Machine Design: An Integrated Approach [6 ed.] 0135184231, 9780135184233

For courses in Machine Design.   An integrated, case-based approach to machine design Machine Design: An Integrated Appr

2,653 438 37MB

English Pages 1120 [1122] Year 2019

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Machine Design: An Integrated Approach [6 ed.]
 0135184231, 9780135184233

Table of contents :
Frontmatter
Preface
Video Contents
Part I. Fundamentals
Chapter 1. Introduction to Design
1.1 Design
Machine Design
1.2 A Design Process
1.3 Problem Formulation and Calculation
Definition Stage
Preliminary Design Stage
Detailed Design Stage
Documentation Stage
1.4 The Engineering Model
Estimation and First-Order Analysis
The Engineering Sketch
1.5 Computer-Aided Design and Engineering
Computer-Aided Design (CAD)
Computer-Aided Engineering (CAE)
Computational Accuracy
1.6 The Engineering Report
1.7 Factors of Safety and Design Codes
Factor of Safety
Choosing a Safety Factor
Design and Safety Codes
1.8 Statistical Considerations
1.9 Units
1.10 Summary
1.11 References
1.12 Web References
1.13 Bibliography
1.14 Problems
Chapter 2. Materials and Processes
2.0 Introduction
2.1 Material-Property Definitions
The Tensile Test
Ductility and Brittleness
The Compression Test
The Bending Test
The Torsion Test
Fatigue Strength and Endurance Limit
Impact Resistance
Fracture Toughness
Creep and Temperature Effects
2.2 The Statistical Nature of Material Properties
2.3 Homogeneity and Isotropy
2.4 Hardness
Heat Treatment
Surface (Case) Hardening
Heat Treating Nonferrous Materials
Mechanical Forming and Hardening
2.5 Coatings and Surface Treatments
Galvanic Action
Electroplating
Electroless Plating
Anodizing
Plasma-Sprayed Coatings
Chemical Coatings
2.6 General Properties of Metals
Cast Iron
Cast Steels
Wrought Steels
Steel Numbering Systems
Aluminum
Titanium
Magnesium
Copper Alloys
2.7 General Properties of Nonmetals
Polymers
Ceramics
Composites
2.8 Selecting Materials
2.9 Summary
2.10 References
2.11 Web References
2.12 Bibliography
2.13 Problems
Chapter 3. Kinematics and Load Determination
3.0 Introduction
3.1 Degree of Freedom
3.2 Mechanisms
3.3 Calculating Degree of Freedom (Mobility)
3.4 Common 1-DOF Mechanisms
Fourbar Linkage and the Grashof Condition
Sixbar Linkage
Cam and Follower
3.5 Analyzing Linkage Motion
Types of Motion
Complex Numbers as Vectors
The Vector Loop Equation
3.6 Analyzing the Fourbar Linkage
Solving for Position in the Fourbar Linkage
Solving for Velocity in the Fourbar Linkage
Angular Velocity Ratio and Mechanical Advantage
Solving for Acceleration in the Fourbar Linkage
3.7 Analyzing the Fourbar Crank-Slider
Solving for Position in the Fourbar Crank-Slider
Solving for Velocity in the Fourbar Crank-Slider
Solving for Acceleration in the Fourbar Crank-Slider
Other Linkages
3.8 Cam Design and Analysis
The Timing Diagram
The svaj Diagram
Polynomials for the Double-Dwell Case
Polynomials for the Single-Dwell Case
Pressure Angle
Radius of Curvature
3.9 Loading Classes For Force Analysis
3.10 Free-body Diagrams
3.11 Load Analysis
Three-Dimensional Analysis
Two-Dimensional Analysis
Static Load Analysis
3.12 Two-Dimensional, Static Loading Case Studies
3.13 Three-Dimensional, Static Loading Case Study
3.14 Dynamic Loading Case Study
3.15 Vibration Loading
Natural Frequency
Dynamic Forces
3.16 Impact Loading
Energy Method
3.17 Beam Loading
Shear and Moment
Singularity Functions
Superposition
3.18 Summary
3.19 References
3.20 Web References
3.21 Bibliography
3.22 Problems
Chapter 4. Stress, Strain, and Deflection
4.0 Introduction
4.1 Stress
4.2 Strain
4.3 Principal Stresses
4.4 Plane Stress and Plane Strain
Plane Stress
Plane Strain
4.5 Mohr’s Circles
4.6 Applied Versus Principal Stresses
4.7 Axial Tension
4.8 Direct Shear Stress, Bearing Stress, and Tearout
Direct Shear
Direct Bearing
Tearout Failure
4.9 Beams and Bending Stresses
Beams in Pure Bending
Shear Due to Transverse Loading
4.10 Deflection in Beams
Deflection by Singularity Functions
Statically Indeterminate Beams
4.11 Castigliano’s Method
Deflection by Castigliano’s Method
Finding Redundant Reactions with Castigliano’s Method
4.12 Torsion
4.13 Combined Stresses
4.14 Spring Rates
4.15 Stress Concentration
Stress Concentration Under Static Loading
Stress Concentration Under Dynamic Loading
Determining Geometric Stress-Concentration Factors
Designing to Avoid Stress Concentrations
4.16 Axial Compression - Columns
Slenderness Ratio
Short Columns
Long Columns
End Conditions
Intermediate Columns
4.17 Stresses in Cylinders
Thick-Walled Cylinders
Thin-Walled Cylinders
4.18 Case Studies in Static Stress and Deflection Analysis
4.19 Summary
4.20 References
4.21 Bibliography
4.22 Problems
Chapter 5. Static Failure Theories
5.0 Introduction
5.1 Failure of Ductile Materials Under Static Loading
The von Mises-Hencky or Distortion-Energy Theory
The Maximum Shear-Stress Theory
The Maximum Normal-Stress Theory
Comparison of Experimental Data with Failure Theories
5.2 Failure of Brittle Materials Under Static Loading
Even and Uneven Materials
The Coulomb-Mohr Theory
The Modified-Mohr Theory
5.3 Fracture Mechanics
Fracture-Mechanics Theory
Fracture Toughness Kc
5.4 Using The Static Loading Failure Theories
5.5 Case Studies in Static Failure Analysis
5.6 Summary
5.7 References
5.8 Bibliography
5.9 Problems
Chapter 6. Fatigue Failure Theories
6.0 Introduction
History of Fatigue Failure
6.1 Mechanism of Fatigue Failure
Crack Initiation Stage
Crack Propagation Stage
Fracture
6.2 Fatigue-Failure Models
Fatigue Regimes
The Stress-Life Approach
The Strain-Life Approach
The LEFM Approach
6.3 Machine-Design Considerations
6.4 Fatigue Loads
Rotating Machinery Loading
Service Equipment Loading
6.5 Measuring Fatigue Failure Criteria
Fully Reversed Stresses
Combined Mean and Alternating Stress
Fracture-Mechanics Criteria
Testing Actual Assemblies
6.6 Estimating Fatigue Failure Criteria
Estimating the Theoretical Fatigue Strength S or Endurance Limit S
Correction Factors—Theoretical Fatigue Strength or Endurance Limit
Corrected Fatigue Strength S or Corrected Endurance Limit S
Creating Estimated S-N Diagrams
6.7 Notches and Stress Concentrations
Notch Sensitivity
6.8 Residual Stresses
6.9 Designing for High-Cycle Fatigue
6.10 Designing for Fully Reversed Uniaxial Stresses
Design Steps for Fully Reversed Stresses with Uniaxial Loading
6.11 Designing for Fluctuating Uniaxial Stresses
Creating the Modified-Goodman Diagram
Applying Stress-Concentration Effects with Fluctuating Stresses
Determining the Safety Factor with Fluctuating Stresses
Design Steps for Fluctuating Stresses
6.12 Designing for Multiaxial Stresses in Fatigue
Frequency and Phase Relationships
Fully Reversed Simple Multiaxial Stresses
Fluctuating Simple Multiaxial Stresses
Complex Multiaxial Stresses
6.13 A General Approach to High-Cycle Fatigue Design
6.14 A Case Study in Fatigue Design
6.15 Summary
6.16 References
6.17 Bibliography
6.18 Problems
7.0 Introduction
Chapter 7. Surface Failure
7.1 Surface Geometry
7.2 Mating Surfaces
7.3 Friction
Effect of Roughness on Friction
Effect of Velocity on Friction
Rolling Friction
Effect of Lubricant on Friction
7.4 Adhesive Wear
The Adhesive-Wear Coefficient
7.5 Abrasive Wear
Abrasive Materials
Abrasion-Resistant Materials
7.6 Corrosion Wear
Corrosion Fatigue
Fretting Corrosion
7.7 Surface Fatigue
7.8 Spherical Contact
Contact Pressure and Contact Patch in Spherical Contact
Static Stress Distributions in Spherical Contact
7.9 Cylindrical Contact
Contact Pressure and Contact Patch in Parallel Cylindrical Contact
Static Stress Distributions in Parallel Cylindrical Contact
7.10 General Contact
Contact Pressure and Contact Patch in General Contact
Stress Distributions in General Contact
7.11 Dynamic Contact Stresses
Effect of a Sliding Component on Contact Stresses
7.12 Surface Fatigue Failure Models—Dynamic Contact
7.13 Surface Fatigue Strength
7.14 Summary
7.15 References
7.16 Problems
Chapter 8. Finite element Analysis
8.0 Introduction
Stress and Strain Computation
8.1 Finite Element Method
8.2 Element Types
Element Dimension and Degree of Freedom (DOF)
Element Order
H-Elements Versus P-Elements
Element Aspect Ratio
8.3 Meshing
Mesh Density
Mesh Refinement
Convergence
8.4 Boundary Conditions
8.5 Applying Loads
8.6 Testing the Model (Verification)
8.7 Modal Analysis
8.8 Case Studies
8.9 Summary
8.10 References
8.11 Bibliography
8.12 Web Resources
8.13 Problems
Part II. Machine Design
Chapter 9. Design Case Studies
9.0 Introduction
9.1 Case Study 8—A Portable Air Compressor
9.2 Case Study 9—A Hay-Bale Lifter
9.3 Case Study 10—A Cam-Testing Machine
9.4 Summary
9.5 References
9.6 Design Projects
Chapter 10. Shafts, Keys, and Couplings
10.0 Introduction
10.1 Shaft Loads
10.2 Attachments and Stress Concentrations
10.3 Shaft Materials
10.4 Shaft Power
10.5 Shaft Loads
10.6 Shaft Stresses
10.7 Shaft Failure in Combined Loading
10.8 Shaft Design
General Considerations
Design for Fully Reversed Bending and Steady Torsion
Design for Fluctuating Bending and Fluctuating Torsion
10.9 Shaft Deflection
Shafts as Beams
Shafts as Torsion Bars
10.10 Keys and Keyways
Parallel Keys
Tapered Keys
Woodruff Keys
Stresses in Keys
Key Materials
Key Design
Stress Concentrations in Keyways
10.11 Splines
10.12 Interference Fits
Stresses in Interference Fits
Stress Concentration in Interference Fits
Fretting Corrosion
10.13 Flywheel Design
Energy Variation in a Rotating System
Determining the Flywheel Inertia
Stresses in Flywheels
Failure Criteria
10.14 Critical Speeds of Shafts
Lateral Vibration of Shafts and Beams—Rayleigh’s Method
Shaft Whirl
Torsional Vibration
Two Disks on a Common Shaft
Multiple Disks on a Common Shaft
Controlling Torsional Vibrations
10.15 Couplings
Rigid Couplings
Compliant Couplings
10.16 Case Study 8B
Designing Driveshafts for a Portable Air Compressor
10.17 Summary
10.18 References
10.19 Problems
Chapter 11. Bearings and Lubrication
11.0 Introduction
A Caveat
11.1 Lubricants
11.2 Viscosity
11.3 Types of Lubrication
Full-Film Lubrication
Boundary Lubrication
11.4 Material Combinations in Sliding Bearings
11.5 Hydrodynamic Lubrication Theory
Petroff’s Equation for No-Load Torque
Reynolds’ Equation for Eccentric Journal Bearings
Torque and Power Losses in Journal Bearings
11.6 Design of Hydrodynamic Bearings
Design Load Factor—The Ocvirk Number
Design Procedures
11.7 Nonconforming Contacts
11.8 Rolling-element bearings
Comparison of Rolling and Sliding Bearings
Types of Rolling-Element Bearings
11.9 Failure of Rolling-element bearings
11.10 Selection of Rolling-element bearings
Basic Dynamic Load Rating C
Modified Bearing Life Rating
Basic Static Load Rating C0
Combined Radial and Thrust Loads
Calculation Procedures
11.11 Bearing Mounting Details
11.12 Special Bearings
11.13 Case Study 10B
11.14 Summary
Important Equations Used in This Chapter
11.15 References
11.16 Problems
Chapter 12. Spur Gears
12.0 Introduction
12.1 Gear Tooth Theory
The Fundamental Law of Gearing
The Involute Tooth Form
Pressure Angle
Gear Mesh Geometry
Rack and Pinion
Changing Center Distance
Backlash
Relative Tooth Motion
12.2 Gear Tooth Nomenclature
12.3 Interference and Undercutting
Unequal-Addendum Tooth Forms
12.4 Contact Ratio
12.5 Gear Trains
Simple Gear Trains
Compound Gear Trains
Reverted Compound Trains
Epicyclic or Planetary Gear Trains
12.6 Gear Manufacturing
Forming Gear Teeth
Machining
Roughing Processes
Finishing Processes
Gear Quality
12.7 Loading on Spur Gears
12.8 Stresses in Spur Gears
Bending Stresses
Surface Stresses
12.9 Gear Materials
Material Strengths
Bending-Fatigue Strengths for Gear Materials
Surface-Fatigue Strengths for Gear Materials
12.10 Lubrication of Gearing
12.11 Design of Spur Gears
12.12 Case Study 8C
12.13 Summary
12.14 References
12.15 Problems
Chapter 13. Helical, Bevel, and Worm Gears
13.0 Introduction
13.1 Helical Gears
Helical Gear Geometry
Helical-Gear Forces
Virtual Number of Teeth
Contact Ratios
Stresses in Helical Gears
13.2 Bevel Gears
Bevel-Gear Geometry and Nomenclature
Bevel-Gear Mounting
Forces on Bevel Gears
Stresses in Bevel Gears
13.3 Wormsets
Materials for Wormsets
Lubrication in Wormsets
Forces in Wormsets
Wormset Geometry
Rating Methods
A Design Procedure for Wormsets
13.4 Case Study 9B
13.5 Summary
13.6 References
13.7 Problems
Chapter 14. Spring Design
14.0 Introduction
14.1 Spring Rate
14.2 Spring Configurations
14.3 Spring Materials
Spring Wire
Flat Spring Stock
14.4 Helical Compression Springs
Spring Lengths
End Details
Active Coils
Spring Index
Spring Deflection
Spring Rate
Stresses in Helical Compression Spring Coils
Helical Coil Springs of Nonround Wire
Residual Stresses
Buckling of Compression Springs
Compression-Spring Surge
Allowable Strengths for Compression Springs
The Torsional-Shear S-N Diagram for Spring Wire
The Modified-Goodman Diagram for Spring Wire
14.5 Designing Helical Compression Springs for Static Loading
14.6 Designing Helical Compression Springs for Fatigue Loading
14.7 Helical Extension Springs
Active Coils in Extension Springs
Spring Rate of Extension Springs
Spring Index of Extension Springs
Coil Preload in Extension Springs
Deflection of Extension Springs
Coil Stresses in Extension Springs
End Stresses in Extension Springs
Surging in Extension Springs
Material Strengths for Extension Springs
Design of Helical Extension Springs
14.8 Helical Torsion Springs
Terminology for Torsion Springs
Number of Coils in Torsion Springs
Deflection of Torsion Springs
Spring Rate of Torsion Springs
Coil Closure
Coil Stresses in Torsion Springs
Material Parameters for Torsion Springs
Safety Factors for Torsion Springs
Designing Helical Torsion Springs
14.9 Belleville Spring Washers
Load-Deflection Function for Belleville Washers
Stresses in Belleville Washers
Static Loading of Belleville Washers
Dynamic Loading
Stacking Springs
Designing Belleville Springs
14.10 Case Study 10C
14.11 Summary
14.12 References
14.13 Problems
Chapter 15. Screws and Fasteners
15.0 Introduction
15.1 Standard Thread Forms
Tensile Stress Area
Standard Thread Dimensions
15.2 Power Screws
Square, Acme, and Buttress Threads
Power Screw Application
Power Screw Force and Torque Analysis
Friction Coefficients
Self-Locking and Back-Driving of Power Screws
Screw Efficiency
Ball Screws
15.3 Stresses in Threads
Axial Stress
Shear Stress
Torsional Stress
15.4 Types of Screw Fasteners
Classification by Intended Use
Classification by Thread Type
Classification by Head Style
Nuts and Washers
15.5 Manufacturing Fasteners
15.6 Strengths of Standard Bolts and Machine Screws
15.7 Preloaded Fasteners in Tension
Preloaded Bolts Under Static Loading
Preloaded Bolts Under Dynamic Loading
15.8 Determining the Joint Stiffness Factor
Joints With Two Plates of the Same Material
Joints With Two Plates of Different Materials
Gasketed Joints
15.9 Controlling Preload
The Turn-of-the-Nut Method
Torque-Limited Fasteners
Load-Indicating Washers
Torsional Stress Due to Torquing of Bolts
15.10 Fasteners in Shear
Dowel Pins
Centroids of Fastener Groups
Determining Shear Loads on Fasteners
15.11 Case Study 8D
15.12 Summary
15.13 References
15.14 Bibliography
15.15 Problems
Chapter 16. Weldments
16.0 Introduction
16.1 Welding Processes
Types of Welding in Common Use
Why Should a Designer Be Concerned with the Welding Process?
16.2 Weld Joints and Weld Types
Joint Preparation
Weld Specification
16.3 Principles of Weldment Design
16.4 Static Loading of Welds
16.5 Static Strength of Welds
Residual Stresses in Welds
Direction of Loading
Allowable Shear Stress for Statically Loaded Fillet and PJP Welds
16.6 Dynamic Loading of Welds
Effect of Mean Stress on Weldment Fatigue Strength
Are Correction Factors Needed For Weldment Fatigue Strength?
Effect of Weldment Configuration on Fatigue Strength
Is There an Endurance Limit for Weldments?
Fatigue Failure in Compression Loading?
16.7 Treating a Weld as a Line
16.8 Eccentrically Loaded Weld Patterns
16.9 Design Considerations for Weldments in Machines
16.10 Summary
16.11 References
16.12 Problems
Chapter 17. Clutches and Brakes
17.0 Introduction
17.1 Types of Brakes and Clutches
17.2 Clutch/Brake Selection and Specification
17.3 Clutch and Brake Materials
17.4 Disk Clutches
Uniform Pressure
Uniform Wear
17.5 Disk Brakes
17.6 Drum Brakes
Short-Shoe External Drum Brakes
Long-Shoe External Drum Brakes
Long-Shoe Internal Drum Brakes
17.7 Summary
17.8 References
17.9 Bibliography
17.10 Problems
Appendices
A. Material Properties
B. Beam Tables
C. Stress-Concentration Factors
D. Answers to Selected Problems
Index
Downloads Index

Citation preview

MACHINE DESIGN An Integrated Approach Sixth Edition

Robert L. Norton P.E. Milton P. Higgins II Distinguished Professor Emeritus Worcester Polytechnic Institute Worcester, Massachusetts

i

Senior Vice President Courseware Portfolio Management: Engineering, Computer Science, Mathematics, Statistics, and Global Editions: Marcia J. Horton Director, Portfolio Management: Engineering, Computer Science, Mathematics, Statistics, and Global Editions: Julian Partridge Team Lead: Engineering and Computer Science Norrin Dias Portfolio Management Assistant: Emily Egan Managing Producer, Engineering, Computer Science, and Mathematics: Scott Disanno Senior Content Producer: Erin Ault Project Manager: Louise Capulli Manager, Rights and Permissions: Ben Ferrini Operations Specialist: Maura Zaldivar-Garcia Inventory Manager: Bruce Boundy Product Marketing Manager: Yvonne Vannatta Field Marketing Manager: Demetrius Hall Marketing Assistant: Jon Bryant Cover Designer: Robert L. Norton Cover Printer: Phoenix Color/Hagerstown Printer/Binder: Lake Side Communications, Inc. (LSC) Copyright © 2020, 2014, 2011, 2006, 2000, 1998 Pearson Education, Inc., Hoboken, NJ 07030. All rights reserved. Manufactured in the United States of America. This publication is protected by copyright, and permission should be obtained from the publisher prior to any prohibited reproduction, storage in a retrieval system, or transmission in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise. For information regarding permissions, request forms and the appropriate contacts within the Pearson Education Global Rights & Permissions department, please visit www.pearsoned.com/permissions/. Many of the designations by manufacturers and seller to distinguish their products are claimed as trademarks. Where those designations appear in this book, and the publisher was aware of a trademark claim, the designations have been printed in initial caps or all caps. The author and publisher of this book have used their best efforts in preparing this book. These efforts include the development, research, and testing of the theories and programs to determine their effectiveness. The author and publisher make no warranty of any kind, expressed or implied, with regard to these programs or the documentation contained in this book. The author and publisher shall not be liable in any event for incidental or consequential damages in connection with, or arising out of, the furnishing, performance, or use of these programs. Mathcad is a registered trademark of PTC Inc. MATLAB™ is a registered trademark of The MathWorks, Inc. Microsoft is a registered trademark or trademark of Microsoft Corporation in the United States and/or other countries. TK Solver is a trademark of UTS Corporation in the U.S. and/or other countries. The cover photograph by the author is of Tom Norton, five-time New England Trail Riders Association (NETRA) Hare Scrambles Champion (and the author’s son). The book, including all art and equations, was written and electronically typeset by the author using InDesign and Mathtype software. All non-photographic figures were drawn in Freehand and Illustrator by the author. Library of Congress Cataloging-in-Publication Data Norton, Robert L. Machine Design An Integrated Approach 6ed / Robert L. Norton p. cm. Includes bibliographical references and index. ISBN: 0-13-518423-1 (hard cover) 1. Machine design. I. Title TJ230.N64 2020 621.8’15--dc22 97-19522 CIP Data on file 1 19

ISBN10: 0-13-518423-1 ISBN13: 978-0-13-518423-3

ABOUT THE AUTHOR Robert L. Norton earned undergraduate degrees in both mechanical engineering and industrial technology at Northeastern University, an MS in engineering design at Tufts University, and was awarded a Doctor of Engineering (h.c.) by Worcester Polytechnic Institute (WPI). He is a registered professional engineer in Massachusetts and Florida. He has over 50 years experience in engineering design and manufacturing and over 40 years experience teaching mechanical engineering, engineering design, computer science, and related subjects at Northeastern University, Tufts University, and WPI. Norton has been on the faculty of WPI since 1981, and is currently the Milton P. Higgins II Distinguished Professor Emeritus. He is also the founder and president of Norton Associates Engineering Consultants since 1970. At Polaroid Corporation for 10 years, he designed cameras, related mechanisms, and high-speed automated machinery. He spent three years at Jet Spray Cooler Inc., designing food-handling machinery and products. For five years he helped develop artificial-heart and noninvasive assisted-circulation devices at the Tufts New England Medical Center and Boston City Hospital. Since leaving industry to join academia in 1974, he has continued as an independent consultant on engineering projects ranging from disposable medical products to high-speed production machinery. He holds thirteen U.S. patents. He is the author of numerous technical papers and journal articles covering kinematics, dynamics of machinery, cam design and manufacturing, computers in education, engineering education, and of the texts Design of Machinery, Kinematics and Dynamics of Machinery, Machine Design: An Integrated Approach, and the Cam Design and Manufacturing Handbook. He is a Life Fellow of the American Society of Mechanical Engineers and a past member of the Society of Automotive Engineers and the American Society for Engineering Education. In 2007, he was selected as a U.S. Professor of the Year by the Council for the Advancement and Support of Education (CASE) and the Carnegie Foundation for the Advancement of Teaching, who jointly present the only national awards for higher-education teaching excellence given in the United States of America. WHAT USERS SAY ABOUT THE BOOK Your text is the best of all the texts I have used—the balance of fundamentals and practice is especially important, and you have achieved that with aplomb! —Professor John P. H. Steele, Colorado School of Mines This book is one of the best-written on any engineering subject that I have come across. We switched to the book because we felt that the students would get a lot out of reading the text. —Professor Ed Howard, Milwaukee School of Engineering (The) writing is clear and meaningful, (and the) text is much more accessible to students because of the practical nature of (its) case studies and problem sets. —Professor Douglas Walcerz, York College of Pennsylvania I’m doing my machine design homework now, and . . . your book is the single best text I have ever used in my educational career. It is straightforward, easy to understand and has good examples. —Josh, student, Colorado School of Mines

This book is dedicated to the memory of: Donald N. Zwiep 1924–2012 Provost, Department Head, and Professor Emeritus Worcester Polytechnic Institute A gentleman and a leader, without whose faith and foresight, this book would never have been written.

PREFACE Introduction This text is intended for the Design of Machine Elements courses typically given in the junior year of most mechanical engineering curricula. The usual prerequisites are a first course in Statics and Dynamics, and one in Strength of Materials. The purpose of this book is to present the subject matter in an up-to-date manner with a strong design emphasis. The level is aimed at junior-senior mechanical engineering students. A primary goal was to write a text that is very easy to read and that students will enjoy reading despite the inherent dryness of the subject matter. This textbook is designed to be an improvement over others currently available and to provide methods and techniques that take full advantage of computer-aided analysis. It emphasizes design and synthesis as well as analysis. Example problems, case studies, and solution techniques are spelled out in detail and are self-contained. All the illustrations are done in two colors. Short problems are provided in each chapter and, where appropriate, longer unstructured design-project assignments are given. The book is independent of any particular computer program. Computer files for the solution of all the examples and case studies written in several different languages (Mathcad, MATLAB, Excel, and TK Solver) are provided on the book’s website at http://www.pearsonhighered.com/norton. Several other programs written by the author are available from the author. These include a matrix solver (MATRIX.exe). An index of the website’s content is on the website. While this book attempts to be thorough and complete on the engineering-mechanics topics of failure theory and analysis, it also emphasizes the synthesis and design aspects of the subject to a greater degree than most other texts in print on this subject. It points out the commonality of the analytical approaches needed to design a wide variety of elements and emphasizes the use of computer-aided engineering as an approach to the design and analysis of these classes of problems. The author’s approach to this course is based on over 50 years of practical experience in mechanical engineering design, both in industry and as a consultant to industry. He has taught mechanical engineering design at the university level for 40 of those years as well.

What’s in the Sixth Edition? • Twenty-one Master Lecture videos on the topics of most chapters are provided on the website. These are taken from the author’s live lectures to classes at WPI. The student can watch these videos to review and enhance their understanding of the book’s topics. • Eight short videos are provided on the website in which the author demonstrates various principles of stress analysis and shows examples of common machine parts such as springs, gears, and bearings. • Six videos that show real machinery in operation are also provided on the website. • Over 80 problems are added with many being in SI units. • In addition to the printed version of the text, digital e-book versions are also available. These have hotlinks to all the videos and to the downloadable content provided. There

viii

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

are 37 videos. All of these are marked in the print version as well, with their URLs provided, and they can be downloaded by print-book users. A Video Contents is provided, and all other downloadable items are listed in the Downloads Index. • The author-written programs that come with the book have been completely rewritten to improve their interface and usability, and they are now compatible with the latest operating systems and computers. Programs Linkages and Dynacam have been completely rewritten and are much improved. Program Matrix is updated. These computer programs undergo frequent revision to add features and enhancements. Once installed, the programs can be updated from the Start Menu. Users should occasionally check for updates. • All the downloadable files are accessible to digital-book users through the publisher’s Mastering website via links in the digital book. Any instructor or student who uses the print book may register on my website, http://www.designofmachinery.com , either as a student or instructor, and I will send them a password to access a protected site where they can download the latest versions of my computer programs, Linkages, Dynacam, and Matrix, all videos, and all files listed in the Downloads Index. Note that I personally review each of these requests for access and approve only those that are filled out completely and correctly according to the provided instructions. I require complete information and only accept university email addresses.

Philosophy This is often the first course that mechanical engineering students see that presents them with design challenges rather than set-piece problems. Nevertheless, the type of design addressed in this course is that of detailed design, which is only one part of the entire design-process spectrum. In detailed design, the general concept, application, and even general shape of the required device are typically known at the outset. We are not trying to invent a new device so much as define the shape, size, and material of a particular machine element such that it will not fail under the loading and environmental conditions expected in service. The traditional approach to the teaching of the Elements course has been to emphasize the design of individual machine parts, or elements, such as gears, springs, shafts, etc. One criticism that is sometimes directed at the Elements course (or textbook) is that it can easily become a “cookbook” collection of disparate topics that does not prepare the student to solve other types of problems not found in the recipes presented. There is a risk of this happening. It is relatively easy for the instructor (or author) to allow the course (or text) to degenerate into the mode “Well, it’s Tuesday, let’s design springs—on Friday, we’ll do gears.” If this happens, it may do the student a disservice because it doesn’t necessarily develop a fundamental understanding of the practical application of the underlying theories to design problems. However, many of the machine elements typically addressed in this course provide superb examples of the underlying theory. If viewed in that light, and if presented in a general context, they can be an excellent vehicle for the development of student understanding of complex and important engineering theories. For example, the topic of preloaded bolts is a perfect vehicle to introduce the concept of prestressing used as a foil against fatigue loading. The student may never be called upon in practice to design a preloaded bolt, but he or she may well utilize the understanding of prestressing gained from the experience. The design of helical gears to withstand time-varying loads provides an excellent vehicle to develop the student’s understanding of combined stresses, Hertzian stresses, and fatigue failure. Thus the elements approach is a valid and defensible one as long as the approach taken in the text is sufficiently global. That is, it should not be allowed to degenerate into a collection of apparently unrelated exercises, but rather provide an integrated approach.

ix

Another area in which the author has found existing texts (and Machine Elements courses) to be deficient is the lack of connection made between the dynamics of a system and the stress analysis of that system. Typically, these texts present their machine elements with (magically) predefined forces on them. The student is then shown how to determine the stresses and deflections caused by those forces. In real machine design, the forces are not always predefined and can, in large part, be due to the accelerations of the masses of the moving parts. However, the masses cannot be accurately determined until the geometry is defined and a stress analysis done to determine the strength of the assumed part. Thus, an impasse exists that is broken only by iteration, i.e., assume a part geometry and define its geometric and mass properties, calculate the dynamic loads due in part to the material and geometry of the part. Then calculate the stresses and deflections resulting from those forces, find out it fails, redesign, and repeat.

An Integrated Approach The text is divided into two parts. The first part presents the fundamentals of stress, strain, deflection, materials properties, failure theories, fatigue phenomena, fracture mechanics, FEA, etc. These theoretical aspects are presented in similar fashion to other texts. The second part presents treatments of specific, common design elements used as examples of applications of the theory but also attempts to avoid presenting a string of disparate topics in favor of an integrated approach that ties the various topics together via case studies. Most Elements texts contain many more topics and more content than can possibly be covered in a one-semester course. Before writing the first edition of this book, a questionnaire was sent to 200 U.S. university instructors of the Elements course to solicit their opinions on the relative importance and desirability of the typical set of topics in an Elements text. With each revision to second through fifth editions, users were again surveyed to determine what should be changed or added. The responses were analyzed and used to influence the structure and content of this book in all editions. One of the strongest desires originally expressed by the respondents was for case studies that present realistic design problems. We have attempted to accomplish this goal by structuring the text around a series of ten case studies. These case studies present different aspects of the same design problem in successive chapters, for example, defining the static or dynamic loads on the device in Chapter 3, calculating the stresses due to the static loads in Chapter 4, and applying the appropriate failure theory to determine its safety factor in Chapter 5. Later chapters present more complex case studies, with more design content. The case study in Chapter 6 on fatigue design is one such example a real problem taken from the author’s consulting practice. Chapter 8 presents FEA analyses of several of these case studies and compares those results to the classical solutions done in prior chapters. The case studies provide a series of machine design projects throughout the book that contain various combinations of the elements normally dealt with in this type of text. The assemblies contain some collection of elements such as links subjected to combined axial and bending loads, column members, shafts in combined bending and torsion, gearsets under alternating loads, return springs, fasteners under fatigue loading, rolling element bearings, etc. This integrated approach has several advantages. It presents the student with a generic design problem in context rather than as a set of disparate, unrelated entities. The student can then see the interrelationships and the rationales for the design decisions that affect the individual elements. These more comprehensive case studies are in Part II of the text. The case studies in Part I are more limited in scope and directed to the engineering mechanics topics of the chapter. In addition to the case studies, each chapter has a selection of worked-out examples to reinforce particular topics.

x

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Chapter 9, Design Case Studies, is devoted to the setup of three design case studies that are used in the following chapters to reinforce the concepts behind the design and analysis of shafts, springs, gears, fasteners, etc. Not all aspects of these design case studies are addressed as worked-out examples since another purpose is to provide material for student-project assignments. The author has used these case study topics as multi-week or term-long project assignments for groups or individual students with good success. Assigning open-ended project assignments serves to reinforce the design and analysis aspects of the course much better than set-piece homework assignments.

Problem Sets Most of the 967 problem sets (767, or 79%) are independent within a chapter, responding to requests by users of the first edition to decouple them. The other 21% of the problem sets are built upon in succeeding chapters. These linked problems have the same dash number in each chapter and their problem number is boldface to indicate their commonality among chapters. For example, Problem 3-4 asks for a static force analysis of a trailer hitch; Problem 4-4 requests a stress analysis of the same hitch based on the forces calculated in Problem 3-4; Problem 5-4 asks for the static safety factor for the hitch using the stresses calculated in Problem 4-4; Problem 6-4 requests a fatigue-failure analysis of the same hitch, and Problem 7-4 requires a surface stress analysis. The same trailer hitch is used as an FEA case study in Chapter 8. Thus, the complexity of the underlying design problem is unfolded as new topics are introduced. An instructor who wishes to use this approach can assign problems with the same dash number in succeeding chapters. If one does not want to assign an earlier problem on which a later one is based, the solution manual data from the earlier problem can be provided to the students. Instructors who do not like interlinked problems can avoid them entirely and select from the 767 problems with nonbold problem numbers that are independent within their chapters.

Text Arrangement Chapter 1 provides an introduction to the design process, problem formulation, safety factors, and units. Material properties are reviewed in Chapter 2 since even the student who has had a first course in material science or metallurgy typically has but a superficial understanding of the wide spectrum of engineering material properties needed for machine design. Chapter 3 presents a discussion of the fundamentals of kinematic linkages and cams. It also provides a review of static and dynamic loading analysis, including beam, vibration, and impact loading, and sets up a series of case studies that are used in later chapters to illustrate the stress and deflection analysis topics with some continuity. The Design of Machine Elements course, at its core, is really an intermediate-level, applied stress-analysis course. Accordingly, a review of the fundamentals of stress and deflection analysis is presented in Chapter 4. Static failure theories are presented in detail in Chapter 5 since the students have typically not yet fully digested these concepts from their first stressanalysis course. Fracture-mechanics analysis for static loads is also introduced. The Elements course is typically the student’s first exposure to fatigue analysis since most introductory stress-analysis courses deal only with statically loaded problems. Accordingly, fatigue-failure theory is presented at length in Chapter 6 with the emphasis on stress-life approaches to high-cycle fatigue design, which is commonly used in the design of rotating machinery. Fracture-mechanics theory is further discussed with regard to crack propagation under cyclic loading. Strain-based methods for low-cycle fatigue analysis are not presented

xi

but their application and purpose are introduced to the reader and bibliographic references are provided for further study. Residual stresses are also addressed. Chapter 7 presents a thorough discussion of the phenomena of wear mechanisms, surface contact stresses, and surface fatigue. Chapter 8 provides an introduction to Finite Element Analysis (FEA). Many instructors are using the machine elements course to introduce students to FEA as well as to instruct them in the techniques of machine design. The material presented in Chapter 8 is not intended as a substitute for education in FEA theory. That material is available in many other textbooks devoted to that subject and the student is urged to become familiar with FEA theory through coursework or self-study. Instead, Chapter 8 presents proper techniques for the application of FEA to practical machine design problems. Issues of element selection, mesh refinement, and the definition of proper boundary conditions are developed in some detail. These issues are not usually addressed in books on FEA theory. Many engineers in training today will, in their professional practice, use CAD solid modeling software and commercial finite element analysis code. It is important that they have some knowledge of the limitations and proper application of those tools. This chapter can be taken up earlier in the course if desired, especially if the students are expected to use FEA to solve assigned tasks. It is relatively independent of the other chapters. Many of various chapters’ problem assignments have Solidworks models of their geometry provided on the website. These eight chapters comprise Part I of the text and lay the analytical foundation needed for design of machine elements. They are arranged to be taken up in the order presented and build upon each other with the exception of Chapter 8 on FEA. Part II of the text presents the design of machine elements in context as parts of a whole machine. The chapters in Part II are essentially independent of one another and can be taken (or skipped) in any order that the instructor desires (except that Chapter 12 on spur gears should be studied before Chapter 13 on helical, bevel, and worm gears). It is unlikely that all topics in the book can be covered in a one-term or one semester course. Uncovered chapters will still serve as a reference for engineers in their professional practice. Chapter 9 presents a set of design case studies to be used as assignments and as example case studies in the following chapters and also provides a set of suggested design project assignments in addition to the detailed case studies as described above. Chapter 10 investigates shaft design using the fatigue-analysis techniques developed in Chapter 6. Chapter 11 discusses fluid-film and rolling-element bearing theory and application using the theory developed in Chapter 7. Chapter 12 gives a thorough introduction to the kinematics, design and stress analysis of spur gears using the latest AGMA recommended procedures. Chapter 13 extends gear design to helical, bevel, and worm gearing. Chapter 14 covers spring design including helical compression, extension and torsion springs, as well as a thorough treatment of Belleville springs. Chapter 15 deals with screws and fasteners including power screws and preloaded fasteners. Chapter 16 presents an up-to-date treatment of the design of weldments for both static and dynamic loading. Chapter 17 presents an introduction to the design and specification of disk and drum clutches and brakes. The appendices contain material-strength data, beam tables, and stress-concentration factors, as well as answers to selected problems.

Supplements A Solutions Manual is available to instructors from the publisher and PowerPoint slides of all figures and tables in the text are available on the publisher’s website (password protected) at: http://www.pearsonhighered.com/

xii

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

To download these resources, choose the Instructor Support tab to register as an instructor and follow instructions on the site to obtain the resources provided. Mathcad files for all the problem solutions are available with the solutions manual. This computerized approach to problem solutions has significant advantages to the instructor who can easily change any assigned problem’s data and instantly solve it. Thus, an essentially infinite supply of problem sets is available, going far beyond those defined in the text. The instructor also can easily prepare and solve exam problems by changing data in the supplied files. As errata are discovered they will be posted on the author’s personal website at: http://www.designofmachinery.com/MD/errata.html

Professors who adopt the book may register at the author’s personal website to obtain additional information relevant to the subject (syllabi, master lectures, project assignments, etc.) and the text and to download web content and updated software (password protected). Go to: http://designofmachinery.com/books/machine-design/professors-using-our-books-md/ Anyone who purchases the book may register at the author’s personal website to request downloads and updated software for the current edition (password protected). Go to: http://designofmachinery.com/books/machine-design/machine-design-6thed-for-students-login/

Acknowledgments The author expresses his sincere appreciation to all those who reviewed the first edition of the text in various stages of development including Professors J. E. Beard, Michigan Tech; J. M. Henderson, U. California, Davis; L. R. Koval, U. Missouri, Rolla; S. N. Kramer, U. Toledo; L. D. Mitchell, Virginia Polytechnic; G. R. Pennock, Purdue; D. A. Wilson, Tennessee Tech; Mr. John Lothrop; and Professor J. Ari-Gur, Western Michigan University, who also taught from a class-test version of the book. Robert Herrmann (WPI-ME ‘94) provided some problems and Charles Gillis (WPI-ME ‘96) solved most of the problem sets for the first edition. Professors John R. Steffen of Valparaiso University, R. Jay Conant of Montana State, Norman E. Dowling of Virginia Polytechnic, and Francis E. Kennedy of Dartmouth made many useful suggestions for improvement and caught many errors. Special thanks go to Professor Hartley T. Grandin of WPI, who provided much encouragement and many good suggestions and ideas throughout the book’s gestation, and also taught from various classtest versions. Two former and the current Prentice Hall editors deserve special mention for their efforts in developing this book: Doug Humphrey, who wouldn’t take no for an answer in persuading me to write it and Bill Stenquist, who usually said yes to my requests and expertly shepherded the book through to completion in its first edition. Norrin Dias’ support has helped marshall the sixth edition into print. Since the book’s first printing in 1995, several users have kindly pointed out errors and suggested improvements. My thanks go to Professors R. Boudreau of U. Moncton, Canada, V. Glozman of Cal Poly Pomona, John Steele of Colorado School of Mines, Burford J. Furman of San Jose State University, and Michael Ward of California State University, Chico.

xiii

Several other faculty have been kind enough to point out errors and offer constructive criticisms and suggestions for improvement in the later editions. Notable among these are: Professors Cosme Furlong of Worcester Polytechnic Institute, Joseph Rencis of University of Arkansas, Annie Ross of Universite de Moncton, Andrew Ruina of Cornell University, Douglas Walcerz of York College, and Thomas Dresner of Mountain City, CA. Dr. Duane Miller of Lincoln Electric Company provided invaluable help with Chapter 16 on weldments and reviewed several drafts. Professor Stephen Covey of St. Cloud State University, and engineers Gregory Aviza and Charles Gillis of P&G Gillette also provided valuable feedback on the weldment chapter. Professor Robert Cornwell of Seattle University reviewed the discussion in Chapter 15 of his new method for the calculation of bolted joint stiffness and his method for computing stress concentration in rectangular wire springs discussed in Chapter 14. Professors Fabio Marcelo Peña Bustos of Universidad Autónoma de Manizales, Caldas, Colombia, and Juan L. Balsevich-Prieto of Universidad Católica Nuestra Señora de la Asunción, Asunción, Paraguay, were kind enough to point out errata in the Spanish translation. Special thanks are due to William Jolley of The Gillette Company who created the FEA models for the examples and reviewed Chapter 8, and to Edwin Ryan, retired Vice President of Engineering at Gillette, who provided invaluable support. Donald A. Jacques of the UTC Fuel Cells division of the United Technologies Company also reviewed Chapter 8 on Finite Element Analysis and made many useful suggestions. Professor Eben C. Cobb of Worcester Polytechnic Institute and his student Thomas Watson created the Solidworks models of many problem assignments and case studies and solved the FEA for the case studies that are on the website. Thanks are due several people who responded to surveys for the fifth edition and made many good suggestions: Steven J. Covey of St. Cloud State University, Yesh P. Singh of University of Texas at San Antonio, César Augusto Álvarez Vargas of Universidad Autonoma de Manizales, Caldas, Colombia, Ardeshir Kianercy of University of Southern California, Kenneth W. Miller of St. Cloud State University, Yannis Korkolis of University of New Hampshire, J. Alex Thomasson of Texas A&M, Timothy Dewhurst of Cedarville University, Jon Svenninggaard of VIA University College, Horsens, Denmark, John D. Landes of University of Tennessee, Peter Schuster of California Polytechnic State University, Amanuel Haque of Penn State University, J. Brian Jordon of University of Alabama, Yabin Liao of Arizona State University, Raymond K. Yee of San Jose State University, Jean-Michel Dhainaut of EmbryRiddle Aeronautical University, Pal Molian of Iowa State University, and John Strenkowski of North Carolina State University. The author is greatly indebted to Thomas A. Cook, Professor Emeritus, Mercer University, who did the Solutions Manual for this book, updated the Mathcad examples, and contributed most of the new problem sets for this edition. Thanks also to Dr. Adriana Hera of Worcester Polytechnic Institute who updated the MATLAB and Excel models of all the examples and case studies and thoroughly vetted their correctness. Finally, Nancy Norton, my infinitely patient wife for the past fifty-nine years, deserves renewed kudos for her unfailing support and encouragement during many summers of “book widowhood.” I could not have done it without her. Every effort has been made to eliminate errors from this text. Any that remain are the author’s responsibility. He will greatly appreciate being informed of any errors that still remain so they can be corrected in future printings. An e-mail to [email protected] will be sufficient.

Robert L. Norton Mattapoisett, MA. June 1, 2019

xv

Contents Preface

_______________________________________________________ vii

video contents __________________________________________________xxix

Part i

fundamentals

chaPter 1 introduction 1.1

to

1

design_________________________________ 3

Design ..........................................................................................................3 Machine Design

3

1.2

A Design Process ..........................................................................................5

1.3

Problem Formulation and Calculation ...........................................................7 Definition Stage Preliminary Design Stage Detailed Design Stage Documentation Stage

1.4

The Engineering Model .................................................................................9 Estimation and First-Order Analysis The Engineering Sketch

1.5

8 8 8 9 9 10

Computer-Aided Design and Engineering ...................................................10 Computer-Aided Design (CAD) Computer-Aided Engineering (CAE) Computational Accuracy

11 12 15

1.6

The Engineering Report...............................................................................16

1.7

Factors of Safety and Design Codes ............................................................16 Factor of Safety Choosing a Safety Factor Design and Safety Codes

16

17 19

1.8

Statistical Considerations............................................................................20

1.9

Units ...........................................................................................................20

1.10

Summary .....................................................................................................25

1.11

References ..................................................................................................26

1.12

Web References ..........................................................................................26

1.13

Bibliography ...............................................................................................27

1.14

Problems .....................................................................................................27

xvi

MACHINE DESIGN

-

An Integrated Approach -

chaPter 2 materials

and

Sixth Edition

Processes ______________________________29

2.0

Introduction ................................................................................................29

2.1

Material-Property Definitions .....................................................................29 The Tensile Test Ductility and Brittleness The Compression Test The Bending Test The Torsion Test Fatigue Strength and Endurance Limit Impact Resistance Fracture Toughness Creep and Temperature Effects

31 33 35 35 35 37 38 40 40

2.2

The Statistical Nature of Material Properties...............................................41

2.3

Homogeneity and Isotropy..........................................................................42

2.4

Hardness .....................................................................................................42 Heat Treatment Surface (Case) Hardening Heat Treating Nonferrous Materials Mechanical Forming and Hardening

2.5

Coatings and Surface Treatments ................................................................48 Galvanic Action Electroplating Electroless Plating Anodizing Plasma-Sprayed Coatings Chemical Coatings

2.6

49 50 50 51 51 51

General Properties of Metals ......................................................................52 Cast Iron Cast Steels Wrought Steels Steel Numbering Systems Aluminum Titanium Magnesium Copper Alloys

2.7

44 45 46 46

52 53 53 54 56 58 59 59

General Properties of Nonmetals ................................................................60 Polymers Ceramics Composites

60 62 62

2.8

Selecting Materials .....................................................................................63

2.9

Summary .....................................................................................................64

2.10

References ..................................................................................................68

2.11

Web References ..........................................................................................68

2.12

Bibliography ...............................................................................................68

2.13

Problems .....................................................................................................69

xvii

chaPter 3 Kinematics

and

load determination ___________________73

3.0

Introduction ................................................................................................73

3.1

Degree of Freedom .....................................................................................73

3.2

Mechanisms................................................................................................74

3.3

Calculating Degree of Freedom (Mobility) ..................................................75

3.4

Common 1-DOF Mechanisms ....................................................................76 Fourbar Linkage and the Grashof Condition Sixbar Linkage Cam and Follower

3.5

Analyzing Linkage Motion ..........................................................................80 Types of Motion Complex Numbers as Vectors The Vector Loop Equation

3.6

82 84 86 87

Analyzing the Fourbar Crank-Slider ............................................................89 Solving for Position in the Fourbar Crank-Slider Solving for Velocity in the Fourbar Crank-Slider Solving for Acceleration in the Fourbar Crank-Slider Other Linkages

3.8

80 81 82

Analyzing the Fourbar Linkage....................................................................82 Solving for Position in the Fourbar Linkage Solving for Velocity in the Fourbar Linkage Angular Velocity Ratio and Mechanical Advantage Solving for Acceleration in the Fourbar Linkage

3.7

76 78 79

89 90 91 92

Cam Design and Analysis ...........................................................................92 The Timing Diagram The svaj Diagram Polynomials for the Double-Dwell Case Polynomials for the Single-Dwell Case Pressure Angle Radius of Curvature

93 94 94 97 99 100

3.9

Loading Classes For Force Analysis ..........................................................101

3.10

Free-body Diagrams..................................................................................103

3.11

Load Analysis............................................................................................103 Three-Dimensional Analysis Two-Dimensional Analysis Static Load Analysis

104 105 106

3.12

Two-Dimensional, Static Loading Case Studies.........................................106

3.13

Three-Dimensional, Static Loading Case Study .........................................121

3.14

Dynamic Loading Case Study....................................................................126

3.15

Vibration Loading .....................................................................................130 Natural Frequency Dynamic Forces

3.16

130 132

Impact Loading .........................................................................................134 Energy Method

135

xviii

MACHINE DESIGN

3.17

-

An Integrated Approach -

Sixth Edition

Beam Loading ...........................................................................................139 Shear and Moment Singularity Functions Superposition

139 141 151

3.18

Summary ...................................................................................................151

3.19

References ................................................................................................154

3.20

Web References ........................................................................................155

3.21

Bibliography .............................................................................................155

3.22

Problems ...................................................................................................156

chaPter 4 stress, strain,

and

deflection ______________________ 173

4.0

Introduction ..............................................................................................173

4.1

Stress ........................................................................................................173

4.2

Strain ........................................................................................................177

4.3

Principal Stresses ......................................................................................177

4.4

Plane Stress and Plane Strain ....................................................................180 Plane Stress Plane Strain

180 180

4.5

Mohr’s Circles ...........................................................................................180

4.6

Applied Versus Principal Stresses..............................................................185

4.7

Axial Tension ............................................................................................186

4.8

Direct Shear Stress, Bearing Stress, and Tearout........................................186 Direct Shear Direct Bearing Tearout Failure

4.9

Beams and Bending Stresses.....................................................................188 Beams in Pure Bending Shear Due to Transverse Loading

4.10

198 204

Castigliano’s Method ................................................................................207 Deflection by Castigliano’s Method Finding Redundant Reactions with Castigliano’s Method

4.12

188 192

Deflection in Beams ..................................................................................196 Deflection by Singularity Functions Statically Indeterminate Beams

4.11

187 187 188

209 209

Torsion ......................................................................................................210

4.13

Combined Stresses....................................................................................216

4.14

Spring Rates ..............................................................................................218

4.15

Stress Concentration .................................................................................219 Stress Concentration Under Static Loading Stress Concentration Under Dynamic Loading Determining Geometric Stress-Concentration Factors Designing to Avoid Stress Concentrations

4.16

220 222 222 224

Axial Compression - Columns ..................................................................226 Slenderness Ratio Short Columns

226 227

xix

Long Columns End Conditions Intermediate Columns

4.17

227 229 230

Stresses in Cylinders .................................................................................237 Thick-Walled Cylinders Thin-Walled Cylinders

237 238

4.18

Case Studies in Static Stress and Deflection Analysis ...............................239

4.19

Summary ...................................................................................................255

4.20

References ................................................................................................261

4.21

Bibliography .............................................................................................261

4.22

Problems ...................................................................................................262

chaPter 5 static failure theories ______________________________ 279 5.0

Introduction ..............................................................................................279

5.1

Failure of Ductile Materials Under Static Loading ....................................281 The von Mises-Hencky or Distortion-Energy Theory The Maximum Shear-Stress Theory The Maximum Normal-Stress Theory Comparison of Experimental Data with Failure Theories

5.2

Failure of Brittle Materials Under Static Loading ......................................294 Even and Uneven Materials The Coulomb-Mohr Theory The Modified-Mohr Theory

5.3

282 288 290 290 294 295 296

Fracture Mechanics ...................................................................................301 Fracture-Mechanics Theory Fracture Toughness Kc

302 305

5.4

Using The Static Loading Failure Theories ................................................309

5.5

Case Studies in Static Failure Analysis ......................................................310

5.6

Summary ...................................................................................................321

5.7

References ................................................................................................324

5.8

Bibliography .............................................................................................325

5.9

Problems ...................................................................................................326

chaPter 6 fatigue failure theories ____________________________ 339 6.0

Introduction ..............................................................................................339 History of Fatigue Failure

6.1

Mechanism of Fatigue Failure ..................................................................342 Crack Initiation Stage Crack Propagation Stage Fracture

6.2

339 343 343 344

Fatigue-Failure Models .............................................................................345 Fatigue Regimes The Stress-Life Approach The Strain-Life Approach The LEFM Approach

345 347 347 347

xx

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

6.3

Machine-Design Considerations ...............................................................348

6.4

Fatigue Loads............................................................................................349 Rotating Machinery Loading Service Equipment Loading

6.5

Measuring Fatigue Failure Criteria ............................................................351 Fully Reversed Stresses Combined Mean and Alternating Stress Fracture-Mechanics Criteria Testing Actual Assemblies

6.6

349 350 352 358 360 363

Estimating Fatigue Failure Criteria ............................................................364 Estimating the Theoretical Fatigue Strength Sf’ or Endurance Limit Se’ Correction Factors—Theoretical Fatigue Strength or Endurance Limit Corrected Fatigue Strength Sf or Corrected Endurance Limit Se Creating Estimated S-N Diagrams

364 366 373 373

6.7

Notches and Stress Concentrations ...........................................................378

6.8

Residual Stresses ......................................................................................383

6.9

Designing for High-Cycle Fatigue ............................................................388

6.10

Designing for Fully Reversed Uniaxial Stresses.........................................388

Notch Sensitivity

Design Steps for Fully Reversed Stresses with Uniaxial Loading

6.11

389

Designing for Fluctuating Uniaxial Stresses .............................................396 Creating the Modified-Goodman Diagram Applying Stress-Concentration Effects with Fluctuating Stresses Determining the Safety Factor with Fluctuating Stresses Design Steps for Fluctuating Stresses

6.12

379

397 400 402 405

Designing for Multiaxial Stresses in Fatigue .............................................412 Frequency and Phase Relationships Fully Reversed Simple Multiaxial Stresses Fluctuating Simple Multiaxial Stresses Complex Multiaxial Stresses

413 413 414 415

6.13

A General Approach to High-Cycle Fatigue Design ..................................417

6.14

A Case Study in Fatigue Design ................................................................422

6.15

Summary ...................................................................................................435

6.16

References ................................................................................................439

6.17

Bibliography .............................................................................................442

6.18

Problems ...................................................................................................443

7.0

Introduction ..............................................................................................457

chaPter 7 surface failure _____________________________________ 457 7.1

Surface Geometry .....................................................................................459

7.2

Mating Surfaces ........................................................................................461

7.3

Friction......................................................................................................462 Effect of Roughness on Friction Effect of Velocity on Friction Rolling Friction Effect of Lubricant on Friction

463 463 463 464

xxi

7.4

Adhesive Wear ..........................................................................................464 The Adhesive-Wear Coefficient

7.5

Abrasive Wear...........................................................................................468 Abrasive Materials Abrasion-Resistant Materials

7.6

467 471 471

Corrosion Wear .........................................................................................472 Corrosion Fatigue Fretting Corrosion

473 473

7.7

Surface Fatigue .........................................................................................474

7.8

Spherical Contact ......................................................................................476 Contact Pressure and Contact Patch in Spherical Contact Static Stress Distributions in Spherical Contact

7.9

Cylindrical Contact ...................................................................................482 Contact Pressure and Contact Patch in Parallel Cylindrical Contact Static Stress Distributions in Parallel Cylindrical Contact

7.10

482 483

General Contact ........................................................................................486 Contact Pressure and Contact Patch in General Contact Stress Distributions in General Contact

7.11

476 478

486 488

Dynamic Contact Stresses .........................................................................491 Effect of a Sliding Component on Contact Stresses

491

7.12

Surface Fatigue Failure Models—Dynamic Contact ..................................498

7.13

Surface Fatigue Strength ...........................................................................501

7.14

Summary ...................................................................................................508

7.15

References ................................................................................................512

7.16

Problems ...................................................................................................514

chaPter 8 finite 8.0

element

analysis ______________________________ 521

Introduction .............................................................................................521 Stress and Strain Computation

522

8.1

Finite Element Method..............................................................................523

8.2

Element Types ...........................................................................................525 Element Dimension and Degree of Freedom (DOF) Element Order H-Elements Versus P-Elements Element Aspect Ratio

8.3

525 526 527 527

Meshing....................................................................................................527 Mesh Density Mesh Refinement Convergence

528 528 528

8.4

Boundary Conditions ................................................................................532

8.5

Applying Loads .........................................................................................542

8.6

Testing the Model (Verification)................................................................543

8.7

Modal Analysis .........................................................................................546

8.8

Case Studies .............................................................................................548

8.9

Summary ...................................................................................................558

8.10

References ................................................................................................559

xxii

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

8.11

Bibliography .............................................................................................559

8.12

Web Resources .........................................................................................559

8.13

Problems ...................................................................................................560

Part ii machine design ______________________________ 561 chaPter 9 design case studies _________________________________ 563 9.0

Introduction ..............................................................................................563

9.1

Case Study 8—A Portable Air Compressor................................................564

9.2

Case Study 9—A Hay-Bale Lifter ..............................................................567

9.3

Case Study 10—A Cam-Testing Machine..................................................571

9.4

Summary ...................................................................................................577

9.5

References ................................................................................................577

9.6

Design Projects .........................................................................................578

chaPter 10 shafts, Keys,

and

couPlings _______________________ 589

10.0

Introduction ..............................................................................................589

10.1

Shaft Loads ...............................................................................................589

10.2

Attachments and Stress Concentrations ....................................................591

10.3

Shaft Materials..........................................................................................593

10.4

Shaft Power...............................................................................................593

10.5

Shaft Loads ...............................................................................................594

10.6

Shaft Stresses ............................................................................................594

10.7

Shaft Failure in Combined Loading ...........................................................595

10.8

Shaft Design .............................................................................................596 General Considerations Design for Fully Reversed Bending and Steady Torsion Design for Fluctuating Bending and Fluctuating Torsion

10.9

596 597 599

Shaft Deflection ........................................................................................606 Shafts as Beams Shafts as Torsion Bars

607 607

10.10 Keys and Keyways ....................................................................................610 Parallel Keys Tapered Keys Woodruff Keys Stresses in Keys Key Materials Key Design Stress Concentrations in Keyways

610 611 612 612 613 613 614

10.11 Splines ......................................................................................................618 10.12 Interference Fits ........................................................................................620 Stresses in Interference Fits Stress Concentration in Interference Fits Fretting Corrosion

620 621 622

xxiii

10.13 Flywheel Design .......................................................................................625 Energy Variation in a Rotating System Determining the Flywheel Inertia Stresses in Flywheels Failure Criteria

626 628 630 631

10.14 Critical Speeds of Shafts ...........................................................................633 Lateral Vibration of Shafts and Beams—Rayleigh’s Method Shaft Whirl Torsional Vibration Two Disks on a Common Shaft Multiple Disks on a Common Shaft Controlling Torsional Vibrations

635 637 639 640 641 642

10.15 Couplings .................................................................................................644 Rigid Couplings Compliant Couplings

645 646

10.16 Case Study 8B ...........................................................................................648 Designing Driveshafts for a Portable Air Compressor

648

10.17 Summary ...................................................................................................652 10.18 References ................................................................................................654 10.19 Problems ...................................................................................................655

chaPter 11 Bearings

and

luBrication ___________________________ 665

11.0

Introduction ..............................................................................................665

11.1

Lubricants ...............................................................................................667

11.2

Viscosity ...................................................................................................669

11.3

Types of Lubrication..................................................................................670

A Caveat

Full-Film Lubrication Boundary Lubrication

667

671 673

11.4

Material Combinations in Sliding Bearings ...............................................673

11.5

Hydrodynamic Lubrication Theory ............................................................674 Petroff’s Equation for No-Load Torque Reynolds’ Equation for Eccentric Journal Bearings Torque and Power Losses in Journal Bearings

11.6

675 676 681

Design of Hydrodynamic Bearings ............................................................682 Design Load Factor—The Ocvirk Number Design Procedures

682 684

11.7

Nonconforming Contacts ..........................................................................688

11.8

Rolling-element bearings ........................................................................695 Comparison of Rolling and Sliding Bearings Types of Rolling-Element Bearings

11.9

696 696

Failure of Rolling-element bearings ..........................................................700

11.10 Selection of Rolling-element bearings ......................................................701 Basic Dynamic Load Rating C Modified Bearing Life Rating Basic Static Load Rating C0

701 702 703

xxiv

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Combined Radial and Thrust Loads Calculation Procedures

704 705

11.11 Bearing Mounting Details .........................................................................707 11.12 Special Bearings ........................................................................................708 11.13 Case Study 10B ........................................................................................710 11.14 Summary ...................................................................................................712 Important Equations Used in This Chapter

713

11.15 References ................................................................................................715 11.16 Problems ...................................................................................................717

chaPter 12 sPur gears _________________________________________ 725 12.0

Introduction ..............................................................................................725

12.1

Gear Tooth Theory ...................................................................................727 The Fundamental Law of Gearing The Involute Tooth Form Pressure Angle Gear Mesh Geometry Rack and Pinion Changing Center Distance Backlash Relative Tooth Motion

727 728 729 730 731 731 733 733

12.2

Gear Tooth Nomenclature .........................................................................733

12.3

Interference and Undercutting ..................................................................736

12.4

Contact Ratio .............................................................................................738

12.5

Gear Trains ................................................................................................740

Unequal-Addendum Tooth Forms

Simple Gear Trains Compound Gear Trains Reverted Compound Trains Epicyclic or Planetary Gear Trains

12.6

737

740 741 742 743

Gear Manufacturing ..................................................................................746 Forming Gear Teeth Machining Roughing Processes Finishing Processes Gear Quality

746 747 747 747 749

12.7

Loading on Spur Gears .............................................................................749

12.8

Stresses in Spur Gears ..............................................................................752 Bending Stresses Surface Stresses

12.9

752 761

Gear Materials ..........................................................................................765 Material Strengths Bending-Fatigue Strengths for Gear Materials Surface-Fatigue Strengths for Gear Materials

766 767 769

12.10 Lubrication of Gearing ..............................................................................775 12.11 Design of Spur Gears ................................................................................776

xxv

12.12 Case Study 8C ..........................................................................................777 12.13 Summary ...................................................................................................783 12.14 References ................................................................................................784 12.15 Problems ...................................................................................................785

chaPter 13 helical, Bevel,

and

Worm gears ___________________ 791

13.0

Introduction .............................................................................................791

13.1

Helical Gears ................................................................................................................................791 Helical Gear Geometry Helical-Gear Forces Virtual Number of Teeth Contact Ratios Stresses in Helical Gears

13.2

Bevel Gears ...............................................................................................804 Bevel-Gear Geometry and Nomenclature Bevel-Gear Mounting Forces on Bevel Gears Stresses in Bevel Gears

13.3

793 794 795 796 796 804 806 806 806

Wormsets..................................................................................................810 Materials for Wormsets Lubrication in Wormsets Forces in Wormsets Wormset Geometry Rating Methods A Design Procedure for Wormsets

813 814 814 814 815 817

13.4

Case Study 9B ...........................................................................................817

13.5

Summary ...................................................................................................821

13.6

References ................................................................................................825

13.7

Problems ...................................................................................................825

chaPter 14 sPring design _____________________________________ 829 14.0

Introduction ..............................................................................................829

14.1

Spring Rate ...................................................................................................................................829

14.2

Spring Configurations .............................................................................................................832

14.3

Spring Materials........................................................................................834 Spring Wire Flat Spring Stock

14.4

834 837

Helical Compression Springs ....................................................................838 Spring Lengths End Details Active Coils Spring Index Spring Deflection Spring Rate Stresses in Helical Compression Spring Coils Helical Coil Springs of Nonround Wire Residual Stresses

838 839 839 840 840 840 840 842 843

xxvi

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Buckling of Compression Springs Compression-Spring Surge Allowable Strengths for Compression Springs The Torsional-Shear S-N Diagram for Spring Wire The Modified-Goodman Diagram for Spring Wire

844 844 846 847 849

14.5

Designing Helical Compression Springs for Static Loading .......................851

14.6

Designing Helical Compression Springs for Fatigue Loading ....................856

14.7

Helical Extension Springs ........................................................................863 Active Coils in Extension Springs Spring Rate of Extension Springs Spring Index of Extension Springs Coil Preload in Extension Springs Deflection of Extension Springs Coil Stresses in Extension Springs End Stresses in Extension Springs Surging in Extension Springs Material Strengths for Extension Springs Design of Helical Extension Springs

14.8

Helical Torsion Springs..............................................................................874 Terminology for Torsion Springs Number of Coils in Torsion Springs Deflection of Torsion Springs Spring Rate of Torsion Springs Coil Closure Coil Stresses in Torsion Springs Material Parameters for Torsion Springs Safety Factors for Torsion Springs Designing Helical Torsion Springs

14.9

863 864 864 864 864 865 865 866 866 866 875 875 875 876 876 876 877 878 878

Belleville Spring Washers .........................................................................881 Load-Deflection Function for Belleville Washers Stresses in Belleville Washers Static Loading of Belleville Washers Dynamic Loading Stacking Springs Designing Belleville Springs

882 884 885 885 885 886

14.10 Case Study 10C ........................................................................................888 14.11 Summary ...................................................................................................893 14.12 References ................................................................................................896 14.13 Problems ...................................................................................................897

chaPter 15 screWs

and

fasteners _____________________________ 903

15.0

Introduction ..............................................................................................903

15.1

Standard Thread Forms ...........................................................................906 Tensile Stress Area Standard Thread Dimensions

15.2

907 908

Power Screws ............................................................................................909 Square, Acme, and Buttress Threads Power Screw Application Power Screw Force and Torque Analysis

910 910 912

xxvii

Friction Coefficients Self-Locking and Back-Driving of Power Screws Screw Efficiency Ball Screws

15.3

Stresses in Threads ...................................................................................918 Axial Stress Shear Stress Torsional Stress

15.4

913 914 915 915 919 919 920

Types of Screw Fasteners .........................................................................920 Classification by Intended Use Classification by Thread Type Classification by Head Style Nuts and Washers

921 921 921 923

15.5

Manufacturing Fasteners ..........................................................................923

15.6

Strengths of Standard Bolts and Machine Screws .....................................925

15.7

Preloaded Fasteners in Tension ................................................................925 Preloaded Bolts Under Static Loading Preloaded Bolts Under Dynamic Loading

15.8

Determining the Joint Stiffness Factor ......................................................939 Joints With Two Plates of the Same Material Joints With Two Plates of Different Materials Gasketed Joints

15.9

929 933 941 941 943

Controlling Preload ...................................................................................948 The Turn-of-the-Nut Method Torque-Limited Fasteners Load-Indicating Washers Torsional Stress Due to Torquing of Bolts

949 949 949 950

15.10 Fasteners in Shear .....................................................................................950 Dowel Pins Centroids of Fastener Groups Determining Shear Loads on Fasteners

952 953 954

15.11 Case Study 8D ..........................................................................................956 15.12 Summary ...................................................................................................961 15.13 References ................................................................................................964 15.14 Bibliography .............................................................................................964 15.15 Problems ...................................................................................................965

chaPter 16 Weldments ________________________________________ 973 16.0

Introduction ..............................................................................................973

16.1

Welding Processes ....................................................................................975 Types of Welding in Common Use Why Should a Designer Be Concerned with the Welding Process?

16.2

Weld Joints and Weld Types .....................................................................977 Joint Preparation Weld Specification

16.3

976 977 979 979

Principles of Weldment Design .................................................................979

xxviii

MACHINE DESIGN

-

An Integrated Approach

16.4

Static Loading of Welds ............................................................................981

16.5

Static Strength of Welds............................................................................982 Residual Stresses in Welds Direction of Loading Allowable Shear Stress for Statically Loaded Fillet and PJP Welds

16.6

983 983 983

Dynamic Loading of Welds .......................................................................986 Effect of Mean Stress on Weldment Fatigue Strength Are Correction Factors Needed For Weldment Fatigue Strength? Effect of Weldment Configuration on Fatigue Strength Is There an Endurance Limit for Weldments? Fatigue Failure in Compression Loading?

986 986 987 991 992

16.7

Treating a Weld as a Line ..........................................................................993

16.8

Eccentrically Loaded Weld Patterns...........................................................999

16.9

Design Considerations for Weldments in Machines ...............................1000

16.10 Summary .................................................................................................1001 16.11 References ..............................................................................................1002 16.12 Problems .................................................................................................1003

chaPter 17 clutches

and

BraKes ______________________________ 1007

17.0

Introduction ............................................................................................1007

17.1

Types of Brakes and Clutches .................................................................1009

17.2

Clutch/Brake Selection and Specification ................................................1014

17.3

Clutch and Brake Materials .....................................................................1016

17.4

Disk Clutches ..........................................................................................1016 Uniform Pressure Uniform Wear

1017 1017

17.5

Disk Brakes .............................................................................................1019

17.6

Drum Brakes ...........................................................................................1020 Short-Shoe External Drum Brakes Long-Shoe External Drum Brakes Long-Shoe Internal Drum Brakes

1021 1023 1027

17.7

Summary .................................................................................................1027

17.8

References ..............................................................................................1030

17.9

Bibliography ...........................................................................................1030

17.10 Problems .................................................................................................1031

aPPendices ____________________________________________________ 1035 A Material Properties ............................................................................1035 B Beam Tables.......................................................................................1043 C Stress-Concentration Factors .............................................................1047 D Answers to Selected Problems...........................................................1055

index

____________________________________________________ 1065

doWnloads index _____________________________________________ 1078

VIDEO CONTENTS The Sixth Edition has a collection of Master Lecture Videos and Tutorials made by the author over a 31-year period while teaching at Worcester Polytechnic Institute. The lectures were recorded in a classroom in front of students in 2011/2012. Tutorials were done in a recording studio and were intended as supplements to class lectures. There are 37 instructional videos in total. One is a short introduction to the master lecture series and 20 are “50-minute” lectures. Eight are short tutorials and eight are demonstrations of machinery. The run times of all videos are noted in the tables. The sixth edition is available both as a print book and as digital media. The digital, ebook versions have active links that allow these videos to be run while reading the book. The print edition notes the names and URLs of all the videos in the text at their links. In addition to the lecture videos, all the digital content that was with the fifth and earlier editions is still available as downloads, including the author-written programs Linkages, Dynacam, and Matrix. An index of all the non-video downloadable files is in the Downloads Index. In the digital e-book versions, these are hotlinked to the text. The URL of each video is also provided for print-book readers to download them. Any instructor or student who uses the book may register on my website, http://www. designofmachinery.com , either as a student or instructor, and I will send them a password to access a protected site where they can download the latest versions of my computer programs, Linkages, Dynacam, and Matrix. They can also download the 33 videos and all the files listed in the Downloads Index. Note that I personally review each of these requests for access and will approve only those that are filled out completely and correctly according to the provided instructions. I require complete information and only accept university email addresses.

LECTURE VIDEOS Chapter Lecture 1 4 4 4 5 5 6 6 6 10 10 7 12 12 14 14 11 11 15 15 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

(Concatenate this URL with any filename below to run a video)

Topic Introduction Stress Review Stress Distribution Combined Stress, Stress Concentration, Columns Ductile Failure Theory Brittle Failure Theory Fatigue Failure Theory Fully Reversed Loads Fluctuating Loads Shaft Design I Shaft Design II Wear and Surface Fatigue Spur Gear Design I Spur Gear Design II Spring Design I Spring Design II Bearings and Lubrication Rolling Element Bearings Power Screws and Fasteners Preloaded Fasteners Finite Element Analysis

http://www.designofmachinery.com/MD/

Run Time

01_Introduction.mp4 02_Stress_Review.mp4 03_Stress_Distribution.mp4 04_Combined_stress_stress_concentration_columns.mp4 05_Ductile_Failure_Theory.mp4 06_Brittle_Failure_Theory.mp4 07_Fatigue_Failure_Theory.mp4 08_Fully_Reversed_Loads.mp4 09_Fluctuating_Loads.mp4 10_Shaft_Design_I.mp4 11_Shaft_Design_II.mp4 12_Wear_and_Surface_Fatigue.mp4 13_Spur_Gear_Design_I.mp4 14_Spur_Gear_Design_II.mp4 15_Spring_Design_I.mp4 16_Spring_Design_II.mp4 17_Bearings_and_Lubrication.mp4 18_Rolling_Element_Bearings.mp4 19_Power_Screws_and_Fasteners.mp4 20_Preloaded_Fasteners.mp4 21_Finite_Element_Analysis.mp4

03:30 53:40 50:52 54:11 46:02 51:02 55:49 52:59 54:14 44:44 47:20 52:25 51:31 50:37 52:20 47:46 50:07 46:54 44:42 48:22 52:28

xxx

MACHINE DESIGN

-

An Integrated Approach -

TUTORIAL VIDEOS Topic Bearings Bending Stress Columns Failure Modes Gears Springs Stress Cube Torsion

DEMONSTRATION VIDEOS Topic Boot Testing Machine Bottle Printing Machine Cam Machine Fourbar Machine Pick and Place Mechanism Spring Manufacturing Machinery Spring Surge and Spring Failure Vibration Testing

Sixth Edition

(Concatenate this URL with any filename below to run a video)

http://www.designofmachinery.com/MD/

Run Time 09:11 05:57 01:52 09:41 22:07 20:06 05:04 03:13

Bearings.mp4 Bending_Stress.mp4 Columns.mp4 Failure_Modes.mp4 Gears.mp4 Springs.mp4 Stress_Cube.mp4 Torsion.mp4

(Concatenate this URL with any filename below to run a video)

http://www.designofmachinery.com/MD/ Boot_Tester.mp4 Bottle_Printing_Machine.mp4 Cam_Machine.mp4 Fourbar_Machine.mp4 Pick_and_Place_Mechanism.mp4 Spring_Manufacturing.mp4 Fatigue_Failure.mp4 Vibration_Testing.mp4

Run Time 19:02 09:49 21:28 35:38 36:35 12:23 03:46 05:51

Note that you can download a PDF file containing hyperlinks to all the video content listed in the above tables. This allows print-book readers to easily access the videos without having to type in each URL as noted in the tables. Download the file: http://www.designofmachinery.com/MD/Video_Links_for_Machine_Design_6ed.pdf

I

Part

FUNDAMENTALS

INTRODUCTION TO DESIGN Learning without thought is labor lost; thought without learning is perilous. ConfuCius, 6th Century B.C.

1.1

DESIGN View the introductory video (03:30)†

What is design? Wallpaper is designed. You may be wearing “designer” clothes. Automobiles are “designed” in terms of their external appearance. The term design clearly encompasses a wide range of meaning. In the above examples, design refers primarily to the object’s aesthetic appearance. In the case of the automobile, all of its other aspects also involve design. Its mechanical internals (engine, brakes, suspension, etc.) must be designed, more likely by engineers than by artists, though even the engineer gets to exhibit some artistry when designing machinery. The word design is from the Latin word designare meaning to designate, or mark out. Design means many things. It can refer to the design of an artistic work or the appearance of a product. We are more concerned here with engineering design than with artistic design. Engineering design can be defined as The process of applying the various techniques and scientific principles for the purpose of defining a device, a process, or a system in sufficient detail to permit its realization. Machine Design This text is concerned with one aspect of engineering design—machine design. Machine design deals with the creation of machinery that works safely, reliably, and well. A machine can be defined as: A system of elements arranged to transmit motion and energy in a predetermined and controlled fashion, or even more simply as: A system to control force and motion. †

http://www.designofmachinery.com/MD/01_Introduction.mp4

3

1

4

1

Title-page photograph courtesy of Boeing Commercial Airplane Co. Inc., Seattle, Wash.

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

The notion of useful work is basic to a machine’s function, as there is almost always some energy transfer involved. The mention of forces and motion is also critical to our concerns, as, in converting energy from one form to another, machines create motion and develop forces. It is the engineer’s task to define and calculate those motions, forces, and changes in energy in order to determine the sizes, shapes, and materials needed for each of the interrelated parts in the machine. This is the essence of machine design. While one must, of necessity, design a machine one part at a time, it is crucial to recognize that each part’s function and performance (and thus its design) are dependent on many other interrelated parts within the same machine. Thus, we are going to attempt to “design the whole machine” here, rather than simply designing individual elements in isolation from one another. To do this we must draw upon a common body of engineering knowledge encountered in previous courses, e.g., statics, dynamics, mechanics of materials (stress analysis), and material properties. Brief reviews and examples of these topics are included in the early chapters of this book. The ultimate goal in machine design is to size and shape the parts (machine elements) and choose appropriate materials and manufacturing processes so that the resulting machine can be expected to perform its intended function without failure. This requires that the engineer be able to calculate and predict the mode and conditions of failure for each element and then design it to prevent that failure. This in turn requires that a stress and deflection analysis be done for each part. Since stresses are a function of the applied and inertial loads, and of the part’s geometry, an analysis of the forces, moments, torques, and the dynamics of the system must be done before the stresses and deflections can be completely calculated. If the “machine” in question has no moving parts, then the design task becomes much simpler, because only a static force analysis is required. But if the machine has no moving parts, it is not much of a machine (and doesn’t meet the definition above); it is then a structure. Structures also need to be designed against failure, and, in fact, large external structures (bridges, buildings, etc.) are also subjected to dynamic loads from wind, earthquakes, traffic, etc., and thus must also be designed for these conditions. Structural dynamics is an interesting subject but one which we will not address in this text. We will concern ourselves with the problems associated with machines that move. If the machine’s motions are very slow and the accelerations negligible, then a static force analysis will suffice. But if the machine has significant accelerations within it, then a dynamic force analysis is needed and the accelerating parts become “victims of their own mass.” In a static structure, such as a building’s floor, designed to support a particular weight, the safety factor of the structure can be increased by adding appropriately distributed material to its structural parts. Though it will be heavier (more “dead” weight), if properly designed it may nevertheless carry more “live” weight (payload) than it did before, still without failure. In a dynamic machine, adding weight (mass) to moving parts may have the opposite effect, reducing the machine’s safety factor, its allowable speed, or its payload capacity. This is because some of the loading that creates stresses in the moving parts is due to the inertial forces predicted by Newton’s second law, F = ma. Since the accelerations of the moving parts in the machine are dictated by its kinematic design and by its running speed, adding mass to moving parts will increase the inertial loads on those same parts unless their kinematic accelerations are reduced by slowing its operation. Even though the added mass may increase the strength of the part, that benefit may be reduced or cancelled by the resultant increases in inertial forces.

Chapter 1

INTRODUCTION TO DESIGN

5

1

Iteration Thus, we face a dilemma at the initial stages of machine design. Generally, before reaching the stage of sizing the parts, the kinematic motions of the machine will have already been defined. External forces provided by the “outside world” on the machine are also often known. Note that in some cases, the external loads on the machine will be very difficult to predict—for example, the loads on a moving automobile. The designer cannot predict with accuracy what environmental loads the user will subject the machine to (potholes, hard cornering, etc.). In such cases, statistical analysis of empirical data gathered from actual testing can provide some information for design purposes. What remain to be defined are the inertial forces that will be generated by the known kinematic accelerations acting on the as yet undefined masses of the moving parts. The dilemma can be resolved only by iteration, which means to repeat, or to return to a previous state. We must assume some trial configuration for each part, use the mass properties (mass, CG location, and mass moment of inertia) of that trial configuration in a dynamic force analysis to determine the forces, moments, and torques acting on the part, and then use the cross-sectional geometry of the trial design to calculate the resulting stresses. In general, accurately determining all the loads on a machine is the most difficult task in the design process. If the loads are known, the stresses can be calculated. Most likely, on the first trial, we will find that our design fails because the materials cannot stand the levels of stress presented. We must then redesign the parts (iterate) by changing shapes, sizes, materials, manufacturing processes, or other factors in order to reach an acceptable design. It is generally not possible to achieve a successful result without making several iterations through this design process. Note also that a change to the mass of one part will also affect the forces applied to parts connected to it and thus require their redesign also. It is truly the design of interrelated parts. 1.2

A DESIGN PROCESS*

The process of design is essentially an exercise in applied creativity. Various “design processes” have been defined to help organize the attack upon the “unstructured problem,” i.e., one for which the problem definition is vague and for which many possible solutions exist. Some of these design process definitions contain only a few steps and others a detailed list of 25 steps. One version of a design process is shown in Table 1-1, which lists ten steps.[2] The initial step, Identification of Need, usually consists of an ill-defined and vague problem statement. The development of Background Research information (step 2) is necessary to fully define and understand the problem, after which it is possible to restate the Goal (step 3) in a more reasonable and realistic way than in the original problem statement. Step 4 calls for the creation of a detailed set of Task Specifications which bound the problem and limit its scope. The Synthesis step (5) is one in which as many alternative design approaches as possible are sought, usually without regard (at this stage) for their value or quality. This is also sometimes called the Ideation and Invention step, in which the largest possible number of creative solutions are generated. In step 6, the possible solutions from the previous step are Analyzed and either accepted, rejected, or modified. The most promising solution is Selected at step 7. Once an acceptable design is selected, the Detailed Design (step 8) can be done, in which all the loose ends are tied up, complete engineering drawings made, vendors identified,

*

Adapted from Norton, Design of Machinery, 5ed. McGrawHill, New York, 2012, with the publisher’s permission.

6

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

1 Table 1-1

A Design Process

1

Identification of need

2

Background research

3

Goal statement

4

Task specifications

5

Synthesis

6

Analysis

7

Selection

8

Detailed design

9

Prototyping and testing

10

Production

manufacturing specifications defined, etc. The actual construction of the working design is first done as a Prototype in step 9 and finally in quantity in Production at step 10. A more complete discussion of this design process can be found in reference 2, and a number of references on the topics of creativity and design are provided in the bibliography at the end of this chapter. The above description may give an erroneous impression that this process can be accomplished in a linear fashion as listed. On the contrary, iteration is required within the entire process, moving from any step back to any previous step, in all possible combinations, and doing this repeatedly. The best ideas generated at step 5 will invariably be discovered to be flawed when later analyzed. Thus, a return to at least the Ideation step will be necessary in order to generate more solutions. Perhaps a return to the Background Research phase may be necessary to gather more information. The Task Specifications may need to be revised if it turns out that they were unrealistic. In other words, anything is “fair game” in the design process, including a redefinition of the problem, if necessary. One cannot design in a linear fashion. It’s three steps forward and two (or more) back, until you finally emerge with a working solution. Theoretically, we could continue this iteration on a given design problem forever, continually creating small improvements. Inevitably, the incremental gains in function or reductions in cost will tend toward zero with time. At some point, we must declare the design “good enough” and ship it. Often someone else (most likely, the boss) will snatch it from our grasp and ship it over our protests that it isn’t yet “perfect.” Machines that have been around a long time and that have been improved by many designers reach a level of “perfection” that makes them difficult to improve upon. One example is the ordinary bicycle. Though inventors continue to attempt to improve this machine, the basic design has become fairly static after more than a century of development. In machine design, the early design-process steps usually involve the Type Synthesis of suitable kinematic configurations, which can provide the necessary motions. Type synthesis involves the choice of the type of mechanism best suited to the problem. This is a difficult task for the student, as it requires some experience and knowledge of the various types of mechanisms that exist and that might be feasible from a performance and manufacturing standpoint. As an example, assume that the task is to design a device

Chapter 1

INTRODUCTION TO DESIGN

to track the constant-speed, straight-line motion of a part on a conveyor belt and attach a second part to it as it passes by. This has to be done with good accuracy and repeatability and must be reliable and inexpensive. You might not be aware that this task could be accomplished by any of the following devices: • a straight-line linkage • a cam and follower • an air cylinder • a hydraulic cylinder • a robot • a solenoid

Each of these solutions, while possible, may not be optimal or even practical. Each has good and bad points. The straight-line linkage is large and may have undesirable accelerations; the cam and follower is expensive but is accurate and repeatable. The air cylinder is inexpensive but noisy and unreliable. The hydraulic cylinder and the robot are more expensive. The inexpensive solenoid has high impact loads and velocities. So, the choice of device type can have a big effect on design quality. A bad choice at the type-synthesis stage can create major problems later on. The design might have to be changed after completion at great expense. Design is essentially an exercise in trade-offs. There is usually no clear-cut solution to a real engineering design problem. Once the type of required mechanism is defined, its detailed kinematics must be synthesized and analyzed. The motions of all moving parts and their time derivatives through acceleration must be calculated in order to be able to determine the dynamic forces on the system. (See reference 2 for more information on this aspect of machine design.) In the context of machine design addressed in this text, we will not exercise the entire design process as described in Table 1-1. Rather, we will propose examples, problems, and case studies that already have had steps 1–4 defined. The type synthesis and kinematic analysis will already be done, or at least set up, and the problems will be structured to that degree. The tasks remaining will largely involve steps 5 through 8, with a concentration on synthesis (step 5) and analysis (step 6). Synthesis and analysis are the “two faces” of machine design, like two sides of the same coin. Synthesis means to put together and analysis means to decompose, to take apart, to resolve into its constituent parts. Thus, they are opposites, but they are symbiotic. We cannot take apart “nothing,” thus we must first synthesize something in order to analyze it. When we analyze it, we will probably find it lacking, requiring further synthesis, and then further analysis ad nauseam, finally iterating to a better solution. You will need to draw heavily upon your understanding of statics, dynamics, and mechanics of materials to accomplish this. 1.3

PROBLEM FORMULATION AND CALCULATION

It is extremely important for every engineer to develop good and careful computational habits. Solving complicated problems requires an organized approach. Design problems also require good record-keeping and documentation habits in order to record the many assumptions and design decisions made along the way so that the designer’s thought process can be later reconstructed if redesign is necessary.

7

1

8

MACHINE DESIGN

*

A suggested procedure for the designer is shown in Table 1-2, which lists a set of subtasks appropriate to most machine-design problems of this type. These steps should be documented for each problem in a neat fashion, preferably in a bound notebook in order to maintain their chronological order.*

1

If there is a possibility of a patentable invention resulting from the design, then the notebook should be permanently bound (not loose-leaf), and its pages should be consecutively numbered, dated, and witnessed by someone who understands the technical content.

-

An Integrated Approach

-

Sixth Edition

Definition Stage In your design notebook, first Define the Problem clearly in a concise statement. The “givens” for the particular task should be clearly listed, followed by a record of the assumptions made by the designer about the problem. Assumptions expand upon the given (known) information to further constrain the problem. For example, one might assume the effects of friction to be negligible in a particular case, or assume that the weight of the part can be ignored because it will be small compared to the applied or dynamic loads expected. Preliminary Design Stage Once the general constraints are defined, some Preliminary Design Decisions must be made in order to proceed. The reasons and justifications for these decisions should be documented. For example, we might decide to try a solid, rectangular cross section for a connecting link and choose aluminum as a trial material. On the other hand, if we recognized from our understanding of the problem that this link would be subjected to significant accelerations of a time-varying nature that would repeat for millions of cycles, a better design decision might be to use a hollow or I-beam section in order to reduce its mass and also to choose steel for its infinite fatigue life. Thus, these design decisions can have significant effect on the results and will often have to be changed or abandoned as we iterate through the design process. It has often been noted that 90% of a design’s characteristics may be determined in the first 10% of the total project time, during which these preliminary design decisions are made. If they are bad decisions, it may not be possible to save the bad design through later modifications without essentially starting over. The preliminary design concept should be documented at this stage with clearly drawn and labeled Design Sketches that will be understandable to another engineer or even to oneself after some time has passed. Detailed Design Stage With a tentative design direction established we can create one or more engineering (mathematical) models of the element or system in order to analyze it. These models will usually include a loading model consisting of free-body diagrams which show all forces, moments, and torques on the element or system and the appropriate equations for their calculation. Models of the stress and deflection states expected at locations of anticipated failure are then defined with appropriate stress and deflection equations. Analysis of the design is then done using these models and the safety or failure of the design determined. The results are evaluated in conjunction with the properties of the chosen engineering materials and a decision made whether to proceed with this design or iterate to a better solution by returning to an earlier step of the process.

Chapter 1

INTRODUCTION TO DESIGN

9

1 Table 1-2

Problem Formulation and Calculation

1

Define the problem

2

State the givens

3

Make appropriate assumptions

4

Preliminary design decisions

5

Design sketches

6

Mathematical models

7

Analysis of the design

8

Evaluation

9

Document results

Definition stage

Preliminary design stage

Detailed design stage Documentation stage

Documentation Stage Once sufficient iteration through this process provides satisfactory results, the documentation of the element’s or system’s design should be completed in the form of detailed engineering drawings, material and manufacturing specifications, etc. If properly approached, a great deal of the documentation task can be accomplished concurrent with the earlier stages simply by keeping accurate and neat records of all assumptions, computations, and design decisions made throughout the process. 1.4

THE ENGINEERING MODEL

The success of any design is highly dependent on the validity and appropriateness of the engineering models used to predict and analyze its behavior in advance of building any hardware. Creating a useful engineering model of a design is probably the most difficult and challenging part of the whole process. Its success depends a great deal on experience as well as skill. Most important is a thorough understanding of the first principles and fundamentals of engineering. The engineering model that we are describing here is an amorphous thing that may consist of some sketches of the geometric configuration and some equations that describe its behavior. It is a mathematical model that describes the physical behavior of the system. This engineering model invariably requires the use of computers to exercise it. Using computer tools for analyzing engineering models is discussed in the next section. A physical model or prototype usually comes later in the process and is needed to prove the validity of the engineering model through experiments. Estimation and First-Order Analysis The value of making even very simplistic engineering models of your preliminary designs cannot be overemphasized. Often, at the outset of a design, the problem is so loosely and poorly defined that it is difficult to develop a comprehensive and thorough model in the form of equations that fully describe the system. The engineering student is used to problems that are fully structured, of a form such as “Given A, B, and C, find D.” If one can identify the appropriate equations (model) to apply to such a problem, it is relatively easy to determine an answer (which might even match the one in the back of the book).

10

MACHINE DESIGN

*

Real-life engineering design problems are not of this type. They are very unstructured and must be structured by you before they can be solved. Also, there is no “back of the book” to refer to for the answer.* This situation makes most students and beginning engineers very nervous. They face the “blank paper syndrome,” not knowing where to begin. A useful strategy is to recognize that:

1 A student once commented that “Life is an odd-numbered problem.” This (slow) author had to ask for an explanation, which was: “The answer is not in the back of the book.”

-

An Integrated Approach

-

Sixth Edition

1. You must begin somewhere. 2. Wherever you begin, it will probably not be the “best” place to do so. 3. The magic of iteration will allow you to back up, improve your design, and eventually succeed.

With this strategy in mind, you can feel free to make some estimation of a design configuration at the outset, assume whatever limiting conditions you think appropriate, and do a “first-order analysis,” one that will be only an estimate of the system’s behavior. These results will allow you to identify ways to improve the design. Remember that it is preferable to get a reasonably approximate but quick answer that tells you whether the design does or doesn’t work rather than to spend more time getting the same result to more decimal places. With each succeeding iteration, you will improve your understanding of the problem, the accuracy of your assumptions, the complexity of your model, and the quality of your design decisions. Eventually, you will be able to refine your model to include all relevant factors (or identify them as irrelevant) and obtain a higher-order, final analysis in which you have more confidence. The Engineering Sketch A sketch of the concept is often the starting point for a design. This may be a freehand sketch, but it should always be made reasonably to scale in order to show realistic geometric proportions. This sketch often serves the primary purpose of communicating the concept to other engineers and even to yourself. It is one thing to have a vague concept in mind and quite another to define it in a sketch. This sketch should, at a minimum, contain three or more orthographic views, aligned according to proper drafting convention, and may also include an isometric or trimetric view. Figure 1-1 shows a freehand sketch of a simple design for one subassembly of a trailer hitch for a tractor. While often incomplete in terms of detail needed for manufacture, the engineering sketch should contain enough information to allow the development of an engineering model for design and analysis. This may include critical, if approximate, dimensional information, some material assumptions, and any other data germane to its function that is needed for further analysis. The engineering sketch captures some of the givens and assumptions made, even implicitly, at the outset of the design process. 1.5

COMPUTER-AIDED DESIGN AND ENGINEERING

The computer has created a true revolution in engineering design and analysis. Problems whose solution methods have literally been known for centuries but that only a generation ago were practically unsolvable due to their high computational demands can now be solved in minutes on inexpensive microcomputers. Tedious graphical solution methods were developed in the past to circumvent the lack of computational power available from slide rules. Some of these graphical solution methods still have value in that they can

Chapter 1

INTRODUCTION TO DESIGN

11

1

FIGURE 1-1 A Freehand Sketch of a Trailer Hitch Assembly for a Tractor

show the results in an understandable form. But one can no longer “do engineering” without using its latest and most powerful tool, the computer. Computer-Aided Design (CAD) As the design progresses, the crude freehand sketches made at the earliest stages will be supplanted by formal drawings made either with conventional drafting equipment or, as is increasingly common, with computer-aided design or computer-aided drafting software. If the distinction between these two terms (both of which share the acronym CAD) was ever clear (a subject for debate that will be avoided here), then that distinction is fading as more sophisticated CAD software becomes available. The original CAD systems of a generation ago were essentially drafting tools that allowed the creation of computergenerated multiview drawings similar to those done for centuries before by hand on a drafting board. The data stored in these early CAD systems were strictly two-dimensional representations of the orthographic projections of the part’s true 3-D geometry. Only the edges of the part were defined in the database. This is called a wireframe model. Some 3-D CAD packages use wireframe representation as well. Present versions of most CAD software packages allow (and sometimes require) that the geometry of the parts be encoded in a 3-D data base as solid models. In a solid model the edges and the faces of the part are defined. From this 3-D information, the conventional 2-D orthographic views can be automatically generated if desired. The major advantage of creating a 3-D solid-model geometric data base for any design is that its mass-property information can be rapidly calculated. (This is not possible in a 2-D or 3-D wireframe model.) For example, in designing a machine part, we need to determine the location of its center of gravity (CG), its mass, its mass moment of inertia, and its cross-sectional geometries at various locations. Determining this information from a 2-D model must be done outside the CAD package. That is tedious to do and can only be approximate when the geometry is complex. But, if the part is designed in a solid modeling CAD system such as ProEngineer,[7] NX,[4] or one of many others, the mass properties can be calculated for the most complicated part geometries.

12

1

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Solid modeling systems usually provide an interface to one or more Finite Element Analysis (FEA) programs and allow direct transfer of the model’s geometry to the FEA package for stress, vibration, and heat transfer analysis. Some CAD systems include a mesh-generation feature that creates the FEA mesh automatically before sending the data to the FEA software. This combination of tools provides an extremely powerful means to obtain superior designs whose stresses are more accurately known than would be possible by conventional analysis techniques when the geometry is complex. While it is highly likely that the students reading this textbook will be using CAD tools including finite element or boundary element analysis (BEA) methods in their professional practice, it is still necessary that the fundamentals of applied stress analysis be thoroughly understood. That is the purpose of this text. FEA techniques will be discussed in Chapters 4 and 8 but will not be emphasized in this text. Rather, we will concentrate on the classical stress-analysis techniques in order to lay the foundation for a thorough understanding of the fundamentals and their application to machine design. FEA and BEA methods are rapidly becoming the methods of choice for the solution of complicated stress-analysis problems. However, there is danger in using those techniques without a solid understanding of the theory behind them. These methods will always give some results. Unfortunately, those results can be incorrect if the problem was not well formulated and well meshed with proper boundary conditions applied. Being able to recognize incorrect results from a computer-aided solution is extremely important to the success of any design. Chapter 8 provides a brief introduction to FEA. The student should take courses in FEA and BEA to become familiar with these tools. Figure 1-2 shows a solid model of the ball bracket from Figure 1-1 that was created in a CAD software package. The shaded, isometric view in the upper right corner shows that the solid volume of the part is defined. The other three views show orthographic projections of the part. Figure 1-3 shows the mass-properties data that are calculated by the software. Figure 1-4 shows a wireframe rendering of the same part generated from the solid geometry data base. A wireframe version is used principally to speed up the screen-drawing time when working on the model. There is much less wireframe display information to calculate than for the solid rendering of Figure 1-2. Figure 1-5 shows a fully dimensioned, orthographic, multiview drawing of the ball bracket that was generated in the CAD software package. Another major advantage of creating a solid model of a part is that the dimensional and tool-path information needed for its manufacture can be generated in the CAD system and sent over a network to a computer-controlled machine on the manufacturing floor. This feature allows the production of parts without the need for paper drawings such as Figure 1-5. Figure 1-6 shows the same part after a finite element mesh was applied to it by the CAD software before sending it to the FEA software for stress analysis. Computer-Aided Engineering (CAE) The techniques generally referred to above as CAD are a subset of the more general topic of computer-aided engineering (CAE), which term implies that more than just the geometry of the parts is being dealt with. However, the distinctions between CAD and CAE continue to blur as more sophisticated software packages become available. In fact, the description of the use of a solid modeling CAD system and an FEA package together as

Chapter 1

INTRODUCTION TO DESIGN

13

1

FIGURE 1-2 A CAD Solid Model of the Ball Bracket from the Trailer Hitch Assembly of Figure 1-1

described in the previous section is an example of CAE. When some analysis of forces, stresses, deflections, or other aspects of the physical behavior of the design is included, with or without the solid geometry aspects, the process is called CAE. Many commercial software packages do one or more aspects of CAE. The FEA and BEA software packages mentioned previously are in this category. See Chapter 8 for more information on FEA. Dynamic force simulations of mechanisms can be done with such packages as ADAMS[5] and Working Model.[6] Some software packages such as ProEngineer,[7] Solidworks,[12] NX,[4] and others combine aspects of a CAD system with general analysis capabilities.

FIGURE 1-3 Mass Properties of the Ball Bracket Calculated Within the CAD System from its Solid Model

14

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

1

FIGURE 1-4 A Wireframe Representation of the Ball Bracket Generated from its Solid Model in a CAD System

These constraint-based programs allow constraints to be applied to the design which can control the part geometry as the design parameters are varied. Other classes of tools for CAE are equation solvers such as MATLAB[11], Mathcad,[9] TK Solver,[8] and spreadsheets such as Excel.[10] These are general-purpose tools that will allow any combination of equations to be encoded in a convenient form and then will manipulate the equation set (i.e., the engineering model) for different trial data and conveniently display tabular and graphic output. Equation solvers are invaluable for the solution of force, stress, and deflection equations in machine-design problems because they allow rapid “what-if” calculations to be done. The effects of dimensional or material

26.0 D 64 R

32 38.0

70 19 70

50

10.0 D TYP. 2 PL.

13 R 90° ± 2°

13

Widgets Perfected Inc. Title: BALL BRACKET

AS-MILLED FINISH

Mat'l: AISI 1020 Hot-Rolled Steel All dims in: mm By: RLN 4/12/04 xx. ± 1 xx.x ± 0.5 Dwg #: B–579328

FIGURE 1-5 A Dimensioned, 3-View Orthographic Drawing Done in a 2-D CAD Drawing Package

Chapter 1

INTRODUCTION TO DESIGN

15

1

FIGURE 1-6 An FEA Mesh Applied to the Solid Model of the Ball Bracket in the CAD System

changes on the stresses and deflections in the part can be seen instantly. In the absence of a true solid modeling system, an equation solver also can be used to approximate the part’s mass properties while iterating the geometry and material properties of trial part designs. Rapid iteration to an acceptable solution is thus enhanced. The website accompanying the text (see Preface for information on how to access this) contains a large number of models for various equation solvers that support the examples and case studies presented in the text. Introductions to the use of TK Solver and Mathcad along with examples of their use are provided as PDF files on the book’s website. In addition, some custom-written computer programs, Mohr, Contact, Linkages, Dynacam, and Matrix are provided on the book’s website to aid in the calculation of dynamic loads and stresses when solving the open-ended design problems assigned. However, one must be aware that these computer tools are just tools and are not a substitute for the human brain. Without a thorough understanding of the engineering fundamentals on the part of the user, the computer will not give good results. Garbage in, garbage out. Caveat Lector.* Computational Accuracy Computers and calculators make it very easy to obtain numerical answers having many significant figures. Before writing down all those digits, you are advised to recall the accuracy of your initial assumptions and given data. If, for example, your applied loads were known to only two significant figures, it is incorrect and misleading to express the calculated stresses to more significant figures than your input data possessed. However, it is valid and appropriate to make all intermediate calculations to the greatest accuracy available in your computational tools. This will minimize computational round-off errors. But, when done, round off the results to a level consistent with your known or assumed data.

* Reader beware.

1

16

MACHINE DESIGN

*

1.6

Excerpted from Norton, Design of Machinery, 5ed McGrawHill, New York, 2012, with the publisher’s permission.

-

An Integrated Approach

-

Sixth Edition

THE ENGINEERING REPORT*

Communication of your ideas and results is a very important aspect of engineering. Many engineering students picture themselves in professional practice spending most of their time doing calculations of a nature similar to those they have done as students. Fortunately, this is seldom the case, as it would be very boring. Actually, engineers spend a large percentage of their time communicating with others, either orally or in writing. Engineers write proposals and technical reports, give presentations, and interact with support personnel. When your design is done, it is usually necessary to present the results to your client, peers, or employer. The usual form of presentation is a formal engineering report. In addition to a written description of the design, these reports will usually contain engineering drawings or sketches as described earlier, as well as tables and graphs of data calculated from the engineering model. It is very important for engineering students to develop their communication skills. You may be the cleverest person in the world, but no one will know that if you cannot communicate your ideas clearly and concisely. In fact, if you cannot explain what you have done, you probably don’t understand it yourself. The design-project assignments in Chapter 9 are intended to be written up in formal engineering reports to give you some experience in this important skill of technical communication. Information on writing engineering reports can be found in the suggested readings listed in the bibliography. 1.7

FACTORS OF SAFETY AND DESIGN CODES

The quality of a design can be measured by many criteria. It is always necessary to calculate one or more factors of safety to estimate the likelihood of failure. There may be legislated, or generally accepted, design codes that must be adhered to as well. †

Also called safety factor. We will use both terms interchangeably in this text.

Factor of Safety† A factor of safety or safety factor can be expressed in many ways. It is typically a ratio of two quantities that have the same units, such as strength/stress, critical load/applied load, load to fail part/expected service overload, maximum cycles/applied cycles, or maximum safe speed/operating speed. A safety factor is always unitless. The form of expression for a safety factor can usually be chosen based on the character of loading on the part. For example, consider the loading on the side wall of a cylindrical water tower that can never be “more than full” of a liquid of known density within a known temperature range. Since this loading is highly predictable over time, a comparison of the strength of the material to the stress in the wall of a full tank might be appropriate as a safety factor. Note in this example that the possibility of rust reducing the thickness of the wall over time must be considered. (See Section 4.17 for a discussion of stresses in cylinder walls and Section 7.6 for a discussion of corrosion.) If this cylindrical water tower is standing on legs loaded as columns, then a safety factor for the legs based on a ratio of the column’s critical buckling load over the applied load from a full water tower would be appropriate. (See Section 4.16 for a discussion of column buckling.)

Chapter 1

INTRODUCTION TO DESIGN

If a part is subjected to loading that varies cyclically with time, it may experience fatigue failure. The resistance of a material to some types of fatigue loading can be expressed as a maximum number of cycles of stress reversal at a given stress level. In such cases, it may be appropriate to express the safety factor as a ratio of the maximum number of cycles to expected material failure over the number of cycles applied to the part in service for its desired life. (See Chapter 6 for a discussion of fatigue-failure phenomena and several approaches to the calculation of safety factors in such situations.) The safety factor of a part such as a rotating sheave (pulley) or flywheel is often expressed as a ratio of its maximum safe speed over the highest expected speed in service. In general, if the stresses in the parts are a linear function of the applied service loads and those loads are predictable, then a safety factor expressed as strength/stress or failure load/applied load will give the same result. Not all situations fit these criteria. Some require a nonlinear ratio. A column is one example, because its stresses are a nonlinear function of the loading (see Section 4.16). Thus, a critical (failure) load for the particular column must be calculated for comparison to the applied load. Another complicating factor is introduced when the magnitudes of the expected applied loads are not accurately predictable. This can be true in virtually any application in which the use (and thus the loading) of the part or device is controlled by humans. For example, there is really no way to prevent someone from attempting to lift a 10-ton truck with a jack designed to lift a 2-ton automobile. When the jack fails, the manufacturer (and designer) may be blamed even though the failure was probably due more to the “nut behind the jack handle.” In situations where the user may subject the device to overloading conditions, an assumed overload may have to be used to calculate a safety factor based on a ratio of the load that causes failure over the assumed service overload. Labels warning against inappropriate use may be needed in these situations as well. Since there may be more than one potential mode of failure for any machine element, it can have more than one value of safety factor N. The smallest value of N for any part is of greatest concern, since it predicts the most likely mode of failure. When N becomes reduced to 1, the stress in the part is equal to the strength of the material (or the applied load is equal to the load that fails it, etc.) and failure occurs. Therefore, we desire N to be always greater than 1. Choosing a Safety Factor Choosing a safety factor is often a confusing proposition for the beginning designer. The safety factor can be thought of as a measure of the designer’s uncertainty in the analytical models, failure theories, and material-property data used, and should be chosen accordingly. How much greater than one N must be depends on many factors, including our level of confidence in the model on which the calculations are based, our knowledge of the range of possible in-service loading conditions, and our confidence in the available material-strength information. If we have done extensive testing on physical prototypes of our design to prove the validity of our engineering model and of the design, and have generated test data on the particular material’s strengths, then we can afford to use a smaller safety factor. If our model is less well proven or the material-property information is less reliable, a larger N is in order. In the absence of any design codes that may specify N for particular cases, the choice of factor of safety involves engineering judgment. A reasonable approach is to determine the largest loads expected in service (including possible overloads) and the minimum expected material strengths and base

17

1

18

1

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

the safety factors on these data. The safety factor then becomes a reasonable measure of uncertainty. If you fly, it may not give you great comfort to know that safety factors for commercial aircraft are in the range of 1.2 to 1.5. Military aircraft can have N < 1.1, but their crews wear parachutes. (Test pilots deserve their high salaries.) Missiles have N = 1 but have no crew and aren’t expected to return anyway. These small factors of safety in aircraft are necessary to keep weight low and are justified by sophisticated analytical modeling (usually involving FEA), testing of the actual materials used, extensive testing of prototype designs, and rigorous in-service inspections for incipient failures of the equipment. The opening photograph of this chapter shows an elaborate test rig used by the Boeing Aircraft Co. to mechanically test the airframe of full-scale prototype or production aircraft by applying dynamic forces and measuring their effects. It can be difficult to predict the kinds of loads that an assembly will experience in service, especially if those loads are under the control of the end-user, or Mother Nature. For example, what loads will the wheel and frame of a bicycle experience? It depends greatly on the age, weight, and recklessness of the rider, whether used on- or off-road, etc. The same problem of load uncertainty exists with all transportation equipment, ships, aircraft, automobiles, etc. Manufacturers of these devices engage in extensive test programs to measure typical service loads. See Figures 3-41 and 6-7 for examples of such service-load data. Some guidelines for the choice of a safety factor in machine design can be defined based on the quality and appropriateness of the material-property data available, the expected environmental conditions compared to those under which the material test data were obtained, and the accuracy of the loading- and stress-analysis models developed for the analyses. Table 1-3 shows a set of factors for ductile materials which can be chosen in each of the three categories listed based on the designer’s knowledge or judgment of the quality of information used. The overall safety factor is then taken as the largest of the three factors chosen. Given the uncertainties involved, a safety factor typically should not be taken to more than 1 decimal place accuracy. N ductile ≅ MAX ( F1, F 2, F 3)

(1.1a)

The ductility or brittleness of the material is also a concern. Brittle materials are designed against the ultimate strength, so failure means fracture. Ductile materials under static loads are designed against the yield strength and are expected to give some visible warning of failure before fracture, unless cracks indicate the possibility of a fracturemechanics failure (see Sections 5-3 and 6-5). For these reasons, the safety factor for brittle materials is often made twice that which would be used for a ductile material in the same situation: N brittle ≅ 2 * MAX ( F1, F 2, F 3)

(1.1b)

This method of determining a safety factor is only a guideline to obtain a starting point and is obviously subject to the judgment of the designer in selecting factors in each category. The designer has the ultimate responsibility to ensure that the design is safe. A larger safety factor than any shown in Table 1-3 may be appropriate in some circumstances.

Chapter 1

INTRODUCTION TO DESIGN

19

1 Table 1-3

Factors Used to Determine a Safety Factor for Ductile Materials

Information

Quality of Information

Factor

The actual material used was tested

F1 1.3

Material-property data

Representative material test data are available

2

available from tests

Fairly representative material test data are available

3

Poorly representative material test data are available Are identical to material test conditions

5+ F2 1.3

Environmental conditions

Essentially room-ambient environment

2

in which it will be used

Moderately challenging environment

3

Extremely challenging environment Models have been tested against experiments

5+ F3 1.3

Analytical models for

Models accurately represent system

2

loading and stress

Models approximately represent system

3

Models are crude approximations

5+

Design and Safety Codes Many engineering societies and government agencies have developed codes for specific areas of engineering design. Most are only recommendations, but some have the force of law. The ASME provides recommended guidelines for safety factors to be used in particular applications such as steam boilers and pressure vessels. Building codes are legislated in most U.S. states and cities and usually deal with publicly accessible structures or their components, such as elevators and escalators. Safety factors are sometimes specified in these codes and may be quite high. (The code for escalators in one state called for a factor of safety of 14.) Clearly, where human safety is involved, high values of N are justified. However, they come with a weight and cost penalty, as parts must often be made heavier to achieve large values of N. The design engineer must always be aware of these codes and standards and adhere to them where applicable. The following is a partial list of engineering societies and governmental, industrial, and international organizations that publish standards and codes of potential interest to the mechanical engineer. Addresses and data on their publications can be obtained in any technical library or from the Internet. American Gear Manufacturers Association (AGMA) http://www.agma.org/ American Institute of Steel Construction (AISC) http://www.aisc.org/ American Iron and Steel Institute (AISI) http://www.steel.org/ American National Standards Institute (ANSI) http://www.ansi.org/ American Society for Metals (ASM International) http://www.asm-intl.org/ American Society of Mechanical Engineers (ASME) http://www.asme.org/

20

MACHINE DESIGN

1

-

An Integrated Approach

-

Sixth Edition

American Society of Testing and Materials (ASTM) http://www.astm.org/ American Welding Society (AWS) http://www.aws.org/ American Bearing Manufacturers Association (ABMA) www.abma-dc.org/ International Standards Organization (ISO) http://www.iso.ch/iso/en National Institute for Standards and Technology (NIST)* http://www.nist.gov/

*

Formerly the National Bureau of Standards (NBS).

Society of Automotive Engineers (SAE) http://www.sae.org/ Society of Plastics Engineers (SPE) http://www.4spe.org/ Underwriters Laboratories (UL) http://www.ul.com/

1.8

STATISTICAL CONSIDERATIONS

Nothing is absolute in engineering any more than in any other endeavor. The strengths of materials will vary from sample to sample. The actual size of different examples of the “same” part made in quantity will vary due to manufacturing tolerances. As a result, we should take the statistical distributions of these properties into account in our calculations. The published data on the strengths of materials may be stated either as minimum values or as average values of tests made on many specimens. If it is an average value, there is a 50% chance that a randomly chosen sample of that material will be weaker or stronger than the published average value. To guard against failure, we can reduce the material-strength value that we will use in our calculations to a level that will include a larger percentage of the population. To do this requires some understanding of statistical phenomena and their calculation. All engineers should have this understanding and should include a statistics course in their curriculum. We briefly discuss some of the fundamental aspects of statistics in Chapter 2. †

Excerpted from Norton, Design of Machinery, 3ed, 2004, McGraw-Hill, New York, with the publisher’s permission.

1.9

UNITS†

Several different systems of units are used in engineering. The most common in the United States are the U.S. foot-pound-second system (fps), the U.S. inch-pound-second system (ips), and the Systeme Internationale (SI). The metric centimeter, gram, second (cgs) system is being used more frequently in the United States, particularly in international companies, e.g., the automotive industry. All systems are created from the choice of three of the quantities in the general expression of Newton’s second law: F=

mL t2

(1.2a)

where F is force, m is mass, L is length, and t is time. The units for any three of these variables can be chosen and the other is then derived in terms of the chosen units. The three chosen units are called base units, and the remaining one is a derived unit. Most of the confusion that surrounds the conversion of computations between either one of the U.S. systems and the SI system is due to the fact that the SI system uses a different set of base units than the U.S. systems. Both U.S. systems choose force, length, and time as the base units. Mass is then a derived unit in the U.S. systems, which are referred to as gravitational systems because the value of mass is dependent on the local gravitational constant. The SI system chooses mass, length, and time as the base units,

Chapter 1

INTRODUCTION TO DESIGN

21

1

and force is the derived unit. SI is then referred to as an absolute system, since the mass is a base unit whose value is not dependent on local gravity. The U.S. foot-pound-second (fps) system requires that all lengths be measured in feet (ft), forces in pounds (lb), and time in seconds (sec). Mass is then derived from Newton’s law as m=

Ft 2 L

(1.2b)

and its units are pounds seconds squared per foot (lb sec2/ft) = slugs. The U.S. inch-pound-second (ips) system requires that all lengths be measured in inches (in), forces in pounds (lb), and time in seconds (sec). Mass is still derived from Newton’s law, equation 1.2b, but the units are now pounds seconds squared per inch (lb sec2/in) = blobs†



This mass unit is not slugs! It is worth twelve slugs or one “blob”! Weight is defined as the force exerted on an object by gravity. Probably the most common units error that students make is to mix up these two unit systems (fps and ips) when converting weight units (which are pounds force) to mass units. Note that the gravitational acceleration constant (g or gc) on earth at sea level is approximately 32.17 feet per second squared, which is equivalent to 386 inches per second squared. The relationship between mass and weight is mass = weight / gravitational acceleration m=

W gc

(1.3)

It should be obvious that if you measure all your lengths in inches and then use g = gc = 32.17 feet/sec2 to compute mass, you will have an error of a factor of 12 in your results. This is a serious error, large enough to crash the airplane you designed. Even worse is the student who neglects to convert weight to mass at all. The results of this calculation will have an error of either 32 or 386, which is enough to sink the ship!* The value of mass is needed in Newton’s second-law equation to determine forces due to accelerations: F = ma

(1.4 a)

The units of mass in this equation are either g, kg, slugs, or blobs depending on the units system used. Thus, in either English system, the weight W (lbf) must be divided by the acceleration due to gravity gc as indicated in equation 1.3 to get the proper mass quantity for equation 1.4a. Adding further to the confusion is the common use of the unit of pounds mass (lbm). This unit, often used in fluid dynamics and thermodynamics, comes about through the use of a slightly different form of Newton’s equation: F=

ma gc

(1.4 b)

It is unfortunate that the mass unit in the ips system has never officially been given a name such as the term slug used for mass in the fps system. The author boldly suggests (with tongue only slightly in cheek) that this unit of mass in the ips system be called a blob (bl) to distinguish it more clearly from the slug (sl), and to help the student avoid some of the common errors listed below. Twelve slugs = one blob. Blob does not sound any sillier than slug, is easy to remember, implies mass, and has a convenient abbreviation (bl), which is an anagram for the abbreviation for pound (lb). Besides, if you have ever seen a garden slug, you know it looks just like a “little blob.”

