
464
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Welding, Bonding, and the Design of Permanent Joints
Figure 97
Reinforcement
461
Reinforcement
A
A typical butt joint.
l
l
F
F
F
F
Throat h
Throat h
(a) Tensile loading
Figure 98
(b) Shear loading
Throat
A transverse llet weld.
D
A
h
C
h
F
B
2F
h
F
Figure 99
x
Free body from Fig. 98.
t
h
Fs
Fn
F
90
y
The average stress in a butt weld due to shear loading (Fig. 97b) is
τ=
F
hl
(92)
Figure 98 illustrates a typical transverse llet weld. In Fig. 99 a portion of the
welded joint has been isolated from Fig. 98 as a free body. At angle θ the forces on
each weldment consist of a normal force Fn and a shear force Fs . Summing forces in
the x and y directions gives
Fs = F sin θ
(a)
Fn = F cos θ
(b)
Using the law of sines for the triangle in Fig. 99 yields
2h
h
t
h
=
=
=
θ + 45 )
θ)
sin 45
sin(90
sin(135
cos θ + sin θ
Solving for the throat length t gives
t=
h
cos θ + sin θ
(c)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
462
III. Design of Mechanical
Elements
9. Welding, Bonding, and
the Design of Permanent
Joints
465
© The McGrawHill
Companies, 2008
Mechanical Engineering Design
The nominal stresses at the angle θ in the weldment, τ and σ , are
τ=
Fs
F sin θ(cos θ + sin θ)
F
=
= (sin θ cos θ + sin2 θ)
A
hl
hl
(d )
σ=
Fn
F cos θ(cos θ + sin θ)
F
=
= (cos2 θ + sin θ cos θ)
A
hl
hl
(e)
The von Mises stress σ at angle θ is
σ = (σ 2 + 3τ 2 )1/2 =
F
[(cos2 θ + sin θ cos θ)2 + 3(sin2 θ + sin θ cos θ)2 ]1/2
hl
(f )
The largest von Mises stress occurs at θ = 62.5 with a value of σ = 2.16 F /(hl ) . The
corresponding values of τ and σ are τ = 1.196 F /(hl ) and σ = 0.623 F /(hl ) .
The maximum shear stress can be found by differentiating Eq. ( d ) with respect to
θ and equating to zero. The stationary point occurs at θ = 67.5 with a corresponding
τmax = 1.207 F /(hl ) and σ = 0.5 F /(hl ) .
There are some experimental and analytical results that are helpful in evaluating Eqs.
( d ) through ( f ) and consequences. A model of the transverse llet weld of Fig. 98 is
easily constructed for photoelastic purposes and has the advantage of a balanced loading
condition. Norris constructed such a model and reported the stress distribution along the
sides AB and BC of the weld.1 An approximate graph of the results he obtained is shown
as Fig. 910a. Note that stress concentration exists at A and B on the horizontal leg and at
B on the vertical leg. Norris states that he could not determine the stresses at A and B with
any certainty.
Salakian2 presents data for the stress distribution across the throat of a llet weld
(Fig. 910b). This graph is of particular interest because we have just learned that it is
the throat stresses that are used in design. Again, the gure shows stress concentration
at point B. Note that Fig. 910a applies either to the weld metal or to the parent metal,
and that Fig. 910b applies only to the weld metal.
Equations (a) through ( f ) and their consequences seem familiar, and we can become
comfortable with them. The net result of photoelastic and nite element analysis of transverse llet weld geometry is more like that shown in Fig. 910 than those given by
mechanics of materials or elasticity methods. The most important concept here is that we
have no analytical approach that predicts the existing stresses. The geometry of the llet
is crude by machinery standards, and even if it were ideal, the macrogeometry is too abrupt
and complex for our methods. There are also subtle bending stresses due to eccentricities.
Still, in the absence of robust analysis, weldments must be specied and the resulting joints
must be safe. The approach has been to use a simple and conservative model, veried by
testing as conservative. The approach has been to
Consider the external loading to be carried by shear forces on the throat area of the
weld. By ignoring the normal stress on the throat, the shearing stresses are inated
sufciently to render the model conservative.
1
C. H. Norris, Photoelastic Investigation of Stress Distribution in Transverse Fillet Welds, Welding J.,
vol. 24, 1945, p. 557s.
2
A. G. Salakian and G. E. Claussen, Stress Distribution in Fillet Welds: A Review of the Literature,
Welding J., vol. 16, May 1937, pp. 124.
466
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Welding, Bonding, and the Design of Permanent Joints
Figure 910
463
C
Stress distribution in llet
welds: (a) stress distribution on
the legs as reported by Norris;
(b) distribution of principal
stresses and maximum shear
stress as reported by Salakian.