22

1

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

where m = mass in lbm, a = acceleration, and gc = the gravitational constant. On earth, the value of the mass of an object measured in pounds mass (lbm) is numerically equal to its weight in pounds force (lbf). However, the student must remember to divide the value of m in lbm by gc when using this form of Newton’s equation. Thus, the lbm will be divided either by 32.17 or by 386 when calculating the dynamic force. The result will be the same as when the mass is expressed in either slugs or blobs in the F = ma form of the equation. Remember that in round numbers at sea level on earth 1 lbm = 1 lbf

1 slug = 32.17 lbf

1 blob = 386 lbf

The SI system requires that lengths be measured in meters (m), mass in kilograms (kg), and time in seconds (sec). This is sometimes also referred to as the mks system. Force is derived from Newton’s law and the units are: kg m/sec2 = newtons * A 125-million-dollar space probe was lost because NASA failed to convert data that had been supplied in ips units by its contractor, Lockheed Aerospace, into the metric units used in the NASA computer programs that controlled the spacecraft. It was supposed to orbit the planet Mars, but instead either burned up in the Martian atmosphere or crashed into the planet because of this units error. Source: The Boston Globe, October 1, 1999, p. 1. †

A valuable resource for information on the proper use of SI units can be found at the U.S. Government NIST site at https:// physics.nist.gov/cuu/Units Another excellent resource on the proper use of metric units in machine design can be found in a pamphlet “Metric Is Simple,” published and distributed by the fastener company Bossard International Inc., 235 Heritage Avenue, Portsmouth, NH 03801 http://www.bossard.com/

In the SI system there are distinct names for mass and force, which helps alleviate confusion.* When converting between SI and U.S. systems, be alert to the fact that mass converts from kilograms (kg) to either slugs (sl) or blobs (bl), and force converts from newtons (N) to pounds (lb). The gravitational constant (gc) in the SI system is approximately 9.81 m/sec2. The cgs system requires that lengths be measured in centimeters (cm), mass in grams (g), and time in seconds (sec). Force is measured in dynes. The SI system is generally preferred over the cgs system. The systems of units used in this textbook are the U.S. ips system and the SI system. Much of machine design in the United States is still done in the ips system, though the SI system is becoming more common.† Table 1-4 shows some of the variables used in this text and their units. Table 1-5 shows a number of conversion factors between commonly used units. The student is cautioned always to check the units in any equation written for a problem solution, whether in school or in professional practice. If properly written, an equation should cancel all units across the equal sign. If it does not, then you can be absolutely sure it is incorrect. Unfortunately, a unit balance in an equation does not guarantee that it is correct, as many other errors are possible. Always double-check your results. You might save a life.

EXAMPLE 1-1 Units Conversion Problem

The weight of an automobile is known in lbf. Convert it to mass units in the SI, cgs, fps, and ips systems. Also convert it to lbm.

Given

The weight = 4500 lbf.

Assumptions

The automobile is on earth at sea level.

Solution

1 Equation 1.4a is valid for the first four systems listed. For the fps system:

Chapter 1

INTRODUCTION TO DESIGN

23

1 Table 1-4

Variables and Units Base Units in Boldface—Abbreviations in ( )

Variable

Symbol

ips unit

Force

F

pounds (lb)

pounds (lb)

newtons (N)

Length

l

inches (in)

feet (ft)

meters (m)

Time

t

seconds (sec)

seconds (sec)

seconds (sec)

Mass

m

lb-sec2/in (bl)

lb-sec2/ft (sl)

kilograms (kg)

Weight

W

pounds (lb)

pounds (lb)

newtons (N)

Pressure

p

psi

psf

N/m2 = Pa

Velocity

v

in/sec

ft/sec

m/sec

a

in/sec2

ft/sec2

m/sec2

psi

psf

N/m2 = Pa

Acceleration

fps unit

SI unit

Stress

σ, τ

Angle

θ

degrees (deg)

degrees (deg)

degrees (deg)

Angular velocity

ω

radians/sec

radians/sec

radians/sec

Angular acceleration

α

radians/sec2

radians/sec2

radians/sec2

Torque

T

lb-in

lb-ft

N-m

Mass moment of inertia

I

lb-in-sec2

lb-ft-sec2

kg-m2

Area moment of inertia

I

in4

ft4

m4

Energy

E

in-lb

ft-lb

joules = N-m

Power

P

in-lb/sec

ft-lb/sec

N-m/sec = watt

Volume

V

in3

ft3

m3

Specific weight

ν

lb/in3

lb/ft3

N/m3

ρ

bl/in3

sl/ft3

kg/m3

Mass density

m=

lb − sec 2 4500 lbf W = = 139.9 f = 139.9 slugs 2 ft g 32.17 ft sec

(a)

m=

lbf − sec 2 4500 lbf W 11.66 = = = 11.66 blobs in g 386 in sec 2

(b)

For the ips system:

For the SI system: W = 4500 lb m=

4.448 N = 20 016 N lb

20 016 N W N − sec 2 = = 2040 = 2040 kg 2 g 9.81 m sec m

(c)

24

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

1 Table 1-5

Selected Units Conversion Factors Note That These Conversion Factors (and Others) are Built Into the TK Solver Files UNITMAST and STUDENT

Multiply this

by

this

to get

this

by

this

to get

this

mass moment of inertia

acceleration in/sec2

x

ft/sec2

x

0.0254 12

=

m/sec2

=

in/sec2

lb-in-sec2

x

57.2958

=

deg

area in2

x

645.16

=

mm2

ft2

x

144

=

in2

area moment of inertia

x x

0.1138 12

=

N-m

x

=

in-lb

x

8.7873

=

in-lb

N-m

x

0.7323

=

ft-lb

power HP

x

550

=

ft-lb/sec

HP

x

33 000

=

ft-lb/min

6600

=

in-lb/sec

=

watts

=

in-lb/sec

x

=

mm4

HP

x

x

4.162E–07

=

m4

HP

x

m4

x

1.0E+12

=

mm4

N-m/sec

x

m4

x

1.0E+08

=

cm4

ft4

x

=

in4

density

N-m-sec2

ft-lb

in4

20 736

=

N-m

in4

416 231

0.1138

moments and energy in-lb

angles radian

Multiply this

745.7 8.7873

pressure and stress psi

x

psi

x

6894.8 6.895E-3

=

Pa

=

MPa

lb/in3

x

27.6805

=

g/cc

psi

x

144

=

psf

g/cc

x

0.001

=

g/mm3

kpsi

x

1000

=

psi

N/m2

x

1

=

Pa

N/mm2

x

1

=

MPa

175.126

lb/ft3

x

=

lb/in3

kg/m3

x

1.0E–06

=

g/mm3

lb

x

4.448

=

N

lb/in

x

=

N/m

N

x

1.0E+05

=

dyne

lb/ft

x

0.08333

=

lb/in

ton (short)

x

=

lb

x

0.909

=

ksi-in0.5

0.0254

=

m/sec

=

in/sec

=

rpm

1728

spring rate

force

2000

length

MPa-m0.5

in

x

25.4

=

mm

ft

x

12

=

in

mass blob

x

=

lb

slug

x

32.17

=

lb

blob

x

12

=

slug

kg

x

2.205

=

lb

kg

x

9.8083

=

N

kg

stress intensity

x

386

1000

=

g

velocity in/sec

x

ft/sec

x

rad/sec

x

12 9.5493

volume in3

x

16 387.2

=

mm3

ft3

x

1728

=

in3

cm3

x

0.061 023 =

in3

m3

x

1.0E+9

mm3

=

Chapter 1

INTRODUCTION TO DESIGN

1

For the cgs system: W = 4500 lb m=

4.448 E 5 dynes = 2.002 E 9 dynes lb

W 2.002 E 9 dynes dynes − sec 2 = = 2.04 E 6 = 2.04 E 6 g 2 g cm 981 cm sec

(d )

2 For mass expressed in lbm, equation 1.4b must be used. m=W

386 in sec 2 gc = 4500 lbf = 4500 lbm g 386 in sec 2

(e)

Note that lbm is numerically equal to lbf and so must not be used as a mass unit unless you are using the form of Newton’s law expressed as equation 1.4b.

1.10

25

SUMMARY

Design can be fun and frustrating at the same time. Because design problems are very unstructured, a large part of the task is creating sufficient structure to make it solvable. This naturally leads to multiple solutions. To students used to seeking an answer that matches the one in the “back of the book” this exercise can be frustrating. There is no “one right answer” to a design problem, only answers that are arguably better or worse than others. The marketplace has many examples of this phenomenon. How many different makes and models of new automobiles are available? Don’t they all do more or less the same task? But you probably have your own opinion about which ones do the task better than others. Moreover, the task definition is not exactly the same for all examples. A four-wheel-drive automobile is designed for a slightly different problem definition than is a two-seat sports car (though some examples incorporate both those features). The message to the beginning designer then is to be open-minded about the design problems posed. Don’t approach design problems with the attitude of trying to find “the right answer,” as there is none. Rather, be daring! Try something radical. Then test it with analysis. When you find it doesn’t work, don’t be disappointed; instead realize that you have learned something about the problem you didn’t know before. Negative results are still results! We learn from our mistakes and can then design a better solution the next time. This is why iteration is so crucial to successful design. The computer is a necessary tool to the solution of contemporary engineering problems. Problems can be solved more quickly and more accurately with proper use of computer-aided engineering (CAE) software. However, the results are only as good as the quality of the engineering models and data used. The engineer should not rely on computer-generated solutions without also developing and applying a thorough understanding of the fundamentals on which the model and the CAE tools are based. Important Equations Used in This Chapter

See the referenced sections for information on the proper use of these equations.

26

1

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Mass (see Section 1.9):

m=

W gc

(1.3)

Dynamic Force—for use with standard mass units (kg, slugs, blobs) (see Section 1.9):

F = ma

(1.4 a)

Dynamic Force—for use with mass in lbm = lbf (see Section 1.9):

F=

1.11

ma gc

(1.4 b)

REFERENCES

1

Random House Dictionary of the English Language. 2e. unabridged, S.B. Flexner, ed., Random House: New York, 1987, p. 1151.

2

R. L. Norton, Design of Machinery: An Introduction to the Synthesis and Analysis of Mechanisms and Machines, 5e. McGraw-Hill: New York, 2012, pp. 7–14.

3

Autocad, Autodesk Inc., http://usa.autodesk.com

4

NX, Siemens Inc, http://www.plm.automation.siemens.com/en_us/products/nx/

5

ADAMS, MSC, http://www.mscsoftware.com/Products/CAE-Tools/Adams.aspx

6

Working Model, DST, http://www.design-simulation.com/wm2d/index.php.

7

Pro/Engineer, PTC, http://www.ptc.com/community/landing/wf3.htm

8

TK Solver, Universal Technical Systems, http://www.uts.com

9

Mathcad, PTC, www.ptc.com/Mathcad

10

Excel, Microsoft Corp., http://office.microsoft.com/en-us/excel/

11

MATLAB, Mathworks Inc, http://www.mathworks.com/products/matlab/

12

Solidworks, Dassault Systemes, http://www.solidworks.com

1.12

WEB REFERENCES

http://www.onlineconversion.com

Convert just about anything to anything else. Over 5000 units, and 50 000 conversions. http://www.katmarsoftware.com/uconeer.htm

Download a units converter program for engineers. http://global.ihs.com

Search a collection of technical standards with over 500 000 documents available for electronic download. http://www.thomasnet.com

An online resource for finding companies and products manufactured in North America.

Chapter 1

1.13

INTRODUCTION TO DESIGN

BIBLIOGRAPHY

For information on creativity and the design process, the following are recommended:

J. L. Adams, The Care and Feeding of Ideas. 3rd ed. Addison Wesley: Reading, Mass., 1986. J. L. Adams, Conceptual Blockbusting. 3rd ed. Addison Wesley: Reading, Mass., 1986. J. R. M. Alger and C. V. Hays, Creative Synthesis in Design. Prentice-Hall: Englewood Cliffs, N.J., 1964. M. S. Allen, Morphological Creativity. Prentice-Hall: Englewood Cliffs, N.J., 1962. H. R. Buhl, Creative Engineering Design. Iowa State University Press: Ames, Iowa, 1960. W. J. J. Gordon, Synectics. Harper and Row: New York, 1962. J. W. Haefele, Creativity and Innovation. Reinhold: New York, 1962. L. Harrisberger, Engineersmanship. 2nd ed. Brooks/Cole: Monterey, Calif., 1982. D. A. Norman, The Psychology of Everyday Things. Basic Books: New York, 1986. A. F. Osborne, Applied Imagination. Scribners: New York, 1963. C. W. Taylor, Widening Horizons in Creativity. John Wiley: New York, 1964. E. K. Von Fange, Professional Creativity. Prentice-Hall: Englewood Cliffs, N.J., 1959. For information on writing engineering reports, the following are recommended:

R. Barrass, Scientists Must Write. Chapman and Hall: New York, 1978. W. G. Crouch and R. L. Zetler, A Guide to Technical Writing. 3rd ed. The Ronald Press Co.: New York, 1964. D. S. Davis, Elements of Engineering Reports. Chemical Publishing Co.: New York, 1963. D. E. Gray, So You Have to Write a Technical Report. Information Resources Press: Washington, D.C., 1970. H. B. Michaelson, How to Write and Publish Engineering Papers and Reports. ISI: Philadelphia, Pa., 1982. J. R. Nelson, Writing the Technical Report. 3rd ed. McGraw-Hill: New York, 1952.

1.14

PROBLEMS

1-1

It is often said, “Build a better mousetrap and the world will beat a path to your door.” Consider this problem and write a goal statement and a set of at least 12 task specifications that you would apply to its solution. Then suggest three possible concepts to achieve the goal. Make annotated, freehand sketches of the concepts.

1-2

A bowling machine is desired to allow quadriplegic youths, who can only move a joystick, to engage in the sport of bowling at a conventional bowling alley. Consider the factors involved, write a goal statement, and develop a set of at least 12 task specifications that constrain this problem. Then suggest three possible concepts to achieve the goal. Make annotated, freehand sketches of the concepts.

27

1

28

1 Table P1-0

MACHINE DESIGN

1-3

Topic/Problem Matrix

-

An Integrated Approach

-

Sixth Edition

A quadriplegic needs an automated page-turner to allow her to read books without assistance. Consider the factors involved, write a goal statement, and develop a set of at least 12 task specifications that constrain this problem. Then suggest three possible concepts to achieve the goal. Make annotated, freehand sketches of the concepts.

1.4 Engineering Model

*1-4

Convert a mass of 1000 lbm to (a) lbf, (b) slugs, (c) blobs, (d) kg.

1-1, 1-2, 1-3, 1-11, 1-13

*1-5

A 250-lbm mass is accelerated at 40 in/sec2. Find the force in lb needed for this acceleration.

*1-6

Express a 100 kg mass in units of slugs, blobs, and lbm. How much does this mass weigh in lbf and in N?

1.9 Units 1-4, 1-5, 1-6, 1-7, 1-8, 1-12, 1-14, 1-15

1-7

Prepare an interactive computer program (using, for example, Excel, Mathcad, MATLAB, or TK Solver) from which the cross-sectional properties for the shapes shown on the inside front cover can be calculated. Arrange the program to deal with both ips and SI units systems and convert the results between those systems.

1-8

Prepare an interactive computer program (using, for example, Excel, Mathcad, MATLAB, or TK Solver) from which the mass properties for the solids shown on the page opposite the inside front cover can be calculated. Arrange the program to deal with both ips and SI units systems and convert the results between those systems.

1-9

Convert the program written for Problem 1-7 to have and use a set of functions or subroutines that can be called from within any program in that language to solve for the cross-sectional properties of the shapes shown on the inside front cover.

1-10

Convert the program written for Problem 1-8 to have and use a set of functions or subroutines that can be called from within any program in that language to solve for the mass properties for the solids shown on the page opposite the inside front cover.

1-11

A fledgling, nonprofit, recycling organization collects paper, cardboard, plastics, and metals from consumers in a rural county. After sorting plastics by grades 1 and 2, they are transported to a plastics recycler in a major city 150 miles away. In order to maximize the weight of the plastics that can be put into their trailer the organization desires to compact the plastic containers for transport. Consider this problem and write a goal statement and a set of at least 10 task specifications that you would apply to its solution. Then suggest three possible concepts to achieve the goal. Make annotated, freehand sketches of the concepts.

1-12

One square foot of ventilation is required for every 150 sq. ft. of floor space on a house with crawl-space under the floor. Prepare an interactive computer program (using, for example Excel, Mathcad, MATLAB, or TK Solver) from which the number of 40 cm by 20 cm vents can be determined that meet the ventilation requirement if only 75% of the nominal vent area is effective. Test the program using a house that is 13.5 m long by 8.25 m wide.

1-13

A start-up company (husband and wife) manufactures over 50 different health and beauty aids that they sell at craft shows and also over the internet. They make their products in small batches and package them in small jars that they buy unlabeled in case lots. They design and print their own labels on peelable, adhesive sheets. It is very difficult to place the labels on to the containers so that they are in the proper position and properly aligned. Consider this problem and write a goal statement and at least 10 task specifications that you would apply to its solution.

1-14

A 360-gram mass is accelerated at 250 cm/sec2. Find the force in N needed for this acceleration.

1-15

Express an 18-blob mass in units of slugs, kilograms, and lbm. How much does this mass weigh?

*

Answers to these problems are provided in Appendix D.

MATERIALS AND PROCESSES There is no subject so old that something new cannot be said about it. Dostoevsky

2.0

INTRODUCTION

Whatever you design, you must make it out of some material and be able to manufacture it. A thorough understanding of material properties, treatments, and manufacturing processes is essential to good machine design. It is assumed that the reader has had a first course in material science. This chapter presents a brief review of some basic metallurgical concepts and a short summary of engineering material properties to serve as background for what follows. This is not intended as a substitute for a text on material science, and the reader is encouraged to review references such as those listed in the bibliography of this chapter for more detailed information. Later chapters of this text will explore some of the common material-failure modes in more detail. Table 2-0 shows the variables used in this chapter and references the equations, figures, or sections in which they are used. At the end of the chapter, a summary section is provided which groups the significant equations from this chapter for easy reference and identifies the chapter section in which they are discussed. 2.1

MATERIAL-PROPERTY DEFINITIONS View a Video on Failure Modes (09:41)†

Mechanical properties of a material are generally determined through destructive testing of samples under controlled loading conditions. The test loadings do not accurately duplicate actual service loadings experienced by machine parts except in certain special cases. Also, there is no guarantee that the particular piece of material you purchase for your part will exhibit the same strength properties as the samples of similar materials tested previously. There will be some statistical variation in the strength of any particular † http://www.designofmachinery.com/MD/Failure_Modes.mp4

29

2

30

2

MACHINE DESIGN

Opening page photograph courtesy of the Aluminum Extruders Council.

Table 2-0

-

An Integrated Approach

-

Sixth Edition

Variables Used in This Chapter

Symbol

Variable

ips units

SI units

See

A A0

area

in2

m2

Sect. 2.1

original area, test specimen

in2

m2

Eq. 2.1a

E

Young's modulus

psi

Pa

Eq. 2.2

el

elastic limit

psi

Pa

Figure 2-2

f

fracture point

none

none

Figure 2-2

G

shear modulus, modulus of rigidity

psi

Pa

Eq. 2.4

HB

Brinell hardness

none

none

Eq. 2.10

HRB

Rockwell B hardness

none

none

Sect. 2.4

HRC

Rockwell C hardness

none

none

Sect. 2.4

HV

Vickers hardness

none

none

Sect. 2.4

polar second moment of area

in4

m4

Eq. 2.5

stress intensity

kpsi-in0.5

MPa-m0.5

Sect. 2.1

Kc

fracture toughness

kpsi-in0.5

MPa-m0.5

Sect. 2.1

l0

gage length, test specimen

in

m

Eq. 2.3

N

number of cycles

none

none

Figure 2-10

P

force or load

lb

N

Sect. 2.1

pl

proportional limit

psi

Pa

Figure 2-2

r Sd

radius

in

m

Eq. 2.5a

standard deviation

any

any

Eq. 2.9

Se

endurance limit

psi

Pa

Figure 2-10

Sel

strength at elastic limit

psi

Pa

Eq. 2.7

Sf

fatigue strength

psi

Pa

Figure 2-10

Sus

ultimate shear strength

psi

Pa

Eq. 2.5

Sut

ultimate tensile strength

psi

Pa

Figure 2-2

Sy

tensile yield strength

psi

Pa

Figure 2-2

Sys

shear yield strength

psi

Pa

Eq. 2.5c

T UR

torque

lb-in

N-m

Sect. 2.1

modulus of resilience

psi

Pa

Eq. 2.7

UT

modulus of toughness

psi

Pa

Eq. 2.8

y

yield point

none

none

Figure 2-2 Eq. 2.1b

J K

ε

strain

none

none

σ

tensile stress

psi

Pa

Sect. 2.1

τ

shear stress

psi

Pa

Eq. 2.3

θ

angular deflection

rad

rad

Eq. 2.3

µ

arithmetic mean value

any

any

Eq. 2.9b

ν

Poisson's ratio

none

none

Eq. 2.4

sample compared to the average tested properties for that material. For this reason, many of the published strength data are given as minimum values. It is with these caveats that we must view all published material-property data, as it is the engineer’s responsibility to ensure the safety of his or her design.

Chapter 2

MATERIALS AND PROCESSES

31

The best material-property data will be obtained from destructive or nondestructive testing under actual service loadings of prototypes of your actual design, made from the actual materials by the actual manufacturing process. This is typically done only when the economic and safety risks are high. Manufacturers of aircraft, automobiles, motorcycles, snowmobiles, farm equipment, and other products regularly instrument and test finished assemblies under real or simulated service conditions. In the absence of such specific test data, the engineer must adapt and apply published material-property data from standard tests to the particular situation. The American Society for Testing and Materials (ASTM) defines standards for test specimens and test procedures for a variety of material-property measurements.* The most common material test used is the tensile test.

* ASTM, 1994 Annual Book of ASTM Standards, Vol. 03.01, Am. Soc. for Testing and Materials, Philadelphia, PA.

The Tensile Test A typical tensile test specimen is shown in Figure 2-1. This tensile bar is machined from the material to be tested in one of several standard diameters do and gage lengths lo. The gage length is an arbitrary length defined along the small-diameter portion of the specimen by two indentations so that its increase can be measured during the test. The larger-diameter ends of the bar are threaded for insertion into a tensile test machine which is capable of applying either controlled loads or controlled deflections to the ends of the bar, and the gage-length portion is mirror polished to eliminate stress concentrations from surface defects. The bar is stretched slowly in tension until it breaks, while the load and the distance across the gage length (or alternatively the strain) are continuously monitored. The result is a stress-strain plot of the material’s behavior under load as shown in Figure 2-2a, which depicts a curve for a low-carbon or “mild” steel. StreSS and Strain Note that the parameters measured are load and deflection, but those plotted are stress and strain. Stress (σ) is defined as load per unit area (or unit load) and for the tensile specimen is calculated from σ=

P Ao

(2.1a)

Strain is the change in length per unit length and is calculated from l − lo lo

(2.1b)

where lo is the original gage length and l is the gage length at any load P. The strain is unitless, being length divided by length. ModuluS of elaSticity This tensile stress-strain curve provides us with a number of useful material parameters. Point pl in Figure 2-2a is the proportional limit below which the stress is proportional to the strain, as expressed by the one-dimensional form of Hooke’s law: E=

σ ε

do

FIGURE 2-1 A Tensile Test Specimen

where P is the applied load at any instant and Ao is the original cross-sectional area of the specimen. The stress is assumed to be uniformly distributed across the cross section. The stress units are psi or Pa.

ε=

lo

(2.2)

32

MACHINE DESIGN

-

Stress σ

true

Sut

2

An Integrated Approach

y

Sy

u

f

-

Sixth Edition

Stress σ

true u

eng'g

y el pl

pl el (a)

(b)

f

eng'g

offset line E

E Elastic Range

Plastic Range

Strain ε

0.002

Strain ε

FIGURE 2-2 Engineering and True Stress-Strain Curves for Ductile Materials: (a) Low-Carbon Steel (b) Annealed High-Carbon Steel

where E defines the slope of the stress-strain curve up to the proportional limit and is called Young’s modulus or the modulus of elasticity of the material. E is a measure of the stiffness of the material in its elastic range and has the units of stress. Most metals exhibit this linear stiffness behavior and also have elastic moduli that vary very little with heat treatment or with the addition of alloying elements. For example, the higheststrength steel has the same E as the lowest-strength steel at about 30 Mpsi (207 GPa). For most ductile materials (defined below), the modulus of elasticity in compression is the same as in tension. This is not true for cast irons and other brittle materials (defined below) or for magnesium. elaStic liMit The point labeled el in Figure 2-2a is the elastic limit, or the point beyond which the material will take a permanent set, or plastic deformation. The elastic limit marks the boundary between the elastic-behavior and plastic-behavior regions of the material. Points el and pl are typically so close together that they are often considered to be the same. yield Strength At a point y slightly above the elastic limit, the material begins to yield more readily to the applied stress, and its rate of deformation increases (note the lower slope). This is called the yield point, and the value of stress at that point defines the yield strength Sy of the material. Materials that are very ductile, such as low-carbon steels, will sometimes show an apparent drop in stress just beyond the yield point, as shown in Figure 2-2a. Many less ductile materials, such as aluminum and medium- to high-carbon steels, will not exhibit this apparent drop in stress and will look more like Figure 2-2b. The yield strength of a material that does not exhibit a clear yield point has to be defined with an offset line, drawn parallel to the elastic curve and offset some small percentage along the strain axis. An offset of 0.2% strain is most often used. The yield strength is then taken at the intersection of the stress-strain curve and the offset line as shown in Figure 2-2b. ultiMate tenSile Strength The stress in the specimen continues to increase nonlinearly to a peak or ultimate tensile strength value Sut at point u. This is considered to be the largest tensile stress the material can sustain before breaking. However, for the ductile steel curve shown, the stress appears to fall off to a smaller value at the fracture

Chapter 2

FIGURE 2-3

MATERIALS AND PROCESSES

Copyright © 2018 Robert L. Norton: All Rights Reserved

A Tensile Test Specimen of Mild, Ductile Steel Before and After Fracture

point f. The drop in apparent stress before the fracture point (from u to f in Figure 2-2a) is an artifact caused by the “necking-down” or reduction in area of the ductile specimen. The reduction of cross-sectional area is nonuniform along the length of the specimen, as can be seen in Figure 2-3. Because the stress is calculated using the original area Ao in equation 2.1a, it understates the true value of stress after point u. It is difficult to accurately monitor the dynamic change in cross-sectional area during the test, so these errors are accepted. The strengths of different materials can still be compared on this basis. When based on the uncorrected area Ao this is called the engineering stress-strain curve, as shown in Figure 2-2. The stress at fracture is actually larger than shown. Figure 2-2 also shows the true stress-strain curve that would result if the change in area were accounted for. The engineering stress-strain data from Figure 2-2 are typically used in practice. The most commonly used strength values for static loading are the yield strength Sy and the ultimate tensile strength Sut. The material stiffness is defined by Young’s modulus, E. In comparing the properties of different materials, it is quite useful to express those properties normalized to the material’s density. Since light weight is nearly always a goal in design, we seek the lightest material that has sufficient strength and stiffness to withstand the applied loads. The specific strength of a material is defined as the strength divided by the density. Unless otherwise specified, strength in this case is assumed to mean ultimate tensile strength, though any strength criterion can be so normalized. The strength-to-weight ratio (SWR) is another way to express the specific strength. Specific stiffness is the Young’s modulus divided by material density. Ductility and Brittleness The tendency for a material to deform significantly before fracturing is a measure of its ductility. The absence of significant deformation before fracture is called brittleness. ductility The stress-strain curve in Figure 2-2a is of a ductile material, mild steel. Take a common paper clip made of mild-steel wire. Straighten it out with your fingers. Bend it into some new shape. You are yielding this ductile steel wire but not

33

34

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Stress σ

2 Sut

f

Sy

y offset line E FIGURE 2-5 0.002

Strain ε

FIGURE 2-4 Stress-Strain Curve of a Brittle Material

Copyright © 2018 Robert L. Norton: All Rights Reserved

A Tensile Test Specimen of Brittle Cast Iron Before and After Fracture

fracturing it. You are operating between point y and point f on the stress-strain curve of Figure 2-2a. The presence of a significant plastic region on the stress-strain curve is evidence of ductility. Figure 2-3 shows a test specimen of ductile steel after fracture. The distortion called necking-down can clearly be seen at the break. The fracture surface appears torn and is laced with hills and valleys, also indicating a ductile failure. The ductility of a material is measured by its percent elongation to fracture, or percent reduction in area at fracture. Materials with more than 5% elongation at fracture are considered ductile. BrittleneSS Figure 2-4 shows a stress-strain curve for a brittle material. Note the lack of a clearly defined yield point and the absence of any plastic range before fracture. Repeat your paper-clip experiment, this time using a wooden toothpick or matchstick. Any attempt to bend it results in fracture. Wood is a brittle material. Brittle materials do not exhibit a clear yield point, so the yield strength has to be defined at the intersection of the stress-strain curve and an offset line, drawn parallel to the elastic curve and offset some small percentage such as 0.2% along the strain axis. Some brittle materials like cast iron do not have a linear elastic region, and the offset line is taken at the average slope of the region. Figure 2-5 shows a cast iron test specimen after fracture. The break shows no evidence of necking and has the finer surface contours typical of a brittle fracture. The same metals can be either ductile or brittle depending on the way they are manufactured, worked, and heat treated. Metals that are wrought (meaning drawn or pressed into shape in a solid form while either hot or cold) can be more ductile than metals that are cast by pouring molten metal into a mold or form. There are many exceptions to this broad statement, however. The cold working of metal (discussed below) tends to reduce its ductility and increase its brittleness. Heat treatment (discussed below) also has a marked effect on the ductility of steels. Thus, it is difficult to generalize about the relative ductility or brittleness of various materials. A careful look at all the mechanical properties of a given material will tell the story.

Chapter 2

(a)

FIGURE 2-6

MATERIALS AND PROCESSES

(b)

Copyright © 2018 Robert L. Norton: All Rights Reserved

Compression Test Specimens Before and After Failure (a) Ductile Steel (b) Brittle Cast Iron

The Compression Test The tensile test machine can be run in reverse to apply a compressive load to a specimen that is a constant-diameter cylinder as shown in Figure 2-6. It is difficult to obtain a useful stress-strain curve from this test because a ductile material will yield and increase its cross-sectional area, as shown in Figure 2-6a, eventually stalling the test machine. The ductile sample will not fracture in compression. If enough force were available from the machine, it could be crushed into a pancake shape. Most ductile materials have compressive strengths similar to their tensile strengths, and the tensile stress-strain curve is used to represent their compressive behavior as well. A material that has essentially equal tensile and compressive strengths is called an even material. Brittle materials will fracture when compressed. A failed specimen of brittle cast iron is shown in Figure 2-6b. Note the rough, angled fracture surface. The reason for the failure on an angled plane is discussed in Chapter 4. Brittle materials generally have much greater strength in compression than in tension. Compressive stress-strain curves can be generated, since the material fractures rather than crushes and the cross-sectional area doesn’t change appreciably. A material that has different tensile and compressive strengths is called an uneven material. The Bending Test A thin rod, as shown in Figure 2-7, is simply supported at each end as a beam and loaded transversely in the center of its length until it fails. If the material is ductile, failure is by yielding, as shown in Figure 2-7a. If the material is brittle, the beam fractures as shown in Figure 2-7b. Stress-strain curves are not generated from this test because the stress distribution across the cross section is not uniform. The tensile test’s σ–ε curve is used to predict failure in bending, since the bending stresses are tensile on the convex side and compressive on the concave side of the beam. The Torsion Test The shear properties of a material are more difficult to determine than its tensile properties. A specimen similar to the tensile test specimen is made with noncircular details on its ends so that it can be twisted axially to failure. Figure 2-8 shows two such samples, one of ductile steel and one of brittle cast iron. Note the painted lines along their lengths.

35

36

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

2

(a )

(b)

FIGURE 2-7

Copyright © 2018 Robert L. Norton: All Rights Reserved

Bending Test Specimens Before and After Failure (a) Ductile Steel (b) Brittle Cast Iron

The lines were originally straight in both cases. The helical twist in the ductile specimen’s line after failure shows that it wound up for several revolutions before breaking. The brittle, torsion-test specimen’s line is still straight after failure as there was no significant plastic distortion before fracture. ModuluS of rigidity The stress-strain relation for pure torsion is defined by

Table 2-1

τ=

Poisson’s Ratio ν

Material

ν

Aluminum

0.34

Copper

0.35

Iron

0.28

Steel

0.28

Magnesium

0.33

Titanium

0.34

Gr θ

(2.3)

lo

where τ is the shear stress, r is the radius of the specimen, lo is the gage length, θ is the angular twist in radians, and G is the shear modulus or modulus of rigidity. G can be defined in terms of Young’s modulus E and Poisson’s ratio ν: G=

E 2 (1 + ν )

(2.4)

Poisson’s ratio (ν) is the ratio between lateral and longitudinal strain and for most metals is around 0.3 as shown in Table 2-1. ultiMate Shear Strength The breaking strength in torsion is called the ultimate shear strength or modulus of rupture Sus and is calculated from

(a)

FIGURE 2-8 Torsion Test Specimens Before and After Failure (a) Ductile Steel (b) Brittle Cast Iron

(b) Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 2

Sus =

Tr J

MATERIALS AND PROCESSES

37

(2.5a)

where T is the applied torque necessary to break the specimen, r is the radius of the specimen, and J is the polar second moment of area of the cross section. The distribution of stress across the section loaded in torsion is not uniform. It is zero at the center and maximum at the outer radius. Thus, the outer portions have already plastically yielded while the inner portions are still below the yield point. This nonuniform stress distribution in the torsion test (unlike the uniform distribution in the tension test) is the reason for calling the measured value at failure of a solid bar in torsion a modulus of rupture. A thin-walled tube is a better torsion-test specimen than a solid bar for this reason and can give a better measure of the ultimate shear strength. In the absence of available data for the ultimate shear strength of a material, a reasonable approximation can be obtained from tension test data:* steels:

Sus ≅ 0.80 Sut

other ductile metals:

Sus ≅ 0.75Sut

(2.5b)

Note that the shear yield strength has a different relationship to the tensile yield strength: S ys ≅ 0.577 S y

(2.5c)

This relationship is derived in Chapter 5, where failure of materials under static loading is discussed in more detail. Fatigue Strength and Endurance Limit The tensile test and the torsion test both apply loads slowly and only once to the specimen. These are static tests and measure static strengths. While some machine parts may see only static loads in their lifetime, most will see loads and stresses that vary with time. Materials behave very differently in response to loads that come and go (called fatigue loads) than they do to loads that remain static. Most of machine design deals with the design of parts for time-varying loads, so we need to know the fatigue strength of materials under these loading conditions. One test for fatigue strength is the R. R. Moore rotating-beam test in which a similar, but slightly smaller, test specimen than that shown in Figure 2-1 is loaded as a beam in bending while being rotated by a motor. Recall from your first course in strength of materials that a bending load causes tension on one side of a beam and compression on the other. (See Sections 4-9 and 4-10 for a review of beams in bending.) The rotation of the beam causes any one point on the surface to go from compression to tension to compression each cycle. This creates a load-time curve as shown in Figure 2-9. The test is continued at a particular stress level until the part fractures, and the number of cycles N is then noted. Many samples of the same material are tested at various stress levels S until a curve similar to Figure 2-10 is generated. This is called a Wohler strength-life diagram or an S-N diagram. It depicts the breaking strength of a particular material at various numbers of repeated cycles of fully reversed stress.

* In Chapter 14 on helical spring design, an empirical relationship for the ultimate shear strength of small diameter steel wire, based on extensive testing of wire in torsion, is presented in equation 14.4 and is Sus = 0.67 Sut. This is obviously different than the general approximation for steel in equation 2.5b. The best data for material properties will always be obtained from tests of the same material, geometry, and loading as the part will be subjected to in service. In the absence of direct test data we must rely on approximations of the sort in equation 2.5b and apply suitable safety factors based on the uncertainty of these approximations.

+

stress time

– FIGURE 2-9 Time-Varying Loading

38

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

2

log fatigue strength

Sf Sut

failure line An endurance limit Se exists for some ferrous metals and titanium alloys. Other materials show no endurance limit.

Se

Se

Sf 100 101

102

103

104

105

106

107

108

109

N

log number of cycles FIGURE 2-10 Wohler Strength-Life or S-N Diagram Plots Fatigue Strength Against Number of Fully Reversed Stress Cycles

Note in Figure 2-10 that the fatigue strength Sf at one cycle is the same as the static strength Sut, and it decreases steadily with increasing numbers of cycles N (on a log-log plot) until reaching a plateau at about 106 cycles. This plateau in fatigue strength exists only for certain metals (notably steels and some titanium alloys) and is called the endurance limit Se. Fatigue strengths of other materials keep falling beyond that point. While there is considerable variation among materials, their raw (or uncorrected) fatigue strengths at about N = 106 cycles tend to be no more than about 40–50% of their static tensile strength Sut. This is a significant reduction and, as we will learn in Chapter 6, further reductions in the fatigue strengths of materials will be necessary due to other factors such as surface finish and type of loading. It is important at this stage to remember that the tensile stress-strain test does not tell the whole story and that a material’s static strength properties are seldom adequate by themselves to predict failure in a machine-design application. This topic of fatigue strength and endurance limit is so important and fundamental to machine design that we devote Chapter 6 exclusively to a study of fatigue failure. The rotating-beam test is now being supplanted by axial-tension tests performed on modern test machines which can apply time-varying loads of any desired character to the axial-test specimen. This approach provides more testing flexibility and more accurate data because of the uniform stress distribution in the tensile specimen. The results are consistent with (but slightly lower-valued than) the historical rotating-beam test data for the same materials. Impact Resistance The stress-strain test is done at very low, controlled strain rates, allowing the material to accommodate itself to the changing load. If the load is suddenly applied, the energy absorption capacity of the material becomes important. The energy in the differential element is its strain energy density (strain energy per unit volume U0), or the area under the stress-strain curve at any particular strain: U0 =

ε

∫0

σ dε

(2.6a)

Chapter 2

MATERIALS AND PROCESSES

39

The strain energy U is equal to the strain energy density integrated over the volume v: U=

∫v

(2.6b)

U 0 dv

The resilience and toughness of the material are measures, respectively, of the strain energy present in the material at the elastic limit or at the fracture point. reSilience The ability of a material to absorb energy per unit volume without permanent deformation is called its resilience UR (also called modulus of resilience) and is equal to the area under the stress-strain curve up to the elastic limit, shown as the color-shaded area in Figure 2-2a. Resilience is defined as: UR = = UR ≅

εel

∫0

σdε =

1 Sel εel 2

S 1 1 Sel2 Sel el = E 2 2 E

(2.7)

2 1 Sy 2 E

where Sel and εel represent, respectively, the strength and strain at the elastic limit. Substitution of Hooke’s law from equation 2.2 expresses the relationship in terms of strength and Young’s modulus. Since the Sel value is seldom available, a reasonable approximation of resilience can be obtained by using the yield strength Sy instead. This relationship shows that a stiffer material of the same elastic strength is less resilient than a more compliant one. A rubber ball can absorb more energy without permanent deformation than one made of glass. toughneSS The ability of a material to absorb energy per unit volume without fracture is called its toughness UT (also called modulus of toughness) and is equal to the area under the stress-strain curve up to the fracture point, shown as the entire shaded area in Figure 2-2a. Toughness is defined as: UT =

εf

∫0

σdε

 S y + Sut  ≅  ε f 2 

(2.8)

where Sut and εf represent, respectively, the ultimate tensile strength and the strain at fracture. Since an analytical expression for the stress-strain curve is seldom available for actual integration, an approximation of toughness can be obtained by using the average of the yield and ultimate strengths and the strain at fracture to calculate an area. The units of toughness and resilience are energy per unit volume (in-lb/in3 or joules/m3). Note that these units are numerically equivalent to psi or Pa. A ductile material of similar ultimate strength to a brittle one will be much more tough. A sheet-metal automobile body will absorb more energy from a collision through plastic deformation than will a brittle, fiberglass body.*

* It is interesting to note that one of the toughest and strongest materials known is that of spider webs! These tiny arachnids spin a monofilament that has an ultimate tensile strength of 200 to 300 kpsi (1380 to 2070 MPa) and 35% elongation to fracture! It also can absorb more energy without rupture than any fiber known, absorbing 3 times as much energy as Kevlar, the manmade fiber used for bullet-proof vests. According to the Boston Globe, January 18, 2002, researchers in Canada and the U.S. have synthesized a material with similar properties to spider silk in strands up to 10-ft long with strengths of 1/4 to 1/3 that of natural silk fiber, “stronger than a steel wire of similar weight,” and one that has greater elasticity than organic silk fiber.

40

2

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

iMpact teSting Various tests have been devised to measure the ability of materials to withstand impact loading. The Izod and the Charpy tests are two such procedures which involve striking a notched specimen with a pendulum and recording the kinetic energy needed to break the specimen at a particular temperature. While these data do not directly correlate with the area under the stress-strain curve, they nevertheless provide a means to compare the energy absorption capacity of various materials under controlled conditions. Materials handbooks, such as those listed in this chapter’s bibliography, give data on the impact resistance of various materials. Fracture Toughness Fracture toughness Kc (not to be confused with the modulus of toughness defined previously) is a material property that defines its ability to resist stress at the tip of a crack. The fracture toughness of a material is measured by subjecting a standardized, pre-cracked test specimen to cyclical tensile loads until it breaks. Cracks create very high local stress concentrations which cause local yielding (see Section 4.15). The effect of the crack on the local stress is measured by a stress intensity factor K which is defined in Section 5.3. When the stress intensity K reaches the fracture toughness Kc, a sudden fracture occurs with no warning. The study of this failure phenomenon is called fracture mechanics and it is discussed in more detail in Chapters 5 and 6. Creep and Temperature Effects The tensile test, while slow, does not last long compared to the length of time an actual machine part may be subjected to constant loading. All materials will, under the right environmental conditions (particularly elevated temperatures), slowly creep (deform) under stress loadings well below the level (yield point) deemed safe in the tensile test. Ferrous metals tend to have negligible creep at room temperature or below. Their creep rates increase with increasing ambient temperature, usually becoming significant around 30–60% of the material’s absolute melting temperature. Low-melt-temperature metals such as lead, and many polymers, can exhibit significant creep at room temperature as well as increasing creep rates at higher temperatures. Creep data for engineering materials are quite sparse due to the expense and time required to develop the experimental data. The machine designer needs to be aware of the creep phenomenon and obtain the latest manufacturer’s data on the selected materials if high ambient temperatures are anticipated or if polymers are specified. The creep phenomenon is more complex than this simple description implies. See the bibliography to this chapter for more complete and detailed information on creep in materials. It is also important to understand that all material properties are a function of temperature, and published test data are usually generated at room temperature. Increased temperature usually reduces strength. Many materials that are ductile at room temperature can behave as brittle materials at low temperatures. Thus, if your application involves either elevated or low temperatures, you need to seek out relevant material-property data for your operating environment. Material manufacturers are the best source of up-to-date information. Most manufacturers of polymers publish creep data for their materials at various temperatures.

Chapter 2

2.2

MATERIALS AND PROCESSES

THE STATISTICAL NATURE OF MATERIAL PROPERTIES

f(x)

Some published data for material properties represent average values of many samples tested. (Other data are stated as minimum values.) The range of variation of the published test data is sometimes stated, sometimes not. Most material properties will vary about the average or mean value according to some statistical distribution such as the Gaussian or normal distribution shown in Figure 2-11. This curve is defined in terms of two parameters, the arithmetic mean µ and the standard deviation Sd. The equation of the Gaussian distribution curve is: f (x) =

 ( x − µ )2  1 , exp  − 2 π Sd 2 Sd2  

−∞ ≤ x ≤ ∞

41

(2.9a)

Sd Sd Sd Sd Sd Sd

x

µ FIGURE 2-11 The Gaussian (Normal) Distribution

where x represents some material parameter, f(x) is the frequency with which that value of x occurs in the population, and µ and Sd are defined as: µ=

Sd =

1 n

n

∑ xi

(2.9b)

i =1

1 n-1

n

∑ ( xi − µ )

2

(2.9c)

i =1

The mean µ defines the most frequently occurring value of x at the peak of the curve, and the standard deviation Sd is a measure of the “spread” of the curve about the mean. A small value of Sd relative to µ means that the entire population is clustered closely about the mean. A large Sd indicates that the population is widely dispersed about the mean. We can expect to find 68% of the population within µ ± 1Sd, 95% within µ ± 2Sd, and 99% within µ ± 3Sd. There is considerable scatter in multiple tests of the same material under the same test conditions. Note that there is a 50% chance that the samples of any material that you buy will have a strength less than that material’s published mean value. Thus, you may not want to use the mean value alone as a predictor of the strength of a randomly chosen sample of that material. If the standard deviation of the test data is published along with the mean, we can “factor it down” to a lower value that is predictive of some larger percentage of the population based on the ratios listed previously. For example, if you want to have a 99% probability that all possible samples of material are stronger than your assumed material strength, you will subtract 3Sd from µ to get an allowable value for your design. This assumes that the material property’s distribution is Gaussian and not skewed toward one end or the other of the spectrum. If a minimum value of the material property is given (and used), then its statistical distribution is not of concern. Usually, no data are available on the standard deviation of the material samples tested. But you can still choose to reduce the published mean strength by a reliability factor based on an assumed Sd. One such approach assumes Sd to be some percentage of µ based on experience. Haugen and Wirsching[1] report that the standard deviations of strengths of steels seldom exceed 8% of their mean values. Table 2-2 shows reliability reduction factors based on an assumption of Sd = 0.08 µ for various reliabilities. Note that a 50% reliability has a factor of 1 and the factor reduces as you choose higher

Table 2-2 Reliability Factors for Sd = 0.08 µ Reliability %

Factor

50

1.000

90

0.897

95

0.868

99

0.814

99.9

0.753

99.99

0.702

99.999

0.659

99.9999

0.620

42

2

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

reliability. The reduction factor is multiplied by the mean value of the relevant material property. For example, if you wish 99.99% of your samples to meet or exceed the assumed strength, multiply the mean strength value by 0.702. In summary, the safest approach is to develop your own material-property data for the particular materials and loading conditions relevant to your design. Since this approach is usually prohibitively expensive in both time and money, the engineer often must rely on published material-property data. Some published strength data are expressed as the minimum strength to be expected in a statistical sample, but other data may be given as the average value for the samples tested. In that case, some of the tested material samples failed at stresses lower than the average value, and your design strength may need to be reduced accordingly. 2.3

HOMOGENEITY AND ISOTROPY

All discussion of material properties so far has assumed that the material is homogeneous and isotropic. Homogeneous means that the material properties are uniform throughout its continuum, e.g., they are not a function of position. This ideal state is seldom attained in real materials, many of which are subject to the inclusion of discontinuities, precipitates, voids, or bits of foreign matter from their manufacturing process. However, most metals and some nonmetals can be considered, for engineering purposes, to be macroscopically homogeneous despite their microscopic deviations from this ideal. An isotropic material is one whose mechanical properties are independent of orientation or direction. That is, the strengths across the width and thickness are the same as along the length of the part, for example. Most metals and some nonmetals can be considered to be macroscopically isotropic. Other materials are anisotropic, meaning that there is no plane of material-property symmetry. Orthotropic materials have three mutually perpendicular planes of property symmetry and can have different material properties along each axis. Wood, plywood, fiberglass, and some cold-rolled sheet metals are orthotropic. One large class of materials that is distinctly nonhomogeneous (i.e., heterogeneous) and nonisotropic is that of composites (also see below). Most composites are man-made, but some, such as wood, occur naturally. Wood is a composite of long fibers held together in a resinous matrix of lignin. You know from experience that it is easy to split wood along the grain (fiber) lines and nearly impossible to do so across the grain. Its strength is a function of both orientation and position. The matrix is weaker than the fibers, and it always splits between fibers. 2.4

HARDNESS

The hardness of a material can be an indicator of its resistance to wear (but is not a guarantee of wear resistance). The strengths of some materials such as steels are also closely correlated to their hardness. Various treatments are applied to steels and other metals to increase hardness and strength. These are discussed below. Hardness is most often measured on one of three scales: Brinell, Rockwell, or Vickers. These hardness tests all involve the forced impression of a small probe into the

Chapter 2

MATERIALS AND PROCESSES

surface of the material being tested. The Brinell test uses a 10-mm hardened steel or tungsten-carbide* ball impressed with either a 500- or 3000-kg load depending on the range of hardness of the material. The diameter of the resulting indent is measured under a microscope and used to calculate the Brinell hardness number, which has the units of kgf/mm2. The Vickers test uses a diamond-pyramid indenter and measures the width of the indent under the microscope. The Rockwell test uses a 1/16-in ball or a 120° cone-shaped diamond indenter and measures the depth of penetration. Hardness is indicated by a number followed by the letter H, followed by letter(s) to identify the method used, e.g., 375 HB or 396 HV. Several lettered scales (A, B, C, D, F, G, N, T) are used for materials in different Rockwell hardness ranges and it is necessary to specify both the letter and number of a Rockwell reading, such as 60 HRC. In the case of the Rockwell N scale, a narrow-cone-angle diamond indenter is used with loads of 15, 30, or 40 kg and the specification must include the load used as well as the letter specification, e.g., 84.6 HR15N. This Rockwell N scale is typically used to measure the “superficial” hardness of thin or case-hardened materials. The smaller load and narrowangle N-tip give a shallow penetration that measures the hardness of the case without including effects of the soft core. All these tests are nondestructive in the sense that the sample remains intact. However, the indentation can present a problem if the surface finish is critical or if the section is thin, so they are actually considered destructive tests. The Vickers test has the advantage of having only one test setup for all materials. Both the Brinell and Rockwell tests require selection of the tip size or indentation load, or both, to match the material tested. The Rockwell test is favored for its lack of operator error, since no microscope reading is required and the indentation tends to be smaller, particularly if the N-tip is used. But, the Brinell hardness number provides a very convenient way to quickly estimate the ultimate tensile strength (Sut) of the material from the relationship Sut ≅ 500 HB ± 30 HB

psi

Sut ≅ 3.45 HB ± 0.2 HB

MPa

(2.10)

where HB is the Brinell hardness number. This gives a convenient way to obtain a rough experimental measure of the strength of any low- or medium-strength carbon or alloy steel sample, even one that has already been placed in service and cannot be truly destructively tested. Microhardness tests use a low force on a small diamond indenter and can provide a profile of microhardness as a function of depth across a sectioned sample. The hardness is computed on an absolute scale by dividing the applied force by the area of the indent. The units of absolute hardness are kgf/mm2. Brinell and Vickers hardness numbers also have these hardness units, though the values measured on the same sample can differ with each method. For example, a Brinell hardness of 500 HB is about the same as a Rockwell C hardness of 52 HRC and an absolute hardness of 600 kgf/mm2. Note that these scales are not linearly related, so conversion is difficult. Table 2-3 shows approximate conversions between the Brinell, Vickers, and Rockwell B and C hardness scales for steels and their approximate equivalent ultimate tensile strengths.

43

* Tungsten carbide is one of the hardest substances known.

44

MACHINE DESIGN

Table 2-3

2

-

An Integrated Approach

-

Sixth Edition

Approximate Equivalent Hardness Numbers and Ultimate Tensile Strengths for Steels

Source: Table 5-10, p.185, in N. E. Dowling, Mechanical Behavior of Materials, Prentice Hall, Englewood Cliffs, N.J., 1993, with permission.

Heat Treatment The steel heat-treatment process is quite complicated and is dealt with in detail in materials texts such as those listed in the bibliography at the end of this chapter. The reader is referred to such references for a more complete discussion. Only a brief review of some of the salient points is provided here. The hardness and other characteristics of many steels and some nonferrous metals can be changed by heat treatment. Steel is an alloy of iron and carbon. The weight percent of carbon present affects the alloy’s ability to be heat-treated. A low-carbon steel will have about 0.03 to 0.30% of carbon, a medium-carbon steel about 0.35 to 0.55%, and a high-carbon steel about 0.60 to 1.50%. (Cast irons will have greater than 2% carbon.) Hardenability of steel increases with carbon content. Low-carbon steel has too little carbon for effective through-hardening so other surface-hardening methods must be used (see below). Medium- and high-carbon steels can be through-hardened by appropriate heat treatment. The depth of hardening will vary with alloy content. Quenching To harden a medium- or high-carbon steel, the part is heated above a critical temperature (about 1400°F [760°C]), allowed to equilibrate for some time, and then suddenly cooled to room temperature by immersion in a water or oil bath. The rapid cooling creates a supersaturated solution of iron and carbon called martensite, which is

Chapter 2

MATERIALS AND PROCESSES

45

extremely hard and much stronger than the original soft material. Unfortunately it is also very brittle. In effect, we have traded off the steel’s ductility for its increased strength. The rapid cooling also introduces strains to the part. The change in the shape of the stress-strain curve as a result of quenching a ductile, medium-carbon steel is shown in Figure 2-12 (not to scale). While the increased strength is desirable, the severe brittleness of a fully quenched steel usually makes it unusable without tempering. teMpering Subsequent to quenching, the same part can be reheated to a lower temperature (400–1300°F [200–700°C]), heat-soaked, and then allowed to cool slowly. This will cause some of the martensite to convert to ferrite and cementite, which reduces the strength somewhat but restores some ductility. A great deal of flexibility is possible in terms of tailoring the resulting combination of properties by varying time and temperature during the tempering process. The knowledgeable materials engineer or metallurgist can achieve a wide variety of properties to suit any application. Figure 2-12 also shows a stress-strain curve for the same steel after tempering.

Stress σ

annealing The quenching and tempering process is reversible by annealing. The part is heated above the critical temperature (as for quenching) but now allowed to cool slowly to room temperature. This restores the solution conditions and mechanical properties of the unhardened alloy. Annealing is often used even if no hardening has been previously done in order to eliminate any residual stresses and strains introduced by the forces applied in forming the part. It effectively puts the part back into a “relaxed” and soft state, restoring its original stress-strain curve as shown in Figure 2-12. norMalizing Many tables of commercial steel data indicate that the steel has been normalized after rolling or forming into its stock shape. Normalizing is similar to annealing but involves a shorter soak time at elevated temperature and a more rapid cooling rate. The result is a somewhat stronger and harder steel than a fully annealed one but one that is closer to the annealed condition than to any tempered condition. Surface (Case) Hardening When a part is large or thick, it is difficult to obtain uniform hardness within its interior by through hardening. An alternative is to harden only the surface, leaving the core soft. This also avoids the distortion associated with quenching a large, through-heated part. If the steel has sufficient carbon content, its surface can be heated, quenched, and tempered as would be done for through hardening. For low-carbon (mild) steels other techniques are needed to obtain a hardened condition. These involve heating the part in a special atmosphere rich in either carbon, nitrogen or both and then quenching it, a process called carburizing, nitriding, or cyaniding. In all situations, the result is a hard surface (i.e., case) on a soft core, referred to as being case-hardened. Carburizing heats low-carbon steel in a carbon monoxide gas atmosphere, causing the surface to take up carbon in solution. Nitriding heats low-carbon steel in a nitrogengas atmosphere and forms hard iron nitrides in the surface layers. Cyaniding heats the part in a cyanide salt bath at about 1500°F (800°C), and the low-carbon steel takes up both carbides and nitrides from the salt. For medium- and high-carbon steels no artificial atmosphere is needed, as the steel has sufficient carbon for hardening. Two methods are in common use. Flame hardening

quenched tempered

annealed E 1 ε FIGURE 2-12 Stress-Strain Curves for Annealed, Quenched, and Tempered Steel

46

2

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

passes an oxyacetylene flame over the surface to be hardened and follows it with a water jet for quenching. This results in a somewhat deeper hardened case than obtainable from the artificial-atmosphere methods. Induction hardening uses electric coils to rapidly heat the part surface, which is then quenched before the core can get hot. Case hardening by any appropriate method is a very desirable hardening treatment for many applications. It is often advantageous to retain the full ductility (and thus the toughness) of the core material for better energy absorption capacity while also obtaining high hardness on the surface in order to reduce wear and increase surface strength. Large machine parts such as cams and gears are examples of elements that can benefit more from case hardening than from through hardening, as heat distortion is minimized and the tough, ductile core can better absorb impact energy. Heat Treating Nonferrous Materials Some nonferrous alloys are hardenable and others are not. Some of the aluminum alloys can be precipitation hardened, also called age hardening. An example is aluminum alloyed with up to about 4.5% copper. This material can be hot-worked (rolled, forged, etc.) at a particular temperature and then heated and held at a higher temperature to force a random dispersion of the copper in the solid solution. It is then quenched to capture the supersaturated solution at normal temperature. The part is subsequently reheated to a temperature below the quenching temperature and held for an extended period of time while some of the supersaturated solution precipitates out and increases the material’s hardness. Other aluminum alloys, magnesium, titanium, and a few copper alloys are amenable to similar heat treatment. The strengths of the hardened aluminum alloys approach those of medium-carbon steels. Since all aluminum is about 1/3 the density of steel, the stronger aluminum alloys can offer better strength-to-weight ratios than low-carbon (mild) steels. Mechanical Forming and Hardening cold Working The mechanical working of metals at room temperature to change their shape or size will also work-harden them and increase their strength at the expense of ductility. Cold working can result from the rolling process in which metal bars are progressively reduced in thickness by being squeezed between rollers, or from any operation that takes the ductile metal beyond the yield point and permanently deforms it. Figure 2-13 shows the process as it affects the material’s stress-strain curve. As the load is increased from the origin at O beyond the yield point y to point B, a permanent set OA is introduced. If the load is removed at that point, the stored elastic energy is recovered and the material returns to zero stress at point A along a new elastic line BA parallel to the original elastic slope E. If the load is now reapplied and brought to point C, again yielding the material, the new stress-strain curve is ABCf. Note that there is now a new yield point y’ which is at a higher stress than before. The material has strain-hardened, increasing its yield strength and reducing its ductility. This process can be repeated until the material becomes brittle and fractures.

Chapter 2

Stress σ

MATERIALS AND PROCESSES

Stress σ

f B

Sy

47

C B

Sy

f

y'

y

(a)

E 1 O

A

1

O permanent set

Strain ε

elastic energy recovered

E

(b)

elastic energy recovered

A

D

permanent set

FIGURE 2-13 Strain Hardening a Ductile Material by Cold Working (a) First Working (b) Second Working

If significant plastic deformation is required for manufacture, such as in making deep-drawn metal pots or cylinders, it is necessary to cold form the material in stages and anneal the part between successive stages to avoid fracture. The annealing resets the material to more nearly the original ductile stress-strain curve, allowing further yielding without fracture. hot Working All metals have a recrystallization temperature below which the effects of mechanical working will be as described previously, e.g., cold worked. If the material is mechanically worked above its recrystallization temperature (hot working), it will tend to at least partially anneal itself as it cools. Thus, hot working reduces the strainhardening problem but introduces its own problems of rapid oxidation of the surface due to the high temperatures. Hot-rolled metals tend to have higher ductility, lower strength and poorer surface finish than cold-worked metals of the same alloy. Hot working does not increase the hardness of the material appreciably, though it can increase the strength by improving grain structure and aligning the “grain” of the metal with the final contours of the part. This is particularly true of forged parts. forging is an automation of the ancient art of blacksmithing. The blacksmith heats the part red-hot in the forge, then beats it into shape with a hammer. When it cools too much for forming, it is reheated and the process is repeated. Forging uses a series of hammer dies shaped to gradually form the hot metal into the final shape. Each stage’s die shape represents an achievable change in shape from the original ingot form to the final desired part shape. The part is reheated between blows from the hammer dies which are mounted in a forging press. The large forces required to plastically deform the hot metal require massive presses for parts of medium to large size. Machining operations are required to remove the large “flash” belt at the die parting line and to machine holes, mounting surfaces, etc. The surface finish of a forging is as rough as any hot-rolled part due to oxidation and decarburization of the heated metal. Virtually any wrought, ductile metal can be forged. Steel, aluminum, and titanium are commonly used. Forging has the advantage of creating stronger parts than casting or machining can. Casting alloys are inherently weaker in tension than wrought alloys. The hot forming of a wrought material into the final forged shape causes the material’s

Strain ε

48

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

2

FIGURE 2-14

Copyright © 2018 Robert L. Norton: All Rights Reserved

Forged Steel Crankshaft for a Diesel-Truck Engine

internal “flow lines” or “grain” to approximate the contours of the part, which can result in greater strength than if a stock shape’s flow lines were severed by machining to the final contour. Forgings are used in highly stressed parts, such as aircraft wing and fuselage structures, engine crankshafts and connecting rods, and vehicle suspension links. Figure 2-14 shows a forged truck crankshaft. In the cross section, the grain lines can be seen to follow the crankshaft’s contours. The high cost of the multiple dies needed for forged shapes makes it an impractical choice unless production quantities are large enough to amortize the tooling cost. extruSion is used principally for nonferrous metals (especially aluminum) as it typically uses steel dies. The usual die is a thick, hardened-tool-steel disk with a tapered “hole” or orifice ending in the cross-sectional shape of the finished part. A billet of the extrudate is heated to a soft state and then rammed at fairly high speed through the die, which is clamped in the machine. The billet flows, or extrudes, into the die’s shape. The process is similar to the making of macaroni. A long strand of the material in the desired cross section is extruded from the billet. The extrusion then passes through a water-spray cooling station. Extrusion is an economical way to obtain custom shapes of constant cross section since the dies are not very expensive to make. Dimensional control and surface finish are good. Extrusion is used to make aluminum mill shapes such as angles, channels, I-beams, and custom shapes for storm-door and -window frames, sliding-door frames, etc. The extrusions are cut and machined as necessary to assemble them into the finished product. Some extruded shapes are shown in Figure 2-15. 2.5 FIGURE 2-15 Extrusions - Courtesy of The Aluminum Extruders Council

COATINGS AND SURFACE TREATMENTS

Many coatings and surface treatments are available for metals. Some have the prime purpose of inhibiting corrosion while others are intended to improve surface hardness and wear resistance. Coatings are also used to change dimensions (slightly) and to alter physical properties such as reflectance, color, and resistivity. For example, piston rings are chrome-plated to improve wear resistance, fasteners are plated to reduce corrosion,

Chapter 2

MATERIALS AND PROCESSES

49

(a) Nonmetallic Coatings

Polymer

Oxide

Chemical Conversion

Anodizing

GlassCeramic

Phosphate

Chromate

(b) Metallic Coatings

Diffusion

Mechanical

Plating

Electroplating

Hot Dip

Electroless

Immersion

Spray

Deposition

Chemical Vapor

Vacuum

FIGURE 2-16 Coating Methods Available for Metals

Table 2-4

and automobile trim is chrome-plated for appearance and corrosion resistance. Figure 2-16 shows a chart of various types of coatings for machine applications. These divide into two major classes, metallic and nonmetallic, based on the type of coating, not substrate. Some of the classes divide into many subclasses. We will discuss only a few of these here. The reader is encouraged to seek more information from the references in the bibliography.

Galvanic Series of Metals in Seawater Least noble Magnesium Zinc Aluminum

Galvanic Action

Cadmium

When a coating of one metal is applied to another dissimilar metal, a galvanic cell may be created. All metals are electrolytically active to a greater or lesser degree and if sufficiently different in their electrolytic potential will create a battery in the presence of a conductive electrolyte such as seawater or even tap water. Table 2-4 lists some common metals ordered in terms of their galvanic action potential from the least noble (most electrolytically active) to the most noble (least active). Combinations of metals that are close to each other in the galvanic series, such as cast iron and steel, are relatively safe from galvanic corrosion. Combinations of metals far apart on this scale, such as aluminum and copper, will experience severe corrosion in an electrolyte or even in a moist environment.

Cast iron

In a conductive medium, the two metals become anode and cathode, with the lessnoble metal acting as the anode. The self-generated electrical current flow causes a loss of material from the anode and a deposition of material on the cathode. The less-noble metal gradually disappears. This problem occurs whenever two metals sufficiently far apart in the galvanic series are present in an electrically conductive medium. Thus, not only coatings but fasteners and mating parts must be made of metal combinations that will not create this problem.

Steel Stainless steel Lead Tin Nickel Brass Copper Bronze Monel Silver Titanium Graphite Gold Platinum Most noble

50

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Electroplating Electroplating involves the deliberate creation of a galvanic cell in which the part to be plated is the cathode and the plating material is the anode. The two metals are placed in an electrolyte bath and a direct current applied from anode to cathode. Ions of the plating material are driven to the plating substrate through the electrolyte and cover the part with a thin coating of the plating material. Allowance must be made for the plating thickness, which is controllable. Plating thickness is uniform except at sharp corners or in holes and crevices. The plating builds up on the outside corners and will not go into holes or narrow crevices. Thus, grinding may be necessary after plating to restore dimensions. Worn parts (or mistakes) can sometimes be repaired by plating on a coating of suitable material, then regrinding to dimension.

2

* It is interesting to note that chromium in the pure form is softer than hardened steel but when electroplated onto steel, it becomes harder than the steel substrate. Nickel and iron also increase their hardness when electroplated on metal substrates. The mechanism is not well understood, but it is believed that internal microstrains are developed in the plating process that harden the coating. The hardness of the plating can be controlled by changes in process conditions.

Steels, nickel- and copper-based alloys, as well as other metals are readily electroplatable. Two approaches are possible. If a more noble (less active) metal is plated onto the substrate, it can reduce the tendency to oxidize as long as the plating remains intact to protect the substrate from the environment. Tin, nickel, and chromium are often used to electroplate steel for corrosion resistance. Chrome plating also offers an increase in surface hardness to HRC 70, which is above that obtainable from many hardened alloy steels.* Unfortunately, any disruptions or pits in the plating can provide nodes for galvanic action if conductive media (such as rainwater) are present. Because the substrate is less noble than the plating, it becomes the sacrificial anode and rapidly corrodes. Electroplating with metals more noble than the substrate is seldom used for parts that will be immersed in water or other electrolytes. Alternatively, a less-noble metal can be plated onto the substrate to serve as a sacrificial anode which will corrode instead of the substrate. The most common example of this is zinc coating of steel, also called galvanizing. (Cadmium can be used instead of zinc and will last longer in saltwater or salt-air environments.) The zinc or cadmium coating will gradually corrode and protect the more noble steel substrate until the coating is used up, after which the steel will oxidize. Zinc coating can be applied by a process called “hot dipping” rather than by electroplating, which will result in a thicker and more protective coating recognizable by its “mother-of-pearl” appearance. Galvanizing is often applied by manufacturers to automobile body panels to inhibit corrosion. Sacrificial zinc anodes are also attached to aluminum outboard motors and aluminum boat hulls to short-circuit corrosion of the aluminum in seawater. A caution about electroplated coatings is that hydrogen embrittlement of the substrate can occur, causing significant loss of strength. Electroplated finishes should not be used on parts that are fatigue loaded. Experience has shown that electroplating severely reduces the fatigue strength of metals and can cause early failure. Electroless Plating Electroless plating puts a coating of nickel on the substrate without any electric current needed. The substrate “cathode” in this case (there is no anode) acts as a catalyst to start a chemical reaction that causes nickel ions in the electrolyte solution to be reduced and deposited on the substrate. The nickel coating also acts as a catalyst and keeps the

Chapter 2

MATERIALS AND PROCESSES

reaction going until the part is removed from the bath. Thus, relatively thick coatings can be developed. Coatings are typically between 0.001 in and 0.002 in thick. Unlike electroplating, the electroless nickel plate is completely uniform and will enter holes and crevices. The plate is dense and fairly hard at around 43 HRC. Other metals can also be electroless plated but nickel is most commonly used. Anodizing While aluminum can be electroplated (with difficulty), it is more common to treat it by anodizing. This process creates a very thin layer of aluminum oxide on the surface. The aluminum oxide coating is self-limiting in that it prevents atmospheric oxygen from further attacking the aluminum substrate in service. The anodized oxide coating is naturally colorless but dyes can be added to color the surface and provide a pleasing appearance in a variety of hues. This is a relatively inexpensive surface treatment with good corrosion resistance and negligible distortion. Titanium, magnesium, and zinc can also be anodized. A variation on conventional anodizing of aluminum is so-called “hard anodizing.” Since aluminum oxide is a ceramic material, it is naturally very hard and abrasion resistant. Hard anodizing provides a thicker (but not actually harder) coating than conventional anodizing and is often used to protect the relatively soft aluminum parts from wear in abrasive contact situations. The hardness of this surface treatment exceeds that of the hardest steel, and hard-anodized aluminum parts can be run against hardened steel though the somewhat abrasive aluminum oxide surface is not kind to the steel. Plasma-Sprayed Coatings A variety of very hard ceramic coatings can be applied to steel and other metal parts by a plasma-spray technique. The application temperatures are high, which limits the choice of substrate. The coatings as sprayed have a rough “orange-peel” surface finish, which requires grinding or polishing to obtain a fine finish. The main advantage is a surface with extremely high hardness and chemical resistance. However, the ceramic coatings are brittle and subject to chipping under mechanical or thermal shock. Chemical Coatings The most common chemical treatments for metals range from a phosphoric acid wash on steel (or chromatic acid on aluminum) that provides limited and short-term oxidation resistance, to paints of various types designed to give more lasting corrosion protection. Paints are available in a large variety of formulations for different environments and substrates. One-part paints give somewhat less protection than two-part epoxy formulations, but all chemical coatings should be viewed as only temporary protection against corrosion, especially when used on corrosion-prone materials such as steel. Baked enamel and porcelain finishes on steel have longer lives in terms of corrosion resistance, though they suffer from brittleness. New formulations of paints and protective coatings are continually being developed. The latest and best information will be obtained from vendors of these products.

51

52

MACHINE DESIGN

2.6 magnesium 6.5 (44.8)

2

aluminum 10.4 (71.8) gray cast iron 15 (104) brass, bronze 16 (110) titanium 16.5 (114) ductile cast iron

-

An Integrated Approach

-

Sixth Edition

GENERAL PROPERTIES OF METALS

The large variety of useful engineering materials can be confusing to the beginning engineer. There is not space enough in this book to deal with the topic of material selection in complete detail. Several references are provided in this chapter’s bibliography, which the reader is encouraged to use. Tables of mechanical property data are also provided for a limited set of materials in Appendix A of this book. Figure 2-17 shows the Young’s moduli for several engineering metals. The following sections attempt to provide some general information and guidelines for the engineer to help identify what types of materials might be suitable in a given design situation. It is expected that the practicing engineer will rely heavily on the expertise and help available from materials manufacturers in selecting the optimum material for each design. Many references are also published that list detailed property data for most engineering materials. Some of these references are listed in the bibliography to this chapter.

24 (166) stainless steel 27.5 (190) steel 30 (207) 0

10

20

0

70

140 210

30

Young's Modulus E Mpsi (GPa) FIGURE 2-17 Young's Moduli for Various Metals

Cast Iron Cast irons constitute a whole family of materials. Their main advantages are relatively low cost and ease of fabrication. Some are weak in tension compared to steels but, like most cast materials, have high compressive strengths. Their densities are slightly lower than steel at about 0.25 lb/in3 (6920 kg/m3). Most cast irons do not exhibit a linear stress-strain relationship below the elastic limit; they do not obey Hooke’s law. Their modulus of elasticity E is estimated by drawing a line from the origin through a point on the curve at 1/4 the ultimate tensile strength and is in the range of 14–25 Mpsi (97–172 MPa). Cast iron’s chemical composition differs from steel principally in its higher carbon content, being between 2 and 4.5%. The large amount of carbon, present in some cast irons as graphite, makes some of these alloys easy to pour as a casting liquid and also easy to machine as a solid. The most common means of fabrication is sand casting with subsequent machining operations. Cast irons are not easily welded, however. White caSt iron is a very hard and brittle material. It is difficult to machine and has limited uses, such as in linings for cement mixers where its hardness is needed. gray caSt iron is the most commonly used form of cast iron. Its graphite flakes give it its gray appearance and name. The ASTM grades gray cast iron into seven classes based on the minimum tensile strength in kpsi. Class 20 has a minimum tensile strength of 20 kpsi (138 MPa). The class numbers of 20, 25, 30, 35, 40, 50, and 60 then represent the tensile strength in kpsi. Cost increases with increasing tensile strength. This alloy is easy to pour, easy to machine, and offers good acoustical damping. This makes it the popular choice for machine frames, engine blocks, brake rotors and drums, etc. The graphite flakes also give it good lubricity and wear resistance. Its relatively low tensile strength recommends against its use in situations where large bending or fatigue loads are present, though it is sometimes used in low-cost engine crankshafts. It runs reasonably well against steel if lubricated. MalleaBle caSt iron has superior tensile strength to gray cast iron but does not wear as well. The tensile strength can range from 50 to 120 kpsi (345 to 827 MPa) depending on formulation. It is often used in parts where bending stresses are present.

Chapter 2

MATERIALS AND PROCESSES

nodular (ductile) caSt iron has the highest tensile strength of the cast irons, ranging from about 70 to 135 kpsi (480 to 930 MPa). The name nodular comes from the fact that its graphite particles are spheroidal in shape. Ductile cast iron has a higher modulus of elasticity (about 25 Mpsi {172 GPa}) than gray cast iron and exhibits a linear stress-strain curve. It is tougher, stronger, more ductile, and less porous than gray cast iron. It is the cast iron of choice for fatigue-loaded parts such as crankshafts, pistons, and cams. Cast Steels Cast steel is similar to wrought steel in terms of its chemical content, i.e., it has much less carbon than cast iron. The mechanical properties of cast steel are superior to cast iron but inferior to wrought steel. Its principal advantage is ease of fabrication by sand or investment (lost wax) casting. Cast steel is classed according to its carbon content into low carbon ( 0.5%). Alloy cast steels are also made containing other elements for high strength and heat resistance. The tensile strengths of cast steel alloys range from about 65 to 200 kpsi (450 to 1380 MPa). Wrought Steels The term “wrought” refers to all processes that manipulate the shape of the material without melting it. Hot rolling and cold rolling are the two most common methods used though many variants exist, such as wire drawing, deep drawing, extrusion, and coldheading. The common denominator is a deliberate yielding of the material to change its shape either at room or at elevated temperatures. hot-rolled Steel is produced by forcing hot billets of steel through sets of rollers or dies which progressively change their shape into I-beams, channel sections, angle irons, flats, squares, rounds, tubes, sheets, plates, etc. The surface finish of hot-rolled shapes is rough due to oxidation at the elevated temperatures. The mechanical properties are also relatively low because the material ends up in an annealed or normalized state unless deliberately heat-treated later. This is the typical choice for low-carbon structural steel members used for building- and machine-frame construction. Hot-rolled material is also used for machine parts that will be subjected to extensive machining (gears, cams, etc.) where the initial finish of the stock is irrelevant and uniform, non-cold-worked material properties are desired in advance of a planned heat treatment. A wide variety of alloys and carbon contents are available in hot-rolled form. cold-rolled Steel is produced from billets or hot-rolled shapes. The shape is brought to final form and size by rolling between hardened steel rollers or drawing through dies at room temperature. The rolls or dies burnish the surface and cold work the material, increasing its strength and reducing its ductility as was described previously in the section on mechanical forming and hardening. The result is a material with good surface finish and accurate dimensions compared to hot-rolled material. Its strength and hardness are increased at the expense of significant built-in strains, which can later be released during machining, welding, or heat treating, then causing distortion. Coldrolled shapes commonly available are sheets, strips, plates, round and rectangular bars, tubes, etc. Structural shapes, such as I-beams, are typically available only as hot-rolled.

53

54

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Steel Numbering Systems

2

* ASTM is the American Society for Testing and Materials, AISI is the American Iron and Steel Institute, and SAE is the Society of Automotive Engineers. AISI and SAE both use the same designations for steels.

Several steel numbering systems are in general use. The ASTM, AISI, and SAE* have devised codes to define the alloying elements and carbon content of steels. Table 2-5 lists some of the AISI/SAE designations for commonly used steel alloys. The first two digits indicate the principal alloying elements. The last two digits indicate the amount of carbon present, expressed in hundredths of a percent. ASTM and the SAE have developed a new Unified Numbering System for all metal alloys, which uses the prefix UNS followed by a letter and a 5-digit number. The letter defines the alloy category, F for cast iron, G for carbon and low-alloy steels, K for special-purpose steels, S for stainless steels, and T for tool steels. For the G series, the numbers are the same as the AISI/SAE designations in Table 2-5 with a trailing zero added. For example, SAE 4340 becomes UNS G43400. See reference 2 for more information on metal numbering systems. We will use the AISI/SAE designations for steels. plain carBon Steel is designated by a first digit of 1 and a second digit of 0, since no alloys other than carbon are present. The low-carbon steels are those numbered AISI 1005 to 1030, medium-carbon from 1035 to 1055, and high-carbon from 1060 to 1095. The AISI 11xx series adds sulphur, principally to improve machinability. These

Table 2-5

AISI/SAE Designations of Steel Alloys A partial list - other alloys are available - consult the manufacturers

Type

AISI/SAE Series

Principal Alloying Elements

Carbon Steels Plain

10xx

Carbon

Free-cutting

11xx

Carbon plus Sulphur (resulphurized)

13xx

1.75% Manganese

15xx

1.00 to 1.65% Manganese

23xx

3.50% Nickel

25xx

5.00% Nickel

31xx

1.25% Nickel and 0.65 or 0.80% Chromium

33xx

3.50% Nickel and 1.55% Chromium

40xx

0.25% Molybdenum

Alloy Steels Manganese Nickel Nickel-Chrome Molybdenum

44xx

0.40 or 0.52% Molybdenum

Chrome-Moly

41xx

0.95% Chromium and 0.20% Molybdenum

Nickel-Chrome-Moly

43xx

1.82% Nickel, 0.50 or 0.80% Chromium, and 0.25% Molybdenum

47xx

1.45% Nickel, 0.45% Chromium, and 0.20 or 0.35% Molybdenum

46xx

0.82 or 1.82% Nickel and 0.25% Molybdenum

48xx

3.50% Nickel and 0.25% Molybdenum

50xx

0.27 to 0.65% Chromium

51xx

0.80 to 1.05% Chromium

52xx

1.45% Chromium

61xx

0.60 to 0.95% Chromium and 0.10 to 0.15% Vanadium minimum

Nickel-Moly Chrome

Chrome-Vanadium

Chapter 2

MATERIALS AND PROCESSES

are called free-machining steels and are not considered alloy steels as the sulphur does not improve the mechanical properties and also makes it brittle. The ultimate tensile strength of plain carbon steel can vary from about 60 to 150 kpsi (414 to 1034 MPa) depending on heat treatment. alloy SteelS have various elements added in small quantities to improve the material’s strength, hardenability, temperature resistance, corrosion resistance, and other properties. Any level of carbon can be combined with these alloying elements. Chromium is added to improve strength, ductility, toughness, wear resistance, and hardenability. Nickel is added to improve strength without loss of ductility, and it also enhances case hardenability. Molybdenum, used in combination with nickel and/or chromium, adds hardness, reduces brittleness, and increases toughness. Many other alloys in various combinations, as shown in Table 2-5, are used to achieve specific properties. Specialty steel manufacturers are the best source of information and assistance for the engineer trying to find the best material for any application. The ultimate tensile strength of alloy steels can vary from about 80 to 300 kpsi (550 to 2070 MPa), depending on its alloying elements and heat treatment. Appendix A contains tables of mechanical property data for a selection of carbon and alloy steels. Figure 2-18 shows approximate ultimate tensile strengths of some normalized carbon and alloy steels and Figure 2-19 shows engineering stress-strain curves from tensile tests of three steels. tool SteelS are medium- to high-carbon alloy steels especially formulated to give very high hardness in combination with wear resistance and sufficient toughness to resist the shock loads experienced in service as cutting tools, dies and molds. There is a very large variety of tool steels available. Refer to the bibliography and to manufacturers’ literature for more information. StainleSS SteelS are alloy steels containing at least 10% chromium and offer much improved corrosion resistance over plain or alloy steels, though their name should not be taken too literally. Stainless steels will stain and corrode (slowly) in severe environments such as seawater. Some stainless-steel alloys have improved resistance to high tempera-

55

AISI # 4340 4140 1095 6150 9255 3140 1050 1040 1030 111 1020 1015 Kpsi 0 MPa 0

100

200

700

1400

tensile strength FIGURE 2-18 Approximate Ultimate Tensile Strengths of Some Normalized Steels

FIGURE 2-19 Tensile Test Stress-Strain Curves of Three Steel Alloys (From Fig. 5.16, p. 160, in N. E. Dowling, Mechanical Behavior of Materials, Prentice-Hall, Englewood Cliffs, N.J., 1993, with permission)

56

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

ture. There are four types of stainless steel, called martensitic, ferritic, austenitic, and precipitation hardening.

2

Martensitic stainless steel contains 11.5 to 15% Cr and 0.15 to 1.2% C, is magnetic, can be hardened by heat treatment, and is commonly used for cutlery. Ferritic stainless steel has over 16% Cr and a low carbon content, is magnetic, soft, and ductile, but is not heat treatable though its strength can be increased modestly by cold working. It is used for deep-drawn parts, such as cookware, and has better corrosion resistance than the martensitic SS. The ferritic and martensitic stainless steels are both called 400 series stainless steel. Austenitic stainless steel is alloyed with 17 to 25% chromium and 10 to 20% nickel. It has better corrosion resistance due to the nickel, is nonmagnetic, and has excellent ductility and toughness. It cannot be hardened except by cold working. It is classed as 300 series stainless steel. Precipitation-hardening stainless steels are designated by their alloy percentages followed by the letters PH, as in 17-4 PH, which contains 17% chromium and 4% nickel. These alloys offer high strength and high temperature and corrosion resistance. The 300 series stainless steels are very weldable but the 400 series are less so. All grades of stainless steel have poorer heat conductivity than regular steel and many of the stainless alloys are difficult to machine. All stainless steels are significantly more expensive than regular steel. See Appendix A for mechanical property data.

alloy 7075-T6

Aluminum

2014-T6 2024-T4 6061-T6 6063-T6 1100-H18 1100-0 0 kpsi

50

0 MPa 345

100 690

tensile strength FIGURE 2-20 Ultimate Tensile Strengths of Some Aluminum Alloys

Aluminum is the most widely used nonferrous metal, being second only to steel in world consumption. Aluminum is produced in both “pure” and alloyed forms. Aluminum is commercially available up to 99.8% pure. The most common alloying elements are copper, silicon, magnesium, manganese, and zinc, in varying amounts up to about 5%. The principal advantages of aluminum are its low density, good strength-to-weight ratio (SWR), ductility, excellent workability, castability, and weldability, corrosion resistance, high conductivity, and reasonable cost. Compared to steel it is 1/3 as dense (0.10 lb/ in3 versus 0.28 lb/in3), about 1/3 as stiff (E = 10.3 Mpsi [71 GPa] versus 30 Mpsi [207 GPa]), and generally less strong. If you compare the strengths of low-carbon steel and pure aluminum, the steel is about three times as strong. Thus, the specific strength is approximately the same in that comparison. However, pure aluminum is seldom used in engineering applications. It is too soft and weak. Pure aluminum’s principal advantages are its bright finish and good corrosion resistance. It is used mainly in decorative applications. The aluminum alloys have significantly greater strengths than pure aluminum and are used extensively in engineering, with the aircraft and automotive industries among the largest users. The higher-strength aluminum alloys have tensile strengths in the 70 to 90 kpsi (480 to 620 MPa) range, and yield strengths about twice that of mild steel. They compare favorably to medium-carbon steels in specific strength. Aluminum competes successfully with steel in some applications, though few materials can beat steel if very high strength is needed. See Figure 2-20 for tensile strengths of some aluminum alloys. Figure 2-21 shows tensile-test engineering stress-strain curves for three aluminum alloys. Aluminum’s strength is reduced at low temperatures as well as at elevated temperatures.

Chapter 2

MATERIALS AND PROCESSES

57

FIGURE 2-21 Tensile Test Stress-Strain Curves of Three Aluminum Alloys (From Fig. 5.17, p. 160, in N. E. Dowling, Mechanical Behavior of Materials, Prentice-Hall, Englewood Cliffs, N.J., 1993, with permission)

Some aluminum alloys are hardenable by heat treatment and others by strain hardening or precipitation and aging. High-strength aluminum alloys are about 1.5 times harder than soft steel, and surface treatments such as hard anodizing can bring the surface to a condition harder than the hardest steel. Aluminum is among the most easily worked of the engineering materials, though it tends to work harden. It casts, machines, welds,* and hot and cold forms† easily. It can also be extruded. Alloys are specially formulated for both sand and die casting as well as for wrought and extruded shapes and for forged parts. Wrought-aluMinuM alloyS are available in a wide variety of stock shapes such as I-beams, angles, channels, bars, strip, sheet, rounds, and tubes. Extrusion allows relatively inexpensive custom shapes as well. The Aluminum Association numbering system for alloys is shown in Table 2-6. The first digit indicates the principal alloying element and defines the series. Hardness is indicated by a suffix containing a letter and up to 3 numbers as defined in the table. The most commonly available and most-used aluminum alloys in machine-design applications are the 2000 and 6000 series. The oldest aluminum alloy is 2024, which contains 4.5% copper, 1.5% magnesium, and 0.8% manganese. It is among the most machinable of the aluminum alloys and is heat treatable. In the higher tempers, such as -T3 and -T4, it has a tensile strength approaching 70 kpsi (483 MPa), which also makes it one of the strongest of the aluminum alloys. It also has high fatigue strength. However, it has poor weldability and formability compared to the other aluminum alloys. The 6061 alloy contains 0.6% silicon, 0.27% copper, 1.0% manganese, and 0.2% chromium. It is widely used in structural applications because of its excellent weldability. Its strength is about 40 to 45 kpsi (276 to 310 MPa) in the higher tempers. It has lower fatigue strength than 2024 aluminum. It is easily machined and is a popular alloy for extrusion, which is a hot-forming process. The 7000 series is called aircraft aluminum and is used mostly in airframes. These are the strongest alloys of aluminum with tensile strengths up to 98 kpsi (676 MPa) and

*

The heat of welding causes localized annealing, which can remove the desirable strengthening effects of cold work or heat treatment in any metal. †

Some aluminum alloys will cold-work when formed to the degree that trying to bend them again (without first annealing) will cause fractures. Some bicycle racers prefer steel frames over aluminum despite their added weight because, once an aluminum frame is bent in a fall, it cannot be straightened without cracking. Damaged steel tube frames can be straightened and reused.

58

MACHINE DESIGN

Table 2-6

-

An Integrated Approach

-

Sixth Edition

Aluminum Association Designations of Aluminum Alloys A partial list - other alloys are available - consult the manufacturers

2 Series Alloys

Major Alloying Elements

Secondary

1xxx

Commercially pure (99%)

None

2xxx

Copper (Cu)

Mg, Mn, Si

3xxx

Manganese (Mn)

Mg, Cu

4xxx

Silicon (Si)

None

5xxx

Magnesium (Mg)

Mn, Cr

6xxx

Magnesium and Silicon

Cu, Mn

7xxx

Zinc (Zn)

Mg, Cu, Cr

Hardness Designations xxxx-F

As fabricated

xxxx-O

Annealed

xxxx-Hyyy

Work hardened

xxxx-Tyyy

Thermal/age hardened

the highest fatigue strength of about 22 kpsi (152 MPa) @ 108 cycles. Some alloys are also available in an alclad form that bonds a thin layer of pure aluminum to one or both sides to improve corrosion resistance. caSt-aluMinuM alloyS are differently formulated than the wrought alloys. Some of these are hardenable but their strength and ductility are less than those of the wrought alloys. Alloys are available for sand casting, die casting, or investment casting. See Appendix A for mechanical properties of wrought- and cast-aluminum alloys. Titanium Though discovered as an element in 1791, commercially produced titanium has been available only since the 1940s, so it is among the newest of engineering metals. Titanium can be the answer to an engineer’s prayer in some cases. It has an upper servicetemperature limit of 1200 to 1400°F (650 to 750°C), weighs half as much as steel (0.16 lb/in3 [4429 kg/m3]), and is as strong as a medium-strength steel (135 kpsi [930 MPa] typical). Its Young’s modulus is 16 to 18 Mpsi (110 to 124 GPa), or about 60% that of steel. Its specific strength approaches that of the strongest alloy steels and exceeds that of medium-strength steels by a factor of 2. Its specific stiffness is greater than that of steel, making it as good or better in limiting deflections. It is also nonmagnetic. Titanium is very corrosion resistant and is nontoxic, allowing its use in contact with acidic or alkaline foodstuffs and chemicals, and in the human body as replacement heart valves and hip joints, for example. Unfortunately, it is expensive compared to aluminum and steel. It finds much use in the aerospace industry, especially in military aircraft structures and in jet engines, where strength, light weight, and high temperature and corrosion resistance are all required.

Chapter 2

MATERIALS AND PROCESSES

Titanium is available both pure and alloyed with combinations of aluminum, vanadium, silicon, iron, chromium, and manganese. Its alloys can be hardened and anodized. Limited stock shapes are available commercially. It can be forged and wrought, though it is quite difficult to cast, machine, and cold form. Like steel and unlike most other metals, some titanium alloys exhibit a true endurance limit, or leveling off of the fatigue strength, beyond about 106 cycles of repeated loading, as shown in Figure 2-10. See Appendix A for mechanical property data. Magnesium Magnesium is the lightest of commercial metals but is relatively weak. The tensile strengths of its alloys are between 10 and 50 kpsi (69 and 345 MPa). The most common alloying elements are aluminum, manganese, and zinc. Because of its low density (0.065 lb/in3 [1800 kg/m3]), its specific strength approaches that of aluminum. Its Young’s modulus is 6.5 Mpsi (45 GPa) and its specific stiffness exceeds those of aluminum and steel. It is very easy to cast and machine but is more brittle than aluminum and thus is difficult to cold form. It is nonmagnetic and has fair corrosion resistance, better than steel, but not as good as aluminum. Some magnesium alloys are hardenable, and all can be anodized. It is the most active metal on the galvanic scale and cannot be combined with most other metals in a wet environment. It is also extremely flammable, especially in powder or chip form, and its flame cannot be doused with water. Machining requires flooding with oil coolant to prevent fire. It is roughly twice as costly per pound as aluminum. Magnesium is used where light weight is of paramount importance such as in castings for chain-saw housings and other hand-held items. See Appendix A for mechanical property data. Copper Alloys Pure copper is soft, weak, and malleable and is used primarily for piping, flashing, electrical conductors (wire) and motors. It cold works readily and can become brittle after forming, requiring annealing between successive draws. Many alloys are possible with copper. The most common are brasses and bronzes which themselves are families of alloys. Brasses, in general, are alloys of copper and zinc in varying proportions and are used in many applications, from artillery shells and bullet shells to lamps and jewelry. Bronzes were originally defined as alloys of copper and tin, but now also include alloys containing no tin, such as silicon bronze and aluminum bronze, so the terminology is somewhat confusing. Silicon bronze is used in marine applications such as ship propellers. Beryllium copper is neither brass nor bronze and is the strongest of the alloys, with strengths approaching those of alloy steels (200 kpsi [1380 MPa]). It is often used in springs that must be nonmagnetic, carry electricity, or exist in corrosive environments. Phosphor bronze is also used for springs but unlike beryllium copper, it cannot be bent along the grain or heat treated.

59

60

2

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Copper and its alloys have excellent corrosion resistance and are nonmagnetic. All copper alloys can be cast, hot or cold formed, and machined, but pure copper is difficult to machine. Some alloys are heat treatable and all will work harden. The Young’s modulus of most copper alloys is about 17 Mpsi (117 GPa) and their weight density is slightly higher than that of steel at 0.31 lb/in3 (8580 kg/m3). Copper alloys are expensive compared to other structural metals. See Appendix A for mechanical property data. 2.7

GENERAL PROPERTIES OF NONMETALS

The use of nonmetallic materials has increased greatly in the last 50 years. The usual advantages sought are light weight, corrosion resistance, temperature resistance, dielectric strength, and ease of manufacture. Cost can range from low to high compared to metals depending on the particular nonmetallic material. There are three general categories of nonmetals of general engineering interest: polymers (plastics), ceramics, and composites. Polymers have a wide variety of properties, principally low weight, relatively low strength and stiffness, good corrosion and electrical resistance, and relatively low cost per unit volume. Ceramics can have extremely high compressive (but not tensile) strengths, high stiffness, high temperature resistance, high dielectric strength (resistance to electrical current), high hardness, and relatively low cost per unit volume. Composites can have almost any combination of properties you want to build into them, including the highest specific strengths obtainable from any materials. Composites can be low or very high in cost. We will briefly discuss nonmetals and some of their applications. Space does not permit a complete treatment of these important classes of materials. The reader is directed to the bibliography for further information. Appendix A also provides some mechanical property data for polymers. Polymers The word polymers comes from poly = many and mers = molecules. Polymers are long-chain molecules of organic materials or carbon-based compounds. (There is also a family of silicon-based polymeric compounds.) The source of most polymers is oil or coal, which contains the carbon or hydrocarbons necessary to create the polymers. While there are many natural polymer compounds (for example, wax, rubber, proteins), most polymers used in engineering applications are man-made. Their properties can be tailored over a wide range by copolymerization with other compounds or by alloying two or more polymers together. Mixtures of polymers and inorganic materials such as talc or glass fiber are also common. Because of their variety, it is difficult to generalize about the mechanical properties of polymers, but compared to metals they have low density, low strength, low stiffness, nonlinear elastic stress-strain curves as shown in Figure 2-22 (with a few exceptions), low hardness, excellent electrical and corrosion resistance, and ease of fabrication. Their apparent moduli of elasticity vary widely from about 10 kpsi (69 MPa) to about 400 kpsi (2.8 GPa), all much less stiff than any metals. Their ultimate tensile strengths range from about 4 kpsi (28 MPa) for the weakest unfilled polymer to about 22 kpsi (152 MPa) for the strongest glass-filled polymers. The specific gravities of most polymers range from about 0.95 to 1.8 compared to about 2 for magnesium, 3 for aluminum, 8 for steel, and

Chapter 2

MATERIALS AND PROCESSES

61

FIGURE 2-22 Tensile Test Stress-Strain Curves of Three Thermoplastic Polymers (From Fig. 5.18, p. 161, in N. E. Dowling, Mechanical Behavior of Materials, Prentice-Hall, Englewood Cliffs, N.J., 1993, with permission)

13 for lead. So, even though the absolute strengths of polymers are low, their specific strengths are respectable due to their low densities. Polymers are divided into two classes, thermoplastic and thermosets. Thermoplastic polymers can be repeatedly melted and solidified, though their properties can degrade due to the high melt temperatures. Thermoplastics are easy to mold and their rejects or leftovers can be reground and remolded. Thermosetting polymers become cross-linked when first heated and will burn, not melt, on reheating. Cross-linking creates connections (like the rungs of a ladder) between the long-chain molecules that wind and twist through a polymer. These cross-connections add strength and stiffness. Another division among polymers can be made between filled and unfilled compounds. The fillers are usually inorganic materials, such as carbon black, graphite, talc, chopped glass fibers, and metal powders. Fillers are added to both thermoplastic and thermosetting resins, though they are more frequently used in the latter. These filled compounds have superior strength, stiffness, and temperature resistance over that of the raw polymers but are more difficult to mold and to fabricate. A confusing array of polymers is available commercially. The confusion is increased by a proliferation of brand names for similar compounds made by different manufacturers. The generic chemical names of polymers tend to be long, complex, and hard to remember. In some cases a particular polymer brand name has been so widely used that it has become generic. Nylon, plexiglass, and fiberglass are examples. Learning the generic chemical names and associated brand names of the main families of engineering polymers will eliminate some of the confusion. Table 2-7 shows a number of important polymer families. The mechanical properties of a few of these that have significant engineering applications are included in Appendix A.

Table 2-7 Families of Polymers Thermoplastics Cellulosics Ethylenics Polyamides Polyacetals Polycarbonates Polyphenyline oxides Polysulfones Thermosets Aminos Elastomers Epoxies Phenolics Polyesters Silicones Urethanes

62

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Ceramics Ceramic materials are finding increasing application in engineering, and a great deal of effort is being devoted to the development of new ceramic compounds. Ceramics are among the oldest known engineering materials; clay bricks are ceramic materials. Though still widely used in building, clay is not now considered an engineering ceramic. Engineering ceramics are typically compounds of metallic and nonmetallic elements. They may be single oxides of a metal, mixtures of metallic oxides, carbides, borides, nitrides, or other compounds such as Al2O3, MgO, SiC, and Si3N4, for example. The principal properties of ceramic materials are high hardness and brittleness, high temperature and chemical resistance, high compressive strength, high dielectric strength, and potentially low cost and weight. Ceramic materials are too hard to be machined by conventional techniques and are usually formed by compaction of powder, then fired or sintered to form bonds between particles and increase their strength. The powder compaction can be done in dies or by hydrostatic pressure. Sometimes, glass powder is mixed with the ceramic and the result is fired to melt the glass and fuse the two together. Attempts are being made to replace traditional metals with ceramics in such applications as cast engine blocks, pistons, and other engine parts. The low tensile strength, porosity, and low fracture toughness of most ceramics can be problems in these applications. Plasma-sprayed ceramic compounds are often used as hard coatings on metal substrates to provide wear- and corrosion-resistant surfaces.

2

Composites Most composites are man-made, but some, such as wood, occur naturally. Wood is a composite of long cellulose fibers held together in a resinous matrix of lignin. Man-made composites are typically a combination of some strong, fibrous material such as glass, carbon, or boron fibers glued together in a matrix of resin such as epoxy or polyester. The fiberglass material used in boats and other vehicles is a common example of a glassfiber reinforced polyester (GFRP) composite. The directional material properties of a composite can be tailored to the application by arranging the fibers in different juxtapositions such as parallel, interwoven at random or particular angles, or wound around a mandrel. Custom composites are finding increased use in highly stressed applications such as airframes due to their superior strength-to-weight ratios compared to the common structural metals. Temperature and corrosion resistance can also be designed into some composite materials. These composites are typically neither homogeneous nor isotropic as was discussed in Section 2.3.

Table 2-8 Iron and Steel Strengths Form Theoretical

Sut kpsi (MPa) 2900 (20E3)

Whisker

1800 (12E3)

Fine wire

1400 (10E3)

Mild steel

60 (414)

Cast iron

40 (276)

It is interesting to note that if one calculates the theoretical strength of any “pure” elemental crystalline material based on the interatomic bonds of the element, the predicted strengths are orders of magnitude larger that those seen in any test of a “real” material, as seen in Table 2-8. The huge differences in actual versus theoretical strengths are attributed to disruptions of the atomic bonds due to crystal defects in the real material. That is, it is considered impossible to manufacture “pure anything” on any realistic superatomic scale. It is presumed that if we could make a “wire” of pure iron only one atom in diameter, it would exhibit its theoretical “super strength.” Crystal “whiskers” have been successfully made of some elemental materials and exhibit very high tensile strengths, which approach their theoretical values (Table 2-8).

Chapter 2

MATERIALS AND PROCESSES

Other empirical evidence for this theory comes from the fact that fibers of any material made in very small diameters exhibit much higher tensile strengths than would be expected from stress-strain tests of larger samples of the same material. Presumably, the very small cross sections are approaching a “purer” material state. For example, it is well known that glass has poor tensile strength. However, small-diameter glass fibers show much larger tensile strength than sheet glass, making them a practical (and inexpensive) fiber for use in boat hulls, which are subjected to large tensile stresses in use. Smalldiameter fibers of carbon and boron exhibit even higher tensile strengths than glass fiber, which explains their use in composites for spacecraft and military aircraft applications, where their relatively high cost is not a barrier. 2.8

SELECTING MATERIALS

One of the most important design decisions is the proper choice of material. Materials limit design and new materials are still being invented that open new design possibilities. It would help if there were a systematic way to select a material for an application. M. F. Ashby has proposed such an approach that plots various material properties against one another to form “materials selection charts.”[3] Materials can be roughly divided into six classes: metals, ceramics, polymers (solid or foam), elastomers, glasses, and composites (which include wood). Members of these classes and subclasses tend to cluster together on a plot of this type. Figure 2-23 shows such a chart that plots Young’s modulus against density, which is called specific stiffness. By drawing lines of constant slope on such a chart, one can see which materials possess similar properties. A line of specific stiffness E / ρ = C has been drawn in color on Figure 2-23 and shows that some woods have equivalent specific stiffness to steel and some other metals. The line also passes through the lower range of the engineering composites’ “bubble” indicating that fiberglass (GFRP) has about the same specific stiffness as wood and steel, while the nonreinforced thermoplastics, such as nylon and polyester, have lower specific stiffness. So if you seek the stiffest/lightest material, you want to move up and to the left on the chart. Other lines are shown that have slopes equal to En / ρ = C where n is a fraction such as 1/2 or 1/3. These represent loading situations, such as beams in bending, for which the parameter of interest is a nonlinear function of specific stiffness. Since the chart is a log-log plot, exponential functions also plot as straight lines allowing simple comparisons to be made. Figure 2-24 shows a chart of strength versus density (called specific strength) for a number of materials. In this chart, the particular material strength used varies with the material depending on its character. For example, ductile metals and polymers show their yield strength, brittle ceramics their crushing compressive strength, and elastomers their tear strength. The vertical elongation of a material’s “bubble” indicates the range of strength values that can obtain due to thermal or work hardening, alloying elements, etc. The colored line drawn on the chart represents a particular value of specific strength or σ / ρ = C and shows that the strength-to-weight ratio of some woods are as good as high-strength steel and better than most other metals. It should be no surprise that wood is a popular material in building construction. Note also the high specific strength of engineering ceramics. Unfortunately, their tensile strengths are at best only about 10% of these compressive strengths, which is why you seldom see them used in structures where tensile stresses are commonly encountered.

63

64

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

2

E/ρ=C

FIGURE 2-23

Copyright © 2018 Robert L. Norton: All Rights Reserved

Young's Modulus Plotted Against Density for Engineering Materials (Adapted from Fig. 4-3, p. 37 in M. F. Ashby, Materials Selection in Mechanical Design, 2ed, Elsevier, 2016)

Ashby’s book[3] is a very useful reference for the practicing engineer. It has dozens of charts of the type shown here that plot various properties against one another in a manner that enhances their comparison and develops good understanding. 2.9

SUMMARY

There are many different kinds of material strengths. It is important to understand which ones are important in particular loading situations. The most commonly measured and reported strengths are the ultimate tensile strength Sut and the tensile yield strength

Chapter 2

MATERIALS AND PROCESSES

65

σ/ρ=C

FIGURE 2-24

Copyright © 2018 Robert L. Norton: All Rights Reserved

Strength Plotted Against Density for Engineering Materials (Adapted from Fig. 4-4, p. 39 in M. F. Ashby, Materials Selection in Mechanical Design, 2ed, Elsevier, 2016)

Sy. The Sut indicates the largest stress that the material will accept before fracture, and Sy indicates the stress beyond which the material will take a permanent set. Many materials have compressive strengths about equal to their tensile strengths and are called even materials. Most wrought metals are in the even category. Some materials have significantly different compressive and tensile strengths and these are called uneven materials. Cast metals are usually in the uneven category, with compressive strengths much greater than their tensile strengths. The shear strengths of even materials tend to

66

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

be about half their tensile strengths, while shear strengths of uneven materials tend to be between their tensile and compressive strengths.

2

One or more of these strengths may be of interest when the loading is static. If the material is ductile, then Sy is the usual criterion of failure, as a ductile material is capable of significant distortion before fracture. If the material is brittle, as are most cast materials, then the Sut is a more interesting parameter, because the material will fracture before any significant yielding distortion takes place. Yield strength values are nevertheless reported for brittle materials, but are usually calculated based on an arbitrary, small value of strain rather than on any measured yielding of the specimen. Chapter 5 deals with the mechanisms of material failure for both ductile and brittle materials in more detail than does this chapter. The tensile test is the most common measure of these static strength parameters. The stress-strain curve (σ–ε) generated in this test is shown in Figure 2-2. The socalled engineering σ–ε curve differs from the true σ–ε curve due to the reduction in area of a ductile test specimen during the failure process. Nevertheless, the engineering σ–ε curve is the standard used to compare materials, since the true σ–ε curve is more difficult to generate. The slope of the σ–ε curve in the elastic range, called Young’s modulus or the modulus of elasticity E, is a very important parameter as it defines the material’s stiffness or resistance to elastic deflection under load. If you are designing to control deflections as well as stresses, the value of E may be of more interest than the material’s strength. While various alloys of a given base material may vary markedly in terms of their strengths, they will have essentially the same E. If deflection is the prime concern, a low-strength alloy is as good as a high-strength one of the same base material. When the loading on the part varies with time it is called dynamic or fatigue loading. Then the static strengths do not give a good indication of failure. Instead, the fatigue strength is of more interest. This strength parameter is measured by subjecting a specimen to dynamic loading until it fails. Both the magnitude of the stress and the number of cycles of stress at failure are reported as the strength criterion. The fatigue strength of a given material will always be lower than its static strength, and often is less than half its Sut. Chapter 6 deals with the phenomenon of fatigue failure of materials in more detail than does this chapter. Other material parameters of interest to the machine designer are resilience, which is the ability to absorb energy without permanent deformation, and toughness or the ability to absorb energy without fracturing (but with permanent deformation). Homogeneity is the uniformity of a material throughout its volume. Many engineering materials, especially metals, can be assumed to be macroscopically homogeneous even though at a microscopic level they are often heterogeneous. Isotropism means having properties that are the same regardless of direction within the material. Many engineering materials are reasonably isotropic in the macro and are assumed so for engineering purposes. However, other useful engineering materials, such as wood and composites, are neither homogeneous nor isotropic and their strengths must be measured separately in different directions. Hardness is important in wear resistance and is also related to strength. Heat treatment, both through and surface, as well as cold working, can increase the hardness and strength of some materials.

Chapter 2

MATERIALS AND PROCESSES

Important Equations Used in This Chapter See the referenced sections for information on the proper use of these equations. Axial tensile stress (Section 2.1):

P Ao

(2.1a)

l − lo lo

(2.1b)

σ= Axial tensile strain (Section 2.1):

ε=

Modulus of elasticity (Young’s modulus) (Section 2.1):

σ ε

(2.2)

E 2 (1 + ν )

(2.4)

E= Modulus of rigidity (Section 2.1):

G= Ultimate shear strength (Section 2.1):

steels:

Sus ≅ 0.80 Sut

other ductile metals:

Sus ≅ 0.75Sut

(2.5b)

Shear yield strength (Section 2.1):

U=

ε

∫0

σ dε

(2.6)

Modulus of resilience (Section 2.1):

UR ≅

2 1 Sy 2 E

(2.7)

Modulus of toughness (Section 2.1):

 S y + Sut  UT ≅   ε f 2 

(2.8)

Arithmetic mean (Section 2.2):

µ=

1 n

n

∑ xi

(2.9b)

i =1

Standard deviation (Section 2.2):

Sd =

1 n-1

n

∑ ( xi − µ )

2

(2.9c)

i =1

Ultimate tensile strength as a function of Brinell hardness (Section 2.4):

Sut ≅ 500 HB ± 30 HB

psi

Sut ≅ 3.45 HB ± 0.2 HB

MPa

(2.10)

67

68

MACHINE DESIGN

2.10

2

-

An Integrated Approach

-

Sixth Edition

REFERENCES

1

E. B. Haugen and P. H. Wirsching, “Probabilistic Design.” Machine Design, v. 47, nos. 10-14, Penton Publishing, Cleveland, Ohio, 1975.

2

H. E. Boyer and T. L. Gall, eds. Metals Handbook. Vol. 1. American Society for Metals: Metals Park, Ohio, 1985.

3

M. F. Ashby, Materials Selection in Mechanical Design, 2ed., Butterworth and Heinemann, 1999.

2.11

WEB REFERENCES The web is a useful resource for up-to-date material property information at these and other sites that can be found with a search engine.

http://www.matweb.com

Properties data sheets for over 41,000 metals, plastics, ceramics, and composites. http://metals.about.com

Material properties and data.   

2.12

BIBLIOGRAPHY

For general information on materials, consult the following:

Metals & Alloys in the Unified Numbering System. 6th ed. ASTM/SAE: Phila., Pa., 1994. H. E. Boyer, ed. Atlas of Stress-Strain Curves. Amer. Soc. for Metals: Metals Park, Ohio, 1987. Brady, ed. Materials Handbook. 13th ed. McGraw-Hill: New York, 1992. K. Budinski, Engineering Materials: Properties and Selection. 4th ed. Reston-Prentice-Hall: Reston, Va., 1992. M. M. Farag, Selection of Materials and Manufacturing Processes for Engineering Design. Prentice-Hall International: Hertfordshire, U.K., 1989. I. Granet, Modern Materials Science. Reston-Prentice-Hall: Reston, Va., 1980. H. W. Pollack, Materials Science and Metallurgy. 2nd ed. Reston-Prentice-Hall: Reston, Va., 1977. M. M. Schwartz, ed. Handbook of Structural Ceramics. McGraw-Hill: New York, 1984. S. P. Timoshenko, History of Strength of Materials. McGraw-Hill: New York, 1983. L. H. V. Vlack, Elements of Material Science and Engineering. 6th ed. Addison-Wesley: Reading, Mass., 1989. For specific information on material properties, consult the following:

H. E. Boyer and T. L. Gall, ed. Metals Handbook. Vol. 1. American Society for Metals: Metals Park, Ohio, 1985. R. Juran, ed. Modern Plastics Encyclopedia. McGraw-Hill: New York, 1988. J. D. Lubahn and R. P. Felgar, Plasticity and Creep of Metals. Wiley: New York, 1961. U. S. Department of Defense. Metallic Materials and Elements for Aerospace Vehicles and Structures MIL-HDBK-5H, 1998.

Chapter 2

MATERIALS AND PROCESSES

69

For information on failure of materials, consult the following:

J. A. Collins, Failure of Materials in Mechanical Design. Wiley: New York, 1981. N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., 1992. R. C. Juvinall, Stress, Strain and Strength. McGraw-Hill: New York, 1967. For information on plastics and composites, consult the following:

ASM, Engineered Materials Handbook: Composites. Vol. 1. American Society for Metals: Metals Park, Ohio, 1987. ASM, Engineered Materials Handbook: Engineering Plastics. Vol. 2. American Society for Metals: Metals Park, Ohio, 1988. Harper, ed. Handbook of Plastics, Elastomers and Composites. 2nd ed. McGrawHill: New York, 1990. J. E. Hauck, “Long-Term Performance of Plastics.” Materials in Design Engineering, pp. 113-128, November, 1965. M. M. Schwartz, Composite Materials Handbook. McGraw-Hill: New York, 1984. For information on manufacturing processes, see:

R. W. Bolz, Production Processes: The Productivity Handbook. Industrial Press: New York, 1974. J. A. Schey, Introduction to Manufacturing Processes. McGraw-Hill: New York, 1977.

2.13 2-1

PROBLEMS Figure P2-1 shows stress-strain curves for three failed tensile-test specimens. All are plotted on the same scale. (a) (b) (c) (d) (e)

Characterize each material as brittle or ductile. Which is the stiffest? Which has the highest ultimate strength? Which has the largest modulus of resilience? Which has the largest modulus of toughness?

2-2 Determine an approximate ratio between the yield strength and ultimate strength for each material shown in Figure P2-1. 2-3 Which of the steel alloys shown in Figure 2-19 would you choose to obtain (a) (b) (c) (d)

Maximum strength Maximum modulus of resilience Maximum modulus of toughness Maximum stiffness

2-4 Which of the aluminum alloys shown in Figure 2-21 would you choose to obtain (a) (b) (c) (d)

Maximum strength Maximum modulus of resilience Maximum modulus of toughness Maximum stiffness

2-5 Which of the thermoplastic polymers shown in Figure 2-22 would you choose in order to obtain (a) (b) (c) (d)

Maximum strength Maximum modulus of resilience Maximum modulus of toughness Maximum stiffness

Table P2-0 Topic/Problem Matrix 2.1 Material Properties 2-1, 2-2, 2-3, 2-4, 2-5, 2-6, 2-7, 2-8, 2-9, 2-10, 2-11, 2-12, 2-18, 2-19, 2-20, 2-21, 2-22, 2-23 2.4 Hardness 2-13, 2-14, 2-41, 2-42, 2-43, 2-44 2.6 General Properties 2-15, 2-16, 2-17, 2-24, 2-25, 2-26 2.8 Selecting Materials 2-37, 2-38, 2-39, 2-40

70

MACHINE DESIGN

*

Answers to these problems are provided in Appendix D.

*2-6

2 Stress σ

-

An Integrated Approach

-

Sixth Edition

A metal has a strength of 414 MPa at its elastic limit and the strain at that point is 0.002. Assume the test specimen is 12.8-mm dia and has a 50-mm gage length. What is its modulus of elasticity? What is the strain energy at the elastic limit? Can you define the type of metal based on the given data?

2-7

A metal has a strength of 41.2 kpsi (284 MPa) at its elastic limit and the strain at that point is 0.004. Assume the test specimen is 0.505-in dia and has a 2-in gage length. What is the strain energy at the elastic limit? Can you define the type of metal based on the given data?

*2-8

A metal has a strength of 134 MPa at its elastic limit and the strain at that point is 0.003. What is its modulus of elasticity? Assume the test specimen is 12.8-mm dia and has a 50-mm gage length. What is its modulus of elasticity? What is the strain energy at the elastic limit? Can you define the type of metal based on the given data?

*2-9

A metal has a strength of 100 kpsi (689 MPa) at its elastic limit and the strain at that point is 0.006. What is its modulus of elasticity? What is the strain energy at the elastic limit? Assume the test specimen is 0.505-in dia and has a 2-in gage length. Can you define the type of metal based on the given data?

Strain ε (a )

2-10 A material has a yield strength of 689 MPa at an offset of 0.6% strain. What is its modulus of resilience?

Stress σ

2-11 A material has a yield strength of 60 kpsi (414 MPa) at an offset of 0.2% strain. What is its modulus of resilience? *2-12

A steel has a yield strength of 414 MPa, an ultimate tensile strength of 689 MPa, and an elongation at fracture of 15%. What is its approximate modulus of toughness? What is its approximate modulus of resilience?

2-13 The Brinell hardness of a steel specimen was measured to be 250 HB. What is the material’s approximate tensile strength? What is its hardness on the Vickers scale? The Rockwell scale? Strain ε

*2-14

(b )

Stress σ

The Brinell hardness of a steel specimen was measured to be 340 HB. What is the material’s approximate tensile strength? What is its hardness on the Vickers scale? The Rockwell scale?

2-15 What are the principal alloy elements of an AISI 4340 steel? How much carbon does it have? Is it hardenable? By what techniques? *2-16

What are the principal alloy elements of an AISI 1095 steel? How much carbon does it have? Is it hardenable? By what techniques?

2-17 What are the principal alloy elements of an AISI 6180 steel? How much carbon does it have? Is it hardenable? By what techniques? 2-18 Which of the steels in Problems 2-15, 2-16, and 2-17 is the stiffest? Strain ε

2-19 Calculate the specific strength and specific stiffness of the following materials and pick one for use in an aircraft wing spar. (a) (b) (c)

(c )

FIGURE P2-1 Stress-Strain Curves

Steel Aluminum Titanium

Sut = 80 kpsi (552 MPa) Sut = 60 kpsi (414 MPa) Sut = 90 kpsi (621 MPa)

2-20 If maximum impact resistance were desired in a part, which material properties would you look for? 2-21

Refer to the tables of material data in Appendix A and determine the strength-toweight ratios of the following material alloys based on their tensile yield strengths: heat-treated 2024 aluminum, SAE 1040 cold-rolled steel, Ti-75A titanium, type 302 cold-rolled stainless steel.

Chapter 2

MATERIALS AND PROCESSES

2-22

Refer to the tables of material data in Appendix A and determine the strength-toweight ratios of the following material alloys based on their ultimate tensile strengths: heat-treated 2024 aluminum, SAE 1040 cold-rolled steel, unfilled acetal plastic, Ti75A titanium, type 302 cold-rolled stainless steel.

2-23

Refer to the tables of material data in Appendix A and calculate the specific stiffnesses of aluminum, titanium, gray cast iron, ductile iron, bronze, carbon steel, and stainless steel. Rank them in increasing order of this property and discuss the engineering significance of these data.

2-24

Call your local steel and aluminum distributors (consult the Yellow Pages or Internet) and obtain current costs per pound for round stock of consistent size in low-carbon (SAE 1020) steel, SAE 4340 steel, 2024-T4 aluminum, and 6061-T6 aluminum. Calculate a strength/dollar ratio and a stiffness/dollar ratio for each alloy. Which would be your first choice on a cost-efficiency basis for an axial-tension-loaded round rod (a) (b)

2-25

71

If maximum strength were needed? If maximum stiffness were needed?

Call your local plastic stock-shapes distributors (consult the Yellow Pages or Internet) and obtain current costs per pound for round rod or tubing of consistent size in plexiglass, acetal, nylon 6/6, and PVC. Calculate a strength/dollar ratio and a stiffness/ dollar ratio for each alloy. Which would be your first choice on a cost-efficiency basis for an axial-tension-loaded round rod or tube of particular diameters. (Note: material parameters can be found in Appendix A.) (a) (b)

If maximum strength were needed? If maximum stiffness were needed?

2-26 A part has been designed and its dimensions cannot be changed. To minimize its deflections under the same loading in all directions irrespective of stress levels, which of these materials would you choose: aluminum, titanium, steel, or stainless steel? Why? *2-27

Assuming that the mechanical properties data given in Appendix Table A-9 for some carbon steels represents mean values, what is the value of the tensile yield strength for 1050 steel quenched and tempered at 400F if a reliability of 99.9% is required?

2-28 Assuming that the mechanical properties data given in Appendix Table A-9 for some carbon steels represents mean values, what is the value of the ultimate tensile strength for 4340 steel quenched and tempered at 800F if a reliability of 99.99% is required? 2-29 Assuming that the mechanical properties data given in Appendix Table A-9 for some carbon steels represents mean values, what is the value of the ultimate tensile strength for 4130 steel quenched and tempered at 400F if a reliability of 90% is required? 2-30 Assuming that the mechanical properties data given in Appendix Table A-9 for some carbon steels represents mean values, what is the value of the tensile yield strength for 4140 steel quenched and tempered at 800F if a reliability of 99.999% is required? 2-31 A steel part is to be plated to give it better corrosion resistance. Two materials are being considered: cadmium and nickel. Considering only the problem of galvanic action, which would you choose? Why? 2-32 A steel part with many holes and sharp corners is to be plated with nickel. Two processes are being considered: electroplating and electroless plating. Which process would you choose? Why? 2-33

What is the common treatment used on aluminum to prevent oxidation? What other metals can also be treated with this method? What options are available with this method?

*2-34

Steel is often plated with a less noble metal that acts as a sacrificial anode that will corrode instead of the steel. What metal is commonly used for this purpose (when the

*

Answers to these problems are provided in Appendix D.

72

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

finished product will not be exposed to saltwater), what is the coating process called, and what are the common processes used to obtain the finished product?

2

2-35

A low-carbon steel part is to be heat-treated to increase its strength. If an ultimate tensile strength of approximately 550 MPa is required, what mean Brinell hardness should the part have after treatment? What is the equivalent hardness on the Rockwell scale?

2-36

A low-carbon steel part has been tested for hardness using the Brinell method and is found to have a hardness of 220 HB. What are the approximate lower and upper limits of the ultimate tensile strength of this part in MPa?

2-37

Figure 2-24 shows “guide lines” for minimum weight design when failure is the criterion. The guide line, or index, for minimizing the weight of a beam in bending is σf2/3 / ρ, where σf is the yield strength of a material and ρ is its mass density. For a given cross-section shape the weight of a beam with given loading will be minimized when this index is maximized. The following materials are being considered for a beam application: 5052 aluminum, cold rolled; CA-170 beryllium copper, hard plus aged; and 4130 steel, Q&T@1200F. The use of which of these three materials will result in the lowest-weight beam?

2-38

Figure 2-24 shows “guide lines” for minimum weight design when failure is the criterion. The guide line, or index, for minimizing the weight of a member in tension is σf / ρ, where σf is the yield strength of a material and ρ is its mass density. The weight of a member with given loading will be minimized when this index is maximized. For the three materials given in Problem 2-37, which will result in the lowest weight tension member?

2-39

Figure 2-23 shows “guide lines” for minimum weight design when stiffness is the criterion. The guide line, or index, for minimizing the weight of a beam in bending is E1/2 / ρ, where E is the modulus of elasticity of a material and ρ is its mass density. For a given cross-section shape the weight of a beam with given stiffness will be minimized when this index is maximized. The following materials are being considered for a beam application: 5052 aluminum, cold rolled; CA-170 beryllium copper, hard plus aged; and 4130 steel, Q&T@1200F. The use of which of these three materials will result in the lowest-weight beam?

2-40

Figure 2-24 shows “guide lines” for minimum weight design when stiffness is the criterion. The guide line, or index, for minimizing the weight of a member in tension is E / ρ, where E is the modulus of elasticity of a material and ρ is its mass density. The weight of a member with a given stiffness will be minimized when this index is maximized. For the three materials given in Problem 2-39, which will result in the lowest-weight tension member?

2-41

Find a relationship between the Vickers and Brinell hardness scales that will give a Vickers number as a function of a Brinell number to an accuracy of at least 99%. Compare your computed Vickers number against that given in Table 2-3 for Brinell values of 159, 375, and 495 and calculate the percent variation for each.

2-42

Find a relationship between the Rockwell C and Brinell hardness scales over the range of 277 ≤ HB ≤ 578 that will give a Rockwell C number as a function of a Brinell number to an accuracy of at least 99%. Compare your computed Rockwell C number against that given in Table 2-3 for Brinell values of 311, 375, and 495 and calculate the percent variation for each.

2-43

After a part in service failed it was found to have a hardness of 34.2 HRC. What was its approximate ultimate tensile strength?

2-44

A part was tested and found to have a hardness of 437 HV. What was its approximate ultimate tensile strength?

KINEMATICS AND LOAD DETERMINATION

Look long on an engine. It is sweet to the eyes. Macknight black

3.0

INTRODUCTION

This chapter provides a brief introduction to the fundamentals of kinematics, which is defined as the study of motion without regard to forces. This approach, first suggested by Ampere, is intended to make it easier to understand motion analysis by temporarily assuming that all bodies are rigid and are not loaded by forces. This chapter also provides a review of the fundamentals of static and dynamic force analysis, impact forces, and beam loading. It also introduces singularity functions for beam calculations. The Newtonian solution method of force analysis is reviewed and a number of case-study examples are presented to reinforce understanding of this subject. The case studies also set the stage for analysis of these same systems for stress, deflection, and failure modes in later chapters. At the end of the chapter, a summary section is provided that groups significant equations from this chapter for easy reference and identifies the chapter section in which their discussion can be found.

3.1

DEGREE OF FREEDOM

The degree of freedom (DOF) of a system is equal to the number of coordinates needed to define its position in space. In three-space, a single rigid body has six DOF, three translational (x, y, z) and three rotational (q, f, g). When multiple bodies are connected together with joints, then either more or fewer DOF may result. A system’s DOF also defines the number of inputs (motors or other actuators) needed to control its motion. If the DOF = 1, the system is a one-DOF mechanism, more than 1, a multi-DOF mechanism, zero, a structure that cannot move, and less than 1, a preloaded structure. This chapter is Copyright © 2018 Robert L. Norton: All Rights Reserved

73

3

74

3

MACHINE DESIGN

Title-page photograph is of a cutaway model of the valve train for a WWII radial aircraft engine taken by the author at the Bristol, England museum. Copyright 2018 Robert L. Norton: All Rights Reserved * A more complete presentation of the material in sections 3.1 to 3.8 inclusive can be found in reference 1.

-

An Integrated Approach

-

Sixth Edition

MECHANISMS*

3.2

Mechanisms come in many types but are all variants of a linkage, which is made of a collection of links and joints, one of which is grounded, and all are interconnected in a way to provide a controlled output in response to one or more inputs. A link, for the purpose of kinematic analysis, is a rigid body of any shape that has some number of attachment points called nodes (which are often holes) that allow multiple links to be connected by joints. Figure 3-1a shows some examples of links. Links with two attachment points or nodes are called binary, with three, ternary, with four quaternary. This naming and adding of nodes can continue indefinitely, but many mechanisms are made with combinations of the three links shown. Note that the links can be of any shape. The number of nodes alone determines the character of the link.

Nodes

Binary link

Quaternary link

Ternary link

(a) Some links—their names reflect the number of nodes

∆θ

∆θ ∆x

∆x

∆x ∆θ

Prismatic (P) joint—1 DOF (form closed)

Pin in slot—2 DOF (form closed)

Link against plane—2 DOF (force closed)

Revolute (R) joint—1 DOF (form closed)

∆φ

∆x

∆θ Cylindric (C) joint—2 DOF (form closed)

∆θ ∆ψ

∆y

∆φ ∆x

∆θ Helical (H) joint—1 DOF (form closed)

Spherical (S) joint—3 DOF (force closed)

Planar (F) joint—3 DOF (force closed)

(b) Some joint types—note their DOF and type of closure

FIGURE 3-1 Examples of Links and Joints

Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 3

KINEMATICS AND LOAD DETERMINATION

Joints are characterized by their geometry, by the number of degrees of freedom they allow between the links they join, and by whether they are held together (closed) by a force or by their form (geometry). Figure 3-1b shows a number of joint types. Clockwise from the upper left of part b in the figure, two links joined with a pin have one DOF (Dq) and the pin joint is form closed because the pin is captured in the hole by its geometry. This is called a revolute joint. A block captive in a slot is a form-closed prismatic joint with one DOF (Dx). But a pin captive in a slot is a form-closed half joint with two DOF (Dx) and (Dq). Likewise, the link contacting a plane has two DOF but is force closed. A planar joint has three DOF and is force closed. The same is true for the spherical (joystick) joint. The helical joint (screw and nut) has one DOF. As you turn the screw it advances in the x direction. The cylindric joint has two DOF, Dx and Dq.

CALCULATING DEGREE OF FREEDOM (MOBILITY)

3.3

The degree of freedom of an assembly of links and joints can be calculated with a simple equation, though in some cases it may give a misleading result. We present the twodimensional version of this equation, which applies to any planar linkage, defined as one whose links move in parallel planes. The Kutzbach equation uses just the number of links and joints for the calculation: M = 3 ( L − 1) − 2 J1 − J2

(3.1)

where M = degree of freedom or mobility, L is the number of links, J1 is the number of full or one-DOF joints, and J2 is the number of half or two-DOF joints. This equation does not account for link lengths and so can predict an incorrect DOF when the link geometries have particular values. Figure 3-2 shows two linkages that each have 5 links, 6 full (J1) joints, and no half (J2) joints. Equation 3.1 says that they both have zero DOF, i.e., they are predicted to be structures, unable to move. This is true for the linkage of Figure 3-2a, but the linkage in Figure 3-2b can obviously move due to the fact that the three binary links are the same length and are also parallel due to the fact that the two ternary links (one of which—the ground link— is shown only as its three pin joints fixed to ground) have the same spacing between their nodes. This special geometry, not reflected in the Kutzbach equation, causes this linkage to have one DOF M = 3(5 – 1) – 2(6) = 0

M = 3(5 – 1) – 2(6) = 0 but is untrue 5

5 2 3

1

2

4

1

(a) A structure with DOF = 0

FIGURE 3-2 A Kutzbach Paradox

1 1

3

4

1

1

(b) A mechanism with DOF = 1 Copyright © 2018 Robert L. Norton: All Rights Reserved

75

3

76

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

despite the prediction of the equation. This is called a paradox of the Kutzbach equation. Many other paradoxes to this equation exist, all due to unique geometries.

3

3.4

COMMON 1-DOF MECHANISMS

Fourbar Linkage and the Grashof Condition The simplest 1-DOF mechanism is the fourbar linkage: four binary links connected by four pin joints. This is an amazingly versatile mechanism and can solve many intricate motion control problems. It will behave differently depending on the relative lengths of its four links. Its behavior can be predicted by the Grashof criterion: S + L ≤ P+Q

(3.2)

where S is the length of the shortest link, L that of the longest link, and P, Q, the lengths of the other two. It matters not the order in which the links are assembled. If the Grashof inequality is false, then no link can make a full revolution and it is said to be a nonGrashof linkage. If true, then at least one link can revolve fully with respect to the others and it is called a Grashof linkage. But if the expression evaluates to equal, then it is a special-case Grashof linkage that has “change-point” positions at which the successive positions of the linkage are indeterminate. Such a linkage either has to be constrained to avoid these change points, or additional links added to carry it through the change points. Figure 3-3 shows the four inversions of a (non-special-case) Grashof linkage. An inversion results from the grounding of a different link in the chain, thus any linkage will have as many inversions as it has links. Note the link numbering. The ground link

ground 1

coupler 3

rocker 4

rocker 4 crank 2

crank 2 ground 1

coupler 3 (a) Crank rockers

rocker 2 crank 2

ground 1

coupler 3

coupler 3 ground 1

rocker 4 crank 4 (b) Double rocker

FIGURE 3-3 Inversions of the Grashof Fourbar Linkage

(c) Double crank

Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 3

KINEMATICS AND LOAD DETERMINATION

is always 1, the input link is always 2, the coupler is 3, and the output link is 4. A crank can fully revolve, a rocker cannot. Each inversion has the potential to deliver different types of motion from the same set of links. Figure 3-3a shows the two crank-rocker inversions of the Grashof fourbar in which the crank rotates fully, the rocker rocks or oscillates through an arc, and the link connecting crank and rocker, called the coupler, has complex motion. The crank-rocker inversion results when either of the links adjacent to the shortest is grounded. This is a very useful mechanism used whenever one wishes to convert the continuous rotation of a motor to an oscillation. An example is a windshieldwiper mechanism on an automobile where the wiper blade is attached to the rocker. Figure 3-3b shows the double-rocker mechanism that results when the link opposite the shortest is grounded. Since it is always the shortest link of a Grashof mechanism that can fully revolve with respect to the others, this inversion is more difficult to motivate since a motor cannot be attached to either of the grounded rocker links, neither of which fully rotates. Figure 3-3c shows the double-crank inversion in which the shortest link is grounded resulting in the other three revolving about it. This can be motor driven and if one of the cranks is driven at constant speed, the other will exhibit non-constant velocity. This is also called a drag-link mechanism and is sometimes used in automotive steering mechanisms. Figure 3-4a shows one inversion of the special case Grashof linkage: the parallelogram mechanism, which is quite useful as long as the links are not allowed to become colinear, which is a change point. How they will emerge from the colinear position is mathematically unpredictable (indeterminate). However, over its useful range of motion, the coupler remains parallel to the ground plane. Figure 3-4b shows a way that the parallelogram mechanism can be carried through the change points by adding a second stage that is out of phase with the first. Both stages have change points but each occurs at a different crank angle so the other stage carries it through. Note that this five-bar linkage is also a Kutzbach paradox like Figure 3-2b. All four inversions of a non-Grashof fourbar linkage are triple rockers and give similar motions. All three moving links just oscillate until they reach extreme positions called toggle positions. They are nevertheless useful as many applications do not require that any link make a full revolution. One example is the suspension linkage in an automobile that controls the up and down motion of the wheel assemblies over bumps coupler 3 crank 2 crank 2

coupler 3

crank 4 ground 1

(a) Parallelogram linkage

FIGURE 3-4 Special-case Grashof linkages

crank 4 ground 1

crank 5

ground 1

(b) Able to make a full revolution Copyright © 2018 Robert L. Norton: All Rights Reserved

77

3

78

MACHINE DESIGN

-

An Integrated Approach

2

-

2

1

3

Sixth Edition

3

Car frame

3

4

4 Springs

FIGURE 3-5

Copyright © 2018 Robert L. Norton: All Rights Reserved

Non-Grashof Fourbar Linkages Used for Automotive Suspension

as shown in Figure 3-5. Here the wheel motion only requires the links to move through relatively small angles and never through a full revolution. So all forms of the fourbar linkage are very useful and are found in all sorts of machinery. It is the basic building block of all mechanisms. Sixbar Linkage The next simplest combination of links is the sixbar linkage, which comes in two forms, the Watt and the Stephenson. Figure 3-6 shows the two inversions of the Watt sixbar and the three inversions of the Stephenson sixbar. Note that these all contain four binary links to which a pair of ternary links has been added. These are all one-DOF linkages. 4

5

3

3

4

6 2

5

2

6

1

1

(a) Watt’s sixbar inversion I

(b) Watt’s sixbar inversion II

3

5

3

5 2

4

3

6

4 6

4

5

2

2

6

1 1 (c ) Stephenson’s sixbar inversion I

FIGURE 3-6 All Inversions of the Sixbar Linkage

1 (d) Stephenson’s sixbar inversion II

(e) Stephenson’s sixbar inversion III Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 3

KINEMATICS AND LOAD DETERMINATION

Slider block

79

Connecting rod Pisto n

2

3

3

4

1

1

2

4

Crank 1

1 Cylinder

Effective link 4

Effective rocker pivot is at infinity

Gas pressure

∞ (a) Crank-slider—crank drives slider

FIGURE 3-7

(b) Slider-crank—slider drives crank Copyright © 2018 Robert L. Norton: All Rights Reserved

Crank-Slider and Slider-Crank Mechanisms

The Watt II in Figure 3-6b is probably the most commonly seen and it can be thought of as two fourbar linkages in series. Fourbar Crank-Slider and Slider-Crank If the rocker of a fourbar crank rocker linkage is increased in length indefinitely, it effectively becomes infinite in length and the linkage is transformed into a fourbar crankslider as shown in Figure 3-7a. The rocker link becomes a slider block that oscillates in a straight line rather than along an arc. (A straight line can be thought of as an arc of infinite radius.) This is a very common linkage in machinery and is the heart of every internal combustion (IC) engine. There it is driven backward as a slider-crank with the piston (slider) driving the crank as shown in Figure 3-7b. In this mode, it has two change points at top and bottom dead center and its crank must be spun to start it moving continuously, which is why you must pull the starter rope on the lawn mower to start it and crank the car engine with a starter motor. The momentum of the rotating crank carries it through the change points once it is put in motion. The crank-slider of Figure 3-7a is widely used to obtain straight-line output motion from a continuously rotating crank. An example is a piston pump for pumping well water or compressing air. All inversions of the crank-slider that can be physically assembled are Grashof linkages in which the crank can fully revolve. Cam and Follower A cam and follower is another variant of the fourbar linkage in which the crank takes on a contoured shape. It is then called a cam and drives a rocker or slider directly. There is no coupler link. Figures 3-8a and 3-8b show cam and follower arrangements that drive a translating slider and an oscillating rocker follower, respectively. Dotted lines show the effective links of an equivalent fourbar crank-slider or fourbar crank-rocker that would give the same motion as the cam for the instantaneous position shown. One major advantage of a cam-follower mechanism over a pure linkage is the variety of follower motion functions possible. You can think of the cam-follower mechanism as a fourbar linkage or crank-slider in which the crank and coupler are able to change their lengths

3

80

MACHINE DESIGN

Instantaneous center of cam curvature Effective link 2

ω2

3

An Integrated Approach

Sixth Edition

Follower

Half joint

Half joint Vfollowe r 4

2

Effective link 2 Spring

Cam 2

Fo llower Spring parallel effect ive links meet@ ∞

-

Effective link 3

Effective link 3

Cam Effect ive link 1

-

ω2

ω4

1

Effective link 4 Instantaneous center of cam curvature

(a) Cam with sliding follower—a variant of a crank-slider

4

Effect ive link 4

(b) Cam with oscillating follower—a variant of a crank-rocker

FIGURE 3-8

Copyright © 2018 Robert L. Norton: All Rights Reserved

Cams and Followers are Variants of Fourbar Linkages

as the cam rotates. The effective links are shown superposed on the mechanism in the figure. The effective coupler (link 3) runs from the instantaneous center of curvature of the cam contour (which changes for each cam position) to the center of curvature of the follower. The effective crank runs from the cam pivot to the instantaneous center of curvature of the cam contour. Thus, the lengths of the effective crank and effective coupler change with cam rotational position. The cam contour must be defined by appropriate mathematical functions, which will be addressed in a later section.

3.5

ANALYZING LINKAGE MOTION

Before we can determine the forces and torques in a linkage, which are needed for a stress analysis, we first need to know the accelerations to use in Newton’s second law, F = ma. In order to find the accelerations, we must first find the position and velocity of the links because acceleration is the derivative of velocity and velocity is the derivative of position. We will show one approach to this kinematic analysis problem and apply it to the fourbar linkage and crank slider. We will limit our discussion to two dimensional or planar mechanisms. Luckily these constitute the majority of practical mechanisms used in machinery. Types of Motion General motion in the plane is called complex motion. A link in the plane has 3-DOF, call them x, y, and q. Complex motion of a link then involves a translation of a point on the link in x and y and a rotation q about a point in the link. So, complex motion then has two components, called translation and rotation. Each of these can also exist without the other, and then it becomes a special case. A crank spinning about a fixed pivot is in pure rotation and a slider block traveling along a slot has pure translation. These are the components of complex motion. In a fourbar crank slider, the crank is in pure rotation, the slider is in pure translation and the coupler is in complex motion, simultaneously rotating and translating.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

81

Complex Numbers as Vectors Position, velocity, and acceleration are vector quantities and so we need a notation to represent a vector. There are several possibilities, but for two-dimensional vectors, complex numbers provide some advantages. Figure 3-9a shows the complex plane with a vector drawn on it. The axes are labeled real and imaginary. Don’t be put off by the terminology; despite its name, the imaginary axis simply depicts the y-directed cartesian component of the vector with the x-directed component on the real axis. The imaginary name comes about because we use the notation j to represent the square root of minus one, which cannot be evaluated numerically. But think of j as an operator rather than as a value. When present on a term, it merely signifies that the term is y-directed.

3

A vector in this notation has a very compact and convenient form, rejq, where r is its magnitude, e is the Naperian base 2.718..., and q is the angle of the vector. This is the polar notation for the vector, i.e., magnitude and angle. We can easily convert from polar to cartesian coordinates with Euler’s identity: e ± jθ = cos θ ± j sin θ

(3.3)

The real advantage of this complex notation comes when we need to differentiate an expression, especially if we use the polar form. The differential of exdx is ex. When it has a constant multiplier in the exponent such as ejxdx, its differential is jejx. So the differential of our vector rejq dq is rjejq provided that r is a constant. Note in Figure 3-9b that each multiplication of the vector RA by the operator j results in a counterclockwise rotation of the vector through 90 degrees. The vector RB  = jRA is directed along the positive imaginary or j axis. The vector RC  = j2 RA is directed along the negative real axis because j2 = –1 and thus RC  = –RA. In similar fashion, RD = j3 RA = –jRA and this component is directed along the negative j axis. j Polar form : R e θ

Imaginary

Cartesian form : R cos θ + j R sin θ R =

Imaginary j

j R sin θ

A

j

RA

B 2

RC = j R = – R

R B = jR +θ A

RA C

θ

O

Real RA

Real R cos θ (a) Complex number representation of a position vector

FIGURE 3-9 Complex Number Representation of Vectors in Two Dimensions

3

R D = j R = – jR D (b) Vector rotations in the complex plane Copyright © 2018 Robert L. Norton: All Rights Reserved

82

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Thus, the appearance of the j operator as a multiplier in the differential indicates that this vector has been rotated 90°. Two differentiations result in a multiplier j2 = –1, which rotates the second differential 180° from the original vector. By this means, the directions of the differentiated vectors are taken care of automatically. After the differentiation is completed, the expression can be converted to cartesian form with Euler’s equation 3.3.

3 The Vector Loop Equation Closed linkages such as those discussed here are conveniently represented with a vector loop in which each link becomes a vector. These vectors close around the loop making their vector sum zero. That addition provides an equation for the linkage that can be evaluated to find position, velocity and acceleration information for any position. The vectors can be arranged in any orientation, that is their heads and tails can be at either end of a link as convenient. Since the angle of a vector must be measured at its root, then the orientation should be chosen such that it defines the angle of most interest for each link. Accordingly the linkage in Figure 3-10 shows the vector R2 rooted at the fixed pivot of link 2, R3 rooted where link 2 joins link 3, and R4 arranged with its root at the fixed pivot of link 4. For convenience, the X axis is aligned with R1 and the origin is placed at O2, the driving link pivot. Any such arrangement can be used and the correct results obtained as long as the signs of the vectors are observed in the summation.

3.6

ANALYZING THE FOURBAR LINKAGE

Solving for Position in the Fourbar Linkage The first step is to sum the vectors around the loop, taking their orientations into account: R 2 + R 3 − R 4 − R1 = 0

(3.4 a)

Then substitute the polar complex number notation for each vector using the letters of Figure 3-10 to represent the link magnitudes:

Y

B

y A

R3 θ3

b x c

R2

R4 θ4

a θ2

O2 FIGURE 3-10 Vector Loop for a Fourbar Linkage

d R1

O4

X

Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 3

KINEMATICS AND LOAD DETERMINATION

a e jθ2 + b e jθ3 − c e jθ4 − d e jθ1 = 0

83

(3.4 b)

Substitute equation 3.3 to convert the equation to cartesian form: a ( cos θ2 + j sin θ2 ) + b ( cos θ3 + j sin θ3 ) − c ( cos θ4 + j sin θ4 ) − d ( cos θ1 + j sin θ1 ) = 0

(3.4c)

Expand and separate the real and imaginary (x and y) terms into two equations. Note that terms without a j multiplier are x components and those with a j are y components: a cos θ2 + b cos θ3 − c cos θ4 − d cos θ1 = 0 but:

θ1 = 0, so: a cos θ2 + b cos θ3 − c cos θ4 − d = 0

(3.4 d )

ja sin θ2 + jb sin θ3 − jc sin θ4 − jd sin θ1 = 0 but:

θ1 = 0, and the j 's divide out, so: a sin θ2 + b sin θ3 − c sin θ4 = 0

(3.4e)

Equations 3.4d and 3.4e can be solved simultaneously to find angles q3 and q4. The link lengths a, b, c, d and the driver angle q2 for this position are known. The complete derivation of this solution is provided on the website as Fourbar Position Derivation.pdf. It contains the equations 3.4f through 3.4o that have been omitted here to save space. The final result is:

where:

 − B ± B 2 − 4 AC  θ41,2 = 2 arctan     2A A = cos θ2 − K1 − K 2 cos θ2 + K 3

(3.4 p)

B = −2 sin θ2

C = K1 − ( K 2 + 1) cos θ2 + K 3 and:

K1 =

d , a

K2 =

d , c

K3 =

a2 − b2 + c2 + d 2 2ac

Note that there are two solutions corresponding to the + / – conditions on the radical. These two solutions, as with any quadratic equation, may be of three types: real and equal, real and unequal, complex conjugate. If the discriminant under the radical is negative, then the solution is complex conjugate, which simply means that the link lengths chosen are not capable of connection for the chosen value of the input angle q2. This can occur either when the link lengths are completely incapable of connection in any position or, in a non-Grashof linkage, when the input angle is beyond a toggle limit position. Then there is no real solution for that value of input angle q2. Excepting this situation, the solution will usually be real and unequal, meaning there are two values of q4 corresponding to any one value of q2. These are referred to as the crossed and open configurations of the linkage and also as the two circuits of the linkage. In the fourbar linkage, the minus solution gives q4 for the open configuration and the positive solution gives q4 for the crossed configuration. Figure 3-11 shows both crossed and open solutions for a Grashof crank-rocker linkage. The terms crossed and open are based on the assumption that the input link 2, for

3

84

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Y

Open B

y

3

θ3 '

3

b A

θ3

c

x

4

a 2 θ2 O2

θ4 '

1 3'

θ4

O4

d

X

4' Crossed B'

FIGURE 3-11

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Two Solutions to θ4 from the Fourbar Position Equation

which q2 is defined, is placed in the first quadrant (i.e., 0 < q2 < π/2). A Grashof linkage is then defined as crossed if the two links adjacent to the shortest link cross one another, and as open if they do not cross one another in this position. Note that the configuration of the linkage, either crossed or open, is solely dependent upon the way that the links are assembled. You cannot predict, based on link lengths alone, which of the solutions will be the desired one. In other words, you can obtain either solution with the same linkage by simply taking apart the pin that connects links 3 and 4 in Figure 3-11, and moving those links to the only other positions at which the pin will again connect them with the same value of q2. In so doing, you will have switched from one position solution, or circuit, to the other. A similar derivation that eliminates q4 from the equations to solve for q3 gives the following solution:

where

 − E ± E 2 − 4 DF  θ 31,2 = 2 arctan     2D D = cos θ2 − K1 + K 4 cos θ2 + K 5 E = −2 sin θ2 F = K1 + ( K 4 − 1) cos θ2 + K 5

and

(3.5)

K1 =

d a

K4 =

d b

K5 =

c2 − d 2 − a2 − b2 2 ab

Solving for Velocity in the Fourbar Linkage We start with the vector loop equation 3.4a and its complex number form 3.4b: R 2 + R 3 − R 4 − R1 = 0

(3.4 a)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

a e jθ2 + b e jθ3 − c e jθ4 − d e jθ1 = 0

85

(3.4 b)

Differentiate equation 3.4b with respect to time to get velocity: ja e jθ2

but, so:

dθ dθ2 dθ4 + jb e jθ3 3 − jc e jθ4 =0 dt dt dt dθ 3 = ω3; dt

dθ2 = ω2; dt

dθ4 = ω4 dt

ja ω 2 e jθ2 + jb ω 3e jθ3 − jc ω 4 e jθ4 = 0

(3.6a)

(3.6b) (3.6 c )

The angle q1 is constant so its derivative is zero and it has dropped out. Equation 3.6c is the velocity difference equation and can be written: VA + VBA − VB = 0 where:

VA = ja ω 2 e

(3.6d ) jθ2

VBA = jb ω 3e jθ3 VB = jc ω 4 e

(3.6e)

jθ4

VA is the linear velocity of point A in Figure 3-11, VB is the linear velocity of point B, and VBA is the velocity difference of point B versus point A. It is read, “the velocity of B with respect to A.” Equation 3.6c is solved for the angular velocities w3 and w4 with these expressions: ω3 =

aω 2 sin ( θ4 − θ2 ) b sin ( θ3 − θ4 )

(3.6l )

ω4 =

aω 2 sin ( θ2 − θ3 ) c sin ( θ4 − θ3 )

(3.6m)

A complete derivation of this solution is provided on the website as Fourbar Velocity Derivation.pdf. It contains equations 3.6f through 3.6k omitted here to save space. Note that the position analysis must be complete before the velocities can be found as they depend on those terms as well as the link lengths and w2, the known driving velocity. Once w3 and w4 are known, we can find the linear velocities from: VA = a ω 2 ( − sin θ2 + j cos θ2 )

VBA = b ω 3 ( − sin θ3 + j cos θ3 )

(3.6n)

VB = c ω 4 ( − sin θ4 + j cos θ4 )

where the real and imaginary terms are the x and y components of the vectors. There are two solutions for the velocities corresponding to the open and crossed position solutions. Find them using the values for q3 and q4 of each of those results. Figure 3-12 shows the velocity vectors on a fourbar linkage and a graphical solution to equation 3.6d. The angle m between links 3 and 4 is called the transmission angle.

3

86

MACHINE DESIGN

-

An Integrated Approach

Y y

ω3

A ω2

VBA

b R3

δ3 θ3

VA θ4

c R2 θ2

O2

VB

x

a –

VB R4

µ

Sixth Edition

µ

B

P

RP p

3

+

-

VA

– d

R1

ω4

VBA

(b)

X

O4

(a)

FIGURE 3-12

Copyright © 2018 Robert L. Norton: All Rights Reserved

Vector Loop and Velocity Components in a Fourbar Linkage

Angular Velocity Ratio and Mechanical Advantage The angular velocity ratio mV is defined as the output angular velocity divided by the input angular velocity. For a fourbar mechanism this is expressed as: mV =

ω4 ω2

(3.6o)

For a rotating system, power P is the product of torque T and angular velocity w that, in two dimensions, have the same (z) direction: P = Tω

(3.6 p)

Linkage systems can be very efficient if they are well made with low friction bearings on all pivots. Losses are often less than 10%. For simplicity in the following analysis we will assume that the losses are zero (i.e., a conservative system). Then, letting Tin and win represent input torque and angular velocity, and Tout and wout represent output torque and angular velocity, Pin = Tin ω in Pout = Tout ω out Pout = Pin Tout ω out = Tin ω in Tout ω = in = mT Tin ω out

(3.6q)

(3.6r )

Note that the torque ratio (mT = Tout /Tin) is the inverse of the angular velocity ratio. Mechanical advantage (mA) can be defined as: mA =

Fout Fin

(3.6s)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

t

ABA

α3 ω3 AB R3

t

AA R2

α2

n

b

θ3

AB

n

c

a

R4

d

ABA

ω4

–AA t

n

AA

ABA

O4 (b)

(a)

FIGURE 3-13

Copyright © 2018 Robert L. Norton: All Rights Reserved

Acceleration Vectors on the Linkage Vector Loop (a), and in an Acceleration Diagram (b)

Assuming that the input and output forces are applied at some radii rin and rout, perpendicular to their respective force vectors, Fout =

Tout rout

(3.6t )

T Fin = in rin

substituting equations 3.6t in 3.6s gives an expression in terms of torque: T  r  m A =  out   in   Tin   rout 

(3.6u)

Substituting equation 3.6r in 3.6u gives  ω  r  m A =  in   in   ω out   rout 

(3.6 v )

These two ratios, angular velocity ratio and mechanical advantage, provide useful, dimensionless indices of merit by which we can judge the relative quality of various linkage designs that may be proposed as solutions. Solving for Acceleration in the Fourbar Linkage We start with the velocity vector loop equation 3.6c ja ω 2 e jθ2 + jb ω 3e jθ3 − jc ω 4 e jθ4 = 0

(3.6 c )

and differentiate it to get an expression for acceleration:

( j 2 aω 22 e jθ

2

n

AA

t

ABA X

O2

n

n

θ4

R1 AA

AB

AB

α4

ABA

x

A

θ2

ω2

t

AB

B

t

y

Y

87

) (

) (

)

+ jaα 2 e jθ2 + j 2 b ω 32 e jθ3 + jb α 3 e jθ3 − j 2 c ω 24 e jθ4 + jc α 4 e jθ4 = 0

(3.7a)

3

88

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Simplifying and grouping terms:

(aα 2 je jθ 3

2

) (

) (

)

− aω 22 e jθ2 + b α 3 je jθ3 − b ω 32 e jθ3 − c α 4 je jθ4 − c ω 24 e jθ4 = 0

(3.7b)

Equation 3.7b is the acceleration difference equation and can be written as: A A + A BA − A B = 0

(3.7c)

( ) ( ) jθ t n 2 jθ A BA = ( A BA + A BA ) = ( b α 3 je − b ω 3 e ) A B = ( AtB + A nB ) = ( c α 4 je jθ − c ω 24 e jθ ) A A = AtA + A nA = aα 2 je jθ2 − aω 22 e jθ2

where:

3

4

3

(3.7d )

4

AA and AB are the linear accelerations of points A and B, respectively, and ABA is the acceleration difference between points B and A. Note that each of these vectors has two components denoted by the superscripts t and n, which refer to the tangential and normal directions, respectively. The normal (or centripetal) component is directed toward the center of rotation and the tangential component is directed as its name indicates. The acceleration difference component ABA has as its center of rotation its reference point A. Figure 3-13 shows these components on the linkage and in a vector diagram. Equations 3.7i are solved simultaneously to get: CD − AF AE − BD CE − BF α4 = AE − BD α3 =

where:

(3.7 k ) (3.7l )

A = c sin θ4 B = b sin θ3 C = aα 2 sin θ2 + aω 22 cos θ2 + bω 32 cos θ3 − cω 24 cos θ4 D = c cos θ4

(3.7m)

E = b cos θ3 F = aα 2 cos θ2 − aω 22 sin θ2 − bω 32 sin θ3 + cω 24 sin θ4

A complete derivation of these equations showing equations 3.7e through 3.7j is provided on the website as Fourbar Acceleration Derivation.pdf. Note that the angular accelerations a3 and a4 are functions of the link lengths, all the link angles, and all the link angular velocities, as well as the input angular acceleration a2. So, the position and velocity analyses must be done before the accelerations can be found. Once a3 and a4 are found, the linear accelerations can be found by substituting the Euler identity (equation 3.3) into equation 3.7d: A Ax

A A = aα 2 ( − sin θ2 + j cos θ2 ) − aω 22 ( cos θ2 + j sin θ2 ) = − aα 2 sin θ2 − aω 22 cos θ2 A Ay = aα 2 cos θ2 − aω 22 sin θ2

(3.7n)

A BAx

A BA = b α 3 ( − sin θ3 + j cos θ3 ) − b ω 32 ( cos θ3 + j sin θ3 ) = − b α 3 sin θ3 − b ω 32 cos θ3 A BAy = b α 3 cos θ3 − b ω 32 sin θ3

(3.7o)

Chapter 3

A Bx

KINEMATICS AND LOAD DETERMINATION

A B = c α 4 ( − sin θ4 + j cos θ4 ) − c ω 24 ( cos θ4 + j sin θ4 ) = − c α 4 sin θ4 − c ω 24 cos θ4 A By = c α 4 cos θ4 − c ω 24 sin θ4

89

(3.7 p)

where the real and imaginary terms are the x and y components, respectively. It is often the case that we need to know the linear acceleration of a point on a link such as its center of gravity (CG). Figure 3-12 shows the CG of link 3 at point P and defines its location with the position vector RP at angle d3. The acceleration of this point is found from the addition of two acceleration vectors, such as AA and APA. Vector AA is already defined from our analysis of the link accelerations. APA is the acceleration difference of point P with respect to point A. Point A is chosen as the reference point because angle q3 is defined at a local coordinate system whose origin is at A. Position vector RPA is defined using its magnitude p, the internal link offset angle d3 and the angle of link 3, q3. Differentiate this position vector twice to get acceleration: j θ +δ R PA = pe ( 3 3 ) j θ +δ VPA = jpe ( 3 3 )ω 3 j θ +δ j θ +δ A PA = pα 3 je ( 3 3 ) − pω 32 e ( 3 3 )

(3.7q)

= pα 3  − sin ( θ3 + δ 3 ) + j cos ( θ3 + δ 3 ) 

− pω 32  cos ( θ3 + δ 3 ) + j sin ( θ3 + δ 3 ) 

and use the acceleration difference equation to find AP: A P = A A + A PA

3.7

(3.7r )

ANALYZING THE FOURBAR CRANK-SLIDER

Figure 3-14 shows an offset fourbar crank-slider linkage. This is the general case in which the slider axis does not pass through the crank pivot. The vector loop uses four vectors, with R1 arranged parallel to the slider axis and through the crank pivot. Vector R4 is perpendicular to the slider axis and represents the offset, which has constant magnitude c. The variable magnitude d of R1 defines the position of the slider. R3 is rooted at the slider with constant magnitude b, and R2 is rooted at the crank pivot with constant magnitude a. Solving for Position in the Fourbar Crank-Slider Write the vector loop equation from Figure 3-14: R 2 − R 3 − R 4 − R1 = 0

(3.8a)

Put it in complex polar form: a e jθ2 − b e jθ3 − c e jθ4 − d e jθ1 = 0

Substitute equation 3.3 to put it in complex cartesian form:

(3.8b)

3

90

MACHINE DESIGN

-

An Integrated Approach

Y

-

Sixth Edition

θ3

y

slider axis

3

b

offset

O2

R2

x

R3

A a

B

c

Rs

θ2

R4 θ4

d

X

R1 FIGURE 3-14

Copyright © 2018 Robert L. Norton: All Rights Reserved

Vector Loop for an Offset Fourbar Crank-Slider Linkage

a ( cos θ2 + j sin θ2 ) − b ( cos θ3 + j sin θ3 )

− c ( cos θ4 + j sin θ4 ) − d ( cos θ1 + j sin θ1 ) = 0

(3.8c)

Separate the real and imaginary terms to get two scalar equations: a cos θ2 − b cos θ3 − c cos θ4 − d = 0

(3.8d )

a sin θ2 − b sin θ3 − c sin θ4 = 0

(3.8e)

Solve equations 3.8d and 3.8e simultaneously for slider position d and q3. Note that q4 is a known constant value:  a sin θ2 − c  θ31 = arcsin    b d = a cos θ2 − b cos θ3

(3.8 f ) (3.8 g)

This solution is for only one circuit of the linkage. The second circuit has:  a sin θ2 − c  θ32 = arcsin  −  + π  b

(3.8h)

and the value of d for the second circuit is found from equation 3.8g using this value of q3. Solving for Velocity in the Fourbar Crank-Slider Start with the vector loop equation 3.8b for this linkage a e jθ2 − b e jθ3 − c e jθ4 − d e jθ1 = 0

(3.8b)

and differentiate it with respect to time to get velocity: ja ω 2 e jθ2 − jb ω 3e jθ3 − d = 0

(3.9a)

Substitute the Euler identity (equation 3.3) into equation 3.9a, separate into real and imaginary components and solve simultaneously to get:

Chapter 3

KINEMATICS AND LOAD DETERMINATION

θ3

Y

91

y VBA

ω3

+

R3 A ω2

R2

θ2

VBA θ4 (b)

d

O2

X

R1 to ∞

(a)

FIGURE 3-15

effective link 4 Copyright © 2018 Robert L. Norton: All Rights Reserved

Velocity Vectors and Vector Loop on a Crank-Slider Linkage

ω3 =

a cos θ2 ω2 b cos θ3

(3.9 g)

d = − a ω 2 sin θ2 + b ω 3 sin θ3

(3.9h)

Note that a position analysis must be done before solving for velocity. The linear velocities are: VA = a ω 2 ( − sin θ2 + j cos θ2 )

(3.9i )

VAB = b ω 3 ( − sin θ3 + j cos θ3 )

(3.9 j )

VB = d

(3.9 k )

Figure 3-15 shows these vectors on the linkage. A complete derivation of these equations showing equations 3.9b through 3.9f can be found in the file Slider Velocity Derivation.pdf on the website. Solving for Acceleration in the Fourbar Crank-Slider Start with the velocity equation 3.9a for this linkage ja ω 2 e jθ2 − jb ω 3e jθ3 − d = 0

(3.9a)

and differentiate it with respect to time to get acceleration:

( ja α 2e jθ

2

) (

)

+ j 2 a ω 22 e jθ2 − jb α 3e jθ3 + j 2 b ω 32 e jθ3 − d = 0

(3.10 a)

Substitute the Euler identity (equation 3.3) into equation 3.10a, separate into real and imaginary components and solve simultaneously to get: α3 =

3

VA c

VA

µ

R4

Rs

a



x

VB

µ

b

VB

4

B

a α 2 cos θ2 − a ω 22 sin θ2 + b ω 32 sin θ3 b cos θ3

d = − a α 2 sin θ2 − a ω 22 cos θ2 + b α 3 sin θ3 + b ω 32 cos θ3

(3.10i ) (3.10 j)

92

MACHINE DESIGN

θ3

α3 ω3

Y

-

t

AA

A

n

AA

a

b

R2

θ2

ω2

4

x AB

n

ABA c

AA

Sixth Edition

y B

R3

-

t

ABA

AB

3

An Integrated Approach

R4

AA

t

ABA

n

AA

θ4

α2 d

O2

t

n

AA

ABA X

R1

(b)

(a)

FIGURE 3-16

Copyright © 2018 Robert L. Norton: All Rights Reserved

Acceleration Vectors and Vector Loop on a Crank-Slider Linkage

A position and velocity analysis must be done before solving for acceleration. The linear accelerations are:

( = (A

) ( + A ) = (b α

A A = AtA + A nA = aα 2 je A BA

t BA

n BA

3

j θ2

je

− aω 22 e

jθ3

j θ2

− b ω 32 e

)

jθ3

)

(3.10 k )

A B = AtB = d

Substitute the Euler identity (equation 3.3) into equation 3.10k to calculate the normal and tangential components of acceleration. Figure 3-16 shows the acceleration vectors on the linkage. A complete derivation of these equations showing equations 3.10b through 3.10h can be found in the file Slider Acceleration Derivation.pdf on the website. Other Linkages The vector-loop method can be used to analyze many other configurations of linkages. Space does not allow the presentation of derivations for other linkages such as the inverted crank-slider, slider-crank, geared fivebar, or the sixbar linkage. These analyses can be found in reference 9. Also, the Student Edition of program Linkages, written by the author, is provided on the website and can be used to solve a number of linkage configurations including the fourbar linkage, fourbar slider, sixbar, and geared fivebar linkages.

† http://www.designofmachinery.com/MD/Cam_Machine.mp4

3.8

CAM DESIGN AND ANALYSIS View a Video of a Cam Machine

(21:28)†

Cam-follower systems are a variant on the fourbar linkage as shown in Figure 3-8. Proper cam design requires that the higher derivatives of position be considered in defining the functions that control follower motion. The position function s defines the cam’s contour

Chapter 3

KINEMATICS AND LOAD DETERMINATION

93

but not the follower’s dynamic behavior. Velocity, acceleration, and its derivative, called jerk, need to be properly defined to control follower dynamics and avoid unwanted forces and vibrations in follower motion. Cams are particularly useful when a dwell in the follower motion is required. A dwell is defined as the cessation of follower motion while the input motion (cam rotation) continues for a portion of the cycle. This is very easy to obtain with a cam because a constant cam radius over some angle keeps the follower stationary. Any combination of motion and dwells is possible but two arrangements are the most common. A single-dwell, or rise-fall-dwell (RFD) cam is used to actuate the valves in an internal combustion engine (ICE). The rise-fall motion opens and closes the valve at the right point in the cycle to allow the air-fuel mixture in or out of the cylinder, and the dwell holds the valve closed during the compression and power strokes of the cycle. A double-dwell, or rise-dwell-fall-dwell (RDFD) cam is common in automated assembly machines in which the follower picks up a part from a feeder during one dwell, moves the part into position during the rise, places it during the second dwell, and returns (falls) to pick up another part. Different mathematical functions are needed for single-dwell and double-dwell motions but both must obey the Fundamental Law of Cam Design, which states that the motion function, including dwells, must be piecewise continuous over the entire cam through the second derivative of displacement (acceleration). In other words, there can be no discontinuities in the position, velocity, or acceleration functions over the full 360° of the cam. Over the last century, many functions have been proposed and used for double dwell cam motions. Some obey the fundamental law and some do not. These are dealt with in detail in references 9 and 10, but space does not permit their exposition here. Here we will show how to apply one type of function to cam motion, the polynomial, which is superior to the classical double-dwell functions in use and is also useful for the single-dwell case. An advantage of polynomials for cam motion is their adaptability to different constraints. We can tailor a polynomial to any situation by properly applying a set of boundary conditions (BC) to the problem and then use those BC to calculate the necessary coefficients of a polynomial that will adhere to them. For situations in which the polynomial functions do not work well, B-Spline functions will solve the problem in a superior way. These are addressed in detail in reference 10. The Timing Diagram A cam design problem typically starts with a timing diagram, which shows the timing of events such as rises, dwells, and falls over one cycle, defined as one full revolution of the cam. The circular cam is represented on a linear time (or angle) axis that corresponds to the circumference of a circle on the cam. The motions start at this prime circle and grow outward from it. Eventually, this linear axis will be “wrapped around” the circle to create the cam. This diagram does not indicate what functions will be used for the motions. That decision is left to the designer. Figure 3-17 shows an example of a timing diagram for a double-dwell cam. The durations, phasing, and maximum excursion (lift) of the various events (rise, fall, dwells) are defined but the rise and fall motions are not.

3

94

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Motion mm or in High dwel l

1

3

Rise

Low dwel l

Fa ll

0 0

90

180

270

0

0.25

0.50

0.75

FIGURE 3-17

360

Cam angle θ deg

1.0

Time

t sec

Copyright © 2018 Robert L. Norton: All Rights Reserved

A Cam Timing Diagram

The svaj Diagram Cam functions are defined and displayed using svaj diagrams where s is displacement, v is velocity, a is acceleration, and j is jerk, each of the last three being the time derivative of the one preceding it. These are drawn one above the other to show their proper phase relationship and derivative character. The cam must be designed piece by piece. That is each motion (rise, fall, rise-fall, or dwell) must be handled as a separate function. They are then connected to create a piecewise continuous function over the entire interval that obeys the fundamental law of cam design. Figure 3-18 shows an example of an svaj diagram for a rise motion that uses a 3-4-5 polynomial displacement function.

s v

a

j

Polynomials for the Double-Dwell Case 0

cam angle θ

β

Copyright © 2018 Robert L. Norton: All Rights Reserved

F I G U R E 3 - 18 An svaj Diagram for a 3-4-5 Polynomial Displacement

The general form of a polynomial displacement function is: s = C0 + C1 x + C2 x 2 + C3 x 3 + C4 x 4 + C5 x 5 + C6 x 6 +  + Cn x n

(3.11)

where s is follower displacement in length units, x is the independent variable, and the Cn are the coefficients to be determined based on our choice of boundary conditions. The order of a polynomial is equal to the number of terms used and its degree is the highest power present, which is always one less than the order because the first term has the variable to the zeroth power. The order needed for a particular design is equal to the number of boundary conditions (BC) chosen. The independent variable for cam work is usually set to x = q/b where q is the local cam angle for the interval and b is the duration angle of the motion interval. This makes x a normalized variable that runs from zero to one over the interval. To restore the proper units in the results, multiply velocity by w / b and acceleration by w2 / b2 with b in radians and w in rad/sec to get them in terms of time. Figure 3-19 shows the minimum number of BC needed for a double-dwell cam in order to satisfy the fundamental law. Because the dwells have zero velocity and acceleration, the rise and fall functions must have zero BC at the start (0°) and end (b1 or b2) to avoid any discontinuities through acceleration. The displacements must also match those of the dwells at each end of rise or fall. Thus, for a rise or fall between dwells, the minimum number of BC is six to obey the fundamental law. This means that the lowest order polynomial possible for this case is 6th order or 5th degree.

Chapter 3

s

0 v (b)

High dwell

Rise

h

(a)

KINEMATICS AND LOAD DETERMINATION

Low dwell

Fall

β1

0

β2

0

θ deg

0

(c)

β1

0

β2

0

0

θ deg

β1

j 0 (d)

3

θ deg

a

β2

0

0

θ deg

90

0

180

FIGURE 3-19

95

270

360

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Minimum Set of Boundary Conditions for a Rise or Fall Between Dwells

EXAMPLE 3-1 The 3-4-5 Polynomial for a Double Dwell Cam Problem

Determine the coefficients of a 6th order, 5th degree polynomial to rise h mm over 90° between a low and high dwell.

Given

q is the local cam angle, b is the duration angle of the rise, and q/b is the normalized independent variable that runs from zero to one over the interval.

Assumptions

The fundamental law of cam design must be observed.

Solution

See Figures 3-18 and 3-19.

1 Figure 3-19 shows the desired boundary conditions for the rise to maintain continuity through accelerations. They are: when

θ = 0;

then

s = 0,

v = 0,

a=0

when

θ = β1;

then

s = h,

v = 0,

a=0

(a)

The corresponding set for the fall would be: when

θ = 0;

then

s = h,

v = 0,

a=0

when

θ = β2 ;

then

s = 0,

v = 0,

a=0

(b)

2 Write six terms of equation 3.11 in terms of the normalized variable, then differentiate it twice because our BC involve velocity and acceleration.

96

MACHINE DESIGN

-

An Integrated Approach

2

-

Sixth Edition

3

4

 θ  θ  θ  θ  θ s = C0 + C1   + C2   + C3   + C4   + C5    β  β  β  β  β v=

 θ  θ  θ  θ 1 C1 + 2C2   + 3C3   + 4C4   + 5C5   β  β  β  β  β 

a=

2 3  θ  θ  θ  1   2 6 12 20 + C + C C C + 2 3  4  5    β  β  β  β 2  

2

3

3

5

4

 

(c)

(d )

(e)

3 Substitute the boundary condition q = 0, s = 0 into equation (c): 0 = C0 + 0 + 0 +  C0 = 0

(f )

4 Substitute q = 0, v = 0 into equation (d ): 1 C1 + 0 + 0 +  β C1 = 0

(

0=

) (g)

5 Substitute q = 0, a = 0 into equation (e ): 0=

1

β2 C2 = 0

(C2 + 0 + 0 + ) (h)

6 Substitute q = b, s = h into equation (c): h = C3 + C4 + C5

(i )

7 Substitute q = b, v = 0 into equation (d ): 0=

1 (3C3 + 4C4 + 5C5 ) β

(j )

8 Substitute q = b, a = 0 into equation (e ): 0=

1 β2

(6C3 + 12C4 + 20C5 )

(k )

9 Three of the six coefficients are zero. Solve equations i, j, and k simultaneously to find the other three: C3 = 10 h;

C4 = −15h;

C5 = 6h

(l )

10 The equation for displacement is then: 4 5   θ3  θ  θ  s = h 10   − 15   + 6     β  β    β  

(m)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

11 The expressions for velocity and acceleration can be found by substituting the coefficients of equation f, g, h, and l into equations d and e. This function is called the 3-4-5 polynomial after its exponents. Figure 3-18 (repeated here) shows this function and its derivatives through jerk. The displacement is 5th degree, the velocity 4th degree, acceleration 3rd degree (a cubic), and jerk is 2nd degree (a parabola).

97

s v

3 We did not constrain the jerk in the previous example, but we could have. Polynomials offer much flexibility in design as we can change the functions simply by changing the boundary conditions. The advantage of making the jerk zero at the interfaces with the dwells is to reduce vibrations, as they are highly influenced by the jerk function. However, doing so will increase the peak acceleration and that may be undesirable. It is the designer’s choice and there are always trade-offs. If we add two boundary conditions to the previous example to force j = 0 at q = 0 and q = b, we will then have eight BC and eight terms in the equation, which becomes 7th degree. Solving for the new coefficients with these constraints gives the following equation for displacement: 7

  θ  θ  θ  θ s = h 35   − 84   + 70   − 20     β  β  β    β   4

5

6

a

j

0

cam angle θ

β

Copyright © 2018 Robert L. Norton: All Rights Reserved

F I G U R E 3 - 18 Repeated An svaj Diagram for a 3-4-5 Polynomial Displacement

(3.12) s

This is called the 4-5-6-7 polynomial and is one of the lowest vibration double dwell functions available. See Figure 3-20. If the larger peak accelerations associated with this function do not cause the cam or follower to be too highly stressed, then this is an excellent choice, especially for high speed mechanisms. Otherwise, the 3-4-5 polynomial is a good all-around choice for double-dwell motions and has reasonably low vibration as compared to many other common double-dwell functions in use today.

v

a

j

Polynomials for the Single-Dwell Case Cam functions that are good for double dwells are not a good choice for a single dwell case. If the specification calls for a symmetrical rise-fall motion, e.g., the same angle of rise as fall, then a sixth degree polynomial provides the best solution if jerk is unconstrained. Adding jerk constraints will raise it to eighth degree and is the best solution for that case as well. The flexibility of polynomials allows a rise-fall motion with a single expression. Consider the following set of boundary conditions for a single-dwell cam: when q = 0,

s = 0,

when q = b/2,

s=h

when q = b,

s = 0,

v = 0,

a=0

v = 0,

a=0

The zero values at q = 0 and q = b are necessary to match the dwell at those points and the lift value at the midpoint (q = b / 2) creates the rise-fall. Because of symmetry, it is not necessary to specify the slope (velocity) to be zero at b / 2, as it will be so by default. The equation for this function is:   θ  θ  θ  θ s = h 64   − 192   + 192   − 64    β  β  β   β  3

4

5

6

 

(3.13)

This function and its derivatives are shown in Figure 3-21 with its seven BC circled.

0

cam angle θ

β

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 3-20 The 4-5-6-7 Polynomial

98

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Note that the above solution did not create separate functions for rise and fall, but rather, used a single polynomial for both rise and fall. But, if the rise and fall are not the same duration, making the function asymmetrical, then you will have to use separate equations for rise and fall. Moreover, to make them work, you need to start with the motion that has the longer duration, whether rise or fall. Also, you need to leave the acceleration unspecified for the first segment calculated (the one of larger b) and determine its value from the function that results. That value is then used as a BC on the second segment to be calculated in order to match the accelerations at the joint between the segments and maintain mathematical continuity.

s

v

3 a

j 0

β/2

β

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 3-21 3-4-5-6 Polynomial for Symmetrical Single-Dwell Cam

EXAMPLE 3-2 An Asymmetric Rise-Fall Single-Dwell Cam Problem

Design a single-dwell cam function to rise h units in 45°, fall h units in 135 and dwell for the remainder. Let h = 1 and cam w = 15 rad/sec.

Given

q is the local cam angle, b is the duration angle of the rise, and q/b is the normalized independent variable that runs from zero to one over an interval.

Assumptions

The fundamental law of cam design must be observed.

Solution

See Figure 3-22.

1 The rise and fall must be handled with separate polynomials. The longer duration motion (here the fall) must be done first. Set the boundary conditions for the fall to be: when

θ = 0;

then

s = h,

v=0

when

θ = β2 ;

then

s = 0,

v = 0,

a=0

(a)

Note that the acceleration is left unspecified at the start point. There are 5 BC for this segment, which will give a 4th degree polynomial. 2 Substitute the BC in the polynomial equation and its derivatives and solve them for the coefficients. The result is:   θ  θ  θ s = h 1 − 6   + 8   − 3    β  β  β  2

3

4

 

(b)

3 Differentiate this equation twice to get the acceleration function and solve it at q = 0 for the given data. Multiply the acceleration equation by w2 / b2 with b in radians and w in rad/sec to get the result in terms of time. The acceleration at the start of the fall is –486.4 in/sec. This value must be used as a boundary condition on acceleration for the rise segment in order to maintain continuity. The BC for the rise then become: when

θ = 0;

then

s = h,

v = 0,

a=0

when

θ = β1;

then

s = 0,

v = 0,

a = −486.4

(c)

4 Substitute these BC in the polynomial equation and its derivatives and solve them for the coefficients for the rise. The result is:

Chapter 3

segment

1

KINEMATICS AND LOAD DETERMINATION

2

99

3

h = 1 at 45 ° s Segment 1 6 boundary conditions

Segment 2 5 boundary conditions

3 v

a

– 486. 4 in/sec 2

seg 1 only

– 2024 in/sec 2

j 0

45

180

FIGURE 3-22

360

Copyright © 2018 Robert L. Norton: All Rights Reserved

Asymmetric Single-Dwell Polynomial Solution to Example 3-2

3 4 5   θ  θ  θ  s = h 9.333   − 13.667   + 5.333     β  β  β   

(d )

5 The result is shown in Figure 3-22.

Pressure Angle Figure 3-23 shows a cam and follower. The kinematic cam is defined by the pitch curve through the center of the roller. The cam surface is found as a normal offset from this curve by the roller radius. The prime circle radius defines the minimum cam radius and its circumference corresponds to the axis of the s diagram. The pressure angle is the angle between the velocity of the follower and the common normal at the contact point between cam and follower. The force between cam and follower lies along the common normal. So the pressure angle determines what percent of the force goes into motion and what percent is trying to jam the follower sideways in its guide. The rule of thumb is to keep the maximum pressure angle below about 30° for a translating follower. The eccentricity of the follower is defined as the offset of the follower axis from the cam centerline, shown as e in Figure 3-23. The pressure angle f varies around the cam and can be found from: φ = arctan

v−ε s + RP2 − ε 2

(3.14)

100

MACHINE DESIGN

-

An Integrated Approach

Pressure angle

φ

-

Sixth Edition

Vfollower Follower

3

Common normal (axis of transmission)

Common tangent (axis of slip )

A

Pitch curve φ

Prime circle radius R p

ω cam

O2

Eccentricity FIGURE 3-23

ε

Fo llower axis of motion Copyright © 2018 Robert L. Norton: All Rights Reserved

Cam Pressure Angle

where v is the follower velocity in length/radian, e is the eccentricity, s is the displacement in length units, and Rp is the prime circle radius. Once the svaj functions are defined, the only things in this equation that the designer can change are Rp and e. Increasing Rp will reduce the pressure angle as it effectively increases the length of the s diagram. Eccentricity will shift the pressure angle function around zero but will have no effect on the peak to peak range. Radius of Curvature All mathematical functions possess a property called radius of curvature, denoted by r, and it can be positive (convex) or negative (concave). At any point on the function, in the limit, the function can be considered to be an infinitesimally short arc of particular radius. The relationship between this instantaneous radius of the cam contour and the radius of the roller follower is important to proper cam function. If the cam’s radius of curvature is smaller than that of the follower, then the follower will not be able to maintain proper contact with the cam. Two cases are possible. If the cam contour is concave (negative) and its r is smaller than the roller radius Rf, the condition shown in Figure 3-24 will obtain. If the cam

Chapter 3

KINEMATICS AND LOAD DETERMINATION

101

contour is convex (positive) and its r is equal to or smaller than the roller radius Rf, the conditions shown in Figure 3-25 will result. These conditions are called undercutting and result in a cusp on the cam surface that will cause the follower to jump off the cam. The radius of curvature of the cam pitch curve, which is the locus of the center of the roller as shown in Figure 3-23, can be calculated from this equation: ρ pitch =

( R + s )2 + v 2    P

3

32

( RP + s )2 + 2v 2 − a ( RP + s )

(3.15)

where s is displacement in length units, v is the follower velocity in length/radian, a is the follower acceleration in length/radian2, and Rp is the prime circle radius. Again, once svaj are defined, the designer can only control radius of curvature by changing Rp. The rule of thumb is to keep the absolute value of the minimum r of the pitch curve greater than the roller radius Rf by some comfort factor: ρmin >> R f

(3.16)

A factor of 2 is good. Less than 1.5 will be troublesome. See reference 10 for more information on cams and cam design. The Student Edition of program Dynacam, written by the author, is included on the book’s website. This program will calculate the polynomial coefficients for any set of boundary conditions and also provides a large number of additional cam functions. It calculates the pressure angle and radius of curvature and draws the cam profile.

3.9

LOADING CLASSES FOR FORCE ANALYSIS

The type of loading on a system can be divided into several classes based on the character of the applied loads and the presence or absence of system motion. Once the general configuration of a mechanical system is defined and its kinematic motions calculated, the next task is to determine the magnitudes and directions of all the forces and couples Rf >

ρmin

ρmin

Vfollower

ω cam

Cam FIGURE 3-24

Fo llower Copyright © 2018 Robert L. Norton: All Rights Reserved

Result of the Roller Radius Being Larger Than the Smallest Negative Radius of Curvature of the Cam

102

MACHINE DESIGN

Follower

3

-

An Integrated Approach

Cusp due to undercutting

-

Sixth Edition

Missing material and cusp due to undercutting Follower Pitch curve

Pitch curve

Cam surface (a) Radius of curvature of pitch curve equals the radius of the roller follower

FIGURE 3-25

(b) Radius of curvature of pitch curve is less than the radius of the roller follower Copyright © 2018 Robert L. Norton: All Rights Reserved

Undercutting Due to Cam Radius of Curvature Being Less Than or Equal to the Follower Radius

present on the various elements. These loads may be constant or may be varying over time. The elements in the system may be stationary or moving. The most general class is that of a moving system with time-varying loads. The other combinations are subsets of the general class. Table 3-1 shows the four possible classes. Class 1 is a stationary system with constant loads. One example of a Class 1 system is the base frame for an arbor press used in a machine shop. The base is required to support the dead weight of the arbor press, which is essentially constant over time, and the base frame does not move. The parts brought to the arbor press (to have something pressed into them) temporarily add their weight to the load on the base, but this is usually a small percentage of the dead weight. A static load analysis is all that is necessary for a Class 1 system. Class 2 describes a stationary system with time-varying loads. An example is a bridge which, though essentially stationary, is subjected to changing loads as vehicles drive over it and wind impinges on its structure. Class 3 defines a moving system with constant loads. Even though the applied external loads may be constant, any significant accelerations of the moving members can create time-varying reaction forces. An example might be a powered rotary lawn mower. Except for the case of mowing the occasional rock, the blades experience a nearly constant external load from mowing the grass. However, the accelerations of the spinning blades can create high loads at their fastenings. A dynamic load analysis is necessary for Classes 2 and 3.

Chapter 3

Table 3-1

KINEMATICS AND LOAD DETERMINATION

103

Load Classes Constant Loads

Time-Varying Loads

Stationary Elements

Class 1

Class 2

Moving Elements

Class 3

Class 4

3

Note however that, if the motions of a Class 3 system are so slow as to generate negligible accelerations on its members, it could qualify as a Class 1 system and then would be called quasi-static. An automobile scissors jack (see Figure 3-5) can be considered to be a Class 1 system since the external load (when used) is essentially constant, and the motions of the links are slow with negligible accelerations. The only complexity introduced by the motions of the elements in this example is that of determining in which position the internal loads on the jack’s elements will be maximal, since they vary as the jack is raised, despite the essentially constant external load. Class 4 describes the general case of a rapidly moving system subjected to timevarying loads. Note that even if the applied external loads are essentially constant in a given case, the dynamic loads developed on the elements from their accelerations will still vary with time. Most machinery, especially if powered by a motor or engine, will be in Class 4. An example of such a system is the engine in your car. The internal parts (crankshaft, connecting rods, pistons, etc.) are subjected to time-varying loads from the gasoline explosions, and also experience time-varying inertial loads from their own accelerations. A dynamic load analysis is necessary for Class 4.

3.10

FREE-BODY DIAGRAMS

In order to correctly identify all potential forces and moments on a system, it is necessary to draw accurate free-body diagrams (FBDs) of each member of the system. These FBDs should show a general shape of the part and display all the forces and moments that are acting on it. There may be external forces and moments applied to the part from outside the system, and there will be interconnection forces and/or moments where each part joins or contacts adjacent parts in the assembly or system. In addition to the known and unknown forces and couples shown on the FBD, the dimensions and angles of the elements in the system are defined with respect to local coordinate systems located at the centers of gravity (CG) of each element.* For a dynamic load analysis, the kinematic accelerations, both angular and linear (at the CG), need to be known or calculated for each element prior to doing the load analysis.

3.11

LOAD ANALYSIS

This section presents a brief review of Newton’s laws and Euler’s equations as applied to dynamically loaded and statically loaded systems in both 3-D and 2-D. The method of

*

While it is not a requirement that the local coordinate system for each element be located at its CG, this approach provides consistency and simplifies the dynamic calculations. Further, most solid modeling CAD/CAE systems will automatically calculate the mass properties of parts with respect to their CGs. The approach taken here is to apply a consistent method that works for both static and dynamic problems and that is also amenable to computer solution.

104

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

solution presented here may be somewhat different than that used in your previous statics and dynamics courses. The approach taken here in setting up the equations for force and moment analysis is designed to facilitate computer programming of the solution.

3

This approach assumes all unknown forces and moments on the system to be positive in sign, regardless of what one’s intuition or an inspection of the free-body diagram might indicate as to their probable directions. However, all known force components are given their proper signs to define their directions. The simultaneous solution of the set of equations that results will cause all the unknown components to have the proper signs when the solution is complete. This is ultimately a simpler approach than the one often taught in statics and dynamics courses, which requires that the student assume directions for all unknown forces and moments (a practice that does help the student develop some intuition, however). Even with that traditional approach, an incorrect assumption of direction results in a sign reversal on that component in the solution. Assuming all unknown forces and moments to be positive allows the resulting computer program to be simpler than would otherwise be the case. The simultaneous equation solution method used is extremely simple in concept, though it requires the aid of a computer to solve. Software is provided with the text to solve the simultaneous equations. See program Matrix on the website. Real dynamic systems are three dimensional and thus must be analyzed as such. However, many 3-D systems can be analyzed by simpler 2-D methods. Accordingly, we will investigate both approaches. Three-Dimensional Analysis Since three of the four cases potentially require dynamic load analysis, and because a static force analysis is really just a variation on the dynamic analysis, it makes sense to start with the dynamic case. Dynamic load analysis can be done by any of several methods, but the one that gives the most information about internal forces is the Newtonian approach based on Newton’s laws. NewtoN’s First Law A body at rest tends to remain at rest and a body in motion at constant velocity will tend to maintain that velocity unless acted upon by an external force. NewtoN’s secoNd Law The time rate of change of momentum of a body is equal to the magnitude of the applied force and acts in the direction of the force. Newton’s second law can be written for a rigid body in two forms, one for linear forces and one for moments or torques:

∑ F = ma

∑ MG = H G

(3.17a)

where F = force, m = mass, a = acceleration, MG = moment about the center of gravity and H G = the time rate of change of the moment of momentum, or the angular momentum about the CG. The left sides of these equations respectively sum all the forces and moments that act on the body, whether from known applied forces or from interconnections with adjacent bodies in the system.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

105

For a three-dimensional system of connected rigid bodies, this vector equation for the linear forces can be written as three scalar equations involving orthogonal components taken along a local x, y, z axis system with its origin at the CG of the body:

∑ Fx = max

∑ Fy = may

∑ Fz = maz

(3.17b)

If the x, y, z axes are chosen coincident with the principal axes of inertia of the body,* the angular momentum of the body is defined as HG = I x ω x ˆi + I y ω y ˆj + I z ω z kˆ

(3.17c)

where Ix, Iy, and Iz are the principal centroidal mass moments of inertia (second moments of mass) about the principal axes. This vector equation can be substituted into equation 3.17a to yield the three scalar equations known as Euler’s equations:

∑ M x = I x α x − ( I y − I z ) ω yω z ∑ M y = I yα y − ( I z − I x ) ω z ω x ∑ Mz = Izαz − (Ix − I y )ω xω y

(3.17d )

where Mx, My, Mz are moments about those axes and ax, ay, az are the angular accelerations about the axes. This assumes that the inertia terms remain constant with time, i.e., the mass distribution about the axes is constant. NewtoN’s third Law When two particles interact, a pair of equal and opposite reaction forces will exist at their contact point. This force pair will have the same magnitude and act along the same direction line, but have opposite sense. We will need to apply this relationship as well as applying the second law in order to solve for the forces on assemblies of elements that act upon one another. The six equations in equations 3.17b and 3.17d can be written for each rigid body in a 3-D system. In addition, as many (third-law) reaction force equations as are necessary will be written and the resulting set of equations solved simultaneously for the forces and moments. The number of second-law equations will be up to six times the number of individual parts in a three-dimensional system (plus the reaction equations), meaning that even simple systems result in large sets of simultaneous equations. A computer is needed to solve these equations, though high-end pocket calculators will solve large sets of simultaneous equations also. The reaction (third-law) equations are often substituted into the secondlaw equations to reduce the total number of equations to be solved simultaneously. Two-Dimensional Analysis All real machines exist in three dimensions but many three-dimensional systems can be analyzed two dimensionally if their motions exist only in one plane or in parallel planes. Euler’s equations 3.17d show that if the rotational motions (w, a) and applied moments or couples exist about only one axis (say the z axis), then that set of three equations reduces to one equation,

∑ Mz = Izαz

(3.18a)

* This is a convenient choice for symmetric bodies but may be less convenient for other shapes. See F. P. Beer and E. R. Johnson, Vector Mechanics for Engineers, 3rd ed., 1977, McGraw-Hill, New York, Chap. 18, “Kinetics of Rigid Bodies in Three Dimensions.”

3

106

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

because the w and a terms about the x and y axes are now zero. Equation 3.17b is reduced to

∑ Fx = max 3

∑ Fy = may

(3.18b)

Equations 3.18 can be written for all the connected bodies in a two-dimensional system and the entire set solved simultaneously for forces and moments. The number of secondlaw equations will now be up to three times the number of elements in the system plus the necessary reaction equations at connecting points, again resulting in large systems of equations for even simple systems. Note that even though all motion is about one (z) axis in a 2-D system, there may still be loading components in the z direction due to external forces or couples. Static Load Analysis The difference between a dynamic loading situation and a static one is the presence or absence of accelerations. If the accelerations in equations 3.17 and 3.18 are all zero, then for the three-dimensional case these equations reduce to

∑ Fx = 0 ∑ Mx = 0

∑ Fy = 0 ∑ My = 0

∑ Fz = 0 ∑ Mz = 0

(3.19a)

∑ Fy = 0

∑ Mz = 0

(3.19b)

and for the two-dimensional case,

∑ Fx = 0

Thus, we can see that the static loading situation is just a special case of the dynamic loading one, in which the accelerations happen to be zero. A solution approach based on the dynamic case will then also satisfy the static one with appropriate substitutions of zero values for the absent accelerations.

3.12

TWO-DIMENSIONAL, STATIC LOADING CASE STUDIES

This section presents a series of three case studies of increasing complexity, all limited to two-dimensional static loading situations. A bicycle handbrake lever, a crimping tool, and a scissors jack are the systems analyzed. These case studies provide examples of the simplest form of force analysis, having no significant accelerations and having forces acting in only two dimensions.

C A S E

S T U D Y

1 A

Bicycle Brake Lever Loading Analysis Problem

Determine the forces on the elements of the bicycle brake lever assembly shown in Figure 3-26 during braking.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

107

Fb 2 Fcable

Fsheath

3

brake lever

2

cable

pivot Px

Mh

3

1 Py

handlebar

handgrip Fb 1

FIGURE 3-26 Bicycle Brake Lever Assembly

Given

The geometry of each element is known. The average human’s hand can develop a grip force of about 267 N (60 lb) in the lever position shown.

Assumptions

The accelerations are negligible. All forces are coplanar and two dimensional. A Class 1 load model is appropriate and a static analysis is acceptable.

Solution

See Figures 3-26, 3-27, and Table 3-2, parts 1 and 2.

1 Figure 3-26 shows the handbrake lever assembly, which consists of three subassemblies: the handlebar (1), the lever (2), and the cable (3). The lever is pivoted to the handlebar and the cable is connected to the lever. The cable runs within a plastic-lined sheath (for low friction) down to the brake caliper assembly at the bicycle’s wheel rim. The sheath provides a compressive force to balance the tension in the cable (Fsheath = –Fcable). The user’s hand applies equal and opposite forces at some points on the lever and handgrip. These forces are transformed to a larger force in the cable by the lever ratio of part 2. Figure 3-26 is a free-body diagram of the entire assembly since it shows all the forces and moments potentially acting on it except for its weight, which is small compared to the applied forces and is thus neglected for this analysis. The “broken away” portion of the handlebar can provide x and y force components and a moment if required for equilibrium. These reaction forces and moments are arbitrarily shown as positive in sign. Their actual signs will “come out in the wash” in the calculations. The known applied forces are shown acting in their actual directions and senses. 2 Figure 3-27 shows the three subassembly elements separated and drawn as free-body diagrams with all relevant forces and moments applied to each element, again neglecting the weights of the parts. The lever (part 2) has three forces on it, Fb2, F32, and F12. The two-character subscript notation used here should be read as, force of element 1 on 2 (F12) or force at B on 2 (Fb2), etc. This defines the source of the force (first subscript) and the element on which it acts (second subscript). This notation will be used consistently throughout this text for both forces and position vectors such as Rb2, R32, and R12 in Figure 3-27, which serve to locate the above three

108

MACHINE DESIGN

-

An Integrated Approach

θ C

F32

Fb2

Sixth Edition

y

R32

F32

-

2

x

B

F12

3

A

F12

Vector diagram for brake lever

R12

3 F23

1 2°

F31 Fsheath Rd Mh

R31 F21

y A

D

R21

Px Rp Py

FIGURE 3-27

Rb2



F13

φ

Fcabl e

Fb2

x

1 Rb1

Fb1

Bicycle Brake Lever Free-Body Diagrams *

Actually, for a simple static analysis such as the one in this example, any point (on or off the element) can be taken as the origin of the local coordinate system. However, in a dynamic force analysis it simplifies the analysis if the coordinate system is placed at the CG. So, for the sake of consistency, and to prepare for the more complicated dynamic analysis problems ahead, we will use the CG as the origin even in the static cases here. †

You may not have done this in your statics class but this approach makes the problem more amenable to a computer solution. Note that regardless of the direction shown for any unknown force on the FBD, we will assume its components to be positive in the equations. The angles of the known (given) forces (or the signs of their components) do have to be correctly input to the equations, however.

forces in a local, nonrotating coordinate system whose origin is at the center of gravity (CG) of the element or subassembly being analyzed.* On this brake lever, Fb2 is an applied force whose magnitude and direction are known. F32 is the force in the cable. Its direction is known but not its magnitude. Force F12 is provided by part 1 on part 2 at the pivot pin. Its magnitude and direction are both unknown. We can write equations 3.3b for this element to sum forces in the x and y directions and sum moments about the CG. Note that all unknown forces and moments are initially assumed positive in the equations. Their true signs will come out in the calculation.† However, all known or given forces must carry their proper signs:

∑ Fx = F12 x + Fb2 x + F32 x = 0 ∑ Fy = F12 y + Fb2 y + F32 y = 0 ∑ Mz = ( R12 × F12 ) + ( R b2 × Fb2 ) + ( R 32 × F32 ) = 0

(a)

The cross products in the moment equation represent the “turning forces” or moments created by the application of these forces at points remote from the CG of the element. Recall that these cross products can be expanded to

∑ M z = ( R12 x F12 y − R12 y F12 x ) + ( Rb2 x Fb2 y − Rb2 y Fb2 x ) + ( R32 x F32 y − R32 y F32 x ) = 0

(b)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

109

We have three equations and four unknowns (F12x, F12y, F32x, F32y) at this point, so we need another equation. It is available from the fact that the direction of F32 is known. (The cable can pull only along its axis.) We can express one component of the cable force F32 in terms of its other component and the known angle, q of the cable: F32 y = F32 x tan θ

(c)

3 We could now solve the four unknowns for this element, but will wait to do so until the equations for the other two links are defined. 3 Part 3 in Figure 3-27 is the cable that passes through a hole in part 1. This hole is lined with a low-friction material that allows us to assume no friction at the joint between parts 1 and 3. We will further assume that the three forces F13, F23, and Fcable form a concurrent system of forces acting through the CG and thus create no moment. With this assumption only a summation of forces is necessary for this element:

∑ Fx = Fcable ∑ Fy = Fcable

x

+ F13 x + F23 x = 0

y

+ F13 y + F23 y = 0

(d )

4 The assembly of elements labeled part 1 in Figure 3-27 can have both forces and moments on it (i.e., it is not a concurrent system), so the three equations 3.19b are needed:

∑ Fx = F21x + Fb1x + F31x + Px + Fsheath = 0 ∑ Fy = F21y + Fb1y + F31y + Py = 0 ∑ Mz = Mh + ( R 21 × F21 ) + ( R b1 × Fb1 ) + ( R 31 × F31 ) + ( R p × P ) + ( R d × Fsheath ) = 0 x

(e)

Expanding cross products in the moment equation gives the moment magnitude as

∑ M z = M h + ( R21x F21y − R21y F21x ) + ( Rb1x Fb1y − Rb1y Fb1x ) + ( R31x F31y − R31y F31x ) + ( RPx Py − RPy Px ) + ( Rdx Fsheath − Rdy Fsheath ) = 0 y

(f)

x

5 The total of unknowns at this point (including those listed in step 2 above) is 21: Fb1x, Fb1y, F12x, F12y, F21x, F21y, F32x, F32y, F23x, F23y, F13x, F13y, F31x, F31y, Fcablex, Fcabley, Fsheathx, Fsheathy,Px, Py, and Mh. We have only nine equations so far, three in equation set (a), one in set (c), two in set (d), and three in set (e). We need twelve more equations to solve this system. We can get seven of them from the Newton’s third-law relationships between contacting elements: F23 x = − F32 x

F23 y = − F32 y

F21x = − F12 x

F21y = − F12 y

F31x = − F13 x

F31y = − F13 y

( g)

Fsheathx = − Fcablex

Two more equations come from the assumption (shown in Figure 3-26) that the two forces provided by the hand on the brake lever and handgrip are equal and opposite:* Fb1x = − Fb 2 x Fb1y = − Fb 2 y

(h)

*

But not necessarily colinear.

110

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

The remaining three equations come from the given geometry and the assumptions made about the system. The direction of the forces Fcable and Fsheath are known to be in the same direction as that end of the cable. In the figure it is seen to be horizontal, so we can set Fcabley = 0;

3

Fsheathy = 0

(i )

Because of our no-friction assumption, the force F31 can be assumed to be normal to the surface of contact between the cable and the hole in part 1. This surface is horizontal in this example, so F31 is vertical and F31x = 0

( j)

F13x

0.0

N

6 This completes the set of 21 equations (equation sets a, c, d, e, g, h, i, and j), and they can be solved for the 21 unknowns simultaneously “as is,” that is, all 21 equations could be put into matrix form and solved with a matrix-reduction computer program. However, the problem can be simplified by manually substituting equations c, g, h, i, and j into the others to reduce them to a set of eight equations in eight unknowns. The known or given data are as shown in Table 3-2, part 1.

Fb2x

0.0

N

7 As a first step, for link 2, substitute equations b and c in equation a to get:

Fb2y

–267.0

N

184.0

deg

180.0

deg

Table 3–2—part 1 Case Study 1A Given Data Variable

θ φ Rb2x

Value

Unit

F12 x + Fb 2 x + F32 x = 0 F12 y + Fb 2 y + F32 x tan θ = 0

( R12 x F12 y − R12 y F12 x ) + ( Rb2 x Fb2 y − Rb2 y Fb2 x ) + ( R32 x F32 x tan θ − R32 y F32 x ) = 0

39.39

mm

Rb2y

2.07

mm

R32x

–50.91

mm

R32y

4.66

mm

R12x

–47.91

mm

Fcablex + F13 x − F32 x = 0

R12y

–7.34

mm

R21x

Fcable y + F13 y − F32 x tan θ = 0

7.0

mm

R21y

19.0

mm

Rb1x

47.5

mm

Rb1y

–14.0

mm

R31x

–27.0

mm

R31y

30.0

mm

Rpx

–27.0

mm

Rpy

0.0

mm

Rdx

–41.0

mm

Rdy

27.0

mm

(k )

8 Next, take equations d for link 3 and substitute equation c and also –F32x for F23x, and –F32y for F23y from equation g to eliminate those variables: (l )

9 For link 1, substitute equation f in e and replace F21x with –F12x, F21y with –F12y, F31x with –F13x , F31y with –F13y, and Fsheathx with –Fcablex from equation g, − F12 x + Fb1x − F13 x + Px − Fcablex = 0

(

) (

− F12 y + Fb1y − F32 x tan θ + Py = 0

M h + − R21x F12 y + R21y F12 x + Rb1x Fb1y − Rb1y Fb1x

(

) (

)

(m)

)

+ − R31x F13 y + R31y F13 x + RPx Py − RPy Px + Rdy Fcablex = 0

10 Finally, substitute equations h, i, and j into equations k, l, and m to yield the following set of eight simultaneous equations in the eight remaining unknowns: F12x, F12y, F32x, F13y, Fcablex, Px, Py, and Mh. Put them in the standard form, which has all unknown terms on the left and all known terms to the right of the equal signs.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

111

F12 x + F32 x = − Fb 2 x F12 y + F32 x tan θ = − Fb 2 y Fcablex − F32 x = 0 F13 y − F32 x tan θ = 0

(n)

− F12 x + Px − Fcablex = Fb 2 x

(

3

− F12 y − F13 y + Py = Fb 2 y

)

R12 x F12 y − R12 y F12 x + R32 x tan θ − R32 y F32 x = − Rb 2 x Fb 2 y + Rb 2 y Fb 2 x M h − R21x F12 y + R21y F12 x − R31x F13 y + RPx Py − RPy Px + Rdy Fcablex = Rb1x Fb 2 y − Rb1y Fb 2 x

11 Form the matrices from equation n:  1  0  0   0  −1  0   − R12 y   R21y 

0 1 0 0 0 −1 R12 x − R21x

R32 x

1 tan θ −1 − tan θ 0 0 tan θ − R32 y 0

0 0 0 1 0 −1 0

0 0 1 0 −1 0 0

0 0 0 0 1 0 0

0 0 0 0 0 1 0

− R31x

Rdy

− RPy

RPx

 F12 x     F12 y     F32 x   F13 y ×   Fcablex     Px   Py 1     M h 0 0 0 0 0 0 0

      =       (o)

 − Fb 2 x  − Fb 2 y   0  0   Fb 2 x  Fb 2 y   −R F + R F b2x b2 y b2 y b2x   Rb1x Fb 2 y − Rb1y Fb 2 x 

            

12 Substitute the known data as shown in Table 3-2 part 1 (repeated opposite):  1 0 1 0 0  0 1 0.070 0 0  −1 0 0 1  0  0 −0.070 1 0 0  −1 0 0 0 −1  −1 −1 0 0  0  7.34 −47.91 −8.22 0 0  19 −7 0 27 27 

0 0 0 0 0 0 0 0 1 0 0 1 0 0 0 −27

0 0 0 0 0 0 0 1

 F12 x   F 12 y     F32 x     F13 y × F   cablex   Px     Py    M h

    0    267    0       0  ( p) = 0    −267      10 517.13       −12 682.50  

112

MACHINE DESIGN

Table 3–2—part 2 Case Study 1A Calculated Data

3

Variable

Value

Unit

F32 x

–1909

N

F32 y

–133

N

F12 x

1909

N

F12 y

400

N

F23 x

1909

N

F23 y

133

N

F13 y

–133

N

Fcabl e x

–1909

N

Fcabl e y

0

N

Fb1x

0

N

Fb1y

267

N

F31 x

0

N

F31 y

133

N

F21 x

–1909

N

F21 y

–400

N

Px

0

N

Py

0

N

Mh

9

N-m

Fsheathx

1909

N

-

An Integrated Approach

-

Sixth Edition

13 The solution is shown in Table 3-2 part 2. This matrix equation can be solved with any one of a number of commercially available matrix solvers such as Mathcad, MATLAB, Maple, or Mathematica, as well as with many engineering pocket calculators. A custom-written program called Matrix is provided on the book’s website that can be used to solve a linear system of up to 16 equations. Equation p was solved with program Matrix to find the 8 unknowns listed in step 10. Those results were then substituted into the other equations to solve for the previously eliminated variables. 14 Table 3-2 part 2 shows the solution data for the data given in Figure 3-27 and Table 3-2 part 1. This assumes a 267-N (60-lb) force is applied by the person’s hand normal to the brake lever. The force generated in the cable (Fcable) is then 1909 N (429 lb) and the reaction force against the handlebar (F21) is 1951 N (439 lb) at –168°.

C A S E

S T U D Y

2 A

Hand-Operated Crimping-Tool Loading Analysis Problem

Determine the forces on the elements of the crimping tool shown in Figure 3-28 during a crimp operation.

Given

The geometry is known and the tool develops a crimp force of 2000 lb (8896 N) at closure in the position shown.

Assumptions

The accelerations are negligible. All forces are coplanar and two dimensional. A Class 1 load model is appropriate and a static analysis acceptable.

Solution

See Figures 3-28 and 3-29, and Table 3-3, parts 1 and 2.

1 Figure 3-28 shows the tool in the closed position in the process of crimping a metal connector onto a wire. The user’s hand provides the input forces between links 1 and 2, shown as the reaction pair Fh. The user can grip the handle anywhere along its length, but we are assuming a nominal moment arm of Rh for the application of the resultant of the user’s grip force (see Figure 3-29). The high mechanical advantage of the tool transforms the grip force to a large force at the crimp. Figure 3-28 is a free-body diagram of the entire assembly, neglecting the weight of the tool, which is small compared to the crimp force. There are four elements, or links, in the assembly, all pinned together. Link 1 can be considered to be the “ground” link, with the other links moving with respect to it as the jaw is closed. The desired magnitude of the crimp force Fc is defined and its direction will be normal to the surfaces at the crimp. The third law relates the action-reaction pair acting on links 1 and 4: Fc1x = − Fc 4 x Fc1y = − Fc 4 y * Again, in a static analysis it is not necessary to take the CG as the coordinate system origin (any point can be used), but we do so to be consistent with the dynamic analysis approach in which it is quite useful to do so.

(a)

2 Figure 3-29 shows the elements of the crimping-tool assembly separated and drawn as free-body diagrams with all forces applied to each element, again neglecting their weights as being insignificant compared to the applied forces. The centers of gravity of the respective elements are used as the origins of the local, nonrotating coordinate systems in which the points of application of all forces on the elements are located.*

Chapter 3

KINEMATICS AND LOAD DETERMINATION

Fh

Table 3–3—part 1 Case Study 2A Given Data

1 2

Fh

113

3

4

force from hand

Fc crimp force

FIGURE 3-28 Wire Connector Crimping Tool

3 We will consider link 1 to be the ground plane and analyze the remaining moving links. Note that all unknown forces and moments are initially assumed positive. Link 2 has three forces acting on it: Fh is the unknown force from the hand, and F12 and F32 are the reaction forces from links 1 and 3, respectively. Force F12 is provided by part 1 on part 2 at the pivot pin and force F32 is provided by part 3 acting on part 2 at their pivot pin. The magnitudes and directions of both these pin forces are unknown. We can write equations 3.3b for this element to sum forces in the x and y directions and sum moments about the CG (with cross products expanded):

∑ Fx = F12 x + F32 x = 0 ∑ Fy = F12 y + F32 y + Fh = 0 ∑ M z = Fh Rh + ( R12 x F12 y − R12 y F12 x ) + ( R32 x F32 y − R32 y F32 x ) = 0

(b)

4 Link 3 has two forces on it, F23 and F43. Write equations 3.3b for this element:

∑ Fx = F23x + F43x = 0 ∑ Fy = F23 y + F43 y = 0 ∑ M z = ( R23x F23 y − R23 y F23x ) + ( R43x F43 y − R43 y F43x ) = 0

(c)

5 Link 4 has three forces acting on it: Fc4 is the known (desired) force at the crimp, and F14 and F34 are the reaction forces from links 1 and 3, respectively. The magnitudes and directions of both these pin forces are unknown. Write equations 3.19b for this element:

∑ Fx = F14 x + F34 x + Fc4 x = 0 ∑ Fy = F14 y + F34 y + Fc4 y = 0 ∑ M z = ( R14 x F14 y − R14 y F14 x ) + ( R34 x F34 y − R34 y F34 x ) + ( Rc 4 x Fc 4 y − Rc 4 y Fc 4 x ) = 0

(d )

6 The nine equations in sets b through d have 13 unknowns: F12x, F12y, F32x, F32y, F23x, F23y, F43x, F43y, F14x, F14y, F34x, F34y, and Fh. We can write the third-law relationships between action-reaction pairs at each of the joints to obtain the four additional equations needed:

Variable

Value

Unit

Fc4x

–1956.30

lb

Fc4y

415.82

lb

R c4x

0.45

in

R c4y

0.34

in

R 12 x

1.40

in

R 12 y

0.05

in

R 32 x

2.20

in

R 32 y

0.08

in

Rh

–4.40

in

R 23 x

–0.60

in

R 23 y

0.13

in

R 43 x

0.60

in

R 43 y

–0.13

in

R 14 x

–0.16

in

R 14 y

–0.76

in

R 34 x

0.16

in

R 34 y

0.76

in

3

114

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

y

Fh

R21 Rc1

x

3 y

R23 F23

B

A

R43

all dimensions in inches

F21 Fc1

3

1

x 1.23

C

R41

F43

F41

D

1.0

y Rh

0.7

F34

F12 A 0.80

R12 R32

Fh

C

R34

B

x 2

y

Rc4

1.55

F32

Fc4

x F14

1.2 4 D

FIGURE 3-29

R14

Free-Body Diagrams of a Wire Connector Crimping Tool

F32 x = − F23 x ;

F34 x = − F43 x ;

F32 y = − F23 y ;

F34 y = − F43 y

(e)

7 The 13 equations b–e can be solved simultaneously by matrix reduction or by iteration with a root-finding algorithm. For a matrix solution, the unknown terms are placed on the left and known terms on the right of the equal signs: F12 x + F32 x = 0 F12 y + F32 y + Fh = 0 Rh Fh + R12 x F12 y − R12 y F12 x + R32 x F32 y − R32 y F32 x = 0 F23 x + F43 x = 0 F23 y + F43 y = 0 R23 x F23 y − R23 y F23 x + R43 x F43 y − R43 y F43 x = 0 F14 x + F34 x = − Fc 4 x F14 y + F34 y = − Fc 4 y R14 x F14 y − R14 y F14 x + R34 x F34 y − R34 y F34 x = − Rc 4 x Fc 4 y + Rc 4 y Fc 4 x F32 x + F23 x = 0 F34 x + F43 x = 0 F32 y + F23 y = 0 F34 y + F43 y = 0

(f)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

115

8 Substitute the given data from Table 3-3 part 1: F12 x + F32 x = 0 F12 y + F32 y + Fh = 0 −4.4 Fh + 1.4 F12 y − 0.05F12 x + 2.2 F32 y − 0.08 F32 x = 0 F23 x + F43 x = 0

3

F23 y + F43 y = 0 −0.6 F23 y − 0.13F23 x + 0.6 F43 y + 0.13F43 x = 0 F14 x + F34 x = 1 956.3 F14 y + F34 y = −415.82

( g)

(

)

−0.16 F14 y + 0.76 F14 x + 0.16 F34 y − 0.76 F34 x = −0.45 ( 415.82 ) − 0.34 1 956.3 = −852.26 F32 x + F23 x = 0 F34 x + F43 x = 0 F32 y + F23 y = 0 F34 y + F43 y = 0

9 Form the matrices for solution:

                

   0  0 0 0 0 0 0 0 0 1 0 1 0 0  0 0 0 0 0 0 1 0 0 1 0 1  0 0 0 0 0 0 0 0  −0.05 1.4 −0.08 2.2 −4.4 0  0 1 0 0 0 0 0 0 0 0 0 1 0  0 0 1 0 1 0 0 0 0 0 0 0  0 0  0 0 −0.13 −0.6 0.13 0.6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0  1  0 0 0 0 1 0 0 0 0 0 0 0  0 0.76 −0.16 −0.76 0.16   0 0 0 0 0 0 0 0 0  0 0 0 0 0 0 0 1 1 0 0 0 0  0 1 0 0 0 1 0 0 0 0 0 0  0  0 0 0 0 1 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1     

 F12 x  F12 y     0  F32 x   0    0 F32 y       0 Fh    0  F23 x   0    F23 y  =  1 956.30  (h)    F43 x   −415.82    −852.26  F43 y    0  F14 x   0    0 F14 y      0  F34 x   F34 y  

10 Table 3-3 part 2 shows the solution to this problem for the data given in Table 3-3 part 1, assuming a 2000-lb (8896-N) force applied at the crimp, normal to the crimp surface. Program Matrix (on the website) was used. The force generated in link 3 is 1547 lb (6888 N), the reaction force against link 1 by link 2 (F21) is 1561 lb (6 943 N) at 166°, the reaction force against link 1 by link 4 (F41) is 451 lb (2008 N) at 169°, and a –233.5 lb-in (–26.6-N-m) moment must be applied to the handles to generate the specified crimp force. This moment can be obtained with a 53.1-lb (236-N) force applied at mid-handle. This force is within the physiological grip-force capacity of the average human.

116

MACHINE DESIGN

C A S E

Table 3–3—part 2 Case Study 2A Calculated Data Variable

3

Fh

-

An Integrated Approach

S T U D Y

-

Sixth Edition

3 A

Automobile Scissors-Jack Loading Analysis

Value

Unit

53.1

lb

F12 x

1513.6

lb

F12 y

–381.0

lb

F32 x

–1513.6

lb

F32 y

327.9

lb

F43 x

–1513.6

lb

F43 y

327.9

lb

F23 x

1513.6

lb

F23 y

–327.9

lb

F34 x

1513.6

lb

F34 y

–327.9

lb

F14 x

442.7

lb

F14 y

–87.9

lb

F21 x

–1513.6

lb

F21 y

381.0

lb

F41 x

–442.7

lb

F41 y

87.9

lb

Problem

Determine the forces on the elements of the scissors jack in the position shown in Figure 3-30.

Given

The geometry is known and the jack supports a force of P = 1000 lb (4448 N) in the position shown.

Assumptions

The accelerations are negligible. The jack is on level ground. The angle of the elevated car chassis does not impart an overturning moment to the jack. All forces are coplanar and two-dimensional. A Class 1 load model is appropriate and a static analysis is acceptable.

Solution

See Figures 3-30 through 3-33 and Table 3-4, parts 1 and 2.

1 Figure 3-30 shows a schematic of a simple scissors jack used to raise a car. It consists of six links, which are pivoted and/or geared together and a seventh link in the form of a lead screw, which is turned to raise the jack. While this is clearly a three-dimensional device, it can be analyzed as a two-dimensional one if we assume that the applied load (from the car) and the jack are exactly vertical (in the z direction). If so, all forces will be in the xy plane. This assumption is valid if the car is jacked from a level surface. If not, then there will be some forces in the yz and xz planes as well. The jack designer needs to consider the more general case, but for our simple example we will initially assume two-dimensional loading. For the overall assembly as shown in Figure 3-30, we can solve for the reaction force Fg, given force P, by summing forces: Fg = –P. 2 Figure 3-31 shows a set of free-body diagrams for the entire jack. Each element or subassembly of interest has been separated from the others and the forces and moments shown acting on it (except for its weight, which is small compared to the applied forces and is thus neglected for this analysis). The forces and moments can be either internal reactions at interconnections with other elements or external loads from the “outside world.” The centers of gravity of the respective elements are used as the origins of the P dimensions in inches

6

3

ty p 2

4

1

y

7

2

x FIGURE 3-30 An Automobile Scissors Jack

5

6 Fg

30 ° typ

Chapter 3

KINEMATICS AND LOAD DETERMINATION

117

y P R43 Rp

R32

x

F32

y

M42

x

F12

3

y

F43

R23 2

F34

3

F23

R12

R34

x M24

4

R14

F41

F21

F14

1

F17

y

R17

x

7

y

F76 R67

y

M75

M57

x

F16

F61

F71

F67

F56 F65

x

6

R65

R56

R76 FIGURE 3-31

5

Fg

Rg

Free-Body Diagrams of the Complete Scissors Jack

local, nonrotating coordinate systems in which the points of application of all forces on the elements are located. In this design, stability is achieved by the mating of two pairs of crude (noninvolute) gear segments acting between links 2 and 4 and between links 5 and 7. These interactions are modeled as forces acting along a common normal shared by the two teeth. This common normal is perpendicular to the common tangent at the contact point. There are 3 second-law equations available for each of the seven elements, allowing 21 unknowns. An additional 10 third-law equations will be needed for a total of 31. This is a cumbersome system to solve for such a simple device, but we can use its symmetry to advantage in order to simplify the problem. 3 Figure 3-32 shows the upper half of the jack assembly. Because of the mirror symmetry between the upper and lower portions, the lower half can be removed to simplify

R16

118

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

P 3

ty p 6"

3 A

y

x

FAy

2 1

4

B

FB y

FIGURE 3-32 Free-Body Diagram of the Symmetrical Upper Half of an Automobile Scissors Jack

the analysis. The forces calculated for this half will be duplicated in the other. If we wished, we could solve for the reaction forces at A and B using equations 3.3b from this free-body diagram of the half-jack assembly. 4 Figure 3-33a shows the free-body diagrams for the upper half of the jack assembly, which are essentially the same as those of Figure 3-31. We now have four elements but can consider the subassembly labeled 1 to be the “ground,” leaving three elements on which to apply equations 3.3. Note that all unknown forces and moments are initially assumed positive in the equations.

*

Note the similarity to equations (b) in Case Study 2A. Only the subscript for the reaction moment is different because a different link is providing it. The consistent notation of this force analysis method makes it easy to write the equations for any system.

5 Link 2 has three forces acting on it: F42 is the unknown force at the gear tooth contact with link 4; F12 and F32 are the unknown reaction forces from links 1 and 3 respectively. Force F12 is provided by part 1 on part 2 at the pivot pin, and force F32 is provided by part 3 acting on part 2 at their pivot pin. The magnitudes and directions of these pin forces and the magnitude of F42 are unknown. The direction of F42 is along the common normal shown in Figure 3-33b. Write equations 3.19b for this element to sum forces in the x and y directions and sum moments about the CG (with the cross products expanded):*

∑ Fx = F12 x + F32 x + F42 x = 0 ∑ Fy = F12 y + F32 y + F42 y = 0 ∑ M z = R12 x F12 y − R12 y F12 x + R32 x F32 y − R32 y F32 x + R42 x F42 y − R42 y F42 x = 0

(a)

6 Link 3 has three forces acting on it: the applied load P, F23, and F43. Only P is known. Writing equations 3.19b for this element gives

∑ Fx = F23x + F43x + Px = 0 ∑ Fy = F23 y + F43 y + Py = 0 ∑ M z = R23x F23 y − R23 y F23x + R43x F43 y − R43 y F43x + RPx Py − RPy Px = 0

(b)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

P

common normal

y R43

Rp

R32

x

2

F12

3

F23

R12

FA x

F42

R23

F34

F24

4

F41

F21

( b) Gear tooth detail

x R24

1

4

B

R14

F14

FB x FB y

FAy (a) Free-body diagrams

FIGURE 3-33 Free-Body Diagrams of Elements of the Half Scissors Jack

7 Link 4 has three forces acting on it: F24 is the unknown force from link 2; F14 and F34 are the unknown reaction forces from links 1 and 3, respectively:

∑ Fx = F14 x + F24 x + F34 x = 0 ∑ Fy = F14 y + F24 y + F34 y = 0 ∑ M z = R14 x F14 y − R14 y F14 x + R24 x F24 y − R24 y F24 x + R34 x F34 y − R34 y F34 x = 0

(c)

8 The nine equations in sets a through c have 16 unknowns in them, F12x, F12y, F32x, F32y, F23x, F23y, F43x, F43y, F14x, F14y, F34x, F34y, F24x, F24y, F42x, and F42y . We can write the third-law relationships between action-reaction pairs at each of the joints to obtain six of the seven additional equations needed: F32 x = − F23 x

F32 y = − F23 y

F34 x = − F43 x

F34 y = − F43 y

F42 x = − F24 x

F42 y = − F24 y

(d )

9 The last equation needed comes from the relationship between the x and y components of the force F24 (or F42) at the tooth/tooth contact point. Such a contact (or half) joint can transmit force (excepting friction force) only along the common normal[4] that is perpendicular to the joint’s common tangent as shown in Figure 3-33b. The common normal is also called the axis of transmission. The tangent of the angle of this common normal relates the two components of the force at the joint: F24 y = F24 x tan θ

θ

2

y

F43

R42 A

x

R34

x

F32

y

119

(e)

3

120

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

10 Equations (a) through (e) comprise a set of 16 simultaneous equations that can be solved either by matrix reduction or by iterative root-finding methods. Putting them in standard form for a matrix solution gives:

3

Table 3–4—part 1 Case Study 3A Given Data Variable

Value

Unit

Px

0.00

lb

Py

–1000.00

lb

–0.50

in

R px R py

0.87

in

θ R 12 x

–45.00 –3.12

in

R 12 y

–1.80

in

R 32 x

2.08

in

R 32 y

1.20

in

R 42 x

2.71

in

R 42 y

1.00

in

R 23 x

–0.78

in

R 23 y

–0.78

in

R 43 x

0.78

in

R 43 y

–0.78

in

R 14 x

3.12

in

R 14 y

–1.80

in

R 24 x

–2.58

in

R 24 y

1.04

in

R 34 x

–2.08

in

R 34 y

1.20

in

deg

F12 x + F32 x + F42 x = 0 F12 y + F32 y + F42 y = 0 R12 x F12 y − R12 y F12 x + R32 x F32 y − R32 y F32 x + R42 x F42 y − R42 y F42 x = 0 F23 x + F43 x = − Px F23 y + F43 y = − Py R23 x F23 y − R23 y F23 x + R43 x F43 y − R43 y F43 x = − RPx Py + RPy Px F14 x + F24 x + F34 x = 0 F14 y + F24 y + F34 y = 0 R14 x F14 y − R14 y F14 x + R24 x F24 y − R24 y F24 x + R34 x F34 y − R34 y F34 x = 0 (f) F32 x + F23 x = 0 F32 y + F23 y = 0 F34 x + F43 x = 0 F34 y + F43 y = 0 F42 x + F24 x = 0 F42 y + F24 y = 0 F24 y − F24 x tan θ = 0

11 Substituting the given data from Table 3-4 part 1 gives: F12 x + F32 x + F42 x F12 y + F32 y + F42 y −3.12 F12 y + 1.80 F12 x + 2.08 F32 y − 1.20 F32 x + 2.71F42 y − 0.99 F42 x F23 x + F43 x F23 y + F43 y −0.78 F23 y + 0.78 F23 x + 0.78 F43 y + 0.78 F43 x F14 x + F24 x + F34 x F14 y + F24 y + F34 y 3.12 F14 y + 1.80 F14 x − 2.58 F24 y − 1.04 F24 x − 2.08 F34 y − 1.20 F34 x F32 x + F23 x F32 y + F23 y F34 x + F43 x F34 y + F43 y F42 x + F24 x F42 y + F24 y F24 y + 1.0 F24 x

=0 =0 =0 = 0.0 = 1 000 = − 500 =0 =0 =0 =0 =0 =0 =0 =0 =0 =0

( g)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

121

12 Put these equations into matrix form.

                   

1

0

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

1

0

0

0

0

0

0

0

0

0

0

1.80

3.12

1.20

2.08

1.00

2.71

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0`

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0.78

0.78

0.78

0.78

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

1

0

0

0

0

0

0

0

0

0

0

0

0

1

0

1

0

1

0

0

0

0

0

0

0

0

0

0

1.80

3.12

1.04

2.58

1.20

2.08

0

0

1

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

0

1

0

0

0

0

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

0

0

0

0

1

1

0

0

                   

                        

F12 x F12 y F32 x F32 y F42 x F42 y F23 x F23 y F43 x F43 y F14 x F14 y F24 x F24 y F34 x F34 y

                        

                   

0 0 0 0 1000 500 0 0 0 0 0 0 0 0 0 0

           (h)         

3

13 Table 3-4 part 2 shows the solution to this problem from program Matrix for the given data in Table 3-4 part 1, which assumes a vertical 1000-lb (4448-N) applied force P.

Table 3–4—part 2

14 The forces on link 1 can also be found from Newton’s third law.

Case Study 3A Calculated Data

FAx = − F21x = F12 x FAy = − F21y = F12 y FBx = − F41x = F14 x FBy = − F41y = F14 y

3.13

(i )

THREE-DIMENSIONAL, STATIC LOADING CASE STUDY

This section presents a case study involving three-dimensional static loading on a bicycle brake caliper assembly. The same techniques used for two-dimensional load analysis also work for the three-dimensional case. The third dimension requires more equations, which are available from the summation of forces in the z direction and the summation of moments about the x and y axes as defined in equations 3.17 and 3.19 for the dynamic and static cases, respectively. As an example, we will now analyze the bicycle brake arm that is actuated by the handbrake lever that was analyzed in Case Study 1A.

Variable

Value

Unit

F12x

877.8

lb

F12y

530.4

lb

F32x

–587.7

lb

F32y

–820.5

lb

F42x

–290.1

lb

F42y

290.1

lb

F23x

587.7

lb

F23y

820.5

lb

F43x

–587.7

lb

F43y

179.5

lb

F14x

–877.8

lb

F14y

469.6

lb

F24x

290.1

lb

F24y

–290.1

lb

F34x

587.7

lb

F34y

–179.5

lb

122

MACHINE DESIGN

C A S E

-

An Integrated Approach

S T U D Y

-

Sixth Edition

4 A

Bicycle Brake Arm Loading Analysis Problem

Determine the forces acting in three dimensions on the bicycle brake arm in its actuated position as shown in Figure 3-34. This brake arm has been failing in service and may need to be redesigned.

Given

The existing brake arm geometry is known and the arm is acted on by a cable force of 1046 N in the position shown. (See Case Study 1A.)

Assumptions

The accelerations are negligible. A Class 1 load model is appropriate and a static analysis is acceptable. The coefficient of friction between the brake pad and wheel rim has been measured and is 0.45 at room temperature and 0.40 at 150°F.

Solution

See Figures 3-34 and 3-35, and Table 3-5.