+
1
D
x
ma
+
+
A
0
D
B
B
2
(a)
(b)
Figure 911
Parallel llet welds.
l
F
h
2F
F
Use distortion energy for signicant stresses.
Circumscribe typical cases by code.
For this model, the basis for weld analysis or design employs
τ=
F
1.414 F
=
0.707hl
hl
(93)
which assumes the entire force F is accounted for by a shear stress in the minimum
throat area. Note that this inates the maximum estimated shear stress by a factor of
1.414/1.207 = 1.17. Further, consider the parallel llet welds shown in Fig. 911
where, as in Fig. 98, each weld transmits a force F. However, in the case of
Fig. 911, the maximum shear stress is at the minimum throat area and corresponds to
Eq. (93).
Under circumstances of combined loading we
Examine primary shear stresses due to external forces.
Examine secondary shear stresses due to torsional and bending moments.
Estimate the strength(s) of the parent metal(s).
Estimate the strength of deposited weld metal.
Estimate permissible load(s) for parent metal(s).
Estimate permissible load for deposited weld metal.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
464
III. Design of Mechanical
Elements
467
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Mechanical Engineering Design
93
Stresses in Welded Joints in Torsion
Figure 912 illustrates a cantilever of length l welded to a column by two llet welds.
The reaction at the support of a cantilever always consists of a shear force V and a
moment M. The shear force produces a primary shear in the welds of magnitude
τ =
V
A
(94)
where A is the throat area of all the welds.
The moment at the support produces secondary shear or torsion of the welds, and
this stress is given by the equation
τ =
Mr
J
(95)
where r is the distance from the centroid of the weld group to the point in the weld of
interest and J is the second polar moment of area of the weld group about the centroid
of the group. When the sizes of the welds are known, these equations can be solved and
the results combined to obtain the maximum shear stress. Note that r is usually the
farthest distance from the centroid of the weld group.
Figure 913 shows two welds in a group. The rectangles represent the throat areas
of the welds. Weld 1 has a throat width b1 = 0.707h 1 , and weld 2 has a throat width
d2 = 0.707h 2 . Note that h 1 and h 2 are the respective weld sizes. The throat area of both
welds together is
A = A1 + A2 = b1 d1 + b2 d2
(a)
This is the area that is to be used in Eq. (94).
The x axis in Fig. 913 passes through the centroid G 1 of weld 1. The second
moment of area about this axis is
Ix =
3
b1 d1
12
Similarly, the second moment of area about an axis through G 1 parallel to the y axis is
Iy =
3
d1 b1
12
Figure 912
This is a moment connection;
such a connection produces
torsion in the welds.
F
O
r
ro
O
l
468
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Welding, Bonding, and the Design of Permanent Joints
Figure 913
465
y
x2
b2
2
G2
1
G
r2
x
G1
O
x1
y2
y
r1
d1
d2
M
b1
x
Thus the second polar moment of area of weld 1 about its own centroid is
JG 1 = Ix + I y =
3
3
b1 d1
d1 b1
+
12
12
(b)
In a similar manner, the second polar moment of area of weld 2 about its centroid is
JG 2 =
3
3
b2 d2
d2 b2
+
12
12
(c)
The centroid G of the weld group is located at
x=
¯
A1 x 1 + A2 x 2
A
y=
¯
A1 y1 + A2 y2
A
Using Fig. 913 again, we see that the distances r1 and r2 from G 1 and G 2 to G,
respectively, are
r1 = [(x x1 )2 + y 2 ]1/2
¯
¯
r2 = [( y2 y )2 + (x2 x )2 ]1/2
¯
¯
Now, using the parallel-axis theorem, we nd the second polar moment of area of the
weld group to be
2
2
J = JG 1 + A1r1 + JG 2 + A2r2
(d )
This is the quantity to be used in Eq. (95). The distance r must be measured from G
and the moment M computed about G.
The reverse procedure is that in which the allowable shear stress is given and we
wish to nd the weld size. The usual procedure is to estimate a probable weld size and
then to use iteration.
3
3
Observe in Eqs. (b) and (c) the quantities b1 and d2 , respectively, which are the
cubes of the weld widths. These quantities are small and can be neglected. This leaves
3
3
the terms b1 d1 /12 and d2 b2 /12, which make JG 1 and JG 2 linear in the weld width.
Setting the weld widths b1 and d2 to unity leads to the idea of treating each llet weld
as a line. The resulting second moment of area is then a unit second polar moment of
area. The advantage of treating the weld size as a line is that the value of Ju is the same
regardless of the weld size. Since the throat width of a llet weld is 0.707h , the relationship between J and the unit value is
J = 0.707h Ju
(96)
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
466
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Mechanical Engineering Design
in which Ju is found by conventional methods for an area having unit width. The transfer formula for Ju must be employed when the welds occur in groups, as in Fig. 912.