3

1 Figure 3-34 shows a center-pull brake arm assembly commonly used on bicycles. It consists of six elements or subassemblies, the frame and its pivot pins (1), the two brake arms (2 and 4), the cable spreader assembly (3), the brake pads (5), and the wheel rim (6). This is clearly a three-dimensional device and must be analyzed as such. 2 The cable is the same one that is attached to the brake lever in Figure 3-26. The 267-N (60-lb) hand force is multiplied by the mechanical advantage of the hand lever and transmitted via this cable to the pair of brake arms as was calculated in Case Study 1A. We will assume no loss of force in the cable guides, thus the full 1046-N (235-lb) cable force is available at this end. 3 The direction of the normal force between the brake pad and the wheel rim is shown in Figure 3-34 to be at q = 172° with respect to the positive x axis, and the friction force is directed along the z axis. (See Figures 3-34 and 3-35 for the xyz axis orientations.) 4 Figure 3-35 shows free-body diagrams of the arm, frame, and cable spreader assembly. We are principally interested in the forces acting on the brake arm. However, we first need to analyze the effect of the cable spreader geometry on the force applied to the arm at A. This analysis can be two dimensional if we ignore the small z-offset between the two arms for simplicity. A more accurate analysis would require that the z-directed components of the cable-spreader forces acting on the arms be included. Note that the cable subassembly (3) is a concurrent force system. Writing equations 3.19b in two dimensions for this subassembly, and noting the symmetry about point A, we can write from inspection of the FBD:

∑ Fx = F23x + F43x = 0 ∑ Fy = F23 y + F43 y + Fcable = 0

(a)

This equation set can be easily solved to yield Fcable 1046 =− = −523 N 2 2 F23 y −523 = = = −353 N tan ( 56° ) 1.483 = − F23 x = 353 N

F23 y = F43 y = − F23 x F43 x

(b)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

123

Fcabl e y

y

cable

cable

3

3

3 x

z

brake arm

brake arm

2 4

4

2 frame

pad

1

pad wheel rim 6

5

θ 172 ° x N

5 wheel rim

FIGURE 3-34

6

V

ω

Center-Pull Bicycle Brake Arm Assembly

Newton’s third law relates these forces to their reactions on the brake arm at point A: F32 x = − F23 x = 353 N F32 y = − F23 y = 523 N

(c)

F32 z = 0

5 We can now write equations 3.19a for the arm (link 2). For the forces:

∑ Fx = F12 x + F32 x + F52 x = 0; ∑ Fy = F12 y + F32 y + F52 y = 0; ∑ Fz = F12z + F32z + F52z = 0;

F12 x + F52 x = −353 F12 y + F52 y = −523

(d )

F12 z + F52 z = 0

For the moments:

∑ M x = M12 x + ( R12 y F12z − R12z F12 y ) + ( R32 y F32z − R32z F32 y ) + ( R52 y F52 z − R52 z F52 y ) = 0 ∑ M y = M12 y + ( R12z F12 x − R12 x F12z ) + ( R32z F32 x − R32 x F32z ) (

)

+ R52 z F52 x − R52 x F52 z = 0

∑ M z = ( R12 x F12 y − R12 y F12 x ) + ( R32 x F32 y − R32 y F32 x ) + ( R52 x F52 y − R52 y F52 x ) = 0

(e)

124

MACHINE DESIGN

-

An Integrated Approach

-

Fcabl e

frame

Sixth Edition

Fcabl e

1 cable

3

F21 z

F21 x

A

3

3

M21 y M12 y

F23 x

F32 y A

cable

F32 x

2

56 °

56 °

F23 y

F43 y

F43 x

R32

R32

2

x

R12 F12 y

brake arm B R52 F52 x

C

frame

z F12 z

F43 y

y

F32 z

y

F23 y

F32 y

F21 y

R12 R52

F12 x

M12 x

M21 x

F52 z

F52 y

F21 z 1

F52 y

FIGURE 3-35 Brake Arm Free-Body Diagrams

Note that all unknown forces and moments are initially assumed positive in the equations, regardless of their apparent directions on the FBDs. The moments M12x and M12y are due to the fact that there is a moment joint between the arm (2) and the pivot pin (1) about the x and y axes. We assume negligible friction about the z axis, thus allowing M12z to be zero. 6 The joint between the brake pad (5) and the wheel rim (6) transmits a force normal to the plane of contact. The friction force magnitude, Ff, in the contact plane is related to the normal force by the Coulomb friction equation, Ff = µN

(f)

where m is the coefficient of friction and N is the normal force. The velocity of the point on the rim below the center of the brake pad is in the z direction. The force components F52x and F52y are due entirely to the normal force being transmitted through the pad to the arm and are therefore related by Newton’s third law. F52 x = − N x = − N cos θ = − N cos172° = 0.990 N F52 y = − N y = − N sin θ = − N sin172° = −0.139 N

( g)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

125

The direction of the friction force Ff must always oppose motion and thus it acts in the negative z direction on the wheel rim. Its reaction force on the arm has the opposite sense. F52 z = − Ff

(h)

7 We now have 10 equations (in the sets labeled d, e, f, g, and h) containing 10 unknowns: F12x, F12y, F12z, F52x, F52y, F52z, M12x, M12y, N, and Ff. Forces F32x, F32y, and F32z are known from equations c. These 10 equations can be solved simultaneously either by matrix reduction or by iterative root-finding methods. First arrange the equations with all unknowns on the left and all known or assumed values on the right.

3

F12 x + F52 x = −353 F12 y + F52 y = −523 M12 x + R12 y F12 z − R12 z F12 y

Table 3–5—part 1

F12 z + F52 z = 0 + R52 y F52 z − R52 z F52 y = R32 z F32 y − R32 y F32 z

M12 y + R12 z F12 x − R12 x F12 z + R52 z F52 x − R52 x F52 z = R32 x F32 z − R32 z F32 x

(i )

R12 x F12 y − R12 y F12 x + R52 x F52 y − R52 y F52 x = R32 y F32 x − R32 x F32 y Ff − µN = 0 F52 x + N cos θ = 0 F52 y + N sin θ = 0 F52 z + Ff = 0

8 Substitute the known and assumed values from Table 3-5 part 1: F12 x + F52 x = −353 F12 y + F52 y = −523 F12 z + F52 z = 0 M12 x − 27.2 F12 z − 23.1F12 y − 69.7 F52 z − 0 F52 y = 0 ( 523) − 38.7 ( 0 ) = 0 M12 y + 23.1F12 x − 5.2 F12 z + 0 F52 x + 13F52 z = −75.4 ( 0 ) − 0 ( 353) = 0 5.2 F12 y + 27.2 F12 x − 13F52 y + 69.7 F52 x = 38.7 ( 353) + 75.4 ( 523) = 53 095 Ff − 0.4 N = 0 F52 x − 0.990 N = 0 F52 y + 0.139 N = 0 F52 z + Ff = 0

9 Form the matrices for solution.

( j)

Case Study 4A Given and Assumed Data Variable

Value

Unit

µ

0.4

none

θ

172.0

deg

R12x

5.2

mm

R12y

–27.2

mm

R12z

23.1

mm

R32x

–75.4

mm

R32y

38.7

mm

R32z

0.0

mm

R52x

–13.0

mm

R52y

–69.7

mm

R52z F32x

0.0

mm

353.0

N

F32y

523.0

N

F32z

0.0

N

M12z

0.0

N–m

126

MACHINE DESIGN

Table 3–5—part 2 Case Study 4A Calculated Data

3

Variable

Value

Unit

F12 x

–1805

N

F12 y

–319

N

F12 z

587

N

F52 x

1452

N

F52 y

–204

N

–587

N

F52 z M12 x

32 304

N–mm

M12 y

52 370

N–mm

N

1467

N

Ff

587

N

-

An Integrated Approach

 1 0 0 0 1 0  0 1 0 0 0 1  0 1 1 0 0  0  0 −23.1 −27.2 0 0 −69.7  23.1 −5.2 13 0 0 0  0 69.7 −13 0  27.2 5.2  0 0 0 0 0 0  0 0 0 0 1 0  1 0 0 0 0  0 0 0 0 1 0  0

0 0 0 1 0 0 0 0 0 0

0 0 0 0 1 0 0 0 0 0

0 0 0 0 0 0 1 0 0 1

-

Sixth Edition

   0   0   0     0   0 × 0   −0.4   −0.990    0.139     0  

F12 x   F12 y   −353  −523 F12 z     0 F52 x   0  F52 y   0 = F52 z   53 095   0 M12 x   0   M12 y   0 Ff   0   N 

      ( k )       

10 Table 3-5 part 2 shows the solution to this problem from program Matrix for the given data in Table 3-5 part 1. This problem can be solved using any one of several commercial equation-solver programs such as Mathcad, MATLAB, Maple, Mathematica or with the program Matrix provided with the text.

3.14

DYNAMIC LOADING CASE STUDY

This section presents a case study involving two-dimensional dynamic loading on a fourbar linkage designed as a dynamic-load demonstration device. A photograph of this machine is shown in Figure 3-11 and a video is on the website (Fourbar Linkage). This machine can be analyzed in two dimensions since all elements move in parallel planes. The presence of significant accelerations on the moving elements in a system requires that a dynamic analysis be done with equations 3.1. The approach is identical to that used in the preceding static load analyses except for the need to include the mA and Ia terms in the equations.

C A S E † http://www.designofmachinery.com/MD/Fourbar_Machine. mp4

S T U D Y

5 A

Fourbar Linkage Loading Analysis View the Video

(35:38) †

Problem

Determine the theoretical rigid-body forces acting in two dimensions on the fourbar linkage shown in Figure 3-36.

Given

The linkage geometry, masses, and mass moments of inertia are known and the linkage is driven at up to 120 rpm by a speed-controlled electric motor.

Assumptions

The accelerations are significant. A Class 4 load model is appropriate and a dynamic analysis is required. There are no external loads on the system; all loads are due to the accelerations of the links. The weight forces are insignificant compared to the inertial forces and will be neglected. The links are assumed to be ideal rigid bodies. Friction and the effects of clearances in the pin joints also will be ignored.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

127

coupler rocker

crank

3

flywheel

counter weights gear motor

FIGURE 3-36

Copyright © 2018 Robert L. Norton: All Rights Reserved

Fourbar Linkage Dynamic Model (Photo by the author)

Solution

See Figures 3-36 through 3-38 and Table 3-6.

1 Figures 3-36 and 3-37 show the fourbar linkage demonstrator model. It consists of three moving elements (links 2, 3, and 4) plus the frame or ground link (1). The motor drives link 2 through a gearbox. The two fixed pivots are instrumented with piezoelectric force transducers to measure the dynamic forces acting in x and y directions on the ground plane. A pair of accelerometers is mounted to a point on the floating coupler (link 3) to measure its accelerations. 2 Figure 3-37 shows a schematic of the linkage. The links are designed with lightening holes to reduce their masses and mass moments of inertia. The input to link 2 can be an angular acceleration, a constant angular velocity, or an applied torque. Link 2 rotates fully about its fixed pivot at O2. Even though link 2 may have a zero angular acceleration a2, if run at constant angular velocity w2, there will still be time-varying angular accelerations on links 3 and 4 since they oscillate back and forth. In any case, the CGs of the links will experience time-varying linear accelerations as the linkage moves. These angular and linear accelerations will generate inertia forces and torques

3

O 2 -O 4 = 18 in (457.2 mm) O 2 -A = 6 in (152.4 mm) A-B = 16 in (406.4 mm)

B

O4 2 -B = 12 in (304.8 mm)

m 4 , IG 4

m 3 , IG 3

m 2 , IG 2 2 A

ω2, α2, T 2

4

O2 1

O4 1

FIGURE 3-37 Fourbar Linkage Schematic and Basic Dimensions (See Table 3-6 for more information)

128

MACHINE DESIGN

Case Study 5A Given and Assumed Data

3

Value

An Integrated Approach

-

Sixth Edition

as defined by Newton’s second law. Thus, even with no external forces or torques applied to the links, the inertial forces will create reaction forces at the pins. It is these forces that we wish to calculate.

Table 3–6—part 1

Variable

-

3 Figure 3-38 shows the free-body diagrams of the individual links. The local, nonrotating, coordinate system for each link is set up at its CG. The kinematic equations of motion must be solved to determine the linear accelerations of the CG of each link and the link’s angular acceleration for every position of interest during the cycle. (See equations 3.7) These accelerations, AGn and an, are shown acting on each of the n links. The forces at each pin connection are shown as xy pairs, numbered as before, and are initially assumed to be positive.

Unit

θ2

30.00

deg

ω2

120.00

rpm

mass2

0.525

kg

mass3

1.050

kg

mass4

1.050

kg

Icg2

0.057

kg-m2

Icg3

0.011

kg-m2

Icg4

0.455

kg-m2

R12x

–46.9

R12y

–71.3

mm

R32x

85.1

mm

R32y

4.9

mm

R23x

–150.7

mm

R23y

–177.6

mm

R43x

185.5

mm

R43y

50.8

mm

R14x

–21.5

mm

R14y

–100.6

mm

R34x

–10.6

mm

R34y

204.0

mm

4 Equations 3.1 can be written for each moving link in the system. The masses and the mass moments of inertia of each link about its CG must be calculated for use in these equations. In this case study, a solid modeling CAD system was used to design the links’ geometries and to calculate their mass properties. 5 For link 2:

mm

∑ Fx = F12 x + F32 x = m2 aG 2 ∑ Fy = F12 y + F32 y = m2 aG 2 ∑ M z = T2 + ( R12 x F12 y − R12 y F12 x ) + ( R32 x F32 y − R32 y F32 x ) = IG 2α 2 x

(a)

y

6 For link 3:

∑ Fx = F23x + F43x = m3aG3 ∑ Fy = F23 y + F43 y = m3aG3 ∑ M z = ( R23x F23 y − R23 y F23x ) + ( R43x F43 y − R43 y F43x ) = IG 3α3 x

(b)

y

y α3 AG 3 F43

R43

y

R32

x

F32

F34

ω2, α2, T 2

x AG 2 F12

4 3

R12

2

α4

R34

F23

R14 AG 4

F21

FIGURE 3-38

y

R23

1

Free-Body Diagrams of Elements in a Fourbar Linkage

F41 1

F14

x

Chapter 3

KINEMATICS AND LOAD DETERMINATION

7 For link 4:

Table 3–6—part 2

∑ ∑ Fy = F14 y + F34 y = m4 aG 4 ∑ M z = ( R14 x F14 y − R14 y F14 x ) + ( R34 x F34 y − R34 y F34 x ) = IG 4 α 4

Case Study 5A Calculated Data

Fx = F14 x + F34 x = m4 aG 4 x

(c)

y

8 There are 13 unknowns in these nine equations: F12x, F12y, F32x, F32y, F23x, F23y, F43x, F43y, F14x, F14y, F34x, F34y, and T2. Four third-law equations can be written to equate the action-reaction pairs at the joints. F32 x = − F23 x F32 y = − F23 y F34 x = − F43 x F34 y = − F43 y

(d )

9 The set of thirteen equations in a through d can be solved simultaneously to determine the forces and driving torque either by matrix reduction or by iterative root-finding methods. This case study was solved by both techniques and files for both are on the website. Note that the masses and mass moments of inertia of the links are constant with time and position, but the accelerations are time-varying. Thus, a complete analysis requires that equations a–d be solved for all positions or time steps of interest. The models use lists or arrays to store the calculated values from equations a–d for 13 values of the input angle q2 of the driving link ( 0 to 360° by 30° increments). The model also calculates the kinematic accelerations of the links and their CGs, which are needed for the force calculations. The largest and smallest forces present on each link during the cycle can then be determined for use in later stress and deflection analyses. The given data and results of this force analysis for one crank position (q2 = 30°) are shown in Table 3-6, parts 1 and 2. Plots of the forces at the fixed pivots for one complete revolution of the crank are shown in Figure 3-39.

x and y forces on joint 12

0

force N

force N

200

y

–200

Variable

Value

Unit

F12x

–255.8

N

F12y

–178.1

N

F32x

252.0

N

F32y

172.2

N

F34x

–215.6

N

F34y

–163.9

N

F14x

201.0

N

F14y

167.0

N

F43x

215.6

N

F43y

163.9

N

F23x

–252.0

N

F23y

–172.2

N

T12

–3.55

α3

56.7

rad/sec2

α4

138.0

rad/sec2

Acg2x

–7.4

rad/sec2

Acg2y

–11.3

rad/sec2

Acg3x

–34.6

rad/sec2

Acg3y

–7.9

rad/sec2

Acg4x

–13.9

rad/sec2

Acg4y

2.9

rad/sec2

N-m

x and y forces on joint 14

100

–100

129

x

100 y 0

x –100

–300 0

60

120

180

240

300

360

crank angle - degrees FIGURE 3-39 Calculated Rigid-Body Dynamic Forces in the Fourbar Linkage of Case Study 5A

0

60

120

180

240

crank angle - degrees

300

360

3

130

MACHINE DESIGN

3.15

3

-

An Integrated Approach

-

Sixth Edition

VIBRATION LOADING

In systems that are dynamically loaded, there will usually be vibration loads superimposed on the theoretical loads predicted by the dynamic equations. These vibration loads can be due to a variety of causes. If the elements in the system were infinitely stiff, then vibrations would be eliminated. But all real elements, of any material, have elasticity and thus act as springs when subjected to forces. The resulting deflections can cause additional forces to be generated from the inertial forces associated with the vibratory movements of elements or, if clearances allow contact of mating parts, to generate impact (shock) loads (see below) during their vibrations. A complete discussion of vibration phenomena is beyond the scope of this text and will not be attempted here. References are provided in the bibliography at the end of this chapter for further study. The topic is introduced here mainly to alert the machine designer to the need to consider vibration as a source of loading. Often the only way to get an accurate measure of the effects of vibration on a system is to do testing of prototypes or production systems under service conditions. The discussion of safety factors in Section 1.7 mentioned that many industries (automotive, aircraft, etc.) engage in extensive test programs to develop realistic loading models of their equipment. This topic will be discussed further in Section 6.4 when fatigue loading is introduced. Modern finite element (FEA) and boundary element (BEA) analysis techniques also allow vibration effects on a system or structure to be modeled and calculated. It is still difficult to obtain a computer model of a complex system that is as accurate as a real, instrumented prototype. This is especially true when clearances (gaps) between moving parts allow impacts to occur in the joints when loads reverse. Impacts create nonlinearities that are very difficult to model mathematically. Natural Frequency When designing machinery, it is desirable to determine the natural frequencies of the assembly or subassemblies in order to predict and avoid resonance problems in operation. Any real system can have an infinite number of natural frequencies at which it will readily vibrate. The number of natural frequencies that are necessary or desirable to calculate will vary with the situation. The most complete approach to the task is to use Finite Element Analysis (FEA) to break the assembly into a large number of discrete elements. See Chapter 8 for more information on FEA. The stresses, deflections, and number of natural frequencies that can be calculated by this technique are mainly limited by time and the computer resources available. If not using FEA, we would like to determine, at a minimum, the system’s lowest, or fundamental, natural frequency, since this frequency will usually create the largest magnitude of vibrations. The undamped fundamental natural frequency wn, with units of rad/sec, or fn, with units of Hz, can be computed from the expressions: k m 1 ωn fn = 2π

ωn =

(3.20)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

131

where wn is the fundamental natural frequency, m is the moving mass of the system in true mass units (e.g., kg, g, blob, or slug, not lbm), and k is the effective spring constant of the system. (The period of the natural frequency is its reciprocal in seconds, Tn = 1 / fn.) Equation 3.20 is based on a single-degree-of-freedom, lumped model of the system. Figure 3-40 shows such a model of a simple cam-follower system consisting of a cam, a sliding follower, and a return spring. The simplest lumped model consists of a mass connected to ground through a single spring and a single damper. All the moving mass in the system (follower, spring) is contained in m and all the “spring” including the physical spring and the springiness of all other parts is lumped in the effective spring constant k.

3

spriNg coNstaNt A spring constant k is an assumed linear relationship between the force, F, applied to an element and its resulting deflection d (see Figure 3-42): k=

F δ

(3.21a)

If an expression for the deflection of an element can be found or derived, it will provide this spring-constant relationship. This topic is revisited in the next chapter. In the example of Figure 3-40, the spring deflection d is equal to the displacement y of the mass. k=

F y

(3.21b)

dampiNg All the damping, or frictional, losses are lumped in the damping coefficient d. For this simple model, damping is assumed to be inversely proportional to the velocity ydot of the mass. d=

F y

(3.22)

followe r spring

Fspring Fdamper spring k

roller

damper d

. ..

y,y,y

cam

m

m

. ..

mass

y,y,y

ω Fcam

( a) Actual system

FIGURE 3-40 Lumped Model of a Cam-Follower Dynamic System

( b ) Lumped model

Fcam

( c ) Free-body diagram

132

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Equation 3.20 simplifies this model even further by assuming the damping d to be zero. If damping is included, the expressions for the fundamental, damped natural frequency wd with units of radians/sec, or fd, with units of Hz, become k  d 2 −  m  2m  1 fd = ωd 2π

ωd =

3

(3.23)

This damped frequency wd will be slightly lower than the undamped frequency wn. eFFective vaLues Determining the effective mass for a lumped model is straightforward and requires only summing all the values of the connected, moving masses in appropriate mass units. Determining the values of the effective spring constant and the effective damping coefficient is more complicated and will not be addressed here. See reference 2 for an explanation. resoNaNce A condition called resonance can be experienced if the operating or forcing frequency applied to the system is the same as any one of its natural frequencies. That is, if the input angular velocity applied to a rotating system is the same as, or close to, wn, the vibratory response will be very large. This can create large forces and cause failure. Thus, it is necessary to avoid operation at or near the natural frequencies if possible. Dynamic Forces If we write equation 3.17 for the simple, one-DOF model of the dynamic system in Figure 3-40 and substitute equations 3.21 and 3.22, we get

∑ Fy = ma = my

Fcam − Fspring − Fdamper = my Fcam

(3.24)

= my + dy + ky

If the kinematic parameters of displacement, velocity, and acceleration are known for the system, this equation can be solved directly for the force on the cam as a function of time. If the cam force is known and the kinematic parameters are desired, then the well-known solution to this linear, constant-coefficient differential equation can be applied. See reference 3 for a detailed derivation of that solution. Though the coordinate system used for a dynamic analysis can be arbitrarily chosen, it is important to note that both the kinematic parameters (displacement, velocity, and acceleration) and the forces in equation 3.24 must be defined in the same coordinate system. As an example of the effect of vibration on the dynamic forces of a system, we now revisit the fourbar linkage of Case Study 5A and see the results of actual measurements of dynamic forces under operating conditions.

Chapter 3

C A S E

S T U D Y

KINEMATICS AND LOAD DETERMINATION

133

5 B

Fourbar Linkage Dynamic Loading Measurement Problem

Determine the actual forces acting on the fixed pivots of the fourbar linkage in Figure 3-36 during one revolution of the input crank.

Given

The linkage is driven at 60 rpm by a speed-controlled electric motor, and force transducers are placed between the fixed pivot bearings and the ground plane.

Assumptions

There are no applied external loads on the system; all loads are due to the accelerations of the links. The weight forces are insignificant compared to the inertial forces and will be neglected. The force transducers measure only dynamic forces.

Solution

See Figures 3-36, 3-37, and 3-41.

3

1 Figure 3-36 shows the fourbar linkage.* It consists of three moving elements (links 2, 3, and 4) plus the frame or ground link (1). The shock-mounted motor drives link 2 through a gearbox and shaft coupling. The two fixed pivots are instrumented with piezoelectric force transducers to measure the dynamic forces acting in x and y directions on the ground plane. 2 Figure 3-41 shows an actual force and torque measured while the linkage was running at 60 rpm and compares them to the theoretical force and torque predicted by equations a–d in Case Study 5A.[5] Only the x component of force at the pivot between link 2 and the ground and the torque on link 2 are shown as examples. The other pin forces and components show similar deviations from their predicted theoretical values. Some of these deviations are due to variations in instantaneous angular velocity of the drive motor; others are due to backlash in the gearbox. The theoretical analysis assumes constant input shaft velocity. Vibrations and impacts account for other deviations.

This example of deviations from theoretical forces in a very simple dynamic system is presented to point out that our best calculations of forces (and thus the resulting stresses) on a system may be in error due to factors not included in a simplified force analysis. It is common for the theoretical predictions of forces in a dynamic system to understate reality, which is, of course, a nonconservative result. Wherever feasible, the testing of physical prototypes will give the most accurate and realistic results. The effects of vibration on a system can cause significant loadings and are difficult to predict without test data of the sort shown in Figure 3-41, where the actual loads are seen to be double their predicted values; this will obviously double the stresses. A traditional, and somewhat crude, approach used in machine design has been to apply overload factors to the theoretical calculated loads based on experience with the same or similar equipment. As an example, see Table 12-17 in the chapter on spur-gear design. This table lists industry-recommended overload factors for gears subjected to various types of shock loading. These sorts of factors should be used only if one cannot develop more accurate test data of the type shown in Figure 3-41.

*

A video of this linkage is in the Demos folder on the website as Fourbar Linkage.

134

MACHINE DESIGN

-

An Integrated Approach

(a) Theoretical and actual dynamic force in x direction at crank pivot

10

Sixth Edition

actual

5 0 lbf

3

-

–5 –10

theoretical

–15 –20 (b) Theoretical and actual dynamic torque at crank pivot

400 300

actual theoretical

lbf – in

200 100 0

–100 –200 –300

FIGURE 3-41

Copyright © 2018 Robert L. Norton: All Rights Reserved

Theoretical and Measured Dynamic Forces and Torques in a Fourbar Linkage

3.16

IMPACT LOADING

The loading considered so far has either been static or, if time-varying, has been assumed to be gradually and smoothly applied, with all mating parts continually in contact. Many machines have elements that are subjected to sudden loads or impacts. One example is the crank-slider mechanism, which forms the heart of an automobile engine. The piston head is subjected to an explosive rise in pressure every two crank revolutions when the cylinder fires, and the clearance between the circumference of the piston and the cylinder wall can allow an impact of these surfaces as the load is reversed each cycle. A more extreme example is a jackhammer, whose purpose is to impact pavement and break it up. The loads that result from impact can be much greater than those that would result from the same elements contacting gradually. Imagine trying to drive a nail by gently placing the hammer head on the nail rather than by striking it. What distinguishes impact loading from static loading is the time duration of the application of the load. If the load is applied slowly, it is considered static; if applied rapidly, then it is impact. One criterion used to distinguish the two is to compare the time of load application tl (defined as the time it takes the load to rise from zero to its peak value) to the period of the natural frequency Tn of the system. If tl is less than half Tn, it is considered to be impact. If tl is greater than three times Tn, it is considered static. Between those limits is a gray area in which either condition can exist. Two general cases of impact loading are considered to exist, though we will see that one is just a limiting case of the other. Burr[6] calls these two cases striking impact

Chapter 3

KINEMATICS AND LOAD DETERMINATION

135

and force impact. Striking impact refers to an actual collision of two bodies, such as in hammering or the taking up of clearance between mating parts. Force impact refers to a suddenly applied load with no velocity of collision, as in a weight suddenly being taken up by a support. This condition is common in friction clutches and brakes (see Chapter 17). These cases can occur independently or in combination. 3

Severe collisions between moving objects can result in permanent deformation of the colliding bodies as in an automobile accident. In such cases the permanent deformation is desirable in order to absorb the large amount of energy of the collision and protect the occupants from more severe harm. We are concerned here only with impacts that do not cause permanent deformation; that is, the stresses will remain in the elastic region. This is necessary to allow continued use of the component after impact. If the mass of the striking object m is large compared to that of the struck object mb and if the striking object can be considered rigid, then the kinetic energy possessed by the striking object can be equated to the energy stored elastically in the struck object at its maximum deflection. This energy approach gives an approximate value for the impact loading. It is not exact because it assumes that the stresses throughout the impacted member reach peak values at the same time. However, waves of stress are set up in the struck body, travel through it at the speed of sound, and reflect from the boundaries. Calculating the effects of these longitudinal waves on the stresses in elastic media gives exact results and is necessary when the ratio of mass of the striking object to that of the struck object is small. The wave method will not be discussed here. The reader is directed to reference 6 for further information. Energy Method The kinetic energy of the striking body will be converted to stored potential energy in the struck body, assuming that no energy is lost to heat. If we assume that all particles of the combined bodies come to rest at the same instant, then just before rebound, the force, stress, and deflection in the struck body will be maximal. The elastic energy stored in the struck body will be equal to the area under the force-deflection curve defined by its particular spring constant. A generalized force-deflection curve for a linear spring element is shown in Figure 3-42. The elastic energy stored is the area under the curve between zero and any combination of force and deflection. Because of the linear relationship, this is the area of a triangle, A = 1/2bh. Thus, the energy stored at the point of peak impact deflection, di, is E=

1 Fi δ i 2

(3.25a)

E=

Fi 2 2k

(3.25b)

F k Fi

Substituting equation 3.21 gives

horizoNtaL impact Figure 3-43a shows a mass about to impact the end of a horizontal rod. This device is sometimes called a slide hammer and is used to remove dents from automobile sheet metal among other uses. At the point of impact, the portion of the kinetic energy of the moving mass that is imparted to the struck mass is

δi FIGURE 3-42 Energy Stored in a Spring

δ

136

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

δi

l

3

m k m k

vi (a) Horizontal

h

( b) Vertical

δi

FIGURE 3-43 Axial Impact on a Slender Rod

1  E = η  mvi 2  2 

(3.26)

where m is its mass and vi its velocity at impact. We need to modify the kinetic energy term by a correction factor η to account for the energy dissipation associated with the particular type of elastic member being struck. If the dissipation is negligible, η will be 1. Assuming that all the kinetic energy transferred from the moving mass is converted to elastic energy stored in the struck member allows us to equate equations 3.25 and 3.26: Fi 2 mv 2 =η i 2k 2 Fi = vi ηmk

(3.27)

If the mass were allowed to statically load the struck member, the resulting static deflection would be dst = W / k where W = mg. Substituting these into equation 3.27 gives a ratio of dynamic force to static force or dynamic deflection to static deflection: Fi δ η = i = vi W δ st gδ st

(3.28)

The term on the right side of this equation is called the impact factor, which provides a ratio of impact to static force or deflection. Thus, if the static deflection can be calculated for the application of a force equal to the weight of the mass, an estimate of the dynamic force and dynamic deflection can be obtained. Note that equations 3.27 and 3.28 are valid for any case of horizontal impact, whether the object is loaded axially as shown here, in bending, or in torsion. Methods for calculating the deflections of various cases are addressed in the next chapter. The spring rate k for any object can be found by rearranging its deflection equation according to equation 3.21.

Chapter 3

KINEMATICS AND LOAD DETERMINATION

137

verticaL impact For the case of a mass falling through a distance h onto a rod as shown in Figure 3-43b, equation 3.27 also applies with the impact velocity vi2 = 2gh. The potential energy for a drop through distance h is: E= η

mvi 2 = ηmgh = Wηh 2

(3.29a)

3

If the deflection at impact is small compared to the drop distance h, this equation is sufficient. But, if the deflection is significant compared to h, the energy of impact needs to include an amount due to the deflection through which the weight falls beyond h. The total potential energy given up by the mass on impact is then

(

E = Wηh + W δ i = W ηh +δ i

)

(3.29b)

Equate this potential energy to the elastic energy stored in the struck member and substitute equation 3-25b and the expression W = kdst Fi 2 = W ηh +δ i 2k W Fi 2 = 2 kW ηh +δ i = 2 W ηh +δ i δ st

(

(

)

)

(

)

2 2 ηh δ i 2 ηh  Fi  F + = + 2 i    = W δ st δ st δ st W 2  Fi   Fi  2 ηh =0   − 2   − δ st W W

(3.30 a)

giving a quadratic equation in Fi / W whose solution is: Fi δ 2 ηh = i = 1+ 1+ W δ st δ st

(3.30 b)

The right side expression is the impact ratio for the falling weight case. Equation 3-30b can be used for any impact case involving a falling weight. For example, if the weight is dropped on a beam, the static deflection of the beam at the point of impact is used. If the distance h to which the mass is raised is set to zero, equation 3.30b becomes equal to 2. This says that if the mass is held in contact with the “struck” member (with the weight of the mass separately supported) and then allowed to suddenly impart its weight to that member, the dynamic force will be twice the weight. This is the case of “force impact” described earlier, in which there is no actual collision between the objects. A more accurate analysis, using wave methods, predicts that the dynamic force will be more than doubled even in this noncollision case of sudden application of load.[6] Many designers use 3 or 4 as a more conservative estimate of this dynamic factor for the case of sudden load application. This is only a crude estimate, however, and, if possible, experimental measurements or a wave-method analysis should be made to determine more suitable dynamic factors for any particular design. Burr derives the correction factors η for several impact cases in reference 6. Roark and Young provide factors for additional cases in reference 7. For the case of a mass axially impacting a rod as shown in Figure 3-44, the correction factor is[6]

η 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0

5

1 0 15 20 m mass ratio mb

FIGURE 3-44 Correction Factor η as a Function of Mass Ratio

138

MACHINE DESIGN

-

An Integrated Approach

η=

-

Sixth Edition

1 m 1+ b 3m

(3.31)

where m is the mass of the striking object, and mb is the mass of the struck object. As the ratio of the striking mass to the struck mass increases, the correction factor η asymptotically approaches one. Figure 3-44 shows η as a function of mass ratio for three cases: an axial rod (equation 3.31) plotted in color, a simply supported beam struck at midspan (black-solid), and a cantilever beam struck at the free end (black-dotted).[6] The correction factor η is always less than one (and > 0.9 for mass ratios > 5) , so assuming it to be one is conservative. However, be aware that this energy method gives approximate and nonconservative results in general, and it needs to be used with larger-than-usual safety factors applied.

3

force ratio 7000 6000

EXAMPLE 3-3

5000 4000

Impact Loading on an Axial Rod

3000

Problem

The axial rod shown in Figure 3-43a is hit by a mass moving at 1 m/sec. a. Determine the sensitivity of the impact force to the length/diameter ratio of the rod for a constant 1-kg moving mass. b. Determine the sensitivity of the impact force to the ratio of moving mass to rod mass for a constant length/diameter ratio of 10.

2000 1000 0 0

10

20

Given

The round rod is 100 mm long. The rod and moving mass are steel with E = 207 Gpa and a mass density of 7.86 g/cm3.

Assumptions

An approximate energy method will be acceptable. The correction factor for energy dissipation will be applied.

Solution

See Figures 3-43a, 3-45, and 3-46.

l/d ratio force in Newtons 70 000

1 Figure 3-43a shows the system. The moving mass strikes the flange on the end of the rod with the stated velocity of 1 m/sec.

60 000 50 000

2 For part (a), we will keep the moving mass constant at 1 kg, and the rod length constant at 100 mm and vary the rod diameter to obtain l / d ratios in the range of 1 to 20. In the equations that follow we will show only one calculation made for the l / d ratio of 10. We can also compute all the values for a list of l / d ratios. The static deflection that would result from application of the weight force of the mass is calculated from the expression for the deflection of a bar in tension. (See equation 4.8 in the next chapter for the derivation.)

40 000 30 000 20 000 10 000 0 0

10

20

δ st =

Wl AE

=

l/d ratio

(

9.81 N 100 mm 78.54 mm

2

)

(2.07 E5 N mm 2 )

= 0.06 µm

(a)

The correction factor η is calculated here for an assumed mass ratio of 16.2, FIGURE 3-45 Dynamic Force and Force Ratio as a Function of l/d Ratio for the System in Example 3-3

η=

1 1 = = 0.98 mb 0.0617 1+ 1+ 3 (1) 3m

(b)

Chapter 3

KINEMATICS AND LOAD DETERMINATION

These values (calculated for each different rod diameter d) are then substituted into equation 3.12 to find the force ratio Fi / W and the dynamic force Fi. For d = 10 mm Fi η m = vi =1 W gδ st s

(

)

0.98 = 1 285.9 m 9.81 2 0.00006 m s

(

Fi = 1 285.9 9.81N = 12 612 N

)

(c)

The variation in force ratio with changes in l / d ratio for a constant amount of moving mass and a constant impact velocity (i.e., constant input energy) is shown in Figure 3-45. As the l / d ratio is reduced, the rod becomes much stiffer and generates much larger dynamic forces from the same impact energy. This clearly shows that impact forces can be reduced by increasing the compliance of the impacted system. 3 For part (b), we will keep the l / d ratio constant at 10 and vary the ratio between the moving mass and the rod mass over the range of 1 to 20. The files EX03-01B on the website calculate all values for a range of mass ratios. The results for a mass ratio of 16.2 are the same as in part (a) above. Figure 3-46a shows that the dynamic force ratio Fi / W varies inversely with the mass ratio. However, the value of the dynamic force is increasing with mass ratio as shown in Figure 3-46b, because the static force W is also increasing with mass ratio.

139

force ratio 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 0

3

0

10

20

mass ratio force in ne wtons 16 000 14 000 12 000

3.17

BEAM LOADING

A beam is any element that carries loads transverse to its long axis and may carry loads in the axial direction as well. A beam supported on pins or narrow supports at each end is said to be simply supported, as shown in Figure 3-47a. A beam fixed at one end and unsupported at the other is a cantilever beam (Figure 3-47b). A simply supported beam that overhangs its supports at either end is an overhung beam (Figure 3-47c). If a beam has more supports than are necessary to provide kinematic stability (i.e., make the kinematic degree of freedom zero), then the beam is said to be overconstrained or indeterminate, as shown in Figure 3-47d. An indeterminate beam problem cannot be solved for its loads using only equations 3.19. Other techniques are necessary. This problem is addressed in the next chapter. Beams are typically analyzed as static devices, though vibrations and accelerations can cause dynamic loading. A beam may carry loads in three dimensions, in which case equations 3.19a apply. For the two-dimensional case, equations 3.19b suffice. The review examples used here are limited to 2-D cases for brevity. Shear and Moment A beam may be loaded with some combination of distributed and/or concentrated forces or moments as in Figure 3-47. The applied forces will create both shearing forces and bending moments in the beam. A load analysis must find the magnitudes and spatial distributions of these shear forces and bending moments on the beam. The shear forces V and the moment M in a beam are related to the loading function q(x) by

10 000 8 000 6 000 4 000 2 000 0 0

10

20

mass ratio FIGURE 3-46 Dynamic Force and Force Ratio as a Function of Mass Ratio for the System in Example 3-3

140

MACHINE DESIGN

-

An Integrated Approach

l

y a

-

Sixth Edition

l

y a

w 〈x–a 〉0

3

〈x–a 〉–1

M1

x R1

F

x

R2

R1

( a) Simply supported beam with uniformly distributed loading

(b) Cantilever beam with concentrated loading

y

y

l

l

w 〈x–a〉1

a

b a

Mz 〈x–0〉–2

w 〈x–a 〉0

x

x

R2

R1

R1

( c) Overhung beam with moment and linearly distributed loading

R2

R3

( d) Statically indeterminate beam with uniformly distributed loading

FIGURE 3-47 Types of Beams and Beam Loadings

q(x) =

dV d 2 M = dx dx 2

(3.32a)

The loading function q(x) is typically known and the shear V and moment M distributions can be found by integrating equation 3.16a: VB

∫V V

V

dV =

A

xB

∫x

A

q dx = VB − VA

(3.32b)

Equation 3.16b shows that the difference in the shear forces between any two points, A and B, is equal to the area under the graph of the loading function, equation 3.32a. Integrating the relationship between shear and moment gives

positive shear

M

M

positive moment FIGURE 3-48 Beam Sign Convention

MB

∫M

A

dM =

xB

∫x

A

V dx = M B − M A

(3.32c)

showing that the difference in the moment between any two points, A and B, is equal to the area under the graph of the shear function, equation 3.32b. sigN coNveNtioN The usual (and arbitrary) sign convention used for beams is to consider a moment positive if it causes the beam to deflect concave downward (as if to collect water). This puts the top surface in compression and the bottom surface in tension. The shear force is considered positive if it causes a clockwise rotation of the section on which it acts. These conventions are shown in Figure 3-48 and result in positive moments being

Chapter 3

KINEMATICS AND LOAD DETERMINATION

141

created by negative applied loads. All the applied loads shown in Figure 3-47 are negative. For example, in Figure 3-47a, the distributed load magnitude from a to l is q = –w. equatioN soLutioN The solution of equations 3.19 and 3.32 for any beam problem may be carried out by one of several approaches. Sequential and graphical solutions are described in many textbooks on statics and mechanics of materials. One classical approach to these problems is to find the reactions on the beam using equations 3.19 and then draw the shear and moment diagrams using a graphical integration approach combined with calculations for the significant values of the functions. This approach has value from a pedagogical standpoint as it is easily followed, but it is cumbersome to implement. The approach most amenable to computer solution uses a class of mathematical functions called singularity functions to represent the loads on the beam. We present the classical approach as a pedagogical reference and also introduce the use of singularity functions that offer some computational advantages. While this approach may be new to some students, when compared to the methods usually learned in other courses, it has significant advantages in computerizing the solution. Singularity Functions Because the loads on beams typically consist of collections of discrete entities, such as point loads or segments of distributed loads that can be discontinuous over the beam length, it is difficult to represent these discrete functions with equations that are valid over the entire continuum of beam length. A special class of functions called singularity functions was invented to deal with these mathematical situations. Singularity functions are often denoted by a binomial in angled brackets as shown in equations 3.33. The first quantity in the brackets is the variable of interest, in our case x, the distance along the beam length. The second quantity a is a user-defined parameter that denotes where in x the singularity function either acts or begins to act. For example, for a point load, the quantity a represents the particular value of x at which the load acts (see Figure 3-47b). The definition of this singularity function, called the unit impulse or Dirac delta function, is given in equation 3.33d. Note that all singularity functions involve a conditional constraint. The unit impulse evaluates to ∞ if x = a and is 0 at any other value of x. The unit step function or Heaviside step function (Eq. 3.33c) evaluates to 0 for all values of x less than a and to 1 for all other x. Since these functions are defined to evaluate to unity, multiplying them by a coefficient creates any magnitude desired. Their application is shown in the following three examples and is explained in the most detail in Example 3-4B. If a loading function both starts and stops within the range of x desired, it needs two singularity functions to describe it. The first defines the value of a1 at which the function begins to act and has a positive or negative coefficient as appropriate to its direction. The second defines the value a2 at which the function ceases to act and has a coefficient of the same magnitude but opposite sign as the first. These two functions will cancel beyond a2, making the load zero. Such a case is shown in Example 4-6 in the next chapter. Quadratically distributed loads can be represented by a unit parabolic function, x−a

2

which is defined as 0 when x ≤ a, and equal to (x – a)2 when x > a.

(3.33a)

3

142

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Linearly distributed loads can be represented by a unit ramp function, x−a

1

(3.33b)

which is defined as 0 when x ≤ a, and equal to (x – a) when x > a.

3

A uniformly distributed load over a portion of a beam can be represented mathematically by a unit step function, x−a

0

(3.33c)

which is defined as 0 when x  a, and is undefined at x = a. A concentrated force can be represented by the unit impulse function, x−a

−1

(3.33d )

which is defined as 0 when x  a. Its integral evaluates to unity at a. A concentrated moment can be represented by the unit doublet function, x−a

−2

(3.33e)

which is defined as 0 when x  a. It generates a unit couple moment at a. This process can be extended to obtain polynomial singularity functions of any order n to fit distributed loads of any shape. Four of the five singularity functions described here are shown in Figure 3-47, as applied to various beam types. A computer program is needed to evaluate these functions. Table 3-7 shows five of the singularity functions implemented in a “Basic-like” pseudocode. The For loops run the variable x from zero to the beam length l. The If test determines whether x has reached the value of a, which is the location of the start of the singularity function. Depending on this test, the value of y(x) is set either to zero or to the specified magnitude of the singularity function. This type of code can easily be implemented in any computer language (e.g., C+, Fortran, basic) or in an equation solver (e.g., Mathcad, MATLAB, TK Solver, EES). The integrals of these singularity functions have special definitions that, in some cases, defy common sense but nevertheless provide the desired mathematical results. For example, the unit impulse function (Eq. 3.17d) is defined in the limit as having zero width and infinite magnitude, yet its area (integral) is defined as equal to one (Eq. 3.18d). (See reference 8 for a more complete discussion of singularity functions.) The integrals of the singularity functions in equations 3.17 are defined as dλ =

x−a 3

3

dλ =

x−a 2

2

dλ = x − a

1

x

2

x

1

x

0

∫−∞ λ − a ∫−∞ λ − a ∫−∞ λ − a

(3.34 a)

(3.34 b)

(3.34c)

Chapter 3

Table 3-7

KINEMATICS AND LOAD DETERMINATION

143

Pseudocode to Evaluate Singularity Functions

Pulse Singularity Function For x = 0 to l * Note: This routine does not generate the infinite value of the Dirac delta function. Rather, it generates the magnitude of a point load applied at location a for use in plotting a beam’s loading function.

If ABS (x – a) < 0.0001 Then y(x) = magnitude, Else y(x) = 0* Next x Step Singularity Function For x = 0 to l If x < a Then y(x) = 0, Else y(x) = magnitude Next x Ramp Singularity Function For x = 0 to l If x < = a Then y(x) = 0, Else y(x) = magnitude * (x – a) Next x Parabolic Singularity Function For x = 0 to l If x < = a Then y(x) = 0, Else y(x) = magnitude * (x – a)^2 Next x Cubic Singularity Function For x = 0 to l If x < = a Then y(x) = 0, Else y(x) = magnitude * (x – a)^3 Next x

x

−1

dλ = x − a

0

(3.34 d )

x

−2

dλ = x − a

−1

(3.34e)

∫−∞ λ − a ∫−∞ λ − a

where λ is just an integration variable running from –∞ to x. These expressions can be used to evaluate the shear and moment functions that result from any loading function that is expressed as a combination of singularity functions.

EXAMPLE 3-4A Shear and Moment Diagrams of a Simply Supported Beam Using a Graphical Method Problem

Determine and plot the shear and moment functions for the simply supported beam with uniformly distributed load shown in Figure 3-47a.

Given

Beam length l = 10 in, and load location a = 4 in. The magnitude of the uniform force distribution is w = 10 lb/in.

Assumptions

The weight of the beam is negligible compared to the applied load and so can be ignored.

3

144

MACHINE DESIGN

y

Solution

l a

An Integrated Approach

-

Sixth Edition

See Figures 3-47a and 3-49.

1 Solve for the reaction forces using equations 3.19. Summing moments about the right end and summing forces in the y direction gives

〈x–a 〉 0 w x

3 R1

-

R2

∑ M z = 0 = R1l −

w ( l − a )2 2

w ( l − a )2 10 (10 − 4 )2 = = 18 R1 = 2l 2 (10 )

Simply supported beam with uniformly distributed loading

FIGURE 3-47a

∑ Fy = 0 = R1 − w (l − a ) + R2

Repeated

R2 = w ( l − a ) − R1 = 10 (10 − 4 ) − 18 = 42 (a) Loading Diagram 40 q R2 30 20 R 1 10 0 w –10 x –20 0 5 10 (b) Shear Diagram 20 V 10 0 –10 –20 –30 –40 –50 0

–w 5.8 5

x 10

(c) Moment Diagram 100 M 88. 2 80

0 0

5

FIGURE 3-49 Example 3-4 Plots

3 If your reflexes are quick enough, you should try to catch the shear diagram (Figure 3-49b) as you fall, climb onto it, and repeat this backward-walking trick across it to create the moment diagram, which is the integral of the shear diagram. Note in Figure 3-49c that from x = 0 to x = a this moment function is a straight line with slope = R1. Beyond point a, the shear diagram is triangular, and so integrates to a parabola. The peak moment will occur where the shear diagram crosses zero (i.e., zero slope on the moment diagram). The value of x at V = 0 can be found with a little trigonometry, noting that the slope of the triangle is –w: R1 18 =4+ = 5.8 w 10

(c)

Positive shear area adds to the moment value and negative area subtracts. So the value of the peak moment can be found by adding the areas of the rectangular and triangular portions of the shear diagram from x = 0 to the point of zero shear at x = 5.8:

40 5.8

(b)

2 The shape of the shear diagram can be sketched by graphically integrating the loading diagram shown in Figure 3-49a. As a “device” to visualize this graphical integration process, imagine that you walk backward across the loading diagram of the beam, starting from the left end and taking small steps of length dx. You will record on the shear diagram (Figure 3-49b) the area (force · dx) of the loading diagram that you can see as you take each step. As you take the first step backward from x = 0, the shear diagram rises immediately to the value of R1. As you walk from x = 0 to x = a, no change occurs, since you see no additional forces. As you step beyond x = a, you begin to see strips of area equal to –w · dx, which subtract from the value of R1 on the shear diagram. When you reach x = l, the total area w · (l – a) will have taken the value of the shear diagram to –R2. As you step backward off the beam’s loading diagram (Figure 3-49a) and plummet downward, you can now see the reaction force R2, which closes the shear diagram to zero. The largest value of the shear force in this case is then R2 at x = l.

x@V = 0 = a +

60 20

(a)

x

M@ x = 5.8 = R1 ( a ) + R1

1.8 1.8 = 18 ( 4 ) + 18 = 88.2 2 2

(d )

10

The above method gives the magnitudes and locations of the maximum shear and moment on the beam and is useful for a quick determination of those values. However,

Chapter 3

KINEMATICS AND LOAD DETERMINATION

145

all that walking and falling can become tiresome, and it would be useful to have a method that can be conveniently computerized to give accurate and complete information on the shear and moment diagrams of any beam-loading case. Such a method will also allow us to obtain the beam’s deflection curve with little additional work. The simple method shown previously is not as useful for determining deflection curves, as will be seen in the next chapter. We will now repeat this example using singularity functions to determine the loading, shear, and moment diagrams.

3

EXAMPLE 3-4B Shear and Moment Diagrams of a Simply Supported Beam Using Singularity Functions Problem

Determine and plot the shear and moment functions for the simply supported beam with uniformly distributed load shown in Figure 3-47a.

Given

Beam length l = 10 in, and load location a = 4 in. The magnitude of the uniform force distribution is w = 10 lb/in.

Assumptions

The weight of the beam is negligible compared to the applied load and so can be ignored.

Solution

See Figures 3-47a and 3-49.



−1

−w x−a

V = qdx = R1 x − 0



0

0

+ R2 x − l

−1

− w x − a 1 + R2 x − l

M = Vdx = R1 x − 0 1 −

w x−a 2

l a

〈x–a 〉 0 w x

1 Write equations for the load function in terms of equations 3.33 and integrate the resulting function twice using equations 3.34 to obtain the shear and moment functions. For the beam in Figure 3-47a, q = R1 x − 0

y

2

(a) 0

+ C1

+ R2 x − l 1 + C1 x + C2

(b) (c)

There are two reaction forces and two constants of integration to be found. We are integrating along a hypothetical infinite beam from –∞ to x. The variable x can take on values both before and beyond the end of the beam. If we consider the conditions at a point infinitesimally to the left of x = 0 (denoted as x = 0–), the shear and moment will both be zero there. The same conditions apply at a point infinitesimally to the right of x = l (denoted as x = l +). These observations provide the four boundary conditions needed to evaluate the four constants C1, C2, R1, R2: when x = 0–, V = 0, M = 0; when x = l+, V = 0, M = 0. 2 The constants C1 and C2 are found by substituting the boundary conditions x = 0–, V = 0, and x = 0–, M = 0 in equations (b) and (c), respectively:

R1

R2

Simply supported beam with uniformly distributed loading

FIGURE 3-47a Repeated

146

MACHINE DESIGN

-

An Integrated Approach

( )

V 0 − = 0 = R1 0 − − 0

0

-

Sixth Edition

− w 0− − a

1

+ R2 0 − − l

0

w − 0 −a 2

2

+ R2 0 − − l

1

+ C1

C1 = 0

( )

M 0 − = 0 = R1 0 − − 0

3

1



C2 = 0

( )

(d )

+ C1 0 − + C2

Note that in general, the constants C1 and C2 will always be zero if the reaction forces and moments acting on the beam are included in the loading function, because the shear and moment diagrams must close to zero at each end of the beam. 3 The reaction forces R1 and R2 can be calculated from equations (c) and (b) respectively by substituting the boundary conditions x = l+, V = 0, M = 0. Note that we can substitute l for l+ to evaluate it since their difference is vanishingly small. (a) Loading Diagram 40 q R2 30 20 R 1 10 0 w –10 x –20 0 5 10 (b) Shear Diagram 20 V 10 0 –10 –20 –30 –40 –50 0

l −0 +

0 = R1l − R1 =

( )

(

1

+

2l +

−w

(

w l+ − a

w l+ − a

)

V l + = R1 l + − 0

l+ − a

)

2

+ R2 l + − l

2

1

=0

2

(e)

2 2

=

0

w ( l − a )2 10 (10 − 4 )2 = = 18 2l 2 (10 )

− w l+ − a

1

+ R2 l + − l

0 = R1 − w ( l − a ) + R2

0

=0 (f)

R2 = w ( l − a ) − R1 = 10 (10 − 4 ) − 18 = 42

Since w, l, and a are known from the given data, equation (e) can be solved for R1, and this result substituted in equation (f) to find R2. Note that equation (f) is just ΣF = 0, and equation (e) is the sum of moments taken about point l and set to 0.

–w 5.8 5

x 10

(c) Moment Diagram 100 M 88. 2 80 60 40 20

5.8

0 0

( ) = R1

M l

+

5

x 10

F I G U R E 3 - 4 9 Repeated Example 3-4 Plots

4 To generate the shear and moment functions over the length of the beam, equations (b) and (c) must be evaluated for a range of values of x from 0 to l, after substituting the above values of C1, C2, R1, and R2 in them. The independent variable x was run from 0 to l = 10 at 0.1 increments. The reactions, loading function, shear-force function, and moment function were calculated from equations (a) through (f) above and are plotted in Figure 3-49. The files EX03-04 that generate these plots are on the website. 5 The largest absolute values of the shear and moment functions are of interest for the calculation of stresses in the beam. The plots show that the shear force is largest at x = l and the moment has a maximum Mmax near the center. The value of x at Mmax can be found by setting V to 0 in equation (b) and solving for x. (The shear function is the derivative of the moment function and so must be zero at each of its minima and maxima.) This gives x = 5.8 at Mmax. The function values at these points of maxima or minima can then be calculated from equations b and c respectively by substituting the appropriate values of x and evaluating the singularity functions. For the maximum absolute value of shear force at x = l,

Chapter 3

KINEMATICS AND LOAD DETERMINATION

Vmax = V@ x = l − = R1 l − − 0

0

(

− w l− − a

1

+ R2 l − − l

147

0

)

= R1 − w l − − a + 0

( g)

= 18 − 10 (10 − 4 ) + 0 = −42

3

Note that the first singularity term evaluates to 1 since l –  > 0 (see Eq. 3.33c), the second singularity term evaluates to (l – a) because l –  > a in this problem (see Eq. 3.33b), and the third singularity term evaluates to 0 as defined in equation 3.33c. The maximum moment is found in similar fashion: M max = M@ x = 5.8 = R1 5.8 − 0 1 − w = R1 5.8 1 − w = 18 ( 5.8 ) − 10

5.8 − a 2

5.8 − 4 2

2

(5.8 − 4 )2 2

2

+ R2 5.8 − l

1

+ R2 5.8 − 10

1

(h)

+ 0 = 88.2

The third singularity term evaluates to 0 because 5.8 10, making it other than a short column. Calculate the slenderness ratio of the tangent point between the Johnson and Euler lines of Figure 4-42:

( Sr )D =

π

2 ( 30 E 6 ) 2E =π = 84.5 83E 3 Sy

(c)

The slenderness ratio of this column is less than that of the tangent point between the Johnson and Euler lines shown in Figure 4-42. It is thus an intermediate-column and the Johnson-column formula (Eq. 4.43) should be used to find the critical load: 2  1  S y Sr    Pcr = A  S y −  E  2 π    

(d )

2  1  83 000 ( 34 )    = 4765 lb = 0.125 (.5) 83 000 −   30 E 6  2π    

Fh 1 2

Fh

force from hand

3

4

Fc crimp force

F I G U R E 3 - 2 8 Repeated Wire Connector Crimping Tool

243

* Even a small amount of clearance in the holes will prevent the pins from acting as a moment joint along their axes, thus creating an effective pinned-pinned connection in two dimensions unless and until the clearance is taken up by deflection and a pin is jammed in its hole.

MACHINE DESIGN

244

-

An Integrated Approach -

Sixth Edition

curved beam approximation

A Fh

1560

52

2000

1

y

1.23

1548

0.6 r P

B

x

3 Fh

0.7

D

452

1548

C

52

1560

1.0 A

B

1548

0.75

C

0.35

1548 2

0.80

All dimensions in inches and forces in pounds. See Table 3-3 for complete force information.

1.55

a

2000 1.2

452 D

4

FIGURE 4-49 Free-Body Diagrams, Dimensions, and Force Magnitudes for a Wire Connector Crimping Tool

The critical load is 3.1 times larger than the applied load. It is safe against buckling. Link 2 is a shorter, wider column than link 3 and has lower axial forces so can be assumed to be safe against buckling based on the link 3 calculations. 2 Since it does not buckle, the deflection of link 3 in axial compression is (Eq. 4.7): x=

1548 (1.23) Pl = = 0.001 in AE 0.0625 ( 30 E 6 )

(e)

3 Any of the links could also fail in bearing in the 0.25-in-dia holes. The largest force on any pin is 1560 lb. This worst-case bearing stress (Eqs. 4.7 and 4.10) is then σb =

P P 1560 = = = 49 920 psi Abearing length ( dia ) 0.125 ( 0.25)

(f)

There is no danger of tearout failure in links 2 or 3, since the loading is toward the center of the part. Link 1 has ample material around the holes to prevent tearout. 4 The 0.25-in-dia pins are in single shear. The worst-case direct shear stress from equation 4.9 is: τ=

P 1560 = = 31 780 psi Ashear π ( 0.25)2 4

( g)

Chapter 4

STRESS, STRAIN, AND DEFLECTION

5 Link 4 is a 1.55-in-long beam, simply supported at the pins and loaded with the 2 000-lb crimp force at 0.35 in from point C. Write the equations for the load, shear, moment, slope, and deflection using singularity functions, noting that the integration constants C1 and C2 will be zero: q = R1 x − 0 V = R1 x − 0

−1 0

−F x−a

−F x−a

1

0

−1

+ R2 x − l 0

+ R2 x − l

1

M = R1 x − 0 − F x − a + R2 x − l

−1

(h)

1

1  R1 x−0  EI  2

2



F x−a 2

2

+

R2 x−l 2

2

 + C3  

1  R1 y= x−0  EI  6

3

F − x−a 6

3

R + 2 x−l 6

3

 + C3 x + C4  

θ=

(i )

6 The reaction forces can be found from ΣM = 0 and ΣF = 0 (see Appendix B): R1 =

F ( l − a ) 2000 (1.55 − 0.35) = = 1548 lb 1.55 l

( j)

R2 =

Fa 2000 ( 0.35) = = 452 lb 1.55 l

(k )

The maximum moment is 1548(0.35) = 541.8 lb-in at the applied load. The shear and moment diagrams for link 4 are shown in Figure 4-50. 7 The beam depth at the point of maximum moment is 0.75 in and the thickness is 0.187. The bending stress is then 0.75  541.8   2  Mc σ= = = 30 905 psi I 0.187 ( 0.75)3 12

(l )

8 The beam slope and deflection functions require calculation of the integration constants C3 and C4, which are found by substituting the boundary conditions x = 0, y = 0 and x = l, y = 0 in the deflection equation. 1  R1 0−0  EI  6 C4 = 0

3

1  R1 l−0  EI  6

3

0=

0=



F 0−a 6

3

+

R2 0−l 6

3

 + C3 ( 0 ) + C4   (m )



F l−a 6

3

+

R2 l−l 6

3

 + C3 ( l ) 

1 1  C3 =  F ( l − a )3 − R1l 3  = 2000 (1.55 − 0.35)3 − 1548 (1.55)3  = −248.4    6l 6 (1.55) 

( n)

9 The deflection equation is found by combining equations i, j, k, m, and n: y=

F 6lEI

{( )(

)

l − a x 3 +  ( l − a )2 − l 2  x − l x − a  

3

+a x−l

3

}

(o)

245

MACHINE DESIGN

246

Loading Diagram (lb) 1500

An Integrated Approach -

Sixth Edition

and the maximum deflection at x = 0.68 in is ymax =

0

-

Fa ( l − a ) 2 a + ( l − a )2 − l 2 6l EI

(

)

2000 ( 0.35)(1.55 − 0.35)  = 0.352 + (1.55 − 0.35)2 − 1.552  = 0.0005 in  6 (1.55)( 30 E 6 )( 0.0066 ) 

–1000

( p)

–2000 0

0.8

1.6

Shear Diagram (lb) 1500 1000 500 0 –500 0

0.8

1.6

Moment Diagram (lb-in) 600 400

Only a very small deflection is allowed here to guarantee the proper crimp stroke, and this amount is acceptable. The slope and deflection diagrams are shown in Figure 4-50. Also see the file CASE2B-1. 10 Link 1 is relatively massive compared to the others and the only area of concern is the jaw, which is loaded by the 2000-lb crimp force and has a hole in the cross section at its root. While the shape of this element is not exactly that of a curved beam with concentric inside and outside radii, this assumption will be acceptably conservative if we use an outer radius equal to the smallest section dimension as shown in Figure 4-49. This makes its inside radius 0.6 in and its approximate outside radius 1.6 in. The eccentricity e of the curved beam’s neutral axis versus the beam’s centroidal axis rc is found from equation 4.12a,while accounting for the section’s hole in the integration: e = rc −

200 0 0

0.8

1.6

Slope Diagram (rad) 0.0010 0.0005 0 –0.0005 –0.0010 –0.0015 0 0.8 1.6 Deflection Diagram (in) 0 –0.0001 –0.0002 –0.0003 –0.0004 –0.0005 0 0.8 1.6 FIGURE 4-50 Link 4 Plots—Case 2B

0.313 (1 − 0.25) A = 0.103 = 1.1 − 1.600 dr  dA  0.975 dr + 0.313   0 r  0.600 r 1.225 r 



ro





(q)

The radius to the neutral axis (rn) and the distances (ci and co) from the inner and outer fiber radii (ri and ro) to the neutral axis are then (see Figure 4-16) rn = rc − e = 1.10 − 0.103 = 0.997 ci = rn − ri = 0.997 − 0.600 = 0.397

(r )

co = ro − rn = 1.600 − 0.997 = 0.603

11 The applied bending moment on the curved beam section is taken as the applied load times its distance to the beam’s centroidal axis: M = Fl = 2000 ( 0.7 − 0.6 + 1.1) = 2400 lb-in

(s )

12 Find stresses at the inner and outer fibers from equations 4.12b and 4.12c. Reduce the beam cross sectional area by the hole area: σi = +

M  ci  2400  0.397  =   = 65 kpsi   eA  ri  0.103 (1.0 − 0.25)( 0.313)  0.60 

(t )

M  co  2400  0.603  =− σo = −   = −37 kpsi eA  ro  0.103 (1.0 − 0.25)( 0.313)   1.60 

13 There is also a direct axial tensile stress, which adds to the bending stress in the inner fiber at point P: σa =

F 2000 = = 8.5 kpsi A (1.0 − 0.25)( 0.313)

σ max = σ a + σ i = 65 + 8.5 = 74 kpsi

(u)

Chapter 4

STRESS, STRAIN, AND DEFLECTION

(a)

(b) FIGURE 4-51

Copyright © 2018 Robert L. Norton: All Rights Reserved

Finite Element Analysis of the Stresses in the Crimping Tool of Case Study 2B

This is the principal stress for point P, since there is no applied shear or other normal stress at this edge point. The maximum shear stress at point P is half this principal stress or 37 kpsi. The bending stress at the outer fiber is compressive and thus subtracts from the axial tensile stress for a net of –37 + 8.5 = –28.5 kpsi. 14 There is significant stress concentration at the hole. The theoretical stress-concentration factor for the case of a circular hole in an infinite plate is Kt = 3 as defined in equation 4.32a and Figure 4-35. For a circular hole in a finite plate, Kt is a function of the ratio of the hole diameter to the plate width. Peterson gives a chart of stress-concentration factors for a round hole in a flat plate under tension[5] from which we find that Kt  = 2.42 for a dia / width ratio = 1/4. The local axial tensile stress at the hole is then 2.42(8.5) = 20.5 kpsi, which is less than the tensile stress at the inner fiber. 15 While this is far from a complete stress and deflection analysis of these parts, the calculations done address the areas judged to be most likely to fail or to have problem deflections. The stresses and deflections in link 1 were also computed using the ABAQUS finite element analysis program, which gave an estimated maximum principal stress at point P of 81 kpsi compared to our estimate of 74 kpsi. The FEA mesh and stress distribution calculated by the FEA model is shown in Figure 4-51. Case Study 2D in Chapter 8 presents the complete FEA analysis of this assembly. Our analysis simplified the part geometry in order to allow the use of a known closedform model (the curved beam) whereas the FEA model included all the material in the actual part but discretized its geometry. Both analyses should be recognized as only estimates of the stress states in the parts, not exact solutions. 16 Redesign may be needed to reduce these stresses and deflections, based on a failure analysis. This case study will be revisited in the next chapter after various failure theories are presented. You may examine the models for this case study by opening the files CASE2B-1, CASE2B-2, and CASE2B-3 in the program of your choice. A stress analysis of this case study is done in Chapter 8 using Finite Element Analysis (FEA).

247

248

MACHINE DESIGN

C A S E

-

An Integrated Approach -

S T U D Y

Sixth Edition

3 B

Automobile Scissors-Jack Stress and Deflection Analysis Problem

Determine the stresses and deflections at critical points in the scissorsjack assembly shown in Figures 3-5 (repeated here) and 4-52.

Given

The geometry and loading are known from Case Study 3A. The design load is 2000 lb total or 1000 lb per side. The width of the links is 1.032 in and their thickness is 0.15 in. The screw is a 1/2-13 UNC thread with root dia = 0.406 in. The material of all parts is ductile steel with E = 30E6 psi and Sy = 60 000 psi.

Assumptions

The most likely failure points are the links as columns, the holes in bearing where the pins insert, the connecting pins in shear, the gear teeth in bending, and the screw in tension. There are two sets of links, one set on each side. Assume the two sides share the load equally. The jack is typically used for very few cycles over its lifetime, so a static analysis is appropriate.

Solution

See Figures 4-52 to 4-53 and the files CASE3B-1 and CASE3B-2.

1 The forces on this jack assembly for the position shown were calculated in the previous installment of this case study (3A) in Chapter 3. Please see that section and Table 3-4 for additional force data. 2 The force in the jack screw is four times the 878-lb component F21x at point A because that force is from the upper half of the jack in one plane only. The lower half exerts an equal force on the screw, and the back side doubles their sum. These forces put the screw in axial tension. The tensile stress is found from equation 4.7 using the 0.406- in root diameter of the thread to calculate the cross-sectional area. This is a conservative assumption, as we shall see when analyzing threaded fasteners in Chapter 15: σx =

4 (878 ) 3512 P = = = 27 128 psi A π ( 0.406 )2 0.129 4

(a)

The axial deflection of the screw is found from equation 4.8: x=

Pl 4 (878 )(12.55) = = 0.011 in 0.129 ( 30 E 6 ) AE

(b)

3 Link 2 is the most heavily loaded of the links due to the applied load P being slightly offset to the left of center, so we will calculate its stresses and deflections. This link is loaded as a beam-column with both an axial compressive force P between points C and D and a bending couple applied between D and E. Note that the force F12 is virtually colinear with the link axis. The axial load is equal to F12 cos(1°) = 1026 lb and the bending couple created by F42 acting about point D is M = 412(0.9) = 371 in-lb. This couple is equivalent to the axial load being eccentric at point D by distance e = M / P  = 371 / 1026 = 0.36 in. The secant-column formula (Eq. 4.46c) can be used with this effective eccentricity e accounting for the applied couple in the plane of bending; c is 1/2 of the 1.032-in width of the link. Since it is a pinned-pinned column, leff = l from Table 4-4. The radius of gyration k is taken in the xy plane of bending for this calculation (Eq. 4.34):

Chapter 4

STRESS, STRAIN, AND DEFLECTION

249

P dimensions in inches

6

3

ty p

2

4

1

y

2

7

x

5

30 ° typ

6 Fg

F I G U R E 3 - 3 0 Repeated An Automobile Scissors Jack

k=

0.15 (1.032 )3 I bh3 = = = 0.298 12bh 12 ( 0.15)(1.032 ) A

(c)

The slenderness ratio is leff  / k = 20.13. The secant formula can now be applied and iterated for the value of P (see Figure 4-53):

common normal

P = 1000 per side

y

x x

F32 = 1009 D

6"

3 F23

E

4

F43

(b) Gear tooth detail

F34 = 613 2 C

0.9

F42 = 412

4

30°

F41 = 996

F21 = 1026

F14 = 996

1

F12 = 1026 @ 31° FAx = –F21x = 878

A

FAy = –F21y = 530 FIGURE 4-52

θ

2

F24 = 412

B (a) Free-body diagrams

Free-Body Diagrams, Dimensions, and Forces for Elements of the Scissors Jack

FBx = –F41x = –878

FBy = –F41y = 470

MACHINE DESIGN

250

-

An Integrated Approach -

P = A

S yc  leff  ec  1 +  2  sec  k   k

(

P  4 EA 

Sixth Edition

= 18 975 psi (d )

)

Pcrit = 0.155 18 975 = 2937 lb

The column also has to be checked for concentric-column buckling in the weaker (z) direction with c = 0.15 / 2. The radius of gyration in the z direction is found from k=

1.032 ( 0.15)3 I bh3 = = = 0.043 12bh 12 (1.032 )( 0.15) A

(e)

The slenderness ratio in the z direction is Sr =

leff k

=

6 = 138.6 0.043

(f)

This needs to be compared to the slenderness ratio (Sr)D at the tangency between the Euler and Johnson lines to determine which buckling equation to use for this column:

( Sr )D = π

2 ( 30 E 6 ) 2E =π = 99.3 60 000 Sy

( g)

The Sr for this column is greater than (Sr)D, making it an Euler column (see Figure 4-53). The critical Euler load is then found from equation 4.38a: Pcr =

π 2 EI l

=

2

π 2 ( 30 E 6 )(1.032 )( 0.15)3 12 ( 6 )2

= 2387 lb

(h)

Thus, it is more likely to buckle in the weaker z direction than in the plane of the applied moment. Its safety factor against buckling is 2.3. unit load psi x 103 70 short 60 50 Euler 40 Johnson 30 fails 20 10 secant 0 20 0 50 100 150 200 slenderness ratio FIGURE 4-53 Solution to Eccentric Column in Case Study 3B

4 The pins are all 0.437-in dia. The bearing stress in the most heavily loaded hole at C is σ bearing =

P Abearing

=

1026 = 15 652 psi 0.15 ( 0.437 )

(i)

The pins are in single shear and their worst shear stress is τ=

1026 P = = 6841 psi Ashear π ( 0.437 )2 4

( j)

5 The gear tooth on link 2 is subjected to a force of 412 lb applied at a point 0.22 in from the root of the cantilevered tooth. The tooth is 0.44 in deep at the root and 0.15 in thick. The bending moment is 412(0.22) = 91 in-lb and the bending stress at the root is σ=

91( 0.22 ) Mc = = 18 727 psi I 0.15 ( 0.44 )3 12

(k )

Chapter 4

STRESS, STRAIN, AND DEFLECTION

6 This analysis could be continued, looking at other points in the assembly and, more importantly, at stresses when the jack is in different positions. We have used an arbitrary position for this case study but, as the jack moves to the lowered position, the link and pin forces will increase due to poorer transmission angles. A complete stress analysis should be done for multiple positions. This case study will be revisited in the next chapter for the purpose of failure analysis. You may examine the models for this case study by opening the files CASE3B-1 and CASE3B-2 in the program of your choice.

C A S E

S T U D Y

4 B

Bicycle Brake Arm Stress Analysis Problem

Determine the stresses at critical points in the bicycle brake arm shown in Figures 3-9 (repeated here) and 4-54.

Given

The geometry and loading are known from Case Study 4A and are shown in Table 3-5. The cast-aluminum arm is a tee-section curved beam whose dimensions are shown in Figure 4-54. The pivot pin is ductile steel. The loading is three-dimensional.

Assumptions

The most likely failure points are the arm as a double-cantilever beam (one end of which is curved), the hole in bearing, and the connecting pin in bending as a cantilever beam. Since this is a marginally ductile cast material (5% elongation to fracture), we can ignore the stress concentration on the basis that local yielding will relieve it.

Solution

See Figures 4-54 to 4-56.

1 Each brake arm is a double-cantilever beam. Each end of an arm can be treated separately. The curved-beam portion has a tee-shaped cross section as shown in Section X-X of Figure 4-54. The neutral axis of a curved beam shifts from the centroidal axis toward the center of curvature a distance e as described in Section 4-9 and in equation 4.12a. To find e requires an integration of the beam cross section and knowledge of its centroidal radius. Figure 4-55 shows the tee-section broken into two rectangular segments, flange and web. The radius of the centroid of the tee is found by summing moments of area for each segment about the center of curvature:

∑ M = A1rc

1

rct = rct =

+ A2 rc2 = At rct

A1rc1 + A2 rc2 At

=

A1 ( ri + y1 ) + A2 ( ri + y2 ) A1 + A2

( a)

( 20 )( 7.5)(58 + 3.75) + (10 )( 7.5)(58 + 11.25) = 64.25 mm ( 20 )( 7.5) + (10 )( 7.5)

See Figures 4-53 and 4-54 for dimensions and variable names. The integral dA/r for equation 4.12a can be found in this case by adding the integrals for web and flange: ro

∫0

dA A1 A2 ( 20 )( 7.5) (10 )( 7.5) = + = + = 3.51 mm r rc1 rc2 58 + 3.75 58 + 11.25

(b)

251

252

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Fcabl e y

y

cable

cable

3

3 x

z

brake arm

brake arm

2 4

4

2 frame

pad

1

pad

θ

wheel rim 6

172 ° x N

5

5 wheel rim

F I G U R E 3 - 3 4 Repeated

6

V

ω

Center-Pull Bicycle Brake Arm Assembly

The radius of the neutral axis and the distance e are then rn =

At 225 = = 64.06 mm dA 3.51 0 r



ro

e = rc − rn = 64.25 − 64.06 = 0.187 mm

(c)

The magnitude of the bending moment acting on the curved section of the beam at Section X-X can be approximated by taking the cross product of the force F32 and its position vector RAB referenced to the pivot at B in Figure 4-54: M AB = RABx F32 y − RAB y F32 x = −80.6 ( 523) − 66 ( 353) = 65 452 N-mm

(d )

The stresses at the inner and outer fibers can now be found using equations 4.12b and 4.12c (with lengths in mm and moments in N-mm for proper unit balance): σi = +

65 452 ( 6.063) M  ci  = = 162 MPa   eA  ri  ( 0.1873)( 225)( 58 )

65 452 (8.937 ) M  co  = = −190 MPa σo = − eA  ro  ( 0.1873)( 225)( 73)

(e)

2 The hub cross section, shown as Section B-B in Figure 4-54, is a possible location of failure, since there is a combination of bending and axial tension stresses here and the pin hole removes substantial material. The bending stress is due to the maximum mo-

Chapter 4

80.6

F32y +523 A

STRESS, STRAIN, AND DEFLECTION

F32y +523

15

F32x

7.5

brake arm

Section X-X

+353 X

F12x

B –1805

F52x

frame

20

11 d

X

58 r

42.5

29

10

2

66

Section B-B

B

F52y –204

+1452

2

F12z

–319

F52z

y

–587

13 18.2

x

FIGURE 4-54

pin

F52y –204

M21x –32 304

y

z

ment acting on the curved beam at its root and the tensile stress is due to the y component of the force at A. There is also a shear stress due to transverse loading, but this will be zero at the outer fiber where the sum of the bending and axial stresses is maximum. The area and area moment of inertia of the hub cross section are needed: Ahub = length ( d out − din ) = 28.5 ( 25 − 11) = 399 mm 2

(

12

) = 28.5 (253 − 113 ) = 33 948 mm 4

(f)

12

The stress on the left half of Section B-B is the sum of the bending and axial stresses: σ hub =

Mc F32 y 65 452 (12.5) 523 + = + = 25.4 MPa 33 948 399 I hub Ahub

( g)

The stress on the right half of Section B-B is lower, because the compression due to bending is reduced by the axial tension. 3 The straight portion of the brake arm is a cantilever beam loaded in two directions, in the xy plane and in the yz plane. The section moduli and moments are different in these bending directions. The z moment in the xy plane is equal and opposite to the moment on the curved section. The cross section at the root of the cantilever is a rectangle of 23 by 12 mm as shown in Figure 4-54. The bending stress at the outer fiber of the 23-mm side due to this moment is

σ y1

Mc = = I

12 65 452    2 23 (12 )3 12

= 118.6 MPa

F21y +319

1

F21x = +1805 M21y = – 52 370

Brake-Arm Free-Body Diagrams, Forces in N, Moments in N-mm, and Dimensions in mm

3 3 − din length d out

F21z –587

+587

23 by 12 mm 28.5

F12y

C

I hub =

253

(h)

MACHINE DESIGN

254

A1

y1

-

An Integrated Approach -

The x moment is due to force F52z acting at the 42.5 radius, bending the link in the z direction. The bending stress at the surface of the 12-mm side is

σ y2

flange

Mc = = I

23 589 ⋅ 42.5    2 12 ( 23)3 12

ri rc1

(a) Centroid of flange

web y2 rc2

A2

ri

Sixth Edition

= 23.7 MPa

(i )

These two y-direction normal stresses add at the corners of the two faces to give σ y = σ y1 + σ y2 = 118.6 + 23.7 = 142.2 MPa

( j)

4 Another possible failure point is the slot in the cantilever arm. Though the moment is zero there, the shear force is present and can cause tearout in the z direction. The tearout area is the shear area between the slot and edge: Atearout = thickness ( width ) = 8 ( 4 ) = 32 mm 2 τ=

F52z Atearout

(k )

589 = = 18.4 MPa 32

(b) Centroid of web

yt

5 The pivot pin is subjected to force F21, which has both x and y components and to a couple M21 due to the forces F12z and F52z. The force F21 creates a bending moment having components F21xl and F21yl in the yz and xz planes, respectively, where l = 29 mm is the length of the pin:

tee

ri

At

rct

(c) Centroid of tee

yt

e

M pin =

(M

21x

− F12 y ⋅ l

) + (− M 2

21y

+ F12 x ⋅ l

)

2

=

( −32 304 + 319 ⋅ 29)2 + (52 370 − 1805 ⋅ 29)2

=

( −32 304 + 9 251)2 + (52 370 − 52 345)2

=

( −23 053)2 + ( 25)2 = 23 053 N-mm

 25  θ M pin = tan −1   ≅ 0°  −23 053  rn

(l )

(m )

rct

Figure 4-56a shows the moment of the couple M21 and Figure 4-56b shows the moment of the force F21. Their combination is shown in Figure 4-56c.

(d) Neutral axis of tee

It is this combined moment that creates the largest bending stresses in the pin at 0° and 180° around its circumference. The maximum bending stress in the pin (with lengths in mm and moments in N-mm for unit balance) is

At

ri

FIGURE 4-55 Finding Neutral Axis of a Tee-Section Curved Beam—Case Study 4B

σ pin =

M pin c pin I pin

=

11 23 053    2 π (11) 64

4

=

23 053 ( 5.5) 718.7

= 176 MPa

(n)

Chapter 4

STRESS, STRAIN, AND DEFLECTION

255

6 A more complete analysis could be done using finite element methods to determine the stresses and deformations at many other locations on the part. You may examine the model for this case study by opening the file CASE4B in the program of your choice. A stress analysis of this case study is also done in Chapter 8 using Finite Element Analysis (FEA).

y

4.19

SUMMARY

The equations used for stress analysis are relatively few and are fairly easy to remember. (See the equation summary later in this section.) The major source of confusion among students seems to be in understanding when to use which stress equation and how to determine where in the part’s continuum to calculate the stresses, since they vary over the part’s internal geometry. There are two types of applied stresses of interest, normal stress σ and shear stress τ. Each may be present on the same stress element and they will combine to create a set of principal normal stresses and maximum shear stress, as evidenced on the Mohr’s circle plane. It is ultimately these principal stresses that we need to find in order to determine the safety of the design. So, regardless of the source of loading or type of stress that may be applied to the part, you should always determine the principal stresses and maximum shear stress that result from their combination. (See Sections 4.3 and 4.5.) There are only a few types of loading that commonly occur on machine parts, but they may occur in combination on the same part. Loading types that create applied normal stresses are bending loads, axial loads, and bearing loads. Bending loads will always create both tensile and compressive normal stresses at different locations within the part. Beams provide the most common example of bending loads. (See Section 4.9.) Axial loads create normal stresses that can be either tensile or compressive (but not both at once), depending on whether the axial load is in a tensile or compressive direction. (See Section 4.7.) Fasteners such as bolts often have significant axialtension loads. If the axial load is compressive, then there may be a danger of column buckling, and the equations of Section 4.16 must also be applied. Bearing loads create compressive normal stresses in the shaft and bushing (bearing). Loading types that create applied shear stresses are torsional loads, direct shear loads, and bending loads. Torsional loads involve the twisting of a part around its long axis by application of a torque. A transmission shaft is a typical example of a torsionally loaded part. (Chapter 10 deals with the design of transmission shafts.) Direct shear can be caused by loads that tend to slice the part transversely. Fasteners such as rivets or pins sometimes experience direct shear loads. A pin trying to tear its way out of its hole also causes direct shear on the tearout area. (See Section 4.8.) Bending loads also cause transverse shear stresses on the cross section of the beam. (See Section 4.9.) Stresses can vary continuously over the internal continuum of a part’s geometry and are calculated as acting at an infinitesimally small point within that continuum. To do a complete analysis of the stresses at all the infinite number of potential sites within the part would require infinite time, which we obviously do not have. So, we must intelligently select a few sites for our calculations such that they represent the worst-case situations.

x

pin

32.3 N-m

–52.4 N-m (a) Moment of couple F12z - F52z

52.4 N-m

pin

9.3 N-m

(b) Moment of force F21

pin

23.0 N-m

(c) Combined moments

FIGURE 4-56 Bending Moments on Pivot Pin - Case Study 4B

MACHINE DESIGN

256

l a

M1

F 〈x—a〉 – 1 x

R1

FIGURE 4-57 Cantilever Beam with Concentrated Load

Loading Diagram 60 40 20 0 –20 –40 –60 0 5 10 Shear Diagram 40 30 20 10 0 5

10

Moment Diagram 50 0 –50 –100 –150 –200 0

5

FIGURE 4-58 Load Distributions

An Integrated Approach -

Sixth Edition

The student needs to understand how the various stresses are distributed within the continuum of a loaded part. There are two aspects to determining appropriate locations on a given part at which to make the stress calculations. The first aspect concerns the load distribution over the part’s geometry and the second concerns the stress distribution within the part’s cross section. For example, consider a straight cantilever beam loaded with a single force at some point along its length as shown in Figure 4-57. The first aspect requires some knowledge of the way the loads on the beam are distributed in response to the applied force. This comes from an analysis of the beam’s shear and moment diagrams as shown in Figure 4-58, which indicate that in this case, the section with greatest load is at the wall. We would then concentrate our attention on a vanishingly thin “bologna slice” taken from this beam at the wall. Note that the presence of stress concentrations at other locations having lower nominal stresses would require their investigation too. The second aspect is then to determine where in this cross-sectional “bologna slice” the stresses will be greatest. Figures in the relevant sections of this chapter show the stress distributions across sections for various types of loadings. These stress-distribution diagrams are collected in Figure 4-59, which also shows the relevant stress equations for each case. Since the loading in this beam example creates bending stresses, we must understand that there will be a normal stress that is maximally compressive at one extreme fiber and maximally tensile at the other extreme fiber as shown in Figures 4-15 and 4-59c. Thus, we would take a stress element at the outer fiber of this slice of the beam to calculate the worst-case bending normal stress from equation 4.11b. Bending loads also cause shear stress, but its distribution is maximum at the neutral plane and zero at the outer fiber as shown in Figures 4-19 and 4-59d. So, a different stress element is taken at the neutral plane of the cross-sectional slice for calculation of the shear stress due to transverse loading, using the appropriate equation such as equation 4.14b for a rectangular cross section. Each of these two stress elements will have its own set of principal stresses and maximum shear stress, which can be calculated from equation 4.6a for this 2-D case.

50

0

-

10

More complicated loadings on more complicated geometries may have multiple stresses applied to the same infinitesimal stress element. It is very common in machine parts to have loadings that create both bending and torsion on the same part. Example 4-9 deals with such a case and should be studied carefully. Stress is only one consideration in design. The deflections of parts must also be controlled for proper function. Often, a requirement for small deflections will dominate the design and require thicker sections than would be necessary to guard against excessive stress. The deflections of a designed part, as well as its stresses, should always be checked. Equations for deflection under various loadings are given in the relevant sections and are also collected in Appendix B for beams of various types and loadings. Figure 4-60 shows a flow chart depicting a set of steps that can be followed to analyze stresses and deflections under static loading.

Chapter 4

STRESS, STRAIN, AND DEFLECTION

257

Static Loading Stress Analysis Materials are assumed to be Homogeneous and Isotropic

y σ (a) Uniaxial tension, stress distribution across section

z

x

σ=

P A

Eq. 4.7

y τ (b) Direct shear, average-stress distribution across section

z

x

P Ashear

τ=

Eq. 4.9

y σ (c) Bending, normalstress distribution across section

z

x

σ=

My I

Eq. 4.11a

y

τ

z

τ

x

y

τ

z τ

VQ Ib

Eq. 4.13d

τ=

Tr J

Eq. 4.23b

τ τ

FIGURE 4-59

τ=

Determine the stress distributions within the cross sections of interest and identify locations of the highest applied and combined stresses.

Calculate the applied stresses acting on each face of every element, and then calculate the principal stresses and maximum shear stress resulting therefrom.

τ

τ (e) Torsion, shearstress distribution across section

Based on the load distributions over the part’s geometry, determine what cross sections of the part are most heavily loaded.

Draw a 3-D stress element for each of the selected points of interest within the section and identify the stresses acting on it.

σ

(d) Bending, shearstress distribution across section

Find all applied forces, moments, torques, etc. and draw the freebody diagrams to show them applied to the part’s geometry.

τ

Distribution of Stresses Across a Cross Section Under Various Types of Loading

Calculate critical deflections of the parts.

FIGURE 4-60 Flow Chart for Static Stress Analysis

258

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Important Equations Used in This Chapter See the referenced sections for information on the proper use of these equations. The Stress Cubic—its roots are the 3-D principal stresses (Section 4.3):

σ 3 − C2 σ 2 − C1σ − C0 = 0

(4.4c)

where C2 = σ x + σ y + σ z C1 = τ 2xy + τ 2yz + τ 2zx − σ x σ y − σ y σ z − σ z σ x C0 = σ x σ y σ z + 2τ xy τ yz τ zx − σ x τ 2yz − σ y τ 2zx − σ z τ 2xy Maximum Shear Stresses (Section 4.3):

τ13 =

σ1 − σ 3 2

σ 2 − σ1

τ 21 =

;

2

;

τ32 =

σ3 − σ 2 2

(4.5)

Two-Dimensional Principal Stresses (Section 4.3):

σ a ,σ b =

σx + σy 2

2  σx − σy  2 ±   + τ xy 2 

σ c =0 σ1 − σ 3

τ max = τ13 =

2

(4.6a)

(4.6b)

Axial Tension Stress (Section 4.7):

σx =

P A

(4.7)

s=

Pl AE

(4.8)

P Ashear

(4.9)

Axial Deflection (Section 4.7):

Direct Shear Stress (Section 4.8):

τ xy = Direct Bearing Area (Section 4.8):

Abearing =

π ld 4

(4.10 b)

Maximum Bending Stress—Straight Beams (Section 4.9):

σ max =

Mc I

(4.11b)

Maximum Bending Stress—Curved Beams (Section 4.9):

σi = +

M  ci  eA  ri 

(4.12b)

Chapter 4

STRESS, STRAIN, AND DEFLECTION

Transverse Shear Stress in Beams—General Formula (Section 4.9):

τ xy =

V bI

c

∫y

y dA

(4.13 f )

1

Maximum Transverse Shear Stress—Rectangular Beam (Section 4.9):

3V 2 A

τ max =

(4.14 b)

Maximum Transverse Shear Stress—Round Beam (Section 4.9):

τ max =

4V 3 A

(4.15c)

Maximum Transverse Shear Stress—I-Beam (Section 4.9):

τ max ≅

V Aweb

(4.16)

The General Beam Equations (Section 4.9):

q d4y = 4 EI dx

(4.18a)

V d3y = 3 EI dx

(4.18b)

M d2y = EI dx 2

(4.18c)

θ=

dy dx

y = f (x)

(4.18d ) (4.18e)

Maximum Torsional Shear Stress—Round Section (Section 4.12):

τ max =

Tr J

(4.23b)

Maximum Torsional Deflection—Round Section (Section 4.12):

θ=

Tl JG

(4.24)

Maximum Torsional Shear Stress—Nonround Section (Section 4.12):

τ max =

T Q

(4.26a)

Maximum Torsional Deflection—Nonround Section (Section 4.12):

θ=

Tl KG

(4.26b)

259

260

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

Spring Rate or Spring Constant—linear (a), angular (b) (Section 4.14):

k=

F y

(4.27a)

k=

T θ

(4.27b)

Stress with Stress Concentration (Section 4.15):

σ max = K t σ nom

(4.31)

τ max = K ts τ nom Column Radius of Gyration (Section 4.16):

I A

(4.34)

l k

(4.33)

k= Column Slenderness Ratio (Section 4.16):

Sr =

Column Critical Unit Load—Euler Formula (Section 4.16):

Pcr π 2 E = 2 A Sr

(4.38c)

Column Critical Unit Load—Johnson Formula (Section 4.16):

2E S yc

( Sr )D = π

(4.42)

2 Pcr 1  S yc Sr  = S yc −  A E  2 π 

(4.43)

Column Critical Unit Load—Secant Formula (Section 4.16):

P = A

S yc  leff  ec  1 +  2  sec  k   k

P  4 EA 

(4.46c)

Pressurized Cylinder (Section 4.17)

σt =

σr =

σa =

pi ri2 − po ro2 ro2 − ri2

pi ri2 − po ro2 ro2 − ri2

pi ri2 − po ro2 ro2 − ri2

+



ri2 ro2 ( pi − po )

(

r 2 ro2 − ri2

)

ri2 ro2 ( pi − po )

(

r 2 ro2 − ri2

)

(4.47a)

(4.47b)

(4.47c)

Chapter 4

4.20

STRESS, STRAIN, AND DEFLECTION

REFERENCES

261

Table P4-0†

1 I. H. Shames and C. L. Dym, Energy and Finite Element Methods in Structural Mechanics. Hemisphere Publishing: New York, Sect. 1.6, 1985.

Topic/Problem Matrix

2 I. H. Shames and F. A. Cossarelli, Elastic and Inelastic Stress Analysis. Prentice-Hall: Englewood Cliffs, N.J., pp. 46-50, 1991.

Sect. 4.1, 4.2 Stress, Strain

3 R. E. Peterson, Stress Concentration Factors. John Wiley & Sons: New York, 1974.

Sect. 4.5 Mohr’s Circles

4 R. J. Roark and W. C. Young, Formulas for Stress and Strain. 6th ed. McGraw-Hill: New York, 1989.

4-1, 4-79, 4-93, 4-100

5 R. E. Peterson, Stress Concentration Factors. John Wiley & Sons: New York, p. 150, 1974.

4-2, 4-18, 4-61, 4-74a, 4-95

6 W. D. Pilkey, Peterson’s Stress Concentration Factors, John Wiley & Sons: New York, 1997. 7 H. T. Grandin and J. J. Rencis, Mechanics of Materials, John Wiley & Sons: New York, pp. 176-177, 2006. 8 N. Troyani, C. Gomes, and G. Sterlacci, “Theoretical Stress Concentration Factors for Short Rectangular Plates With Centered Circular Holes.” ASME J. Mech. Design, V. 124, pp. 126-128, 2002. 9 N. Troyani, et al., “Theoretical Stress Concentration Factors for Short Shouldered Plates Subjected to Uniform Tensions.” IMechE J. Strain Analysis, V. 38, pp. 103-113, 2003. 10 N. Troyani, G. Sterlacci, and C. Gomes, “Simultaneous Considerations of Length and Boundary Conditions on Theoretical Stress Concentration Factors.” Int. J. Fatigue, V. 25, pp. 353-355, 2003.

4.21

BIBLIOGRAPHY For general information on stress and deflection analysis, see:

F. P. Beer and E. R. Johnston, Mechanics of Materials. 2nd ed. McGraw-Hill: New York, 1992. J. P. D. Hartog, Strength of Materials. Dover: New York, 1961.

Sect. 4.8 Direct Shear, Bearing, and Tearout 4-4, 4-5, 4-6, 4-7, 4-9, 4-15, 4-19, 4-20, 4-22, 4-47, 4-59, 4-60, 4-74f, 4-87 Sect. 4.9 Straight Beams 4-10, 4-11, 4-12, 4-13, 4-14, 4-27, 4-40, 4-43a, 4-64, 4-65, 4-66, 4-67, 4-68, 4-74b,4-74 c, 4-74g, 4-91 Sect. 4.9 Curved Beams 4-17, 4-37, 4-62, 4-63, 4-69 to 4-72, 4-73, 4-74e Sect. 4.10 Deflection 4-8, 4-16, 4-23, 4-24, 4-25, 4-26, 4-28, 4-43b, 4-44, 4-48, 4-98 Sect. 4.11 Castigliano’s Method 4-84, 4-85, 4-86, 4-96, 4-97 Sect. 4.12 Torsion 4-21, 4-34, 4-46, 4-74d, 4-74h, 4-81, 4-82, 4-89, 4-90 Sect. 4.13 Combined Stresses

I. H. Shames, Introduction to Solid Mechanics. Prentice-Hall: Englewood Cliffs, N.J., 1989.

Sect. 4.14 Spring Rates

S. Timoshenko and D. H. Young, Elements of Strength of Materials. 5th ed. Van Nostrand: New York, 1968.

*4-1

Sect. 4.7 Axial Tension

R. J. Roark and W. C. Young, Formulas for Stress and Strain. 6th ed. McGraw-Hill: New York, 1989.

I. H. Shames and F. A. Cossarelli, Elastic and Inelastic Stress Analysis. Prentice-Hall: Englewood Cliffs, N.J., 1991.

4.22

4-55, 4-56, 4-57, 4-58, 4-94

PROBLEMS A differential stress element has a set of applied stresses on it as indicated in each row of Table P4-1. For the row(s) assigned, draw the stress element showing the applied

4-3, 4-34, 4-46, 4-88, 4-92

4-29, 4-30, 4-31, 4-32, 4-35, 4-38, 4-39 Sect. 4.15 Stress Conc. 4-36, 4-75 to 4-78 Sect. 4.16 Columns 4-45, 4-49, 4-50, 4-51, 4-52, 4-53, 4-54 Sect. 4.17 Cylinders 4-41, 4-42, 4-80, 4-83, 4-99 * Answers

to these problems are provided in Appendix D.

MACHINE DESIGN

262

Table P4-1

-

An Integrated Approach -

Sixth Edition

Data for Problem 4-1 (in psi) Rows a–g and k–m are two-dimensional, others are 3-D problems

60 mm

170 mm

F

T

Row

σx

a

1000

0

0

500

0

0

b

–1000

0

0

750

0

0

σy

σz

τ xy

τ yz

τ zx

c

500

–500

0

1000

0

0

d

0

–1500

0

750

0

0

e

750

250

0

500

0

0

f

–500

1000

0

750

0

0

g

1000

0

–750

0

0

250

h

750

500

250

500

0

0

i

1000

–250

–750

250

500

750

j

–500

750

250

100

250

1000

FIGURE P4-1

k

1000

0

0

0

0

0

Problem 4-3 (A Solidworks model

l

1000

0

0

0

500

0

m

1000

0

0

0

0

500

n

1000

1000

1000

500

0

0

o

1000

1000

1000

500

500

0

p

1000

1000

1000

500

500

500

of this is on the book’s website)

40 mm

stresses, find the principal stresses and maximum shear stress analytically, and check the results by drawing Mohr’s circles for that stress state. 4-2 A 400-lb chandelier is to be hung from two 10-ft-long solid steel cables in tension. Choose a suitable diameter of cable which will not exceed an allowable stress of 5000 psi. What will be the deflection of the cables? State all assumptions. FIGURE P4-2

†4-3

For the bicycle pedal arm assembly in Figure P4-1 with a rider-applied force of 1500 N at the pedal, determine the maximum principal stress in the pedal arm if its cross section is 15 mm in dia. The pedal attaches to the pedal arm with a 12-mm screw thread. What is the stress in the pedal screw?

†*4-4

The trailer hitch shown in Figure P4-2 and Figure 1-1 has loads applied as defined in Problem 3-4. The 100-kgf (980 N) tongue weight acts downward and the pull force of 4905 N acts horizontally. Using the dimensions of the ball bracket shown in Figure 1-5, determine:

Problems 4-4, 4-5, 4-6

(A Solidworks model of this is on the book’s website)

* Answers

to these problems are provided in Appendix D.

(a) (b) (c) (d) (e)

Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. †

Problems with numbers in italics are design problems.

The principal stresses in the shank of the ball where it joins the ball bracket. The bearing stress in the ball bracket hole. The tearout stress in the ball bracket. The normal and shear stresses in the attachment bolts if they are 19-mm dia. The principal stresses in the ball bracket as a cantilever.

†4-5

Repeat Problem 4-4 for the loading conditions of Problem 3-5.

†*4-6

Repeat Problem 4-4 for the loading conditions of Problem 3-6.

†*4-7

Design the wrist pin of Problem 3-7 for a maximum allowable principal stress of 20 kpsi if the pin is hollow and loaded in double shear.

†*4-8

A paper mill processes rolls of paper having a density of 984 kg/m3. The paper roll is 1.50 m outside dia (OD) X 220 mm inside dia (ID) X 3.23-m long and is on a simply supported, hollow, steel shaft. Find the shaft ID needed to obtain a maximum deflection

Chapter 4

STRESS, STRAIN, AND DEFLECTION

263

P F

P

F 0.5-cm grid

FIGURE P4-3 Problem 4-9 (A Solidworks model of this is on the website)

at the center of 3 mm if the shaft OD is 220 mm. Assume the shaft to be the same length between supports as the length of the paper roll. 4-9 For the ViseGrip® plier-wrench drawn to scale in Figure P4-3, and for which the forces were analyzed in Problem 3-9, find the stresses in each pin for an assumed clamping force of P = 4000 N in the position shown. The pins are 8-mm dia and are all in double shear. *4-10

*4-11

The overhung diving board of Problem 3-10 is shown in Figure P4-4a. Assume crosssection dimensions of 305 mm X 32 mm. The material has E = 10.3 GPa. Find the largest principal stress at any location in the board when a 980 N person is standing in the center of the width of the board at the free end. What is the maximum deflection? Repeat Problem 4-10 using the loading conditions of Problem 3-11. Assume the board weighs 284.4 N and deflects 131 mm statically when the person stands on it. Find the largest principal stress anywhere in the board when the 980 N person in Problem 4-10 jumps up 250 mm and lands back on the board. Find the maximum deflection.

* Answers

to these problems are provided in Appendix D. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. †

Problems with numbers in italics are design problems.

4-12 Repeat Problem 4-10 using the cantilevered diving board design in Figure P4-4b. 4-13 Repeat Problem 4-11 using the diving board design shown in Figure P4-4b. Assume the board weighs 186.3 N and deflects 85 mm statically when the person stands on it. † 4-14

Figure P4-5 shows a child’s toy called a pogo stick. The child stands on the pads, applying half her weight on each side. She jumps up off the ground, holding the pads up 2m

2m 0.7 m

( a ) Overhung diving board

FIGURE P4-4 Problems 4-10 through 4-13

P

0.7 m

( b) Cantilevered diving board

P

MACHINE DESIGN

264

W/2

-

An Integrated Approach -

Sixth Edition

W/2 FIGURE P4-6 Problem 4-16

against her feet, and bounces along with the spring cushioning the impact and storing energy to help each rebound. Assume a 60-lb (266.9 N) child and a spring constant of 100 lb/in (17.5 N/mm). The pogo stick weighs 5 lb (22.2 N). Design the aluminum cantilever beam sections on which she stands to survive jumping 2 in (50.8 mm) off the ground. Allowable stress = 20 kpsi (138 MPa). Size the beam and define its shape. * † 4-15

P FIGURE P4-5 Problem 4-14

F F

A

Design a shear pin for the propeller shaft of an outboard motor if the shaft through which the pin is placed is 25-mm diameter, the propeller is 200-mm diameter, and the pin must fail when a force > 400 N is applied to the propeller tip. Assume an ultimate shear strength for the pin material of 100 MPa.

4-16 A track to guide bowling balls is designed with two round rods as shown in Figure P4-6. The rods are not parallel to one another but have a small angle between them. The balls roll on the rods until they fall between them and drop onto another track. The angle between the rods is varied to cause the ball to drop at different locations. Each rod’s unsupported length is 30 in and the angle between them is 3.2°. The balls are 4.5-in dia and weigh 2.5 lb. The center distance between the 1-in-dia rods is 4.2 in at the narrow end. Find the maximum stress and deflection in the rods. (a) Assume the rods are simply supported at each end. (b) Assume the rods are fixed at each end.

r

4-17 A pair of ice tongs is shown in Figure P4-7. The ice weighs 50 lb and is 10 in wide across the tongs. The distance between the handles is 4 in, and the mean radius r of a tong is 6 in. The rectangular cross-sectional dimensions are 0.75 in deep X 0.312 in wide. Find the stress in the tongs.

W

*4-18

FIGURE P4-7 Problem 4-17 * Answers

to these problems are provided in Appendix D. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. †

Problems with numbers in italics are design problems.

* † 4-19

A set of steel reinforcing rods is to be stretched axially in tension to create a tensile stress of 30 kpsi prior to being cast in concrete to form a beam. Determine how much force will be required to stretch them the required amount and how much deflection is required. There are 10 rods; each is 0.75-in dia and 30 ft long. The clamping fixture used to pull the rods in Problem 4-18 is connected to the hydraulic ram by a clevis like that shown in Figure P4-8. Determine the size of the clevis pin needed to withstand the applied force. Assume an allowable shear stress of 20 kpsi and an allowable normal stress of 40 kpsi. Determine the required outside radius of the clevis end to not exceed the above allowable stresses in either tearout or bearing if the clevis flanges are each 0.8 in thick.

Chapter 4

† 4-20

STRESS, STRAIN, AND DEFLECTION

265

Repeat Problem 4-19 for 12 rods each 10-mm dia and 10 m long. The desired rod stress is 200 MPa. The allowable normal stress in the clevis and pin is 280 MPa and their allowable shear stress is 140 MPa. Each clevis flange is 20 mm wide.

P

4-21 Figure P4-9 shows an automobile wheel with two common styles of lug wrench being used to tighten the wheel nuts, a single-ended wrench in (a), and a double-ended wrench in (b). In each case two hands are required to provide forces, respectively, at A and B as shown. The distance between points A and B is 1 ft in both cases and the handle diameter is 0.625 in. The wheel nuts require a torque of 70 ft-lb. Find the maximum principal stress and maximum deflection in each wrench design. *4-22

An in-line “roller-blade” skate is shown in Figure P4-10. The polyurethane wheels are 72-mm dia and spaced on 104-mm centers. The skate-boot-foot combination weighs 2 kg. The effective “spring rate” of the person-skate system is 6000 N/m. The axles are 10-mm-dia steel pins in double shear. Find the stress in the pins for a 100-kg person landing a 0.5-m jump on one foot. (a) Assume that all four wheels land simultaneously. (b) Assume that one wheel absorbs all the landing force.

*4-23

A beam is supported and loaded as shown in Figure P4-11a. Find the reactions, maximum shear, maximum moment, maximum slope, maximum bending stress, and maximum deflection for the data given in the assigned row(s) in Table P4-2.

*4-24

A beam is supported and loaded as shown in Figure P4-11b. Find the reactions, the maximum shear, maximum moment, maximum slope, maximum bending stress, and maximum deflection for the data given in the assigned row(s) in Table P4-2.

*4-25

A beam is supported and loaded as shown in Figure P4-11c. Find the reactions, the maximum shear, maximum moment, maximum slope, maximum bending stress, and maximum deflection for the data given in the assigned row(s) in Table P4-2.

*4-26

A beam is supported and loaded as shown in Figure P4-11d. Find the reactions, maximum shear, maximum moment, maximum slope, maximum bending stress, and maximum deflection for the data given in the assigned row(s) in Table P4-2.

† 4-27

A storage rack is to be designed to hold the paper roll of Problem 4-8 as shown in Figure P4-12. Determine suitable values for dimensions a and b in the figure. Consider bending, shear, and bearing stresses. Assume an allowable tensile/compressive stress of 100 axle

axle

3 in

3 in A

B

lug wrench

A

lug wrench B F

F

F

FIGURE P4-9 Problem 4-21

(a)

tire

F

tire (b)

P FIGURE P4-8 Problems 4-19 and 4-20

* Answers

to these problems are provided in Appendix D. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. †

Problems with numbers in italics are design problems.

266

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

l

l b

a

a

F

F w x

R1

w

M1

x

R2

R1

( a)

( b)

l

l b

b

F

a

a

F w

w x R1

R2

x R1

R2

(c)

R3

(d)

FIGURE P4-11 Beam Loadings for Problems 4-23 to 4-26, 4-29 to 4-32 and 4-96—See Table P4-2 for Data

MPa and an allowable shear stress of 50 MPa for both stanchion and mandrel, which are steel. The mandrel is solid and inserts halfway into the paper roll. Balance the design to use all of the material strength. Calculate the deflection at the end of the roll.



Problems with numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

†4-28

Figure P4-13 shows a forklift truck negotiating a 15° ramp to drive onto a 4-ft-high loading platform. The truck weighs 5000 lb and has a 42-in wheelbase. Design two (one for each side) 1-ft-wide ramps of steel to have no more than 1-in deflection in the worst case of loading as the truck travels up them. Minimize the weight of the ramps by using a sensible cross-sectional geometry.

*4-29

Find the spring rate of the beam in Problem 4-23 at the applied concentrated load for the row(s) assigned in Table P4-2.

*4-30

Find the spring rate of the beam in Problem 4-24 at the applied concentrated load for the row(s) assigned in Table P4-2.

* Answers

to these problems are provided in Appendix D.

b stanchion paper roll

a mandrel FIGURE P4-12 Problem 4-27

base

Chapter 4

Table P4-2

*

STRESS, STRAIN, AND DEFLECTION

267

Data for Problems 4-23 through 4-26, 4-29 through 4-32, and 4-96

Use only data relevant to the particular problem. Lengths in m, forces in N, I in m4.

Row

l

a

b

w*

F

I

c

E

a

1.00

0.40

0.60

200

500

2.85E–08

2.00E–02

steel

b

0.70

0.20

0.40

80

850

1.70E–08

1.00E–02

steel

c

0.30

0.10

0.20

500

450

4.70E–09

1.25E–02

steel

d

0.80

0.50

0.60

65

250

4.90E–09

1.10E–02

steel

e

0.85

0.35

0.50

96

750

1.80E–08

9.00E–03

steel

f

0.50

0.18

0.40

450

950

1.17E–08

1.00E–02

steel

g

0.60

0.28

0.50

250

250

3.20E–09

7.50E–03

steel

h

0.20

0.10

0.13

400

500

4.00E–09

5.00E–03

alum

i

0.40

0.15

0.30

50

200

2.75E–09

5.00E–03

alum

j

0.20

0.10

0.15

150

80

6.50E–10

5.50E–03

alum

k

0.40

0.16

0.30

70

880

4.30E–08

1.45E–02

alum

l

0.90

0.25

0.80

90

600

4.20E–08

7.50E–03

alum

m

0.70

0.10

0.60

80

500

2.10E–08

6.50E–03

alum

n

0.85

0.15

0.70

60

120

7.90E–09

1.00E–02

alum

Note that w is a unit force of N/m

*4-31

Find the spring rate of the beam in Problem 4-25 at the applied concentrated load for the row(s) assigned in Table P4-2.

*4-32

Find the spring rate of the beam in Problem 4-26 at the applied concentrated load for the row(s) assigned in Table P4-2.

*4-33

For the bracket shown in Figure P4-14 and the data in the row(s) assigned from Table P4-3, determine the bending stress at point A and the shear stress due to transverse loading at point B. Also find the torsional shear stress at both points. Then determine the principal stresses at points A and B.

*4-34

For the bracket shown in Figure P4-14 and the data in the row(s) assigned from Table P4-3, determine the deflection at load F.

*4-35

For the bracket shown in Figure P4-14 and the data in the row(s) assigned from Table P4-3, determine the spring rate of the tube in bending, the spring rate of the arm in bend-

ramp

FIGURE P4-13 Problem 4-28

1-ft grid

* Answers

to these problems are provided in Appendix D.

MACHINE DESIGN

268

-

An Integrated Approach -

Sixth Edition

F

l a A B

t

y tube

wall

arm

z

h

x od

id

FIGURE P4-14 Problems 4-33 to 4-36 (A Solidworks model of this is on the book’s website)

ing, and the spring rate of the tube in torsion. Combine these into an overall spring rate in terms of the force F and the linear deflection at force F. 4-36 For the bracket shown in Figure P4-14 and the data in the row(s) assigned from Table P4-3, redo Problem 4-33 considering the stress concentration at points A and B. Assume a stress-concentration factor of 2.5 in both bending and torsion. * Answers

to these problems are provided in Appendix D.

*4-37

A semicircular curved beam as shown in Figure P4-15 has OD = 150 mm, ID = 100 mm, and t = 25 mm. For a load pair F = 14 kN applied along the diameter, find the eccentricity of the neutral axis and the stress at the inner and outer fibers.

† 4-38

Design a solid, straight, steel torsion bar to have a spring rate of 10 000 in-lb per radian per foot of length. Compare designs of solid round and solid square cross sections. Which is more efficient in terms of material use?

† 4-39

Design a 1-ft-long steel, end-loaded cantilever spring for a spring rate of 10 000 lb/in at the load. Compare designs of solid round and solid square cross sections. Which is more efficient in terms of material use?

† 4-40

Redesign the roll support of Problem 4-8 to be like that shown in Figure P4-16. The stub mandrels insert to 10% of the roll length at each end. Choose appropriate dimensions a and b to fully utilize the material’s strength, which is the same as in Problem 4-27. See Problem 4-8 for additional data.

*4-41

A 10-mm ID steel tube carries liquid at 7 MPa. Determine the principal stresses in the wall if its thickness is: (a) 1 mm, (b) 5 mm.



Problems with numbers in italics are design problems.

t

b typ OD ID

F

stanchion paper roll

F

a typ

mandrel

FIGURE P4-15 Problem 4-37 (A Solidworks model is on the book’s website)

FIGURE P4-16

base

Problem 4-40 (A Solidworks model of this is on the book’s website)

Chapter 4

Table P4-3

STRESS, STRAIN, AND DEFLECTION

Data for Problems 4-33–4-36, 4-49–4-52, and 4-89–4-92 Incl. Use only data that are relevant to the particular problem. Lengths in mm, forces in N.

Row

l

a

t

h

F

OD

ID

E

a

100

400

10

20

50

20

14

steel

b

70

200

6

80

85

20

6

steel

c

300

100

4

50

95

25

17

steel

d

800

500

6

65

160

46

22

alum

e

85

350

5

96

900

55

24

alum

f

50

180

4

45

950

50

30

alum

g

160

280

5

25

850

45

19

steel

h

200

100

2

10

800

40

24

steel

i

400

150

3

50

950

65

37

steel

j

200

100

3

10

600

45

32

alum

k

120

180

3

70

880

60

47

alum

l

150

250

8

90

750

52

28

alum

m

70

100

6

80

500

36

30

steel

n

85

150

7

60

820

40

15

steel

4-42 A cylindrical tank with hemispherical ends is required to hold 150 psi of pressurized air at room temperature. Find the principal stresses in the 1-mm-thick wall if the tank diameter is 0.5 m and its length is 1 m. 4-43 Figure P4-17 shows an off-loading station at the end of a paper rolling machine. The finished paper rolls are 0.9-m OD by 0.22-m ID by 3.23-m long and have a density of 984 kg/m3. The rolls are transferred from the machine conveyor (not shown) to the forklift truck by the V-linkage of the off-load station, which is rotated through 90° by an air cylinder. The paper then rolls onto the waiting forks of the truck. The forks are 38-mm thick by 100-mm wide by 1.2-m long and are tipped at a 3° angle from the horizontal.

V-links

1m crank arm

paper rolling machine forks A rod FIGURE P4-17 Problems 4-43 to 4-47

off-loading station

air cylinder

fork lift truck

269

270

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

rollers FIGURE P4-18 Problem 4-48

Find the stresses in the two forks on the truck when the paper rolls onto it under two different conditions (state all assumptions):



Problems with numbers in italics are design problems.

(a) The two forks are unsupported at their free end. (b) The two forks are contacting the table at point A. † 4-44

Determine a suitable thickness for the V-links of the off-loading station of Figure P4-17 to limit their deflections at the tips to 10 mm in any position during their rotation. Assume that there are two V-links supporting the roll, arranged at the 1/4 and 3/4 points along the roll’s length, and that each of the V arms is 10 cm wide by 1 m long. The V arms are welded to a steel tube that is rotated by the air cylinder. See Problem 4-43 for more information.

4-45 Determine the critical load on the air-cylinder rod in Figure P4-17 if the crank arm that it rotates is 0.3 m long and the rod has a maximum extension of 0.5 m. The 25-mm-dia rod is solid steel with a yield strength of 400 MPa. State all assumptions. 4-46 The V-links of Figure P4-17 are rotated by the crank arm through a shaft that is 60-mm dia by 3.23 m long. Determine the maximum torque applied to this shaft during the motion of the V-linkage and find the maximum stress and deflection for the shaft. See Problem 4-43 for more information. 4-47 Determine the maximum forces on the pins at each end of the air cylinder of Figure P4-17. Determine the stress in these pins if they are 30-mm dia and in single shear. † 4-48

A 100-kg wheelchair marathon racer wants an exerciser that will allow indoor practicing in any weather. The design shown in Figure P4-18 is proposed. Two free-turning rollers on bearings support the rear wheels. A platform supports the front wheels. Design the 1-m-long rollers as hollow tubes of aluminum to minimize the height of the platform and also limit the roller deflections to 1 mm in the worst case. The wheelchair has 65-cm-dia drive wheels separated by a 70-cm track width. The flanges shown on the rollers limit the lateral movement of the chair while exercising and thus the wheels can be anywhere between those flanges. Specify suitably sized steel axles to support the tubes on bearings. Calculate all significant stresses.

*4-49

A hollow, square column has a length l and material E, as shown in the row(s) assigned in Table P4-3. Its cross-sectional dimensions are 4 mm outside and 3 mm inside. Use Sy = 150 MPa for aluminum and 300 MPa for steel. Determine if it is a Johnson or an Euler column and find the critical load:

* Answers

to these problems are provided in Appendix D.

(a) (b) (c) (d) *4-50

If its boundary conditions are pinned-pinned. If its boundary conditions are fixed-pinned. If its boundary conditions are fixed-fixed. If its boundary conditions are fixed-free.

A hollow, round column has a length of 1.5 m, material E, and cross-sectional dimensions OD and ID as shown in the row(s) assigned in Table P4-3. Use Sy = 150 MPa for

Chapter 4

STRESS, STRAIN, AND DEFLECTION

aluminum and 300 MPa for steel. Determine if it is a Johnson or an Euler column and find the critical load: (a) (b) (c) (d) *4-51

to these problems are provided in Appendix D.

If its boundary conditions are pinned-pinned. If its boundary conditions are fixed-pinned. If its boundary conditions are fixed-fixed. If its boundary conditions are fixed-free.

P

A

If its boundary conditions are pinned-pinned. If its boundary conditions are fixed-pinned. If its boundary conditions are fixed-fixed. If its boundary conditions are fixed-free.

B

A solid, circular column has a length l, material E, diameter OD, and an eccentricity t as shown in the row(s) assigned in Table P4-3. Use Sy = 150 MPa for aluminum and 300 MPa for steel. Determine if it is a Johnson or an Euler column and find the critical load: (a) (b) (c) (d)

†4-53

* Answers

A solid, rectangular column has a length l, material E, and cross-sectional dimensions h and t as shown in the row(s) assigned in Table P4-3. Use Sy = 150 MPa for aluminum and 300 MPa for steel. Determine if it is a Johnson or an Euler column and find the critical load: (a) (b) (c) (d)

*4-52

271

If its boundary conditions are pinned-pinned. If its boundary conditions are fixed-pinned. If its boundary conditions are fixed-fixed. If its boundary conditions are fixed-free.

Design an aluminum, hollow, circular column for the following data: 3 m long, 5 mm wall thickness, 900 N concentric load, material yield strength of 150 MPa, and safety factor of 3: (a) If its boundary conditions are pinned-pinned. (b) If its boundary conditions are fixed-free.

4-54 Three round, 1.25-in-dia bars are made of SAE 1030 hot-rolled steel but are of different lengths, 5 in, 30 in, and 60 in, respectively. They are loaded axially in compression. Compare the load-supporting capability of the three bars if the ends are assumed to be: (a) (b) (c) (d)

Pinned-pinned. Fixed-pinned. Fixed-fixed. Fixed-free.

4-55 Figure P4-19 shows a 1.5-in-dia, 30-in-long steel rod subjected to tensile loads P = 10 000 lb applied at each end of the rod, acting along its longitudinal Y axis and through the centroid of its circular cross section. Point A is 12 in below the upper end and point B is 8 in below A. For this bar with this loading find: (a) All components of the stress tensor matrix (equation 4.1a) for a point midway between A and B. (b) The displacement of point B relative to point A. (c) The elastic strain in the section between A and B. (d) The total strain in the section between A and B. 4-56 The rod in Figure P4-19, with the loading of Problem 4-55, is subjected to a reduction in temperature from 80°F to 20°F after the load is applied. The coefficient of thermal expansion for steel is approximately 6 µin/in/°F. Find: (a) All components of the stress tensor matrix (equation 4.1a) for a point midway between A and B. (b) The displacement of point B relative to point A. (c) The elastic strain in the section between A and B. (d) The total strain in the section between A and B. 4-57 Figure P4-20 shows a steel bar fastened to a rigid ground plane with two 0.25-in-dia hardened steel dowel pins. For P = 1500 lb, find:

P Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P4-19 Problems 4-55, 4-56 and 4-94 †

Problems with numbers in italics are design problems.

P h 0.25"

P

bar 4"

pin

2" pin ground Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P4-20 Problems 4-57 and 4-58

(A Solidworks model of this is on the book’s website)

MACHINE DESIGN

272

-

An Integrated Approach -

P

Sixth Edition

A C 2.5"

40

B

A 10"

B

3"

7"

10

5"

3"

D 1.5" 8" FIGURE P4-22 35

12" P

3"

0.5"

Copyright © 2018 Robert L. Norton: All Rights Reserved

P

Problems 4-60 and 4-61 (A Solidworks model of this is on the book’s website)

P dimensions in mm Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P4-21 Problem 4-59 (A Solidworks model is on the book’s website)

P

20"

20"

(a) The shear stress in each pin. (b) The direct bearing stress in each pin and hole. (c) The minimum value of dimension h to prevent tearout failure if the steel bar has a shear strength of 32.5 kpsi. 4-58 Repeat Problem 4-57 for P = 2200 lb. 4-59 Figure P4-21 shows a rectangular section aluminum bar subjected to off-center forces P = 4000 N applied as shown: (a) Solve for the maximum normal stress in the mid-region of the bar well away from the eyes where the loads are applied. (b) Plot the normal stress distribution across the cross section at this mid-region of the bar. (c) Sketch a “reasonable” plot of the normal stress distribution across the cross section at the ends, close to the applied loads. 4-60 Figure P4-22 shows a bracket machined from 0.5-in-thick steel flat stock. It is rigidly attached to a support and loaded with P = 5000 lb at point D. Find: (a) The magnitude, location, and the plane orientation of the maximum normal stress at section A-A. (b) The magnitude, location, and the plane orientation of the maximum shear stress at section A-A. (c) The magnitude, location, and the plane orientation of the maximum normal stress at section B-B. (d) The magnitude, location, and the plane orientation of the maximum shear stress at section B-B. 4-61 For the bracket of Problem 4-60, solve for the deflection and slope of point C. 4-62 Figure P4-23 shows a 1-in-dia steel bar supported and subjected to the applied load P = 500 lb. Solve for the deflection at the load and the slope at the roller support.

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P4-23 Problem 4-62

4-63 Figure P4-24 shows a 1.25-in-dia solid steel shaft with several twisting couples applied in the directions shown. For TA = 10 000 lb-in, TB = 20 000 lb-in, TC = 30 000 lb-in, find: (a) The magnitude and location of the maximum torsional shear stress in the shaft. (b) The corresponding principal stresses for the location determined in part (a). (c) The magnitude and location of the maximum shear strain in the shaft.

Chapter 4

TA

STRESS, STRAIN, AND DEFLECTION

TB

TC

273

TD

l l 2 l 4

B

A 18"

C 12"

F

D 10"

B FIGURE P4-24

Copyright © 2018 Robert L. Norton: All Rights Reserved

Problems 4-63 and 4-64

d A

4-64 If the shaft of Problem 4-63 were rigidly attached to fixed supports at each end (A and D) and loaded only by the applied couples TB, and TC, then find: (a) The reactions TA and TD at each end of the shaft. (b) The rotation of section B with respect to section C. (c) The magnitude and location of the maximum shear strain. †4-65

Copyright © 2018 Robert L. Norton: All Rights Reserved

Figure P4-25 shows a pivot pin that is a press fit in part A and a slip fit in part B. If F = 100 lb and l = 1.5 in, what pin diameter is needed to limit the maximum stress in the pin to 50 kpsi?

†4-66

Figure P4-25 shows a pivot pin that is a press fit in part A and a slip fit in part B. If F = 100 N and l = 64 mm, what pin diameter is needed to limit the maximum stress in the pin to 250 MPa?

†4-67

Figure P4-25 shows a pivot pin that is a press fit in part A and a slip fit in part B. Determine the l/d ratio that will make the pin equally strong in shear and bending if the shear strength is equal to one-half the bending strength.

75.4

FIGURE P4-25 Problems 4-65 to 4-67



Problem numbers in italics are design problems.

6.4

28.4 (b)

(c)

31.8

3.2 typ.

63.5 A 9.6 (a)

31.8

9.6 (d)

(e)

all dimensions in mm

31.8 3.2

A FIGURE P4-26 Problems 4-69 to 4-73 (A Solidworks model of this is on the book’s website)

Sections A-A Copyright © 2018 Robert L. Norton: All Rights Reserved

274



Problems with numbers in italics are design problems.

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

†4-68

Write a computer program in any language, or use an equation-solver program or spreadsheet to calculate and plot the variation in cross-sectional area, area moment of inertia, radius of gyration, slenderness ratio, and critical load with respect to the inside diameter of both a Euler and a Johnson column having round-hollow cross sections. Assume that the outside diameter of each column is 1 in. The effective length of the Euler column is 50 in. The effective length of the Johnson column is 10 in. Both are made of steel with Sy = 36 000 psi. Let the inside diameter vary from 10% to 90% of the outside diameter for this parametric study. Comment on the advantages of hollow-round columns over solid-round columns of each type (Euler and Johnson) that have the same outer diameters, respective lengths, and materials.

*4-69

Figure P4-26a shows a C-clamp with an elliptical body dimensioned as shown. The clamp has a T-section with a uniform thickness of 3.2 mm at the throat as shown in Figure P4-26b. Find the bending stress at the inner and outer fibers of the throat if the clamp force is 2.7 kN.

4-70 A C-clamp as in Figure P4-26a has a rectangular cross section as in Figure P4-26c. Find the bending stress at the inner and outer fibers of the throat if the clamp force is 1.6 kN. * Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems.

4-71 A C-clamp as in Figure P4-26a has an elliptical cross section as in Figure P4-26d. Dimensions of the major and minor axes of the ellipse are given. Determine the bending stress at the inner and outer fibers of the throat if the clamping force is 1.6 kN. 4-72 A C-clamp as in Figure P4-26a has a trapezoidal cross section as in Figure P4-26e. Determine the bending stress at the inner and outer fibers of the throat if the clamping force is 1.6 kN. †4-73

We want to design a C-clamp with a T-section similar to the one shown in Figures P4-26a and b. The depth of the section will be 31.8 mm as shown but the width of the flange (shown as 28.4 mm) is to be determined. Assuming a uniform thickness of 3.2 mm and a factor of safety against static yielding of 2, determine a suitable value for the width of the flange if the C-clamp is to be made from 60-40-18 ductile iron and the maximum design load is 1.6 kN.

4-74 A round steel bar is 10 in long and has a diameter of 1 in. (a) Calculate the stress in the bar when it is subjected to a 1000-lb force in tension. (b) Calculate the bending stress in the bar if it is fixed at one end (as a cantilever beam) and has a 1000-lb transverse load at the other end. (c) Calculate the transverse shear stress in the bar of part (b). (d) Determine how short the bar must be when loaded as a cantilever beam for its maximum flexural bending stress and its maximum transverse shear stress to provide equal tendency to failure. Find the length as a fraction of the diameter if the failure stress in shear is half the failure stress in bending. (e) Calculate the torsional shear stress when a 10 000 in-lb couple is applied about the centerline (axis) of the cantilever beam at its free end. (f) If the force on the cantilever beam in (b) is eccentric, inducing torsional as well as bending stress, what fraction of the diameter would the eccentricity need to be in order to give a torsional stress equal to the transverse shear stress? (g) Calculate the direct shear stress that would result in the bar of (a) if it were the pin in a pin-and-clevis connection as in Figure P4-8 that is subjected to a 1000-lb pull. (h) Calculate the direct bearing stress that would result on the bar of (a) if it were the pin in a pin-and-clevis connection as in Figure P4-8 that is subjected to a 1000-lb pull if the center part (the eye or tongue) is 1-in wide. (i) Calculate the maximum bending stress in the bar if it is formed into a semicircle with a centroidal radius of 10/π in and 1000-lb opposing forces are applied at the ends and in the plane of the ends similar to Figure P4-15. Assume that there is no distortion of the cross section during bending.

Chapter 4

Table P4-4

STRESS, STRAIN, AND DEFLECTION

275

Data for Problems 4-75 through 4-78 Use only data that are relevant to the particular problem Lengths are in mm, forces in N, and moments in N-m

Row

D

d

h

M

P

a

40

20

4

10

80

8000

b

26

20

1

12

100

9500

c

36

30

1.5

8

60

6500

d

33

30

1

8

75

7200

e

21

20

1

10

50

5500

1.5

7

80

8000

5

8

400

15 000

f

51

50

g

101

100

r

Copyright © 2018 Robert L. Norton: All Rights Reserved

*4-75

For a filleted flat bar loaded in tension similar to that shown in Figure C-9 (Appendix C) and the data from the row(s) assigned from Table P4-4, determine the nominal and maximum axial stress in the bar.

4-76 For a filleted flat bar loaded in bending similar to that shown in Figure C-10 (Appendix C) and the data from the row(s) assigned from Table P4-4, determine the nominal and maximum bending stress in the bar. 4-77 For a shaft, with a shoulder fillet, loaded in tension similar to that shown in Figure C-1 (Appendix C) and the data from the row(s) assigned from Table P4-4, determine the nominal and maximum axial stress in the shaft. 4-78 For a shaft, with a shoulder fillet, loaded in bending similar to that shown in Figure C-2 (Appendix C) and the data from the row(s) assigned from Table P4-4, determine the nominal and maximum bending stress in the shaft. 4-79 A differential stress element has a set of applied stresses on it as shown in Figure 4-1. For σx = 850, σy = –200, σz = 300, τxy = 450, τyz = –300, and τzx = 0; find the principal stresses and maximum shear stress and draw the Mohr’s circle diagram for this threedimensional stress state. 4-80 Write expressions for the normalized (stress/pressure) tangential stress as a function of the normalized wall thickness (wall thickness/outside radius) at the inside wall of a thick-wall cylinder and for a thin-wall cylinder, both with internal pressure only. Plot the percent difference between these two expressions and determine the range of the wall thickness to outside radius-ratio for which the stress predicted by the thin-wall expression is at least 5% greater than that predicted by the thick-wall expression. 4-81 A hollow square torsion bar such as that shown in Table 4-3 has dimensions a = 25 mm, t = 3 mm, and l = 300 mm. If it is made of steel with a modulus of rigidity of G = 80.8 GPa, determine the maximum shear stress in the bar and the angular deflection under a torsional load of 500 N-m. 4-82 Design a hollow rectangular torsion bar such as that shown in Table 4-3 that has dimensions a = 45 mm, b = 20 mm, and l = 500 mm. It is made of steel with a shear yield strength of 90 MPa and has an applied torsional load of 135 N-m. Use a factor of safety against yielding of 2. 4-83 A pressure vessel with closed ends has the following dimensions: outside diameter, OD = 450 mm, and wall thickness, t = 6 mm. If the internal pressure is 690 kPa, find the

* Answers

to these problems are provided in Appendix D.

MACHINE DESIGN

276

-

An Integrated Approach -

l

Sixth Edition

l

a

a w

w x

R1

R2 (a)

FIGURE P4-27

R3

x R1

R2

R3 ( b)

Copyright © 2018 Robert L. Norton: All Rights Reserved

Beams and Beam Loadings for Problem 4-85 to 4-86—see Table P4-2 for Data

principal stresses on the inside surface away from the ends. What is the maximum shear stress at the point analyzed? 4-84 A simply supported steel beam of length l with a concentrated load, F, acting at midspan has a rectangular cross-section of width, b, and depth, h. If the strain energy due to transverse shear loading is Us and that due to bending loading is Ub, derive an expression for the ratio Us / Ub and plot it as a function of h / l over the range 0.0 to 0.10. 4-85 A beam is supported and loaded as shown in Figure P4-27a. Find the reactions for the data given in row a from Table P4-2. 4-86 A beam is supported and loaded as shown in Figure P4-27b. Find the reactions for the data given in row a from Table P4-2. a

30°

w A leg t wheel assembly t = 20 mm w = 45 mm a = 75 mm P Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P4-28 Problems 4-87 and 4-88

4-87 Figure P4-28 shows a portion of the landing gear on a light-weight, experimental airplane. The landing gear attaches to the fuselage at A. Find the maximum shear stress on the cross-section of the leg at A if P = 3.5 kN. 4-88 Figure P4-28 shows a portion of the landing gear on a light-weight, experimental airplane. The landing gear attaches to the fuselage at A. Find the principal stresses on the inner and outer edges of the cross-section of the leg at A if P = 5 kN. 4-89 For the foot-pedal and push-rod linkage shown in Figure P4-29 and the data in the row(s) assigned from Table P4-3, determine the angular deflection of the tube between sections B and C. 4-90 For the foot-pedal and push-rod linkage shown in Figure P4-29 and the data in the row(s) assigned from Table P4-3, determine the maximum torsional shear stress in the tube. 4-91 For the foot-pedal and push-rod linkage shown in Figure P4-29 and the data in the row(s) assigned from Table P4-3, determine the maximum bending stress in the tube. 4-92 For the foot-pedal and push-rod linkage shown in Figure P4-29 and the data in the row(s) assigned from Table P4-3, determine the maximum applied and principal stresses in the tube. 4-93 A differential stress element has a set of applied stresses on it as shown in Figure 4-1. For σx = −2100, σy = 1575, σz = 0, τxy = −1100, τyz = 650, and τzx = 800; find the principal stresses and maximum shear stress and draw the Mohr’s circle diagram for this three-dimensional stress state.

Chapter 4

A

a

B

STRESS, STRAIN, AND DEFLECTION

C

l

a

277

D

Z foot pedal

foot pedal

F

10 h

y tube

bearing

bearing z

x

OD

ID link

link push rod

P Z

push rod Section Z-Z

FIGURE P4-29

Copyright © 2018 Robert L. Norton: All Rights Reserved

Problem 4-89 to 4-92—see Table P4-3 for Data

4-94 Figure P4-19 shows a 12-mm-dia, 250-mm-long steel rod subjected to tensile loads applied at each end of the rod, acting along its longitudinal Y axis and through the centroid of its circular cross section. Before the tensile load was applied the points A and B were 100-mm apart. After the load was applied they were 100.03 mm apart. For this bar with its loading, find: (a) All components of the strain tensor matrix (equation 4.3a) for a point midway between A and B. (b) All components of the stress tensor matrix (equation 4.1a) for a point midway between A and B. (c) The magnitude of the tensile load. 4-95 A 15-m-long, vertical pole supports a 500-N antenna at the top. Four steel guy wires are attached to the pole at an elevation of 12 m. The angle between the guy wires and the pole is 30° and they oppose each other in pairs at the point of attachment. The total load on the pole must not exceed 20 kN. What is the maximum tension allowed in each guy wire? If they have a 12-mm-dia, what is the stress in a cross-section far from either end and how much must each wire stretch to achieve the required load? 4-96 A beam is supported and loaded as shown in Figure P4-11a. Using Castigliano’s method find the deflection under the load F for the data given in the assigned row(s) from Table P4-2. 4-97 A beam is supported and loaded as shown in Figure P4-11b. Using Castigliano’s method find the deflection under the load F for the data given in the assigned row(s) from Table P4-2. 4-98 A beam is supported and loaded as shown in Figure P4-11b. Using the Beam Tables in Appendix B find the reactions, maximum shear, maximum moment, and maximum deflection for the data given in the assigned row(s) from Table P4-2. 4-99 A pressure vessel with closed ends has the following dimensions: outside diameter, OD = 4 in, and wall thickness, t = 0.5 in. If the internal pressure is 10 000 psi, find the principal stresses on the inside surface away from the ends. What is the maximum shear stress at the point analyzed?

10 t

278

MACHINE DESIGN

-

An Integrated Approach -

Sixth Edition

4-100 A differential stress element has a set of applied stresses on it as shown in Figure 4-1. For σx = 1880, σy = −1255, σz = 240, τxy = 0, τyz = 1230, and τzx = 940; find the principal stresses and maximum shear stress and draw the Mohr’s circle diagram for this three-dimensional stress state.

STATIC FAILURE THEORIES The whole of science is nothing more than a refinement of everyday thinking. Albert einstein

5.0

INTRODUCTION

Why do parts fail? This is a question that has occupied scientists and engineers for centuries. Much more is understood about various failure mechanisms today than was known even a few decades ago, largely due to improved testing and measuring techniques. If you were asked to respond to the above question based on what you have learned so far, you would probably say something like “Parts fail because their stresses exceed their strength,” and you would be right up to a point. The follow-up question is the critical one; what kind of stresses cause the failure: Tensile? Compressive? Shear? The answer to this one is the classic, “It depends.” It depends on the material in question and its relative strengths in compression, tension, and shear. It also depends on the character of the loading (whether static or dynamic) and on the presence or absence of cracks in the material. Table 5-0 shows the variables used in this chapter and references the equations or sections in which they are used. At the end of the chapter, a summary section is provided that also groups all the significant equations from this chapter for easy reference and identifies the chapter section in which their discussion can be found. Two master lecture videos on this chapter’s topics and a demonstration video are linked. Lecture 5 covers sections 5.1, and Lecture 6 covers sections 5.2–5.5. The demonstration video, Failure Modes, shows the effects of the failure theories in this chapter on test specimens. Figure 5-1a shows the Mohr’s circle for the stress state in a tensile test specimen. The tensile test (see Section 2-1) slowly applies a pure tensile loading to the part and causes a tensile, normal stress. However, the Mohr’s circle shows that a shear stress is also present, which happens to be exactly half as large as the normal stress. Which stress failed the part, the normal stress or the shear stress? 279

5

280

MACHINE DESIGN

Opening page public domain photograph courtesy of the NASA-Lewis Research Center

Table 5-0 Symbol

-

An Integrated Approach

-

Sixth Edition

Variables Used in This Chapter Variable

ips units

SI units

See

a

half-width of crack

in

m

Sect. 5.3

b

half-width of cracked plate

in

m

Sect. 5.3

E

Young's modulus

psi

Pa

Sect. 5.1

K

stress intensity

kpsi-in0.5

MPa-m0.5

Sect. 5.3

Kc

fracture toughness

kpsi-in0.5

MPa-m0.5

Sect. 5.3

N NFM

safety factor

none

none

Sect. 5.1

safety factor for fracture-mechanics failure

none

none

Sect. 5.3

Suc

ultimate compressive strength

psi

Pa

Sect. 5.2

Sut

ultimate tensile strength

psi

Pa

Sect. 5.2

Sy

tensile yield strength

psi

Pa

Eq. 5.8a, 5.9b

Sys

shear yield strength

psi

Pa

Eq. 5.9b, 5.10

U Ud

total strain energy

in-lb

Joules

Eq. 5.1

distortion strain energy

in-lb

Joules

Eq. 5.2

Uh

hydrostatic strain energy

in-lb

Joules

Eq. 5.2

β

stress-intensity geometry factor

none

none

Eq. 5.14c

ε

strain

none

none

Sect. 5.1

ν σ1

Poisson's ratio

none

none

Sect. 5.1

principal stress

psi

Pa

Sect. 5.1

τ

σ2

principal stress

psi

Pa

Sect. 5.1

σ3

principal stress

psi

Pa

Sect. 5.1

σ

σ

Modified-Mohr effective stress

psi

Pa

Eq. 5.12

σ′

von Mises effective stress

psi

Pa

Eq. 5.7

3

τ

σx

σ σ

σ

Figure 5-1b shows the Mohr’s circle for the stress state in a torsion test specimen. The torsion test (see Section 2-1) slowly applies a pure torsion loading to the part and causes a shear stress. However, the Mohr’s circle shows that a normal stress is also present, which happens to be exactly equal to the shear stress. Which stress failed the part, the normal stress or the shear stress?

(a)

τ τxy σ

σ τyx

(b)

FIGURE 5-1 Mohr's Circles for Unidirectional Tensile Stress (a) and Pure Torsion (b)

σ

σ

In general, ductile, isotropic materials in static tensile loading are limited by their shear strengths while brittle materials are limited by their tensile strengths (though there are exceptions to this rule when ductile materials can behave as if brittle). This situation requires that we have different failure theories for the two classes of materials, ductile and brittle. Recall from Chapter 2 that ductility can be defined in several ways, the most common being a material’s percent elongation to fracture, which, if >5%, is considered ductile. Most ductile metals have elongations to fracture >10%. Most importantly, we must carefully define what we mean by failure. A part may fail if it yields and distorts sufficiently to not function properly. Also, a part may fail by fracturing and separating. Either of these conditions is a failure, but the mechanisms causing them can be very different. Only ductile materials may yield significantly before fracturing. Brittle materials proceed to fracture without significant shape change. The

Chapter 5

Stress σ

STATIC FAILURE THEORIES

true

Sut y

Sy

u

f

281

Stress σ

true u

eng'g

y el pl

pl el (a)

f

(b)

eng'g

3 offset line E

E Elastic Range

Plastic Range

Strain ε

Strain ε

0.002

F I G U R E 2 - 2 Repeated Engineering and True Stress-Strain Curves for Ductile Materials: (a) Low-Carbon Steel (b) Annealed High-Carbon Steel

Stress σ

stress-strain curves of each type of material reflect this difference, as shown in Figures 2-2 and 2-4, which are reproduced here for your convenience. Note that if cracks are present in a ductile material, it can suddenly fracture at nominal stress levels well below the yield strength, even under static loads. Another significant factor in failure is the character of the loading, whether it is static or dynamic. Static loads are slowly applied and remain essentially constant with time. Dynamic loads are either suddenly applied (impact loads) or repeatedly varied with time (fatigue loads), or both. The failure mechanisms are quite different in either case. Table 3-1 defined four classes of loading based on the motion of the loaded parts and the time dependence of the loading. By that definition, only Class 1 loading is static. The other three classes are dynamic to a greater or lesser degree. When the loading is dynamic, the distinction between ductile and brittle materials’ failure behavior blurs, and ductile materials fail in a “brittle” manner. Because of the significant differences in failure mechanisms under static and dynamic loading, we will consider them each separately, discussing failures due to static loading in this chapter and failures due to dynamic loading in the next chapter. For the static loading case (Class 1), we will consider the theories of failure separately for each type of material, ductile and brittle. 5.1

FAILURE OF DUCTILE MATERIALS UNDER STATIC LOADING View the Lecture 5 Video (46:02)* View a Demonstration Video (09:41)†

While ductile materials will fracture if statically stressed beyond their ultimate tensile strength, their failure in machine parts is generally considered to occur when they yield under static loading. The yield strength of a ductile material is appreciably less than its ultimate strength. Historically, several theories have been formulated to explain this failure: the maximum normal-stress theory, the maximum normal-strain theory, the total strain-energy theory, the distortion-energy (von Mises-Hencky) theory, and the maximum shear-stress theory. Of these only the last two agree closely with experimental data for this case, and of those, the von Mises-Hencky theory is the most accurate. We will discuss only the last two in detail, starting with the most accurate (and preferred) approach.

Sut

f

Sy

y offset line E

0.002

Strain ε

F I G U R E 2 - 4 Repeated Stress-Strain Curve of a Brittle Material

*

http://www.designofmachinery. com/MD/05_Ductile_Failure_ Theory.mp4 †

http://www.designofmachinery. com/MD/Failure_Modes.mp4

282

MACHINE DESIGN

An Integrated Approach

-

Sixth Edition

σ

The von Mises-Hencky or Distortion-Energy Theory

σi

The microscopic yielding mechanism is now understood to be due to relative sliding of the material’s atoms within their lattice structure. This sliding is caused by shear stress and is accompanied by distortion of the shape of the part. The energy stored in the part from this distortion is an indicator of the magnitude of the shear stress present.

E

3

-

U εi

ε

FIGURE 5-2

ToTal STrain EnErgy It was once thought that the total strain energy stored in the material was the cause of yield failure, but experimental evidence did not bear this out. The strain energy U in a unit volume (strain energy density) associated with any stress is the area under the stress-strain curve up to the point of the applied stress, as shown in Figure 5-2 for a unidirectional stress state. Assuming that the stress-strain curve is essentially linear up to the yield point, then we can express the total strain energy in a unit volume at any point in that range as

Internal Strain Energy Density in a Deflected Part

U=

1 σε 2

(5.1a)

Extending this to a three-dimensional stress state gives U=

1 (σ1ε1 + σ 2 ε2 + σ 3 ε3 ) 2

(5.1b)

using the principal stresses and principal strains that act on planes of zero shear stress. This expression can be put in terms of principal stresses alone by substituting the relationships 1 (σ1 − νσ 2 − νσ 3 ) E 1 ε 2 = ( σ 2 − νσ1 − νσ 3 ) E 1 ε3 = ( σ 3 − νσ1 − νσ 2 ) E ε1 =

(5.1c)

where ν is Poisson’s ratio, giving U=

1 σ12 + σ 22 + σ 32 − 2 ν ( σ1σ 2 + σ 2 σ 3 + σ1σ 3 )   2E 

(5.1d )

HydroSTaTic loading Very large amounts of strain energy can be stored in materials without failure if they are hydrostatically loaded to create stresses that are uniform in all directions. This can be done in compression very easily by placing the specimen in a pressure chamber. Many experiments have shown that materials can be hydrostatically stressed to levels well beyond their ultimate strengths in compression without failure, as this just reduces the volume of the specimen without changing its shape. P. W. Bridgman subjected water ice to 1 Mpsi hydrostatic compression with no failure. The explanation is that uniform stresses in all directions, while creating volume change and potentially large strain energies, cause no distortion of the part and thus no shear stress. Consider the Mohr’s circle for a specimen subjected to σx = σy = σz = 1 Mpsi compressive stress. The Mohr’s “circle” is a point on the σ axis at –1 Mpsi and σ1 = σ2 = σ3. The shear stress is zero, so there is no distortion and no failure. This is true for ductile or brittle materials when the principal stresses are identical in magnitude and sign.

Chapter 5

STATIC FAILURE THEORIES

Den Hartog[1] describes the condition of rocks at great depth in the earth’s crust where they withstand uniform, hydrostatic compressive stresses of 5500 psi/mile of depth due to the weight of the rock above. This is in excess of their typical 3000 psi ultimate compressive strength as measured in a compression test. While it is much more difficult to create hydrostatic tension, Den Hartog[1] also describes such an experiment done by the Russian scientist Joffe in which he slowly cooled a glass marble in liquid air, allowed it to equilibrate to a stress-free state at the low temperature, then removed it to a warm room. As the marble warmed from the outside in, the temperature differential versus its cold core created uniform tensile stresses calculated to be well in excess of the material’s tensile strength, but it did not crack. Thus, it appears that distortion is the culprit in tensile failure as well. componEnTS of STrain EnErgy The total strain energy in a loaded part (Eq. 5.1d) can be considered to consist of two components—one due to hydrostatic loading, which changes its volume, and one due to distortion, which changes its shape. If we separate the two components, the distortion-energy portion will give a measure of the shear stress present. Let Uh represent the hydrostatic or volumetric component and Ud the distortion-energy component, then U = Uh + Ud

(5.2)

We can also express each of the principal stresses in terms of a hydrostatic (or volumetric) component σh that is common to each face and a distortion component σιd that is unique to each face, where the subscript i represents the principal stress direction, 1, 2, or 3: σ1 = σ h + σ1d σ 2 = σ h + σ 2d

(5.3a)

σ 3 = σ h + σ 3d

Adding the three principal stresses in equation 5.3a gives: σ1 + σ 2 + σ 3 = σ h + σ1d + σ h + σ 2d + σ h + σ 3d

(

σ1 + σ 2 + σ 3 = 3σ h + σ1d + σ 2d + σ 3d

(

)

3σ h = σ1 + σ 2 + σ 3 − σ1d + σ 2d + σ 3d

(5.3b)

)

For a volumetric change with no distortion, the term in parentheses in equation 5.3b must be zero, giving an expression for the volumetric or hydrostatic component of stress σh: σh =

σ1 + σ 2 + σ 3 3

(5.3c)

which you will note is merely the average of the three principal stresses. Now, the strain energy Uh associated with the hydrostatic volume change can be found by replacing each principal stress in equation 5.1d with σh: 1  2 σ h + σ 2h + σ 2h − 2 ν ( σ h σ h + σ h σ h + σ h σ h )  2E  1  2 = 3σ h − 2 ν 3σ 2h   2E  − ν 1 2 3( ) 2 Uh = σh E 2 Uh =

( )

(5.4 a)

283

3

284

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

and substituting equation 5.3c: 2 3 (1 − 2 ν )  σ1 + σ 2 + σ 3    E 2 3 1 − 2ν  2 = σ1 + σ 22 + σ 32 + 2 ( σ1σ 2 + σ 2 σ 3 + σ1σ 3 )  6E 

Uh =

3

(5.4 b)

diSTorTion EnErgy The distortion energy Ud is now found by subtracting equation 5.4b from 5.1d in accordance with equation 5.2: Ud = U − Uh

{

}

1 σ12 + σ 22 + σ 32 − 2 ν ( σ1σ 2 + σ 2 σ 3 + σ1σ 3 )   2E  1 − 2ν  2 − σ1 + σ 22 + σ 32 + 2 ( σ1σ 2 + σ 2 σ 3 + σ1σ 3 )  6E  1+ ν  2 σ1 + σ 22 + σ 32 − σ1σ 2 − σ 2 σ 3 − σ1σ 3  Ud = 3E  =

{

} (5.5)

To obtain a failure criterion, we will compare the distortion energy per unit volume given by equation 5.5 to the distortion energy per unit volume present in a tensile test specimen at failure, because the tensile test is our principal source of material-strength data. The failure stress of interest here is the yield strength Sy. The tensile test is a uniaxial stress state where, at yield, σ1 = Sy and σ2 = σ3 = 0. The distortion energy associated with yielding in the tensile test is found by substituting these values in equation 5.5: Ud =

1+ ν 2 Sy 3E

(5.6a)

and the failure criterion is obtained by equating the general expression 5.5 with the specific failure expression 5.6a to get 1+ ν 2 1+ ν  2 σ1 + σ 22 + σ 32 − σ1σ 2 − σ 2 σ 3 − σ1σ 3  S y = Ud = 3E 3E  S y2 = σ12 + σ 22 + σ 32 − σ1σ 2 − σ 2 σ 3 − σ1σ 3

(5.6b)

S y = σ12 + σ 22 + σ 32 − σ1σ 2 − σ 2 σ 3 − σ1σ 3

which applies to the three-dimensional stress state. * Note that this assumption will be consistent with the conventional ordering of principal stresses in the 3-D case (σ1 > σ2 > σ3) only if σ3  σ3, σ2 = 0, then only the first and fourth quadrants of Figure 5-12 need to be drawn, as shown in Figure 5-13, which plots the stresses normalized by N / Sut where N is the safety factor. Figure 5-13 also depicts three plane-stress conditions labeled A, B, and C. Point A represents any stress state in which the two nonzero principal stresses, σ1, σ3 are positive. Failure will occur when the load line OA crosses the failure envelope at A’. The safety factor for this situation can be expressed as N=

Sut σ1

(5.12a)

Chapter 5

STATIC FAILURE THEORIES

297

3

FIGURE 5-12 Biaxial Fracture Data of Gray Cast Iron Compared to Various Failure Criteria (From Fig 7.13, p. 255, in Mechanical Behavior of Materials by N. E. Dowling, Prentice-Hall, Englewood Cliffs, NJ, 1993.)

If the two nonzero principal stresses have opposite sign, then two possibilities exist for failure, as depicted by points B and C in Figure 5-13. The only difference between these two points is the relative values of their two stress components σ1, σ3. The load line OB exits the failure envelope at B’ above the point Sut, –Sut, and the safety factor for this case is given by equation 5.12a above. If the stress state is as depicted by point C, then the intersection of the load line OC and the failure envelope occurs at C’ below point Sut, –Sut. The safety factor can be found by solving for the intersection between the load line OC and the failure line. Write the equations for these lines and solve simultaneously to get the modified Mohr equation: N=

Sut Suc

Suc σ1 − Sut ( σ1 + σ 3 )

(5.12b)

If the stress state is in the fourth quadrant, both equations 5.12a and 5.12b should be checked and the smaller resulting safety factor used. Compare equation 5.12b to the less-accurate equation for the unmodified CoulombMohr theory (which is not recommended for use): N=

Sut Suc Suc σ1 − Sut σ 3

To use the preferred modified Mohr theory of equation 5.12b, it would be convenient to have expressions for an effective stress that would account for all the applied stresses and allow direct comparison to a material-strength property, as was done for ductile materials with the von Mises stress. Dowling[5] develops a set of expressions for this effective stress involving the three principal stresses:*

σ3

N ⋅ stress / Sut Sut, Sut

1.0 0

A’

A

O

σ1

B

–1.0

C

–2.0

B’ Sut, —Sut C’

–3.0 0 —Suc

–4.0 0

1.0

N ⋅ stress / Sut FIGURE 5-13 Modified-Mohr Failure Theory for Brittle Material

* See reference 5 for a complete derivation for both two- and three-dimensional CoulombMohr and modified-Mohr theories and the effective stress.

298

MACHINE DESIGN

-

An Integrated Approach

-

C1 =

 2 Sut − Suc 1 σ1 + σ 2 )   σ1 − σ 2 + ( 2  − Suc 

C2 =

 2 Sut − Suc 1 σ 2 + σ 3 )  σ 2 − σ3 + ( 2  − Suc 

C3 =

 2 Sut − Suc 1  σ 3 − σ1 + ( σ 3 + σ1 )  2  − Suc 

3

Sixth Edition

(5.12c)

The largest of the set of six values (C1, C2, C3, plus the three principal stresses) is the desired effective stress as suggested by Dowling:  = MAX (C1 , C2 , C3 , σ1 , σ 2 , σ 3 ) σ  = 0 if MAX < 0 σ

(5.12d )

where the signed function max denotes the algebraically largest of the six supplied arguments. If all of the arguments are negative, then the effective stress is zero. This modified-Mohr effective stress can now be compared to the ultimate tensile strength of the material to determine a safety factor: N=

Sut  σ

(5.12e)

This approach allows easy computerization of the process.

EXAMPLE 5-2 Failure of Brittle Materials Under Static Loading Problem

Determine the safety factors for the bracket rod shown in Figure 5-9 (repeated on next page) based on the modified-Mohr theory.

Given

The material is class 50 gray cast iron with Sut = 52 500 psi and Suc = –164 000 psi. The rod length l = 6 in and arm a = 8 in. The rod outside diameter d = 1.5 in. Load F = 1000 lb.

Assumptions

The load is static and the assembly is at room temperature. Consider shear due to transverse loading as well as other stresses.

Solution

See Figures 5-9 and 4-33 and Examples 4-9 and 5-1.

1 The rod of Figure 5-9 is loaded both in bending (as a cantilever beam) and in torsion. The largest tensile bending stress will be in the top outer fiber at point A. The largest torsional shear stress will be all around the outer circumference of the rod. First take a differential element at point A where both of these stresses combine. Find the normal bending stress and torsional shear stress on point A using equations 4.11b and 4.23b: σx =

Mc 1000 ( 6 )( 0.75) = = 18 108 psi I 0.249

(a)

Chapter 5

STATIC FAILURE THEORIES

299

y l A

wall

d

rod

B

z

arm

3 x

F

A y

a

F I G U R E 5 - 9 Repeated

z

Bracket for Examples 5-1 and 5-2

B

τ xz =

Tr 1000 (8 )( 0.75) = = 12 072 psi J 0.497

T

(b)

2 Find the maximum shear stress and principal stresses that result from this combination of applied stresses using equations 4.6:

(a) Two points of interest for stress calculations

2 2  18 108 − 0   σ − σz  τ max =  x + τ 2xz =  + 12 0722 = 15 090 psi      2  2

σ1 =

σx + σz

σ2 = 0

2

+ τ max =

18 108 2

F

y

+ 15 090 = 24 144 psi

x z

(c)

τxz

Tr/J

18 108 σ + σz σ3 = x − τ max = − 15 090 = −6036 psi 2 2

3 The principal stresses for point A can now be plotted on a modified-Mohr diagram as shown in Figure 5-14a. This shows that the load line crosses the failure envelope above the Sut,–Sut point, making equation 5.12a appropriate for the safety-factor calculation: N=

Sut 52 400 = = 2.2 σ1 24 144

=

y x z

 2 Sut − Suc 1 σ1 + σ 2 )   σ1 − σ 2 + ( 2  − Suc 

(

)

V A

2 52 500 − 164 000 1  24 144 − 0 + 24 144 + 0 −164 000 2 

(



) = 16 415 psi 

(e)

)

 2 52 500 − 164 000 1 =  0 − ( −6036 ) + ( 0 − 6036) = 1932 psi −164 000 2  

Tr/J (c) Stress element at point B Repeated

 2 Sut − Suc 1 C2 =  σ 2 − σ 3 + (σ 2 + σ 3 ) 2  − Suc 

(

Mc/I

(b) Stress element at point A

(d )

4 An alternative approach that does not require drawing the modified-Mohr diagram is to find the Dowling factors C1, C2, C3 using equation 5.12c: C1 =

σx

τzx

Note that these stresses are identical to those of Example 5-1.

FIGURE 4-33 (f)

Stress Elements at Points A and B within Cross Section of Rod for Example 4-10

300

MACHINE DESIGN

C3 = σ3

3

52.5 kpsi

= σ1

0

-

An Integrated Approach

-

Sixth Edition

 2 Sut − Suc 1 σ 3 + σ1 )   σ 3 − σ1 + ( 2  − Suc 

(

)

2 52 500 − 164 000 1  24 144 − ( −6036 ) + 24 144 − 6036 −164 000 2 

(



) = 18 348

( g)



5 Then find the largest of the six stresses C1, C2, C3, σ1, σ2, σ3:  = MAX (C1 , C2 , C3 , σ1 , σ 2 , σ 3 ) σ

A

(

(h)

)

 = MAX 16 415, 1932, 18 348, 24 144, 0, − 6036 = 24 144 σ –52.5

which is the modified-Mohr effective stress. 6 The safety factor for point A can now be found using equation 5.12e: N=

Sut 52 500 = = 2.2  σ 24 144

(i )

which is the same as was found in step 3. –164 0

7 Since the rod is a short beam, we need to check the shear stress due to transverse loading at point B on the neutral axis. The maximum transverse shear stress at the neutral axis of a solid round rod was given as equation 4.15c:

24.1 52.5 kpsi

(a) Stresses at point A

τ bending =

σ3 52.5 kpsi σ1 B

–12.8

4V 4 (1000 ) = = 755 psi 3A 1.767

( j)

Point B is in pure shear. The total shear stress at point B is the algebraic sum of the transverse shear stress and the torsional shear stress, which both act on the same planes of the differential element and, in this case, act in the same direction as shown in Figure 4-33b. τ max = τtorsion + τ bending = 12 072 + 755 = 12 827 psi

(k )

8 Find the principal stresses for this pure shear loading: –52.5

σ1 = τ max = 12 827 psi σ2 = 0

(l )

σ 3 = −τ max = −12 827 psi

9 These principal stresses for point B can now be plotted on a modified-Mohr diagram as shown in Figure 5-14b. Because this is a pure shear loading, the load line crosses the failure envelope at the Sut’–Sut point, making equation 5.12a appropriate for the safety-factor calculation:

–164 0 12.8

(b) Stresses at point B

FIGURE 5-14 Example 5-2

N=

52.5 kpsi

Sut 52 400 = = 4.1 σ1 12 827

(m)

10 To avoid drawing the modified-Mohr diagram, find the Dowling factors C1, C2, C3 using equations 5.12c: C1 = 8721 psi;

C2 = 4106 psi;

C3 = 12 827 psi;

(n)

Chapter 5

STATIC FAILURE THEORIES

301

11 And find the largest of the six stresses C1, C2, C3, σ1, σ2, σ3:  = 12 827 psi σ

P

(o)

which is the modified-Mohr effective stress. 12 The safety factor for point B can now be found using equation 5.12e: 52 500 S N = ut = = 4.1  σ 12 827

( p)

and it is the same as was found in step 9.

c

13 The files EX05-02 are on the book’s website.

5.3

P

FRACTURE MECHANICS

The static failure theories discussed so far have all assumed that the material is perfectly homogeneous and isotropic, and thus free of any defects such as cracks, voids, or inclusions, which could serve as stress-raisers. This is seldom true for real materials. Actually, all materials are considered to contain microcracks too small to be seen with the naked eye. Dolan[6] says that “... every structure contains small flaws whose size and distribution are dependent upon the material and its processing. These may vary from nonmetallic inclusions and microvoids to weld defects, grinding cracks, quench cracks, surface laps, etc.” Scratches or gouges in the surface due to mishandling can also serve as incipient cracks. Functional geometric contours that are designed into the part may raise local stresses in predictable ways and can be taken into account in the stress calculations as was discussed in Chapter 4 (and will be further discussed in the next chapter). Cracks that occur spontaneously in service, due to damage or material flaws, are more difficult to predict and account for. The presence of a sharp crack in a stress field creates stress concentrations that theoretically approach infinity. See Figure 4-35 and equation 4.32a, which are repeated here for your convenience:  a Kt = 1 + 2    c

3

a

(4.32a)

Kt

10 9 8 7 6 5 4 3 2 1 0 0 2 4

6 8 10 c/a

F I G U R E 4 - 3 5 Repeated Stress Concentration at the Edge of an Elliptical Hole in a Plate

Note that when the value of c approaches zero, the stress concentration, and thus the stress, approaches infinity. Since no material can sustain such high stresses, local yielding (for ductile materials), local microfracture (for brittle materials), or local crazing (for polymers) will occur at the crack tip[7]. If stresses are high enough at the tip of a crack of sufficient size, a sudden, “brittle-like” failure can result, even in ductile materials under static loads. The science of fracture mechanics has been developed to explain and predict this sudden-failure phenomenon. Cracks commonly occur in welded structures, bridges, ships, aircraft, land vehicles, pressure vessels, etc. Many catastrophic failures of tankers and Liberty Ships built during World War II occurred.* [8], [9] Twelve of these failures occurred shortly after the ships were launched and before they had sailed anywhere. They simply split in half while tied to the pier. One such ship is shown in Figure 5-15. The hull material was welded,

* “Nearly 80 ships broke in two, and almost 1000 were found to have deck plates with long brittle fractures.” D. A. Canonico, “Adjusting the Boiler Code,” Mechanical Engineering, Feb. 2000, p.56.

302

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

3

FIGURE 5-15 WW II Tanker Cracked in Two While Berthed Prior to Being Placed in Service, Portland, Oregon, January 16, 1943 (Public Domain—Courtesy of the Ship Structures Committee, U.S. Government)

ductile steel and the ship had not yet been dynamically loaded to any significant degree. The nominal stresses were well below the material’s yield strength. Other examples of sudden failure at stresses below yield strength have occurred in this century, such as the Boston molasses-tank rupture in January 1919, which drowned 21 people and many horses under 2.3 million gallons of the sticky liquid.[10] A more recent example is the failure of a 22-ft-dia rocket-motor case while being pressure tested by the manufacturer. Figure 5-16 shows the pieces of the rocket case after failure. It “… was designed to stand proof pressures of 960 psi (but) failed … at 542 psi.”[11] These and other sudden “brittle-like” failures of ductile materials under static loading led researchers to seek better failure theories, since the ones then available could not adequately explain the observed phenomena. Where human life is at risk, as in bridges, aircraft, etc., periodic structural-safety inspections for cracks are required by law or government regulation. These inspections can be by X-ray, ultrasonic energy, or just be visual. When cracks are discovered, an engineering judgment must be made whether to repair or replace the flawed part, retire the assembly, or to continue it in service for a further time subject to more frequent inspection. (Many commercial aircraft currently flying contain structural cracks.) These decisions can now be made sensibly through the use of fracture-mechanics theory. Fracture-Mechanics Theory Fracture mechanics presumes the presence of a crack. The stress state in the region of the crack may be one of plane strain or plane stress (see Section 4.4). If the zone of yielding around the crack is small compared to the dimensions of the part, then linear-elastic fracture-mechanics (LEFM) theory is applicable. LEFM assumes that the bulk of the material is behaving according to Hooke’s law. However, if a significant portion of the

Chapter 5

STATIC FAILURE THEORIES

303

3

FIGURE 5-16 Failed Rocket-Motor Case (Public Domain—Courtesy of NASA-Lewis Research Center)

bulk material is in the plastic region of its stress-strain behavior, then a more complicated approach is required than that described here. For the following discussion, we will assume that LEFM applies. modES of crack diSplacEmEnT Depending on the orientation of the loading versus the crack, the applied loads may tend to pull the crack open in tension (Mode I), shear the crack in-plane (Mode II), or shear (tear) it out-of-plane (Mode III) as shown in Figure 5-17. Most of the fracture-mechanics research and testing has been devoted to the tensile loading case (Mode I), and we will limit our discussion to it. STrESS inTEnSiTy facTor k Figure 5-18a shows a plate (not to scale) of width 2b under tension with a through crack of width 2a in the center. The crack is assumed to be sharp at its ends, and b is much larger than a. The crack’s cross section is in the xy plane. An r-θ polar coordinate system is also set up in the xy plane with its origin at the crack tip as shown in Figure 5-18b. From the theory of linear elasticity, for b >> a the stresses around the crack tip, expressed as a function of the polar coordinates, are σx =

3θ  θ θ K cos 1 − sin sin  +  2 2 2  2 πr 

(a) Mode I

(b) Mode II

(5.13a) (c) Mode III

σy =

τ xy =

3θ  θ θ K cos 1 + sin sin  +  2 2 2  2 πr

(5.13b)

θ θ 3θ K sin cos cos + 2 2 2 2 πr

(5.13c)

Copyright © 2018 Robert L. Norton All Rights Reserved

FIGURE 5-17 Three Modes of Crack Displacement

304

MACHINE DESIGN

P

or

-

An Integrated Approach

σz = 0

-

Sixth Edition

for plane stress

(

σz = ν σx + σ y

)

τ yz = τ zx = 0 2a

3

(5.13e)

with higher-order terms of small value omitted. Note that, when the radius r is zero, the xy stresses are infinite, which is consistent with equation 4.32b and Figure 4-36. The stresses diminish rapidly as r increases. The angle θ defines the geometric distribution of the stresses around the crack tip at any radius. The quantity K is called the stress intensity factor. (A subscript can be added to designate the mode I, II, III of loading as in KI, KII, KIII. Since we are dealing only with mode I loading, we will eliminate the subscript and let K = KI.)

a

b 2b

If we take the plane-stress case and compute the von Mises stress σ' from the x, y, and shear components (Eqs. 5.13a, -b, -c), we can plot the distribution of σ' versus θ for any chosen r as shown in Figure 5-19a for r = 1E–5 in and K = 1. The maximum occurs at 71°. If we set θ to that angle and compute the distribution of σ' as a function of r, it looks like Figure 5-19b, which plots r from 1E–5 to 1 in on a log scale.

P (a)

y

The high stresses near the crack tip cause local yielding and create a plastic zone of radius ry as shown (not to scale) in Figure 5-19c. For any radius r and angle θ, the stress state in this plastic zone at the crack tip is directly proportional to the stress intensity factor K. If b >> a, then K can be defined for the center-cracked plate as

r θ

crack tip

x

(b)

FIGURE 5-18 A Through-Crack in a Plate in Tension *

(5.13d )

for plane strain

The nominal stress for a fracture-mechanics analysis is calculated based on the gross cross-sectional area, without any reduction for the crack area. Note that this is different from the procedure used for calculation of nominal stress when using stress-concentration factors in a regular stress analysis. Then, the net cross section is used to find the nominal stress.

K = σ nom π a

a 0.025b

(b) Channel

b

2

h

A95 = 0.05bh

6

Sixth Edition

nonrotating

1

A95 = 0.076 6 d 2

d

-

2

nonrotating A951−1 = 0.10 bt

t

A952 − 2 = 0.05bh, t > 0.025b

1

(c) Solid rectangular

(d) I beam

FIGURE 6-25

Formulas for 95% Stressed Areas of Various Sections Loaded in Bending (Data from: Fig. 7-15, p. 294, Shigley and Mitchell, Mechanical Engineering Design, 4th ed., McGraw-Hill, New York, 1983)

Brinell hardness (HB) 120

160

200

240

280

320

1.0

360

400

440

520

240

260

mirror-polished

0.9

fine-ground or commercially polished

0.8 0.7

mac

hine

0.6 surface factor, Csurf

480

d

0.5

hot-

0.4

rolle

d

as-fo

0.3

rged

0.2 corroded in tap water

0.1

corroded in salt water 0 60

80

FIGURE 6-26

100

120

140

160

180

200

220

tensile strength, Sut (kpsi)

Surface Factors for Various Finishes on Steel (Data from R. C. Juvinall, Stress, Strain, and Strength,

McGraw-Hill, New York, 1967, p. 234)

Chapter 6

Csurf

FATIGUE FAILURE THEORIES

1.0 4 2 1 0.9 16 8 32 63 0.8 125 250 0.7 500 1000 2000 0.6 surface finish Ra (µin) 0.5 0.4 40 60 80 100 120 140 160 180 200 220

369

240

Sut (kpsi) FIGURE 6-27 Surface Factor as a Function of Surface Roughness and Ultimate Tensile Strength (Data from R. C. Johnson, Machine Design, vol. 45, no. 11, 1967, p. 108, Penton Publishing, Cleveland, Ohio)

6

materials are more sensitive to the stress concentrations introduced by surface irregularities. In Figure 6-26 corrosive environments are seen to drastically reduce strength. These surface factors have been developed for steels and should only be applied to aluminum alloys and other ductile metals with the caution that testing of actual parts under realistic loading conditions be done in critical applications. Cast irons can be assigned a Csurf = 1, since their internal discontinuities dwarf the effects of a rough surface. R. C. Johnson[25] provides the data shown in Figure 6-27 that gives more detail for machined and ground surfaces by relating Csurf to tensile strength based on the measured surface-average-roughness Ra in microinches.* If Ra is known for a machined or ground part, Figure 6-27 can be used to determine a suitable surface factor Csurf. The surface-factor curves in Figure 6-26 for hot-rolled, forged, and corroded surfaces should still be used, as they account for decarburization and pitting effects as well as surface roughness. Shigley and Mischke[39] suggest using an exponential equation of the form Csurf ≅ A ( Sut )

b

if Csurf > 1.0, set Csurf = 1.0

(6.7e)

to approximate the surface factor with Sut in either kpsi or MPa. The coefficients A and exponents b for various finishes were determined from data similar to those in Figure 6-26 and are shown in Table 6-3. This approach has the advantage of being computer programmable and eliminates the need to refer to charts such as Figures 6-26 and 6-27. Table 6-3

Coefficients for Surface-Factor Equation 6.7e Some data taken from Shigley, Mischke and Budynas, Mechanical Engineering Design, 7th ed., McGraw-Hill, New York, 2004, p. 329 For Sut in MPa, use

Surface Finish

A

b

For Sut in psi, use

A

b

Ground

1.58

–0.085

2.411

–0.085

Machined or cold-rolled

4.51

–0.265

16.841

–0.265

Hot-rolled As-forged

57.7 272

–0.718

2052.9

–0.718

–0.995

38 545.0

–0.995

*

There are many parameters used to characterize surface roughness, all of which are typically measured by passing a sharp, conical diamond stylus over the surface with controlled force and velocity. The stylus follows and encodes the microscopic contours and stores the surface profile in a computer. A number of statistical analyses are then done on the profile, such as finding the largest peak-to-peak distance (Rt), the average of the 5 highest peaks (Rpm), etc. The most commonly used parameter is Ra (or Aa), which is the arithmetic average of the absolute values of the peak heights and valley depths. It is this parameter that is used in Figure 6-27. See Section 7.1 for more information on surface roughness.

370

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

150 120 alternating stress, 90 σa (kpsi)

unplated

60

plated

30 0

100

101

102

103

104

105

106

107

108

cycles to failure, N FIGURE 6-28

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Effect of Chrome Plating on Fatigue Strength of Steel (Data from C. C. Osgood, Fatigue Design, Pergamon Press, London, 1982)

Note that slightly different values of Csurf will be obtained from equation 6.7e than from the chart in Figure 6-26 due to the different data used to develop equation 6.7e and its factors in Table 6-3. Surface treatments such as electroplating with certain metals can severely reduce the fatigue strength, as shown in Figure 6-28 for chrome plating because electroplating typically creates residual tensile stress in the surface. Plating with soft metals such as cadmium, copper, zinc, lead, and tin appears not to severely compromise fatigue strength. Electroplating with chrome and nickel is generally not recommended for parts stressed in fatigue unless additional surface treatments such as shot peening (see below) are also applied. An exception may be when the part is in a corrosive environment and the corrosion protection afforded by the plating outweighs its strength reduction. The strength lost to plating can be recovered by shot-peening the surface after plating to introduce compressive stresses as shown in Figure 6-29, which shows the advantage of plating before shot peening. Choosing that order of application increases endurance strength above 70 alternating stress, σa (kpsi)

6

nickel plated, then peened

60

unplated, unpeened

50 40

peened, then nickel plated

30 nickel plated, unpeened 20 104

105

106

107

108

cycles to failure, N FIGURE 6-29

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Effect of Nickel Plating and Shot Peening on Fatigue Strength of Steel (Data from Almen and Black, Residual Stress and Fatigue in Metals, McGraw-Hill, 1963, p. 148)

Chapter 6

FATIGUE FAILURE THEORIES

the unpeened material, but peening before plating has the opposite effect. Shot peening and other means of creating residual stresses are addressed in Section 6.8. temperature Fatigue tests are most commonly done at room temperature. The fracture toughness decreases at low temperatures and increases at moderately high temperatures (up to about 350°C). But, the endurance-limit knee in the S-N diagram disappears at high temperatures, making the fatigue strength continue to decline with number of cycles, N. Also, the yield strength declines continuously with temperatures above room ambient and, in some cases, this can cause yielding before fatigue failure. At temperatures above about 50% of the material’s absolute melting temperature, creep becomes a significant factor and the stress-life approach is no longer valid. The strain-life approach can account for the combination of creep and fatigue under high-temperature conditions and should be used in those situations. Several approximate formulas have been proposed to account for the reduction in endurance limit at moderately high temperatures. A temperature factor Ctemp can be defined. Shigley and Mitchell[26] suggest the following: for T ≤ 450°C (840°F ) :

Ctemp = 1

for 450°C < T ≤ 550°C:

Ctemp = 1 − 0.0058 ( T − 450 )

for 840°F < T ≤ 1 020°F:

Ctemp = 1 − 0.0032 ( T − 840 )

(6.7 f )

Note that these criteria are based on data for steels and should not be used for other metals such as Al, Mg, and Cu alloys. reLiaBiLity Many of the reported strength data are mean values. There is considerable scatter in multiple tests of the same material under the same test conditions. Haugen and Wirsching[27] report that the standard deviations of endurance strengths of steels seldom exceed 8% of their mean values. Table 6-4 shows reliability factors for an assumed 8% standard deviation. Note that a 50% reliability has a factor of 1 and the factor reduces as you choose higher reliability. For example, if you wish to have 99.99% probability that your samples meet or exceed the assumed strength, multiply the mean

vacuum alternating stress, σa

air presoak corrosion fatigue 103

104

105 106 cycles to failure, N

FIGURE 6-30

107

108

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Effect of Environment on Fatigue Strength of Steel (Data from Fuchs and Stephens, Metal Fatigue in Engineering, New York, 1980, p. 220)

371

Table 6-4 Reliability Factors for Sd = 0.08 µ Reliability % Creliab 50

1.000

90

0.897

95

0.868

99

0.814

99.9

0.753

99.99

0.702

99.999

0.659

99.9999

0.620

6

MACHINE DESIGN

endurance limit, Se (kpsi)

372

-

An Integrated Approach

-

Sixth Edition

100 80

in air (all steels)

60 (chromium steels)

40

in fresh water

20 0

(carbon and low-alloy steels) 0

40

80

120

160

200

240

tensile strength, Sut (kpsi) FIGURE 6-31

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Effect of Fresh Water on Fatigue Strength of Steel (Data from P. G. Forrest, Fatigue of Metals, Pergamon Press, Oxford, 1962, p. 212)

6

strength value by 0.702. The values in Table 6-4 provide strength-reduction factors Creliab for chosen reliability levels. environment The environment can have significant effects on fatigue strength as is evidenced by the curves for corroded surfaces in Figure 6-26. Figure 6-30 shows schematically the relative effects of various environments on fatigue strength. Note that even room air reduces strength compared to vacuum. The higher the relative humidity and temperature, the larger will be the reduction of strength in air. The presoak line represents parts soaked in a corrosive environment (water or seawater) and then tested in room air. The increased roughness of the corroded surface is thought to be the reason for the loss of strength. The corrosion fatigue line shows a drastic reduction of strength and elimination of the endurance-limit knee. The corrosion-fatigue phenomenon is not yet fully understood but empirical data such as those in Figures 6-30 and 6-31 depict its severity. Figure 6-31 shows the effect of operation in fresh water on the S-N curves of carbon and low-alloy steels. The relationship between Se’ and Sut becomes constant at about 15 kpsi. So low-strength carbon 250

σm = 0

air 3% NaCl solution

200 σa (MPa)

150

30

20

100 Al - 7.5

50

σa (kpsi)

10

Zn - 2.5 Mg

0

0 104

105

106

107

108

cycles to failure, N FIGURE 6-32

Copyright © 2018 Robert L. Norton: All Rights Reserved

The Effect of Saltwater on the Fatigue Strength of Aluminum (Data from Stubbington and Forsyth,

“Some Corrosion-Fatigue Observations on a High-Purity Aluminum-Zinc-Magnesium Alloy and Commercial D.T.D. 683 Alloy," J. of the Inst. of Metals, London, U.K., vol. 90, 1961–1962, pp. 347–354)

Chapter 6

FATIGUE FAILURE THEORIES

373

steel is as good as high-strength carbon steel in this environment. The only steels that retain some strength in water are chromium steels (including stainless steels), since that alloying element provides some corrosion protection. Figure 6-32 shows the effects of saltwater on the fatigue strength of one alloy of aluminum. Only limited data are available for material strengths in severe environments. Thus, it is difficult to define any universal strength-reduction factors for environmental conditions. The best approach is to extensively test all designs and materials in the environment that they will experience. (This is difficult to do for situations in which the long-term effects of low-frequency loading are desired because it will take an unreasonable time to obtain the data.) Based on Figure 6-31, for carbon and low-alloy steels in fresh water, the relationship between Se’ and Sut in equation 6.5a should be modified to Se ' ≅ 15 kpsi (100 MPa )

for carbon steel in fresh water

(6.8)

Presumably, a saltwater environment would be even worse.

6

Calculating the Corrected Fatigue Strength Sf or Corrected Endurance Limit Se The strength-reduction factors can now be applied to the uncorrected endurance limit S e’ or to the uncorrected fatigue strength Sf’ using equation 6.6 to obtain corrected values for design purposes. Creating Estimated S-N Diagrams Equations 6.6 provide information about the material’s strength in the high-cycle region of the S-N diagram. With similar information for the low-cycle region, we can construct an S-N diagram for the particular material and application as shown in Figure 6-33. The bandwidth of interest is the HCF regime from 103 to 106 cycles and beyond. Let the material strength at 103 cycles be called Sm. Test data indicate that the following estimates of Sm are reasonable:[28] bending:

Sm = 0.9 Sut

axial loading:

Sm = 0.75Sut

(6.9)

The estimated S-N diagram can now be drawn on log-log axes as shown in Figure 6-33. The x axis runs from N = 103 to N = 109 cycles or beyond. The appropriate Sm Sn

Sn

(a)

Sm

(b)

Sm Se

Sf N

103 104 105 106 107 108 109 N1 N2

N 103 104 105 106 107 108 109 N1 N2

FIGURE 6-33 Estimated S-N Curves for (a) Materials with Knee, (b) Materials Without Knee

374

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

from equation 6.9 for the type of loading is plotted at N = 103. Note that the correction factors from equation 6.6 are not applied to Sm. If the material exhibits a knee, then the corrected Se from equation 6.6 is plotted at Ne = 106 cycles and a straight line is drawn between Sm and Se. The curve is continued horizontally beyond that point. If the material does not exhibit a knee, then the corrected Sf from equation 6.6 is plotted at the number of cycles for which it was generated (shown at Nf = 5 x 108) and a straight line is drawn between Sm and Sf. This line may be extrapolated beyond that point, but its accuracy is questionable, though probably conservative (see Figure 6-10). The equation of the line from Sm to Se or Sf can be written as S (N ) = a N b

(6.10 a)

log S ( N ) = log a + b log N

(6.10 b)

or

6

where S(N) is the fatigue strength at any N and a, b are constants defined by the boundary conditions. For all cases, the y intercept is S(N) = Sm at N = N1 =103. For the endurancelimit case (Figure 6-33), S(N) = Se at N = N2 = 106. For a material that does not exhibit an endurance-limit knee, the fatigue strength is taken at some number of cycles: S(N) = Sf at N = N2 (Figure 6-33b). Substitute the boundary conditions in equation 6.10b and solve simultaneously for a and b: S  1 b = log  m  z  Se 

where

z = log N1 − log N 2

log ( a ) = log ( Sm ) − b log ( N1 ) = log ( Sm ) − 3b

(6.10c)

Note that N1 is always 1000 cycles and its log10 is 3. For a material with a knee at N2 = 106, z = (3 – 6) = –3 as shown in Table 6-5. This curve is valid only to the knee, beyond which S(N) = Se as shown in Figure 6-33a. For a material with no knee and S(N) = Sf at N = N2 (Figure 6-33b), the values of z corresponding to various values of N2 are easily calculated. For example, the curve in Figure 6-33b shows Sf at N2 = 5E8 cycles. The value for z is then

Table 6-5 z-factors for Eq. 6-10c

1.0E6 5.0E6 1.0E7 5.0E7 1.0E8 5.0E8 1.0E9 5.0E9

–3.000 –3.699 –4.000 –4.699 –5.000 –5.699 –6.000 –6.699

z@5 E 8 = log(1000) − log(5E 8) = 3 − 8.699 = −5.699 S  1 b@5 E 8 =– log  m  5.699  Sf 

(6.10 d )

for S f @ N 2 = 5E 8 cycles

The constants for any other boundary conditions are determined in the same way. Some values of z for a range of values of N2 with N1 set to 103 are shown in Table 6-5. These equations for the S-N curve allow the estimated finite life N to be found for any fully reversed fatigue strength S(N), or the estimated fatigue strength S(N) can be found for any N.

Chapter 6

FATIGUE FAILURE THEORIES

375

EXAMPLE 6-1 Determining Estimated S-N Diagrams for Ferrous Materials Problem

Create an estimated S-N diagram for a steel bar and define its equations. How many cycles of life can be expected if the alternating stress is 100 MPa?

Given

The Sut has been tested at 600 MPa. The bar is 150 mm square and has a hot-rolled finish. The operating temperature is 500°C maximum. The loading will be fully reversed bending.

Assumptions

Infinite life is required and is obtainable since this ductile steel will have an endurance limit. A reliability factor of 99.9% will be used.

Solution

6

1 Since no endurance-limit or fatigue strength information is given, we will estimate Se’ based on the ultimate strength using equation 6.5a. Se ' ≅ 0.5 Sut = 0.5 ( 600 ) = 300 MPa

(a)

2 The loading is bending so the load factor from equation 6.7a is Cload = 1.0

(b)

3 The part is larger than the test specimen and is not round, so an equivalent diameter based on its 95% stressed area (A95) must be determined and used to find the size factor. For a rectangular section in nonrotating bending loading, the A95 area is defined in Figure 6-25c and the equivalent diameter is found from equation 6.7d: A95 = 0.05bh = 0.05 (150 )(150 ) = 1125 mm 2 dequiv =

1125 mm 2 A95 = = 121.2 mm 0.0766 0.0766

(c)

and the size factor is found for this equivalent diameter from equation 6.7b: Csize = 1.189 (121.2 )−0.097 = 0.747

(d )

4 The surface factor is found from equation 6.7e and the data in Table 6-3 for the specified hot-rolled finish. b Csurf = A Sut = 57.7 ( 600 )−0.718 = 0.584

(e)

5 The temperature factor is found from equation 6.7f: Ctemp = 1 − 0.0058 ( T − 450 ) = 1 − 0.0058 ( 500 − 450 ) = 0.71

(f)

6 The reliability factor is taken from Table 6-4 for the desired 99.9% and is Creliab = 0.753

7 The corrected endurance limit Se can now be calculated from equation 6.6:

( g)

376

MACHINE DESIGN

1000 500 300 200

σ (MPa)

100

-

An Integrated Approach

Sut

-

Sixth Edition

Sm

failure point

σa

σa Se

50 30 20 10

100

101

102

103

104

105

106

107

108

109

number of cycles, N FIGURE 6-34

6

S-N Diagram and Alternating Stress Line Showing Failure Point for Example 6-1

Se = Cload Csize Csurf Ctemp Creliab Se ' = 1.0 ( 0.747 )( 0.584 )( 0.71)( 0.753)( 300 )

(h)

Se =70 MPa

8 To create the S-N diagram, we also need a number for the estimated strength Sm at 103 cycles based on equation 6.9 for bending loading: Sm = 0.90 Sut = 0.90 ( 600 ) = 540 MPa

(i )

9 The estimated S-N diagram is shown in Figure 6-34 with the above values of Sm and Se. The expressions of the two lines are found from equations 6.10a through 6.10c assuming that Se begins at 106 cycles: S  1 1  540  b = – log  m  = – log  = −0.295 765  70  3 3  Se 

(

)

log ( a ) = log ( Sm ) − 3b = log [ 540 ] − 3 −0.295 765 :

( j)

a = 4165.707

S ( N ) = a N b = 4165.707 N −0.295 765 MPa

103 ≤ N ≤ 106

S ( N ) = Se = 70 MPa

N > 106

(k )

10 The number of cycles of life for any alternating stress level can now be found from equations (k). For the stated stress level of 100 MPa, we get 100 = 4165.707 N −0.295 765

103 ≤ N ≤ 106

log100 = log 4165.707 − 0.295 765 log N 2 = 3.619 689 − 0.295 765 log N log N =

2 − 3.619 689 −0.295 765

(l )

= 5.476 270

N = 105.476 270 = 3.0 E 5 cycles

Figure 6-34 shows the intersection of the alternating stress line with the failure line at N = 3E5 cycles. 11 The files EX06-01 are on the website.

Chapter 6

FATIGUE FAILURE THEORIES

377

EXAMPLE 6-2 Determining Estimated S-N Diagrams for Nonferrous Materials Problem

Create an estimated S-N diagram for an aluminum bar and define its equations. What is the corrected fatigue strength at 2E7 cycles?

Given

The Sut for this 6061-T6 aluminum has been tested at 45 000 psi. The forged bar is 1.5 in round. The maximum operating temperature is 300°F. The loading is fully reversed torsion.

Assumptions

A reliability factor of 99.0% will be used. The uncorrected fatigue strength will be taken at 5E8 cycles.

Solution

1 Since no fatigue-strength information is given, we will estimate Sf ’ based on the ultimate strength using equation 6.5c: S f ' ≅ 0.4 Sut

(

)

for Sut < 48 kpsi

(a)

S f ' ≅ 0.4 45 000 = 18 000 psi

This value is at N = 5E8 cycles. There is no knee in an aluminum S-N curve. 2 The loading is pure torsion, so the load factor from equation 6.7a is Cload = 1.0

(b)

because the applied torsional stress will be converted to an equivalent von Mises normal stress for comparison to the S-N strength. 3 The part size is greater than the test specimen and it is round, so the size factor can be estimated with equation 6.7b, noting that this relationship is based on steel data:

(

Csize = 0.869 dequiv

)−0.097 = 0.869 (1.5)−0.097 = 0.835

(c)

4 The surface factor is found from equation 6.7e using the data in Table 6-3 for the specified forged finish, again with the caveat that these relationships were developed for steels and may be less accurate for aluminum: b Csurf = A Sut = 39.9 ( 45)−0.995 = 0.904

(d )

5 Equation 6.7f is only for steel so we will assume: Ctemp = 1

(e)

6 The reliability factor is taken from Table 6-4 for the desired 99.0% and is Creliab = 0.814

(f)

7 The corrected fatigue strength Sf at N = 5E8 can be calculated from equation 6.6: S f = Cload Csize Csurf Ctemp Creliab S f '

(

)

= 1.0 ( 0.835)( 0.904 )(1.0 )( 0.814 ) 18 000 = 11 063 psi

( g)

6

378

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

8 To create the S-N diagram, we also need a number for the estimated strength Sm at 103 cycles based on equation 6.9. Note that the bending value is used for torsion:

(

)

Sm = 0.90 Sut = 0.90 45 000 = 40 500 psi

(h)

9 The coefficient and exponent of the corrected S-N line and its equation are found using equations 6.10a through 6.10c. The value of z is taken from Table 6-5 for Sf at 5E8 cycles: b=–

S   40 500  1 1 log  m  = – log   = −0.0989 5.699 5.699  11 063   Sf 

log ( a ) = log ( Sm ) − 3b = log  40 500  − 3 ( −0.0989 ) :

6

(i )

a = 80 193

10 The fatigue strength at the desired life of N = 2E7 cycles can now be found from the equation for the corrected S-N line: S ( N ) = aN b = 80 193 N −0.0989 = 80 193 ( 2e 7 )−0.0989 = 15 209 psi

( j)

S(N) is larger than Sf because it is at a shorter life than the published fatigue strength. 11 Note the order of operations. We first found an uncorrected fatigue strength Sf’ at some “standard” cycle life (N = 5E8), then corrected it for the appropriate factors from equations 6.7. Only then did we create equation 6.10a for the S-N line so that it passes through the corrected Sf at N = 5E8. If we had created equation 6.10a using the uncorrected Sf’, solved it for the desired cycle life (N = 2E7), and then applied the correction factors, we would get a different and incorrect result. Because these are exponential functions, superposition does not hold. 12 The files EX06-02 are on the website.

6.7

NOTCHES AND STRESS CONCENTRATIONS

Notch is a generic term in this context and refers to any geometric contour that disrupts the “force flow” through the part as described in Section 4.15. A notch can be a hole, a groove, a fillet, an abrupt change in cross section, or any disruption to the smooth contours of a part. The notches of concern here are those that are deliberately introduced to obtain engineering features such as O-ring grooves, fillets on shaft steps, fastener holes, etc. It is assumed that the engineer will follow good design practice and keep the radii of these notches as large as possible and reduce stress concentrations as described in Section 4.15. Notches of extremely small radii are poor design practice and, if present, should be treated as cracks and the principles of fracture mechanics used to predict failure (see Sections 5.3 and 6.57). A notch creates a stress concentration that raises the stresses locally and may even cause local yielding. In the discussion of stress concentration in Chapters 4 and 5 where only static loads were being considered, the effects of stress concentrations were only of concern for brittle materials. It was assumed that ductile materials would yield at the local stress concentration and lower the stress to acceptable levels. With dynamic loads, the situation is different, since ductile materials behave as if brittle in fatigue failures. The geometric (theoretical) stress-concentration factors Kt for normal stress and Kts for shear stress were defined and discussed in Section 4.15. (We will refer to them both here as Kt.) These factors give an indication of the degree of stress concentration

Chapter 6

FATIGUE FAILURE THEORIES

379

at a notch having a particular contour and are used as a multiplier on the nominal stress present in the cross section containing the notch (see equation 4.31). Many of these geometric or theoretical stress-concentration factors have been determined for various loadings and part geometries and are published in various references.[30], [31] For dynamic loading, we need to modify the theoretical stress-concentration factor based on the notch sensitivity of the material to obtain a fatigue stress-concentration factor, Kf, which can be applied to the nominal dynamic stresses. Notch Sensitivity Materials have different sensitivity to stress concentrations, which is referred to as the notch sensitivity of the material. In general, the more ductile the material, the less notch sensitive it is. Brittle materials are more notch sensitive. Since ductility and brittleness in metals are roughly related to strength and hardness, low-strength, soft materials tend to be less notch sensitive than high-strength, hard ones. Notch sensitivity is also dependent on the notch radius (which is a measure of notch sharpness). As notch radii approach zero, the notch sensitivity of materials decreases and also approaches zero. This is quite serendipitous, since you will recall from Section 4.15 that the theoretical stress-concentration factor Kt approaches infinity as the notch radius goes to zero. If not for the reduced notch sensitivity of materials at radii approaching zero (i.e., cracks), engineers would be at a loss to design parts able to withstand any nominal stress level when notches are present. Neuber[32] made the first thorough study of notch effects and published an equation for the fatigue stress-concentration factor in 1937. Kuhn[33] later revised Neuber’s equation and experimentally developed data for the Neuber constant (a material property) needed in his equation. Peterson[30] subsequently refined the approach and developed the concept of notch sensitivity q defined as q=

K f −1 Kt − 1

(6.11a)

where Kt is the theoretical (static) stress-concentration factor for the particular geometry and Kf is the fatigue (dynamic) stress-concentration factor. The notch sensitivity q varies between 0 and 1. This equation can be rewritten to solve for Kf : K f = 1 + q ( K t − 1)

(6.11b)

The procedure is to first determine the theoretical stress concentration Kt for the particular geometry and loading, then establish the appropriate notch sensitivity for the chosen material and use them in equation 6.11b to find the dynamic stress-concentration factor Kf. The nominal dynamic stress for any situation is then increased by the factor Kf for tensile stress (Kfs for shear stress) in the same manner as was done for the static case: σ = K f σ nom τ = K fs τ nom

(6.12)

Note in equation 6.11 that when q = 0, Kf = 1 and it does not increase the nominal stress in equation 6.12. When q = 1, Kf = Kt and the entire effect of the geometric stressconcentration factor is felt in equation 6.12.

6

380

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Neuber Constant,

a ( in0.5 )

0.5

0.4

Wrought Aluminums heat treated (-T )

0.3

annealed or strain-hardened (-O and -H )

0.2 Low-alloy Steels loaded in torsion 0.1

loaded in tension

6 0 0

40

80

120

160

200

240

Ultimate Tensile Strength, Sut (kpsi) FIGURE 6-35

Copyright © 2018 Robert L. Norton: All Rights Reserved

Neuber Constants for Steel and Aluminum (Data from P. Kuhn, el al,"NACA Technical Note 2805, 1952 and NACA Technical Note D-1259, 1962)

The notch sensitivity q can also be defined from the Kuhn-Hardrath formula[33] in terms of Neuber’s constant a and the notch radius r, both expressed in inches. Sut kpsi (MPa)

Notch-Sensitivity Factors for Steels (mm)

1.0

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

200 160 140 120 100 80 70 60 50

5.0

0.9 0.8 0.7 q

0.6

Note: For torsional loading, use a curve for an Sut that is 20 kpsi higher than that of the material selected

0.5 0.4 0.3 0.2 0.1

(in)

0

1379 1103 965 827 689 552 483 414 345

0

FIGURE 6-36

0.02 Part 1

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

notch radius, r Copyright © 2018 Robert L. Norton: All Rights Reserved

Notch-Sensitivity Curves for Steels Calculated from Equation 6.13 Using NACA Data from Figure 6-35 and Table 6-6

Chapter 6

FATIGUE FAILURE THEORIES

1

q= 1+

381

(6.13)

a r

Notch-Sensitivity Factors for Heat-Treated Aluminum (-T) (mm)

1.0

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

kpsi (MPa)

0.9 0.8 0.7 q

Sut

90

621

60

414

40

276

0.6

30

207

0.5

20

138

0.4 0.3 0.2 0.1 (in)

0

0

0.02

0.04

0.06

0.08

0.10

0.12

0.14

0.16

0.18

0.20

notch radius, r

(mm)

1.0

0

Notch-Sensitivity Factors for Annealed and Strain-Hardened Aluminum (-O & -H) 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

kpsi (MPa)

0.9

45

310

35

241

0.7

25

172

0.6

20

138

0.5

15

103

0.8

q

Sut

0.4 0.3 0.2 0.1 (in)

0

0

0.02

FIGURE 6-36

Part 2

0.04

0.06

0.08 0.10 0.12 notch radius, r

0.14

0.16

0.18

0.20

Copyright © 2018 Robert L. Norton: All Rights Reserved

Notch-Sensitivity Curves for Aluminums Calculated from Equation 6.13 Using NACA Data from Figure 6-35 and Tables 6-7 and 6-8

6

382

MACHINE DESIGN

Table 6-6 Neuber's Constant for Steels Sut (ksi)

6

a (in0.5)

50 55 60 70 80 90 100 110 120 130 140

0.130 0.118 0.108 0.093 0.080 0.070 0.062 0.055 0.049 0.044 0.039

160 180 200 220 240

0.031 0.024 0.018 0.013 0.009

Table 6-7 Neuber's Constant for Annealed Aluminum Sut (kpsi)

a (in0.5)

10 15 20 25

0.500 0.341 0.264 0.217

30 35 40 45

0.180 0.152 0.126 0.111

Table 6-8 Neuber's Constant for Hardened Aluminum Sut (kpsi)

a (in0.5)

15 20 30

0.475 0.380 0.278

40 50 60 70 80

0.219 0.186 0.162 0.144 0.131

90

0.122

-

An Integrated Approach

-

Sixth Edition

Note that the Neuber constant is defined as the square root of a, not as a, so it is directly substituted in equation 6.13, while the value of r must have its square root taken. A plot of the Neuber constant √a for three types of materials is shown in Figure 6-35, and Tables 6-6 to 6-8 show data taken from that figure. Note in Figure 6-35 that, for torsional loads on steel, the value of √a should be read for Sut 20 kpsi higher than that of the material. Figure 6-36, parts 1 and 2 show sets of notch sensitivity curves for steels and aluminums (respectively) generated with equation 6.13 using the data in Figure 6-35 and Tables 6-6 through 6-8. These curves are for notches whose depth is less than four times the root radius and should be used with caution for deeper notches. The full value of the fatigue stress-concentration factor Kf applies only to the high end of the HCF regime (N = 106 – 109 cycles). Some authors [10], [30], [34] recommend applying a reduced portion of Kf, designated Kf’, to the alternating stress at N = 103 cycles. For high-strength or brittle materials Kf’ is nearly equal to Kf, but for low-strength, ductile materials, Kf’ approaches 1. Others  [35] recommend applying the full value of Kf even at 103 cycles. The latter approach is more conservative since data indicate that the effects of stress concentration are less pronounced at lower N. We will adopt the conservative approach and apply Kf uniformly across the HCF range, since the uncertainties surrounding the estimates of fatigue strength and its collection of modifying factors encourage conservatism.

EXAMPLE 6-3 Determining Fatigue Stress-Concentration Factors Problem

A rectangular, stepped bar similar to that shown in Figure 4-36 is to be loaded in bending. Determine the fatigue stress-concentration factor for the given dimensions.

Given

Using the nomenclature in Figure 4-36, D = 2, d = 1.8, and r = 0.25. The material has Sut = 100 kpsi.

Solution

1 The geometric stress-concentration factor Kt is found from the equation in Figure 4-36: b r Kt = A   d

(a)

where A and b are given in the same figure as a function of the D / d ratio, which is 2 / 1.8 = 1.111. For this ratio, A = 1.014 7 and b = –0.217 9, giving  0.25  K t = 1.014 7   1.8 

−0.217 9

= 1.56

(b)

2 The notch sensitivity q of the material can be found by using the Neuber factor √a from Figure 6-35 and Tables 6-6 to 6-8 in combination with equation 6.13, or by reading q directly from Figure 6-36. We will do the former. The Neuber factor from Table 6-6 for Sut = 100 kpsi is 0.062. Note that this is the square root of a:

Chapter 6

q=

1 a 1+ r

=

FATIGUE FAILURE THEORIES

1 = 0.89 0.062 1+ 0.25

383

(c)

3 The fatigue stress-concentration factor can now be found from equation 6.11b: K f = 1 + q ( K t − 1) = 1 + 0.89 (1.56 − 1) = 1.50

(d )

4 The files EX06-03 are on the website.

6.8

RESIDUAL STRESSES

Residual stress refers to stresses that are “built-in” to an unloaded part. Most parts will have some residual stresses from their manufacturing processes. Any procedure such as forming or heat treatment that creates localized strains above the yield point will leave stresses behind when the strain is removed. Good design requires that the engineer try to tailor the residual stresses to, at a minimum, not create negative effects on the strength and preferably to create positive effects. Fatigue failure is a tensile-stress phenomenon. Figures 6-17 and 6-18 show the beneficial effects of mean compressive stresses on fatigue strength. While the designer has little or no control over the presence or absence of a mean compressive stress in the loading pattern to which the part will be subjected, there are techniques that allow the introduction of compressive residual stresses in a part prior to its being placed in service. Properly done, these compressive residual stresses can make significant improvements in fatigue life. There are several methods for introducing compressive residual stresses: thermal treatments, surface treatments, and mechanical prestressing treatments. Most of them create biaxial compressive stresses at the surface, triaxial compressive stresses below the surface, and triaxial tensile stresses in the core. Since the part is in equilibrium, the compressive stresses near the surface have to be balanced by tensile stresses in the core. If the treatment is overdone, the increased tensile core stresses can cause failure, so a balance must be struck. These treatments have the greatest value when the applied stress distribution due to loading is nonuniform and is maximally tensile at the surface, as in reversed bending. Bending in one direction will have peak tensile stress on one side only and the treatment can then be applied just to that side. Axial-tension loading is uniform across the section and so will not benefit from a nonuniform residual stress pattern, unless there are notches at the surface to cause local increases in tensile stress. Then compressive residual stresses at the surface are very helpful. In fact, regardless of loading, the net tensile stresses at notches will be reduced by adding residual compressive stresses in those locations. Since designed notches are usually at the surface, a treatment can be applied to them. The deliberate introduction of residual compressive stresses is most effective on parts made of high-yield-strength materials. If the yield strength of the material is low, then the residual stresses may not stay long in the part due to later yielding from high applied stresses in service. Faires [36] found that steels with Sy  1.0, set Csurf = 1.0

for T ≤ 450°C (840°F ) :

Ctemp = 1

for 450°C < T ≤ 550°C:

Ctemp = 1 − 0.0058 ( T − 450 )

for 840°F < T ≤ 1 020°F:

Ctemp = 1 − 0.0032 ( T − 840 )

437

(6.7e)

(6.7 f )

Corrected Fatigue Strength Estimates (Section 6.6):

Se = Cload Csize Csurf Ctemp Creliab Se ' S f = Cload Csize Csurf Ctemp Creliab S '

(6.6)

f

Approximate Strength at 1000 Cycles (Section 6.6):

bending:

Sm = 0.9 Sut

axial loading:

Sm = 0.75Sut

(6.9)

S-N Diagram (Section 6.6):

log S ( N ) = log a + b log N S  1 b = log  m  z  Se 

(6.10 b)

z = log N1 − log N 2

where

(6.10c)

log ( a ) = log ( Sm ) − b log ( N1 ) = log ( Sm ) − 3b Notch Sensitivity (Section 6.7):

1

q= 1+

(6.13)

a r

Fatigue Stress-Concentration Factors (Sections 6.7 and 6.11):

K f = 1 + q ( K t − 1)

(6.11b)

if K f σ maxnom < S y then:

K fm = K f

if K f σ maxnom > S y then:

K fm =

if K f σ maxnom − σ minnom > 2 S y then:

K fm = 0

S y − K f σ anom σ mnom

(6.17)

Safety Factor - Fully Reversed Stresses (Section 6.10):

Nf =

Sn σ'

(6.14)

Modified-Goodman Diagram (Section 6.11):



σ 'm σ 'a + =1 S yc S yc

(6.16a)

6

438

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

σ 'a = S f

(6.16b)

σ 'm σ 'a + =1 Sut Sf

(6.16c)

σ 'm σ 'a + =1 Sy Sy

(6.16d )



σ'  Sy 

Safety Factor - Fluctuating σ 'Stresses − a  S y 6.11): m @Q =  1 (Section



Case 1:

Sy  σ 'a  = 1 −  σ σ Sy  ' '  m@ σZ'm  m  = 1 − Sf S   σ 'm @ Q

Nf = σ 'a @ P

6

ut

Case 2:

(

Case 4:

(6.18b)

)

Sut S f 2 − S f σ 'a + Sut σ 'm 2 S f Sut σ' 2 N f = Smf @+R S=ut ' ' σ σ S m@ Z a ut + σ 'm S f Sf =− σ 'm @ S ) + S f ( Sut

σ 'm @ S = σ 'a @ S

 σ 'm   1 − S  ut

Sf σ 'a @ P = σ 'a @ Z σ 'a

Nf = Case 3:

(6.18a)

(6.18e)

( σ ' m − σ ' m @ S )2 + ( σ ' a − σ ' a @ S )2

ZS =

(6.18 f )

( σ ' a )2 + ( σ ' m )2

OZ =

OZ + ZS OZ

Nf =

(6.18 g)

Sines Method for Multiaxial Stresses in Fatigue - 3-D (Section 6.12):

σ 'a =



xa

− σ ya

) + (σ 2

ya

− σ za

) + (σ 2

− σ xa

za

) + 6( τ 2

2 xya

+ τ 2yza + τ 2zxa

)

2

(6.21a)

σ 'm = σ xm + σ ym + σ zm Sines Method for Multiaxial Stresses in Fatigue - 2-D (Section 6.12):

σ 'a = σ 2xa + σ 2ya − σ xa σ ya + 3τ 2xya

(6.21b)

σ 'm = σ xm + σ ym General Multiaxial Stresses in Fatigue - 3-D (Section 6.12):

σ 'a =

σ 'm =



xa

− σ ya

) + (σ 2

ya

− σ za

) + (σ 2

za

− σ xa

) + 6( τ 2

2 xya

+ τ 2yza + τ 2zxa

)

2



xm

− σ ym

) + (σ 2

ym

− σ zm

) + (σ 2

zm

2

− σ xm

) + 6( 2

τ 2xym

+ τ 2yzm

+ τ 2zxm

)

(6.22a)

Chapter 6

FATIGUE FAILURE THEORIES

439

General Multiaxial Stresses in Fatigue - 2-D (Section 6.12):

σ 'a = σ 2xa + σ 2ya − σ xa σ ya + 3τ 2xya

(6.22b)

σ 'm = σ 2xm + σ 2ym − σ xm σ ym + 3τ 2xym SEQA Method for Complex Multiaxial Stresses in Fatigue (Section 6.12): 1

σ SEQA = 2

 3 2 3 2 9 4 2 1 + Q + 1 + Q cos 2φ + Q  2 16  4 

(6.23)

Fracture Mechanics in Fatigue (Section 6.5):

K = βσ max πa − βσ min πa = β πa ( σ max − σ min ) da = A (∆K )n dN

6.16

(6.3b)

(6.4 a)

REFERENCES

1 R. P. Reed, J. H. Smith, and B. W. Christ, The Economic Effects of Fracture in the United States: Part I, Special Pub. 647-1, U. S. Dept. of Commerce, National Bureau of Standards, Washington, D.C., 1983. 2 N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., p. 340, 1993. 3 J. W. Fischer and B. T. Yen, Design, Structural Details, and Discontinuities in Steel, Safety and Reliability of Metal Structures, ASCE, Nov. 2, 1972. 4 N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., p. 355, 1993. 5 D. Broek, The Practical Use of Fracture Mechanics. Kluwer Academic Publishers: Dordrecht, The Netherlands, p. 10, 1988. 6 N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., p. 347, 1993. 7 R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. McGrawHill: New York, p. 280, 1967. 8 J. E. Shigley and C. R. Mischke, Mechanical Engineering Design. 5th ed. McGrawHill: New York, p. 278, 1989. 9 A. F. Madayag, Metal Fatigue: Theory and Design. John Wiley & Sons: New York, p. 117, 1969. 10 N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., p. 418, 1993. 11 R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. McGrawHill: New York, p. 231, 1967.

6

440

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

12 J. A. Bannantine, J. J. Comer, and J. L. Handrock, Fundamentals of Metal Fatigue. Prentice-Hall: Englewood Cliffs, N. J., p. 13, 1990. 13 P. C. Paris and F. Erdogan, A Critical Analysis of Crack Propagation Laws. Trans. ASME, J. Basic Eng., 85(4): p. 528, 1963. 14 J. M. Barsom, Fatigue-Crack Propagation in Steels of Various Yield Strengths. Trans. ASME, J. Eng. Ind., Series B(4): p. 1190, 1971. 15 H. O. Fuchs and R. I. Stephens, Metal Fatigue in Engineering. John Wiley & Sons: New York, p. 88, 1980. 16 J. A. Bannantine, J. J. Comer, and J. L. Handrock, Fundamentals of Metal Fatigue. Prentice-Hall: Englewood Cliffs, N. J., p. 106, 1990. 17 J. M. Barsom and S. T. Rolfe, Fracture and Fatigue Control in Structures, 2nd ed. Prentice-Hall: Englewood Cliffs, N. J., p. 256, 1987.

6

18 H. O. Fuchs and R. I. Stephens, Metal Fatigue in Engineering. John Wiley & Sons: New York, p. 89, 1980. 19 J. M. Barsom and S. T. Rolfe, Fracture and Fatigue Control in Structures. 2nd ed. Prentice-Hall: Englewood Cliffs, N. J., p. 285, 1987. 20 P. G. Forrest, Fatigue of Metals. Pergamon Press: London, 1962. 21 J. E. Shigley and L. D. Mitchell, Mechanical Engineering Design. 4th ed. McGrawHill: New York, p. 293, 1983. 22 R. Kuguel, A Relation Between Theoretical Stress-Concentration Factor and Fatigue Notch Factor Deduced from the Concept of Highly Stressed Volume. Proc. ASTM, 61: pp. 732-748, 1961. 23 R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. McGrawHill: New York, p. 233, 1967. 24 Ibid., p. 234. 25 R. C. Johnson, Machine Design, vol. 45, p. 108, 1973. 26 J. E. Shigley and L. D. Mitchell, Mechanical Engineering Design. 4th ed. McGrawHill: New York, p. 300, 1983. 27 E. B. Haugen and P. H. Wirsching, “Probabilistic Design.” Machine Design, vol. 47, pp. 10-14, 1975. 28 R. C. Juvinall and K. M. Marshek, Fundamentals of Machine Component Design. 2nd ed. John Wiley & Sons: New York, p. 270, 1967. 29 Ibid., p. 267. 30 R. E. Peterson, Stress-Concentration Factors. John Wiley & Sons: New York, 1974. 31 R. J. Roark and W. C. Young, Formulas for Stress and Strain. 5th ed. McGraw-Hill: New York, 1975. 32 H. Neuber, Theory of Notch Stresses. J. W. Edwards Publisher Inc.: Ann Arbor Mich., 1946. 33 P. Kuhn and H. F. Hardrath, An Engineering Method for Estimating Notch-size Effect in Fatigue Tests on Steel. Technical Note 2805, NACA, Washington, D.C., Oct. 1952.

Chapter 6

FATIGUE FAILURE THEORIES

441

34 R. B. Heywood, Designing Against Fatigue. Chapman & Hall Ltd.: London, 1962. 35 R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. McGrawHill: New York, p. 280, 1967. 36 V. M. Faires, Design of Machine Elements. 4th ed. Macmillan: London, p. 162, 1965. 37 R. B. Heywood, Designing Against Fatigue of Metals. Reinhold: New York, p. 272, 1962. 38 H. O. Fuchs and R. I. Stephens, Metal Fatigue in Engineering. John Wiley & Sons: New York, p. 130, 1980. 39 J. E. Shigley and C. R. Mischke, Mechanical Engineering Design. 5th ed. McGrawHill: New York, p. 283, 1989. 40 N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N J., p. 416, 1993. 41 R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. McGrawHill: New York, p. 280, 1967. 42 G. Sines, Failure of Materials under Combined Repeated Stresses Superimposed with Static Stresses, Technical Note 3495, NACA, 1955. 43 F. S. Kelly, A General Fatigue Evaluation Method, Paper 79-PVP-77, ASME, New York, 1979. 44 Y. S. Garud, A New Approach to the Evaluation of Fatigue Under Multiaxial Loadings, in Methods for Predicting Material Life in Fatigue, W. J. Ostergren and J. R. Whitehead, ed. ASME: New York. pp. 249-263, 1979. 45 M. W. Brown and K. J. Miller, A Theory for Fatigue Failure under Multiaxial StressStrain Conditions. Proc. Inst. Mech. Eng., 187(65): pp. 745-755, 1973. 46 J. O. Smith, The Effect of Range of Stress on the Fatigue Strength of Metals. Univ. of Ill., Eng. Exp. Sta. Bull., (334), 1942. 47 J. E. Shigley and L. D. Mitchell, Mechanical Engineering Design. 4th ed. McGrawHill: New York, p. 333, 1983. 48 B. F. Langer, Design of Vessels Involving Fatigue, in Pressure Vessel Engineering, R. W. Nichols, ed. Elsevier: Amsterdam. pp. 59-100, 1971. 49 J. A. Collins, Failure of Materials in Mechanical Design. 2nd ed. J. Wiley & Sons: New York, pp. 238-254, 1993. 50 S. M. Tipton and D. V. Nelson, Fatigue Life Predictions for a Notched Shaft in Combined Bending and Torsion, in Multiaxial Fatigue, K. J. Miller and M. W. Brown, Editors. ASTM: Philadelphia, PA. pp. 514-550, 1985. 51 R. C. Rice, ed. Fatigue Design Handbook. 2nd ed. Soc. of Automotive Engineers: Warrendale, PA. p. 260, 1988. 52 T. Nishihara and M. Kawamoto, The Strength of Metals under Combined Alternating Bending and Torsion with Phase Difference. Memoirs College of Engineering, Kyoto Univ., Japan, 11(85), 1945. 53 Shot Peening of Gears. American Gear Manufacturers Association AGMA 938-A05, 2005.

6

442

MACHINE DESIGN

Table P6-0† Topic/Problem Matrix Sect. 6.4 Fatigue Loads 6-1 Sect. 6.5 Fracture Mechanics 6-51, 6-52, 6-53, 6-72, 6-73, 6-74, 6-82, 6-83 Sect. 6.6 S/N Diagrams 6-2, 6-4, 6-5, 6-54, 6-55, 6-56, 6-57, 6-75, 6-84, 6-88, 6-92

6

Sect. 6.7 Stress Conc. 6-15, 6-58, 6-59, 6-60, 6-63 to 6-66, 6-76, 6-77, 6-78, 6-85, 6-89, 6-93 Sect. 6.10 Fully Reversed 6-6, 6-7, 6-8, 6-20, 6-26, 6-29, 6-33, 6-35, 6-37, 6-46, 6-48, 6-49, 6-50, 6-86, 6-87, 6-90 Sect. 6.11 Fluctuating 6-3, 6-9, 6-10, 6-11, 6-12, 6-13, 6-14, 6-16, 6-17, 6-18, 6-19, 6-21, 6-22, 6-23, 6-24, 6-25, 6-27, 6-28, 6-30, 6-31, 6-32, 6-34, 6-36, 6-38, 6-40, 6-43, 6-44, 6-45, 6-70, 6-79, 6-80, 6-81, 6-91 Sect. 6.12 Multiaxial 6-39, 6-41, 6-42, 6-61, 6-62, 6-67, 6-68, 6-69 †

Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problem numbers in italics are design problems.

6.17

-

An Integrated Approach

-

Sixth Edition

BIBLIOGRAPHY

For more information on fatigue design, see:

J. A. Bannantine, J. J. Comer, and J. L. Handrock, Fundamentals of Metal Fatigue. Prentice-Hall: Englewood Cliffs, N.J., 1990. H. E. Boyer, ed., Atlas of Fatigue Curves. Amer. Soc. for Metals: Metals Park, Ohio. 1986. H. O. Fuchs and R. I. Stephens, Metal Fatigue in Engineering. John Wiley & Sons: New York, 1980. R. C. Juvinall, Engineering Considerations of Stress, Strain and Strength. A. J. McEvily, ed. Atlas of Stress-Corrosion and Corrosion Fatigue Curves. Amer. Soc. for Metals: Metals Park, Ohio. 1990. McGraw-Hill: New York, 1967. For more information on the strain-life approach to low-cycle fatigue, see:

N. E. Dowling, Mechanical Behavior of Materials. Prentice-Hall: Englewood Cliffs, N.J., 1993. R. C. Rice, ed. Fatigue Design Handbook. 2nd ed. Soc. of Automotive Engineers: Warrendale, PA. 1988. For more information on the fracture mechanics approach to fatigue, see:

J. M. Barsom and S. T. Rolfe, Fracture and Fatigue Control in Structures. 2nd ed. Prentice-Hall: Englewood Cliffs, N.J., 1987. D. Broek, The Practical Use of Fracture Mechanics. Kluwer Academic Publishers: Dordrecht, The Netherlands, 1988. For more information on residual stresses, see:

J. O. Almen and P. H. Black, Residual Stresses and Fatigue in Metals. McGraw-Hill: New York, 1963. For more information on multiaxial stresses in fatigue, see:

A. Fatemi and D. F. Socie, A Critical Plane Approach to Multiaxial Fatigue Damage Including Out-of-Phase Loading. Fatigue and Fracture of Engineering Materials and Structures, 11(3): pp. 149-165, 1988. Y. S. Garud, Multiaxial Fatigue: A Survey of the State of the Art. J. Test. Eval., 9(3), 1981. K. F. Kussmaul, D. L. McDiarmid and D. F. Socie, eds. Fatigue Under Biaxial and Multiaxial Loading. Mechanical Engineering Publications Ltd.: London. 1991. G. E. Leese and D. Socie, eds. Multiaxial Fatigue: Analysis and Experiments. Soc. of Automotive Engineers: Warrendale, Pa., 1989. K. J. Miller and M. W. Brown, eds. Multiaxial Fatigue. Vol. STP 853. ASTM: Philadelphia, Pa., 1985. G. Sines, Behavior of Metals Under Complex Static and Alternating Stresses, in Metal Fatigue, G. Sines and J. L. Waisman, eds. McGraw-Hill: New York. 1959. R. M. Wetzel, ed., Fatigue Under Complex Loading: Analyses and Experiments. SAE Pub. No. AE-6, Soc. of Automotive Engineers: Warrendale, Pa., 1977.

Chapter 6

6.18 *6-1

FATIGUE FAILURE THEORIES

PROBLEMS For the data in the row(s) assigned in Table P6-1, find the stress range, alternating stress component, mean stress component, stress ratio, and amplitude ratio.

6-2 For the steel-material strength data in the row(s) assigned in Table P6-2, calculate the uncorrected endurance limit and draw a strength-life (S-N) diagram. *6-3

For the bicycle pedal arm assembly in Figure P6-1 assume a rider-applied force that ranges from 0 to 1 500 N at the pedal each cycle. Determine the fluctuating stresses in the 15-mm-dia pedal arm. Find the fatigue safety factor if Sut = 500 MPa.

6-4 For the aluminum-material strength data in the row(s) assigned in Table P6-2, calculate the uncorrected fatigue strength at 5E8 cycles and draw a strength-life (S-N) diagram for the material.

*6-7 *6-8

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

Table P6-1 Data for Problem 6-1

σ max

σ min

a

1000

0

b

1000

–1000

c

1500

500

d

1500

–500

e

500

–1000

Design the wrist pin of Problem 3-7 for infinite life with a safety factor of 1.5 if the 2500-g acceleration is fully reversed and Sut = 130 kpsi.

f

2500

–1200

g

0

–4500

A paper machine processes rolls of paper having a density of 984 kg/m3.

h

2500

1500

For the trailer hitch from Problem 3-6 (also see Figure P6-2 and Figure 1-5), find the infinite-life fatigue safety factors for all modes of failure, assuming that the horizontal impact force of the trailer on the ball is fully reversed. Use steel with Sut = 600 MPa and Sy = 450 MPa.

The paper roll is 1.50-m outside dia (OD) x 0.22-m inside dia (ID) x 3.23 m long and is on a simply supported, hollow, steel shaft with Sut = 400 MPa. Find the shaft ID needed to obtain a dynamic safety factor of 2 for a 10-year life running three 8-hour shifts per day if the shaft OD is 22 cm and the roll turns at 50 rpm.

6-9 For the ViseGrip® plier-wrench drawn to scale in Figure P6-3, and for which the forces were analyzed in Problem 3-9 and stresses in Problem 4-9, find the safety factors for

Table P6-3 Row

* Answers

Row

6-5 For the data in the row(s) assigned in Table P6-3, find the corrected endurance strength (or limit), create equations for the S-N line, and draw the S-N diagram. *6-6

443

Data for Problem 6-5

Material

Sut kpsi

Shape

Size inches

Surface Finish

Loading

Temp °F

Reliability

a

steel

110

round

2

ground

torsion

b

steel

90

square

4

mach.

axial

c

steel

80

I-beam

16 x 18*

hot rolled

bending

d

steel

200

round

5

forged

torsion

e

steel

150

square

7

cold rolled

axial

f

aluminum

70

round

9

mach.

bending

ground

torsion

room

99.9

cold rolled

axial

room

99.0

900

50

room

90

round

4

ground

bending

room

99.99

40

square

6

forged

torsion

room

99.999

k

ductile iron

70

round

5

cast

axial

room

50

room

90

50

90

bronze

80

square

6

cast

axial

alum.

alum.

60

n

steel

25 000

35 000

99.999

aluminum

torsion

150 000

e

h

–50

aluminum

bending

d

alum.

j

forged

steel

99.99

i

cast

steel

120 000

room

9

9

250 000

c

alum.

24 x 36*

7

b

40 000

I-beam

round

steel

70 000

square

square

90 000

g

85

60

a

f

50

90

mat'l

99.0

aluminum

bronze

sut (psi)

99.9

aluminum

ductile iron

Row

600

g

l

Data for Problems 6-2, 6-4

room

h

m

Table P6-2

212

99.999

6

444

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

60 mm

P F

170 mm

F

P F

6

0.5-cm grid

FIGURE P6-3

T

Problem 6-9 (A Solidworks model of this is on the website)

FIGURE P6-1 Problem 6-3 (A Solidworks

each pin for an assumed clamping force of P = 4000 N in the position shown. The steel pins are 8-mm dia with Sy = 400 MPa, Sut = 520 MPa, and are all in double shear. Assume a desired finite life of 5E4 cycles.

model is on the book’s website)

40 mm *6-10

An overhung diving board is shown in Figure P6-4a. A 100-kg person is standing on the free end. Assume cross-sectional dimensions of 305 mm x 32 mm. What is the fatigue safety factor for finite life if the material is brittle fiberglass with Sf = 39 MPa @ N = 5E8 cycles and Sut = 130 MPa in the longitudinal direction?

*6-11

Repeat Problem 6-10 assuming the 100-kg person in Problem 6-10 jumps up 25 cm and lands back on the board. Assume the board weighs 29 kg and deflects 13.1 cm statically when the person stands on it. What is the fatigue safety factor for finite life if the material is brittle fiberglass with Sf = 39 MPa @ N = 5E8 cycles and Sut = 100 MPa in the longitudinal direction?

FIGURE P6-2 Problem 6-6 (A Solidworks model is on the book’s website)

6-12 Repeat Problem 6-10 using the cantilevered diving board design in Figure P6-4b. 6-13 Repeat Problem 6-11 using the diving board design shown in Figure P6-4b. Assume the board weighs 19 kg and deflects 8.5 cm statically when the person stands on it.

* Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

6-14 Figure P6-5 shows a child’s toy called a pogo stick. The child stands on the pads, applying half her weight on each side. She jumps up off the ground, holding the pads up against her feet, and bounces along with the spring cushioning the impact and storing energy to help each rebound. Assume a 60-lb child and a spring constant of 100 lb/in. The pogo stick weighs 5 lb. Design the aluminum cantilever beam sections on which she stands to survive jumping 2 in off the ground with a dynamic safety factor of 2 for a finite life of 5E4 cycles. Use 2000 series aluminum. Define the beam shape and size. 2m

0.7 m

(a) Overhung diving board

FIGURE P6-4 Problems 6-10 through 6-13

2m P

0.7 m

(b) Cantilevered diving board

P

Chapter 6

Table P6-4

FATIGUE FAILURE THEORIES

445

Data For Problem 6-15 (kpsi)

Row

(in)

Material

Loading

a

100

3.3

0.25

steel

bending

b

90

2.5

0.55

steel

torsion

c

150

1.8

0.75

steel

bending

d

200

2.8

1.22

steel

torsion

e

20

3.1

0.25

soft aluminum

bending

f

35

2.5

0.28

soft aluminum

bending

g

45

1.8

0.50

soft aluminum

bending

h

50

2.8

0.75

hard aluminum

bending

i

30

3.5

1.25

hard aluminum

bending

j

90

2.5

0.25

hard aluminum

bending

6 W/2

*6-15

W/2

For a notched part having notch dimension r, geometric stress-concentration factor Kt, material strength Sut, and loading as shown in the assigned row(s) of Table P6-4, find the Neuber factor a, the material’s notch sensitivity q, and the fatigue stress-concentration factor Kf.

6-16 A track to guide bowling balls is designed with two round rods as shown in Figure P6-6. The rods have a small angle between them. The balls roll on the rods until they fall between them and drop onto another track. Each rod’s unsupported length is 30 in and the angle between them is 3.2°. The balls are 4.5-in-dia and weigh 2.5 lb. The center distance between the 1-in-dia rods is 4.2 in at the narrow end. Find the infinite-life safety factor for the 1-in-dia SAE 1010 cold-rolled steel rods. (a) Assume rods are simply supported at each end. (b) Assume rods are fixed at each end. *6-17

A pair of ice tongs is shown in Figure P6-7. The ice weighs 50 lb and is 10-in wide across the tongs. The distance between the handles is 4 in, and the mean radius r of a steel tong is 6 in. The rectangular cross-sectional dimensions are 0.750 in x 0.312 in. Find the safety factor for the tongs for 5E5 cycles if their Sut = 50 kpsi.

6-18 Repeat Problem 6-17 with the tongs made of Class 40 gray cast iron.

FIGURE P6-6 Problem 6-16

P FIGURE P6-5 Problem 6-14 * Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

446

MACHINE DESIGN

F F

A

An Integrated Approach

-

Sixth Edition

Determine the size of the clevis pin shown in Figure P6-8 needed to withstand an applied repeated force of 0 to 130 000 lb for infinite life. Also determine the required outside radius of the clevis end to not fail in either tearout or bearing if the clevis flanges each are 2.5 in thick. Use a safety factor of 3. Assume Sut = 140 kpsi for the pin and Sut = 80 kpsi for the clevis.

6-20 A ±100 N-m torque is applied to a 1-m-long, solid, round steel shaft. Design it to limit its angular deflection to 2° and select a steel alloy to have a fatigue safety factor of 2 for infinite life.

r

W

6

*6-19

-

FIGURE P6-7

6-21 Figure P6-9 shows an automobile wheel with two styles of lug wrench, a single-ended wrench in (a), and a double-ended wrench in (b). The distance between points A and B is 1 ft in both cases and the handle diameter is 0.625 in. How many cycles of tightening can be expected before a fatigue failure if the average tightening torque is 100 ft-lb and the material Sut = 60 kpsi? *6-22

Problem 6-17

P

An in-line “roller-blade” skate is shown in Figure P6-10. The polyurethane wheels are 72-mm dia and spaced on 104-mm centers. The skate-boot-foot combination weighs 2 kgf. The effective “spring rate” of the person-skate system is 6000 N/m. The 10-mmdia axle pins are in double shear and of steel with Sut = 550 MPa. Find the infinite-life fatigue safety factor for the pins when a 100-kg person lands a 0.5-m jump on one foot. (a) Assume that all four wheels land simultaneously. (b) Assume that one wheel absorbs all the landing force.

P FIGURE P6-8

*6-23

The beam in Figure P6-11a is subjected to a sinusoidal force time function with Fmax = F and Fmin = –F / 2, where F and the beam’s other data are given in the row(s) assigned from Table P6-5. Find the stress state in the beam due to this loading and choose a material specification that will give a safety factor of 3 for N = 5E8 cycles.

*6-24

The beam in Figure P6-11b is subjected to a sinusoidal force time function with Fmax = F N and Fmin = F / 2, where F and the beam’s other data are given in the row(s) assigned from Table P6-5. Find the stress state in the beam due to this loading and choose a material specification that will give a safety factor of 1.5 for N = 5E8 cycles.

*6-25

The beam in Figure P6-11c is subjected to a sinusoidal force time function with Fmax = F and Fmin = 0, where F and the beam’s other data are given in the row(s) assigned from

Problem 6-19

axle

axle

3 in

3 in A

B

lug wrench

A

lug wrench B F

F FIGURE P6-10 Problem 6-22 * Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

F

FIGURE P6-9 Problem 6-21

(a)

tire

F

tire (b)

Chapter 6

Table P6-5

FATIGUE FAILURE THEORIES

447

l

Data for Problems 6-23 through 6-26

Use Only Data Relevant to a Particular Problem. Lengths in m, Forces in N, I in m4

b

F

Row a

1.00

0.40

0.60

500

2.85E–08

2.00E–02

steel

b

0.70

0.20

0.40

850

1.70E–08

1.00E–02

steel

c

0.30

0.10

0.20

450

4.70E–09

1.25E–02

steel

d

0.80

0.50

0.60

250

4.90E–09

1.10E–02

steel

e

0.85

0.35

0.50

750

1.80E–08

9.00E–03

steel

f

0.50

0.18

0.40

950

1.17E–08

1.00E–02

steel

g

0.60

0.28

0.50

250

3.20E–09

7.50E–03

steel

h

0.20

0.10

0.13

500

4.00E–09

5.00E–03

alum.

i

0.40

0.15

0.30

200

2.75E–09

5.00E–03

alum.

j

0.20

0.10

0.15

80

6.50E–10

5.50E–03

alum.

k

0.40

0.16

0.30

880

4.30E–08

1.45E–02

alum.

l

0.90

0.25

0.80

600

4.20E–08

7.50E–03

alum.

m

0.70

0.10

0.60

500

2.10E–08

6.50E–03

alum.

n

0.85

0.15

0.70

120

7.90E–09

1.00E–02

alum.

x R1

R2 (a )

l F

M1

x R1

l b

Table P6-5. Find the stress state in the beam due to this loading and choose a material specification that will give a safety factor of 2.5 for N = 5E8 cycles. *6-26

*6-27

The beam in Figure P6-11d is subjected to a sinusoidal force time function with Fmax = F lb and Fmin = –F, where F and the beam’s other data are given in the row(s) assigned from Table P6-5. Find the stress state in the beam due to this loading and choose a material specification that will give a safety factor of 6 for N = 5E8 cycles.

F

x R2

R1 (c)

A storage rack is to be designed to hold the paper roll of Problem 6-8 as shown in Figure P6-12. Determine a suitable value for dimension a in the figure for an infinite-life fatigue safety factor of 2. Assume dimension b = 100 mm and that the mandrel is solid and inserts halfway into the paper roll: (more overleaf)

l b

(a) If the beam is a ductile material with Sut = 600 MPa. (b) If the beam is a cast-brittle material with Sut = 300 MPa.

a

6-28 Figure P6-13 shows a forklift truck negotiating a 15° ramp to drive onto a 4-ft-high loading platform. The truck weighs 5000 lb and has a 42-in wheelbase. Design two (one for each side) 1-ft-wide ramps of steel to have a safety factor of 2 for infinite life in the worst case of loading as the truck travels up them. Minimize the weight of the ramps

(b )

F x

R1

R2

R3

(d )

b stanchion

FIGURE P6-11

paper roll

Beams and Beam Loadings for Problems 6-23 to 6-26: See Table P6-5 for Data

a mandrel FIGURE P6-12 Problem 6-27

* Answers

base

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

6

448

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

ramp 1-ft grid

FIGURE P6-13 Problem 6-28

by using a sensible cross-sectional geometry. Choose an appropriate steel or aluminum alloy.