Table 91 lists the throat areas and the unit second polar moments of area for the most
common llet welds encountered. The example that follows is typical of the calculations normally made.
Table 91
Torsional Properties of Fillet Welds*
Weld
Throat Area
Ju
d 3 /12
Ju =
d (3b2 + d 2 )
6
b2
2(b + d )
Ju =
d2
2(b + d )
(b + d )4 6b 2 d 2
12(b + d )
¯
y=
A
G
Unit Second Polar
Moment of Area
Location of G
¯
x=
b2
2b + d
Ju =
8b3 + 6bd 2 + d 3
b4
12
2b + d
Ju =
(b + d )3
6
Ju
2π r3
¯
x
0.70 hd
0
¯
y = d /2
d
y
b
A
¯
x = b/2
1.41 hd
¯
y = d /2
d
G
y
x
b
A
0.707h(2b
d)
d
G
y
¯
x=
x
b
A
0.707h(2b
d)
¯
y = d /2
d
G
y
x
b
A
G
1.414h(b
d)
¯
x = b/2
¯
y = d /2
d
y
x
A
r
1.414 π hr
G
*G is centroid of weld group; h is weld size; plane of torque couple is in the plane of the paper; all welds are of unit width.
469
470
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Welding, Bonding, and the Design of Permanent Joints
467
EXAMPLE 91
A 50-kN load is transferred from a welded tting into a 200-mm steel channel as illustrated in Fig. 914. Estimate the maximum stress in the weld.
Solution3
(a) Label the ends and corners of each weld by letter. Sometimes it is desirable to label
each weld of a set by number. See Fig. 915.
(b) Estimate the primary shear stress τ . As shown in Fig. 914, each plate is welded to
the channel by means of three 6-mm llet welds. Figure 915 shows that we have
divided the load in half and are considering only a single plate. From case 4 of
Table 91 we nd the throat area as
A = 0.707(6)[2(56) + 190] = 1280 mm2
Then the primary shear stress is
V
25(10)3
=
= 19.5 MPa
A
1280
τ =
(c) Draw the τ stress, to scale, at each lettered corner or end. See Fig. 916.
(d ) Locate the centroid of the weld pattern. Using case 4 of Table 91, we nd
x=
¯
(56)2
= 10.4 mm
2(56) + 190
This is shown as point O on Figs. 915 and 916.
Figure 914
6
200
6
Dimensions in millimeters.
50 kN
6
100
6
56
200-mm
190
6
Figure 915
Diagram showing the weld
geometry; all dimensions in
millimeters. Note that V and M
represent loads applied by the
welds to the plate.
25 kN
100
110.4
C
D
V
56
y
O
45.6
M
B
A
95
x
3
We are indebted to Professor George Piotrowski of the University of Florida for the detailed steps,
presented here, of his method of weld analysis R.G.B, J.K.N.
BudynasNisbett: Shigleys
Mechanical Engineering
Design, Eighth Edition
468
III. Design of Mechanical
Elements
© The McGrawHill
Companies, 2008
9. Welding, Bonding, and
the Design of Permanent
Joints
Mechanical Engineering Design
Figure 916
F
D
D
Free-body diagram of one of
the side plates.
C
A
D
C
rC
D
A
rD
O
rA
rB
B
A
B
A
C
C
B
B
(e) Find the distances ri (see Fig. 916):
r A = r B = [(190/2)2 + (56 10.4)2 ]1/2 = 105 mm
rC = r D = [(190/2)2 + (10.4)2 ]1/2 = 95.6 mm
These distances can also be scaled from the drawing.
( f ) Find J. Using case 4 of Table 91 again, we get
J = 0.707(6)
(g) Find M:
8(56)3 + 6(56)(190)2 + (190)3
(56)4
12
2(56) + 190
= 7.07(10)6 mm4
M = Fl = 25(100 + 10.4) = 2760 N · m
(h) Estimate the secondary shear stresses τ at each lettered end or corner:
τ A = τB =
Mr
2760(10)3 (105)
= 41.0 MPa
=
J
7.07(10)6
τC = τ D =
2760(10)3 (95.6)
= 37.3 MPa
7.07(10)6
(i) Draw the τ stress, to scale, at each corner and end. See Fig. 916. Note that this is a freebody diagram of one of the side plates, and therefore the τ and τ stresses represent what
the channel is doing to the plate (through the welds) to hold the plate in equilibrium.
( j) At each letter, combine the two stress components as vectors. This gives
τ A = τ B = 37 MPa
τC = τ D = 44 MPa
(k) Identify the most highly stressed point:
Answer
τmax = τC