6 * Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

*6-29

A bar 22 mm x 30 mm in cross section is loaded axially in tension with F(t) = ±8 kN. A 10-mm hole passes through the center of the 30-mm side. Find the safety factor for infinite life if the material has Sut = 500 MPa.

6-30 Repeat Problem 6.29 with Fmin = 0, Fmax = 16 kN. *6-31

Repeat Problem 6.29 with Fmin = 8 kN, Fmax = 24 kN.

6-32 Repeat Problem 6.29 with Fmin = –4 kN, Fmax = 12 kN. *6-33

The bracket in Figure P6-14 is subjected to a sinusoidal force time function with Fmax = F and Fmin = –F, where F and the beam’s other data are given in the row(s) assigned from Table P6-6. Find the stress states at points A and B due to this fully reversed loading and choose a ductile steel or aluminum material specification that will give a safety factor of 2 for infinite life if steel or N = 5E8 cycles if aluminum. Assume a geometric stress-concentration factor of 2.5 in bending and 2.8 in torsion.

*6-34

The bracket in Figure P6-14 is subjected to a sinusoidal force time function with Fmax = F and Fmin = 0, where F and the beam’s other data are given in the row(s) assigned from Table P6-6. Find the stress states at points A and B due to this repeated loading and choose a ductile steel or aluminum material specification that will give a safety factor of 2 for infinite life if steel or N = 5E8 cycles if aluminum. Assume a geometric stressconcentration factor of 2.8 in bending and 3.2 in torsion.

6-35 Repeat Problem 6-33 using a cast iron material. 6-36 Repeat Problem 6-34 using a cast iron material.

F

l a A

tube

y

B wall

t arm

z x od

id

FIGURE P6-14 Problems 6-33 to 6-36 (A Solidworks model of this is on the book’s website)

h

Chapter 6

Table P6-6

FATIGUE FAILURE THEORIES

449

Data for Problems 6-33 through 6-36 Use only data that are relevant to the particular problem. Lengths in mm, forces in N.

Row

l

a

t

h

F

od

id

E

a

100

400

10

20

50

20

14

steel

b

70

200

6

80

85

20

6

steel

c

300

100

4

50

95

25

17

steel

d

800

500

6

65

160

46

22

alum.

e

85

350

5

96

900

55

24

alum.

f

50

180

4

45

950

50

30

alum.

g

160

280

5

25

850

45

19

steel

h

200

100

2

10

800

40

24

steel

i

400

150

3

50

950

65

37

steel

j

200

100

3

10

600

45

32

alum.

k

120

180

3

70

880

60

47

alum.

l

150

250

8

90

750

52

28

alum.

m

70

100

6

80

500

36

30

steel

n

85

150

7

60

820

40

15

steel

*6-37

A semicircular curved beam as shown in Figure P6-15 has od = 150 mm, id = 100 mm and t = 25 mm. For a load pair F = ±3 kN applied along the diameter, find the fatigue safety factor at the inner and outer fibers: (a) If the beam is steel with Sut = 700 MPa. (b) If the beam is cast iron with Sut = 420 MPa.

6-38 A 42-mm-dia steel shaft with a 19-mm transverse hole is subjected to a sinusoidal combined loading of σ = ±100 MPa bending stress and a steady torsion of 110 MPa. Find its safety factor for infinite life if Sut = 1 GPa. *6-39

A 42-mm-dia steel shaft with a 19-mm transverse hole is subjected to a combined loading of σ = ±100 MPa bending stress and an alternating torsion of ±110 MPa, which are 90° out-of-phase. Find its safety factor for infinite life if Sut = 1 GPa.

6-40 Redesign the roll support of Problem 6-8 to be like Figure P6-16. The mandrels insert to 10% of the roll length. Design dimensions a and b for an infinite-life safety factor of 2. (a) If the beam is steel with Sut = 600 MPa. (b) If the beam is cast iron with Sut = 300 MPa. *6-41

A 10-mm-ID steel tube carries liquid at 7 MPa. The pressure varies periodically from zero to maximum. The steel has Sut = 400 MPa. Determine the infinite-life fatigue safety factor for the wall if its thickness is: (a) 1 mm. (b) 5 mm.

6-42 A cylindrical tank with hemispherical ends is required to hold 150 psi of pressurized air at room temperature. The pressure cycles from zero to maximum. The steel has Sut = 500 MPa. Determine the infinite-life fatigue safety factor if the tank diameter is 0.5 m, the wall thickness is 1 mm, and its length is 1 m.

t

OD ID

F F

6 FIGURE P6-15 Problem 6-37 (A Solidworks

model is on the book’s website)

* Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

450

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

b typ stanchion paper roll

a typ

mandrel base

FIGURE P6-16

Problem 6-40 (A Solidworks model of this is on the book’s website)

6-43 The paper rolls in Figure P6-17 are 0.9-m-OD x 0.22-m-ID x 3.23 m long and have a density of 984 kg/m3. The rolls are transferred from the machine conveyor (not shown) to the forklift truck by the V-linkage of the off-load station, which is rotated through 90° by an air cylinder. The paper then rolls onto the waiting forks of the truck. The forks are 38-mm-thick by 100-mm-wide by 1.2 m long and are tipped at a 3° angle from the horizontal and have Sut = 600 MPa. Find the infinite-life fatigue safety factor for the two forks on the truck when the paper rolls onto it under two different conditions (state all assumptions):

6

(a) The two forks are unsupported at their free end. (b) The two forks are contacting the table at point A. *

Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

6-44 Determine a suitable thickness for the V-links of the off-loading station of Figure P6-17 to limit their deflections at the tips to 10 mm in any position during their rotation. Two V-links support the roll, at the 1/4 and 3/4 points along the roll’s length, and each of the V arms is 10 cm wide by 1 m long. What is their infinite-life fatigue safety factor when designed to limit deflection as above? Sut = 600 MPa. See Problem 4-43 for more information. 6-45 Determine the infinite-life fatigue safety factor based on the tension load on the air-cylinder rod in Figure P6-17. The tension load cycles from zero to maximum (compression

V-links

1m crank arm

paper rolling machine forks A rod FIGURE P6-17 Problems 6-43 to 6-47

off-loading station

air cylinder

forklift truck

Chapter 6

FATIGUE FAILURE THEORIES

451

rollers FIGURE P6-18 Problem 6-48

loads below the critical buckling load will not affect fatigue life). The crank arm that it rotates is 0.3 m long and the rod has a maximum extension of 0.5 m. The 25-mm-dia rod is solid steel with Sut = 600 MPa. State all assumptions.

6

6-46 The V-links of Figure P6-17 are rotated by the crank arm through a shaft that is 60-mm dia by 3.23-m long. Determine the maximum torque applied to this shaft during the motion of the V-linkage and find the infinite-life fatigue safety factor for the shaft if its Sut = 600 MPa. See Problem 6-43 for more information. *6-47

Determine the maximum forces on the pins at each end of the air cylinder of Figure P6-17. Determine the infinite-life fatigue safety factor for these pins if they are 30-mm dia and in single shear. Sut = 600 MPa. See Problem 6-43 for more information.

6-48 Figure P6-18 shows an exerciser for a 100-kg wheelchair racer. The wheelchair has 65-cm-dia drive wheels separated by a 70-cm track width. Two free-turning rollers on bearings support the rear wheels. The lateral movement of the chair is limited by the flanges. Design the 1-m-long rollers as hollow tubes of aluminum (select alloy) to minimize the height of the platform and also limit the roller deflections to 1 mm in the worst case. Specify suitably sized steel axles to support the tubes on bearings. Calculate the fatigue safety factors at a life of N = 5E8 cycles.

* Answers

to these problems are provided in Appendix D. Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number. Problems in succeeding chapters may also continue and extend these problems.

6-49 Figure P6-19 shows a machined pivot pin that is a press-fit into part A and is a slip fit in part B. If F = 100 lb, l = 2 in, and d = 0.5 in, what is the pin’s safety factor against fatigue when made of SAE 1020 cold-rolled steel? The loading is fully reversed and a reliability of 90% is desired. There is a bending stress concentration factor Kt = 1.8 at the section where the pin leaves part A on the right-hand side.

l l 2 l 4

6-50 Figure P6-19 shows a machined pivot pin that is a press-fit into part A and is a slip fit in part B. If F = 100 N, l = 50 mm, and d = 16 mm, what is the pin’s safety factor against fatigue when made of class 50 cast iron? The loading is fully reversed and a reliability of 90% is desired. There is a bending stress concentration factor Kt = 1.8 at the section where the pin leaves part A on the right-hand side. 6-51 A component in the shape of a large sheet is to be fabricated from 7075-T651 aluminum, which has fracture toughness Kc = 24.2 MPa-m0.5 and a tensile yield strength of 495 MPa. Determine the number of loading cycles that can be endured if the nominal stress varies from 0 to one-half the yield strength and the initial crack had a total length of 1.2 mm. The values of the coefficient and exponent in equation 6.4 for this material n n  are A = 5 x 10–11 ( mm/cycles)  MPa ( m )  and n = 4. *6-52

A component in the shape of a large sheet is to be fabricated from SAE 4340 steel, which has a fracture toughness Kc = 98.9 MPa-m0.5. The sheets are inspected for crack flaws after fabrication, but the inspection device cannot detect flaws smaller than 5 mm. Determine the minimum thickness required for the sheet to have a minimum cycle life

F

B d A

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P6-19 Problems 6-49 and 6-50

452

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

of 106 cycles (using fracture-mechanics criteria) if its width is 400 mm and the load normal to the crack varies from 20 to 170 kN. The values of the coefficient and exponent in n n  equation 6.4 for this material are A = 4 x 10–9 ( mm/cycles)  MPa ( m )  and n = 3. 6-53 A closed, thin-wall cylinder is made from an aluminum alloy that has a fracture toughness of 38 MPa-m0.5 and has the following dimensions: length = 200 mm, OD = 84 mm, and ID = 70 mm. A 2.8-mm-deep semicircular crack is discovered on the inner diameter away from the ends, oriented along a line parallel to the cylinder axis. If the cylinder is repeatedly pressurized from 0 to 75 MPa, how many pressure cycles can it withstand? The values of the coefficient and exponent in equation 6.4 for this material are A = 5 x n n  10–12 ( mm/cycles)  MPa ( m )  and n = 4. (Hint: The value of the geometry factor for a semicircular surface flaw is β = 2 / p and the crack grows in the radial direction.) 6-54 A nonrotating, hot-rolled, steel beam has a channel section with h = 64 mm and b = 127 mm. It is loaded in repeated bending with the neutral axis through the web. Determine its corrected fatigue strength with 90% reliability if it is used in an environment that has a temperature that is below 450°C and has an ultimate tensile strength of 320 MPa.

6

6-55 A nonrotating, machined, steel rod has a round section with d = 50 mm. It is loaded with a fluctuating axial force. Determine its corrected fatigue strength with 99% reliability if it is used in an environment that has a temperature below 450°C and has an ultimate tensile strength of 480 MPa. 6-56 A nonrotating, cold-drawn, steel rod has a round section with d = 76 mm. It is loaded in repeated torsion. Determine its corrected fatigue strength with 99% reliability if it is used in an environment that has a temperature of 500°C and has an ultimate tensile strength of 855 MPa. 6-57 A nonrotating, ground, steel rod has a rectangular section with h = 60 mm and b = 40 mm. It is loaded in repeated bending. Determine its corrected fatigue strength with 99.9% reliability if it is used in an environment that has a temperature that is below 450°C and has an ultimate tensile strength of 1550 MPa. 6-58 A grooved steel shaft similar to that shown in Figure C-5 (Appendix C) is to be loaded in bending. Its dimensions are: D = 57 mm, d = 38 mm, r = 3 mm. Determine the fatigue stress-concentration factor if the material Sut = 1130 MPa. 6-59 A steel shaft with a transverse hole similar to that shown in Figure C-8 (Appendix C) is to be loaded in torsion. Its dimensions are: D = 32 mm, d = 3 mm. Determine the fatigue stress-concentration factor if the material Sut = 808 MPa. 6-60 A hardened aluminum filleted flat bar similar to that shown in Figure C-9 (Appendix C) is to be loaded axially. Its dimensions are: D = 1.20 in, d = 1.00 in, r = 0.10 in. Determine the fatigue stress-concentration factor if the material Sut = 76 kpsi.

D

d

r Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P6-20 Problem 6-61, 6-86, 6-90

6-61 A rotating shaft with a shoulder fillet seated in the inner race of a rolling contact bearing with the shoulder against the edge of the bearing is shown in Figure P6-20. The bearing has a slight eccentricity that induces a fully reversed bending moment in the shaft as it rotates. Measurements indicate that the resulting alternating stress amplitude due to bending is σa = 57 MPa. The torque on the shaft fluctuates from a high of 90 N-m to a low of 12 N-m and is in phase with the bending stress. The shaft is ground and its dimensions are D = 23 mm, d = 19 mm, and r = 1.6 mm. The shaft material is SAE 1040 cold-rolled steel. Determine the infinite-life fatigue safety factor for the shaft for a reliability of 99%. 6-62 A tension member in a machine is filleted as shown in Figure P6-21. The member has a manufacturing defect that causes the fluctuating tension load to be applied eccentrically resulting in a fluctuating bending load as well. Measurements indicate that the

Chapter 6

FATIGUE FAILURE THEORIES

maximum bending stress is 16.4 MPa and the minimum is 4.1 MPa. The tensile load fluctuates from a high of 3.6 kN to a low of 0.90 kN and is in phase with the bending stress. The member is machined and its dimensions are D = 33 mm, d = 25 mm, h = 3 mm and r = 3 mm. The material is SAE 1020 cold-rolled steel. Determine the infinite-life fatigue safety factor for the member for 90% reliability. 6-63 For a filleted flat bar in tension similar to that shown in Figure C-9 (Appendix C) and the data from the row(s) assigned from Table P6-7, determine the alternating and mean axial stresses as modified by the appropriate stress concentration factors in the bar. 6-64 For a filleted flat bar in bending similar to that shown in Figure C-10 (Appendix C) and the data from the row(s) assigned from Table P6-7, determine the alternating and mean bending stresses as modified by the appropriate stress concentration factors in the bar. 6-65 For a shaft, with a shoulder fillet, in tension similar to that shown in Figure C-1 (Appendix C) and the data from the row(s) assigned from Table P6-7, determine the alternating and mean axial stresses as modified by the appropriate stress concentration factors in the shaft. 6-66 For a shaft, with a shoulder fillet, in bending similar to that shown in Figure C-2 (Appendix C) and the data from the row(s) assigned from Table P6-7, determine the alternating and mean bending stresses as modified by the appropriate stress concentration factors in the shaft. 6-67 A machine part is subjected to fluctuating, simple, multiaxial stresses. The fully corrected nonzero stress ranges are: σxmin = 50 MPa, σxmax = 200 MPa, σymin = 80 MPa, σymax = 320 MPa, τxzmin = 120 MPa, τxymax = 480 MPa. The material properties are: Se = 525 MPa and Sut = 1200 MPa. Using a Case 3 load line, calculate and compare the infinite-life safety factors given by the Sines and von Mises methods. 6-68 A cylindrical tank with hemispherical ends has been built. It was made from hot rolled steel that has Sut = 380 MPa. The tank outside diameter is 300 mm with 20 mm wall thickness. The pressure may fluctuate from 0 to an unknown maximum. For an infinitelife fatigue safety factor of 4 with 99.99% reliability, what is the maximum pressure to which the tank may be subjected? 6-69 A rotating shaft has been designed and fabricated from SAE 1040 HR steel. It is made from tubing that has an outside diameter of 60 mm and a wall thickness of 5 mm. Strain gage measurements indicate that there is a fully reversed axial stress of 68 MPa and a torsional stress that fluctuates from 12 MPa to 52 MPa in phase with the axial stress at the critical point on the shaft. Determine the infinite-life fatigue safety factor for the shaft for a reliability of 99%.

Table P6-7

Data for Problems 6-63 through 6-66 Use only data that are relevant to the particular problem. Lengths in mm, forces in N, and moments in N-m.

Row

D

d

a

40

20

r

h

Mmin

Mmax

Pmin

4

10

80

320

12

100

500

9500

47 500

SAE 1040 CR

8

60

180

6500

19 500

SAE 1020 CR

8000

Pmax

Material

32 000

SAE 1020 CR

b

26

20

1

c

36

30

1.5

d

33

30

1

8

75

300

7200

28 800

SAE 1040 CR

e

21

20

1

10

50

150

5500

16 500

SAE 1050 CR

f

51

50

1.5

7

80

320

8000

32 000

SAE 1020 CR

g

101

100

5

8

400

800

15 000

60 000

SAE 1040 CR

Copyright © 2018 Robert L. Norton: All Rights Reserved

453

F d r h

D F Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P6-21 Problem 6-62

6

454

MACHINE DESIGN

Data for Problem 6-70

6

σm’

An Integrated Approach

-

Sixth Edition

6-70 For the mean and alternating stress values (in MPa) in the row(s) assigned in Table P6-8, find the safety factor for each of the four loading variation cases based on the ModifiedGoodman diagram if Se = 100, Sy = 150, and Sut = 200 MPa.

Table P6-8

Row

-

σa’

a

50

30

b

70

30

c

100

10

d

20

60

e

80

40

f

40

40

g

120

50

h

80

80

Copyright © 2018 Robert L. Norton: All Rights Reserved

6-71 A rotating shaft with a shoulder fillet, in torsion similar to that shown in Appendix C Figure C-3 is made from SAE 1020 CR steel and has dimensions D = 40 mm, d = 20 mm, and r = 4 mm. The shaft is ground and is subjected to a fully reversed torque of +/– 80 N-m. Determine the shaft’s infinite-life safety factor for 99.9% reliability. 6-72 A 1.6-mm-long central crack is found on a fuselage panel of a commuter aircraft during routine inspection. The material has the same properties as the panel in Problem 6-51. The plane takes off and lands 32 times each day (creating a pressure cycle each time), 260 days per year. On the ground the panel is unstressed. In the air the maximum tensile stress is 300 MPa. If the next inspection will be done six months later, estimate the expected crack length at that time. 6-73 Steel sheets 400 mm wide by 2000 mm long by 2-mm thick are fabricated from the same material given in Problem 6-52 with the inspection device used in that problem. Using fracture-mechanics criteria, estimate the minimum potential cycle life of these sheets if they will be subjected to a lengthwise tensile load that continuously varies from 50 to 500 kN. 6-74 Repeat problem 6-53 for a cylinder with dimensions: length = 180 mm, OD = 80 mm, and ID = 72 mm; a 2.4-mm-deep semicircular crack; and a pressure that varies from 0 to 70 MPa. 6-75 A nonrotating, hot-rolled, steel beam has an I-beam section with h = 200 mm, b = 200 mm, and t = 25 mm. It is loaded in repeated bending with the neutral axis through the web. Determine its corrected fatigue strength with 95% reliability if it is used in an environment that has a temperature that is below 450C and has an ultimate tensile strength of 365 MPa. 6-76 A steel, shaft with shoulder fillet similar to that shown in Figure C-3 (Appendix C) is to be loaded in torsion. Its dimensions are: D = 52 mm, d = 39 mm, r = 4 mm. Determine the fatigue stress-concentration factor if the material Sut = 1330 MPa. 6-77 A steel, grooved shaft similar to that shown in Figure C-4 (Appendix C) is to be loaded in axial tension. Its dimensions are: D = 60 mm, d = 40 mm, r = 5 mm. Determine the fatigue stress-concentration factor if the material Sut = 1200 MPa. 6-78 A steel, grooved shaft similar to that shown in Figure C-6 (Appendix C) is to be loaded in torsion. Its dimensions are: D = 48 mm, d = 40 mm, r = 2 mm. Determine the fatigue stress-concentration factor if the material Sut = 1450 MPa. 6-79 A cylindrical tank with hemispherical ends is made from hot rolled steel that has Sut = 450 MPa. The tank outside diameter is 350 mm with 25 mm wall thickness. The pressure fluctuates from 0 to 15 MPa. Determine the infinite-life fatigue safety factor for 99.9% reliability. 6-80 A cylindrical tank with hemispherical ends is to be fabricated from cold-rolled steel that has Sut = 550 MPa. The tank outside diameter is to be 250 mm. The pressure will fluctuate from 0 to 15 MPa. For an infinite-life fatigue safety factor of 2 with 99% reliability, determine the minimum wall thickness that may be used. 6-81 A ground steel shaft with a shoulder fillet has dimensions: D = 30 mm, d = 25 mm, and r = 2 mm. It is subjected to a combined nominal alternating loading of σσ = 50 to 180 MPa bending stress and torsional stress of στ = 10 to 50 MPa, which are in phase. Find the safety factor for infinite life if Sut = 965 MPa and the reliability is 95%.

Chapter 6

FATIGUE FAILURE THEORIES

455

6-82 A component in the shape of a large sheet is to be fabricated from 2024-T351 aluminum. Estimate the number of loading cycles that can be endured if the nominal stress varies from 0 to six tenths of the yield strength and the initial crack had a total length of 1 mm. The values of the coefficient and exponent in equation 6.4 for this material are A = 4.8 x 10-11 (mm/cycle) and n = 3.5. 6-83 A machine part is made from 2024-T351 aluminum plate. For each cycle of the machine the part undergoes a tensile stress that varies from 0 to 1/2 of the material’s tensile yield strength. The machine cycles 20 times per minute, 24 hours per day, 365 days per year. At the last inspection of the machine the part was found to have a central crack 1.7 mm long. The machine is inspected every six months. Estimate the expected crack length at next inspection. The values of the coefficient and exponent in equation 6.4 for this material are A = 4.8 x 10-11 (mm/cycle) and n = 3.5. 6-84 A non rotating, hot-rolled, SAE 1010 steel beam has a rectangular section with h = 80 mm and b = 40 mm. It is loaded in repeated bending with the neutral axis through the long dimension. Determine its corrected fatigue strength with 95% reliability if it is used in an environment that has a temperature that is below 450C. 6-85 A hardened aluminum, notched, flat bar similar to that shown in Figure C-11 (Appendix C) is to be loaded axially. Its dimensions are: D = 1.625 in, d = 1.250 in, r = 0.125 in. Determine the fatigue stress-concentration factor if the material Sut = 55 kpsi. 6-86 A rotating shaft with a shoulder fillet seated in the inner race of a rolling contact bearing with the shoulder against the edge of the bearing is shown in Figure P6-20. The bearing has a slight eccentricity that induces a fully reversed bending moment in the shaft as it rotates. Measurements indicate that the resulting fully reversed bending moment at the shaft shoulder (in N-m) is (4.3x103)d where d is in mm . Design a steel shaft for this application that has a dynamic factor of safety of 2 for an infinite life. Bearings are available with IDs of 10 mm to 100 mm in increments of 2 mm. To seat the bearing properly the maximum fillet radius on the shaft is rmax = d x 0.05 and the maximum shoulder diameter is Dmax = d + (4 x rmax). 6-87 A rotating, steel shaft with a shoulder fillet similar to that shown in Figure C-2 (Appendix C) is to be loaded in bending. The dimensions of the fillet radius and shoulder diameter are related to the small diameter as r = d x 0.05 and D = d + (4 x r). For a fully reversed bending moment at the fillet of 880d lb-in, where d is in in, a material with Sut = 100 kpsi, and an infinite life calculate and plot the dynamic safety factor as a function of the diameter d. 6-88 A nonrotating, cold-rolled, SAE 1040 steel rod has a square section with h = 50 mm. It is loaded with a fluctuating axial force. Determine its corrected fatigue strength with 99% reliability if it is used in an environment that has a temperature below 450C. 6-89 An annealed aluminum, flat bar with a transverse hole similar to that shown in Figure C-14 (Appendix C) is to be loaded in bending. Its dimensions are: W = 1.625 in, d = 0.125 in, h = 0.125 in. Determine the fatigue stress-concentration factor if the material Sut = 34 kpsi. 6-90 A rotating shaft with a shoulder fillet seated in the inner race of a rolling contact bearing with the shoulder against the edge of the bearing is shown in Figure P6-20. The bearing has a slight eccentricity that induces a fully reversed bending moment in the shaft as it rotates. Measurements indicate that the resulting fully reversed bending moment at the shaft shoulder (lb-in) is 880d where d is in inches. Design a steel shaft for this application that has a dynamic factor of safety of 2 for an infinite life. Bearings are available with IDs of 0.5 in to 5.0 in in increments of 0.125 in. To seat the bearing properly the maximum fillet radius on the shaft is rmax = d x 0.05 and the maximum shoulder diameter is Dmax = d + (4 x rmax).

6

456

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

6-91 A 1.625-in-dia SAE 1050 HR steel shaft with a 1.250-in-dia groove is subjected to a combined loading of σσ = ±8 kpsi bending stress and an alternating torsional stress that varies from −2 kpsi to +6 kpsi, which are 45 deg out of phase. Find the safety factor for infinite life for a reliability of 99% if the shaft is ground and operates at room temperature. The groove radius is 0.125 in. 6-92 For the part in Problem 6-54, find the corrected endurance strength (or limit), create equations for the S-N line, draw the S-N diagram, and find the fatigue strength at 105 cycles. 6-93 A steel, flat bar with a transverse hole similar to that shown in Figure C-13 (Appendix C) is to be loaded in bending. Its dimensions are: W = 1.625 in and d = 0.125 in. Determine the fatigue stress-concentration factor if the material Sut = 127 kpsi.

6

SURFACE FAILURE

7

Use it up, wear it out; Make it do, or do without. New eNglaNd MaxiM

7.0

INTRODUCTION

There are only three ways in which parts or systems can “fail”: obsolescence, breakage, or wearing out. My old computer still works well but is obsolete and no longer of any use to me. My wife’s favorite vase is now in pieces since I dropped it on the floor, and it is irrecoverable. However, my 123 000-mile automobile is still quite serviceable and useful despite showing some signs of wear. Most systems are subject to all three types of possible failure. Failure by obsolescence is somewhat arbitrary. (My grandchild is now getting good use of the old computer.) Failure by breakage is often sudden and may be permanent. Failure by “wearing out” is generally a gradual process and is sometimes repairable. Ultimately, any system that does not fall victim to one of the other two modes of failure will inevitably wear out if kept in service long enough. Wear is the final mode of failure, which nothing escapes. Thus, we should realize that we cannot design to avoid all types of wear completely, only to postpone them. The previous chapters have dealt with failure of parts by distortion (yielding) and breakage (fracture). Wear is a broad term that encompasses many types of failures, all of which involve changes to the surface of the part. Some of these so-called wear mechanisms are still not completely understood, and rival theories exist in some cases. Most experts describe five general categories of wear: adhesive wear, abrasive wear, erosion, corrosion wear, and surface fatigue. The following sections discuss these topics in detail. In addition, there are other types of surface failure that do not fit neatly into one of the five categories or that can fit into more than one. Corrosion fatigue has aspects of the last two categories as does fretting corrosion. For simplicity, we will discuss these hybrids in concert with one of the five main categories listed above. Failure from wear usually involves the loss of some material from the surfaces of solid parts in the system. The wear motions of interest are sliding, rolling, or some 457

457

458

MACHINE DESIGN

Opening page photograph courtesy of V. S. Makhijani.

Table 7-0 Symbol

a Aa

7

An Integrated Approach

-

Sixth Edition

Variables Used in This Chapter Variable

ips units

SI units

See

half-width of contact patch

in

m

Sect. 7.8–7.10

apparent area of contact

in2

m2

Sect. 7.2

Ar

real area of contact

in2

m2

Eq. 7.1

B b d E F, P f f max

geometry factor

1/in

1/m

Eq. 7.9b

half-length of contact patch

in

m

Sect. 7.8–7.10

depth of wear

in

m

Eq. 7.7

Young's modulus

psi

Pa

all

force or load

lb

N

all

friction force

lb

N

Eq. 7.2

maximum tangential force

lb

N

Eq. 7.22f

K l L m1, m2

wear coefficient

none

none

Eq. 7.7

length of linear contact

in

m

Eq. 7.7

length of cylindrical contact

in

m

Eq. 7.14

material constants

1/psi

m2/N

Eq. 7.9a

penetration hardness

psi

kg/mm2

Eq. 7.7

number of cycles

none

none

Eq. 7.26

H N Nf

* A 1977 study sponsored by the ASME estimated that the energy cost to the U.S. economy associated with the replacement of equipment that failed from wear accounted for 1.3% of total U.S. energy consumption. This was then equivalent to about 160 million barrels of oil per year. See O. Pinkus and D. F. Wilcock, Strategy for Energy Conservation through Tribology, ASME, New York, 1977, p. 93.

-

safety factor in surface fatigue

none

none

Example 7-5

p p avg

pressure in contact patch

psi

N/m2

Sect. 7.8–7.10

avg pressure in contact patch

psi

N/m2

Sect. 7.8–7.10

p max

max pressure in contact patch

psi

N/m2

Sect. 7.8–7.10

R,R

radii of curvature

in

m

Eq. 7.9b

Sus

ultimate shear strength

psi

Pa

Sect. 7.3

Sut

ultimate tensile strength

psi

Pa

Sect. 7.3

Sy

yield strength

psi

Pa

Sect. 7.3

Syc

yield strength in compression

psi

Pa

Sect. 7.3

V x, y, z µ ν σ σ1

volume

in3

m3

Eq. 7.7

generalized length variables

in

m

all

coefficient of friction

none

none

Eq. 7.2–7.6

Poisson's ratio

none

none

all

normal stress

psi

Pa

all

principal stress

psi

Pa

Sect. 7.11

σ2

principal stress

psi

Pa

Sect. 7.11

σ3

principal stress

psi

Pa

Sect. 7.11

τ τ 13

shear stress

psi

Pa

all

maximum shear stress

psi

Pa

Sect. 7.11

τ 21

principal shear stress

psi

Pa

Sect. 7.11

τ 32

principal shear stress

psi

Pa

Sect. 7.11

combination of both. Wear is a serious cost to the national economy.* It only requires the loss of a very small volume of material to render the entire system nonfunctional. Rabinowicz[1] estimated that a 4000-lb automobile, when completely “worn out,” will

Chapter 7

SURFACE FAILURE

459

have lost only a few ounces of metal from its working surfaces. Moreover, these damaged surfaces will not be visible without extensive disassembly, so it is often difficult to monitor and anticipate the effects of wear before failure occurs. Table 7-0 shows the variables used in this chapter and references the equations, tables, or sections in which they are used. At the end of the chapter, a summary section is provided that also groups all the significant equations from this chapter for easy reference and identifies the chapter section in which their discussion can be found. A master lecture video on the topics in sections 7.1–7.13 is linked as Lecture 12.

7.1

SURFACE GEOMETRY View a Video of Lecture 12 (52:25)*

*

Before discussing the types of wear mechanisms in detail it will be useful to define the characteristics of an engineering surface that are relevant to these processes. (Material strength and hardness will also be factors in wear.) Most solid surfaces that are subject to wear in machinery will be either machined or ground, though some will be as-cast or asforged. In any case, the surface will have some degree of roughness that is concomitant with its finishing process. Its degree of roughness or smoothness will have an effect on both the type and degree of wear that it will experience. Even an apparently smooth surface will have microscopic irregularities. These can be measured by any of several methods. A profilometer passes a lightly loaded, hard (e.g., diamond) stylus over the surface at controlled (low) velocity and records its undulations. The stylus has a very small (about 0.5 µm) radius tip that acts, in effect, as a low-pass filter, since contours smaller than its radius are not sensed. Nevertheless, it gives a reasonably accurate profile of the surface with a resolution of 0.125 µm or better. Figure 7-1 shows the profiles and SEM* photographs (100x) of both ground (a) and machined

(a)

FIGURE 7-1

http://www.designofmachinery. com/MD/12_Wear_and_Surface_Fatigue.mp4

7

*

Scanning Electron Microscope

(b) Copyright © 2018 Robert L. Norton: All Rights Reserved

Scanning Electron Microscope Surface-Replica Photographs (100x) and Profiles of Ground (a) and Milled (b) Cam Surfaces

460

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

(b) surfaces of hardened steel cams. The profiles were measured with a Hommel T-20 profilometer that digitizes 8000 data points over the sample length (here 2.5 mm). The microscopic “mountain peaks” on the surfaces are called asperities. From these profiles a number of statistical measures may be calculated. ISO defines at least 19 such parameters. Some of them are shown in Figure 7-2 along with their mathematical definitions. Perhaps the most commonly used parameters are Ra, which is the average of the absolute values of the measured points, or Rq, which is their rms average. These are very similar in both value and meaning. Unfortunately, many engineers specify only one of these two parameters, neither of which tells enough about the surface. For example, the two surfaces shown in Figures 7-3a and b have the same Ra and Rq values but are clearly different in nature. One has predominantly positive, and the other predominantly negative, features. These two surfaces will react quite differently to sliding or rolling against another surface.

7

FIGURE 7-2 DIN and ISO Surface Roughness, Waviness, and Skewness Parameter Definitions

Chapter 7

Ra or Rq

SURFACE FAILURE

461

Ra or Rq

(a )

( b)

FIGURE 7-3 Different Surface Contours Can Have the Same Ra or Rq Values

In order to differentiate these surfaces that have identical Ra or Rq values, other parameters should be calculated. Skewness Sk is a measure of the average of the first derivative of the surface contour. A negative value of Sk indicates that the surface has a predominance of valleys (Figure 7-3a) and a positive Sk defines a predominance of peaks (Figure 7-3b). Many other parameters can be computed (see Figure 7-2). For example, Rt defines the largest peak-to-valley dimension in the sample length, Rp the largest peak height above the mean line, and Rpm the average of the 5 largest peak heights. All the roughness measurements are calculated from an electronically filtered measurement that zeros out any slow-changing waves in the surface. An average line is computed from which all peak/valley measurements are then made. In addition to these roughness measurements (denoted by R), the waviness Wt of the surface can also be computed. The Wt computation filters out the high-frequency contours and preserves the long-period undulations of the raw surface measurement. If you want to completely characterize the surface-finish condition, note that using only Ra or Rq is not sufficient.

7.2

MATING SURFACES

When two surfaces are pressed together under load, their apparent area of contact Aa is easily calculated from geometry but their real area of contact Ar is affected by the asperities present on their surfaces and is more difficult to accurately determine. Figure 7-4 shows two parts in contact. The tops of the asperities will initially contact the mating part and the initial area of contact will be extremely small. The resulting stresses in the asperities will be very high and can easily exceed the compressive yield strength of the material. As the mating force is increased, the asperity tips will yield and spread until their combined area is sufficient to reduce the average stress to a sustainable level, i.e., some compressive penetration strength of the weaker material. We can get a measure of a material’s compressive penetration strength from conventional hardness tests (Brinell, Rockwell, etc.), which force a very smooth stylus into

FIGURE 7-4 The Actual Contact Between Two Surfaces Is Only at the Asperity Tips

7

462

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

the material and deform (yield) the material to the stylus’ shape. (See Section 2.4.) The penetration strength Sp is easily calculated from these test data and tends to be of the order of 3 times the compressive yield strength Syc of most materials.[2] The real area of contact can then be estimated from Ar ≅

F F ≅ S p 3S yc

(7.1)

where F is the force applied normal to the surface and the strengths are as defined in the above paragraph, taken for the weaker of the two materials. Note that the contact area for a material of particular strength under a given load will be the same regardless of the apparent area of the mating surfaces.

7.3

FRICTION

Note that the real area of contact Ar (equation 7.1) is independent of the apparent area Aa that is defined by the geometry of the mating parts. This is the reason that Coulomb friction between two solids is also independent of the apparent area of contact Aa. The equation for Coulomb sliding friction is

7

f = µF

(7.2a)

where f is the force of friction, µ is the coefficient of dynamic friction, and F is the normal force. The normal force presses the two surfaces together and creates elastic deformations and adhesions (see next section) at the asperities’ tips. We can define the dynamic Coulomb friction force f as being the force necessary to shear the adhered and elastically interlocked asperities in order to allow a sliding motion. This shearing force is equal to the product of the shear strength of the weaker material and the actual contact area Ar plus a “plough force” P: f = Sus Ar + P

* This

will be true only if the two surfaces have about the same hardness. If one surface is harder and rougher than the other, there could be a significant plough force.

(7.2b)

The plough force P is due to loose particles digging into the surfaces and is negligible compared to the shear force,* so can be ignored. Recalling equation 7.1 gives Ar ≅

F 3S yc

(7.2c)

Substituting equation 7.2c in 7.2b (ignoring P) gives f ≅F

Sus 3S yc

(7.2d )

Combining equations 7.2a and 7.2d gives µ=

S f ≅ us F 3S yc

(7.3)

which indicates that the coefficient of friction µ is a function only of a ratio of material strengths of the weaker of the two materials in contact.

Chapter 7

SURFACE FAILURE

463

The ultimate shear strength can be estimated based on the ultimate tensile strength of the material. steels:

Sus ≅ 0.8 Sut

other ductile metals:

Sus ≅ 0.75Sut

(7.4)

The yield strength in compression as a fraction of the ultimate tensile strength varies with the material and alloy over a fairly broad range, perhaps 0.5Sut < S yc < 0.9 Sut

(7.5)

Substituting equations 7.4 and 7.5 in equation 7.3 gives 0.75Sut 0.8 Sut  Fi as shown in Figure 15-22e). The additional deflection ∆δ creates a new load situation in both the bolt and material as shown in Figure 15-24b. The load in the material is reduced by Pm and moves down the material stiffness line to point D with a new value Fm. The load in the bolt is increased by Pb and moves up the bolt stiffness line to point C with a new value Fb. Note that the applied load P is split into two components, one (Pm) taken by the material and one (Pb) taken by the bolt. P = Pm + Pb

(15.12a)

The compressive load Fm in the material is now Fm ≥ 0*

Fm = Fi − Pm :

(15.12b)

and the tensile load Fb in the bolt becomes Fb = Fi + Pb

* If Fm has a negative value, set Fm = 0, because the material cannot support a tensile force—it will separate.

(15.12c)

Note what has happened as a result of the preload force Fi. The “spring” of the material was “wound up” under preloading. Any applied loads are partially supported by the “unwinding” of this spring. If the relative stiffness of bolt and material are as shown in Figure 15-24 (i.e., material stiffer than bolt), the material supports the majority

F A

B

A

Fi km δm

kb

∆δ δb

(a) Preload force and initial deflections

δ

Pb

F Fb Fi D

Pm

C B

P

Fm

δm

15

∆δ δb

(b) Load deflection and resulting forces

FIGURE 15-24 Effects on Bolt and Material from Preload: (a) Preload and (b) Applied Load

δ

930

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

of the applied load and the bolt feels little additional load above that of the initial preload. This is one aspect of the justification for the earlier statement that “if the bolt doesn’t fail when preloaded, it probably won’t fail in service.” There is also another reason for this being true, which will be discussed in a later section. Note, however, that if the applied load P is large enough to cause the component Pm to exceed the preload force Fi, then the joint will separate, Fm will be zero, and the bolt will feel the full value of the applied load P. The material can no longer contribute to supporting the load if the joint is separated. This is one reason for the very large recommended preloads as a percentage of bolt proof strength. In order to get the full benefit of material load sharing, the preload should be high. We can summarize the information in Figure 15-24 in the following way. The common change in deflection ∆δ due to the applied load P is δ=

Pb Pm = kb km

(15.13a)

kb Pm km

(15.13b)

Pb =

or:

Substitute equation 15.12a to get Pb = or

kb P km + kb

Pb = CP

where C =

kb km + kb

(15.13c)

The term C is called the joint’s stiffness constant or just the joint constant. Note that C is typically < 1, and if kb is small compared to km, C will be a small fraction. This confirms that the bolt will see only a portion of the applied load P. In like fashion, Pm =

km P = (1 − C ) P kb + km

(15.13d )

These expressions for Pb and Pm can be substituted into equations 15.12b and 15.12c to get expressions for the bolt and material loads in terms of the applied load P:

15

Fm = Fi − (1 − C ) P

(15.14 a)

Fb = Fi + CP

(15.14 b)

Equation 15.14b can be solved for the preload Fi needed for any given combination of applied load P and maximum allowable bolt (proof) load Fb, provided that the joint constant C is known. The load P0 required to separate the joint can be found from equation 15.14a by setting Fm to zero. P0 =

Fi (1 − C )

(15.14c)

Chapter 15

SCREWS AND FASTENERS

931

A safety factor against joint separation can be found from N separation =

P0 Fi = P P (1 − C )

(15.14 d )

EXAMPLE 15-2 Preloaded Fasteners in Static Loading Problem

Determine a suitable bolt size and preload for the joint shown in Figure 15-23 (repeated here). Find its safety factor against yielding and separation. Determine the optimum preload as a percentage of proof strength to maximize the safety factors.

Given

The joint dimensions are D = 1 in and l = 2 in. The applied load P = 2000 lb.

Assumptions

Both of the clamped parts are steel. The effects of the flanges on the joint stiffness will be ignored. A preload of 90% of the bolt’s proof strength will be applied as a first trial.

Solution

See Figure 15-25.

1 As with most design problems, there are too many unknown variables to solve the necessary equations in one pass. Trial values must be chosen for various parameters and iteration used to find a good solution. We actually made several iterations to solve this problem, but will only present two in the interest of brevity. Thus, the trial values used here have already been massaged to reasonable values. 2 The bolt diameter is the principal trial value to be chosen along with a thread series and a bolt class to define the proof strength. We choose a 5/16-18 UNC-2A steel bolt of SAE class 5.2. (This was actually our third trial choice.) For a clamp length of 2 in, assume a bolt length of 2.5 in to allow sufficient protrusion for the nut. The preload is taken at 90% of proof strength as assumed above.

d P/2

P/2

3 Table 15-6 shows the proof strength of this bolt to be 85 kpsi. The tensile-stress area from equation 15-1a is 0.052 431 in2. The preload is then

(

)(

)

Fi = 0.9 S p At = 0.9 85 000 0.052 431 = 4011 lb

(a)

4 Find the lengths of thread lthd and shank ls of the bolt as shown in Figure 15-21: lthd = 2d + 0.25 = 2 ( 0.3125) + 0.25 = 0.875 in ls = lbolt − lthd = 2.5 − 0.875 = 1.625 in

l1 l l2

(b)

15

from which we can find the length of thread lt that is in the clamp zone: lt = l − ls = 2.0 − 1.625 = 0.375 in

(c)

P/2 D

5 Find the stiffness of the bolt from equation 15.11a: l l 1.625 ( 4 ) 1 0.375 = t + s = + kb At E Ab E 0.052 431( 30 E 6 ) π ( 0.3125)2 ( 30 E 6 ) k b = 1.059 E 6 lb/in

P/2

FIGURE 15-23 (d )

Repeated

A Preloaded Bolt Compressing a Cylinder to Which External Loads Are Applied

932

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

6 The calculation for the stiffness of the clamped material is simplified in this example by its relatively small diameter. We can assume in this case that the entire cylinder of material is compressed by the bolt force. (We will soon address the problem of finding the clamped area in a continuum of material.) The material stiffness from equation 15.11d is km =

(

)

(

π D2 − d 2 E π 1.0 2 − 0.3122 m = 4 4 l

) (30 E6) = 1.063E 7 lb/in 2.0

(e)

7 The joint stiffness factor from equation 15.13c is C=

kb 1.059 E 6 = = 0.090 56 km + kb 1.063E 7 + 1.059 E 6

(f)

8 The portions of the applied load P felt by the bolt and the material can now be found from equations 15.13: Pb = CP = 0.090 56(2000) = 181 lb

(

)

Pm = (1 − C ) P = 1 − 0.090 56 (2000) = 1819 lb

( g)

9 Find the resulting loads in bolt and material after the load P is applied: Fb = Fi + Pb = 4011 + 181 = 4192 lb Fm = Fi − Pm = 4011 − 1 819 = 2192 lb

(h)

Note how little the applied load adds to the preload in the bolt. 10 The maximum tensile stress in the bolt is σb =

Fb 4192 = = 79 953 psi At 0.052 431

(i )

Note that no stress-concentration factor is used because it is static loading. 11 This is a uniaxial stress situation, so the principal stress and von Mises stress are identical to the applied tensile stress. The safety factor against yielding is then Ny =

Sy

=

σb

92 000 79 953

= 1.15

( j)

The yield strength is found from Tables 15-6 and 15-7. 12 The load required to separate the joint and the safety factor against joint separation are found from equations 15.14c and 15.14d:

15

P0 =

Fi

(1 − C )

=

(

4011

1 − 0.090 56

N separation =

)

= 4410 lb

P0 4410 = = 2.2 P 2000

(k )

(l )

13 The safety factor against separation is acceptable. The yielding safety factor is low but this is to be expected, since the bolt is deliberately preloaded to a level close to its yield strength.

Chapter 15

SCREWS AND FASTENERS

933

5.0 Ny

4.5 4.0 safety factor

3.5 3.0

Nseparation

2.5

B

2.0

A

1.5

B’

1.0 .05 0 0

0.1

0.2

0.3 0.4 0.5 0.6 0.7 0.8 preload as decimal % of proof strength

FIGURE 15-25

0.9

1.0

Copyright © 2018 Robert L. Norton: All Rights Reserved

Safety Factors Versus Preload for the Statically Loaded Bolt in Example 15-2

14 The model was solved for the range of possible preloads from zero to 100% of proof strength and the safety factors plotted versus preload percentage. The results are shown in Figure 15-25. The separation safety factor rises linearly with increasing preload, but is S y then:

K fm =

S y − K f σ anom σ mnom

K f σ maxnom = K f σ anom + σ mnom = 5.9 864 + 77 364 = 462 548 psi

15

462 548 psi > S y = 92 000 psi K fm =

S y − K f σ anom σ mnom

=

92 000 − 5.9 (864 ) 77 364

= 1.12

8 The local mean and alternating stresses in the bolt are then:

(e)

Chapter 15

F

SCREWS AND FASTENERS

Pb

B

force (lb X 103)

5.0

Fb Pm

4.0

937

Fi A

Fm 3.0 2.0 1.0

km

kb δ

0 –0.0005

0

0.001

0.002

0.003

0.004

deflection (in)

FIGURE 15-27

Dynamic Forces in Bolt and Material for Example 15-3, Drawn to Scale

σ a = K f σ anom = 5.9 (864 ) = 5107 psi

(

(f)

)

σ m = K fm σ mnom = 1.12 77 364 = 86 893 psi

9 The stress at the initial preload is σ i = K fm

Fi 4011 = 1.12 = 85 923 psi 0.052 431 At

( g)

10 An endurance strength must be found for this material. Using the methods of Section 6.6 we find

(

)

Se' = 0.5Sut = 0.5 120 000 = 60 000 psi Se = Cload Csize Csurf Ctemp Creliab Se'

(

(h)

)

= 0.70 ( 0.995)( 0.76 )(1)( 0.81) 60 000 = 25 726 psi

(i )

where the strength reduction factors are taken from the tables and formulas in Section 6.6 for, respectively, axial loading, the bolt size, a machined finish, room temperature, and 99% reliability. 11 The corrected endurance strength and the ultimate tensile strength are used in equation 15.16 to find the safety factor from the Goodman line: Nf = =

Se ( Sut − σ i )

Se ( σ m − σ i ) + Sut σ a

(

(

25 837 120 000 − 85 923

)

)

25 726 86 893 − 85 923 + 120 000 ( 5107 )

= 1.38

The modified-Goodman diagram for this stress state is shown in Figure 15-28.

( j)

15

938

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

30 alternating stress (kpsi)

S 25

to

Goo

dma

20

n li

ne

load line for 90% preload

15 10 σa 5

0

10

20

30

40

50

70

60

80

mean stress (kpsi)

FIGURE 15-28

Su

S

σi

0

90 σm

100

11

120

Modified-Goodman Diagram for Example 15-3

12 The static bolt stress after initial local yielding and the yielding safety factor are: σs =

Fbolt 4102 = = 78 227 psi 0.045 36 At

Ny =

Sy σb

=

92 000 78 227

= 1.18

(k )

13 The preload required to obtain these safety factors is

(

)(

)

Fi = 0.90 S p At = 0.90 85 000 0.052 431 ≅ 4011 lb

(l )

14 The safety factor against joint separation is found from equation 15.14d: N separation =

Fi 4011 = = 4.4 P (1 − C ) 1000 1 − 0.090 56

(

)

(m )

15 The fatigue and separation safety factors are acceptable. The yielding safety factor is low but is nevertheless acceptable, since the bolt is being deliberately preloaded to a level close to its yield strength.

15

16 The model was solved for the range of possible preloads from zero to 100% of proof strength and the safety factors plotted versus percent preload. The results are shown in Figure 15-29. The fatigue and separation safety factors are 2x the screw diameter (10 threads), which is the minimum recommended length for a steel screw in aluminum threads. For U.S. standard bolts up to 6-in long, the thread length is 2d + 0.25 = 0.875 in of this bolt’s length,[2] which allows the desired penetration. The trial clamp length for the stiffness calculations is then 1.25 in, since the entire cap screw is engaged.

132

1000

p psig

Fg lb 0

180

360

crank angle (deg) F I G U R E 9 - 2 Repeated Pressure and Force Within Cylinder During One Cycle

15

958

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

5 Try an SAE grade 7 screw preloaded to 70% of its proof strength. Table 15-6 shows the proof strength of this screw to be 105 kpsi. The tensile-stress area from equation 15-1 is 0.052 431 in2. The required preload is then

(

)(

)

Fi = 0.7 S p At = 0.7 105 000 0.052 431 = 3853.66 lb

(c)

6 Find the joint aspect ratio and plate to screw modulus from equations 15.18a and b: j=

d 0.313 = = 0.25 1.25 l

E 10.4 E 6 = 0.347 r = material = 30 E 6 Ebolt

(d )

7 Since this joint has the same material (aluminum) throughout, equation 15.19 is all that is needed to calculate the joint constant Cng for the metal without the gasket. The parameters needed for the equation are found by interpolating for j = 0.25 in Table 15-8. They are: p0 = 0.653, p1 = –1.207, p2 = 1.103, and p3 = –0.383: Cng = p3r 3 + p2 r 2 + p1r + p0 = –0.383 ( 0.347 )3 + 1.103 ( 0.347 )2 − 1.207 ( 0.347 ) + 0.653 = 0.351

(e)

8 We can approximate the stiffness of the screw kb’ with equation 15.17, then estimate material stiffness km with no gasket by using equation 15.13c, given kb’ and Cng : length of thread:

lthd = 2d + 0.25 = 2 ( 0.313) + 0.25 = 0.875 in

length of shank:

ls = lbolt − lthd = 1.25 − 0.875 = 0.375 in

length of clamped thread:

lt = l − ls = 1.25 − 0.375 = 0.875 in

d  −1 At Ab  Eb kb ′ ≅  1 +   l  Ab lt + At ls

(f)

0.052 ( 0.077 ) 0.313  −1  kb ′ ≅  1 + 30 E 6 = 1.112 E 6 lb/in   1.25  0.077 ( 0.875) + 0.052 ( 0.375) 1 − Cng kb ′  1 − 0.351  Cng = km1 = kb ′ = 1.112 E 6  = 2.060 E 6 lb/in  0.351  km + kb ′ Cng 1

9 Now consider the gasket. The area of the unconfined gasket subjected to the clamp force can be assumed to be that of one “bolt’s worth” of the total clamped area which extends from the outside diameter of the cylinder head to the bore: Atotal =

15

(

) (

)

π 2 π Do − D12 = 5.6252 − 3.1252 ≅ 17.18 in 2 4 4

( g)

Dividing by the number of bolts and subtracting the screw hole area gives the area of the clamped gasket around any one screw as: Ag =

Atotal π 2 17.18 π − d = − 0.3132 = 2.071 in 2 4 8 4 nb

(h)

10 The stiffness of this piece of gasket is then km2 =

Ag Eg lg

=

2.071(13.5E 6 ) ≅ 4.659 E 8 lb/in 0.06

(i )

Chapter 15

SCREWS AND FASTENERS

959

The modulus of elasticity of the gasket material is found in Table 15-10. 11 The material and gasket stiffness combine according to equation 14.2b. The combined stiffness of the gasketed joint is then 1

k mg =

1 km1

+

1

=

km2

1 1 1 + 2.060 E 6 4.659 E 8

kmg = 2.051E 6 lb/in

( j)

Note that the combined stiffness in this case is dominated by the aluminum because the copper-asbestos gasket is stiffer. 12 The joint constant with the unconfined gasket is C=

kb ' 1.112 E 6 = ≅ 0.352 kmg + kb ' 2.051E 6 + 1.112 E 6

(1 − C ) = 0.648

and

(k )

13 The 1000-lb load is assumed to be divided equally among the 8 bolts at 125 lb each. The portions of the applied load felt by each screw and material (Eq. 15.13) are: Pb = CP = 0.352(125) ≅ 43.95 lb Pm = (1 − C ) P = 0.614(125) ≅ 81.05 lb

(l )

14 The resulting peak loads in screw and material are Fb = Fi + Pb = 3853.66 + 43.95 = 3897.61 lb Fm = Fi − Pm = 3853.66 − 81.05 = 3772.61 lb

(m)

15 The alternating and mean components of force on the screw are Falt = Fmean =

Fb − Fi 3897.61 − 3853.66 = ≅ 21.98 lb 2 2 Fb + Fi 3897.61 + 3853.66 = ≅ 3875.63 lb 2 2

(n)

16 The nominal mean and alternating stresses in the screw are σ anom =

Falt 21.98 = ≅ 419.2 psi At 0.052 431

σ mnom =

Fmean 3875.63 = ≅ 73 919 psi At 0.052 431

(o)

17 The fatigue stress-concentration factor for this diameter thread is found from equation 15.15c and the mean stress-concentration factor Kfm is found from equation 6.17.

15

960

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

K f = 5.7 + 0.6812d = 5.7 + 0.6812 ( 0.313) = 5.9 if K f σ maxnom > S y then:

K fm =

S y − K f σ anom σ mnom

K f σ maxnom = K f σ anom + σ mnom = 5.9 460.2 + 73 960 = 440 039 psi 440 039 psi > S y = 115 000 psi K fm =

S y − K f σ anom σ mnom

=

115 000 − 5.9 ( 460.2 ) 73 960

= 1.52

( p)

18 The local mean and alternating stresses in the screw are Falt 21.98 = 5.9 ≅ 2478 psi At 0.052 431

σa = K f

σ m = K fm

(q)

Fmean 3875.63 = 1.52 ≅ 112 522 psi At 0.052 431

19 The stresses at the initial preload and at the maximum screw force are σ i = K fm

Fi 3853.66 = 1.52 ≅ 111 883 psi At 0.052 431

σ b = K fm

Fb 3897.61 = 1.52 = 113 160 psi At 0.052 431

(r )

20 The endurance strength for this material is found with the methods of Section 6.6:

(

)

Se' = 0.5Sut = 0.5 133 000 = 66 500 psi Se = Cload Csize Csurf Ctemp Creliab Se'

(

(s )

)

= 0.70 (.995)( 0.739 )(1)( 0.753) 66 500 = 25 778 psi

(t )

where the strength-reduction factors are taken from the tables and formulas in Section 6.6 for, respectively, axial loading, the screw size, a machined finish, room temperature, and 99.9% reliability. 21 The corrected endurance strength and the ultimate tensile strength are used in equation 15.16 to find the fatigue safety factor from the Goodman line: Nf =

15

=

Se ( Sut − σ i )

Se ( σ m − σ i ) + Sut σ a

(

(

25 778 133 000 − 111 883

)

)

25 778 112 522 − 111 883 + 133 000 ( 2478 )

≅ 1.6

(u)

22 The static screw stress after initial local yielding and the yielding safety factor are: σs =

Fbolt 3897.61 = = 74 338 psi 0.052 431 At

Ny =

Sy σb

=

115 000 74 338

≅ 1.5

23 The safety factor against joint separation is found from equation 15.14d:

(v)

Chapter 15

N separation =

SCREWS AND FASTENERS

Fi 3853.66 = ≅ 47 P (1 − C ) 125 (1 − 0.352 )

961

( w)

24 The joint will leak unless the clamping forces are sufficient to create more pressure at the gasket than exists in the cylinder. The minimum clamping pressure can be found from the total area of the gasketed joint and the minimum clamping force Fm : pavg =

4 Fm Fm 4 ( 3772.6 ) = = ≅ 228 psi A j π Do2 − Di2 − nb Ab π 5.6252 − 3.1252 − 8(0.077)

(

)

(

N leak =

pavg pcyl

)

(x)

228 = ≅ 1.7 130

This ratio of clamp pressure vs. cylinder pressure makes the screw spacing acceptable. 25 The torque required to obtain the preload of 3853.66 lb found in step 5 is: Ti ≅ 0.21Fi d = 0.21( 3853.66 )( 0.313) ≅ 253 lb-in

( y)

26 This design uses eight 5/16-18 UNC-2A, grade 7 hex-head cap screws, 1.25 in long, preloaded to 70% of proof strength and equispaced on a 4.375-in-dia bolt circle. This design has a safety factor against leakage of 1.7, a fatigue safety factor of 1.6, and can stand 47 times the operating pressure before joint separation will occur. These safety factors are acceptable. The files CASE8D can be found on the book’s website.

15.12 SUMMARY This chapter has dealt with only a small sample of commercially available fasteners. An extremely varied collection of fasteners is available. The “right” fastener can usually be found for any application, and if not (and the required quantity is high enough), some vendor will make a new one for you. Many standards exist that define the configurations, sizes, strengths, and tolerances of fasteners. Threaded fasteners are made to one or another of these standards, which provides good interchangeability. Unfortunately, metric and English threads are not interchangeable, and both are in wide use in the United States. Power screws are threaded devices used primarily to move loads or accurately position objects. They have low efficiency due to their large friction losses unless the ballscrew variety is used, which lowers the friction significantly. However, low-friction screws give up one of their advantages, which is self-locking or the ability to hold a load in place with no input of energy (as in a jack). Back-drivable screws are the opposite of self-locking and can be used as a linear-to-rotary motion converter. Threaded fasteners (bolts, nuts, and screws) are the standard means of holding machinery together. These fasteners are capable of supporting very large loads, especially if they are preloaded. Preloading tightens the fastener to a high level of axial tension before any working loads are applied. The tension in the fastener causes compression in the clamped parts. This compression has several salutary effects. It keeps the joint tightly together and thus able to contain fluid pressure and resist shear loads with its interfacial friction. The compressive forces in the clamped material also serve to protect the fastener from fluctuating fatigue loads by absorbing most of the applied load oscillations. The

15

962

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

high clamping forces also guard against vibratory loosening of the fastener by creating high friction forces in the threads. Threaded fasteners are also capable of resisting shear loads and are used extensively in that manner in structural applications. In machine design it is more common to rely on close-fitted dowel pins to take shear loads and let the threaded fasteners provide the tension to hold the joint together. The interested reader is referred to the publications listed in the bibliography of this chapter for more information on the diverse and fascinating world of fasteners. Important Equations Used in This Chapter See the referenced sections for information on the proper use of these equations. Torque Required to Raise the Load with a Power Screw (Section 15.2):

Tu = Tsu + Tc =

( ) + µ P dc ( π d p cos α − µL ) c 2

Pd p µπ d p + L cos α 2

(15.5a)

Self-Locking of a Power Screw Will Occur if (Section 15.2):

µ≥

L cos α πd p

µ ≥ tan λ cos α

or

(15.6a)

Efficiency of a Power Screw (Section 15.2):

e=

Wout PL = 2 πT Win

(15.7c)

Spring Constant of a Threaded Fastener (Section 15.7):

l l − lt l l 1 = t + = t + s kb At Eb Ab Eb At Eb Ab Eb



kb =

At Ab Eb Ab lt + At ls

d  −1  kb ′ =  1 +  kb  l

(15.11a)

(15.17alt )

Spring Constant of the Clamped Material if Am is known (Section 15.7):

km =

15

Am Em l

(15.11c)

Load Taken by the Preloaded Material (Section 15.7):

Pm =

km P = (1 − C ) P kb + km

(15.13d )

Load Taken by a Preloaded Bolt and the Joint Constant C (Section 15.7):

Pb = or

kb P km + kb

Pb = CP

kb where C = km + kb

(15.13c)

Chapter 15

SCREWS AND FASTENERS

963

Minimum Load in Material and Maximum Load in Bolt (Section 15.7):

Fm = Fi − (1 − C ) P

(15.14 a)

Fb = Fi + CP

(15.14 b)

Load Required to Separate a Preloaded Joint (Section 15.7):

Fi

P0 =

(15.14c)

(1 − C )

Mean and Alternating Loads Felt by a Preloaded Bolt (Section 15.7):

Falt =

Fb − Fi , 2

Fmean =

Fb + Fi 2

(15.15a)

Mean and Alternating Stresses in a Preloaded Bolt (Section 15.7):

σa = K f

Falt , At

σ m = K fm

Fmean At

(15.15b)

Fatigue Stress Concentration Factor in Threads:

or

K f = 5.7 + 0.6812d

d in inches

K f = 5.7 + 0.026 82d

d in mm

(15.15c)

Preload Stress in a Bolt (Section 15.7):

σ i = K fm

Fi At

(15.15d )

Fatigue Safety Factor for a Preloaded Bolt (Section 15.7):

Nf =

Se ( Sut − σ i )

(15.16)

Se ( σ m − σ i ) + Sut σ a

Approximate Torque Needed to Preload a Bolt (Section 15.9):

Ti ≅ 0.21Fi d

(15.23d )

Centroid of a Group of Fasteners (Section 15.10):

∑ 1 Ai xi , x = n ∑ 1 Ai

∑ 1 Ai yi y = n ∑ 1 Ai

n

n

(15.24)

15

Forces on Fasteners Eccentrically Loaded in Shear (Section 15.10):

F1i = F2i =

P n

Mri



n

r2 j =1 j

(15.25a) =

Plri

∑ j =1 rj2 n

(15.25b)

964

Table P15-0†

-

An Integrated Approach

-

Sixth Edition

15.13 REFERENCES

Topic/Problem Matrix

1 Product Engineering, vol. 41, p. 9, Apr. 13, 1970.

15.2 Power Screws

2 H. L. Horton, ed. Machinery’s Handbook, 21st ed. Industrial Press, Inc.: New York. p. 1256, 1974.

15-1, 15-2, 15-3, 15-37, 15-38 15.3 Stresses in Threads

3 ANSI/ASME Standard B1.1-1989, American National Standards Institute, New York, 1989.

15-15, 15-16, 15-39, 15-47, 15-53

4 ANSI/ASME Standard B1.13-1983 (R1989), American National Standards Institute, New York, 1989.

15.6 Strengths of Bolts

5 T. H. Lambert, Effects of Variations in the Screw Thread Coefficient of Friction on Clamping Force of Bolted Connections, J. Mech Eng. Sci., 4: p. 401, 1962.

15-28, 15-29, 15-40, 15-41, 15-48 15.8 Joint Stiffness Material Only 15-17, 15-18, 15-19, 15-31, 15-32, 15-33, 15-42, 15-43 No Gasket-Static Load 15-7, 15-8, 15-23, 15-30, 15-49 No Gasket-Dynamic Load 15-9, 15-10, 15-24, 15-25, 15-27 Unconfined Gasket 15-26, 15-42, 15-43, 15-44, 15-45, 15-46 Different Materials in Joint 15-42, 15-43, 15-44, 15-45, 15-46 15.9 Controlling Preload Torque Only

6 R. E. Peterson, Stress-Concentration Factors. John Wiley & Sons: New York, p. 253, 1974. 7 H. H. Gould and B. B. Mikic, Areas of Contact and Pressure Distribution in Bolted Joints, Trans ASME, J. Eng. for Industry, 94: pp. 864–869, 1972. 8 N. Nabil, Determination of Joint Stiffness in Bolted Connections, Trans. ASME, J. Eng. for Industry, 98: pp. 858–861, 1976. 9 Y. Ito, J. Toyoda and S. Nagata, Interface Pressure Distribution in a Bolt-Flange Assembly, Trans. ASME, J. Mech. Design, 101: pp. 330–337, 1979. 10 J. F. Macklin and J. B. Raymond, Determination of Joint Stiffness in Bolted Connections using FEA, Major Qualifying Project, Worcester Polytechnic Institute, Worcester, Mass., Dec. 31, 1994. 11 B. Houle, An Axisymmetric FEA Model Using Gap Elements to Determine Joint Stiffness, Major Qualifying Project, Worcester Polytechnic Institute, Worcester, Mass., Dec. 31, 1995. 12 J. E. Shigley and C. H. Mischke, Mechanical Engineering Design, 5th ed. McGrawHill: New York, p. 354, 1989. 13 J. Wileman, M. Choudhury, and I. Green, “Computation of Member Stiffness in Bolted Connections,” Trans. ASME, J. Mech. Design, 113, pp. 432–437, 1991.

15-11, 15-12, 15-13, 15-14, 15-54

14 T. F. Lehnhoff, K. I. Ko, and M. L. McKay, “Member Stiffness and Contact Pressure Distribution of Bolted Joints,” Trans. ASME, J. Mech. Design, 116, pp. 550-557, 1994.

No Gasket-Static Load

15 R. E. Cornwell, “Computation of Load Factors in Bolted Connections,” Proc. IMechE, J. Mech. Eng. Sci., 223c, pp. 795-808, 2009.

15-4, 15-5, 15-6, 15-50, 15-56, 15-59 No Gasket-Dynamic Load 15-22, 15-51, 15-52, 15-57, 1560, 15-62

15

MACHINE DESIGN

16 A. R. Kephart, “Fatigue Acceptance Test Limit Criterion for Larger Diameter Rolled Thread Fasteners,” ASTM Workshop on Fatigue and Fracture of Fasteners, May 6, 1997, St. Louis, MO. 17 C. Crispell, “New Data on Fastener Fatigue,” Machine Design, pp. 71-74, April 22, 1982.

Unconfined Gasket 15-20, 15-21, 15-55, 15-58, 15-61

15.14 BIBLIOGRAPHY

15.10 Fasteners in Shear

American Institute of Steel Construction Handbook. AISI: New York.

15-34, 15-35, 15-36

Helpful Hints for Fastener Design and Application. Russell, Burdsall & Ward Corp.: Mentor, Ohio, 1976.

Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash no.



SAE Handbook. Soc. of Automotive Engineers: Warrendale, Pa., 1982. “Fastening and Joining Reference Issue,” Machine Design, vol. 55, Nov. 17, 1983.

Chapter 15

SCREWS AND FASTENERS

965

J. H. Bickford, An Introduction to the Design and Behavior of Bolted Joints, 2nd ed. Marcel Dekker, New York, 1990. H. L. Horton, ed., Machinery’s Handbook, 21st ed. Industrial Press: New York, 1974.

40 mm

R. O. Parmley, ed. Standard Handbook of Fastening and Joining. McGraw-Hill: New York. 1977. H. A. Rothbart, ed., Mechanical Design and Systems Handbook. McGraw-Hill: New York, Sections 20, 21, 26, 1964.

15.15 PROBLEMS 15-1 Compare the tensile load capacity of a 5/16-18 UNC thread and a 5/16-24 UNF thread made of the same material. Which is stronger? Make the same comparison for M8 x 1.25 and M8 x 1 ISO threads. Compare them all to the strength of a 5/16-14 Acme thread. *15-2

A 3/4-6 Acme thread screw is used to lift a 2 kN load. The mean collar diameter is 4 cm. Find the torque to lift and to lower the load using a ball-bearing thrust washer. What are the efficiencies? Is it self-locking?

FIGURE P15-1 Problems 15-4 to 15-6

* Answers

to these problems are provided in Appendix D.

15-3 A 1 3/8-4 Acme thread screw is used to lift a 1-ton load. The mean collar diameter is 2 in. Find the torque to lift and to lower the load using a ball-bearing thrust washer. What are the efficiencies? Is it self-locking? *†15-4

The trailer hitch from Figure 1-1 (p. 11) has loads applied as shown in Figure P15-1. The tongue weight of 100 kgf acts downward and the pull force of 4905 N acts horizontally. Using the dimensions of the ball bracket in Figure 1-5 (p. 14), draw a free-body diagram of the ball bracket and find the tensile and shear loads applied to the two bolts that attach the bracket to the channel in Figure 1-1. Size and specify the bolts and their preload for a safety factor of at least 1.7.

15-5 For the trailer hitch of Problem 3-4, determine the horizontal force that will result on the ball from accelerating a 2000-kg trailer to 60 m/s in 20 sec. Assume a constant acceleration. Size and specify the bolts and their preload for a safety factor of at least 1.7.‡ *15-6

For the trailer hitch of Problem 3-4, determine the horizontal force that will result on the ball from an impact between the ball and the tongue of the 2000-kg trailer if the hitch deflects 1 mm on impact. The tractor weighs 1000 kg. The velocity at impact is 0.3 m/s. Size and specify the bolts and their preload for a safety factor of at least 1.7.‡

*15-7

A 1/2-in-dia UNC, class 7 bolt with rolled threads is preloaded to 80% of its proof strength when clamping a 3-in-thick sandwich of solid steel. Find the safety factors against static yielding and joint separation when a static 1000-lb external load is applied.‡

15-8 An M14 x 2, class 8.8 bolt with rolled threads is preloaded to 75% of its proof strength when clamping a 3-cm-thick sandwich of solid aluminum. Find the safety factors against static yielding and joint separation when a static 5-kN external load is applied.‡ *15-9

A 7/16-in-dia UNC, class 7 bolt with rolled threads is preloaded to 70% of its proof strength when clamping a 2.75-in-thick sandwich of solid steel. Find the safety factors against fatigue failure, yielding, and joint separation when a 1000-lb (peak) fluctuating external load is applied.‡

15-10 An M12 x 1.25, class 9.8 bolt with rolled threads is preloaded to 85% of its proof strength when clamping a 5-cm-thick sandwich of aluminum. Find the safety factors



Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash no.



For these problems, assume that the nut and washer, together, have a thickness equal to the bolt diameter, and assume that bolts are available in length increments of either 0.25 in or 5 mm.

15

966

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

against fatigue failure, yielding, and joint separation when a 2.5-kN (peak) fluctuating external load is applied.‡ to these problems are provided in Appendix D.

*15-11



*15-13

* Answers

Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number.

Find the tightening torque required for the bolt in Problem 15-7.

15-12 Find the tightening torque required for the bolt in Problem 15-8. Find the tightening torque required for the bolt in Problem 15-9.

15-14 Find the tightening torque required for the bolt in Problem 15-10. 15-15 An automobile manufacturer would like a feasibility study of the concept of buildingin electric-motor-powered screw jacks at each end of the car to automatically jack the car wheels off the ground for service. Assuming a 2-ton vehicle with a 60/40 front/ rear weight distribution, design a self-locking screw jack capable of lifting either end of the car. The jack body will be attached to the car frame and the screw will extend downward to engage the ground. Assume a minimum installed clearance of 8 in under the retracted screw in the up position. It must lift the car frame at least 8 additional inches. Use rolling-element thrust bearings. Determine a minimum screw size safe against column buckling. Determine its required lifting torque and efficiency and the power required to lift it to full height in 45 sec. What is your recommendation as to the feasibility of this idea? 15-16 Design a manual screw jack similar to that shown in Figure 15-4 for a 20-ton lift capacity and a 10-cm lift stroke. Assume that the operator can apply a 400-N force at the tip of its bar handle to turn either the screw or nut depending on your design. Design the cylindrical bar handle to fail in bending at the design load before the jack screw fails, so that one cannot lift an overload and fail the screw. Use rolling-element thrust bearings. Seek a safety factor of 3 for thread or column failure. State all assumptions. *15-17

(a) (b) (c) (d) (e)

Table P15-1 Data for Problem 15-19 Row

Member Mat'l (mm)

m8 x 1

steel

30

b

m8 x 1

alum

40

c

m14 x 2

steel

38

d

m14 x 2

alum

45

e

m24 x 3

steel

75

f

m24 x 3

alum

90

a

15

Bolt Thd

Determine the effective spring constant of the following sandwiches of materials under compressive load. All are uniformly loaded over their 10-cm2 area. The first- and third-listed materials are each 10 mm thick and the middle one is 1 mm thick, together making a 21-mm-thick sandwich. aluminum, copper-asbestos, steel steel, copper, steel steel, rubber, steel steel, rubber, aluminum steel, aluminum, steel

In each case determine which material dominates the calculation. *15-18

Determine the effective spring constant of the following sandwiches of materials under compressive load. All are uniformly loaded over their 1.5-in2 area. The first and thirdlisted materials are each 0.4 in thick and the middle one is 0.04 in thick, together making a 0.84-in-thick sandwich. (a) (b) (c) (d) (e)

Copyright © 2018 Robert L. Norton: All Rights Reserved

aluminum, copper-asbestos, steel steel, copper, steel steel, rubber, steel steel, rubber, aluminum steel, aluminum, steel

In each case determine which material dominates the calculation. ‡

For these problems, assume that the nut and washer, together, have a thickness equal to the bolt diameter, and assume that bolts are available in length increments of either 0.25 in or 5 mm.

15-19 A preloaded steel bolt similar to that shown in Figure 15-21 clamps two flanges of total thickness l. Using the data given in the rows assigned in Table P15-1, find the joint stiffness constant.‡ *15-20

A single-cylinder air-compressor head sees forces that range from 0 to 18.5 kN each cycle. The head is 80-mm-thick aluminum, the unconfined gasket is 1-mm-thick Teflon,

Chapter 15

SCREWS AND FASTENERS

967

and the block is aluminum. The effective clamp length of the cap screw is 120 mm. The piston is 75-mm dia and the cylinder is 140-mm outside dia. Specify a suitable number, class, preload, and bolt circle for the cylinder-head cap screws to give a minimum safety factor of 1.2 for any possible failure mode. 15-21 A single-cylinder engine head sees explosive forces that range from 0 to 4000 lb each cycle. The head is 2.5-in-thick cast iron, the unconfined gasket is 0.125-in-thick copperasbestos, and the block is cast iron. The effective clamp length of the cap screw is 3.125 in. The piston is 3-in dia and the cylinder is 5.5-in outside dia. Specify a suitable number, class, preload, and bolt circle for the cylinder-head cap screws to give a minimum safety factor of 1.2 for any possible failure mode. †15-22

The forged-steel connecting rod for the engine of Problem 15-21 is split around the 38-mm-dia crankpin and fastened with two bolts and nuts that hold its two halves together. The total load on the two bolts varies from 0 to 8.5 kN each cycle. Design these bolts for infinite life. Specify their size, class, and preload.‡

*15-23

(See also Problem 4-33.) The bracket in Figure P15-2 is fastened to the wall by 4 cap screws equispaced on a 10-cm-dia bolt circle and arranged as shown. The wall is the same material as the bracket. The bracket is subjected to a static force F, where F and the beam’s other data are given in the row(s) assigned from Table P15-2. Find the forces acting on each of the 4 cap screws due to this loading and choose a suitable cap screw diameter, length, and preload that will give a minimum safety factor of 2 for any possible mode of failure.‡

*15-24

(See also Problem 6-33.) The bracket in Figure P15-2 is fastened to the wall by 4 cap screws equispaced on a 10-cm-dia bolt circle and arranged as shown. The wall is the same material as the bracket. The bracket is subjected to a sinusoidal force-time function with Fmax = F and Fmin = –F, where F and the beam’s other data are given in the row(s) assigned from Table P15-2. Find the forces acting on each of the 4 cap screws due to this fully reversed loading and choose a suitable cap screw diameter, length, and preload that will give a minimum safety factor of 1.5 for any possible mode of failure for N = 5E8 cycles.‡

*15-25

(See also Problem 6-34.) The bracket in Figure P15-2 is fastened to the wall by 4 cap screws equispaced on a 10-cm-dia bolt circle and arranged as shown. The bracket is subjected to a sinusoidal force-time function with Fmax = F and Fmin = 0, where F and the beam’s other data are given in the row(s) assigned from Table P15-2. Find the forces acting on each of the 4 cap screws due to this fluctuating loading and choose a suitable cap screw diameter, length, and preload that will give a minimum safety factor of 1.5 for any possible mode of failure for N = 5E8 cycles.‡

†15-26

(See also Problem 6-42.) A cylindrical, steel tank with hemispherical ends is required to hold 425 kPa of pressurized air at room temperature. The pressure cycles from zero

y

wall

t

z x FIGURE P15-2 Problems 15-23 through 15-25

* Answers

to these problems are provided in Appendix D.



For these problems, assume that the nut and washer, together, have a thickness equal to the bolt diameter, and assume that bolts are available in length increments of either 0.25 in or 5 mm.

15

a

od

Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number.

F

l

t



id

arm

h

968

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Table P15-2 Data for Problems 15-23 through 15-25 Use Only Data Relevant to the Particular Problem—Lengths in mm, Forces in N Row



Problem numbers in italics are design problems. Problem numbers in boldface are extended from similar problems in earlier chapters with the same dash number.

a

100

400

10

20

50

10

4

steel

b

70

200

6

80

85

12

6

steel

c

300

100

4

50

95

15

7

steel

d

800

500

6

65

250

25

15

alum

e

85

350

5

96

900

40

30

alum

f

50

180

4

45

950

30

25

alum

g

160

280

5

25

850

45

40

steel

h

200

100

2

10

800

40

35

steel

i

400

150

3

50

950

45

38

steel

j

200

100

3

10

600

30

20

alum

k

120

180

3

70

880

60

55

alum

l

150

250

8

90

750

45

30

alum

m

70

100

6

80

500

20

12

steel

n

85

150

7

60

820

25

15

steel

to maximum. The tank diameter is 0.5 m and its length is 1 m. The hemispherical ends are attached by some number of bolts through mating flanges on each part of the tank. A 0.5-mm-thick, compressed asbestos, unconfined gasket is used between the 10-mmthick steel flanges. Determine a suitable number, class, preload for, and size of bolts to fasten the ends to the tank. Specify the bolt circle and outside diameter of the flange needed to prevent leakage. A minimum safety factor of 2 is desired against leakage and a safety factor of 1.5 against bolt failure for infinite life.‡ 15-27 Repeat Problem 15-26 using a confined O-ring gasket.‡ 15-28 Calculate the proof load (load that causes a tensile stress equal to the proof strength) for 1/2-13 UNC bolts in each SAE grade listed in Table 15-6. 15-29 Calculate the proof load (load that causes a tensile stress equal to the proof strength) for M20 x 2.50 bolts in each class listed in Table 15-7.



For these problems, assume that the nut and washer, together, have a thickness equal to the bolt diameter, and assume that bolts are available in length increments of either 0.25 in or 5 mm.

15

15-30 Determine the joint stiffness constant for the bolt and members in Problem 15-7.‡ 15-31 Determine the joint stiffness constant for the bolt and members in Problem 15-8.‡ 15-32 Determine the joint stiffness constant for the bolt and members in Problem 15-9.‡ 15-33 Determine the joint stiffness constant for the bolt and members in Problem 15-10.‡ 15-34 Figure P15-3 shows a bolted and doweled joint eccentrically loaded in shear. The shear loads are taken by the dowel pins, the number and size of which are given in Table P15-3. Though the figure shows 5 dowel pins, that is not the case for every row in the table. For a = 4 in, b = 4 in, l = 10 in, P = 2500 lb, and the data in the row(s) assigned from Table P15-3, find the magnitude and direction of the total shear force acting on each dowel. 15-35 Figure P15-3 shows a bolted and doweled joint eccentrically loaded in shear. The shear loads are taken by the dowel pins, the number and size of which are given in Table P15-3. Though the figure shows 5 dowel pins, that is not the case for every row in the

Chapter 15

SCREWS AND FASTENERS

969

l A

B b/2

C

D

b E

a/2 P

a FIGURE P15-3

Copyright © 2018 Robert L. Norton: All Rights Reserved

Problems 15-34 through 15-36

table. For a = 4 in, b = 4 in, l = 10 in, P = 2500 lb, and the data in the row(s) assigned from Table P15-3, find the total shear stress in each dowel. 15-36 Figure P15-3 shows a bolted and doweled joint eccentrically loaded in shear. The shear loads are taken by the alloy steel dowel pins, the number and size of which are given in Table P15-3. Though the figure shows 5 dowel pins, that is not the case for every row in the table. For a = 4 in, b = 4 in, l = 10 in, P = 2500 lb, and the data in the row(s) assigned from Table P15-3, find the factor of safety against yielding in shear for each dowel pin. See Table 15-12 for strength data. 15-37 The coefficient of friction for an oil-lubricated, single-start, power screw thread-nut combination is 0.10. Which of the American Standard Acme Threads in Table 15-3 will be self-locking for this thread-nut combination? Which will be the least likely to back down in the presence of dynamic loading? Which will be the most likely to back down in the presence of a dynamic load? 15-38 The coefficient of friction for an oil-lubricated, single-start, power screw thread-nut combination is 0.20. Which of the American Standard Acme Threads in Table 15-3 will have the greatest efficiency (neglecting thrust-collar friction)?

Table P15-3 Data for Problems 15-34 through 15-36 Dowels with Diameters of Zero Are Not Present Row

Number of Dowels, n

a

5

0.250

0.250

0.250

0.250

0.250

b

4

0.250

0.250

0

0.250

0.250

c

5

0.375

0.250

0.250

0.375

0.250 0.250

d

5

0.375

0.375

0.250

0.250

e

3

0.375

0.375

0

0.375

0

f

3

0.375

0

0

0.375

0.375

Copyright © 2018 Robert L. Norton: All Rights Reserved

15

970

MACHINE DESIGN

* Answers

*15-39

to these problems are provided in Appendix D.

-

An Integrated Approach

-

Sixth Edition

Determine the number of engaged screw threads required to make the total strippingshear area of the engaged threads equal to twice the tensile stress area for each of the fine pitch screw threads in Table 15-2.

15-40 Calculate the ultimate tensile loads (load that causes a tensile stress equal to the tensile strength) for 1/2-13 UNC bolts in each SAE grade listed in Table 15-6. *15-41

Calculate the ultimate tensile loads (load that causes a tensile stress equal to the tensile strength) for M20 x 2.50 bolts in each class listed in Table 15-7.

15-42 3/8-16 UNC bolts and nuts clamp 0.75-in-thick aluminum cover plate to a 0.50-in-thick steel flange. Determine the stiffness factor of the joint at each bolt. 15-43 M14 x 2.0 bolts and nuts clamp a 16-mm-thick aluminum cover plate to a 12-mm-thick steel flange. Determine the stiffness factor of the joint at each bolt. 15-44 M16 x 2.0 bolts and nuts clamp a 50-mm-thick aluminum cover plate to a 30-mm-thick steel flange. Determine the stiffness factor of the joint at each bolt. 15-45 5/16-18 UNC bolts and nuts clamp a 1.625-in-thick cast-iron cover plate to a 1.5-inthick steel flange. Determine the stiffness factor of the joint at each bolt. 15-46 M16 x 1.5 bolts and nuts clamp an 8-mm-thick titanium cover plate to a 8-mm-thick stainless steel flange. Determine the stiffness factor of the joint at each bolt. 15-47 An M12 X 1.25 soft steel nut is assembled with a hardened steel bolt. The nut is 11 mm thick and has a shear yield-strength of 120 MPa. Determine the axial force that will cause stripping failure of the nut if the nut threads fail before the bolt fails. 15-48 Compare the yield loads (load that causes a tensile stress equal to the yield strength) to the proof loads (load that causes a tensile stress equal to the proof strength) for M12 x 1.25 bolts in each class listed in Table 14-7. 15-49 An M16 x 1.50, class 4.8 bolt with cut threads is preloaded to 85% of its proof strength when clamping a 20-mm-thick sandwich of solid steel. Find the safety factors against static yielding and joint separation when a static 3-kN external load is applied. 15-50 A 15-mm-thick steel cap is to be fastened to a 15-mm-thick steel flange with 6 bolts and nuts. The external load on the cap is 30 kN. Size and specify the bolts for a safety factor of at least 1.5 and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-51 Repeat Problem 15-50 with a total external load on the six bolts that varies from 0 to 30 kN per cycle. Design these bolts for infinite life with a factor of safety of at least 1.5. Specify their size, class, preload, and tightening torque.

15

15-52 A 20-mm-thick aluminum cap is to be fastened to a 20-mm-thick aluminum flange with 8 bolts and nuts. The external load on the cap varies from 0 to 40 kN per cycle. Size and specify the bolts for infinite life and a safety factor of at least 1.5 and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-53 An M8 X 1.00 soft steel nut is assembled with a hardened steel bolt. The nut is 8 mm thick and has a shear yield-strength of 140 MPa. Determine the axial force that will cause stripping failure of the nut if the nut threads fail before the bolt fails. 15-54 An M8 x 1.00, class 9.8 bolt is preloaded to 80% of its proof strength. Find the tightening torque required to obtain this condition. 15-55 A cylinder with an unconfined gasket between head and body applies a static load of 15.5 kN to its head. The head is aluminum, the unconfined gasket is 2-mm-thick plain rubber, and the cylinder flange is aluminum. The pressure diameter is 80 mm and the outside

Chapter 15

SCREWS AND FASTENERS

971

flange diameter is 130 mm. The effective clamp length of a cap screw is 80 mm. Determine the safety factor against static failure of the cap screws if there are 12 M8 x 1.25, class 8.8 screws preloaded to 90% of proof strength on a 110-mm-diameter bolt circle. 15-56 A 0.600-in-thick steel cap is to be fastened to a 0.600-in-thick steel flange with 5 bolts and nuts. The external load on the cap is 5500 lb. Size and specify the bolts for a safety factor of at least 1.5 and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-57 A 3/4-in-thick aluminum cap is to be fastened to a 3/4-in-thick aluminum flange with 6 bolts and nuts. The external load on the cap varies from 0 to 6750 lb per cycle. Size and specify the bolts for infinite life and a safety factor of at least 1.5 and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-58 A cylinder with an unconfined gasket between head and body applies a static load of 14 kN to its head. The head is steel, the copper-asbestos, unconfined gasket is 3-mm-thick and the cylinder flange is steel. The pressure diameter is 80 mm and the outside flange diameter is 140 mm. The effective clamp length of a cap screw is 80 mm. Specify a suitable number, class, preload, and bolt circle for the cylinder head cap screws to give a minimum safety factor of 1.5 for any possible failure mode and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-59 Experiments have shown that, with a certain lubricant, the thread coefficient of friction is µ = 0.12 and the clamping coefficient of friction is µc = 0.10 for UNF-2 threads in the range of 1/4 to 1/2 in. Calculate the torque coefficient for the standard bolt sizes in this range when this lubricant is used. 15-60 Two steel machine members will be held together by a single bolt and nut. The total clamped length is 2.75 in. The bolt will be subjected to a fluctuating peak load of 2000 lb. Size and specify the bolt for infinite life and a fatigue safety factor of at least 1.4 and specify the torque required on the bolt to obtain the preload if the threads are lubricated. 15-61 A cylinder with an unconfined gasket between head and body applies a static load of 3150 lb to its head. The head is steel, the copper-asbestos, unconfined gasket is 0.125-in-thick and the cylinder flange is steel. The pressure diameter is 3 in and the outside flange diameter is 5.5 in. The effective clamp length of a cap screw is 3 in. Specify a suitable number, class, preload, and bolt circle for the cylinder head cap screws to give a minimum safety factor of 1.5 for any possible failure mode and specify the torque required on each bolt to obtain the preload if the threads are lubricated. 15-62 Two steel machine members will be held together by a single bolt and nut. The total clamped length is 70 mm. The bolt will be subjected to a fluctuating peak load of 9 kN. Size and specify the bolt for infinite life and a fatigue safety factor of at least 1.3 and specify the torque required on the bolt to obtain the preload if the threads are lubricated.

15

972

MACHINE DESIGN

-

An Integrated Approach

NOTES

15

-

Sixth Edition

WELDMENTS To invent, you need a good imagination and a pile of junk. Thomas a. Edison

16.0

INTRODUCTION

Weldments (welded assemblies) are used in many applications such as machine frames, machine parts, building structures, bridges, ships, vehicles, construction equipment, and many other systems. Our focus here will be on their use in machine design rather than as structural elements in buildings and bridges, though the principles of weldment design are similar across applications. We also do not address welded pressure vessels that see elevated temperatures and can have corrosion issues. ASME publishes detailed codes for these applications. A half-century ago, machine frames were commonly made as grey-iron castings, which have good damping. Now it is as common to see machines with welded steel frames. One reason for the change is the superior stiffness of steel over grey cast iron (30E6 psi versus 15E6 psi). A steel frame can then be lighter than a cast-iron casting and have equivalent stiffness or be the same weight and much stiffer. Unlike structural steel applications such as buildings, which have relatively loose tolerances on their dimensions, machine frames and parts usually have to be made to tight dimensional tolerances. But it is difficult to hold tight tolerances when welding an assembly. This requires (as it does for castings) that the weldment be machined as an assembly once it is welded into the desired configuration on any surfaces whose dimensions are critical. Noncritical surfaces can be left in the as-welded condition. Most metals can be welded though some are easier to weld than others. Low carbon steel is one of the easiest to weld. High carbon and alloy steels are more difficult to weld and, if parts are hardened or cold-rolled before welding to improve strength, the high localized heat of welding will tend to anneal it locally, reducing its strength. For these reasons, it is generally recommended to limit steel weldments to low carbon and low-alloy steels. Aluminum can be welded but requires proper attention to choice of process and This chapter is Copyright © 2018 Robert L. Norton: All Rights Reserved

973

16

16

974

MACHINE DESIGN

Opening page photograph A man gas metal arc welding by William M. Plate Jr., USAF (photograph is in public domain).

Table 16-0 Symbol

-

An Integrated Approach

-

Sixth Edition

Variables Used in This Chapter Variable

ips units

SI units

See

A

amplitude ratio

none

none

Eq. 6.1d

Ashear

shear area of weld

in2

mm2

Eq. 16.4

Aw

shear area per unit length of weld

in

mm

Eq. 16.4

Cf

Coefficient for Sfr equation

none

none

Table 16-4

Exx

min. ultimate tensile strength of electrode

kpsi



fb

bending load per unit length of weld

lb/in

N/mm

Eq. 16-4

fn

normal load per unit length of weld

lb/in

N/mm

Eq. 16-4

fs

shear load per unit length of weld

lb/in

N/mm

Eq. 16-4

ft

torsional load per unit length of weld

lb/in

N/mm

Eq. 16-4

Fr

torsional load per unit length of weld

lb/in

N/mm

Eq. 16-4

Jw

polar 2nd moment per unit length of weld

in3

mm3

Eq. 16.4

N Nf

number of cycles

none

none

Ex. 16-2

fatigue stress safety factor

none

none

Ex. 16-4

Nfr

fatigue stress-range safety factor

none

none

Eq. 16.3

Nyield

static yield-stress safety factor

none

none

Eq. 16.3

R Ser

stress ratio

none

none

Eq. 6.1d

tensile stress-range endurance strength

psi

MPa

Table 16-4

Sers

shear stress-range endurance strength

psi

MPa

Ex. 16-2

Sfr

tensile stress-range fatigue strength

psi

MPa

Eqs. 16.2

Sf r s

shear stress-range fatigue strength

psi

MPa

Ex. 16-4

Sw

section modulus per unit length of weld

in2

mm2

Eq. 16.4

t

weld throat dimension

in

mm

Fig. 16-4

w

weld leg dimension

in

mm

Fig. 16-4

Z

section modulus

in3

mm3

Eq. 4.11d

∆σ σx σ1,2,3 σa σm σ' σmax σmin ∆τ τxy τallow

Table 16-4

stress range

psi

MPa

Eq. 6.1a

normal stress

psi

MPa

Eq. 4.7

principal stresses

psi

MPa

Ex. 16-4

alternating normal stress

psi

MPa

Eq. 6.1b

mean normal stress

psi

MPa

Eq. 6.1c

von Mises effective stress

psi

MPa

Ex. 16-4

maximum applied normal stress

psi

MPa

Eq. 6.1a

minimum applied normal stress

psi

MPa

Eq. 6.1a

shear stress range

psi

MPa

Ex. 16-2

shear stress

psi

MPa

Eq. 4.9

allowable shear stress

psi

MPa

Ex. 16-1, -2

technique. Machine frames are usually made of steel for strength and stiffness. Their large mass is not of concern since they are static. Machine parts that move, such as links, may be better made from aluminum if they experience large accelerations. Many links will nevertheless need to be of steel for strength, stiffness, or wear issues.

Chapter 16

WELDMENTS

The reader should consider this chapter to be but a brief introduction into what is actually a quite fascinating and complicated technology. This short treatment will not make you an expert weldment designer by any means. But it will get you started and point you to the extensive literature and code publications available on the topic.* Several organizations do research in welding and publish recommendations and requirements. Some of these are: American Association of State Highway and Transportation Officials (AASHTO)— http://www.transportation.org/

975

*

The Lincoln Electric Company (www.lincolnelectric.com) offers a short course called “Blodgett’s Welding Design,” which is an excellent introduction to welding processes and weldment design for any engineer.

American Institute of Steel Construction (AISC)—http://www.aisc.org/ American Petroleum Institute (API)—http://www.api.org/ American Society of Mechanical Engineers (ASME)—http://www.asme.org/ American Welding Society (AWS)—http://www.aws.org/ James F. Lincoln Arc Welding Foundation—http://www.jflf.org/ Welding Research Council—http://www.forengineers.org

See the references at the end of this chapter for some of their relevant publications.

16.1

WELDING PROCESSES

Arc welding of metals requires the localized application of sufficient heat to melt the base material while adding compatible filler material to join the two parts. A properly applied weld can be as strong as the material adjacent to it but if improperly done can leave the assembly severely weakened. The heat is typically supplied by bringing an electrode close to or in contact with the surface, causing an arc to jump between electrode and workpiece. The “arc-welding” machine supplies either DC or AC current at sufficient voltage to create the arc which has a temperature of 6000–8000 °F, well above the melting point of steel.† The weld filler metal is supplied either as the electrode itself or as a separate wire fed into the arc and are consumed in the process. A good weld requires fusion of the metal on both sides of the joint with the added weld metal and fusion requires atomic cleanliness. The oxygen in air will contaminate the surface with metal oxide rapidly at these temperatures. The nitrogen in air also can compromise weld quality and trap bubbles in the molten metal as it cools causing porosity that weakens the weld. Moisture in the air or on the metal will cause hydrogen embrittlement and weaken the weld. To prevent contamination of the heated metal, either a flux material is supplied that covers the molten pool of metal with slag while it cools, or a stream of inert gas, such as argon or helium, is used to displace the air. If slag is present, it is chipped off when the weld cools. A good weld also requires that the molten mass penetrate into the parent metal, making the final weld metal a combination of the filler material and the parent material. There will also be a heat affected zone or HAZ formed at the edges of the weld as shown in Figure 16-1. This HAZ can be weaker than the parent material in higher strength steel (above 50 kpsi tensile strength), or stronger and harder than the parent in low-strength steel, which promotes crack formation. Aluminum strength is reduced up to 50% in the HAZ. Figure 16-1 shows typical weld terminology. The toes are at the interface between the weld and the base material on the weld face. The root is at the base of the weld.



Oxyacetylene gas welding is not generally used where highstrength welds are required. It is slow and the gas flame causes too much oxidization that contaminates and weakens the weld. It is sometimes used for field repairs but not generally for new work.

16

976

MACHINE DESIGN

reinforcement

face

-

An Integrated Approach

throat

heat affected zone (HAZ) backing

Sixth Edition

reinforcement

toe

throat

-

weld outline toe penetration

root root opening (a) General weld terminology

note flaw (crack) in throat

HAZ (b) A fillet weld cross section showing the HAZ and a flaw

FIGURE 16-1

Copyright © 2018 Robert L. Norton: All Rights Reserved

Weld Cross Section and Terminology Photo Courtesy of Lincoln Electric Co, Cleveland, OH

Preparation of the parts for welding may involve contouring the edges and leaving a root opening to allow the heat and weld metal to penetrate fully. A root opening may require a backing strip be used to keep the molten puddle in place until it solidifies. The backing can be the same or different material. If the same, it will become welded to the joint. It can be left in place or removed by grinding. The latter is usually recommended if the joint is dynamically loaded as the backing creates stress concentrations. The reinforcement is the amount of weld material that protrudes above the surface of the base metal. This can be left in place for statically loaded joints but should be ground flush to remove the stress concentration at the toes if dynamically loaded. The material in the reinforcement is not considered to contribute to the strength of the weld regardless of loading. The throat dimension, which is used to determine the stress area excludes any weld material outside the part thickness or weld outline. Types of Welding in Common Use Shielded Metal Arc Welding (SMAW), also called “stick welding,” uses discrete lengths of electrode (sticks) that are coated with flux on the outside. As the arc melts the electrode, liquefied flux flows onto the pool to cover and protect it from air contact. This method is commonly used outdoors or for field repairs as it has no gas stream that could be blown away by wind. Flux Cored Arc Welding (FCAW) uses hollow electrode wire with flux trapped in its hollow core. This allows it to be spooled in long lengths. The welding machine has a wire feeder that drives the wire through the welding head at a rate controlled by the welder, making it a continuous and more efficient process. FCAW wire electrodes are available for use either with or without an inert gas stream. The gas stream makes it easier to use indoors, but with the right electrode it can also be used without gas.

16

Gas Metal Arc Welding (GMAW), also called MIG (Metal Inert Gas) welding, uses a wire electrode with no flux. An inert gas is directed onto the weld to displace the air. This makes cleaner welds because of the absence of slag and its necessary cleanup but cannot be used outside if the wind blows over 5 m.p.h. Gas Tungsten Arc Welding (GTAW), also called TIG (Tungsten Inert Gas) or Heliarc welding, uses a nonconsumable tungsten wire electrode. A separate metal filler rod or wire is fed into the molten pool. An inert gas, such as argon, is directed onto the weld to protect it. The original gas used was helium, which gave it the Heliarc name. This method is often used on aluminum, titanium, and magnesium, and for fine repairs. It gives a clean weld but has the same wind limitation outside as does GMAW.

Chapter 16

WELDMENTS

977

Submerged Arc Welding (SAW) uses powdered flux that is applied in such quantity as to bury or “submerge” the entire weld in a blanket so thick that the arc cannot be seen. The operator needs only untinted eye protection. The flux is delivered to the arc as a powder flowing through a tube adjacent to or concentric with the electrode. After the weld cools, the unmelted powder is brushed or vacuumed away and can be reused. The melted slag is chipped away to expose the weld. This process is limited to welds done on a top surface and is best suited to production welds in a shop where the motion and path of the electrode can be controlled automatically by a robot or semiautomatically by guides. It gives a good appearing weld that is free of spatter. Resistance welds are created in thin sheet metal by a similar electrical process. Electrodes pinch the metal sandwich and a high current is passed through, fusing the two materials together in a “spot.” If the electrodes are moved along the parts with the current on, it will create a “seam” weld. A laser can be used to create the needed heat instead of electrodes. No filler material is added in these welds. They are commonly used to weld automotive bodies together as well as sheet metal enclosures and other thin-walled structures but are not used for thicker sections. Why Should a Designer Be Concerned with the Welding Process? It is useful for the design engineer to have a basic understanding of these welding processes and their limitations, just as one needs to understand how a part can (and cannot) be machined on a lathe or milling machine. But most design engineers are not also machinists nor are they often certified welders. So just as engineers don’t try to tell an expert machinist how to make the part, they should leave the detailed decisions on welding process to the expert welder. But they nevertheless must be familiar with its limitations. The design engineer’s task is to design the weldment according to good and accepted engineering practice, size the welds based on the techniques to be explained here (and in the references of this chapter) so that they are safe against failure in expected use, choose the needed strength of weld material, and specify this information on the drawing.

16.2

WELD JOINTS AND WELD TYPES

There are five types of weld joints as shown in Figure 16-2, butt, tee, corner, lap, and edge. The choice of joint type will to some degree be dictated by the desired geometry of the weldment and a given weldment may have several types within it. There are three general types of welds that can be used with one or more of the five joint types: groove welds, fillet welds, and plug/slot welds as shown in Figure 16-3. Groove welds break into two subcategories, having either complete or partial penetration. It is generally recommended that plug and slot welds be avoided as they tend to be weaker than the others. We will focus here on the two subtypes of groove welds and fillet welds. Groove welds are suitable for butt joints, outside corner joints, and edge joints in materials with sufficient thickness. Fillet welds are suitable for tee joints, lap joints, and inside corner joints. Groove welds can have complete joint penetration (CJP) or partial joint penetration (PJP) as shown in Figures 16-3 and 16-4. A CJP groove butt weld statically loaded in tension will be as strong as the thinner of the two materials joined. The strength of a PJP joint depends on the depth of its “throat” as defined in Figure 16-4. PJP welds

16

978

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

butt tee

CJP Groove edge

PJP Groove

corner lap FIGURE 16-2

Fillets

Copyright © 2018 Robert L. Norton: All Rights Reserved

Types of Welded Joints

are typically used on both sides of thick sections where a CJP weld would be larger than necessary. Note that any “reinforcement” in the bead that protrudes outside the surface of the welded parts is not included in the throat measurement. In fatigue loaded applications, unless the stress range is sufficiently low, the reinforcement may need to be ground flush with the material to eliminate stress concentrations in the rippled surface of the weld and at the toes. The total weld throat area is the throat dimension times the length of weld. The fusion area is the area of the junction between the weld and the base material as shown in Figure 16-4.

Plug

Slot Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 16-3 Types of Welds

Fillet welds are defined by their leg dimension w, but the weld strength is limited by the throat dimension t as shown in Figure 16-4. Fillet welds are typically at 45° between two orthogonal parts but can join parts at any angle. If the pieces joined are orthogonal and the fillet at 45°, then the throat width t is 0.707 times the leg width w. Any reinforcement is ignored. The total weld area is the throat width t times the length of weld but the fusion area that determines whether the weld tears away from the base metal is the leg width w times the length of weld on each leg for a fillet weld. The stress on each leg’s fusion area may be the same or different depending on part loading.

reinforcement

throat (t)

CJP

reinforcement fusion area

fusion area leg (w)

16

face

Fillet

reinforcement

fusion area

fusion areas

FIGURE 16-4 Throat Dimensions of Welded Joints Data from Lincoln Arc Welding Co., Cleveland, OH

(t)

root

at

ro

th

throat (t)

PJP

toe

leg (w) Copyright © 2018 Robert L. Norton: All Rights Reserved

Chapter 16

WELDMENTS

979

Joint Preparation Better welds will result if the joint is properly prepared to allow the heat and weld metal to reach and fuse to all portions of the joint area. Unless the sections are thin, the weld joint should be prepared by removing metal from one or both sides of the joint. Several joint shapes are recommended, U, J, and V, as shown in Figure 16-5. The J or U joint detail leaves a small amount of base material in the bottom of the slot to prevent the molten weld metal from running out but is thin enough to allow good penetration. A V groove is easier to machine but may need to have a gap at the bottom to get good penetration. This gap can be closed with a backing strip of metal or ceramic to contain the weld metal until it cools. If the backing strip is the same material as the parts joined, it too will become welded to the joint. It can be removed, but if left in place it should be continuous over the entire length of the joint. If any splices are allowed in the backing strip and if they are not fully welded, high stress concentrations will be created that could cause failure. In like fashion, one can choose to weld a long seam only intermittently rather than put down a continuous bead over its length. But, it is often better to use a continuous weld because every interruption of the bead causes a stress concentration, undesirable especially with dynamic loading as is common in machine parts. Moreover, for the same strength a bead that runs only half the length of a joint must have twice the throat dimension, which quadruples its cross-sectional area. A longer, smaller crosssection bead will use less weld material and be more economical. The designer needs to select and specify the size of the groove to be machined into the parts before welding. Recommended sizes based on part thickness can be found in welding handbooks and specifications such as references 1–4. It is highly recommended that these handbooks and codes be used when designing weldments. If one is designing buildings or bridges, then the AISC and AWS codes must be adhered to. They contain very specific rules and procedures that must be followed when designing structures whose failure can put people’s lives in danger. The machine designer is usually not required to follow these codes but would be wise to do so anyway to be sure of good results.

Square edge

Single bevel

Single V

Single U

Single J

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 16-5 Joint Preparation

Weld Specification Welds and their joint preparation are specified on a drawing using a standard form of welding symbol as shown in Figure 16-6a. At a minimum, it has a reference line and an arrow. There can also be an optional tail at the end opposite the arrow. The arrow points to the joint and the weld symbol defines the type of weld (fillet, CJP, PJP, etc.). Weld symbols above the reference line refer to the side opposite the arrow and those below the line refer to the arrow side. The arrow may point either up or down. The symbols are always read right to left regardless of which end the arrow is on. Figure 16-6b shows some of the possible weld symbols. Refer to AWS A2.4[3] for details.

16.3

PRINCIPLES OF WELDMENT DESIGN

It is just as important to pay attention to the geometry of the weldment as it is to consider the sizes of the welds if one wants the design to succeed. It is also very important to arrange the welds so that a sensible and safe load path exists to take the applied loads to their reaction points. The “force-flow” concept introduced in Chapter 4 applies here.

16

980

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

(a)

(b)

FIGURE 16-6

Copyright © 2018 Robert L. Norton: All Rights Reserved

Standard Welding Symbols

The following rules have been developed by welding design experts over many years and more detail on them can be found in reference 1: 1 Provide a path for applied forces to enter into the section(s) of the weldment that lie parallel to the direction of the applied force. 2 Forces will follow the stiffest path to ground so it is better to have relatively uniform stiffness to distribute reaction loads evenly in the weldment.

16

3 There are no secondary members in weldments. It is essentially one piece. But the welds can be either primary or secondary. A primary weld carries the entire load directly and its failure causes the weldment to fail. Secondary welds just hold the parts together and have low forces on them. 4 Do not put welds in bending if possible. If bending moments exist, try to place welds near locations of zero or low moment, or arrange the welds to handle shear or axial tension/compression. 5 Where possible, do not apply tensile loads across a transverse thickness of the parent

Chapter 16

WELDMENTS

981

material (i.e., “across the grain”). Wrought metals do have a grain in the direction of rolling and are slightly weaker across the grain than with the grain. (See Section 14.3 for a discussion of grain in metals.) Welds loaded in tension on a transverse cross section (see Figure 16-10) may cause “lamellar tearing” of the material under the weld. Shear loads applied to such a surface through a weld are preferable. 6 Where section size changes across a joint, taper the material around the joint to improve force flow and reduce stress concentration.

Figure 16-7 shows alternate designs for a hangar welded to a beam. Part a shows an inferior arrangement in which the hangar is turned 90° and so loads the flange of the beam. Since the stiffest path will carry the bulk of the load, the small length of weld that spans the central web will handle most of the force. If the entire length of the weld is sized to have acceptable stress levels, it may fail because the nonuniform stress distribution across the flange overloads the center section of the weld. The flange compliance near its ends reduces its ability to transfer load to the central web. Note the force flow arrows for each design. Part b shows an acceptable arrangement wherein the load from the hangar is transferred directly into the web of the I-beam through the weld. Figure 16-8 shows a similar assembly that uses a rectangular section beam. In part a, the solution that worked in Figure 16-7 is now inferior because the load applied to the center of the bottom surface of the section must travel across it to reach the side webs that carry the forces into the beam. This weld experiences bending. A better arrangement is shown in part b where the hangar is redesigned to weld directly to the side webs, put the welds in shear, and carry the forces into the sections that lie parallel to the load through the hangar, which is stiffer in bending than the section’s web. Note the force flow arrows for each design. Figure 16-9 shows parts of differing width or thickness welded together. The AWS welding code D1.1 calls for a taper of at least 2.5:1 (≤ 22°) be provided at all such junctions across nonuniform-size butt joints when cyclically loaded.

(a) Poor design—load taken through flange into web and the center of weld takes the brunt of load

(b) Good design—load taken directly into web and whole weld shares the load Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 16-7 Examples of Good and Poor Welded Joints Data from

Lincoln Arc Welding Co., Cleveland, OH

16.4

STATIC LOADING OF WELDS

Compared to the calculation of stresses in general machine parts of complex geometry, calculation of weld stresses is fairly straightforward. If we can avoid loading welds in bending, the loading will typically be direct tension/compression or direct shear. In either

16 (a) Poor design—load taken through bottom into side webs (b) Good design—load taken directly into side webs

FIGURE 16-8

Copyright © 2018 Robert L. Norton: All Rights Reserved

Examples of Good and Poor Welded Joints Data from Lincoln Arc Welding Co., Cleveland, OH

982

MACHINE DESIGN

-

2.5

An Integrated Approach

22°

-

Sixth Edition

2.5

1.0

(a) Tapered width dimension

22° 1.0

(b) Tapered thickness dimension

FIGURE 16-9

Copyright © 2018 Robert L. Norton: All Rights Reserved

Mismatched Joints Must Be Tapered Data from Lincoln Electric Company, Cleveland, OH

case the equation for stress is simple. For axial tension or compression the normal stress was defined in equation 4.7 as: σx =

P A

(4.7)

Direct shear, which is common in welds, is defined in equation 4.9 as: τ xy =

P Ashear

(4.9)

In both cases P is the applied load and A or Ashear is the area of the weld. For a butt joint with a CJP weld in tension or compression, the weld throat area is equal to the cross section of the smaller part joined. For a PJP weld in tension or compression or a fillet weld in tension, compression, or shear, the area is simply the throat dimension t times the length of the weld. Figure 16-4 defines throat size for a variety of welds. Note that whether a fillet-welded tee joint is loaded in tension or shear, the weld will experience shear stress and may also have tensile stress. The fusion areas between weld and base metal may have one or both types of stresses applied depending on loading.

16.5

16

* The yield strength of steel electrode material is usually taken as 75% of Sut.

STATIC STRENGTH OF WELDS

In Chapter 5 we discussed static failure theories in depth and concluded that static failures were due to shear stress and that the distortion energy theory best defines a safe stress level for a ductile material using any desired safety factor. The metals used in welding and the weld filler materials are ductile, so this theory should apply, and it does. However, extensive testing done on welded assemblies over the last half-century by the welding industry has developed good data on allowable strengths of welds and weldments. An electrode’s part number appears as E followed by 4 or 5 digits, the first 2 or 3 of which define its minimum ultimate tensile strength in kpsi and the rest indicate the position in which it can be used and its coating. We note its strength generally as Exx. For example, an E70 electrode has a minimum Sut = 70 kpsi and E110 has a minimum Sut = 110 kpsi.* Note that a given sample may actually exceed the stated value. It is recommended that the strength of the selected electrode be approximately matched to that of the base metal to be welded, and this is a requirement with CJP welds loaded in tension. It can be undermatched (weld metal weaker than base metal) in some cases, particularly if higher strength steels are welded or better crack resistance is

Chapter 16

WELDMENTS

983

lo ng itu di na l

needed. Undermatching is often done with fillet welds. Overmatching (weld metal stronger than base metal) is generally not recommended. Residual Stresses in Welds Welds will always have large residual tensile stresses in them. This is because weld metal expands to about six times its elongation at yield when melted. When it cools, it shrinks by that same factor. No solid metal has 600% elongation to yield, so this means that the weld material is guaranteed to yield in tension as it shrinks, since the solid base metal adjacent to the weld cannot move with it. There will be a balancing compressive stress formed in the adjacent base metal and the residual stress distribution will look like Figure 16-10 when cooled. Using an undermatched (i.e., weaker) weld metal will reduce the residual stress due to its lower yield strength. It is not recommended that a significantly overmatched weld metal be used in any case as this could result in a nonconservative design. Direction of Loading

transverse (a) CJP butt-welded plates

tensile stress

compressive stress

The direction of applied load versus the weld axis has a significant effect on fillet weld strength. Tests show that fillet welds loaded orthogonal (transverse) to the long axis of the weld, have 50% greater strength than the same weld loaded along (longitudinal to) the weld axis (see Figure 16-10a). This is due in part to the effective throat of a transversely loaded fillet weld between two orthogonal parts being at 67.5° versus 45° when loaded longitudinally. The 67.5° plane has 30% more shear area. Nevertheless, the AWS D1.1 code specifies that the effective throat area, defined by the shortest distance from root to weld face, be used for any direction of applied load.[9] However, longitudinal welds have the advantage of allowing more deformation before yield than transverse welds do.[9] Nevertheless, welds are designed against failure by rupture rather than yield because, even though ductile, the weld volume is so small compared to the entire part, the magnitude of weld deformation between yield and fracture is too small to provide any warning of incipient failure.

tensile stress

base metal

weld compressive stress (c) Longitudinal residual stress Copyright © 2018 Robert L. Norton: All Rights Reserved

Allowable Shear Stress for Statically Loaded Fillet and PJP Welds For static loading, the AWS[3] recommends that shear stresses in fillet or PJP welds be limited to 30% of the electrode tensile strength, Exx : τ allow = 0.30 Exx

(b) Transverse residual stress

FIGURE 16-10 Residual Stresses in Welds

(16.1)

This value has a built-in minimum safety factor against fracture ranging from 2.21 to 4.06 for various weld loadings using electrode strengths from E60xx to E110xx in extensive testing.* Table 16-1 shows more detail on these safety factors. These safety factors and equation 16.1 are based on extensive testing done on welded assemblies by AISC[6] and assume that the weld material strength approximately matches that of the base metal in each case.† Equation 16.1 is included in the AWS D1.1 Structural Welding Code.[3] The nominal safety factor for equation 16.1 is usually stated as 2.5, which is within the range of longitudinal weld safety factor values in Table 16-1. Equation 16.1 is a bit unusual in that it compares applied shear stress to tensile ultimate strength as a reference value. It also appears to give a safety factor of 3.33 (the

* Note that E60XX electrodes are now considered essentially obsolete and the lowest strength electrode now in common use is E70XX. † Equation 16.1 has been verified by extensive FEA modeling that shows good correlation with this estimate.[8]

16

984

MACHINE DESIGN

Table 16-1

-

An Integrated Approach

-

Sixth Edition

Safety Factors Against Static Failure When Using Equation 16.1

As reported by Testing Engineers, Inc., 1968 (Data from Reference 16) Factor of Safety Using Stress on Throat Area Equal to 0.3 Tensile Strength of Electrode

Table 16-2

Minimum Fillet Weld Sizes* Base Metal Thickness (T)

Minimum Weld Size

sizes in inches

T ≤ 1/4

1/8

1/4 < T ≤ 1/2 1/2 < T ≤ 3/4 3/4 < T

3/16 1/4 5/16

sizes in mm

T≤6

3

6 < T ≤ 12

5

12 < T ≤ 20

6

20 < T

8

* Data from AWS D1.1 Copyright © 2018 Robert L. Norton: All Rights Reserved

Table 16-3 Minimum Strengths of Some ASTM Structural Steels

16

ASTM Number

Sy kpsi (MPa)

Sut kpsi (MPa)

A36

36 58-80 (250) (400-500)

A572 Gr42

42 (290)

60 (415)

A572 Gr50

50 (345)

65 (450)

A514

100 (690)

120 (828)

Copyright © 2018 Robert L. Norton: All Rights Reserved

Longitudinal Welds

Transverse Welds

Base Metal

Electrode Class

Average

Minimum

Average

Minimum

A36

E60xx

2.88

2.67





A441

E70xx

2.95

2.67

4.62

4.06

A514

E110xx

2.41

2.21

3.48

3.30

Copyright © 2018 Robert L. Norton: All Rights Reserved

reciprocal of 0.30) rather than the stated 2.5. This anomaly is explained by the mixing of shear stress and tensile strength in the equation. Equation 2.5b gives an approximate ratio between ultimate shear strength and ultimate tensile strength of ductile metals of 0.75 to 0.8. Multiplying 0.75 by 3.33 gives 2.5 as a factor against shear rupture. But, it is more convenient to use the stated minimum tensile strength of the electrode and adjust the factor in the equation from 1 / 2.5 to 1 / 3.33 = 0.30. The AWS D1.1 Structural Welding Code defines minimum sizes for welds based on thickness of the material being welded. Some of these data are shown in Table 16-2. The minimum weld size is to ensure that sufficient heat is applied to achieve good fusion.

EXAMPLE 16-1 Design of a Statically-Loaded Fillet Weld Problem

A tee section of 0.5-in-thick by 4-in wide ASTM 36 hot-rolled steel on both legs as shown in Figure 16-11 is to be fillet welded on both sides. Determine the required weld throat size t.

Given

The tee is statically loaded in tension on the center web with P = 16 800 lb applied at the 1-in-dia hole. Strengths of ASTM steels are shown in Table 16-3.

Assumptions

Use matching strength electrode material and run the welds the full width of the parts. The weld is directly loaded and will fail in shear along a 45° plane in its throat.

Solution

1 Table 16-3 gives the tensile strength of this material as 58-80 kpsi. Select an electrode with approximately the same strength as the average of the material strength. Electrodes come in increments of 10 kpsi and the closest available is an E70 with 70 kpsi tensile strength. (E60xx electrodes are now considered obsolete.) 2 Determine the allowable strength based on 30% of the Exx value for this electrode using equation 16.1:

Chapter 16

WELDMENTS

τ allow = 0.30 Exx = 0.30 ( 70 ) = 21 kpsi

985

(a)

3 Determine the shear area in the throat needed to limit the stress to this value: 16 800 P = Ashear Ashear

τ xy = τ allow = 21 kpsi = Ashear =

16 800 21 000

= 0.8 in 2

(b) 0.5"

4 Determine the throat dimension of two full length fillet welds (one on each side of the joint) that will give the required area:

L = 4" P

Ashear = 2 Lt = 2(4)t = 0.8 in 2 t = 0.1 in

(c)

5 Convert this throat dimension t to a leg width w assuming an equal-leg fillet in a 90° tee joint: w=

t

( )

cos 45

=

0.1 = 0.141 in 0.707

(d )

6 Check this against the recommended minimum weld size for this thickness part. Table 16-2 indicates that a 0.5-in-thick part needs at least a 3/16-in weld leg width, so increase the weld’s leg width to 0.187 in. 7 Now check whether the part will fail in the fused base metal. There are two areas of concern, the areas between the welds and the base, labeled areas ’A’, which are in tension, and the areas between welds and web, labeled areas ’B’,which are in shear. Since both are the same total area and the shear strength of the base metal is about half that of its tensile strength, we need only check the shear areas ’B’ against failure. The material’s minimum tensile yield strength from Table 16-3 is 36 kpsi: τ xy =

16 800 P P = = = 11 230 psi A fusion 2 Lw 2 ( 4 ) (0.187)

N yield =

Ss y τ xy

=

36 000 ( 0.577 ) 11 230

=

20 772 11 230

A

0.5"

= 1.85

P/2

P/2

A

areas 'B' web

t

fillet weld

throat base areas 'A' w

leg width

View A-A (e)

This is acceptable, especially since the yield strength is a guaranteed minimum value.

Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 16-11 Examples 16-1 and 16-2

8 Check the part’s strength against tensile failure across the section at the centerline of the 1-in-dia hole: P 16 800 = = 11 200 psi A 3(0.5) S y 36 000 N yield = = = 3.2 σ x 11 200

σx =

(f)

The part is safe against yielding in tension at the hole and against failure in the weld, which limits the design. To complete the design, it needs to be checked for tearout and bearing failure at the hole, and preloaded bolts need to be designed to clamp the base of the tee. These tasks are left to the reader.

16

986

MACHINE DESIGN

16.6

-

An Integrated Approach

-

Sixth Edition

DYNAMIC LOADING OF WELDS

As described in Chapter 6, parts loaded dynamically fail at much lower stress levels than parts loaded statically, In that chapter we learned about fully reversed, repeated, and fluctuating stresses (see Figure 6-6) and also learned that the presence of a mean stress component in addition to an alternating stress component made the situation worse and required that a Goodman-, Gerber-, or Soderberg-line analysis be done (Figures 6-42, 6-43, and 6-44). The mean and alternating stress components σm and σa, the stress range ∆σ, the stress ratio R, and the amplitude ratio A, were defined in equations 6.1a – 6.1d, repeated here for your convenience:

R=

σ min σ max

σ = σ max − σ min

(6.1a)

σa =

σ max − σ min 2

(6.1b)

σm =

σ max + σ min 2

(6.1c) A=

σa σm

(6.1d )

Effect of Mean Stress on Weldment Fatigue Strength

* The 2005 AISC specification[2] states on p. 16.1-400: Extensive test programs using full-size specimens, substantiated by theoretical stress calculations have confirmed the following general conclusions ... : (1) Stress range and notch severity are the dominant stress variables for welded details and beams; (2) Other variables such as minimum stress, mean stress, and maximum stresses are not significant for design purposes; and (3) Structural steels with yield points of 36 to 100 kpsi (250 to 690 MPa) do not exhibit significantly different fatigue strengths for given welded details fabricated in the same manner.

16



Fatigue crack propagation and fatigue life of most weldments is proportional to the third power of the stress range[10]

Weldments loaded dynamically behave in a surprisingly different way than nonwelded parts, which makes the mean stress irrelevant to their potential fatigue failure.* Figure 16-12 shows fatigue test data for both unwelded and welded test specimens. The unwelded specimens were rectangular cross-section, hot-rolled steel bars loaded in axial tension/compression. The welded specimens were pieces of the same bar cut, then joined with transverse CJP butt welds, making their overall geometry, material, and finish the same as the unwelded ones. Samples were tested with axial loading at stress ratios R of 1/4 (fluctuating), 0 (repeated), and –1 (fully reversed). The first two have nonzero mean stress and the last has zero mean stress. Note that nonzero mean stress reduces the strength of unwelded parts as we saw in Figure 6-16. But, the data for welded parts show no significant difference between fully reversed data and data with mean stress added. The stress range (equation 6.1a) is the single determining factor in failure of dynamically loaded weldments. This makes the calculation for dynamically loaded weldments simpler than for nonwelded machine parts. A Goodman-line analysis is not needed here. Instead, the stress range that a weldment sees in the cycle is compared to an acceptable stress-range fatigue strength Sfr obtained from test data.† Are Correction Factors Needed For Weldment Fatigue Strength? Recall from Chapter 6 that the fatigue strength Sf of a steel part is never more than 50% of its ultimate tensile stress Sut and is usually much less due to a number of factors that involve its surface finish, size, type of loading, and other factors (equation 6.6). The data that give the uncorrected endurance limit Se’ = 0.5 Sut come from testing small-diameter, rotating-beam specimens with polished surfaces and these are reported as average values. Thus, the uncorrected value Se’ must be reduced by the factors noted to account for differences in size, surface finish, etc. between the test specimen and your part and by a statistical reliability factor to obtain a corrected endurance strength Se.

WELDMENTS

987

∆σ

∆σ

Chapter 16

FIGURE 16-12 Fatigue-Strength Data for Unwelded and Welded Parts at Various Stress Ratios[4]

Fatigue and endurance strength data for weldments are not obtained from polished laboratory specimens but rather from actual welded assemblies in a variety of configurations. Also, these test specimens are large (think building- and bridge-sized parts) and are made from hot-rolled steel with rough surfaces, residual stresses from the rolling process, and real welds containing stress concentrations and tensile residual stress. So, we don’t have to factor these fatigue test data down to match our parts on the basis of size, surface finish, etc., as our parts are similar to the test specimens in those respects. Effect of Weldment Configuration on Fatigue Strength The fatigue resistance of a weldment also will vary according to the presence or absence of interruptions in the geometry of the assembly and of weld beads, both of which create stress concentrations.* So, the test specimens were made with deliberate stress concentrations in the form of interrupted welds, added stiffeners, and all varieties of weld joints and welds listed above. AISC defined and tested many different welded configurations, and based on the fatigue strength of the base metal, grouped them into eight categories labeled A, B, B’, C, D, E, E’, and F in order of decreasing resistance to dynamic loading. Category A is most resistant and E’ least resistant to fatigue. Note that F is the shear strength of the weld metal itself, and the others are for tensile strength of the fusion area between the weld and base material. Sketches of these configurations and their associated letter categories and fatigue strengths for various numbers of cycles for each letter category are all published in reference 2. The chart is too large to reproduce here in its entirety, but a few selected examples are shown in Figure 16-13.

*

According to Barsom and Rolfe:[10] One of the most sensitive weld details is a fillet weld termination oriented perpendicular (transverse) to the applied cyclic stress field. In this case, fatigue cracking initiates from the toe of the fillet weld and propagates through the adjacent base metal. In fact, the majority of weldrelated fatigue failures initiate at the surface, generally at the weld toe. Note that several examples in this chapter have fillet welds loaded in this manner.

16

988

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Figure 16-13a shows a category A part. Note that it has no welds. This is the strongest category and is in effect a reference against which the other categories can be compared. The relatively low fatigue strength of this specimen (24 kpsi) compared to a rotating beam specimen of the same material (about 30 kpsi) has to do with its much larger size, rough surface, residual stresses from the hot-rolling process, and flame cut edges. Figure 16-13b has uninterrupted welds running the full length or width of the part and is a category B. Figure 16-13c is similar to 16-13b but has attachments added. In the direction of stress, the attachments are short (just the thickness of the attached member), but some stress will flow from the main member into the attachment, creating

Category B Category A

(b)

(a)

Category C

Category B Category B'

Table 16-4

Category C

Reliability Factors Versus 95% for Sd = 0.08 µ

(c )

(d)

Reliability % Creliab

16

50

1.152

90

1.033

95

1.000

99

0.938

99.9

0.868

99.99

0.809

99.999

0.759

Copyright © 2018 Robert L. Norton: All Rights Reserved

Category C

(e)

FIGURE 16-13

Copyright © 2018 Robert L. Norton: All Rights Reserved

AISC Strength Categories for Welded Parts Subject to Fatigue Loading Adapted from [2]

Chapter 16

WELDMENTS

989

a stress concentration and the stress category is therefore reduced to category C. Figure 16-13d can be any one of three categories depending on the base material and whether or not the weld reinforcement is ground flush. Figure 16-13e has very different geometry from 16-13c but is also a category C with a Note C attached that applies if the web size is larger than 0.5 in. Note C is also shown in Figure 16-13e. Note the level of detail in these test specimen types—there are many more in the AISC Specification.* The weld metal itself is nearly always category F except in the case of a transverse CJP weld, which can be a category B or C depending on reinforcement removal. Extensive testing of weldments in each of these categories was done by the Highway Research Board in the 1960s.[11] Testing of 374 beams of 10-foot span and depths of about 15 inches having various welded details were done at two separate university laboratories. The data from both labs had close statistical correlation. Figure 16-14a shows test data for category A beams and Figure 16-14b shows data for category E beams. The three sets of data in each plot are from loadings with different values of minimum stress but with the same stress range. All cluster together showing that the only stress parameter that had any effect was stress range. The maximum, mean, and minimum stresses were not factors in the failures.

Table 16-5a[2] U.S. Coefficient Cf and Ser for Equation 16.2

Regression lines are fitted to the data on log-log axes. The equations are shown on the plots. One has a slope of 3.372 and the other a slope of 2.877. The lines above and below the mean line represent ±2 standard deviations, which includes 95% of the population. The lower band line is taken as the fatigue strength line. Exponential relationships for weldment fatigue strength as a function of number of cycles were developed from the data and are shown below. The average slope of all the data for categories A to E’ was rounded to 1/3 for the design equation 16.2a below. Figure 16-14c shows the S-N diagrams for each of the AISC weldment categories based on these test data. These are different from the S-N diagrams in Chapter 6 in that the ordinate shows stress range ∆σ (labelled as fatigue strength range Sfr) rather than alternating stress σa (denoted as fatigue strength Sf) They also differ in that they do not show average fatigue strength values, but rather use values that are two standard deviations below the average. While not truly minimum values, they are close to minimum because 95% of the population will be within plus or minus two standard deviations from the average, meaning only 2.5% lie below these values. So it is not necessary to apply reliability factors to reduce them further unless one wants to have more than a 95% confidence level in the data, then use Table 16-4.

S fr = Creliab  ≥ Ser  N 

Cf ips

Ser kpsi

A

250 E08 120 E08 61 E08 44 E08 22 E08 11 E08 3.9 E08 150 E10

24.0 16.0 12.0 10.0 10.0 4.5 2.6 8.0

B B' C D E E' F

Copyright © 2018 Robert L. Norton: All Rights Reserved

Table 16-5b[2] SI Coefficient Cf and Ser for Equation 16.2 AISC Category

Cf

Ser

SI

MPa

A

170 E10 83 E09 42 E09 30 E09 15 E09 7.6 E09 2.7 E09 10 E12

165 110 82 69 48 31 18 55

All categories except F have the same slope of 1/3 and their intercepts decrease with each higher category letter. Category F, which is for the weld metal rather than the base metal near the weld, has a shallower slope of 1/6 and a low intercept. Note that the knees where infinite life begins also vary with the category from 2E6 to over 1E7 cycles. These curves are linear on a log-log plot up to the knee so exponential equations can be fitted to them. For all categories except F the allowable fatigue stress range Sfr is 1  Cf  3

AISC Category

B B' C D E E' F

(16.2a)

where N is the required number of stress cycles. CF and Ser (the stress range endurance strength) are shown in Tables 16-5a and b for U.S. and SI units, respectively. These values are for hot-rolled steels with tensile yield strengths from 36 to 110 kpsi.

Copyright © 2018 Robert L. Norton: All Rights Reserved

*

The AISC Specification For Structural Steel Buildings is a free download from www.aisc.org.

16

990

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

For category F, which is for shear stress in the weld metal, it becomes: 1

S frs * The Welding Research Council[12] recommends that the exponent in equation 16.2b be 1/5 rather than 1/6.

 Cf  6 = Creliab  ≥ Sers  N 

and CF and Sers for category F are also shown in Tables 16-5a and b. Creliab is given in Table 16-4.*

(b) Experimental data for Category E

Stress-range fatigue strength Sfr or Sfrs (kpsi)

(a) Experimental data for Category A

16

(16.2b)

60 50

weldment tensile strength range Sfr weld metal shear strength range Sfrs Category

30

A

20

B B' C

10

D

5

F

E

3

E' near horizontal lines are stressrange endurance strength Ser or Sers

2

1 5

10

6

10

10

7

4.0 E7

cycles to failure, N (c) Regression lines fitted to experimental fatigue strength data for all categories

FIGURE 16-14

Copyright © 2018 Robert L. Norton: All Rights Reserved

Experimental Fatigue-Strength Data for Welded Parts in the AISC Categories and for Weld Metal Some data from Ref. 11

Chapter 16

WELDMENTS

991

The values in Tables 16-5 can be reduced by a factor of three for use with aluminum. To use these strength data, calculate your applied stress range ∆σ or ∆τ and determine the safety factor for the fusion area in tension or weld metal in shear, respectively, as: N fr =

S fr σ

or

N frs =

S frs τ

(16.3)

Is There an Endurance Limit for Weldments? Until recently it was assumed that once the knee of a curve in Figure 16-14c was reached, the curve remained horizontal out to infinite cycles for steel and a few other materials such as titanium. This allowed one to use that value as an endurance limit for infinite life as was done in Chapter 6 for unwelded parts. More recent research[12] indicates that the fatigue strength of weldments continues to decline beyond the knee. Figure 16-15 shows sample stress-range fatigue curves for both steel and aluminum from reference 12. These data result from extensive testing done by the Welding Research Council (WRC) on weldment assemblies of various geometries. The WRC has defined categories similar to those of the AISC shown in Figure 16-13 except that WRC has many more categories that are denoted by numbers rather than letters. See reference 12 for their definitions. The numbers assigned to the symbols in Figure 16-15 refer to the strength of that number WRC welding category taken at 2 million cycles. The knee for steel and aluminum in tension is taken at 1E7 cycles, but shear fatigue data (not shown) have the knee at 1E8 cycles. These figures show the tensile-stress-range fatigue strength of steel and aluminum weldments declining with stress cycles beyond the curve’s knee but with a much lower slope. The exponent of the curve’s equation becomes 1/22 beyond the knee.[12] Figure 16-14 also reflects this fact by showing those lines with a small negative slope beyond the knee. We also avoid calling the value at the knee an endurance limit, instead referring to Sers as an endurance strength.

16 (a) Stress-range curves for steel weldments

(b) Stress-range curves for aluminum weldments

FIGURE 16-15 Experimental Tensile-Stress-Range Fatigue-Strength Data for Welded Parts in Various Welding Research Council Categories [12]

992

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

lo ng itu di na l

Fatigue Failure in Compression Loading?

transverse (a) CJP butt-welded plates

tensile stress

compressive stress (b) Transverse residual stress

tensile stress

base metal

weld compressive stress

Another difference between unwelded and welded parts in fatigue has to do with the tensile residual stress in the welds. Recall from Chapter 6 that fatigue failures were stated to be due only to oscillating tensile stress. Compressive stress oscillations could be safely ignored. In fact, we saw in Chapter 15 how to use residual compressive stress in the clamp zone of preloaded bolts to partially “hide” the applied oscillating tensile stress from the tension-loaded bolt. We regard compressive stress as a friend in dynamically loaded, unwelded parts. When welds are introduced to a part this is no longer true. Oscillating compressive stress can also cause fatigue cracks. How can this be? The answer is residual tensile stress. As described earlier and as depicted in Figure 16-10, a weld will always have residual tensile stress at the material’s yield point. Consider two different loading cases on a given welded part whose yield strength is 50 kpsi. In the first case, an oscillating tensile stress that varies between zero and 10 kpsi is applied in the region of the weld. On the first cycle, the stress in the weld area will go above the yield strength. The material will yield locally relieving some of the residual stress, about 10 kpsi of it. When the load cycles back to zero it has only 40 kpsi of residual stress there. The next and all successive cycles will oscillate from 40 to 50 kpsi tensile stress at that point with a stress range of 10 kpsi. Now get a fresh specimen with the same 50 kpsi residual stress in its weld and change the applied load to a compressive oscillation from zero to negative 10 kpsi. The local stress in the weld now goes from 50 to 40 kpsi tensile stress each cycle. Since the phase of the stress oscillation does not matter, this is the same tensile stress oscillation in both cases. Applied compression loading stresses that location with a varying tensile stress and so can develop cracks in the weld. These cracks only grow in the residual tensile stress zone and don’t propagate into the base metal but they still weaken the weld and can cause failure there. Chapter 6 warned against allowing residual tensile stress in fatigue-loaded parts. Unfortunately, the nature of welds guarantees high tensile residual stress. This residual stress can be reduced by hammer-peening as discussed in Section 6.8, but shot peening is less effective on welds.

(c) Longitudinal residual stress Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE 16-10

Repeated

Residual Stresses in Welds

16

EXAMPLE 16-2 Design of a Dynamically-Loaded Fillet Weld Problem

The welded tee section of Figure 16-11 (repeated here) is going to be loaded dynamically with repeated stress ranging from zero to a maximum. Determine the largest repeated tensile load that the welds can safely withstand for infinite life with a fatigue safety factor Nfrs = 1.5.

Given

The tee is 0.5-in-thick by 4-in wide ASTM A36 hot-rolled steel on both legs, and has 3/16-in fillet welds full length on both sides. This matches the minimum weld size specified in Table 16-2.

Assumptions

Matching strength electrode material was used. Shear stress in the weld throat governs the design (see Example 16-1). The load is evenly distributed along the weld length.

Solution

1 A repeated stress will range from zero to a maximum value each cycle. Adapting equation 6.1a, the shear stress range ∆τ is τ = τ max − τ min = τ max

(a)

Chapter 16

WELDMENTS

993

2 The geometry of this weldment is similar to that of Figure 16-13e. Note C in Figure 16-13e indicates that this part is a category C based on its thickness. However, this design has the welds taking the load directly. In that case we must use the fatigue strength for category F, the weld metal itself, in the weld throat area. From Table 16-5a, the tensile stress range endurance strength Sers for category F and infinite life is 8000 psi. 3 Calculate an allowable shear stress range based on Sers and the desired safety factor: τ allow

S 8000 = ers = = 5333 psi 1.5 N frs

(b)

P

4 The part as designed has two 4-in-long 3/16-in-leg fillet welds. We need the throat area for this calculation which is 0.707 as wide as the leg. The load to create the allowable shear stress from equation 4.9 is then: Pmax = τ allow Ashear = 5333 ( 2 )( 4 )( 0.187 )( 0.707 ) = 5641 lb

Pmax 5641 = = 3770 psi A fusion 2 ( 4 )( 0.187 )

A

0.5" (c)

5 Check that the fusion area of weld to base metal is safe. The most likely point of failure is at the weld toe. The weld leg area rather than throat area is used here: σ toe =

0.5" L = 4"

P/2 (d )

A category C weldment has Ser = 10 kpsi from Table 16-5a, giving a safety factor of 10 000 / 3770 = 2.65. The weld throat limits the design, as is typical in fillet welds. 6 We still need to check that this repeated load will not fail the part statically or in fatigue at locations away from the welds. For example, a Goodman line analysis (Chapter 6) should be done for this repeated load applied to the smallest tensile area across the centerline of the 1-in-dia hole and for the tearout shear areas of the same hole. Any fasteners used to attach the base will need to be checked for fatigue under their applied preload (Chapter 15). These tasks are left to the reader.

P/2

A

areas 'B' t

web fillet weld

throat base areas 'A' w

leg width

View A-A Copyright © 2018 Robert L. Norton: All Rights Reserved

16.7

TREATING A WELD AS A LINE

The designer is usually trying to determine the size of weld (cross-section and length) needed to withstand the applied loads. One approach is to assume a weld size, calculate the safety factor and if inadequate, change the weld size assumption, recalculate and iterate this process until satisfied with the result. A more direct and somewhat simpler approach as defined by Blodgett[1] is to treat the weld as a line. Instead of stress, this calculation will give a load/length (lb/in or N/m) number that is easily converted to a required weld cross section area using the allowable stress for the weld from either equation 16.1 for static loads or equations 16.2 for dynamic loads. The weld area and the load are essentially normalized to one unit of weld throat length. Allowable unit load factors relating fillet weld leg width w to electrode strength for static loads are shown in Table 16-6. These are calculated as 0.707(0.30)Exx[5] and so are for 90° corners. Loading on a given weld location is typically one or a combination of direct tension or compression, direct shear, bending, and torsion. The stresses associated with each of these loading conditions were defined in earlier chapters. The stress equations can be converted into fx = load per length of weld throat t as follows:

F I G U R E 1 6 - 1 1 Repeated Examples 16-1 and 16-2

Table 16-6

Data from [5]

Allowable Static Unit Force on Fillet Welds as a Function of Weld Leg Length w. Electrode Allowable Unit Number Force lb/in E60 E70 E80 E90 E100 E110 E120

12 730 w 14 850 w 16 970 w 19 090 w 21 210 w 23 330 w 25 450 w

Copyright © 2018 Robert L. Norton: All Rights Reserved

16

994

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

P Aw V fs = Aw

fn =

direct tension or compression direct shear

M fb = Sw Tc ft = Jw

bending torsion

(16.4)

The units of Aw are area / length = length. Sw is the section modulus Z / length = length2 and Jw is the polar area moment of inertia J / length = length3. This makes fx = load per length of weld throat in all cases. Figure 16-16 shows nine configurations of weldments and equations for calculating the factors Aw, Sw, and Jw. Their application will be shown in an example.

EXAMPLE 16-3 Design of a Statically Loaded Weldment Assembly Problem

The weldment assembly in Figure 16-17 has fillet welds all around between the pipe and each end plate. Determine the weld size needed to withstand a static load P = 2700 lb.

Given

The material is ASTM A36 structural steel and an E70xx welding electrode is used. The Schedule 40 pipe is 4.5-in OD by 0.24-in wall. Dimension a = 15 in and r = 10 in.

Assumptions

The weld takes the load directly. Stress in the weld will limit the design since the fusion areas for fillet welds are larger than their throat areas. Ignore weight of arm and pipe.

Solution

See Figure 16-17.

1 The offset load puts the pipe section and weld in combined bending, torsion, and direct shear at the root of the cantilever beam where the moment and torque are both maximum. The torsion and direct shear are assumed to be uniformly distributed along the weld. The location of highest stress will be on the top of the pipe at the weld toe (labeled A) where the bending stress is maximum tensile. We first need to calculate the unit loads on the weld due to each mode of loading and then find their vector sum. 2 This is Case 9 in Figure 16-16, which shows the direct shear component factor Aw = πd. Use this to find the unit load fs at point A due to direct shear:

16

fs =

P P 2700 = = = 191.0 lb/in Aw πd 4.5π

(a)

3 Find the unit load fb at point A due to the bending moment using Sw from Case 9: fb =

2700 (15) M Pa = = = 2546.5 lb/in Sw π d 2 4 π 4.52 4

(

)

(b)

Chapter 16

WELDMENTS

995

16

FIGURE 16-16 Geometry Factors to Analyze a Weld as a Line (reprinted from [7])

996

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

y pipe

A

ft

fb

fb

fs

x

wall

z FR

ft

FR

y

y r

a A wall z

fb

x

arm

fs FR P

P FIGURE 16-17

Copyright © 2018 Robert L. Norton: All Rights Reserved

Examples 16-3 and 16-4

4 Find the unit load ft at point A due to the torsional moment using Jw from Case 9: ft =

(

)

Tc Pr d 2 2700 (10 ) 4.5 2 = = = 848.8 lb/in Jw π d 3 4 π 4.53 4

(

)

(c)

5 Find the magnitude of the resultant force at point A (the maximum weld load): FR =

fs2 + fb2 + ft2 = 2691 lb/in

(d )

6 This is the load per inch of weld. The throat area of one linear inch of weld is equal to the throat dimension. So, if we set the throat stress equal to the allowable value from equation 16.1, use this unit load and calculate the area needed to achieve that allowable stress, we will define the required throat dimension. From equation 16.1, an E70 electrode has an allowable stress of 0.30(70 000) = 21 000 psi: t=

FR τ allow

=

2691 lb/in 21 000 lb/in 2

= 0.128 in 2 /in

(e)

7 This is the throat dimension, but fillet welds are specified by their leg dimension. Assuming an equal-leg fillet in a 90° joint, the leg dimension will be: w = 1.414t = 1.414 ( 0.128 ) = 0.181 in

16

(f)

8 Specify a 3/16” fillet weld. This meets the minimum weld size specified in Table 16-2 and will have a safety factor of approximately 2.5 based on equation 16.1. 9 The weld between the pipe and arm is stressed at a lower level than the weld at A because the bending moment is zero at the end of the cantilever. It sees only direct shear and torsional shear which are 32% of the stress at A. Use a 3/16” weld here also for consistency of fabrication. It is also the minimum for this wall thickness.

Chapter 16

WELDMENTS

997

EXAMPLE 16-4 Design of a Dynamically Loaded Weldment Assembly Problem

The weldment assembly in Figure 16-17 has fillet welds all around between the pipe and each end plate. Determine the weld size needed to withstand a dynamic load that varies between Fmin = –80 lb and Fmax = 600 lb for infinite life with a fatigue safety factor Nfr = 1.5.

Given

The material is ASTM A36 structural steel and an E70xx welding electrode is used. The Schedule 80 pipe is 4.5-in OD by 0.337-in wall (3.83-in ID). Dimension a = 15 in and r = 10 in.

Assumptions

The weldment is a Category C but the weld may limit as Category F.

Solution

See Figure 16-17.

1 Failure due to dynamic loading of weldments has been shown to depend only on the stress range or oscillation between the minimum and maximum values of stress seen during the cycle.[2] The force range is Fmax – Fmin = 600 –(–80) = 680 lb. 2 This is Case 9 in Figure 16-16, which shows the direct shear component factor Aw = πd. Use this to find the unit load fs at point A due to direct shear. fs =

680 P P = = = 48.1 lb/in Aw πd 4.5π

(a)

3 Find the unit load fb at point A due to the bending moment using Sw from Case 9: fb =

680 (15) M Pa = = = 641.3 lb/in Sw π d 2 4 π 4.52 4

(

)

(b)

4 Find the unit load ft at point A due to the torsional moment using Jw from Case 9: ft =

(

)

Tc Pr d 2 680 (10 ) 4.5 2 = = = 213.8 lb/in Jw π d 3 4 π 4.53 4

(

)

(c)

5 Find the magnitude of the resultant unit force at point A: FR =

fs2 + fb2 + ft2 = 678 lb/in

(d )

This is the load per inch of weld length. 6 From Table 16.5a, a Category F joint has a shear stress-range endurance strength Sers = 8000 psi. Apply the safety factor to this strength to get an allowable stress: τ allow =

Sers 8000 = = 5333 psi 1.5 N fr

(e)

7 The throat area of one linear inch of weld length is equal to the throat dimension. So, if we set the throat stress equal to the allowable value from equation (e), use this unit load and calculate the unit area needed to achieve that allowable stress, we will define the required throat dimension.

16

998

MACHINE DESIGN

-

An Integrated Approach

t=

FR

=

τ allow

678 lb/in 5333 lb/in 2

-

Sixth Edition

= 0.127 in 2 /in

(f)

8 This is the throat dimension (in), but fillet welds are specified by their leg dimension. Assuming a 45° fillet, the leg dimension will be: w = 1.414t = 1.414 ( 0.127 ) = 0.180 in

( g)

9 Specify a 3/16” (0.187”) fillet weld to obtain the specified safety factor of 1.5. This meets the minimum weld size specified in Table 16-2. 10 The weld between the pipe and arm has lower stress than the weld at A because the bending moment is zero at the cantilever end. That weld sees only direct shear and torsional shear which are 32% of the stress at A. Table 16-2 gives a minimum 3/16” weld for this wall thickness—the same size as at A, which simplifies fabrication. 11 We also need to check that the stress in the fusion area between weld and pipe will not fail. Remember that only the stress range is of concern with weldments and it is due to the range of force oscillation, here 680 lb from step 1. Find the normal bending stress range and torsional shear stress range on point A using equations 4.11b and 4.23b, respectively: σx =

680 (15)( 2.25) Mc ( Pa )c = = = 2388 psi I I π 4.54 − 3.834 64

τ xz =

Pr c 680 (10 )( 2.25) Tr = = = 796 psi J J π 4.54 − 3.834 32

(

( )

)

(

)

(h)

(i )

12 Find the maximum shear stress, principal stresses, and von Mises stress that result from this stress combination using equations 4.6 and 5.7c: 2 2  σ − σz   2388 − 0  τ max =  x + τ 2xz =  + 7962 = 1435 psi      2  2

σ1 =

σx + σz 2

+ τ max =

2388 + 1435 = 2629 psi 2

σ2 = 0 σ + σz 2388 σ3 = x − τ max = − 1435 = −241 psi 2 2

σ ' = σ12 − σ1σ 3 + σ 32 = 2 6292 − 2 629 ( −24 ) + ( −24 )2 = 2 757 psi

( j)

(k )

13 Assume this assembly is an AISC Category C. Table 16-5a shows the infinite-life fatigue strength for a Category C weldment as 10 000 psi. The safety factor is then:

16

Nf =

Sf σ'

=

The weld material limits this design.

10 000 2757

= 3.62

(l )

Chapter 16

16.8

WELDMENTS

999

ECCENTRICALLY LOADED WELD PATTERNS

Weldments are often used to support offset or eccentric loads as in Examples 16-3 and 16-4. There are many other common arrangements as shown in Figure 16-16. Some of these such as the one shown in Figure 16-18 require that the centroid of the weld pattern be found. This is the same procedure as was described in Section 15-10 and Example 15-6 where a pattern of bolts and dowels was used to support an eccentric load and its centroid was needed. The arrangement of Figure 16-18 is defined in part 5 of Figure 16-16. An equation for the location of the weld pattern’s centroid is defined there as well as expressions for the unit loads due to direct shear and either bending or torsional loading depending on the location of the applied force being either in- or out-of-plane. This application will be shown in an example.

EXAMPLE 16-5 Design of an Eccentrically Loaded Weldment Assembly Problem

The weldment assembly in Figure 16-18 has fillet welds on three sides of the joint. Determine the weld size needed to withstand a static load of P = 4000 lb. Check for yielding in the beam also.

Given

The material is ASTM A36 structural steel and an E70xx welding electrode is used. The plates are 1/2” thick. Dimension a = 12”, b = 3”, and d = 6”.

Assumptions

Consider the weld as a line. The weld pattern matches that of part 5 of Figure 16-16. The weld will limit the design as Category F.

Solution

See Figure 16-18.

1 Find the centroid of the weld pattern with the equation given in Figure 16-16: x=

32 9 b2 = = = 0.75 in 2b + d 2 ( 3) + 6 12

(a)

2 Find the radius from applied load to centroid: r = a + b − x = 12 + 3 − 0.75 = 14.25 in

(b)

3 The weld joint is loaded in direct shear and also has a torsional moment about its centroid. Find the unit load in the weld due to direct shear: fs =

4000 P P = = = 333.3 lb/in Aw 2b + d 6 + 2 ( 3)

(c)

4 As far as the weld is concerned, the torsional moment acts about its centroid which is at radius r from the applied load P: T = Pr = 4000 (14.25) = 57 000 lb-in

(d )

5 The torsional geometry factor for this weld versus the centroid from Figure 16-16 is: Jw =

( 2b + d )3 12



b 2 ( b + d )2 (12 )3 9 ( 9 )2 = − = 83.25 in3 2b + d 12 12

(e)

16

1000

MACHINE DESIGN

-

An Integrated Approach

b

Sixth Edition

P

A

x

-

f th

A centroid of weld pattern

d

ftv FR fs

a r

(b) Vector sum of unit forces at point A

(a) Weld geometry

FIGURE 16-18

Copyright © 2018 Robert L. Norton: All Rights Reserved

Example 16-5

6 The highest stress in the weld due to the torsional moment will occur at the point furthest from the centroid, labelled A in Figure 16-18. It will be simplest if we compute the horizontal and vertical components at point A and then resolve them in combination with the vertical direct shear component into a resultant unit force. For the horizontal component, the radius from the centroid is d / 2. For the vertical component, the radius from the centroid is r – a: fth = ftv = FR =

Td 2 Jw

=

( ) = 2054.1 lb/in

57 000 6 2 83.25

T (r − a ) = Jw

57 000 (14.25-12 ) 83.25

(

ft2h + ftv + fs

)

2

= 1540.5 lb/in

(f)

= 2054.12 + (1540.5 + 333.3)2 = 2780 lb/in

7 The unit area of the throat will then be t=

FR τ allow

=

2780 lb/in

(

)

0.30 70 000 lb/in 2

= 0.132 in 2 /in

( g)

using the Sut of E70 electrode and equation 16.1 for τallow. The leg dimension is w = 1.414t = 1.414 ( 0.132 ) = 0.187 in

(h)

8 The minimum recommended weld size for 1/2-in plate is 3/16”, which matches this number. Use a 3/16” fillet weld.

16.9 16

DESIGN CONSIDERATIONS FOR WELDMENTS IN MACHINES

Weldments can be a practical choice for odd-shaped subassemblies that locate and support machine parts. They may not prove to be an economic choice if not properly designed, however. A significant fraction of their cost is in fixturing and setup of the pieces to hold them in position while welded. In some cases, it can be less expensive to machine a complex part from solid billet with modern numerical control (CNC) machines, which can machine complex shapes defined with solid-modeling CAD systems rapidly and relatively inexpensively and can run unattended after being programmed, creating parts while you sleep. The CNC tool path information generated from the CAD model can

Chapter 16

WELDMENTS

1001

be sent directly to the machining center electronically and the part made in one or a few setups with very little operator interaction. Welded assemblies, on the other hand, often are labor-intensive. It may be worth seeking quotes on direct machining of parts from billet that were initially designed as weldments. That said, what can be done during its design to reduce the cost of a weldment? It may be possible to design the weldment assembly to be fully or partially “self-fixturing.” The degree that the individual pieces to be welded are designed to fit together easily in the correct orientation and alignment for welding as opposed to requiring external fixtures to be made to hold them will reduce cost. Obviously, minimizing the number and size of welds will also reduce cost. Be cautious about using weldments for parts subject to dynamic loading. Try to place welds in locations of lower stress when possible. The residual tensile stress in all welds is a danger when oscillating stress is applied to those regions. If that situation cannot be designed around, then weld reinforcements and any backing strips should be removed and weld surfaces ground smooth to reduce stress concentration at the toes and on the weld surface. Hammer peening the welds will reduce tensile residual stress and improve fatigue resistance, but this can add significant cost. Parts machined from solid billet will often prove less expensive than a weldment in these cases and should be investigated as an alternative. If quantities permit, forging will give parts with better fatigue resistance, but high tooling cost eliminates this option for small batches that are typical with custom machine parts. If a weldment requires post-weld machining, it will probably be necessary to thermally stress relieve the weldment before machining it. This will relieve the residual stresses globally and improve part life. Otherwise the part will likely distort when metal removal during machining releases the residual stress locally where machining is done.

16.10 SUMMARY This brief introduction to the design of weldments probably raises as many questions in the mind of the designer as it provides answers. Welding is a complex subject and is based on a large body of research and experimental data. It has advanced tremendously since its application to ship construction in World War II. Some of the problems encountered then are described in Chapters 5 and 6. If the reader needs to design weldments, they are strongly encouraged to delve more deeply into the subject than can be attempted here. See the references noted. One of the most interesting results of extensive weldment fatigue testing is the independence of failure to the presence of mean stress, which is a serious issue in fatigue failure of unwelded parts. This simplifies the weldment designer’s task since it eliminates the need for a Goodman line analysis. Treating the weld as a line greatly simplifies the task of determining appropriate weld sizes for various loading situations. The existence of design codes both simplifies and complicates the designer’s task. It simplifies by providing rules and guidelines for design. But it also complicates the situation because those rules are far from simple and require much effort to fully understand and use properly. The AISC provides design guidelines and failure stress data based on extensive experiment that should be used for weldment design. The AISC and AWS specifications and codes should be studied by the serious designer.

16

1002

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Weldments work best when made of low-carbon hot-rolled steel and statically loaded, which is typical of machine frames and structures. Parts that are highly stressed or loaded dynamically may not be not good candidates for weldment design. One does not find engine crankshafts and connecting rods made as weldments, for example. Those highly loaded, dynamically stressed parts are usually hot forged. Important Equations Used in This Chapter Allowable Shear Stress in Statically Loaded Weldments (Section 16.5):

τ allow = 0.30 Exx

(16.1)

Tensile Stress-Range Fatigue Strength—Cat. A-E (Section 16.6): 1

S fr

 Cf  3 = Creliab  ≥ Ser  N 

(16.2a)

Shear Stress-Range Fatigue Strength—Cat. F (Section 16.6): 1

S frs

 Cf  6 = Creliab  ≥ Sers  N 

(16.2b)

See Table 16-5 for values of Cf and Ser, Sers. Safety Factor in Fatigue for Weldments

N fr =

S fr σ

or

N frs =

S frs τ

(16.3)

16.11 REFERENCES 1 O. Blodgett, Design of Weldments, J. F. Lincoln Foundation, Cleveland, OH, 1963.

Table P16-0† Topic/Problem Matrix 16.5 Static Strength of Welds

3 American Welding Society, Miami, FL.

16-1, 16-2, 16-3, 16-13, 16-17

4 SAE Fatigue Design Handbook AE-22 3ed., p. 95, Society of Automotive Engineers, Warrandale. PA, 1997.

16.6 Dynamic Strength of Welds

5 O. Blodgett, “Stress Allowables Affect Welding Design,” J. F. Lincoln Foundation, Cleveland, OH, 1998.

16-5, 16-7, 16-9, 16-14, 16-18

6 T. R. Higgins and F. R. Preece, “Proposed Working Stresses for Fillet Welds in Building Construction,” AISC Engineering Journal, Vol. 6, No. 1, pp. 16-20, 1969.

16.7 Weld as a Line

16

2 ANSI/AISC 360-05, Specification for Structural Steel Buildings, p. 16-1–400, American Institute of Steel Construction, March 9, 2005.

16-6, 16-7, 16-8, 16-9, 16-19 16.8 Eccentrically Loaded Weld Patterns 16-4, 16-5, 16-15, 16-16, 16-20 Problem numbers in italics are design problems.



7 R. L. Mott, Machine Elements in Mechanical Design, 4ed., p. 786, Prentice-Hall, Upper Saddle Brook, NJ, 2004. 8 M. A. Weaver, “Determination of Weld Loads and Throat Requirements Using Finite Element Analysis with Shell Element Models—A Comparison with Classical Analysis,” Welding Research Supplement, pp. 1s-11s, 1999. 9 D. K. Miller, “Consider Direction of Loading When Sizing Fillet Welds,” Welding Innovation, Vol. XV. No. 2, 1998.

Chapter 16

WELDMENTS

1003

10 J. M. Barsom and S. T. Rolfe, Fracture and Fatigue Control in Structures, 3ed., pp. 238, 269, Prentice-Hall, Upper Saddle Brook, NJ, 1999. 11 J. W. Fisher et al. , Effect of Weldments on the Fatigue Strength of Steel beams, National Cooperative Highway Research Program Report 102, Highway Research Board, National Research Council, 1970.

P 2"

P

0.25"

12 Recommendations for Fatigue Design of Welded Joints and Components, WRC Bulletin 520, The Welding Research Council Inc., 2009. bar

6"

16.12 PROBLEMS *16-1

A submerged-arc complete joint penetration (CJP) butt weld was made between two sections of A36 hot-rolled steel plate. The plate is 10-in wide by 1/2-in-thick. E70 electrode was used. How much tensile load can the assembly withstand across the weld without yielding in either base metal or weld?

3"

16-2 The plate of problem 16-1 has partial joint penetration welds applied from each side. Each weld throat is 1/4 in. What is the maximum allowable tensile load across the weld. *†16-3

A tee bracket similar to Figure 16-11 has 1/2-in-thick A572 Grade 42 steel welded with a 3/16 in fillet weld along both inside corners using an E70 electrode. It will be subjected to a 20 kip tensile load on the leg of the tee. Determine the minimum required length L of the bracket based on full-length welds.

16-4 Figure P16-1 shows a bar welded to a base with 3/16 in fillet welds on three sides using an E70 electrode. Material is A572 Grade 50 hot-rolled steel. What is your recommended maximum static load P that can safely be applied? *16-5

*†16-6

*†16-7

Figure P16-1 shows a bar welded to a base with 3/16 in fillet welds on three sides using an E70 electrode. Material is A572 Grade 50 hot-rolled steel. What is your recommended maximum dynamic repeated load, zero to Pmax that can be applied for 10E8 cycles with a safety factor of 1.6? Figure P16-2 shows a bracket welded to a wall with a fillet weld using an E70 electrode. For the row(s) assigned in Table P16-1, determine the fillet weld size needed between the tube and wall for a static load F and h = 1.2OD, a = 2OD, and l = 2.5OD. The pipe and wall material are A36 steel.

base Copyright © 2018 Robert L. Norton: All Rights Reserved

FIGURE P16-1 Problems 16-4 and 16-5

Table P16-1 Data for Problems 16-6 to 16-7 Lengths in inches, forces in kip. Row F

Figure P16-2 shows a bracket welded to a wall with a fillet weld using an E70 electrode. For the row(s) assigned in Table P16-1, determine the fillet weld size needed between the tube and wall for a dynamic load that varies from – 0.1F to + 0.2F. The pipe and wall material are A36 steel, h = 1.2OD, a = 2OD, and l = 2.5OD. Use a safety factor of 1.5.

x

id

wall

FIGURE P16-2 Problems 16-6 to 16-7

arm

b

a

2.5

3.500

3.068 1/2

b

3.8

4.500

4.026 1/2

c

4.8

5.563

5.047 1/2

d

8.0

6.625

6.065 1/2

e

9.0

8.625

7.981 3/4

f 11.0 10.750 10.020 3/4

to these problems are provided in Appendix D.

z

pipe

b

* Answers

a od

ID

Copyright © 2018 Robert L. Norton: All Rights Reserved

y

F

l

OD

h

Copyright © 2018 Robert L. Norton: All Rights Reserved



Problem numbers in italics are design problems.

16

1004

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

A

75

support

75

425

75

B 200

300

b

P FIGURE P16-3

P Copyright © 2018 Robert L. Norton: All Rights Reserved

Problems 16-8 and 16-9 * Answers

to these problems are provided in Appendix D.

†16-8

Figure P16-3 shows a bracket machined from 12-mm-thick A572 Grade 50 hot-rolled steel flat stock. It is welded to a support with a fillet weld all around using an E80 electrode. Determine the fillet weld size needed between the bracket and the support for a static load of P = 12 kN.

*†16-9

Figure P16-3 shows a bracket machined from 12-mm-thick A572 Grade 50 hot-rolled steel flat stock. It is welded to a support with a fillet weld all around using an E90 electrode. Determine the fillet weld size needed between the bracket and the support for a dynamic load that varies from 0 to +3 kN using a safety factor of 1.8.



Problem numbers in italics are design problems.

16-10 Two 8-mm-thick by 50-mm-wide A572 Grade 42 steel straps are welded together with fillet welds in a lap joint using E70 electrode. The tensile load of 45 kN on the straps is orthogonal (transverse) to the welds. What is your recommended weld size for the two full-length welds? 16-11 Two 8-mm-thick by 50-mm-wide A572 Grade 42 steel straps are welded together with fillet welds in a lap joint using E70 electrode. Determine the fillet weld size needed for a dynamic load that varies from 0 to +12 kN using a safety factor of 1.5 for infinite life of the two full-length welds. 16-12 Two 12-mm-thick by 50-mm-wide weldable aluminum straps are welded together with fillet welds in a lap joint using an aluminum electrode. Determine the fillet weld size needed for a dynamic load that varies from 0 to +5 kN using a safety factor of 2.0 for infinite life of the two full-length welds. 16-13 Two 10-mm-thick by 60-mm-wide, hot rolled SAE 1050 steel plates are to be butt-welded together with a CJP weld. What electrode part number would you recommend if the weldment is loaded in tension? 16-14 Calculate and plot the 106-cycle stress-range fatigue strength for AISC Categories A through E’ at a reliability of 99%. Use SI units.

16

16-15 Figure P16-1 shows a bar welded to a base on three sides with 4-mm fillet welds using an E80 electrode. The material is SAE 1035 hot-rolled steel. Determine the safety factor against static failure if the load P = 5 kN. 16-16 Figure P16-1 shows a bar welded to a base on three sides with 4-mm fillet welds using an E80 electrode. The material is SAE 1035 hot-rolled steel. Determine the safety factor against dynamic failure for 10E8 cycles if the repeated load varies from zero to 2.5 kN

Chapter 16

WELDMENTS

1005

16-17 A tee bracket similar to that shown in Figure 16-11 is to have 1/2-in-thick A572 Grade 50 steel welded with fillet welds along both inside corners. The length of the bracket is L = 6 in. If it will be subjected to a 30-kip tensile load on the leg of the tee, specify a suitable electrode and weld size for a minimum factor of safety of 1.4 against any failure mode in the weld. 16-18 Figure P16-3 shows a bracket machined from 16-mm-thick A572 Grade 42 hot-rolled steel flat stock. It is welded to a support with a fillet weld all around. Determine the fillet weld size needed between the bracket and the support for a dynamic load that varies from 0 to +700 lb using a safety factor of 1.8. 16-19 Case 6 of Figure 16-16 shows a channel section welded to a wall with a fillet weld using an E70 electrode. Determine the fillet weld size needed between the channel and the wall for a static bending load P = 3200 lb. The channel and wall material are A36 steel and a = 6 in, b = 3 in, and d = 1.410 in. 16-20 Case 6 of Figure 16-16 shows a channel section welded to a wall with a fillet weld using an E70 electrode. Determine the fillet weld size needed between the channel and the wall for a static torsion load P = 3200 lb at a = 16 in. The channel and wall material are A36 steel and b = 3 in and d = 1.410 in.

16

1006

MACHINE DESIGN

-

An Integrated Approach

NOTES

16

-

Sixth Edition

CLUTCHES AND BRAKES A big book is a big nuisance. CallimaChas, 260 bC

17.0

INTRODUCTION

Clutches and brakes are essentially the same device. Each provides a frictional, magnetic, hydraulic, or mechanical connection between two elements. If both connected elements can rotate, then it is called a clutch. If one element rotates and the other is fixed, it is called a brake. A clutch thus provides an interruptible connection between two rotating shafts as, for example, the crankshaft of an automobile engine and the input shaft of its transmission. A brake provides an interruptible connection between one rotating element and a nonrotating ground plane as, for example, the wheel of an automobile and its chassis. The same device may be used as either clutch or brake by fixing its output element to a rotatable shaft or by fixing it to ground. Brakes and clutches are used extensively in production machinery of all types, not just in vehicle applications where they are needed to stop motion and allow the internalcombustion engine to continue turning (idling) when the vehicle is stopped. Clutches also allow a high-inertia load to be started with a smaller electric motor than would be required if it were directly connected. Clutches are commonly used to maintain a constant torque on a shaft for tensioning of webs or filaments. A clutch may be used as an emergency disconnect device to decouple the shaft from the motor in the event of a machine jam. In such cases, a brake will also be fitted to bring the shaft (and machine) to a rapid stop in an emergency. To minimize injuries, many U.S. manufacturers require their production machinery to stop within one or fewer revolutions of the main driveshaft if a worker hits the “panic bar” that typically spans the length of the machine. This can be a difficult specification to achieve on large machines (10 to 100 feet long) driven by multihorsepower electric motors. Manufacturers provide clutch-brake combinations in the same package for such applications. Applying power disengages the brake and engages the clutch, making it fail-safe. Fail-safe brakes are engaged (typically by inter1007

17

1008

MACHINE DESIGN

Title page photograph courtesy of the Logan Clutch Corporation, Cleveland, Ohio.

Table 17-0

-

An Integrated Approach

-

Sixth Edition

Variables Used in this Chapter

Symbol

Variable

ips units

SI units

See

a

length

in

m

Sect. 17.6

b

length

in

m

Sect. 17.6

c

length

in

m

Sect. 17.6

d

diameter

in

m

various

F

force

lb

N

various

Fa

applied force

lb

N

Sect. 17.6

Ff

friction force

lb

N

Sect. 17.6

Fn

normal force

lb

N

Sect. 17.6

Rx

reaction force

lb

N

Sect. 17.6

Ry

reaction force

lb

N

Sect. 17.6

K

arbitrary constant

none

none

various

l

length

in

m

various

M

moment

lb-in

N-m

various

N

number of friction surfaces

none

none

Eq. 17.2

P

power

hp

Watts

Ex. 17-1

p

pressure

psi

N/m2

Sect. 17.4

pmax

maximum pressure

psi

N/m2

Sect. 17.4

r

radius

in

m

various

ri

inside radius of disk lining

in

m

various

ro

outside radius of disk lining

in

m

various

T

torque

lb-in

N-m

various

V

linear velocity

in/sec

m/sec

Eq. 17.4

W

wear rate

psi-in/sec

Pa-m/sec

Eq. 17.4

w

width

in

m

various

θ

angular position

rad (deg)

rad (deg)

Sect. 17.6

µ

coefficient of friction

none

none

Eq. 17.2

ω

angular velocity

rad/sec

rad/sec

Ex. 17-1

nal springs) unless power is applied to disengage them. Thus, they “fail safe” and stop the load if the power fails. Highway truck and railroad-car air brakes are of this type. Air pressure releases the brake, which is normally engaged. If the railroad car or truck trailer breaks loose and severs its air-hose connection to the engine or tractor, the brakes automatically engage.

17

This chapter will describe a number of types of commercially available clutches and brakes and their typical applications, and will also discuss the theory and design of a few particular types of friction clutches/brakes. Table 17-0 lists the variables used in this chapter and indicates the equation or section in which they can be found.

Chapter 17

17.1

CLUTCHES AND BRAKES

1009

TYPES OF BRAKES AND CLUTCHES

Brakes and clutches can be classified in a number of ways, by the means of actuation, the means of energy transfer between the elements, and the character of the engagement. Figure 17-1 shows a flow chart depicting these characteristics. The means of actuation may be mechanical, as in the pushing of an automobile’s clutch pedal, pneumatic or hydraulic, in which fluid pressure drives a piston to mechanically engage or disengage as with vehicle brakes, electrical, which is typically used to excite a magnetic coil, or automatic, as in an antirunaway brake that engages by relative motion between the elements. Positive ContaCt ClutChes The means of energy transfer may be a positive mechanical contact, as in a toothed or serrated clutch, which engages by mechanical interference as shown in Figure 17-2. The character of engagement is mechanical interference obtained with jaws of square or saw-toothed shape, or with teeth of various shapes. These devices are not as useful for brakes (except as holding devices) because they cannot dissipate large amounts of energy as can a friction brake, and as clutches they can be engaged only at low relative velocities (about 60 rpm max for jaw clutches and 300 rpm max for toothed clutches). Their advantage is positive engagement, and, once

17 FIGURE 17-1 Classification of Clutches and Brakes Source: Z. Hinhede, Machine Design Fundamentals: A Practical Approach, Prentice-Hall, 1983.

1010

MACHINE DESIGN

-

An Integrated Approach

FIGURE 17-2

-

Sixth Edition

Copyright © 2018 Robert L. Norton: All Rights Reserved

A Positive Contact Clutch Photo by author

* Automotive transmissions typically use helical gears, for quiet operation, as was noted in Chapter 13. The helical gears cannot be easily shifted in and out of engagement in manual transmissions because of their helix angle. So, they are all kept in constant mesh and clutched/ declutched from the transmission shaft to engage a particular ratio. Each gear has a synchromesh clutch connecting it to its shaft. This clutch actually consists of conical friction surfaces that drag the two elements (shaft and gear) into near zero relative velocity before the teeth of its companion positive-contact clutch engage. The shift lever moved by the driver is shifting these synchromesh clutches into and out of engagement, rather than moving gears around in the transmission.

coupled, they can transmit large torque with no slip. They are sometimes combined with a friction-type clutch, which drags the two elements to nearly the same velocity before the jaws or teeth engage. This is the principle of a synchromesh clutch in a manual automotive transmission.* FriCtion ClutChes and Brakes are the most common types used. Two or more surfaces are pressed together with a normal force to create a friction torque. The friction surfaces may be flat and perpendicular to the axis of rotation, in which case the normal force is axial (disk brake or clutch), as shown in Figure 17-3, or they may be cylindrical with the normal force in a radial direction (drum brake or clutch), as shown in Figures 17-9 and 17-10, or conical (cone brake or clutch). Cone clutches can tend to grab or refuse to release and are not now used very much in the United States, but are popular in Europe.[1]

inlet port

piston

needle thrust bearing

cylinder cylinder bearing

17

pressure plate FIGURE 17-3 A Multiplate Disk Clutch Actuated by Fluid Pressure Courtesy of Logan Clutch Corporation, Cleveland, Ohio

Chapter 17

CLUTCHES AND BRAKES

1011

At least one of the friction surfaces is typically metal (cast iron or steel) and the other is usually a high-friction material, referred to as the lining. If there are only two elements, there will be either one or two friction surfaces to transmit the torque. A cylindrical arrangement (drum brake or clutch) has one friction surface, and an axial arrangement (disk brake or clutch) has one or two friction surfaces, depending on whether the disk is sandwiched between two surfaces of the other element or not. For higher torque capacity, disk clutches and brakes are often made with multiple disks to increase the number of friction surfaces (see Figure 17-3). A clutch or brake’s ability to transfer the heat of friction generated can become the limiting factor on its capacity. Multidisk clutches are more difficult to cool so are appropriate for high-load, low-speed applications. For highspeed dynamic loads, fewer friction surfaces are better.[1] Friction clutches may be operated either dry or wet, the latter running in an oil bath. While the oil severely reduces the coefficient of friction, it greatly increases the heat transfer. Friction coefficients of clutch/brake material combinations typically range from 0.05 in oil to 0.60 dry. Wet clutches often use multiple disks to make up for the lower friction coefficient. Automatic transmissions for automobiles and trucks contain many wet clutches and brakes operating in oil that is circulated out of the transmission for cooling. Manual transmissions for off-road vehicles, such as motorcycles, use sealed, oil-filled, multidisk wet clutches to protect the friction surfaces from dust, water, and dirt. Manual-transmission automobiles and trucks typically use single-disk, dry clutches. overrunning ClutChes (also called one-way clutches) operate automatically based on the relative velocity of the two elements. They act on the circumference and allow relative rotation in only one direction. If the rotation attempts to reverse, the internal geometry of the clutch mechanism grabs the shaft and locks up. These backstop clutches can be used on a hoist to prevent the load from falling back if power to the shaft is interrupted, for example. These clutches are also used as indexing mechanisms. The input shaft can oscillate back and forth but the output turns intermittently in only one direction. Another common application of an overrunning clutch is in the rear hub of a bicycle to allow freewheeling when the wheel speed exceeds that of the drive sprocket. Several different mechanisms are used in one-way clutches. Figure 17-4a shows a sprag clutch, which has an inner and outer race like a ball bearing. But, instead of balls, the space between the races is filled with odd-shaped sprags, which allow motion in one direction but jam and lock the races together in the other, transmitting a one-way torque. A similar result is obtained with balls or rollers captured in wedge-shaped chambers between the races, then called a roller clutch. Figure 17-4b shows another type of one-way or overrunning clutch called a spring clutch, which uses a spring wrapped tightly around the shaft. Rotation in one direction wraps the spring tighter on the shaft and transmits torque. Counterrotation unwinds the spring slightly, allowing it to slip. CentriFugal ClutChes engage automatically when the shaft speed exceeds a certain magnitude. Friction elements are thrown radially outward against the inside of a cylindrical drum to engage the clutch. Centrifugal clutches are sometimes used to couple an internal-combustion engine to the drive train. The engine can idle decoupled from the wheels and when the throttle is opened, its increased speed automatically engages the clutch. These are common on go-karts. Used on chain saws for the same purpose, they also serve as an overload release that slips to allow the engine to continue running when the chain jams in the wood.

17

1012

MACHINE DESIGN

outer race

sprag

-

An Integrated Approach

-

Sixth Edition

energizing spring

output hub wrap spring input hub inner race

spring retainer (a)

( b)

FIGURE 17-4 Overrunning Clutches (a) Sprag Clutch (b) Wrap-Spring Clutch Courtesy of Warner Electric, South Beloit, Ill. 61080

MagnetiC ClutChes and Brakes are made in several types. Friction clutches are commonly electromagnetically operated, as shown in Figure 17-5. These have many advantages, such as rapid response times, ease of control, and smooth starts and stops, and are available powered on or powered off (fail-safe). Both clutch and brake versions are supplied as well as combined clutch-brake modules. Magnetic-particle clutches and brakes (not shown) have no direct frictional contact between the clutch disk and the housing and no friction material to wear. The gap between the surfaces is filled with a fine ferrous powder. When the coil is energized, the powder particles form chains along the magnetic field’s flux lines and couple the disk to the housing with no slip. The torque can be controlled by varying the current to the coil and the device will slip when the applied torque exceeds the value set by coil current, providing a constant tension. Magnetic hysteresis clutches and brakes have no mechanical contact between the rotating elements and so have zero friction when disengaged. The rotor, also called the drag cup, is dragged along (or braked) by the magnetic field set up by the field coil (or permanent magnet). These devices are used to control torque on shafts in applications such as winding machines, where a constant force must be applied to a web or filament of material as it is wound up. The torque on a hysteresis clutch is controllable independent of speed. These devices are extremely smooth, quiet, and long-lived, as there is no mechanical contact within the clutch except in its bearings.

17

Eddy-current clutches (not shown) are similar in construction to hysteresis devices in that they have no mechanical contact between rotor and pole. The coil sets up eddy currents, which magnetically couple the clutch together. There will always be some slip in this type of clutch, as there has to be relative motion between rotor and pole to generate

Chapter 17

armature

CLUTCHES AND BRAKES

1013

coil

magnet or field bushings

armature spline

friction material

FIGURE 17-5 Magnetically Operated Friction Clutch Courtesy of Warner Electric, South Beloit, Ill. 61080,

the eddy currents that supply the coupling force; so an eddy-current brake cannot hold a load stationary, only slow it from one speed to another. These have similar advantages to the hysteresis devices and are used for similar applications, such as coil or filament winders, etc. Fluid CouPlings transmit torque through a fluid, typically an oil. An impeller having a set of blades is turned by the input shaft and imparts angular momentum to the oil that surrounds it. A turbine (or runner) with similar blades is attached to the output shaft and is turned by the moving oil impinging on it. The principle of operation is similar to placing two electric fans face to face and turning only one on. The airflow from the powered fan blades will cause the facing, unpowered blades to windmill, passing power without any mechanical contact. Using incompressible oil in a confined volume is much more efficient than two open-air fans, especially when the impeller and turbine blades are optimally shaped to pump the oil. A fluid coupling provides extremely smooth starts and absorbs shocks, as the fluid simply shears when there is a speed differential, then gradually accelerates (or decelerates) the output turbine to nearly match the speed of the impeller. There will always be some slip,* meaning the turbine can never reach 100% of the impeller’s speed (0% slip), but it can operate at 100% slip when the turbine is stalled. All the input energy will then be converted to heat in shearing the oil. Heat transfer is an important consideration when designing a fluid coupling. The outside case is often finned to improve heat transfer. A fluid coupling transmits the input torque to the output shaft at any speed including stall, so it can never be totally decoupled like a friction clutch. The output must be braked to hold it stationary when the input shaft is turning.† The horsepower rating of a fluid coupling varies as the fifth power of its diameter. A 15% increase in diameter doubles its power capacity. If used as a brake, a fluid coupling can only provide a drag to slow a device, as in a dynamometer, but cannot hold a load stationary.

*

This slip is the reason that autombiles with fluid-coupled, automatic transmissions get worse mileage per gallon than a comparable standard tranmission auto. To counter this, many automatic transmissions are now fitted with a mechanical lock-up clutch that engages above some road speed (like 30 mph) to lock the impeller and output turbine together. This clutch is released automatically when the car slows below the set speed. †

This is why you must keep your foot on the brake when stopped at a traffic light in an automobile that has an automatic transmission if the engine is running and the transmission is in “drive.” The fluid coupling between the engine and transmission is always transmitting torque and the car will creep forward at idle unless the brakes are applied.

17

1014

MACHINE DESIGN

-

An Integrated Approach

flywheel

oil

turbine attached to transmission input shaft

-

Sixth Edition

impeller or pump attached to flywheel oil

oil oil

stator

direction of oil flow one-way clutch between stator and transmission case

engine crankshaft

transmission input shaft pilot bearing

oil oil oil

FIGURE 17-6

transmission case oil stator is stationary at stall and during acceleration but rotates in same direction as pump and turbine at steady state Copyright © 2018 Robert L. Norton: All Rights Reserved

Schematic of a Torque Converter

If a third, stationary, element with a set of curved blades, called a reactor or stator, is placed between impeller and turbine, additional angular momentum is imparted to the fluid and the device is then called a torque converter as shown in Figure 17-6. Torque converters are used in vehicles to couple the engine to an automatic transmission. The engine can idle with the vehicle stopped (turbine stalled—100% slip). At stall, the impeller and reactor blades create about a 2:1 torque multiplication, which is available to help accelerate the vehicle when the brakes are released and the engine speed increased. As the vehicle (and thus turbine) speed increases, the turbine will approach the impeller speed and the torque multiplication will decrease to essentially zero at a slip of a few percent. If more torque is needed momentarily (as in passing), the slip between impeller and turbine will automatically increase when the engine is sped up to provide more torque and power to accelerate the vehicle.

17.2

17

CLUTCH/BRAKE SELECTION AND SPECIFICATION

Manufacturers of specialty clutches and brakes such as those described above provide extensive information on the torque and power capacities of their various models in catalogs, many of which are as informative as a textbook on the particular subject. They also define procedures for selection and specification, usually based on the anticipated torque and power for the application plus suggested service factors that attempt to accommodate different loading, installation, or environmental factors than those under which the products were tested. For example, the manufacturer’s standard rating for a clutch model may be based on a smooth driver, such as an electric motor. If the particular application uses an internal-combustion engine of the same power, there will be impulse loads, and

Chapter 17

CLUTCHES AND BRAKES

1015

a clutch or brake of larger capacity than dictated by the average power will need to be selected. This is sometimes referred to as derating the clutch (or brake), meaning that its actual capacity under the anticipated conditions is considered to be less than the rated capacity of the chosen device. serviCe FaCtors According to many clutch manufacturers, a common cause of clutch trouble is the designer’s failure to properly apply adequate service factors to account for the particular conditions of the application.[1] This may, in part, be due to confusion engendered by a lack of standardization for definitions of service factors. One manufacturer may recommend a service factor of 1.5 for a particular condition while another manufacturer recommends 3.0 for the same condition. Both will be correct for their particular clutch designs because, in one case, the manufacturer may have already built-in a safety factor while the other applies it in the service factor. The wise designer will carefully follow each manufacturer’s recommended selection procedures for their products, realizing that they are typically based on extensive (and expensive) testing programs as well as on field-service experience with that particular product. A clutch even slightly too small for the applied loading will slip and overheat. A clutch too large for the load is also bad, as it adds unnecessary inertia and may overload the motor that must accelerate it. Most manufacturers of machine parts are generous in providing engineering help to properly size and specify their products for any application. The machine designer’s principal concern should be to accurately define the loading and environmental conditions that the device must accommodate. This may require extensive and tedious calculations of such things as the moments of inertia for all the elements in the drive train actuated by the clutch or brake. The load-analysis methods of Chapter 3 are applicable to such a task. ClutCh loCation When a machine has both high- and low-speed shafts (as it will whenever a speed reducer is used, such as in Case Studies 7 and 8), and a clutch is needed in the system, the question immediately arises, should the clutch be placed on the low- or high-speed side of the gear reducer? Sometimes the answer is dictated by function. For example, it would make little sense to put an automotive clutch on the output shaft of the transmission instead of the input side, since in this instance the principal purpose of the clutch is to interrupt the connection between engine and transmission, ergo, it must go on the high-speed side. In other cases, function does not dictate the clutch location, as, for example, in Case Study 8, where the coupling on either shaft could be replaced by a clutch if it were desired to decouple the compressor from the engine. (See Figure 9-1.) The choice is less clear in these situations, and there are two competing schools of thought. The torque (and any shock load) is larger on the low-speed shaft than on the highspeed shaft by the gear ratio. The power is essentially the same at both locations (neglecting losses in the gear train), but the kinetic energy at the high-speed shaft is larger by the square of the gear ratio. A clutch on the low-speed side must be larger (and thus more expensive) in order to handle the larger torque. However, a smaller and cheaper clutch on the high-speed side must dissipate the greater kinetic energy at that location and thus may more readily overheat. Some manufacturers recommend always using the highspeed side for the clutch location if function allows, opting for its better initial economy. Other clutch manufacturers suggest that the higher initial cost of the larger, low-speed clutch will be counterbalanced by lower maintenance cost in the long run. The balance of expert opinion seems to tilt toward the high-speed location with the caveat that each situation should be individually evaluated on its own merits.[1]

17

1016

MACHINE DESIGN

17.3

-

An Integrated Approach

-

Sixth Edition

CLUTCH AND BRAKE MATERIALS

Materials for the structural parts of brakes and clutches, such as the disks or drums, are typically made of gray cast iron or steel. The friction surfaces are usually lined with a material having a good coefficient of friction and sufficient compressive strength and temperature resistance for the application. Asbestos fiber was once the most common ingredient in brake or clutch linings, but is no longer used in many applications because of its danger as a carcinogen. Linings may be molded, woven, sintered, or of solid material. Molded linings typically use polymeric resins to bind a variety of powdered fillers or fibrous materials. Brass or zinc chips are sometimes added to improve heat conduction and wear resistance and reduce scoring of drums and disks. Woven materials typically used long asbestos fibers. Sintered metals provide higher-temperature resistance and compressive strength than molded or woven materials. Materials such as cork, wood, and cast iron are sometimes used as linings as well. Table 17-1 shows some frictional, thermal, and mechanical properties of a few friction-lining materials.

17.4

DISK CLUTCHES

The simplest disk clutch consists of two disks, one lined with a high-friction material, pressed together axially with a normal force to generate the friction force needed to transmit torque, as shown in Figure 17-7. The normal force can be supplied mechanically, pneumatically, hydraulically, or electromagnetically and is typically quite large. The pressure between the clutch surfaces can approach a uniform distribution over the surface if the disks are flexible enough. In such cases, the wear will be greater at larger diameters because wear is proportional to pressure times velocity (pV) and the velocity increases linearly with radius. However, as the disks wear preferentially toward the outside, the loss of material will change the pressure distribution to a nonuniform one and the clutch will approach a uniform wear condition of pV = constant. Thus, the two extremes are a uniform-pressure and a uniform-wear condition. A flexible clutch may be close to a uniform-pressure condition when new, but will tend toward a uniform-wear condition with use. A rigid clutch will more rapidly approach the uniform-wear condition with use. The calculations for each condition are different, and the uniform-wear assumption gives a more conservative clutch rating, so is favored by some designers. A

lining

dr

ro F

r

F ri

17

A FIGURE 17-7 A Single-Surface Axial Disk Clutch

Section A-A

Chapter 17

Table 17-1

CLUTCHES AND BRAKES

1017

Properties of Common Clutch/Brake Lining Materials

Friction Material Against Steel or CI

Dynamic Coefficient of Friction

Maximum Pressure psi

Maximum Temperature °F

°C

400–500

204–260

400–500

204–260

dry

in oil

kPa

Molded

0.25–0.45

0.06–0.09

Woven

0.25–0.45

0.08–0.10

Sintered metal

0.15–0.45

0.05–0.08

150–300 1 030–2 070

450–1 250 232–677

Cast iron or hard steel

0.15–0.25

0.03–0.06

100–250

500

150–300 1 030–2 070 50–100

345–690

690–720

260

Uniform Pressure Consider an elemental ring of area on the clutch face of width dr as shown in Figure 17-7. The differential force acting on this ring is dF = pθr dr

(17.1a)

where r is the radius, θ is the angle of the ring in radians and p is the uniform pressure on the clutch face. For a full circumference clutch as shown in Figure 17-7, θ will be 2π. The total axial force F on the clutch is found by integrating this expression between the limits ri and ro: F=

ro

∫r

pθr dr =

i

(

pθ 2 ro − ri2 2

)

(17.1b)

The friction torque on the differential ring element is dT = pθµr 2 dr

(17.2a)

where µ is the coefficient of friction. The total torque for one clutch disk is T=

ro

∫r

pθµr 2 dr =

i

(

pθµ 3 3 ro − ri 3

)

(17.2b)

For a multiple-disk clutch with N friction faces: T=

(

)

pθµ 3 3 ro − ri N 3

(17.2c)

Equations 17.1b and 17.2c can be combined to give an expression for torque as a function of axial force: T = NµF

( (

3 3 2 ro − ri 3 ro2 − ri2

) )

(17.3)

Uniform Wear The constant-wear rate W is assumed to be proportional to the product of pressure p and velocity V:

17

1018

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

W = pV = constant

(17.4 a)

And the velocity at any point on the face of the clutch is V = rω

(17.4 b)

Combine these equations, assuming a constant angular velocity ω: pr = cons tan t = K

(17.4c)

The largest pressure pmax must then occur at the smallest radius ri : K = pmax ri

(17.4 d )

Combining equations 17.4c and 17.4d gives an expression for pressure as a function of the radius r: p = pmax

ri r

(17.4e)

where the maximum allowable pressure pmax will vary with the lining material used. Table 17-1 shows recommended values of pmax and coefficients of friction for various clutch/brake lining materials. The axial force F is found by integrating equation 17.1a for the differential force on the ring element of Figure 17-7 with equation 17.4e substituted for p: F=

ro

∫r

pθr dr =

i

ro

∫r

i

r  θ  pmax i  r dr = ri θpmax ( ro − ri )  r

(17.5a)

The torque is found by integrating equation 17.2a with the same substitution: T=

ro

∫r

pθµr 2 dr =

i

(

θ µri pmax ro2 − ri2 2

)

(17.5b)

Combine equations 17.2a and b for an expression relating torque to axial force for the uniform-wear case: T = NµF

(ro + ri ) 2

(17.6)

where N is the number of friction surfaces in the clutch. From equations 17.5a and 17.5b, it can be shown that the maximum torque for any outside radius ro will be obtained when the inside radius is: ri = 1 3ro = 0.577ro

17

(17.7)

Note that the uniform-wear assumption gives a lower torque capacity for the clutch than does the uniform-pressure assumption. The higher initial wear at the larger radii shifts the center of pressure radially inward, giving a smaller moment arm for the resultant friction force. Clutches are usually designed based on uniform wear. They will have a greater capacity when new, but will end up close to the predicted design capacity after they are worn in.

Chapter 17

CLUTCHES AND BRAKES

1019

EXAMPLE 17-1 Design of a Disk Clutch Problem

Determine a suitable size and required force for an axial disk clutch.

Given

The clutch must pass 7.5 hp at 1725 rpm with a service factor of 2.

Assumptions

Use a uniform-wear model. Assume a single dry disk with a molded lining.

Solution

1 The service factor of 2 requires derating the clutch by that factor, so we will design for 15 hp instead of 7.5. Find the torque required for that power at the design rpm: in-lb sec   15 hp  6600 hp   P = 548.05 lb-in T= = ω  2π rad sec  1725 rpm   60 rpm 

(a)

2 Find the coefficient of friction and maximum recommended pressure for a dry, molded material from Table 17-1. Use the average of the ranges of values shown: pmax = 225 psi and µ = 0.35. 3 Substitute equation 17.7 relating ri to ro for maximum torque into equation 17.5b to get

(

)

1   T = πµri pmax ro2 − ri2 = πµ ( 0.577ro ) pmax  ro2 − ro2  = 0.3849ro3 πµpmax  3  1

 T ro =   0.3849 πµp

1

3  3 548.05  =  0.3849 π ( 0.35)( 225)  = 1.79 in max

( b)

4 From equation 17.7: ri = 0.577ro = 0.577 (1.79 ) = 1.03 in

(c)

5 The axial force needed (from equation 17.5a) is: F = 2 πri pmax ( ro − ri ) = 2 π (1.034 )( 225)(1.792 − 1.034 ) = 1108 lb

(d )

6 The clutch specification is then a single disk with 3.6-in outside diameter and 2-in inside diameter, a molded lining with µdry ≥ 0.35, and an actuating force ≥ 1 108 lb. 7 The files EX17-01 can be found on the book’s website.

17.5

DISK BRAKES

The disk clutch equations also apply to disk brakes. However, disk brakes are seldom made with linings covering the entire circumference of the face, because they would then overheat. Use the appropriate value of θ for the included angle of the brake pads in equations 17-1 through 17-5 to calculate the force and torque in a disk brake. Note that

17

1020

MACHINE DESIGN

-

An Integrated Approach

cable

-

Sixth Edition

cable

brake arm

brake arm frame

pad 1

pad 2 wheel rim

FIGURE 17-8

wheel rim θ

Bicycle Disk-Type Brake

N will be at least 2 for a disk brake as it has opposing pads in pairs as shown in Figure 17-8. While clutches are often used with light duty cycles (engagement time a small fraction of total time), brakes often must absorb large amounts of energy in repeated applications. Caliper disk brakes, such as those used on automobiles, use friction pads applied against only a small fraction of the disk circumference, leaving the rest exposed for cooling. The disk is sometimes ventilated with internal air passages to promote cooling. The caliper typically straddles the disk and contains one or more pairs of pads, each pair rubbing both sides of the disk. This cancels the axial force and reduces axial loads on the bearings. The common bicycle caliper brake as shown in Figure 17-8 is another example in which the wheel rim is the disk and the calipers pinch against only a small fraction of the circumference. Disk brakes are now commonly used on automobiles, particularly on the front wheels, which provide more than half the stopping force. Some advantages of disk over drum brakes are their good controllability and linearity (braking torque directly proportional to applied axial force).

17.6

17

DRUM BRAKES

Drum brakes (or clutches) apply the friction material to the circumference of a cylinder, either externally, internally, or both. These devices are more often used as brakes than as clutches. The part to which the friction material is riveted or bonded with adhesive is called the brake shoe, and the part against which it rubs, the brake drum. The shoe is forced against the drum to create the friction torque. The simplest configuration of a drum brake is the band brake, in which a flexible shoe is wrapped around a majority of the outer circumference of the drum and squeezed against it. Alternatively a relatively rigid, lined shoe (or shoes) can be pivoted against the outer or inner circumference (or both) of the drum. If the shoe contacts only a small angular portion of the drum, it is a

Chapter 17

CLUTCHES AND BRAKES

1021

short-shoe brake, otherwise a long-shoe brake. The geometry of the short versus long shoe contact requires a different analytical treatment in each case. We will examine the cases of external short-shoe and external long-shoe drum brakes to illustrate their differences and features, particularly in contrast to disk brakes. The principles are the same for internal-shoe brakes as well. Short-Shoe External Drum Brakes Figure 17-9a shows a schematic of a short-shoe external drum brake. If the angle θ subtended by the arc of contact between shoe and drum is small (< about 45°), then we can consider the distributed force between shoe and drum to be uniform, and it can be replaced by the concentrated force Fn in the center of the contact area, as shown in Figure 17-9b. For any maximum allowable lining pressure pmax (Table 17-1) the force Fn can be estimated as Fn = r θ w pmax

(17.8)

where w is the width of the brake shoe in the z direction and θ is the subtended angle in radians. The frictional force Ff is Ff = µFn

(17.9)

where µ is the coefficient of friction for the brake lining material (Table 17-1). The torque on the brake drum is then T = Ff r = µFn r

(17.10)

Summing moments about point O on the free-body diagram of Figure 17-9b and substituting equation 17.9 gives

∑ M = 0 = aFa − bFn + cFf

Fa =

bFn − cFf a

=

(17.11a)

bFn − µcFn b − µc = Fn a a

Fa

Fa

shoe θ

(17.11b)

a

r

Ff

Ry

c

O

Rx

O

Fn drum

ω b (a) Brake assembly

FIGURE 17-9 Geometry and Forces for a Short-Shoe External Drum Brake

(b) Free-body diagram

17

1022

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

The reaction forces at the pivot are found from a summation of forces: Rx = − Ff Ry = Fa − Fn

(17.12)

selF-energizing Note in Figure 17-9b that with the direction of drum rotation shown, the friction moment cFf adds to the actuating moment aFa. This is referred to as self-energizing. Once any application force Fa is applied, the friction generated at the shoe acts to increase the braking torque. However, if the brake-drum rotation is reversed from that shown in Figure 17-9a, the sign of the friction-moment term cFf of equation 17.11a becomes negative, and the brake is then self-deenergizing. This self-energizing feature of drum brakes is a potential advantage, as it reduces the required application force compared to a disk brake of the same capacity. Drum brakes typically have two shoes, one of which can be made self-energizing in each direction, or both in one direction. The latter arrangement is commonly used in automobile brakes to aid in stopping forward motion at the expense of stopping backward motion, which is normally at lower speeds. selF-loCking Note in equation 17.11 that if the brake is self-energizing and the product µc ≥ b, the force Fa needed to actuate the brake becomes zero or negative. The brake is then said to be self-locking. If the shoe touches the drum, it will grab and lock. This is usually not a desired condition except in so-called backstopping applications, as were described under overrunning clutches above. In effect, a self-locking brake can function as an overrunning clutch to backstop a load and prevent it from running away if power is lost. Such brakes are sometimes used in hoists for that purpose.

EXAMPLE 17-2 Design of a Short-Shoe Drum Brake

* Files EX17-02 are on the book’s website.

Problem

For the drum-brake arrangement shown in Figure 17-9, determine the ratio c / r that will give a self-energizing ratio Fn / Fa of 2. Also find the c / r ratio that will cause self-locking.

Given

The dimensions are a = b = 6, r = 5.

Assumptions

Coefficient of friction µ = 0.35.

Solution*

See Figure 17-9.

1 Rearrange equation 17.11 to form the desired ratio: Fn a = Fa b − µc

(a)

2 Substitute the desired self-energizing ratio, the given dimensions, and solve for c:

17

Fn 6 =2= Fa 6 − 0.35c −3 = 8.571 c= −0.35

(b)

Chapter 17

CLUTCHES AND BRAKES

1023

3 Form the c / r ratio for a self-energizing ratio of 2 with the given brake geometry: c 8.571 = = 1.71 5 r

(c)

4 For self-locking to begin, Fa becomes zero, making Fn / Fa = ∞ and Fa / Fn = 0. The second of these ratios will need to be used to avoid division by zero. Rearrange equation 17.11 to form the desired ratio and solve for c: Fa b − µc 6 − 0.35c = = =0 Fn a 6 c = 17.143

(d )

5 Form the c / r ratio for self-locking with the given brake geometry: c 17.143 = = 3.43 5 r

(e)

6 Note that these ratios are specific to the dimensions of the brake. The length a was set equal to b in this example in order to eliminate the effect of the lever arm ratio a / b, which further reduces the application force Fa required for any normal force Fn.

Long-Shoe External Drum Brakes If the angle of contact θ between shoe and drum in Figure 17-9 exceeds about 45°, then the assumption of uniform pressure distribution over the shoe surface will be inaccurate. Most drum brakes have contact angles of 90° or more, so a more accurate analysis than the short-shoe assumption is needed. Since any real brake shoe will not be infinitely rigid, its deflections will affect the pressure distribution. An analysis that accounts for deflection effects is very complicated and is not really warranted here. As the shoe wears, it will pivot about point O in Figure 17-10, and point B will travel farther than point A because of its greater distance from O. The pressure at any point on the shoe will also

a Fa

Fa shoe B

dFf



drum

dFn C A

y

θ

ω

r – b cos θ

b cos θ

θ2 θ1 x

B

dFn

C A

y θ

dFf

(a) Brake assembly

FIGURE 17-10 Geometry and Forces for a Long-Shoe External Drum Brake

x Rx

O r

Ry

b sin θ

O

b ( b) Free-body diagram and dimensions

17

1024

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

vary in proportion to its distance from O. Assume that the drum turns at constant velocity and that wear is proportional to the friction work done, i.e., the product pV. Then, at any arbitrary point on the shoe, such as C in Figure 17-10, the normal pressure p will be proportional to its distance from point O: p ∝ b sin θ ∝ sin θ

(17.13a)

Since the distance b is constant, the normal pressure at any point is just proportional to sin θ. Call its constant of proportionality K: p = K sin θ

(17.13b)

If the maximum allowable pressure for the lining material is pmax (Table 17-1) then the constant K can be defined as K=

pmax p = sin θ sin θmax

(17.13c)

where θmax is the smaller of θ2 and 90°. Then p=

pmax sin θ sin θmax

(17.13d )

Equation 17.13d defines the normal pressure at any point on the shoe and it varies as sin θ, since pmax and θ2 are constant for any particular brake. Thus, the friction force is small at small θ, is optimum at θ = 90°, and diminishes at angles larger than 90°. Little is gained by using θ1  120°. To obtain the total force on the shoe, the pressure function must be integrated over the angular range of the shoe. Consider the differential element dθ shown in Figure 17-9. Two differential forces act on it, dFn and dFf. They have respective moment arms about point O of b sin θ and r – b cos θ, as shown in Figure 17-10b. Integrating to create their moments about O for the entire surface gives, for the moment due to the normal force: θ2

M Fn =

∫θ

=

∫θ

pwr d θ b sin θ =

1

θ2

∫θ

wrb p sin θ d θ

1

θ2

wrb

1

M Fn = wrb

pmax sin 2 θ d θ sin θmax

pmax sin θmax

1 1   2 ( θ2 − θ1 ) − 4 ( sin 2θ2 − sin 2θ1 ) 

(17.14 a)

where w is the drum width in the z direction and the other variables are as defined in Figure 17-10. For the moment due to the frictional force: θ2

M Ff =

∫θ

=

∫θ

1

θ2 1

17

M Ff = µwr

µpwr d θ ( r − b cos θ ) µwr

pmax sin θ ( r − b cos θ ) d θ sin θmax

pmax sin θmax

(

)

b   2 2  − r ( cos θ2 − cos θ1 ) − 2 sin θ2 − sin θ1 

(17.14 b)

Chapter 17

CLUTCHES AND BRAKES

1025

Summing moments about point O gives Fa =

M Fn  M Ff

(17.14c)

a

where the upper sign is for a self-energizing brake and the lower one for a self-deenergizing brake. Self-locking can occur only if the brake is self-energizing and MFf  > MFn. The torque for the brake is found by integrating the expression for the product of the friction force Ff and drum radius r : T=

θ2

∫θ

µpwr d θ r

1

=

θ2

∫θ

µwr 2

1

T = µwr 2

pmax sin θ d θ sin θmax

pmax ( cos θ1 − cos θ2 ) sin θmax

(17.15)

The reaction forces Rx and Ry are found by summing forces in the x and y directions (see Figure 17-10b):

∫ θ =∫ θ



Rx = cos θ dFn + sin θ dFf 2

wrp cos θ dθ + µ

1

=

θ2 1

∫ θ =∫ θ

wrp sin θ dθ

1

∫θ

wr

pmax sin θ cos θ dθ + µ sin θmax

pmax sin θmax

Rx = wr

θ2

∫θ

θ2

∫θ

1

wr

pmax sin 2 θ dθ sin θmax

  sin 2 θ  sin 2 θ1  2 − −    2    2    1  1   +µ  2 ( θ2 − θ1 ) − 4 ( sin 2θ2 − sin 2θ1 )    



(17.16a )

Ry = cos θ dFf + sin θ dFn − Fa 2

1

Ry = wr

µwr

pmax sin θ cos θ dθ + sin θmax

pmax sin θmax

θ2

∫θ

1

µwr

pmax sin 2 θ dθ − Fa sin θmax

  sin 2 θ  sin 2 θ1  2 − −µ    2    2    − Fa 1  1  +  ( θ2 − θ1 ) − ( sin 2θ2 − sin 2θ1 )    4 2  

(17.16b)

17

1026

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

EXAMPLE 17-3 Design of a Long-Shoe Drum Brake Problem

For the drum-brake arrangement shown in Figure 17-10, determine the friction torque T, application force Fa, and reaction forces Rx, Ry.

Given

The dimensions are a = 180 mm, b = 90 mm, r = 100 mm, w = 30 mm, θ1 = 30°, θ2 = 120°, θmax = 90°.

Assumptions

Coefficient of friction µ = 0.35, maximum lining pressure pmax = 1.5 MPa, and the brake is self-energizing.

Solution

See Figure 17-10.

1 Convert the given angles θ1 and θ2 to radians: θ1 = 0.524 rad, θ2 = 2.094 rad. 2 Calculate the moment MFn about O due to the normal force using equation 17.14a: M Fn = wrb

pmax sin θmax

1 1   2 ( θ2 − θ1 ) − 4 ( sin 2θ2 − sin 2θ1 ) 

1 ( 2.094 − 0.524 ) 1.5  2 = 30 (100 )( 90 )  1 sin (1.57 )  − sin {2 ( 2.094 )} − sin {2 ( 0.524 )}  4 = 493 N-m

(

)

     (a)

3 Calculate the moment MFf about O due to the friction force using equation 17.14b: M Ff = µwr

pmax sin θmax

(

)

b   2 2  − r ( cos θ2 − cos θ1 ) − 2 sin θ2 − sin θ1 

 −100 ( cos {2.094} − cos {0.524}) 1.5  = 0.35 ( 30 )(100 ) 90 sin (1.57 )  − sin 2 {2.094} − sin 2 {0.524}  2 = 180 N-m

(

)

    (b)

4 Find the application force from equation 17.14c: Fa =

M Fn  M Ff a

=

497 − 181 = 1743 N 0.180

(c)

5 Find the friction torque from equation 17.15: pmax ( cos θ1 − cos θ2 ) sin θmax 1.5 = 0.35 ( 30 )(100 )2 ( cos {2.094} − cos {0.524}) sin (1.57 ) = 215 N-m

T f = µwr 2

17

6 The reaction forces are found from equations 17.16:

(d )

Chapter 17

Rx = wr

pmax sin θmax

CLUTCHES AND BRAKES

  sin 2 θ  sin 2 θ1  2 − −    2    2    1  1   +µ  2 ( θ2 − θ1 ) − 4 ( sin 2θ2 − sin 2θ1 )    

 sin 2 {2.094} sin 2 {0.524}  −   2 2   1.5  1 = 30 (100 )  sin (1.57 )   2 ( 2.094 − 0.524 )  −0.35  1   − sin {2 ( 2.094 )} − sin {2 ( 0.524 )}   4 = 794 N

(

Ry = − Fa + wr

1027

pmax sin θmax

)

          (e)

  sin 2 θ  sin 2 θ1  2 − −µ    2 2       1 1    +  ( θ2 − θ1 ) − ( sin 2θ2 − sin 2θ1 )    4 2  

  sin 2 {2.094} sin 2 {0.524}  − −0.35   2 2    1.5  1  = −2013 + 30 (100 )  sin (1.57 )   2 ( 2.094 − 0.524 )  +  − 1 sin {2 ( 2.094 )} − sin {2 ( 0.524 )}    4  Ry = 4134 N

(

)

          (f)

7 The files EX17-03 can be found on the book’s website.

Long-Shoe Internal Drum Brakes Most drum brakes (and virtually all automotive ones) use internal shoes that expand against the inside of the drum. Typically two shoes are used, pivoted against the ends of an adjusting screw and forced against the drum by a double-ended hydraulic cylinder. Light springs hold the shoes against the pistons of the wheel cylinder and pull the shoes away from the drum when not activated. Typically, one shoe is self-energizing in the forward direction and the other is self-energizing in the reverse direction of drum rotation. The automobile wheel attaches directly to the brake drum. The analysis of an internal-shoe brake is the same as that of an external-shoe one.

17.7

SUMMARY

Clutches and brakes are used extensively in all kinds of machinery. Vehicles all need brakes to stop their motion, as do many stationary machines. Clutches are needed to interrupt the flow of power from a prime mover (motor, engine, etc.) to the load so the

17

1028

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

load can be stopped (by a brake) and the prime mover continue running. A clutch and a brake are essentially the same device, the principal difference being that both sides of a clutch (input and output) are capable of rotation, but the output side of a brake is fixed to some nonrotating “ground plane,” which itself may have some other motion, as in the case of an automobile chassis. Many different styles of clutches/brakes are made, but the most common style uses frictional contact between two or more surfaces to couple the input and output sides together. The friction surfaces can be moved into and out of engagement by any of several means including direct mechanical, electromagnetic, pneumatic, hydraulic, or combinations thereof. Other styles include direct magnetic (magnetic particle, hysteresis, and eddy current), some of which have no mechanical contact (and thus zero drag) when disengaged, and fluid couplings, which are commonly used to couple vehicle engines to automatic transmissions. Except for high-volume, specialized applications such as vehicle design, a machine designer seldom designs a clutch or brake from scratch. For the typical machine-design application, one usually selects a commercially available clutch or brake assembly from the many manufacturers’ offerings. The design problem then becomes one of properly defining the torque, speed, and power requirements and the character of the load, whether smooth or shocky, continuous or intermittent, etc. The inertia of the rotating elements to be accelerated by a clutch or decelerated by a brake can have a significant effect on the required size of that device and must be carefully calculated. Any gear ratios present in the system will cause the reflected or effective inertia to vary as the square of the gear ratio and this effect must be carefully included when calculating the inertia. (See Chapter 3 or any text on dynamics of machinery.) The clutch/brake manufacturers’ catalogs contain extensive engineering data that rate each device on its torque and power capacity and also suggest empirical derating factors for situations with shock loads, high duty cycles, etc. Once the loading is well defined, a suitable device can be specified using the manufacturers’ rating data modified by their suggested service factors. The designer’s (nontrivial) task then becomes that of proper load definition for the application, followed by proper use of the manufacturers’ rating data. Engineering assistance is usually available from the manufacturer for the latter task, but the result can be only as suitable to the design requirements as the accuracy of the load analysis allows. The mechanical configurations of several clutch designs are briefly described within this chapter. Manufacturers’ catalogs and applications engineers can provide more detailed information on the capabilities and limitations of the various clutch/brake styles.

17

Commercial friction clutches and brakes are most often made in a single- or multipledisk configuration. Vehicle brakes are typically made in either disk or drum configurations. Disk configurations provide a friction torque that is linearly proportional to the applied actuating force, and this can be an advantage from a control standpoint. Drum configurations can be designed to be self-energizing, meaning that once the brake or clutch is initially engaged, the friction force tends to increase the normal force, thus nonlinearly increasing the friction torque in a positive feedback fashion. This can be an advantage when braking large loads, as it decreases the required application force but it makes control of braking torque more difficult. The analysis of both friction disk devices and friction drum devices is developed in this chapter.

Chapter 17

CLUTCHES AND BRAKES

1029

Clutches and brakes are essentially energy-transfer or energy-dissipation devices and as such generate a great deal of heat in operation. They must be designed to absorb and transfer this heat without damage to themselves or their surroundings. Often, the heat-transfer ability of a device rather than its mechanical torque-transmission ability limits its capacity. The thermal design of clutches and brakes is a very important consideration, but it goes beyond the scope of this text and space does not permit its treatment here. Nevertheless, the designer needs to be aware of the heat transfer aspect of clutch/ brake design and take it into account. See any text on heat transfer for the theoretical background and see the references mentioned in the bibliography to this chapter as well as other manufacturers’ catalogs for more specific information. Friction clutches may be operated either dry or wet (typically in oil). Dry friction is obviously more effective, as the coefficient of friction is severely reduced with lubrication. However, running in oil can significantly improve the heat transfer situation especially when the oil is circulated and/or cooled. More friction surfaces (e.g., multiple disks) are needed to achieve the same torque capacity wet that can be obtained with a single dry disk, but the trade-off can be positive because of enhanced cooling. Modern vehicle automatic transmissions use many internal clutches and brakes to interconnect or stop various members of their epicyclic (planetary) gear trains in order to shift among gear ratios. These are either multidisk clutches or band brakes and are run immersed in the transmission oil that is continuously circulated through a heat exchanger in the vehicle’s radiator for cooling. Important Equations Used in This Chapter Torque in a Disk Clutch with Uniform Pressure (Section 17.4):

T = NµF

( (

3 3 2 ro − ri 3 ro2 − ri2

) )

(17.3)

Torque in a Disk Clutch with Uniform Wear (Section 17.4):

T = NµF

(ro + ri ) 2

(17.6)

Forces and Torque on a Short-Shoe Drum Brake (Section 17.6):

bFn + cFf ∑ M =Fn0==raFθ aw−pmax

bFn − cFf

bFn − µcFn b − µc = Fn a a a θ2 θ2 M Fn = pwr d θ b sin θ = wrb p sin θ d θ T =θ1Ff r = µFn r θ1 θ2 p = wrb max sin 2 θ d θ Forces and Torque Long-Shoe Drum Brake (Section 17.6): θ1 on asin θmax Fa =



=



(17.8) (17.11a) (17.11b) (17.10)



M Fn = wrb

pmax sin θmax

1 1   2 ( θ2 − θ1 ) − 4 ( sin 2θ2 − sin 2θ1 ) 

(17.14 a)

17

1030



θ2

M Ff = µpwr d θ ( r − b cos θ ) θ1 MACHINE DESIGN - An Integrated Approach θ2 p = µwr max sin θ ( r − b cos θ ) d θ θ1 sin θmax

-

Sixth Edition



pmax sin θmax

(

)

b   θ2 − cos θ1 ) − sin 2 θ2 − sin 2 θ1   − r ( cos 2  θ2 T= µpwr d θ r θ1 Rx = cos θ dFn + sin θ dFf M F  M F θ2 n p f = Faθ=µwr 2 a max sin θ d θ θ2 2 sin = wrp cos θ dθ + µ θ1 wrp sin θθ dθmax θ1 θ1 p 2 max T = µwr − cos θ2 ) θ2 θ(2cos θ1 p p = wr max sin θ cos θsin dθ θ+max wr max sin 2 θ dθ µ θ1 θ1 sin θmax sin θmax M Ff = µwr

∫ ∫







∫ ∫



  sin 2 θ  sin 2 θ1  2 − −    p 2    2  Rx = wr max   Ry = cos + sin θ dF − F sinθθdF f n a max  1 1  +µ θ − θ − θ − θ sin 2 sin 2 ( ) ( ) 2 1   2 2 1 θ42 θ2 p  p  = µwr max sin θ cos θ dθ + µwr max sin 2 θ dθ − Fa θ1 θ1 sin θmax sin θmax

∫ ∫

Ry = wr

17.8



(17.14 b)

(17.14c) (17.15)

(17.16a )



pmax sin θmax

  sin 2 θ  sin 2 θ1  2 − −µ    2    2    − Fa 1  1  +  ( θ2 − θ1 ) − ( sin 2θ2 − sin 2θ1 )    4 2  

(17.16b)

REFERENCES

1 J. Proctor, “Selecting Clutches for Mechanical Drives,” Product Engineering, pp. 43–58, June 19, 1961.

17.9

BIBLIOGRAPHY Mechanical Drives Reference Issue. Penton Publishing: Cleveland, Ohio, Deltran, Electromagnetic Clutches and Brakes. American Precision Industries: Buffalo, N.Y., 716-631-9800. Logan, Multiple Disk Clutches and Brakes. Logan Clutch Co.: Cleveland, Ohio, 216431-4040. Magtrol, Hysteresis Brakes and Clutches. Magtrol: Buffalo, N.Y., 800-828-7844. MCC, Catalog 18e. Machine Components Corporation: Plainview, NY, 516-694-7222. MTL, Eddy Current Clutch Design. Magnetic Technologies Ltd.: Oxford, Mass., 508987-3303. W. C. Orthwein, Clutches and Brakes: Design and Selection. Marcel Dekker: New York, 1986. Placid, Magnetic Particle Clutches and Brakes. Placid Industries Inc.: Lake Placid, N.Y., 518-523-2422.

17

Warner, Clutches, Brakes, and Controls: Master Catalog. Warner Electric: South Beloit, Ill., 815-389-2582.

Chapter 17

CLUTCHES AND BRAKES

17.10 PROBLEMS *17-1

Table P17-0†

Find the torque that a 2-surface, dry disk clutch can transmit if the outside and inside lining diameters are 120 mm and 70 mm, respectively, and the applied axial force is 10 kN. Assume uniform wear and µ = 0.4. Is the pressure on the lining acceptable? What lining material(s) would be suitable?

Topic/Problem Matrix 17.4 Disk Clutches 17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-27, 17-28, 17-35

17-2 Repeat Problem 17-1 assuming uniform pressure. *17-3

1031

Design a single-surface disk clutch to transmit 100 N-m of torque at 750 rpm using a molded lining with a maximum pressure of 1 MPa and µ = 0.25. Assume uniform wear. Find the outside and inside diameters required if ri = 0.577ro. What is the power transmitted?

17.5 Disk Brakes

†17-4

Repeat Problem 17-3 assuming uniform pressure.

17.6 Drum Brakes

*17-5

How many surfaces are needed in a wet disk clutch to transmit 120 N-m of torque at 1000 rpm using a sintered lining with a maximum pressure of 1.8 MPa and µ = 0.06? Assume uniform wear. Find the outside and inside diameters required if ri = 0.577ro. How many disks are needed? What is the power transmitted?

Short Shoe

17-6 Repeat Problem 17-5 assuming uniform pressure. *17-7

Figure P17-1 shows a single short-shoe drum brake. Find its torque capacity and required actuating force for a = 100, b = 70, e = 20, r = 30, w = 50 mm, and θ = 35°. What value of c will make it self-locking? Assume pmax = 1.3 MPa and µ = 0.3.

17-8 Repeat Problem 17-7 with the drum rotating clockwise.

17-29, 17-30, 17-31, 17-32, 17-33, 17-34, 17-36, 17-37, 17-38

17-7, 17-8, 17-9, 17-10, 17-11, 17-12, 17-21, 17-22 Long Shoe 17-13, 17-14, 17-15, 17-16, 17-17, 17-18, 17-19, 17-20, 17-23, 17-24, 17-25, 17-26 Problem numbers in italics are design problems. †

17-9 Figure P17-1 shows a single short-shoe drum brake. Find its torque capacity and required actuating force for a = 8, b = 6, e = 4, r = 5, w = 1.5 in, and θ = 30°. What value of c will make it self-locking? Assume pmax = 250 psi and µ = 0.35. 17-10 Repeat Problem 17-9 with the drum rotating clockwise. *17-11

Figure P17-2 shows a double short-shoe drum brake. Find its torque capacity and required actuating force for a = 90, b = 80, e = 30, r = 40, w = 60 mm, and θ = 25°. What value of c will make it self-locking? Assume pmax = 1.5 MPa and µ = 0.25. Hint: Calculate the effects of each shoe separately and superpose them.

* Answers

to these problems are provided in Appendix D. †

Problem numbers in italics are design problems.

17-12 Figure P17-2 shows a double short-shoe drum brake. Find its torque capacity and required actuating force for a = 12, b = 8, e = 3, r = 6 in, and θ = 25°. What value of a b shoe Fa

Y

θ

drum

c

r O1

e X

ω

17 FIGURE P17-1 Geometry for a Short-Shoe External Drum Brake

1032

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

a b shoe Fa

Y

θ

c

r O1

e

drum

X ω

e

O2

Fa

c

shoe FIGURE P17-2 Geometry for a Double Short-Shoe External Drum Brake

c will make it self-locking? Assume pmax = 200 psi, width = 2 in, and µ = 0.28. Hint: Calculate the effects of each shoe separately and superpose them. *

Answers to these problems are provided in Appendix D.

*17-13

Figure P17-3 shows a single long-shoe drum brake. Find its torque capacity and required actuating force for aX = 100, bX = 70, bY = 20, r = 30, w = 50 mm, and θ1 = 25°, θ2 = 125°. Assume pmax = 1.3 MPa and µ = 0.3.

17-14 Figure P17-3 shows a single long-shoe drum brake. Find its torque capacity and required actuating force for aX = 8, bX = 6, bY = 4, r = 5, w = 1.5 in, and θ1 = 35°, θ2 = 155°. Assume pmax = 250 psi and µ = 0.35. *17-15

Figure P17-4 shows a double long-shoe drum brake. Find its torque capacity and required actuating force for aX = 90, bX = 80, bY = 30, r = 40, w = 30 mm, and θ1 = 30°, θ2 = 160°. Assume pmax = 1.5 MPa and µ = 0.25. Hint: Calculate the effects of each shoe separately and superpose them.

17-16 Figure P17-4 shows a double long-shoe drum brake. Find its torque capacity and required actuating force for aX = 12, bX = 8, bY = 3, r = 6, w = 2 in, and θ1 = 25°, θ2 = 145°. Assume pmax = 200 psi and µ = 0.28. Hint: Calculate the effects of each shoe separately and superpose them. aX bX shoe Fa

r

y1

Y

drum

θ1

x1

θ2 O

bY X

ω

17 FIGURE P17-3 Geometry for a Long-Shoe External Drum Brake

Chapter 17

CLUTCHES AND BRAKES

1033

aX bX shoe Fa

r y1

Y

y2

θ1

x1

θ2 O1

bY

drum

X ω

Fa

θ1

bY

O2 θ2

shoe

x2

(a) Brake assembly

FIGURE P17-4 Geometry for a Double Long-Shoe External Drum Brake *17-17

The short-shoe approximation is considered to be valid for brake shoes with an included angle of up to about 45°. For the brake shown in Figure P17-3, calculate its torque capacity and required force by both the short-shoe method and the long-shoe method and compare the results for the following data: aX = 90, bX = 80, bY = 30, r = 40, w = 30 mm. Assume pmax = 1.5 MPa and µ = 0.25. Note that θ = θ2 – θ1 for the short-shoe approximation. (a) (b) (c)

*17-18

*

Answers to these problems are provided in Appendix D.

θ1 = 75°, θ2 = 105°. θ1 = 70°, θ2 = 110°. θ1 = 65°, θ2 = 115°.

Repeat Problem 17-17 for the brake design shown in Figure P17-4.

17-19 The short-shoe approximation is considered to be valid for brake shoes with an included angle of up to about 45°. For the brake shown in Figure P17-3, calculate its torque capacity and required force by both the short-shoe method and the long-shoe method and compare the results for the following data: aX = 8, bX = 6, bY = 4, r = 5, w = 2 in. Assume pmax = 250 psi and µ = 0.35. Note that θ = θ2 – θ1 for the short-shoe approximation. (a) (b) (c)

θ1 = 75°, θ2 = 105°. θ1 = 70°, θ2 = 110°. θ1 = 65°, θ2 = 115°.

17-20 Repeat Problem 17-19 for the brake design shown in Figure P17-4. *17-21

Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-11.

17-22 Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-12. *17-23

Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-13.

17-24 Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-14. *17-25

Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-15.

17

1034

* Answers

to these problems are provided in Appendix D. †

Problem numbers in italics are design problems.

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

17-26 Find the reaction forces at the arm pivot in the global XY system for the brake of Problem 17-16. †17-27

A clutch is needed for an electric motor that transmits 20 kW at 1100 rpm. The clutch will attach directly to the motor housing faceplate and is to have the same housing diameter as the motor, which is 125 mm. The minimum radial clearance between the housing OD and the clutch disk OD is 5 mm. The clutch output shaft will have the same diameter as the motor shaft, which is 15 mm. Design a multiple disk clutch for this application. State all assumptions and design choices. Specify the clutch material, outside disk radius, inside disk radius, and the required actuation force.

17-28 A clutch is needed for an electric motor that transmits 25 hp at 800 rpm. The clutch will attach directly to the motor housing faceplate and is to have the same housing diameter as the motor, which is 5.5 in. The minimum radial clearance between the housing OD and the clutch disk OD is 0.25 in. The clutch output shaft will have the same diameter as the motor shaft, which is 0.625 in. Design a multiple disk clutch for this application. State all assumptions and design choices. Specify the clutch material, outside disk radius, inside disk radius, and the required actuation force. *17-29

Find the torque that a dual-pad, caliper disc brake with pad angle of 60 deg can transmit if the outside and inside lining diameters are 160 mm and 90 mm, respectively, and the applied axial force is 3 kN. Assume uniform wear and µ = 0.35. Is the pressure on the lining acceptable? What lining materials would be suitable?

17-30 Repeat Problem 17-29 assuming uniform pressure. *17-31

Design a dual-pad caliper disc brake to provide a braking force of 240 N at the periphery of a 750-mm-dia wheel that is rotating at 670 rpm. Use an inside radius to outside radius ratio of 0.577. Assume uniform wear. State all assumptions and design choices. Specify the brake material, outside pad radius, inside pad radius, pad angle, and the required actuation force.

17-32 Repeat Problem 17-31 assuming uniform pressure. 17-33 An ultra-light solar racecar weighs 500 lb with driver. It has two 20-in-dia bicycle wheels in front that are to have dual-pad caliper disk brakes on each wheel. The brakes must be capable of bringing the car to a stop in a distance of 150 feet from a speed of 45 mph. Neglecting aerodynamic and rolling resistance forces, design dual-pad caliper disc brakes for the car. Use an inside radius to outside radius ratio of 0.577. Assume uniform wear. State all assumptions and design choices. Specify the brake material, outside pad radius, inside pad radius, pad angle, and the required actuation force. 17-34 Repeat Problem 17-33 assuming uniform pressure. 17-35 Determine the torque that a 4-surface, dry disk clutch can transmit if the outside and inside lining diameters are 150 mm and 80 mm, respectively, and the applied axial force is 15 kN. Assume uniform wear and μ = 0.35. Is the pressure on the lining acceptable? What lining materials would be suitable? 17-36 Find the torque that a dual-pad, caliper disc brake with pad angle of 30 deg can transmit if the outside and inside lining diameters are 150 mm and 120 mm, respectively, and the applied axial force is 300 N. Assume uniform wear and μ = 0.4. Is the pressure on the lining acceptable? What lining materials would be suitable? 17-37 Repeat Problem 17-36 assuming uniform pressure.

17

17-38 Design a dual-pad caliper disc brake to provide a braking force of 100 lb at the periphery of a 30-in-dia wheel that is rotating at 670 rpm. Use an inside radius to outside radius ratio of 0.577. Assume uniform wear. State all assumptions and design choices. Specify the brake material, outside pad radius, inside pad radius, pad angle, and the required actuation force.

Appendix MATERIAL PROPERTIES The following tables contain approximate values for strengths and other specifications of a variety of engineering materials compiled from various sources. In some cases the data are minimum recommended values and in other cases they are from a single test specimen. These data are suitable for use in the engineering exercises contained in this text, but should not be considered as statistically valid representations of specifications for any particular alloy or material. The designer should consult the materials’ manufacturers for more accurate and up-to-date strength information on materials used in engineering applications, or conduct independent tests of the selected materials to determine their ultimate suitability to any application. Much more information on material properties is available on the World Wide Web. Some sites are: http://www.matweb.com http://metals.about.com Table No.

Description

A-1

Physical Properties of Some Engineering Materials

A-2

Mechanical Properties for Some Wrought-Aluminum Alloys

A-3

Mechanical Properties for Some Aluminum Casting Alloys

A-4

Mechanical Properties for Some Wrought- and Cast-Copper Alloys

A-5

Mechanical Properties for Some Titanium Alloys

A-6

Mechanical Properties for Some Magnesium Alloys

A-7

Mechanical Properties for Some Cast Iron Alloys

A-8

Mechanical Properties for Some Stainless Steel Alloys

A-9

Mechanical Properties for Some Carbon Steels

A-10

Mechanical Properties for Some Alloy and Tool Steels

A-11

Mechanical Properties for Some Engineering Plastics

1035

A

1036

MACHINE DESIGN

Table A-1

-

An Integrated Approach

-

Sixth Edition

Physical Properties of Some Engineering Materials

Data from Various Sources.* These Properties Are Essentially Similar for All Alloys of the Particular Material

Material

Modulus of Elasticity Mpsi

Poisson’s Ratio ν

Modulus of Rigidity

GPa

Mpsi

GPa

Weight Density γ

Mass Specific Density ρ Gravity

lb/in3

Mg/m3

Aluminum Alloys

10.4

71.7

3.9

26.8

0.34

0.10

2.8

2.8

Beryllium Copper

18.5

127.6

7.2

49.4

0.29

0.30

8.3

8.3

Brass, Bronze

16.0

110.3

6.0

41.5

0.33

0.31

8.6

8.6

Copper

17.5

120.7

6.5

44.7

0.35

0.32

8.9

8.9

Iron, Cast, Gray

15.0

103.4

5.9

40.4

0.28

0.26

7.2

7.2

Iron, Cast, Ductile

24.5

168.9

9.4

65.0

0.30

0.25

6.9

6.9

Iron, Cast, Malleable

25.0

172.4

9.6

66.3

0.30

0.26

7.3

7.3

6.5

44.8

2.4

16.8

0.33

0.07

1.8

1.8

Nickel Alloys

30.0

206.8

11.5

79.6

0.30

0.30

8.3

8.3

Steel, Carbon

30.0

206.8

11.7

80.8

0.28

0.28

7.8

7.8

Steel, Alloys

30.0

206.8

11.7

80.8

0.28

0.28

7.8

7.8

Steel, Stainless

27.5

189.6

10.7

74.1

0.28

0.28

7.8

7.8

Titanium Alloys

16.5

113.8

6.2

42.4

0.34

0.16

4.4

4.4

Zinc Alloys

12.0

82.7

4.5

31.1

0.33

0.24

6.6

6.6

Magnesium Alloys

* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.; Metals Handbook, American Society for Metals, Materials Park, Ohio.

Table A-2

Mechanical Properties for Some Wrought-Aluminum Alloys Data from Various Sources.* Approximate Values. Consult Material Manufacturers for More Accurate Information

WroughtAluminum Alloy 1100 2024 3003 5052 6061 7075

Condition

Tensile Yield Strength (0.2% offset) kpsi MPa

Ultimate Tensile Strength kpsi MPa

Fatigue Strength at 5E8 cycles kpsi MPa

Elongation Brinell over 2 in Hardness

%

-HB

5

34

13

90

35

23

cold rolled

22

152

24

165

5

44

sheet annealed

11

76

26

179

20



heat treated

42

290

64

441

19



sheet annealed

20

138

6

41

16

110

30

28

cold rolled

27

186

29

200

4

55

sheet annealed

13

90

28

193

25

47

cold rolled

37

255

42

290

7

77

25

30

12

95

sheet annealed

8

55

18

124

heat treated

40

276

45

310

bar annealed

15

103

33

228

heat treated

73

503

83

572

sheet annealed

14 14

97 97

16

60

11

150

*Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.; Metals Handbook, American Society for Metals, Materials Park, Ohio.

A

Appendix A

Table A-3

1037

Mechanical Properties for Some Aluminum Casting Alloys

Data from INCO.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Aluminum Casting Alloy

Condition

43

permanent mold casting—as cast

195

MATERIAL PROPERTIES

Tensile Yield Strength (0.2% offset) kpsi MPa

sand casting—as cast

Ultimate Tensile Strength kpsi MPa

Elongation Brinell over 2 in Hardness

%

-HB

9

62

23

159

10

45

24

165

36

248

5

– 75

220

sand casting—solution heat treated

26

179

48

331

16

380

die casting—as cast

24

165

48

331

3

A132

permanent mold casting—heat treated + 340°F

43

296

47

324

0.5

125

A142

sand casting—heat treated + 650°F

30

207

32

221

0.5

85



* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.

Table A-4

Mechanical Properties for Some Wrought- and Cast-Copper Alloys

Data from INCO.* Approximate Values. Consult Material Manufacturers for More Accurate Information Tensile Yield Strength (0.2% offset) kpsi MPa

Ultimate Tensile Strength kpsi MPa

Elongation Brinell or over 2 in Rockwell Hardness

Copper Alloy

Condition

CA110—Pure Copper

strip annealed

10

69

32

221

45

40HRF

spring temper

50

345

55

379

4

60HRB

strip annealed plus age

145

1000

165

1138

7

35HRC

hard plus age

170

1172

190

1310

3

40HRC

strip annealed

10

69

37

255

45

53HRF

spring temper

62

427

72

496

3

78HRB

strip annealed

15

103

40

276

50

50HB

hard temper

60

414

75

517

7

135HB

strip annealed

11

76

44

303

66

54HRF

spring temper

65

448

94

648

3

91HRB

strip annealed

14

97

46

317

65

58HRF

spring temper

62

427

91

627

30

90HRB

annealed

19

131

47

324

64

73HRF

spring temper

80

552

100

689

4

95HRB

soft

45

310

82

565

40

84HRB

hard

60

414

89

614

32

87HRB

annealed

21

145

56

386

63

76HRF

spring temper

62

427

110

758

4

97HRB

CA675—Manganese Bronze

soft

30

207

65

448

33

65HRB

Leaded-Tin Bronze

half-hard as cast

60 19

414 131

84 34

579 234

19 18

90HRB 60HB

Nickel-Tin Bronze

as cast

20

138

50

345

40

85HB

cast and heat treated

55

379

85

586

10

180HB

CA170—Beryllium Copper CA220—Commercial Bronze CA230—Red Brass CA260—Cartridge Brass CA270—Yellow Brass CA510—Phosphor Bronze CA614—Aluminum Bronze CA655—High Silicon Bronze

%

* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.

index flag for Table C3 index flag for Table C4

B A

1038

Table A-5

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

Mechanical Properties for Some Titanium Alloys

Data from INCO.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Titanium Alloy

Condition

Ti-35A

Tensile Yield Strength (0.2% offset)

sheet annealed

Ultimate Tensile Strength

Elongation over 2 in

kpsi

MPa

kpsi

MPa

%

30

207

40

276

30

Brinell or Rockwell Hardness 135HB

Ti-50A

sheet annealed

45

310

55

379

25

215HB

Ti-75A

sheet annealed

75

517

85

586

18

245HB

Ti-0.2Pd Alloy

sheet annealed

45

310

55

379

25

215HB

Ti-5 Al-2.5 Sn Alloy

annealed

125

862

135

931

13

39HRC

Ti-8 Al-1 Mo-1 V Alloy

sheet annealed

130

896

140

965

13

39HRC

Ti-8 Al-2 Sn-4 Zr-2 Mo Alloy

bar annealed

130

896

140

965

15

39HRC

Ti-8 Al-6 V-2 Sn Alloy

sheet annealed

155

1069

165

1138

12

41HRC

Ti-6 Al-4 V Alloy

sheet annealed

130

896

140

13

Ti-6 Al-4 V Alloy

heat treated

165

1138

175

1207

12

2.5

39HRC –

T1-13 V-11 Cr-3 Al Alloy

sheet annealed

130

896

135

931

13

37HRC

T1-13 V-11 Cr-3 Al Alloy

heat treated

170

1172

180

1241

6



* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.

Table A-6

Mechanical Properties for Some Magnesium Alloys

Data from INCO.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Magnesium Alloy

Condition

Tensile Yield Strength (0.2% offset)

Elongation over 2 in

Brinell or Rockwell Hardness

kpsi

MPa

kpsi

MPa

%

sheet annealed

22

152

37

255

21

56HB

hard sheet

32

221

42

290

15

73HB

as forged

33

228

48

331

11

69HB

forged and aged

36

248

50

345

6

72HB

AZ91A & AZ91B

die cast

22

152

33

228

3

63HB

AZ91C

as cast

14

97

24

165

2.5

60HB

cast, solution treated and aged

19

131

40

276

5

70HB

as cast

14

97

25

172

2

65HB

cast, solution treated

14

97

40

276

10

63HB

cast, solution treated and aged

22

152

40

276

3

81HB

EZ33A

cast and aged

16

110

23

159

3

50HB

HK31A

strain hardened

29

200

37

255

8

68HB

cast and heat treated

15

103

32

221

8

66HRB

HZ32A

cast - solution treated and aged

13

90

27

186

4

55HB

ZK60A

as extruded

38

262

49

338

14

75HB

extruded and aged

44

303

53

365

11

82HB

AZ 31B AZ 80A

AZ92A

* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.

A

Ultimate Tensile Strength

Appendix A

Table A-7

MATERIAL PROPERTIES

1039

Mechanical Properties for Some Cast-Iron Alloys Data from Various Sources.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Cast Iron Alloy

Condition

Tensile Yield Strength (0.2% offset)

Ultimate Tensile Strength

Compressive Strength

Brinell Hardness

kpsi

MPa

kpsi

MPa

kpsi

MPa

Gray Cast Iron—Class 20

as cast





22

152

83

572

-HB 156

Gray Cast Iron—Class 30

as cast





32

221

109

752

210

Gray Cast Iron—Class 40

as cast





42

290

140

965

235

Gray Cast Iron—Class 50

as cast





52

359

164

1131

262

Gray Cast Iron—Class 60

as cast





62

427

187

1289

302

Ductile Iron 60-40-18

annealed

47

324

65

448

52

359

160

Ductile Iron 65-45-12

annealed

48

331

67

462

53

365

174

Ductile Iron 80-55-06

annealed

53

365

82

565

56

386

228

Ductile Iron 120-90-02

Q&T

120

827

140

965

134

924

325

* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.; Metals Handbook, American Society for Metals, Materials Park, Ohio.

index flag for Table C7 index flag for Table C8 Table A-8

Mechanical Properties for Some Stainless Steel Alloys

Data from INCO.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Stainless Steel Alloy Type 301

Condition

strip annealed cold rolled

Type 302

sheet annealed cold rolled

Type 304

sheet annealed cold rolled

Tensile Yield Strength (0.2% offset) kpsi MPa

Ultimate Tensile Strength kpsi MPa

Elongation over 2 in

%

Brinell or Rockwell Hardness

40

276

110

758

60

85HRB

165

1138

200

1379

8

41HRC

40

276

90

621

50

85HRB

165

1138

190

1310

5

40HRC

35

241

85

586

50

80HRB

160

1103

185

1276

4

40HRC

100

689

45

180HB

Type 314

bar annealed

50

345

Type 316

sheet annealed

40

276

90

621

50

85HRB

Type 330

hot rolled

55

379

100

689

35

200HB

annealed

35

241

80

552

50

150HB

sheet annealed

45

310

70

483

25

80HRB

heat treated

140

965

180

1241

15

39HRC

bar annealed

50

345

95

655

25

92HRB

heat treated

195

1344

230

1586

8

500HB

bar annealed

95

655

125

862

25

260HB

heat treated

150

1034

195

1344

15

400HB

bar annealed

65

448

110

758

14

230HB

Q & T @ 600°F

275

1896

285

1965

2

57HRC

17-4 PH (AISI 630)

hardened

185

1276

200

1379

14

44HRC

17-7 PH (AISI 631)

hardened

220

1517

235

1620

6

48HRC

Type 410 Type 420 Type 431 Type 440C

* Properties of Some Metals and Alloys, International Nickel Co., Inc., N.Y.

B A

1040

MACHINE DESIGN

Table A-9 SAE / AISI Number 1010 1020 1030

1035 1040

1045 1050

1060

1095

-

An Integrated Approach

-

Sixth Edition

Mechanical Properties for Some Carbon Steels

Data from Various Sources.* Approximate Values. Consult Material Manufacturers for More Accurate Information Condition

Tensile Yield Strength (0.2% offset) kpsi MPa

Ultimate Tensile Strength kpsi MPa

Elongation over 2 in

Brinell Hardness

%

-HB

hot rolled

26

179

47

324

28

95

cold rolled

44

303

53

365

20

105

hot rolled

30

207

55

379

25

111

cold rolled

57

393

68

469

15

131

hot rolled

38

259

68

469

20

137

normalized @ 1650°F

50

345

75

517

32

149

cold rolled

64

441

76

524

12

149

quench & temper @ 1000°F

75

517

97

669

28

255

quench & temper @ 800°F

84

579

106

731

23

302

quench & temper @ 400°F

94

648

123

848

17

495

hot rolled

40

276

72

496

18

143

cold rolled

67

462

80

552

12

163

hot rolled

42

290

76

524

18

149

normalized @ 1650°F

54

372

86

593

28

170

cold rolled

71

490

85

586

12

170

quench & temper @ 1200°F

63

434

92

634

29

192

quench & temper @ 800°F

80

552

110

758

21

241

quench & temper @ 400°F

86

593

113

779

19

262

hot rolled

45

310

82

565

16

163

cold rolled

77

531

91

627

12

179

hot rolled

50

345

90

621

15

179

normalized @ 1650°F

62

427

108

745

20

217

cold rolled

84

579

100

689

10

197

quench & temper @ 1200°F

78

538

104

717

28

235

quench & temper @ 800°F

115

793

158

1089

13

444

quench & temper @ 400°F

117

807

163

1124

9

514

hot rolled

54

372

98

676

12

200

normalized @ 1650°F

61

421

112

772

18

229

quench & temper @ 1200°F

76

524

116

800

23

229

quench & temper @ 1000°F

97

669

140

965

17

277

quench & temper @ 800°F

111

765

156

1076

14

311

hot rolled

66

455

120

827

10

248

normalized @ 1650°F

72

496

147

1014

9

13

quench & temper @ 1200°F

80

552

130

896

21

269

quench & temper @ 800°F

112

772

176

1213

12

363

quench & temper @ 600°F

118

814

183

1262

10

375

* SAE Handbook, Society of Automotive Engineers, Warrendale, Pa.; Metals Handbook, American Society for Metals, Materials Park, Ohio.

index flags for Table C9

A

Appendix A

Table A-10 SAE / AISI Number 1340

Data from Various Sources.* Approximate Values. Consult Material Manufacturers for More Accurate Information Condition

annealed annealed

4140

4340

6150

%

Brinell or Rockwell Hardness

63

434

102

703

25

204HB

109

752

125

862

21

250HB

47

324

75

517

30

150HB

779

132

910

12

264HB

359

81

558

28

156HB

normalized @ 1650°F

63

434

97

669

25

197HB

quench & temper @ 1200°F

102

703

118

814

22

245HB

quench & temper @ 800°F

173

1193

186

1282

13

380HB

quench & temper @ 400°F

212

1462

236

1627

10

41HB

annealed @ 1450°F

61

421

95

655

26

197HB

normalized @ 1650°F

95

655

148

1020

18

302HB

quench & temper @ 1200°F

95

655

110

758

22

230HB

quench & temper @ 800°F

165

181

1248

13

370HB

quench & temper @ 400°F

238

1138 1641

257

1772

8

510HB

quench & temper @ 1200°F

124

855

140

965

19

280HB

quench & temper @ 1000°F

156

1076

170

1172

13

360HB

quench & temper @ 800°F

198

1365

213

1469

10

430HB

quench & temper @ 600°F

230

1586

250

1724

10

486HB

59

407

96

662

23

192HB

148

1020

157

1082

16

314HB

60

414

95

655

25

190HB

133

917

144

993

18

288HB

53

365

100

689

25

96HRB

250

1724

295

2034

9

55HRC

74

510

103

710

25

96HRB

260

1793

290

1999

5

54HRC

55

379

95

655

25

93HRB

260

1793

290

1999

4

54HRC

75

517

100

689

17

97HRB

205

1413

270

1862

10

52HRC

60

414

100

689

24

96HRB

275

1896

300

2068

4

57HRC

64

441

105

724

25

96HRB

280

1931

340

2344

5

59HRC

annealed annealed

H-11

annealed @ 1600°F

L-2

annealed @ 1425°F

L-6

annealed @ 1425°F

P-20

annealed @ 1425°F

S-1

annealed @ 1475°F

S-5

annealed @ 1450°F

quench & temper @ 1000°F quench & temper @ 400°F quench & temper @ 600°F quench & temper @ 400°F quench & temper @ 400°F quench & temper @ 400°F annealed @ 1525°F quench & temper @ 400°F A-8

Elongation over 2 in

52

quench & temper

S-7

Ultimate Tensile Strength kpsi MPa

113

quench & temper 8740

Tensile Yield Strength (0.2% offset) kpsi MPa

annealed @ 1450°F

quench & temper 4130

1041

Mechanical Properties for Some Alloy and Tool Steels

quench & temper 4027

MATERIAL PROPERTIES

annealed @ 1550°F quench & temper @ 1050°F

55

379

93

641

25

95HRB

210

1448

315

2172

7

58HRC

65

448

103

710

24

97HRB

225

1551

265

1827

9

52HRC

* Machine Design Materials Reference Issue, Penton Publishing, Cleveland Ohio; Metals Handbook, ASM, Materials Park, Ohio.

B A

1042

MACHINE DESIGN

Table A-11

-

An Integrated Approach

-

Sixth Edition

Mechanical Properties of Some Engineering Plastics Data from Various Sources.* Approximate Values. Consult Material Manufacturers for More Accurate Information

Material

Approximate Modulus of Elasticity E

Ultimate Tensile Strength

Ultimate Compressive Strength

Elongation over 2 in

Max Temp

Specific Gravity

Mpsi

GPa

kpsi

MPa

kpsi

MPa

%

°F

0.3

2.1

6.0

41.4

10.0

68.9

5 to 25

160–200

1.05

0.6

4.1

10.0

68.9

12.0

82.7

3

200–230

1.30

0.5

3.4

8.8

60.7

18.0

124.1

60

220

1.41

1.0

6.9

10.0

68.9

18.0

124.1

7

185–220

1.56

Acrylic

0.4

2.8

10.0

68.9

15.0

103.4

5

140–190

1.18

Fluoroplastic (PTFE)

0.2

1.4

5.0

34.5

6.0

41.4

100

350–330

2.10

Nylon 6/6

0.2

1.4

10.0

68.9

10.0

68.9

60

180–300

1.14

Nylon 11

0.2

1.3

8.0

55.2

8.0

55.2

300

180–300

1.04

0.4

2.5

12.8

88.3

12.8

88.3

4

250–340

1.26

0.4

2.4

9.0

62.1

12.0

82.7

100

250

1.20

ABS 20—40% glass filled Acetal 20—30% glass filled

20—30% glass filled Polycarbonate

1.0

6.9

17.0

117.2

17.0

117.2

2

275

1.35

HMW Polyethylene

0.1

0.7

2.5

17.2





525



0.94

Polyphenylene Oxide

0.4

2.4

9.6

66.2

16.4

113.1

20

212

1.06

1.1

7.8

15.5

106.9

17.5

120.7

5

260

1.23

0.2

1.4

5.0

34.5

7.0

48.3

500

250–320

0.90

0.7

4.8

7.5

51.7

6.2

42.7

2

300–320

1.10

0.3

2.1

4.0

27.6

6.0

41.4

2 to 80

140–175

1.07

0.1

0.7

12.0

82.7

16.0

110.3

1

180–200

1.25

0.4

2.5

10.2

70.3

13.9

95.8

50

300–345

1.24

10—40% glass filled

20—30% glass filled Polypropylene 20—30% glass filled Impact Polystyrene 20—30% glass filled Polysulfone

* Modern Plastics Encyclopedia, McGraw-Hill, New York; Machine Design Materials Reference Issue, Penton Publishing, Cleveland, Ohio.

index flags for Table C11

A

Appendix BEAM TABLES Loading, shear, moment, slope, and deflection functions for a selection of common beam configurations and loadings are presented in these tables. Cantilever, simply supported, and overhung beams with either a concentrated load at any point or a uniformly distributed load across any portion of the span are defined. A general set of equations is derived for each beam. Special cases, such as those with the load at center span, are accommodated by appropriate choice of dimensions in the general formulas. In all cases, singularity functions are used to write the beam equations, which gives a single expression for the entire span for each function. See Section 3.17 for a discussion of singularity functions. The equations for the beam cases in this appendix have been encoded in TK Solver files, which are provided on the website for this text. In some cases, these files allow multiple loads to be applied at different locations on the beam, but the derivations in this appendix each accommodate only one load per beam. Use superposition to combine various beam cases when more than one type of load is present on a beam. For a more complete collection of beam formulas, see Roark and Young, Formulas for Stress and Strain, 6th ed., McGraw-Hill, New York, 1989. A key to the figures in this appendix and their related files follows.

Figure No.

Case

File Name

B-1a

Cantilever beam with concentrated load

cantconc

B-1b

Cantilever beam with uniformly distributed load

cantunif

B-2a

Simply supported beam with concentrated load

simpconc

B-2b

Simply supported beam with uniformly distributed load

simpunif

B-3a

Overhung beam with concentrated load

ovhgconc

B-3b

Overhung beam with uniformly distributed load

ovhgunif

1043

B

1044

MACHINE DESIGN

-

(a) Cantilever beam with concentrated loading

An Integrated Approach

M1

l

F 〈x–a〉 – 1

R1 = F x

q = M1 x

Loading

−2

+ R1 x

V

−1

−1

−F x−a

x

V = M1 x

x

−1

(

+ R1 − F x − a

= F 1− x − a

Shear

0

0

)

0

x

Shear

0

Mmax

(

= F −a + x − x − a

Moment

0

x

θ=

Slope

x

0

y

1

)

(

F −2ax + x 2 − x − a 2 EI

ymax =

)

Fa 2 ( a − 3l ) 6 EI

when a = l: ymax = −

ymax

2

M

Mmax

Fl 3 3EI

1  M1 2 R1 3 F  x + x − x − a 3 −  EI  2 6 6 F = x 3 − 3ax 2 − x − a 3 6 EI

0

(

)

(

w 2 l − a2 2

)

wl 2 2

θ=

(

(

)

w 3 (l − a ) x 2 − 3 l 2 − a 2 x − x − a 6 EI

x

0

)

R   − M1 x + 1 x 2  1  2 θ=  w EI  x − a 3 −   6

θmax

θ

y

1

+ R1 − w x − a

M max = M1 =

ymax =

3

)

(

w −3l 4 + 4 a3l − a 4 24 EI

when a = 0: ymax = −

ymax

)

wl 4 8 EI

1  M1 2 R1 3 w  x + x − x − a 4 −  EI  2 6 24 w 3 2 2 2 = 4 (l − a ) x − 6 l − a x − x − a 24 EI

y= Deflection

(

FIGURE B-1

B

−1

= w ( l − a ) − x − a 1 

x

y= Deflection

V = M1 x

(

Slope

0

−w x−a

M = − M1 + R1 x − w x − a 2 w =  2 (l − a ) x − l 2 − a 2 − x − a 2   2

Moment

R   − M1 x + 1 x 2  1  2 θ=  F EI  x − a 2 −   2

θmax

θ

−1

when a = 0: M max =

M = − M1 + R1 x − F x − a 1

+ R1 x

)

Vmax = R1 = w ( l − a )

x

when a = l: M max = − Fl M

−2

q = M1 x

Vmax

M max = − Fa

x

R1 Loading

(

w 2 l − a2 2

M1 =

V

Vmax = R1 = F

Vmax

0

a

M1

R1 = w ( l − a )

w 〈x–a〉 0

M1 = Fa

R1

0

Sixth Edition

(b) Cantilever beam with uniformly distributed loading

l a

-

Cantilever Beams with Concentrated or Distributed Loading. Note: < > Denotes a Singularity Function

(

)

4

)

Appendix B

(a) Simply supported beam with concentrated loading

a

F 〈x–a〉

–1

R1

l a

R1

q = R1 x

−1

V

−F x−a

−1

−1

+ R2 x − l

0

 M max = Fa  1 − 

Mmax

a  l

M = R1 x − F x − a + R2 x − l

θ

θmax

0

x

FIGURE B-2

0

−1

+ R2 x − l

Mmax when a = 0: M =

0

wx (l − x ) 2

x M = R1 x − =

Moment

0 a 2  2  1 −  x − x − a  l  a  + − a 2 + 3al − 2l 2  3l

ymax

Deflection

−w x−a

1  = w  ( l − a )2 − x − a 1   2l 

1

(

)

    

x ymax

)

w x−a 2

2

+ R2 x − l

1

w x (l − a )2 − x − a 2  2  l 

θmax

F θ= 2 EI

0

y

M

a   = F  1 −  x − x − a 1    l  

Moment

−1

V = R1 − w x − a 1 + R2 x − l

0

1

(

w 2 l − a2 2l

Vmax = MAX( R1 , R2 )

x

Shear

l Fl when a = : M max = 2 4

x

Slope

q = R1 x

0

+ R2 x − l

a   = F 1 − − x − a 0    l

Shear

0

Loading

w ( l − a )2 2l

R2

0

x V = R1 − F x − a

R2 =

V

Vmax = MAX( R1 , R2 )

0

R1 =

w 〈x–a〉 0 x

R2

Loading

M

a  l

 a R2 = F   l

x

1045

(b) Simply supported beam with uniformly distributed loading

 R1 = F  1 − 

l

BEAM TABLES

F = 3EI

 3 a4 2  2a − l − la   

a   1 −  x3 − x − a 3  F  l y=   6 EI  a + − a 2 + 3al − 2l 2 x    l

(

)

θ Slope

0

x   6x 2 l − a )2 − 4 x − a 3   w  l ( θ=   24 EI  1 + ( l − a )4 − 2l 2 ( l − a )2      l x ymax

when a = 0 : wx y= 2lx 2 − x 3 − l 3 24 EI

(

)

y

Deflection

w y= 24 EI

 2x3  ( l − a )2 − x − a 4   l    + x ( l − a )4 − 2l 2 ( l − a )2      l 

Simply Supported Beams with Concentrated or Distributed Loading. Note: < > Denotes a Singularity Function

B B

1046

MACHINE DESIGN

-

(a) Overhung beam with concentrated loading l F 〈x–a〉 – 1 x

R2

R1

q = R1 x

Loading

l

−F x−a

−1

+ R2 x − b

−1

w 〈x–a〉 0

x

b

wa 2 2b

R1

R2 Loading q = R1 x −1 − w x 0 + w x − a 0 + R2 x − b −1

x

0 x

0

b−a V =F x  b

Shear

0

+

a x−b b

0

 − x−a 0 

 a2 V = w a − x + x − a 1 + 2b 

Shear

x

0 Mmax

Moment

θ

θ 0

0

Slope

a b−a x 2 + x−b F  b b θ=  b 2 EI  + (a − b)  3

0 y

Deflection

FIGURE B-3

2

 − x−a 2   

x

θmin Slope

0

ymax a b−a x 3 + x−b F  b b y= 6 EI  +b (a − b) x 

)

 −1  

y 3

 − x −a 3  

    1 

θmax

x θmax

0

 a2    2a −  x − x 2 b  w  M =  2 2  + x−a 2 + a x−b  b

M b−a 1 a  M =F x + x − b 1 − x − a 1 b  b 

Moment

( x−b

Mmax

x

0

B

R2 =

V

V

M

Sixth Edition

a   R1 = wa  1 −   2b 

a

 a R2 = F    b −1

-

(b) Overhung beam with uniformly distributed loading

 b− a R1 = F   b 

a b

An Integrated Approach

 1  2 4 2 3 1  24  2a b − 4 ab + b − b b − a       a a2  2 1 3 w    θ= + −  x − x EI  6  2 4b     2 a 2 1  3  + x−b + x−a   4b 6  x

ymax

Deflection

 2 4  2 3 1  2a b − 4 ab + b − b b − a  x      w  2a 2  3 4   y= x −x +  4a −  b  24 EI      2 2a 3  4 x−b  + x−a +  b 

Overhung Beams with Concentrated or Distributed Loading. Note: < > Denotes a Singularity Function

Appendix STRESSCONCENTRATION FACTORS Stress-concentration factors for 14 common cases are presented in this appendix as listed below. Data for all curves were manually read, by the author, from curves in R. E. Peterson, “Design Factors for Stress Concentration, Parts 1 to 5, ” Machine Design, February–July, 1951, Penton Publishing, Cleveland, Ohio. These data were then typed into Excel, plotted, and approximate equations for their functions fitted. The r-square values were all above 0.97. These equations were used to generate the curves for each figure in this appendix, and the equation parameters are defined in each figure. These equations also have been encoded as computerized functions (noted below) that can be incorporated in computer models to allow automatic generation of approximate stressconcentration factors during calculations. Figure

Case

File Name

C-1

Shaft with Shoulder Fillet in Axial Tension

APP_C-01

C-2

Shaft with Shoulder Fillet in Bending

APP_C-02

C-3

Shaft with Shoulder Fillet in Torsion

APP_C-03

C-4

Shaft with Groove in Axial Tension

APP_C-04

C-5

Shaft with Groove in Bending

APP_C-05

C-6

Shaft with Groove in Torsion

APP_C-06

C-7

Shaft with Transverse Hole in Bending

APP_C-07

C-8

Shaft with Transverse Hole in Torsion

APP_C-08

C-9

Flat Bar with Fillet in Axial Tension

APP_C-09

C-10

Flat Bar with Fillet in Bending

APP_C-10

C-11

Flat Bar with Notch in Axial Tension

APP_C-11

C-12

Flat Bar with Notch in Bending

APP_C-12

C-13

Flat Bar with Transverse Hole in Axial Tension

APP_C-13

C-14

Flat Bar with Transverse Hole in Bending

APP_C-14

1047

C

1048

MACHINE DESIGN

-

An Integrated Approach

3.0

2.6

1.50 1.30

2.4

P

P d

D

1.20

2.2

1.15

1.05

Kt 2.0

1.10

1.8

1.02

1.6

1.01

1.07

1.05

1.4 1.2 1.0 0

0.05

0.10

0.15

0.20

Sixth Edition

r Kt ≅ A   d where:

r

D / d = 2.0

2.8

-

0.25

b

D/d

A

b

2.00 1.50 1.30 1.20 1.15 1.10 1.07 1.05 1.02 1.01

1.014 70 0.999 57 0.996 82 0.962 72 0.980 84 0.984 50 0.984 98 1.004 80 1.012 20 0.984 13

–0.300 35 –0.282 21 –0.257 51 –0.255 27 –0.224 85 –0.208 18 –0.195 48 –0.170 76 –0.124 74 –0.104 74

0.30

r/d FIGURE C-1

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Shaft with a Shoulder Fillet in Axial Tension

3.0

b r Kt ≅ A   d where:

r

2.8 2.6

D

D / d = 6.0

2.4

3.0 M 2.0 1.50 1.20

2.2 Kt 2.0 1.8

d M

1.10

1.05

1.03

1.6

1.02

1.01

1.4 1.2 1.0 0

0.05

0.10

0.15

0.20

0.25

0.30

D/d

A

b

6.00 3.00 2.00 1.50 1.20 1.10 1.07 1.05 1.03 1.02 1.01

0.878 68 0.893 34 0.908 79 0.938 36 0.970 98 0.951 20 0.975 27 0.981 37 0.980 61 0.960 48 0.919 38

–0.332 43 –0.308 60 –0.285 98 –0.257 59 –0.217 96 –0.237 57 –0.209 58 –0.196 53 –0.183 81 –0.177 11 –0.170 32

r/d FIGURE C-2 Geometric Stress-Concentration Factor Kt for a Shaft with a Shoulder Fillet in Bending

C

Copyright © 2018 Robert L. Norton: All Rights Reserved

Appendix C

STRESS-CONCENTRATION FACTORS

1049

3.0 r

2.8 2.6

T

2.4

b r Kt ≅ A   d

T

where:

d

D

2.2 D / d = 2.0

Kt 2.0

1.33

1.8 1.6

1.20

D/d

A

b

2.00 1.33 1.20 1.09

0.863 31 0.848 97 0.834 25 0.903 37

–0.238 65 –0.231 61 –0.216 49 –0.126 92

1.09

1.4 1.2 1.0 0

0.05

0.30

0.25

0.20

0.15

0.10

r/d FIGURE C-3

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Shaft with a Shoulder Fillet in Torsion

1.05

D / d = 1.07 1.10

1.03

1.15

1.02

Kt

1.20

b r Kt ≅ A   d where:

r

D

P

P d 1.30

1.01

1.50

2.0 ∞

D/d

A

b

∞ 2.00 1.50 1.30 1.20 1.15 1.10 1.07 1.05 1.03 1.02 1.01

0.993 72 0.993 83 0.998 08 1.004 90 1.010 70 1.026 30 1.027 20 1.023 80 1.027 20 1.036 70 1.037 90 1.000 30

–0.393 52 –0.382 31 –0.369 55 –0.355 45 –0.337 65 –0.316 73 –0.294 84 –0.276 18 –0.252 56 –0.216 03 –0.187 55 –0.156 09

r/d FIGURE C-4

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Grooved Shaft in Axial Tension

B C

1050

MACHINE DESIGN

-

An Integrated Approach

3.0 r

D

2.8

d

D / d = 1.05

2.6

1.07

1.03

1.10

2.4 1.02

2.2

1.30

1.01

Kt 2.0

M

M

1.15 2.0



1.8 1.6 1.4 1.2 1.0 0

0.05

0.15

0.10

0.25

0.20

0.30

r/d FIGURE C-5

-

Sixth Edition

r Kt ≅ A   d where:

b

D/d

A

b

∞ 2.00 1.50 1.30 1.20 1.15 1.12 1.10 1.07 1.05 1.03 1.02 1.01

0.948 01 0.936 19 0.938 94 0.942 99 0.946 81 0.953 11 0.955 73 0.954 54 0.967 74 0.987 55 0.990 33 0.977 53 0.993 93

–0.333 02 –0.330 66 –0.323 80 –0.315 04 –0.305 82 –0.297 39 –0.288 86 –0.282 68 –0.264 52 –0.241 34 –0.215 17 –0.197 93 –0.152 38

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Grooved Shaft in Bending

3.0 D

2.8 D / d = 1.10

2.6

r

d

T

b r Kt ≅ A   d

T

where:

1.20

2.4

1.30

2.2

2.0

Kt 2.0



1.05

1.8

1.02 1.01

1.6 1.4 1.2

D/d

A

b

∞ 2.00 1.30 1.20 1.10 1.05 1.02 1.01

0.881 26 0.890 35 0.894 60 0.901 82 0.923 11 0.938 53 0.968 77 0.972 45

–0.252 04 –0.240 75 –0.232 67 –0.223 34 –0.197 40 –0.169 41 –0.126 05 –0.101 62

1.0 0

0.05

0.10

0.15

0.20

0.25

0.30

r/d FIGURE C-6 Geometric Stress-Concentration Factor Kt for a Grooved Shaft in Torsion

C

Copyright © 2018 Robert L. Norton: All Rights Reserved

Appendix C

STRESS-CONCENTRATION FACTORS

1051

3.0 2.9

M

M

2.8 D

2.7 2.6 Kt

2.5

1.50

d 2.0

2.4

d K t ≅ 1.589 90 − 0.635 50 log    D

2.3 2.2

on s

urfac

2.1

e of

2.0

shaf

t at h

ole

1.9 0

0.05

0.10

0.20

0.15

0.30

0.25

d/D FIGURE C-7

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Shaft with a Transverse Hole in Bending

4.0 K t B ≅ 3.9702 − 9.292

3.9 T

3.8

T

2 3 d d +27.159   + 30.231    D  D

3.7 D

3.6

4 5 d d −393.19   + 650.39    D  D

3.5 3.4

B

Kt 3.3

6 d +15.451    D

d

3.2 below

3.1 A

3.0

shaft

surfa

ce in

2.9

on surfa

2.8

K t A ≅ 3.921 50 − 24.435 hole

ft at hole

4 5 d d +3059.5   − 3042.4    D  D

2.6 0

0.05

0.10

0.15

0.20

0.2

d D

2 3 d d +234.06   − 1200.5    D  D

ce of sha

2.7

d D

.30

d/D FIGURE C-8

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Shaft with a Transverse Hole in Torsion

B C

1052

MACHINE DESIGN

-

An Integrated Approach

3.0 D / d = 2.0

2.8

1.15

1.30

1.10

1.20

1.07

2.2

r

D

1.05

Kt 2.0

3.0

1.8

1.02

1.6

1.01

1.4 1.2 1.0 0

Sixth Edition

b r Kt ≅ A   d where:

d

1.50

2.6 2.4

h

-

0.05

0.10

0.15

0.20

0.25

D/d

A

b

2.00 1.50 1.30 1.20 1.15 1.10 1.07 1.05 1.02 1.01

1.099 60 1.076 90 1.054 40 1.035 10 1.014 20 1.013 00 1.014 50 1.026 10 1.025 90 0.976 62

–0.320 77 –0.295 58 –0.270 21 –0.250 84 –0.239 35 –0.215 35 –0.193 66 –0.170 85 –0.169 78 –0.106 56

0.30

r/d FIGURE C-9

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Filleted Flat Bar in Axial Tension

b r Kt ≅ A   d

3.0 h

D / d = 6.0

2.8

where:

3.0

2.6 2.4

2.0 1.30

2.2

d

D M

M

r

1.20

Kt 2.0

1.10

1.8

1.07

1.6 1.4

1.01

1.2

1.02

1.03

1.05

1.0 0

0.05

0.10

0.15

0.20

0.25

D/d

A

b

6.00 3.00 2.00 1.30 1.20 1.10 1.07 1.05 1.03 1.02 1.01

0.895 79 0.907 20 0.932 32 0.958 80 0.995 90 1.016 50 1.019 90 1.022 60 1.016 60 0.995 28 0.966 89

–0.358 47 –0.333 33 –0.303 04 –0.272 69 –0.238 29 –0.215 48 –0.203 33 –0.191 56 –0.178 02 –0.170 13 –0.154 17

0.30

r/d FIGURE C-10

C

Geometric Stress-Concentration Factor Kt for a Filleted Flat Bar in Bending

Copyright © 2018 Robert L. Norton: All Rights Reserved

Appendix C

STRESS-CONCENTRATION FACTORS

D / d = 2.0 3.0 2.8

1.30

2.6

1.20

2.4

1.10

Kt 2.2

1.07

b r Kt ≅ A   d where:

h

D

1.50 P

r ∞

1.05

2.0

P

d

1.15

1.03

1.8

1.02

1.6 1.01

1.4 1.2 0.02

0.10

0.05

0.15

0.20

1053

0.25

0.30

r/d FIGURE C-11

D/d

A

b

∞ 3.00 2.00 1.50 1.30 1.20 1.15 1.10 1.07 1.05 1.03 1.02 1.01

1.109 50 1.113 90 1.133 90 1.132 60 1.158 60 1.147 50 1.095 20 1.085 10 1.091 20 1.090 60 1.051 80 1.054 00 1.042 60

–0.417 12 –0.409 23 –0.385 86 –0.365 92 –0.332 60 –0.315 07 –0.325 17 –0.299 97 –0.268 57 –0.241 63 –0.222 16 –0.188 79 –0.141 45

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Notched Flat Bar in Axial Tension

b r Kt ≅ A   d where:

3.0 2.8 2.6

1.07

D / d = 1.10

1.05

1.15

h

D

1.20

2.4

1.03

2.2

1.02

d

1.30 M

Kt 2.0

1.50 1.01

1.8

M

r 2.0



1.6 1.4 1.2 1.0 0

0.05

0.10

0.15

0.20

r/d FIGURE C-12 Geometric Stress-Concentration Factor Kt for a Notched Flat Bar in Bending

0.25

0.30

D/d

A

b

∞ 3.00 2.00 1.50 1.30 1.20 1.15 1.10 1.07 1.05 1.03 1.02 1.01

0.970 79 0.971 94 0.968 01 0.983 15 0.982 88 0.990 55 0.993 04 1.007 10 1.014 70 1.025 00 1.029 40 1.037 40 1.060 50

–0.356 72 –0.350 47 –0.349 15 –0.333 95 –0.326 06 –0.313 19 –0.302 63 –0.283 79 –0.261 45 –0.240 08 –0.211 61 –0.184 28 –0.133 69

Copyright © 2018 Robert L. Norton: All Rights Reserved

B C

1054

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

3.0 2.9

d

W P

2.8

for

P

h

K t ≅ 3.0039 − 3.753

2.7

d W

2  d +7.9735   W

2.6 Kt

d ≤ 0.65 : W

3  d −9.2659   W

2.5 2.4

4  d +1.8145   W

2.3

5  d +2.9684   W

2.2 2.1 0

0.10

0.20

0.30

0.40

0.60

0.50

0.70

d/W FIGURE C-13

Copyright © 2018 Robert L. Norton: All Rights Reserved

Geometric Stress-Concentration Factor Kt for a Flat Bar with Transverse Hole in Axial Tension

3.0

d ≤ 0.65 : W d K t ≅ 2.9947 − 3.4833 W

for

2.8

d

W

2.6

2.2

h

M

0

2.4

M

0.50

1.8

1.0 1.5 2.0

1.6 1.4

d/h

1.2 1.0 0.10

0.20

0.30

2 3  d  d +5.8268   − 4.1986   W W

for



0

0 and

 b( d W )  d  ≥ 0.25 : K t ≅ A e  h where:

0.25

Kt 2.0

d h

0.40

0.50

0.60

0.70

d/h

A

b

0.25 0.50 1.00 1.50 2.00 ∞

2.687 50 2.466 20 2.240 00 2.024 30 2.105 60 1.808 20

–0.751 28 –0.772 15 –0.787 39 –0.808 21 –0.798 78 –0.667 02

d/W FIGURE C-14 Geometric Stress-Concentration Factor Kt for a Flat Bar with Transverse Hole in Bending

C

Copyright © 2018 Robert L. Norton: All Rights Reserved

Appendix ANSWERS TO SELECTED PROBLEMS The solutions manual (pdf) and a complete set of Mathcad problem-solution files are downloadable (only by instructors) from http://www.pearsonhighered.com/norton under the Instructor Support option. A family tree of related problems in various chapters is shown in the solutions manual. Lecture slides and other materials are available to instructors who adopt the book as a required text from the author’s website at: http://designofmachinery.com/books/machine-design/professors-using-our-booksmd/. CHAPTER 1

INTRODUCTION TO DESIGN

1-4 1000 lbf, 31.081 slug, 2.59 blob, 453.592 kg, 4448.2 N. 1-5 25.9 lbf. 1-6 220.5 lbf, 220.5 lbm, 6.85 slug, 0.571 blob, 980.7 N.

CHAPTER 2

MATERIALS AND PROCESSES

2-6 E = 207 GPa, U = 2.7 N-m, steel. 2-8 E = 207 GPa, U = 1.3 N-m, magnesium. 2-9 E = 16.7 Mpsi, Uel = 300 psi, titanium. 2-12 UT = 82.7 MPa, UR = 0.41 MPa. 2-14 Sut = 170 kpsi, 359HV, 36.5HRC. 2-16 Iron and carbon, 0.95% carbon, can be through hardened or surface hardened without carburization. 2-27 Sy = 88.1 kpsi, Sy = 607 MPa. 2-34 The most commonly used metal is zinc. The process is called “galvanizing” and it is accomplished by electroplating or hot dipping.

1055

D

1056

MACHINE DESIGN

CHAPTER 3

-

An Integrated Approach

-

Sixth Edition

KINEMATICS AND LOAD DETERMINATION

3-3 T = 255 N-m on sprocket, T = 90 N-m on arm, M = 255 N-m on arm. 3-6 55 114 N. 3-7 12 258 N. 3-8 ωn = 31.6 rad/sec, ωd = 30.1 rad/sec. 3-10 R1 = V = –1821 N @ 0 to 0.7 m, R1 = 2802 N, M = –1275 N-m @ 0.7 m. 3-11 3056-N dynamic force and 408-mm deflection. V = –5677 N @ 0 to 0.7 m, M = –3973 N-m @ 0.7 m, R1 = 5676 N, R2 = 8733 N. 3-15 µ = 0.025. 3-18 Tipover begins at 18.7 mph, load slides at 14.8 to 18.3 mph. 3-22 (a) 897 N, (b) 3592 N. 3-23 (a) R1 = 264 N, R2 = 316 N, V = -316 N, M = 126 N-m. 3-24 (a) R1 = 620 N, M1 = 584 N-m, V = 620 N, M = –584 N-m. 3-25 (a) R1 = –353 N, R2 = 973 N, V = 580 N, M = –216 N-m. 3-27 (a) R1 = 53 895 N, V = 53 895 N, M = –87 040 N-m. 3-34 Row (a) Vmax = 1000 lb at x = 16 to 18 in, Mmax = 2000 lb at x = 16 in. 3-48 ωn = 10.1 Hz.

CHAPTER 4

STRESS, STRAIN, AND DEFLECTION

4-1 σ1 = 1207 psi, σ2 = 0, σ3 = –207 psi, τmax = 707 psi. 4-4 (a) σ1 = 114 MPa, σ2 = 0, σ3 = 0 MPa, τmax = 57 MPa. (b) 9.93 MPa. (c) 4.41 MPa. (d) σ = 53.6 MPa, τ = 1.73 MPa. (e) σ1 = 72.8 MPa, σ2 = 0, σ3 = 0, τmax = 36.4 MPa. 4-6 (a) σ1 = 1277.8 MPa, σ2 = 0, σ3 = 0, τmax = 639 MPa. (b) 111.6 MPa. (c) 49.6 MPa. (d) σ = 540 MPa, τ = 1.7 MPa. (e) σ1 = 636 MPa, σ2 = 0, σ3 = 0, τmax = 318 MPa. 4-7 OD = 0.375 in, ID = 0.230 in. 4-8 199 mm. 4-10 24.5-MPa principal stress, –128-mm deflection. 4-11 76-MPa stress, –400-mm deflection.

D

4-15 4.5-mm-dia pin.

Appendix D ANSWERS TO SELECTED PROBLEMS

1057

4-18 13 254-lb force per rod, 132 536-lb force total, 0.36-in deflection. 4-19 2.125-in-dia pin, 2.375-in outside radius. 4-22 (a) 5.72 MPa.(b) 22.87 MPa. 4-23 Row (a) R1 = 264 N, R2 = 316 N, V = –316 N over b ≤ x ≤ l, M = 126 N-m @ x = b, θ = 0.33 deg,, y = –1.82 mm, σmax = 88.7 MPa. 4-24 Row (a) R1 = 620 N, M1 = 584 N-m, V = 620 N @ x = 0, M = –584 N-m @ x = 0, θ = –2.73 deg,, y = – 32.2 mm, σmax = 410 MPa. 4-25 Row (a) R1 = –353 N, R2 = 973 N, V = 578 N @ x = b, M = –216 N-m @ x = b, θ = –0.82 deg, y = –4.81 mm, σmax = 152 MPa. 4-26 Row (a) R1 = 112 N, R2 = 559 N, R3 = –52 N,V = –428 N @ x = b, M = 45 N-m @ x = a, θ = 0.06 deg, y = –0.02 mm, σmax = 31.5 MPa. 4-29 Row (a) 307.2 N/mm. 4-30 Row (a) 17.7 N/mm. 4-31 Row (a) 110.6 N/mm. 4-32 Row (a) 2844 N/mm. 4-33 Row (a) σ1 = 21.5 MPa @ A, σ1 = 16.1 MPa @ B. 4-34 Row (a) y = –1.62 mm. 4-35 Row (a) k = 31 N/mm. 4-37 e = 0.84 mm, σi = 410 MPa, σo = –273 MPa. 4-41 (a) 38.8 MPa, (b) 11.7 MPa. 4-49 Row (a) Johnson—part: (a) 1.73 kN, (b) 1.86 kN, (c) 1.94 kN, (d) Euler 676 N. 4-50 Row (a) Euler—part: (a) 5.42 kN, (b) 8.47 kN, (c) 12.8 kN, (d) 1.23 kN. 4-51 Row (a) Johnson—part: (a) 57.4 kN, (b) 58.3 kN, (c) 58.9 kN, (d) 48.3 kN. 4-52 Row (a) part: (a) 18.6 kN, (b) 18.7 kN, (c) 18.8 kN, (d) 17.9 kN. 4-69 σi = 132 MPa, σo = –204 MPa. 4-75 Row (a) σnom = 40 Mpa, Kt = 1.838, σmax = 73.5 Mpa.

CHAPTER 5

STATIC FAILURE THEORIES

5-1 Row (a) σ1 = 1207 psi, σ2 = 0 psi, σ3 = –207 psi, τ13 = 707 psi, σ’ = 1323 psi. Row (h) σ1 = 1140 psi, σ2 = 250 psi, σ3 = 110 psi, τ13 = 515 psi, σ’ = 968 psi. 5-4 (a) N = 2.6, (b) N = 30.2, (c) N = 39.3, (d) N = 5.6, (e) N = 4.1. 5-6 (a) N = 0.23, (b) N = 2.7, (c) N = 3.5, (d) N = 0.56, (e) N = 0.47. 5-7 for N = 3.5, OD = 0.375 in, ID = 0.281 in. 5-8 ID = 198 mm. 5-10 N = 5.3. 5-11 N = 1.7. 5-15 N = 1.0 by definition if stress = strength.

B D

1058

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

5-17 N = 3.5. 5-19 2.250-in-dia pin and 2.250-in outside radius. 5-22 (a) N = 40.4 (b) N = 10.1. 5-23 Row (a)—part: (a) N = 3.4, (b) N = 1.7. 5-24 Row (a)—part: (a) N = 0.73, (b) N = 0.37. 5-25 Row (a)—part: (a) N = 2, (b) N = 1. 5-26 Row (a)—part: (a) N = 9.5, (b) N = 4.8. 5-27 (a) a = 166 mm, b = 94 mm, N = 1.5, (b) a = 208 mm, b = 70 mm, N = 1.5. 5-32 Modified-Mohr, N = 1.6. 5-33 Row (a) σ’ = 30.2 MPa at point A, σ’ = 27.9 MPa at point B. 5-34 Row (a) Distortion-energy theory: N = 13.2 at point A, N = 14.3 at point B, Max shear theory: N = 11.6 at point A, N = 12.4 at point B, Max normal stress theory: N = 18.6 at point A, N = 24.8 at point B. 5-35 Row (a) Coulomb-Mohr theory: N = 13.4 at point A, N = 16.1 at point B, ModifiedMohr theory: N = 16.3 at point A, N = 21.7 at point B. 5-37 (a) N = 1.7 at inner fiber, N = 2.6 at outer fiber; (b) N = 1.0 at inner fiber, N = 4.4 at outer fiber. 5-38 Ν = 1.5. 5-39 Crack half-width = 0.216 in. 5-41 (a) Ν = 9.4, (b) Ν = 24.5. 5-65 (a) Na = 1.8, (b) Nb = 2.6. 5-68 d = 1.500 in.

CHAPTER 6

FATIGUE FAILURE THEORIES

6-1 Row (a) ∆σ = 1000, σa = 500, σm = 500, R = 0, A = 1.0. Row (c) ∆σ = 1000, σa = 500, σm = 1000, R = 0.33, A = 0.50. Row (e) ∆σ = 1500, σa = 750, σm = –250, R = –2.0, A = –3.0. 6-3 Nf = 0.31. 6-6 (a) 0.14, (b) 1.17, (c) 1.6, (d) 0.24, (e) 0.25. 6-7 for Nf = 1.5, OD = 0.375 in, ID = 0.299 in. Round to ID = 0.281 in for Nf = 1.8. 6-8 ID = 190 mm assuming machined, 99.9% reliability, and room temperature. 6-10 Nf = 2.4. 6-11 Nf = 0.79. 6-15 Row (a) a = 0.062 in0.5, q = 0.89, Kf = 3.05. 6-17 Nf = 2.7 assuming forged, 99.99% reliability, and room temperature.

D

6-19 2.750-in-dia pin and 2.625-in outside radius (machined, 90% reliability, and 100 °F). 6-22 (a) Nf = 21.3, (b) Nf = 5.3 assuming machined, 99.999% reliability, and 37 °C.

Appendix D ANSWERS TO SELECTED PROBLEMS

1059

6-23 Row (a) Use a material with Sut = 468 MPa (assuming Ctemp = Csurf = Creliab = 1). 6-24 Row (a) Use a material with Sut = 676 MPa (assuming Ctemp = Csurf = Creliab = 1). 6-25 Row (a) Use a material with Sut = 550 MPa (assuming Ctemp = Csurf = Creliab = 1). 6-26 Row (a) Use a material with Sut = 447 MPa (Ctemp = Creliab = 1, Csurf = 0.895). 6-27 (a) a = 190 mm, b = 100 mm, N = 2.1, (b) a = 252 mm, b = 100 mm, N = 2 (both assuming machined, 90% reliability, and 40 °C). 6-29 Nf = 2.6 assuming machined, 99.999% reliability, and 37 °C. 6-31 Nf = 1.8 assuming machined, 99.999% reliability, and 37 °C. 6-33 Row (a) Use a material with Sut = 362 MPa (machined, 50% reliability, and 37 °C). 6-34 Row (a) Use a material with Sut = 291for Nf = 1.5 MPa (machined, 50% reliability, and 37 °C). 6-37 (a) Nf = 1.8, (b) Nf = 0.92 assuming machined, 90% reliability, and 37 °C. 6-39 Nf = 1.9 using SEQA method and assuming ground shaft, 50% reliability, and 37 °C. 6-41 (a) Nf = 3.3, (b) Nf = 8.6 assuming machined, 99.999% reliability, and 37 °C. 6-47 Nf = 1.5 assuming machined, 90% reliability, and 60 °C. 6-52 tmin = 3.2 mm. 6-64 Row (a) σm = 0.0 MPa, σa = 251.9 MPa.

CHAPTER 7

SURFACE FAILURE

7-1 Ar = 0.333 mm2. 7-2 µ = 0.4. 7-3 N = 4.6E6. 7-4 σ1 = –61 kpsi, σ2 = –61 kpsi, σ3 = –78 kpsi. 7-8 64.4-mm total width. 7-10 0.15-mm total width. 7-13 (a) 19.6 min dry, (b) 9.8 min wet. 7-16 1-mm dia contact patch, σzball = –1900 MPa, σzplate = –1900 MPa. 7-18 0.166-mm total contact width, σzcylinder = σxcylinder = –123 MPa, σzplate = σxplate = –123 MPa. 7-20 Contact-patch half-dimensions: 0.933 x 0.713 mm, σ1 = –5.39 GPa, σ2 = –5.81 GPa, σ3 = –7.18 GPa. 7-22 (a) σ1 = –66.9 MPa, σ2 = –75.2 MPa, σ3 = –79.0 MPa, (b) σ1 = –106 MPa, σ2 = –119 MPa, σ3 = –125 MPa. 7-23 σ1 = –24 503 psi, σ2 = –30 043 psi, σ3 = –57 470 psi. 7-39 t = 4.7 min. 7-42 The principal stresses are maximum at the surface. They are: σ1 = –276.7 MPa, σ2 = –393.3 MPa, σ3 = –649.0 MPa. The maximum shear stress is τ13 = 186.1 MPa.

B D

1060

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

CHAPTER 10 SHAFTS, KEYS, AND COUPLINGS 10-1 Row (a) d = 1.188 in, assuming machined, 99% reliability, and 100°F. 10-2 Row (a) d = 48.6 mm, assuming machined, 99% reliability, and 30°C. 10-4 Row (a) y = 0.0036 in, θ = 0.216 deg. 10-5 Row (a) y = – 5.7 µm, θ = 1.267 deg. 10-6 Row (a) 3/8-in square key, 0.500 in long, Nf = 2.1, Nbearing = 2.1. 10-8 Shaft ID = 191 mm, assuming machined, 99.9% reliability, and 30°C. 10-9 Row (a) d = 1.188 in, assuming a notch radius of 0.015 in, machined, 99% reliability, and 100°F. 10-11 Row (a) 0.0007 to 0.0021 in of interference over tolerance range. 10-13 Row (a) 2102 rad/sec or 20 075 rpm, or 334.5 Hz. 10-15 Row (a) min = 0, avg = 11.9 hp, max = 23.8 hp. 10-16 Row (a) min = 0, avg = 5.2 kW, max = 10.5 kW. 10-17 Row (a) N = 0.61 at key on right end of roller, θ = 0.20 deg, fn = 1928 Hz. 10-18 Row (a) y = –30.0 µm to 22.9 µm. 10-19 Row (a) d = 1.337 in, NA = 2.0, NB = 3.1. 10-37 δmin = 0.06 mm, δmax = 0.12 mm. 10-38 ri = 1.00 in, ro = 14.66 in, t = 0.800 in.

CHAPTER 11 BEARINGS AND LUBRICATION 11-1 Row (a)—part (a) d = 1.188 in, l = 1.485 in, Cd = 1.8E–3 in, RL = 125 lb, RR = 1125 lb, ηL = 0.204 µreyn, ηR = 1.84 µreyn, pavgL = 71 psi, pavgR = 638 psi, TrL = 0.15 lb-in, TrR = 1.38 lb-in, ΦL = 0.004 hp, ΦR = 0.033 hp. Part (b) #6300 bearing at left end gives 1.4E9 cycles L10 life on left bearing and #6306 bearing at left end gives 8.8E7 cycles L10 life on right bearing. 11-3 267 cP. 11-5 0.355 in-lb. 11-6 10.125 µm. 11-7 Tr = 3.74 N-m, T0 = 2.17 N-m, Ts = 2.59 N-m, Φ = 979 W. 11-8 d = 220 mm, l = 165 mm, Cd = 0.44 mm, RL =  RR = 26.95 kN, η = 181 cP, pavg = 743 kPa, Tr = 12.9 N-m, Φ = 67.7 W. 11-10 hmin = 4.94 µm. 11-14 η = 13 cP, Ts = 519 N-mm, T0 = 325 N-mm, Tr = 699 N-mm, Φ = 183 W, P = 19.222 kN. 11-17 Row (a)—part (a) d = 40 mm, l = 30mm, Cd = 0.04 mm, RL = 6275 N, RR = 7 525 N, ηL = 20.7 cP, ηR = 24.8 cP, pavgL = 5229 kPa, pavgR = 6271 kPa, TrL = 468 N-m, TrR = 561 N-m, ΦL = 88.2 W, ΦR = 106 W.

D

Part (b) #6308 bearing at left end gives 1.41E8 cycles L10 life on left bearing and #6309 bearing at left end gives 1.58E8 cycles L10 life on right bearing.

Appendix D ANSWERS TO SELECTED PROBLEMS

1061

11-20 Specific film thickness = 0.53—boundary lubrication. 11-33 Row (a) left, #6300; right, #6314. 11-36 Row (a) left, #6300; right, #6320.

CHAPTER 12 SPUR GEARS 12-1 dp = 5.4 in, addendum = 0.2 in, dedendum = 0.25 in, OD = 5.8 in, pc = 0.628 in. 12-3 1.491. 12-5 30.33 deg. 12-7 7.159 : 1. 12-9 96:14 and 96:14 compounded give 47.02:1. 12-11 Nring = 75 t, ratio between arm and sun gear = 1 : 3.273. 12-14 7878 in-lb on pinion shaft, 19 524 in-lb on gear shaft. 12-16 pd = 3 and F = 4.25 in gives Npinion = 5.4 and Ngear = 2.0. 12-18 pd = 4, F = 4.125 in, gives Npinion = 3.5 and Ngear = 2.0. 12-20 202 N-m (1786 in-lb) on sun shaft, 660 N-m (5846 in-lb) on arm shaft. 12-23 pd = 3, F = 3.500 in, gives Npinion = 7.7 and Ngear = 2.8. 12-25 pd = 4, F = 4.000 in, gives Npinion = 4.8 and Ngear = 1.8. 12-27 T1 = 1008 N-m, T2 = 9184 N-m, T3 = 73 471 N-m, T4 = 661 236 N-m. 12-28 pd = 3, F = 4.500 in. 12-29

pd = 1.5, F = 9.375 in.

12-30

pd = 0.75, F = 17 in.

12-31 104:12 and 144:16 compounded give exactly 78:1. 12-52 T1 = 40.9 N-m, T2 = 295 N-m, T3 = 2172 N-m. 12-53 F = 1.250 in, pd = 8.

CHAPTER 13 HELICAL, BEVEL, AND WORM GEARS 13-1 dp = 5.4 in, addendum = 0.200 in, dedendum = 0.250 in, OD = 5.8 in, pt = 0.628 in, pn = 0.544 in, pa = 1.088 in. 13-3 mp = 1.491, mF = 0.561. 13-5 αg = 83.66°, αp = 6.34°, dg = 21 in, dp = 2.33 in, Wag = Wrp = 25 lb, Wrg = Wap = 2.8 lb. 13-7 αg = 78.69°, αp = 11.31°, dg = 11.429 in, dp = 2.286 in, Wag = 60.5 lb, Wap = –169.64 lb, Wrp = 136.28 lb, Wrg = 209.01 lb. 13-9 l = 20 mm, λ = 7.26°, λ per tooth = 3.63°, dg = 140 mm, c = 95 mm, self-locking. 13-11 l = 5 mm, λ = 2.28°, λ per tooth = 2.28°, dg = 130.5 mm, c = 85.3 mm, self-locking. 13-12 Tw = 22.4 N-m, Tg = 492 N-m rated, Wt = 7 028 N, friction = 215 N, output power available = 2.34 kW, rated input power = 2.91 kW.

B D

1062

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

13-14 7878 in-lb on pinion, 19 524 in-lb on gear. 13-16 pd = 3 and F = 3.25 in gives Npinion = 5.6 and Ngear = 2.0. 13-18 pd = 4 and F = 2.75 in gives Npinion = 2.8 and Ngear = 1.6. 13-20 pd = 18 and F = 1 in gives Npinion = 2.0—bending, with Ngear = 13.6—bending. 13-23 pd = 16 and F = 1 in gives Npinion = 1.4—surface failure, which limits the design. 13-27 Rated input power = 2.11 hp, output power available = 1.69 hp, rated output torque = 1290 lb-in. 13-49 F = 1.375 in, pd = 10.

CHAPTER 14 SPRING DESIGN 14-1 k = 1.6 N/mm. 14-3 Sys = 110 931 psi, Sus = 148 648 psi. 14-4 Sfs’ = 85.6 kpsi. 14-6 C = 10, k = 1.01 N/mm. 14-7 fn = 363.4 Hz. 14-10 k = 7614 N/m, fn = 1.39 Hz. 14-11 d = 0.125 in, D = 0.94 in, Lf = 3.16 in, k = 36 lb/in, 13.75 coils RH, music wire, squared and ground ends, unpeened, set. 14-13 Na = 19.75, D = 1.37 in, Lf = 7.84 in, Lshut = 6.79 in, k = 266.6 lb/in, yinitial = 0.19 in, hole = 1.75 in. 14-17 d = 3.5 mm, D = 28 mm, Lf = 93.63 mm, k = 5876 N/m, 12.75 coils, music wire, std. hooks. Ny = 2.2 torsion fatigue in hook, Nf = 1.7 bending fatigue in hook, Nsurge = 5.5. 14-19 d = 6.5 mm, Do = 65 mm, Lf = 171 mm, Ntot = 10.75 coils, Ny = 1.6 shut, Nf = 1.3, Nsurge = 9.3. 14-21 d = 5 mm, D = 40 mm, Lf = 116.75 mm, k = 8 967 N/m, 13 coils RH, music wire, S&G ends, unpeened, set. 14-22 d = 16 mm, D = 176 mm, k = 1600 N-m/rev, 4.5 coils RH, 40-mm straight ends, A229 oil-tempered wire, unpeened, relieved. 14-24 d = 15 mm, D = 124.5 mm, k = 248 N-m/rev, 31 coils RH, 40-mm straight ends, A229 oil-tempered, unpeened, relieved. 14-26 do = 39.55 mm, di = 19.77 mm, t = 0.76 mm, h = 1.075 mm, h / t = 1.414, 1-mm working deflection, Sut = 1700 MPa, Ns = 1.11. 14-42 Do = 3.000 in, Di = 1.500 in, t = 0.125 in, h = 0.050 in. 14-44 A228 wire, d = 0.125 in, Do = 1.000 in, Nt = 15, Lf = 3.600 in.

CHAPTER 15 SCREWS AND FASTENERS

D

15-2 Lifting torque = 42.68 lb-in, lowering torque = 18.25 lb-in, lifting efficiency = 27.95%, lowering efficiency = 65.36%, screw is self-locking.

Appendix D ANSWERS TO SELECTED PROBLEMS

1063

15-4 Two M12 x 1.75 bolts, ISO class 8.8, Fpreload = 59% proof strength, Ny = 1.7. Nsep = 2.5. 15-6 Two M24 x 3 bolts, ISO class 12.9, Fpreload = 55% proof strength, Ny = 1.7, Nsep = 1.6. 15-7 Ny = 1.4, Nsep = 13.7. 15-9 Nf = 1.3, Ny = 1.5, Nsep = 8.9. 15-11 1252 in-lb. 15-13 718 in-lb. 15-17 (a) Keff = 5.04E9 N-m, aluminum dominates. (b) Keff = 9.52E9 N-m, steel dominates. (c) Keff = 2.73E8 N-m, rubber dominates. (d) Keff = 2.66E8 N-m, rubber dominates. (e) Keff = 9.04E9 N-m, no one material dominates. 15-18 (a) Keff = 2.74E7 in-lb, aluminum dominates. (b) Keff = 5.18E7 in-lb, steel dominates. (c) Keff = 3.73E5 in-lb, rubber dominates. (d) Keff = 3.70E5 in-lb, rubber dominates. (e) Keff = 4.92E7 in-lb, no one material dominates. 15-20 Use 10 M12 x 1.75, ISO class 8.8 cap screws, torqued to 90% of proof strength on a 107.5-mm-dia bolt circle. Nf  = 1.3, Nsep = 34, Ny = 1.2 dynamic and 1.2 static. 15-23 Row (a)—Four M5 x 0.8 x 20-mm-long cap screws, class 4.6, Fpreload =1.72 kN, (54% of proof), load on top bolt: 1.73 kN, Nsep = 58, Ny = 2.0. 15-24 Row (a)—Four M4 x 0.7 x 20-mm-long cap screws, class 4.8, Fpreload = 2.04 kN, (75% of proof), load on top bolts when force is maximum and on bottom bolts when force is minimum: 2.05 kN, load on top bolts when force is minimum and on bottom bolts when force is maximum: 2.05 kN, Ny  = 1.5, Nsep = 69, Nf  = 10. 15-25 Row (a)—Four M4 x 0.7 x 20-mm-long cap screws, class 4.8, Fpreload = 2.04 kN, (75% of proof), load on top bolts when force is maximum and on bottom bolts when force is minimum: 2.05 kN, load on top bolts when force is minimum and on bottom bolts when force is maximum: 2.05 kN, Ny  = 1.5, Nsep = 69, Nf  = 10. 15-39 d = 8 mm: number of threads = 4.6. 15-41 Class 4.6: Fut = 98 kN.

CHAPTER 16 WELDMENTS 16-1 A CJP butt weld in tension develops the full strength of the section: Pmax = 180 000 lb. 16-3 Bracket (weld) length = 3.592 in. 16-5 Maximum dynamic load = 525 lb. 16-6a Weld size required = 3/16 in. 16-7a Weld size required = 1/4 in. 16-9 Weld size required = 10 mm.

B D

1064

MACHINE DESIGN

-

An Integrated Approach

-

Sixth Edition

CHAPTER 17 CLUTCHES AND BRAKES 17-1 T = 380 N-m, pmax = 1.819 MPa, molded or sintered metal lining will work. 17-3 do = 140 mm, di = 80 mm, Φ = 7.85 kW. 17-5 N = 7, do = 104 mm, di = 60 mm, Φ = 12.6 kW. 17-7 (a) T = 10.7 N-m, Fa = 798 N, (b) Will self-lock when c = 233 mm. 17-11 (a) T = 30.5 N-m (15.7 top shoe, 14.8 bottom shoe), Fa = 1353 N, (b) Will self-lock when c = 320 mm. 17-13 T = 26 N-m, Fa = 1689 N. 17-15 T = 56.5 N-m (32.5 top shoe, 24 bottom shoe), Fa = 2194 N. 17-17 (a) (b) (c) 17-18 (a) (b) (c)

short shoe:

T = 11.3 N-m,

Fa = 806 N,

long shoe:

T = 11.2 N-m,

Fa = 750 N.

short shoe:

T = 15.1 N-m,

Fa = 1075 N,

long shoe:

T = 14.8 N-m,

Fa = 982 N.

short shoe:

T = 18.8 N-m,

Fa = 1344 N,

long shoe:

T = 18.3 N-m,

Fa = 1197 N.

short shoe:

T = 21.8 N-m,

Fa = 806 N,

long shoe:

T = 19.6 N-m,

Fa = 750 N.

short shoe:

T = 29.1 N-m,

Fa = 1075 N,

long shoe:

T = 25.8 N-m,

Fa = 982 N.

short shoe:

T = 36.3 N-m,

Fa = 1344 N,

long shoe:

T = 31.9 N-m,

Fa = 1197 N.

17-21 Top pivot: Bottom pivot:

Rx = –392.7 N, Rx = –368.9 N,

Ry = –218.2 N, Ry = –123.0 N.

17-23 Rx = 1005 N, Ry = –808 N. 17-25 Top pivot: Bottom pivot:

Rx = 1694 N, Rx = 325 N,

Ry = –45.3 N, Ry = –147.7 N.

17-29 T = 131 N-m. 17-31 Sintered metal, m = 0.30, ri = 40 mm, ro = 70 mm, θ = 90 deg, F = 2.83 kN.

D

INDEX A abrasion 468. See also wear controlled 470 grinding

470

three-body 468 two-body 468 uncontrolled 468 abrasive particles 470

arm (epicyclic) 743 asperities 460, 673, 689 ASTM wire alloy numbers 834 autofrettage 386 automeshing 527 axial tension 186, 220 axis of transmission gear teeth 729 axisymmetric 526

hardness 471 sharpness 471

wear 503 absolute hardness 43. See also hardness absolute units system 21 accuracy 15 Acme thread 910–964, 913 stub 910 addendum 729, 733 circle 733 modification coefficients 737 adhesion. See wear: adhesive adhesive wear in gear teeth 752 AGMA 725, 735, 737, 752, 791 backup ratio 759 quality index 754 air compressor 564 air cylinder 7 Almen number 385 aluminum 56 aircraft 58 alloys 56 hardenable

57

cast 58 table of properties 1037

wrought 57 aluminum oxide 471 thickness of 472 analysis closed-form 521 definition 7 first-order 9 angle of approach 730 of recess 730 angular velocity ratio 87, 727 definition 86 anisotropic 42 annealing 45 anode 50 anodizing 51, 471, 472 hard-coating 51 answers to selected problems 1055 aquaplaning 671, 672 arc of action 730 area moment of inertia 227

B backdrive 731, 813, 914 backlash 733 ball bearings 696 ball screws 916 base circle of gear 731, 733, 736 of involute 728 base units 20 beam 37, 139, 188 assumptions 189 cantilever 139, 219 deflection

219

centroidal axis curved 191

191

stress distribution in 191

deflection 196 deflection function of dummy load 209 hollow

197

shear stress in 195

I-beams 195 indeterminate 139, 207 loading 139 long 194 neutral axis 191 neutral plane 190 overhung 139 pure bending 188 rectangular 194 shear stress in 194

round

195

shear stress in 195

section modulus 190 shear, transverse 192 sign convention 140 simply supported 139 spring rate 219 statically indeterminate 204 straight 189 stress distribution in 190

tables 1043 bearing 187, 589, 665 air 712 area 187

1065

ball

695

angular-contact 697 Conrad 697 deep-groove 697 thrust 697

cam follower 708 cleanliness 509 flange units 708 journal 671, 677 clearance ratio 685 coefficient of friction in 682 eccentricity 676 eccentricity ratio 677, 683 lubrication in 671 power lost in 682 torque in 681

linear 708 long journal solution 678 Sommerfeld equation 678

materials

673

babbitt 673 bronze 674 gray cast iron 674 nonmetallic 674

needle 697 pillow blocks 708 plain 665 rod ends 708 roller 482, 695, 697 tapered

697

rolling-element 474, 592, 597, 665, 695 advantages 696 basic dynamic load rating 701 basic static load rating 703 calculation procedure 705 disadvantages 696 endurance limit 501, 695 equivalent load 704 failure in 700 L10 life 700 linear motion 708 manufacturing 695 materials for 695 mounting of 707 rating life 700 selection of 701 tolerance classes 695

self-aligning 697 short journal 667 load factor 683 solution 679

sleeve 593, 597, 665, 674 thrust 597, 665 cylindrical-roller 697 hydrostatic 671 bearings

1066

MACHINE DESIGN

Belleville washers 881. See also springs: Belleville; See also washers: Belleville belt drive 741 bending moment 190 of shaft 589 bimetallic strips 466 blobs 21, 22 Boeing Aircraft Co. 363 bolts 903, 921 preloaded dynamic loading 933 static loading 929

stiffness torsional stress due to torquing 950 boron carbide 471 boundary conditions 522, 532 boundary element analysis 12 boundary lubrication 673. See also lubrication; See also lubrication: boundary brake 743, 1007 band 1020 disc 1011 disk 1019 automobile 1020 caliper 1020

1011, 1020

external 1021 internal shoe 1027 long-shoe 1021, 1023 self-deenergizing 1022 self-energizing 1022 self-locking 1022 short-shoe 1021

eddy current 1013 friction 1010 magnetic hysteresis 1012 magnetic particle 1012 torque 1025 Bridgman, W.P. 282 Brinell test 43 British Comet. See fatigue failure: of British Comet brittleness 18, 34 bronze 674 Buckingham equation 762 bushing 665 bronze 674 buttress thread 910

C

D

CAD 11, 12 multiview drawing 12 solid model 11 used with FEA 527 wireframe model 11 cam and follower 7, 474 as effective linkage 80 boundary conditions 94 minimum number needed 94

An Integrated Approach

B-Spline functions dwell

93

93 double 93 single 93

eccentricity 100 follower jump 101 functions 3-4-5 polynomial 94 4-5-6-7 polynomial 97 double dwell 93 jerk 97 normalized variable 94 piecewise continuous 93 single dwell symmetric

equation for 940

drum

-

97

single-dwell asymmetrical

98

fundamental law polynomial 94

93

advantages 93 boundary conditions degree 94 double-dwell 94 order 94 single dwell 97

pressure angle maximum

prime circle radius

94

99

99

Sixth Edition

Case Study 09A 568 Case Study 09B 818 Case Study 10A 572 Case Study 10B 710 Case Study 10C 888 Castigliano’s method for deflection 209 redundant reactions 209 Castigliano’s theorem 207 cast iron strengths of, table 1039 cathode 49 ceramics 62 chain drive 741 Charpy impact test 40 circuit of a linkage 83, 84 circular pitch 733 clearance 733 Clerk, James Maxwell 287 closed-form analysis 521 clutch 973, 1007 backstop 1011 centrifugal 1011 cone 1010 disk 1011, 1016 uniform pressure 1016, 1017 uniform wear 1016, 1017

93

99

program Dynacam 101 radius of curvature 100 undercutting 101 versus roller radius

-

101

roller radius 99 svaj diagram 94 timing diagram 93 cam followers 708 cam-testing machine 571 cap screw 921 carborundum 470 carburizing 384 Case Study 01A 106 Case Study 01B 239 Case Study 01C 310 Case Study 01D 549 Case Study 02A 112 Case Study 02B 242 Case Study 02C 313 Case Study 02D 551 Case Study 03A 116 Case Study 03B 247 Case Study 03C 316 Case Study 04A 122 Case Study 04B 251 Case Study 04C 318 Case Study 04D 553 Case Study 05A 126 Case Study 05B 133 Case Study 06 423 Case Study 07 556 Case Study 08A 566 Case Study 08B 648 Case Study 08C 777 Case Study 08D 956

eddy current 1012 friction 1010 dry 1011 electromagnetic 1012 material 1016 wet 1011

location 1015 magnetic hysteresis 1012 magnetic particle 1012 multiple disk 1011 one-way 1011 overrunning 1011 positive contact 1009 roller 1011 service factors 1015 sprag 1011 spring-wrapped 1011 synchromesh 1010 coatings 51 ceramic 471 plasma-sprayed

471

chemical 51 plasma-spray 51 coining 386 cold forming 384, 386 cold working 46, 837 collar, clamp 591 column 226 buckling 226 of springs. See springs: helical compression: buckling of

critical unit load 228 eccentrically-loaded 230, 233 eccentricity ratio 236 end conditions 229 fixed-fixed

229

INDEX fixed-free 229 fixed-pinned 229

Euler formula 228 intermediate 226, 230 Johnson 230 long 227 radius of gyration 227 secant formula 236 short 226, 227 slenderness ratio 227, 230 Comet aircraft. See fatigue failure: of British Comet common normal 729 tangent 729 communication 16 complex numbers 81 composites 42, 62 compound gear train 741 compression test 35 compressive strength of cast irons, table 1039 computer-aided design 11 conjugates 728 constant-life diagram 398 contact patch 693 contact pressure 476 contact ratio 738, 750. See also gearset axial 796 low

796

minimum 738 transverse 796 controlling preload 948 copper 59 alloys 59 pure 59 Cornwell, R. 842, 940 corrosion 472 fatigue 344, 372, 457, 473 fretting 622 wear 457, 472, 508 Coulomb friction. See friction couplings 644 compliant 646 bellows 647 constant velocity (Rzeppa) 647 flexible-disk 646 gear/spline 646 helical 647 jaw 646 linkage (Schmidt) 647 Oldham 647 universal joints 647

fluid 1013 Hooke 647 rigid 645 clamp collars 645 keyed 645 setscrew 645

Rzeppa 647 crack growth 343 in a corrosive environment initiation 342, 343, 347

343

1067

micro 343 propagation 305, 342, 343 creativity 27 creep 40 critical frequency 634 critical speed 634 crossed mechanism 84 cycloidal gear tooth 728

D damping 131 dedendum 733, 736 circle 733 deflection 31 angular 211 cantilever beam 219 of helical torsion spring 875 spring 218 degree of freedom 73 Den Hartog, J. P. 283 derived unit 20 design 3, 7, 563 analysis 8 decisions 8 documentation 9 process 5, 27, 563 sketches 8 diametral pitch 734 diamond 471 differential 743 direct bearing 187. See also bearing distortion energy 283, 284. See also von Mises ellipse 288 comparison to experiments

290

theory 282 DOF kinematic removal of

finite 521 H-elements 527 hexahedral 525 line 525 linear 526 order 526 P-elements 527 quadrilateral 525 rigid-body 539 shell 526 skew of 527 surface 525 taper of 527 tetrahedral 525 three-dimensional 526 triangular 525 truss 525 two-dimensional 526 volume 525 warp of 527 wedge 528 zero-DOF 525 elongation of alloy steels, table 1041 of aluminum, table 1037 of carbon steels, table 1040 of copper alloys, table 1037 of plastics, table 1042 of stainless steel, table 1039 endurance 37 limit 38, 347, 352 corrected 373 correction factors for 366 estimating 364, 766

strength 38, 353 endurance strength. See fatigue: strength energy kinetic in shaft

524

Dolan 301 dowel pins 951, 952 press-fit 952 Dowling, N. E. 297, 341, 400 ductility 18, 33 dynamic force 22, 132

E effective mass, spring, damping 132 Eichinger 287 elastic behavior 32 limit 32 elastohydrodynamic lubrication 672 electroplating 50 chrome 50 element 521 aspect ratio of 527 automeshers 527 boundaries 526 discrete 522

627

method 135 variation in rotating system 626 engineering model 9 engineering report 16 engineering stress-strain curve 33 equation solver 14, 563. See also TKSolver estimation 9 Euler’s equations 105 Euler’s identity 81 even material 35 Example 01-01 22 Example 03-01 95, 98, 138 Example 03-02A 143 Example 03-02B 145 Example 03-03A 147 Example 03-03B 148 Example 03-04 149 Example 04-01 181 Example 04-02 183 Example 04-03 184 Example 04-04 198 Example 04-05 200 Example 04-06 202

B D

1068

D

Example 04-07 Example 04-08 Example 04-09 Example 04-10 Example 04-11 Example 05-01 Example 05-02 Example 05-03 Example 06-01 Example 06-02 Example 06-03 Example 06-04 Example 06-05 Example 06-06 Example 07-01 Example 07-02 Example 07-03 Example 07-04 Example 07-05 Example 08-01 Example 08-02 Example 08-03 Example 08-04 Example 08-05 Example 08-06 Example 09-06 Example 10-01 Example 10-02 Example 10-03 Example 10-04 Example 10-05 Example 10-06 Example 10-07 Example 10-08 Example 11-01 Example 11-02 Example 11-03 Example 11-04 Example 12-01 Example 12-02 Example 12-03 Example 12-04 Example 12-05 Example 12-06 Example 12-07 Example 13-01 Example 13-02 Example 14-01 Example 14-02 Example 14-03 Example 14-04 Example 14-05 Example 14-06 Example 14-07 Example 15-01 Example 15-02 Example 15-03 Example 15-04 Example 15-05 Example 15-06 Example 16-01 Example 16-02 Example 16-03

MACHINE DESIGN 205 209 212 216 232 292 298 308 375 377 382 390 406 419 480 484 488 496 507 529 533 536 540 543 547 628 600 605 608 614 623 627 631 643 685 692 703 706 739 742 745 751 760 763 772 800 808 848 849 852 858 867 878 886 917 931 935 944 950 955 984 992 994

-

Example 16-04 Example 16-05 Example 17-01 Example 17-02 Example 17-03

An Integrated Approach

997 999 1019 1022 1026

face width 733, 758, 776 facewidth factor 758, 776 factor of safety 16, 17. See also safety factor for gearsets 777 guidelines 18 of commercial aircraft 18 with fluctuating stresses 402 fail safe brake 1007 definition 1008 failure mechanisms 281

strength

294

290

287

linear-elastic

301

302

hydrostatic loading 282 maximum normal stress 290 comparison to experiments

291

maximum shear stress 288 comparison to experiments

291

modified Mohr 296 static loading 309 total strain energy 282 von Mises-Hencky 282 false brinelling 476 fasteners centroid of group 953 head forming 925 in shear 950 manufacturing 923 preloaded 925 screw classification by head style 921 classification by intended use 921 classification by thread type 921 types of 920

shear loads on 954 strengths of 925 stress concentration in 934

choice of 348 linear elastic fracture mechanics 347, 348 strain-life 345, 347, 348 stress-life 345, 347, 348

stages of 342 sudden fracture testing 351

281

fatigue 339 fracture mechanics

1037

stress concentration factor 379, 382 fatigue-crack-growth life 361 fatigue failure 339 of British Comet 341 combined mean and alternating stress 358 cost to the economy 341 crack initiation 342, 343 crack propagation 342, 343 estimating criteria for 364 Gerber line 359 Goodman line 359 history of 339 in rolling bearings 700 mechanism of 342 models 345

ductile material static loading

347, 352

aluminum, table of surface 501

molasses-tank 302 of Liberty Ships 301 of rocket-motor case 302 failure theories 281 assumptions 309 brittle material

historical note

344, 349,

high-cycle fatigue 345, 348 low-cycle fatigue 345, 348

catastrophic 301 shear 294 tensile 294

static loading

Sixth Edition

torque-limited 949 fatigue 339, 341 electroplating in 50 endurance limit 352 fracture 344 in a corrosive environment 372, 473 loads 349 regimes 345

F

comparison to experimental data Coulomb-Mohr 295 distortion-energy 282

-

345,

342

actual assemblies 363 axial fatigue test 355, 362 cantilever beam 356 fully-reversed stress 352 rotating-beam test 352 torsional loading 357

unstable crack growth 342 fatigue strength 37, 353, 374. See also endurance strength corrected 373 effects of loading type on 366 effects of reliability on 371 effects of size on 367 effects of surface finish on 367 effects of temperature on 371 estimating 364, 766 correction factors for

366

fatty acids 464, 668 FEA applying loads 542 automeshers 527 boundary conditions 532 buckling in 525 contact constraints 536 direct stiffness method 523 dynamic analysis 546

INDEX dynamic stresses 547 eigenvalues 547 eigenvectors 547 importing part geometry 527 loading models 542 mass units in 542 mathematical formulations 523 model verification 543 structural analysis 522 used for 523 finite element analysis 12, 130, 565, 635 degree of freedom 525 mesh 522 method theory of

523

in threads

Coulomb

913

288, 462

effect of velocity on

521

node 522 types 525 finite-life 347 fit expansion 620 interference 620 stresses in

620

press 620 shrink 620 flame hardening 384 fluctuation coefficient of 628, 629 fluid couplings 1013 flywheel 625 design 625 failure criteria 631 inertia 628 physical 629 stresses in 630 foot-pound-second (fps) system 21 force normal 462 plough 462 force-flow analogy 225 force ratio 856 forcing frequency 634 fourbar crank-slider offset 89 fracture 34, 35, 280, 344 fracture mechanics 40, 301, 360 fracture toughness 40, 305, 306 free-body diagrams 73, 103 free vibration 633 frequency critical 634, 638 forcing 634 fretting 473, 509 fretting corrosion 457, 473, 622 friction 462 coefficient of 462 in boundary lubrication 673 in clutches/brakes 1011 in hydrodynamic bearing 672, 682 in roll-slide contact 492

463

materials 1016 rolling 463 functions singularity 198. See also singularity functions unit doublet 142 unit impulse 142 unit parabolic 141 unit ramp 142 unit step 142

fundamental law of gearing 727 See also gearing: fundamental law of fusion 975

model 521 models examples of

1069

G gage length 31 galling 467 galvanic action 49 cell 49 coatings 50 series 49 gasketed joints 943 Gaussian distribution gear 727, 750 antibacklash 733 base pitch 738 bevel

41

back cone 805 crowning factor 807 design pinion torque 807 forces on 806 geometry factors I & J 808 operating pinion torque 807 spiral 804 straight 804 stress in 806, 807

blank 731 face width. See face width (gears) helical 791 crossed 791, 793 double 795 parallel 793 stresses in 796

herringbone 795 idler 741, 750 idler factor 759 manufacturing 746 burnishing 748 finishing 747 forming 746 grinding 748 hobbing 747 lapping and honing 748 machining 747 rack generation 747 shaping 747 shaving 748

materials

765

bronzes 766 cast irons 766

nonmetallics 766 steels 766 strengths 766

mesh geometry 730 pinion 727 quality 749 quality index 749 rack 731 rack cutter 731 ratio 728 rim thickness factor 759 shaper 736, 747 teeth 474, 728 AGMA bending stress equation 753 cycloidal 728 elastic coefficient 762 fatigue fracture 752 friction forces 753 full-depth 735 geometry factor J 753, 796, 806 interference 754 involute 728, 732 Lewis equation 752 load sharing in 796 long addendum 805 minimum number 737 minimum number of 795, 815 radius of curvature 762 root fillets 753 standard, full-depth 735 stresses in 752 surface fatigue 752 surface finish factor 763 surface geometry factor I 762, 796 surface stresses in 761, 762 undercutting 795 unequal addendum 737, 738. See also gears: profile-shifted virtual 795

tooth theory 727 train ratio 740 virtual 795 whine 793 worm 811 lead 812 pitch diameter 814 single start 818

Zerol 804 gearbox 775 gearing fundamental law of 727, 730 gears 563, 727 antibacklash 733 bending stress 752 bevel 725, 821 cold drawing 746 form milling 747 helical 474, 725, 821 advantages of

821

hypoid 474 injection molded 746 machining 746 profile-shifted 737, 738 spiral bevel 474

B D

1070 spur

MACHINE DESIGN

776

surface-contact stresses 752 worm 474, 725 wormsets 810, 817, 822

hammer peening 386 hard-anodizing 471. See also anodizing: hard-coating hardening 384 age 46 case 45, 384

AGMA power rating 815 design procedure 817 double-enveloping 813 geometry 814 lubrication 814 materials 813 ratios 812 self-locking 813 single-enveloping 813

gearset 727, 731 angle of approach 730 angle of recess 730 application factor 758 arc of action 730 backlash 733, 753 contact ratio 738, 749, 753 definition 740 dynamic factor 754 external 727, 728, 735 highest point of single-tooth contact 750 internal 727, 728, 735 length of action 730, 738 load distribution factor 758 loading 749 fatigue

750

lubrication 775 pitch line velocity 754 pressure angle 729 variation with center distance 732

ratio limit 740 size factor 759 torque 749 transmission error 754 calculation of

755

gear train 740 compound 740 nonreverted 742 reverted 742

epicyclic 740, 743, 744 idler 740, 750 moment on

751

kinematic design of 740 simple 740 Gerber line 359 Goodman diagram modified 398, 599 torsional

plastics, table of

Guest, J.

case crushing 501 subcase fatigue 501

738,

cold working 46 cyaniding 45 flame 45 induction 46 mechanical 46, 53 nitriding 45 precipitation 46 strain 46 surface 45 through 384 hardness 42, 44 absolute 43 of alloy steels, table 1041 of aluminum, table 1037 of carbon steels, table 1040 of cast irons, table 1039 of copper alloys, table 1037 of stainless steel, table 1039 surface 48, 461 hay-bale lifter 567 heat treating 44 nonferrous 46 tempering. See tempering helical overlap 796 helix angle 791, 794 Hencky 282, 287. See also von Mises herringbone gear 795 Hertz, H. 476 Hertzian stress 475, 690 high-cycle fatigue 345 hob (gear) 736 Holzer’s method 637, 642 homogeneity 42, 288 Hooke’s law 31, 39 hot-working 47 hovercraft 671, 712 Hueber 287 hunting 733 hydraulic cylinder 7 hydrogen embrittlement 50 hydrostatic loading 282

856

Goodman line 359, 398, 752 graphite 674 Grashof 423 Grashof condition 76 gravitational constant 20, 22 system 20 gravity specific

D

288

An Integrated Approach

H

474, 725

design of

-

1042

-

Sixth Edition

indices of merit 87 induction hardening infinite life 347 interference 736 fit 620

stress concentration in 621 stresses in 620 torque transmitted 620

investment casting 746. See also manufacturing methods; involute 728, 731, 732, 736 definition 728 teeth 731 isotropic 288 isotropy 42, 288 iteration 5, 563 Izod impact test 40

J jacks 909 joint aspect ratio 940 constant 930, 943 separation 931 stiffness factor determining

joints gasketed

939

943

confined 943 unconfined 943

of different materials

force impact 135 striking impact 135

resistance 38 testing 40 inch-pound-second (ips) 21 indeterminate beam 207 indexing mechanisms 1011

941

K key 591, 610 design 613 materials 613 parallel 610 stresses 612 bearing 613 shear 613

tapered 611 woodruff 225, 612 keyway 610 stress concentration 614 kinematics definition 73 Kuhn-Hardrath formula 380 Kutzbach 75 equation 75 paradox 76

I impact 134 loading 134

384

L Lanchester damper laser peening 386 laybar 423 lead 812, 906 angle power screw worm 812

912

of a thread 906 lead screw 903, 909

642

INDEX length of action 730, 738 Liberty ships. See failure: of Liberty Ships linear actuators 910 links 665 Littman and Widner 500 loading classes 73, 74, 101, 281 dynamic 281 fatigue 281, 590, 591 impact 281

absolute 669 kinematic 669

lubrication. See lubrication: boundary boundary 667, 670, 671, 689, 712, 761, 775 673

667, 670, 712

definition 672 film thickness 672, 677, 690, 691 partial 689 specific film thickness 689, 690

full-film 670, 671 hydrodynamic 667, 670, 671, 761, 775 hydrostatic 667, 670, 671 water film 671

mixed film 670 670

hydrodynamic

lumped mass

674

637

M machine definition 3 screw 921 magnesium 59 malleable iron 52

221

221

667, 672,

acceleration equation 91 analysis 89 position equation 89 velocity equation 90

221

ductility 18 even and uneven 35, 294 failure 280 for springs 834 fracture toughness 305 hardness 42 homogeneous 42, 288, 301 isotropic 42, 288, 301 properties 1035

geared fivebar Grashof 76

92

crank-rocker 77 double-crank 77 double-rocker 77 equation 76

internal combustion engine 79 inversions 77

affected by temperature 310 mechanical 29 of whiskers 63 pure 62 statistical nature of 41 tables of 1035

definition

76

inverted crank-slider joint

92

definition 75 full 75 half 75 prismatic 75 revolute 75

shaft 593 sintered 674 test

Kutzbach

bending 35 compression 35 impact 40 rotating-beam 37 shear 35 tensile 31, 40 torsion 35, 36

equation 75 paradoxes 76

Kutzbach equation links 74

testing 31 uneven and even 35 materials abrasion-resistant 471 mathematical models 8 matrix-reduction computer program maximum shear stress in Hertzian contact 483, 488 mean 41 mechanical advantage 86, 728 prestressing 386 properties 29 mechanism acceleration of a point on a link

88

fourbar crank-slider

strength of 280 stress concentration in

solid 668, 674 solid-film 668 viscosity 669

of gearsets 672, 775 of nonconforming contacts 688 oil whirl 667 squeeze-film 667 theory 667

acceleration difference equation acceleration equation 87 analysis 82 crank-slider 79 crossed 83 open 83 position equation 83 slider-crank 79 velocity equation 85

18

stress concentration in

extreme pressure (EP) 464, 472, 503, 668, 669, 673, 775, 814

definition

material 29 anisotropic 42 bearing 673 brittle

compatibility 465 composites 42, 309 cracks in 301, 310 ductile

rotating machinery 349 service equipment 350 static 281 load sharing 738 ratio 797 low-cycle fatigue 345, 348 lubricant 509 gaseous 668 grease 668 liquid 464, 667, 668

definition

change point 76, 77 circuits 83, 84 crossed 84 definition 74 effective links 80 efficiency 86 fourbar

drawing 47 extrusion 48 forging 47

stress concentration in

210

elastohydrodynamic

manufacturing methods forming

brittleness cast

impact. See impact loading pure shear 287 pure torsional 210, 287 assumptions for

1071

89

angular velocity ratio 86 auto suspension 78 cam and follower 79 cam-follower systems 92

75

binary 74 quaternary 74 ternary 74

110

mechanical advantage 87 multi-DOF 73 nodes 74 non-constant velocity 77 one-DOF 73 open 84 other linkages 92 parallelogram 77 piston pump 79 power in 86 program Linkages 92 sixbar 78 Stephenson Watt 78

78

special-case Grashof 76 torque ratio 86 vector-loop method 92 mesh convergence 528 density 528 refinement 528

B D

1072 metal cast iron

MACHINE DESIGN

of shaft 639

coatings 48 corrosion 48 electroplatable 50 grain 837 47

noble 49 properties 52 sintered 674, 746 surface treatments 48 microcrack 343 microhardness tests 43 mks system 22 mobility equation 75 modal analysis 635 model engineering 9 physical 9 modified Goodman diagram 398, 402. See also Goodman diagram: modified; See also Goodman line module 734, 753, 776 modulus of elasticity 32 of elasticity, table of plastics

1042

of rigidity 36, 639 of rupture 36 MOHR computer program 182 Mohr plane 181 Mohr’s circle 180, 279, 282, 294, 480 for even and uneven materials 294 for hydrostatic test 282 for tensile test brittle material 294 ductile material 279

for torsion test 280, 294 moisture 310 molasses tank rupture. See failure: molasses-tank moment 139 M/EI function 607 moment of inertia definition 227 polar of shaft

639

motion complex 80 rotation 80 translation 80

N

D

An Integrated Approach

of compression springs 845 torsional

52

ductile 53 gray 52 nodular 53 white 52

structure

-

Nadai 287 natural frequency fundamental 634 multiple 130, 597

Neuber constant 382 Neuber’s equation 379 neutral axis 191, 194 Newton’s laws first law 104 second law 21, 104 third law 105 nitriding 384 node 640 nodes 522 nodular iron 593 nonmetallic materials 60 ceramics 62 composites 62 polymers 60 thermoplastic thermosetting

61 61

properties 60, 1042 normal force 462 normalizing 45 notch 343, 378 sensitivity 379 nuts 903, 910, 921, 923 acorn 923 castle 923 hex 923 jam 923 lock 923 minimum length 920 wing 923

O octahedral stresses. See distortion energy; See also von Mises Ocvirk equation 679 number 682, 684, 685 open mechanism 83 oxidation 472

P Paris equation 360 particle size 470 Peterson, R. E. 379 Petroff’s equation 675 phosphates 472 photoelastic stress analysis 480 pillow blocks 708 pin 591, 665 taper 592, 593 pinion 727, 731 pitch axial 794 circle 728, 729, 731, 733 virtual 795

circular

794. See also circular pitch

normal 795

-

Sixth Edition

diameters 728, 731 diametral 734, 735, 749, 752, 753, 776, 794. See also diametral pitch in normal plane

794

point 728, 729 pitting 476, 508 of gear teeth 752 of rolling bearings 700 plane strain 526 stress 526 plane strain 483 plane stress 483 planet gear 743 plastic-behavior 32 plastics maximum temperature, table 1042 plating electroless 50 Poisson’s ratio 36, 478 Poncelet 341 pounds force (lbf) 22 pounds mass (lbm) 21 power screw 903, 909 back driven 914 self-locking 914 torque in 912 power, shaft 593 preload 930 controlling in bolts 948 preloaded bolts 929 fasteners 925 preloaded structure 73 press-fit 593 dowel pins 952 pressure angle 796 cam-follower flat-faced

of gearsets

729

729, 732, 794

normal 793 transverse 793

of wormsets 813 principal stresses 483, 488. See also stress: principal profilometer 459 proof strength 925 proportional limit 31 prototype 9

Q quenching

44

R Rabinowicz, E. 458 rack 731 and pinion 731 steering

731

cutter 747 helical 793 radius of curvature 100 undercutting 101 radius of gyration 227

INDEX Rayleigh-Ritz method 637 Rayleigh’s method 635, 638 residual stresses 383, 387 compressive 359 resilience 39 resonance 132, 634 Reynolds’ equation 676, 678 ring gear 744 rocket motor case. See failure: of rocket-motor case Rockwell test 43 rod end 708 roll-bonding 466 roller bearings 697 rollers 482 crowned 500 logarithmic curve

500

stress concentration in 500 rolling cylinders 727, 728 rolling contact combined rolling and sliding gear teeth

contact patch

491

761

475

half-width 476, 483 semi-ellipsoid 486

contact pressure contact stress

design

subsurface shear stress

479, 498

cylinder-on-cylinder 474 cylindrical contact 482 geometry constant 477 Hertzian stress 483 stress distribution

SAE wire alloy numbers 834 safety factor 17, 286 fluctuating stress 405, 599 preloaded bolts dynamic loading

935

screw 903 efficiency 915 lead 903, 909 power 903, 909 torque to lower load 913 torque to raise load 912

torsional stress in

920

594

608

stresses 594, 596 transmission 589, 665 vibration 635 lateral 635, 639 torsional 639

37

Sommerfeld number 678 space width (gears) 733 spalling 476, 501, 509 of rolling bearings 700 specific film thickness 689 gravity plastics, table of

1042

stiffness 33 strength 33 splines 618 sprag 1011 spreadsheet 563 spring constant 131 torsional

607, 639

flat bend factor

837

flat-strip stock 837 materials 834 rectangular wire 842 wire 834 spring index 840, 864 spring rate 216, 218, 829, 840 combined 831 in parallel 831 in series 831

helical compression spring 840 helical extension springs 864 springs 218, 563 beam 832 Belleville 881

597

effective spring constant

487

S

ASME method

hardened 593 hollow 597, 744 key 592 loading 594 natural frequency 597, 633 stepped 591, 592, 607

material constants 477 pit formation 479 pitting 495 semi-ellipsoid sphere-on-sphere 474 rolling-element bearings 695 rolls nip 474, 499 rotating-beam test 37, 38 rotating machinery 348 loading in 349 R. R. Moore rotating-beam test

636

596

time-varying

483

pressure distribution

screws ball 915 self-drilling 921 self-tapping 921 slotted 921 socket cap 922 tapping 921 thread-cutting 921 thread-forming 921 second moment of area 227 polar 639 section modulus 190 self-locking 817, 818 of a lead screw 914 of a worm 813 SEMS 923 SEQA equation 415 servo mechanisms 733 motor 910 shaft 563 as a beam 607 as a torsion bar 607 critical speed 633 deflections 596, 597, 606 dynamic

476

1073

whirl 637, 639, 643 shear 139 area 187 direct 187 double 187 single 187 torsional 210 transverse, in beams 192, 194 shot peening 384, 385, 386, 844 significant figures 15 silicon carbide 471 Simpson’s rule 607 Sines equation 413, 414 singularity functions 141, 198. See also functions: singularity SI system 20 slop 733 slugs 21, 22 Smith and Lui 491 snap-ring 591, 592 S-N diagram 37 estimated 373 knee of 374

designing 886 dynamic loading 885 load-deflection relationship stacking 885 static loading 885 stresses in 884

cantilever 832 clock 834 helical compression

882

832, 838

active coils 839 assembled length 839 barrel 832 buckling of 844 clash allowance 839 conical 832 designing for dynamic loads 856 designing for static loads 851 direct shear factor 841 end details 839 free length 838 hourglass 832 mean coil diameter 838 natural frequency of 845 residual stresses 843 setting 843 shut height 839, 840 spring rate 840 stresses in 840 surging 845 torsional fatigue strength 846 torsional yield strength 846

helical extension

832, 863

active coils 863 coil preload 864

B D

1074

MACHINE DESIGN

deflection 864 design of 866 drawbar 832 hooks and loops 863 spring rate 864 stresses in 865

helical torsion

832, 874

active coils 875 coil closure 876 design of 878 spring rate 876 stress in 876

54

strengths of, table 1040

55

300 series 56 400 series 56 austenitic 56 martensitic 56 strengths of, table 1039

tool

55

strengths of, table 1041

wrought 53 stepping motor 910 stiffness of a joint determining

939

specific 33, 597, 642 stiffness constant of a joint 930 stiffness matrix 524 reduced 525 straight-eight engine 642 straight-line linkage 7 strain 31, 177, 526 plane 180 strain energy 38, 207, 282, 283 components of 283 strength bending fatigue of spring wire

D

compressive 35 creep 40 endurance 37

877

fatigue 37, 347 impact 38 of screw fasteners 925 of various materials, tables 1035 proof. See proof strength shear yield 37, 287 specific 33 tensile 32 torsional fatigue of spring wire

strengths of, table 1041

stainless

An Integrated Approach

846, 866

torsional yield

in series 607 load reversal in 844 motor 834 power 834 shot-peening of 844 variable rate 832 volute 832 Wahl’s factor for 841 washers 832 square thread 910 standard deviation 41 static load analysis 106 static strength 38 statistical considerations 20 steel alloy 55 cast 53 cold-rolled 53 hot-rolled 53 numbering systems plain carbon 54

-

of helical spring wire

846

to-weight ratio 33 ultimate tensile 290 as function of hardness

ultimate torsion

836

32, 280

brittle material

32

strength-to-weight ratio 33 stress 31, 173, 527 2-D 175, 179 3-D 175 alternating component 350 amplitude ratio 350 applied 178, 185 as a function of time 349, 350 bearing 187 combined 285 concentration 220, 309, 343, 591, 592 at notches 378 designing to avoid 224 due to notches 220 dynamic loads 222, 595 geometric 221, 222 static loads 220 torsional 595

concentration factor in threads

934

concentration factors 1047 contact 475 corrosion 343, 473 cubic polynomial 178 cyclic 343 effective 285, 297 Dowling 298 von Mises 285

fluctuating

349

fully reversed designing for

349, 591 388

Hertzian 475, 478, 502 induced 185 in gears helical 796 spur 796

intensity factor

multiaxial designing for in fatigue fluctuating 414 fully-reversed 413

40, 303

range 360

maximum bending 190 maximum shear 177, 178 mean component 350

412, 594

nominal 220, 281 normal 173, 280 octahedral 287 plane 180 principal 177, 178, 283 raisers 220, 597 range 350 ratio 350 repeated 349 residual 383, 387 383

residual compressive 359, 383, 384, 387 shafts 594 shear 36, 175, 280 tensile 343 thermal 384 von Mises 304, 483, 599 stress analysis photoelastic 480 stress concentration 343, 378, 509 factors 378 fatigue 379 geometric 379, 500

with fluctuating stress 400 stress-corrosion 343, 473 stress intensity factor. See stress: intensity factor stress-strain curve 35 stress-time functions 349, 350 structure 73 stud 921 stylus 459 sulfides 472 sun gear 743 surface asperities 461 coatings 509 compressive stresses 384, 387 contaminants 466 crack initiation inclusion origin

fatigue

designing for 396 design steps for 405

Sixth Edition

methods for introducing

838

36

as function of tensile strength

yield

-

500

457, 474, 508, 509, 752

strength 501 subcase failure 501 USM Corp test data 501

peeling 501 pitting 498 effect of lubricant 499 fatigue 503 point surface origin 501 subsurface cracks 500

polishing 470 roughness 459, 670, 690 composite 689 effect on friction 463 parameters 460 skewness 461 waviness 461

scoring 465 scuffing 465

INDEX thrust bearing 665 ball or roller 697 hydrostatic 671 titanium 58 tooth thickness 733 torque 749 coefficient 948 converter 1014 fluctuating 591 needed for preload 948 pinion 807 ratio 86, 728 repeated 591 -time function 629 wrench 948 torsion 36, 210, 220, 589 in circular cross-sections 211 in noncircular cross-sections 212 test 35, 280, 294 torsional damper 642 toughness 39 trade-offs 7 train ratio 741 transmission 728 automotive 793, 1010, 1011, 1014 trapezoidal rule 607 Tresca 288 true stress-strain curve 33 tumbling 470 tuned absorber 642 turn-of-the-nut method 949

spalling 498 treatments 384 autofrettage 386 coining 386 cold forming 384, 386 mechanical prestressing 386 shot peening 384, 385

synthesis 7 system discrete 634

T tapping screws 921 tear-out 188 teeth virtual 795 Teflon 674 temperature 310 effects 40 maximum for plastics 1042 recrystallization 47 tempering 45 tensile strength 32 of alloy steels, table 1041 of aluminums, table 1037 of carbon steels, table 1040 of cast irons, table 1039 of copper alloys, table 1037 of plastics, table 1042 of stainless steel, table 1039 tensile test 31, 35, 284, 290 thread 906 Acme 910, 913 Acme stub 910 buttress 910 class of fit 907 cutting 923 lead angle 912 minimum nut length 920 minimum tapped hole engagement multiple 906 multiple-start 906 pitch 907 rolling advantages of

U

920

924

specification 907 square 910, 912 standard 906 dimensions 908

stress 907 stripping-shear area 919 tensile-stress area 907 Unified National Standard (UNS) coarse series 906 extra fine series 907 fine series 906

threads stress-concentration in stresses in 918 axial 919 shear 919

934

ultimate compressive strength of cast irons, table 1039 of plastics, table 1042 ultimate tensile strength 32 of alloy steels, table 1041 of aluminum, table 1037 of carbon steels, table 1040 of cast irons, table 1039 of copper alloys, table 1037 of plastics, table 1042 of stainless steel, table 1039 undercutting 736, 795 uneven material 35 uniaxial stress state 284 units 542 units systems 20 unstructured problem 10 U.S. units system 20

906

V vector loop equation 82 vee-eight engine 642 velocity ratio 728, 735 of involute gears

vibration 130 self-excited 639 torsional controlling

642

732

1075 torsional damper 642 tuned absorber 642 Vickers hardness test 43 virtual gear 795 teeth 795 viscosity 669, 676 absolute 669, 690 units of

669

kinematic units of

von Mises

669 669

282, 285, 287, 297

W washers 921, 923 Belleville 832, 881, 923 fender 923 load-indicating 949 lock 923 water-jet 423 Way, S. 501 wear 457 abrasive 457, 470, 503, 670 adhesive 457, 464, 670 in gear teeth

752

corrosive 472 Weibull distribution 700 weight 21 weld area 978 as a line 993 atomic cleanliness 975 backing strip 979 eccentric loading 999 electrode numbering

982

failure from compression stress 992

filler metal 975 fusion 975 groove size

979

HAZ 975 heat affected zone 975 hydrogen embrittlement 975 joint butt 977 corner 977 edge 977 lap 977 preparation 979 shapes 979 tee 977

leg width 978 metal 975 fatigue strength of

989

overmatching 983 parent metal 975 penetration 975 complete 977 partial 977

reinforcement safety factor static

983

978

B D

1076 slag 975 specification strength

MACHINE DESIGN

yield 280 point 32 strength 32 of alloy steels, table 1041 of aluminum, table 1037 of carbon steels, table 1040 of copper alloys, table 1037 of stainless steel, table 1039

stress dynamic 986 mean 986 range 986 residual 983, 992 static 981 static allowable 983

Young’s modulus 32, 33, 39, 197 tables 52, 1042

Z Zimmerli, F. P.

978

fillet 978 groove 977 laser 977 plug 977 seam 977 slot 977 spot 977

undermatching 983 welding codes 979 electrode 975 gas 975 symbol 979 type arc 975 FCAW 976 GMAW 976 GTAW 976 MIG 976 resistance 977 SAW 977 SMAW 976 TIG 976

welding, cold 466 weldment categories 987 cost 1001 design considerations 1000 design principles 979 wire rectangular 842 square 842 strength as function of size 834 Wohler August 339 strength-life diagram 37, 341 worm gear 811 wheel 811 wrench pneumatic impact 948 torque 948 error in preload

An Integrated Approach

Y

979

endurance 987 fatigue 986 fatigue safety factor 991 reliability factors 989 static 982 testing 989

throat 977 throat width type 977

-

948

writing engineering reports wrought 34

27

846

-

Sixth Edition

DOWNLOADS INDEX for Machine Design by Norton Sixth Edition © 2020. All rights reserved.

The book’s website contains over 400 model files that encode most of the Example and Case-Study solutions in the text. These are encoded separately in formats for the programs MATLAB, Mathcad, Excel, and TK Solver for use with Windows/NT/2000/XP/ Vista/7. In addition, over 100 general TK Solver solution files for beams, springs, shafts, fasteners, and a collection of TK Solver rule, list, and procedure functions that can be merged into other files are provided. Solidworks CAD models are provided of a number of Case Studies and Problem sets in the book. The Case Studies also have files for FEA solutions in Solidworks provided. The free eDrawings viewer can display these files if Solidworks is not available. The web address for its download is given in this index. Twenty-one Master Lectures by the author on the topics in many of the book’s chapters are provided as videos. In addition, videos that demonstrate particular topics such as bending stress, transverse and torsional shear stress, columns, and the result of these stresses as applied to failed test specimens are included. Videos that demonstrate common machine parts such as bearings, gears, and springs are also supplied. Several videos of actual machinery in operation are included as well. Programs for kinematic, dynamic, and stress analysis are also included on the website. These are called Dynacam, Linkages, Matrix, and Mohr, and are installable as executable files. Some data files for examples and case studies that were solved with these programs are also provided in the corresponding programs’ folders. Tutorials and examples for the use of MATLAB, Excel, Mathcad and TK Solver are provided on the website, but the MATLAB, Excel, TK Solver, and Mathcad programs are NOT included with this text. To run their files will require access to one or more of these programs. Instructors who adopt the text can obtain solution files for the book’s problem sets encoded for Mathcad directly from the publisher along with a PDF solutions manual. 1077

1078

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

VIDEOS See the Video Contents. TUTORIALS Excel

Path: Tutorials \ Excel Tutorial

Excel Intro.doc

An introduction to Excel with 10 example files and detailed explanations.

Mathcad

Path: Tutorials \ Mathcad Tutorial

Mathcad Intro.pdf

An introduction to Mathcad with 6 example files and detailed explanations.

MATLAB

Path: Tutorials \ MATLAB Tutorial

MATLAB Intro.doc An introduction to MATLAB with 14 example m-files and detailed explanations.

TKSolver

Path: Tutorials \ TKSolver Tutorial

TKSolver Intro.pdf

An introduction to TK Solver with 9 example files and detailed explanations.

MODEL FILES - EXAMPLES

*



No TK Solver file for this one.

These examples are also solved with program Mohr. Their files (EX04-xx.MOH) can be found in the folder PROGRAM FILES \ MOHR.

Examples

Path: Excel Files \ Excel Examples \ Chap_No Path: Mathcad Files \ Mathcad Examples \ Chap_No Path: MATLAB Files \ MATLAB Examples \ Chap_No Path: TKSolver Files \ TKSolver Examples \ Chap_No Path: PDF Files \ Examples \ Chap_No

EX02‑01*

An example that determines a material’s modulus of elasticity and yield strength from test data.

EX03‑01A

Example of impact of a mass against a horizontal rod in axial tension. Examines and plots the sensitivity of the impact force to the length/diameter ratio of the rod for a constant mass ratio. (See Figure 3-18.)

EX03‑01B

Example of impact of a mass against a horizontal rod in axial tension. Examines and plots the sensitivity of the impact force to the mass ratio of the rod for a constant length/diameter ratio. (See Figure 3-18.)

EX03‑02

Calculates the loading, shear, and moment functions for a simply supported beam with a uniformly distributed load over a portion of its length ending at one support. Finds reactions and plots the beam functions. (See Figure 3-22a.)

EX03‑03

Calculates the loading, shear, and moment functions for a cantilever beam with a concentrated load at any point along its length. Finds reactions and plots the beam functions. (See Figure 3-22b.)

EX03‑04

Calculates the loading, shear, and moment functions for an overhung beam with a moment load at any point along its length and with a ramp load over a portion of its length beginning at one support. Finds reactions and plots the beam functions. (See Figure 3-22c.)

EX04‑01†

Solves the stress cubic and finds the principal stresses and maximum shear for given values. Is same program as STRESS3D with data for this example. (See Figure 4-5.)

EX04‑02†

Solves the stress cubic and finds the principal stresses and maximum shear for given values. Is same program as STRESS3D with data for this example. (See Figure 4-6.)

EX04‑03†

Solves the stress cubic and finds the principal stresses and maximum shear for given values. Is same program as STRESS3D with data for this example. (See Figure 4-7.)

INDEX OF SOFTWARE ON BOOK’S WEBSITE

EX04‑04

Calculates the shear, moment, slope, and deflection functions for a simply supported beam with a uniformly distributed load over a portion of its length ending at one support. Finds reactions, plots the beam functions, and finds max and min values. (See Figure 3-22a.)

EX04‑05

Calculates the shear, moment, slope, and deflection functions for a cantilever beam with a concentrated load at any point along its length. Finds reactions, plots the beam functions, and finds their max and min values. Is the same program as CANTCONC with data for this example. (See Figure 4-22b.)

EX04‑06

Calculates the shear, moment, slope, and deflection functions for an overhung beam with a concentrated load at any point along its length and with a uniformly distributed load over a portion of its length beginning at one support. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 4-22c.)

EX04‑07

Calculates the shear, moment, slope, and deflection functions for a statically indeterminate beam with a uniformly distributed load over a portion of its length. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 4-22d.)

EX04‑08

Calculates the shear, moment, slope, and deflection functions for a statically indeterminate beam with a uniformly distributed load over its length using Castigliano’s method. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 4-22e.)

EX04‑09

Determines the best cross-sectional shape for a hollow bar loaded in pure torsion. (See Figure 4-29.)

EX04‑10

Calculates the stresses due to combined bending and torsional loading. (See Figure 4-30.)

EX04‑11C*

Designs columns in circular cross sections for concentric loading using both Johnson and Euler criteria to find critical load, weight, and safety factor. Is the same program as COLMNDES with data for this example. (See Figure 4-42.)

EX04‑11S

Designs columns in square cross sections for concentric loading using both Johnson and Euler criteria to find critical load, weight, and safety factor. Is same program as COLMNDES with data for this example. (See Figure 4-42.)

EX05‑01

Calculates the principal and von Mises stresses for a bracket made of ductile material and loaded in combined bending and torsion. Finds the safety factors based on the distortion-energy and maximum-shear-stress theories. (See Figure 5-9.)

EX05‑02

Calculates the principal and von Mises stresses for a bracket made of brittle material and loaded in combined bending and torsion. Finds the safety factors based on the modified-Mohr theory. (See Figure 5-9.)

EX05‑03

Calculates the fracture mechanics failure criteria for a cracked part. Compares the fracture-mechanics failure stress with a yield failure. (See Figure 5-19.)

EX06‑01

Calculates the corrected endurance strength of ferrous metals based on supplied data about finish, size, strength, etc., and draws an estimated S‑N diagram for supplied levels of alternating and mean stresses. Is the same program as S_NDIAGM with data for this example. (See Figure 6-34.)

EX06‑02

Calculates the corrected endurance strength of nonferrous metals based on supplied data about finish, size, strength, etc., and draws an estimated S‑N diagram for supplied levels of alternating and mean stresses. Is the same program as S_NDIAGM with data for this example. (See Figure 6-33.)

EX06‑03

Finds the fatigue stress-concentration factor for a part of known material and geometry. (See Figure 4-36 or Figure E-10.)

EX06‑04A†

Design of a cantilever bracket for fully reversed bending—part a: an unsuccessful design. (See Figure 6-41.)

EX06‑04B

Design of a cantilever bracket for fully reversed bending—part b: a successful design. (See Figure 6-41.)

EX06‑05A‡

Design of a cantilever bracket for fluctuating bending—part a: an unsuccessful design. (See Figure 6-47.)

1079

* The Mathcad solution to this example is labeled EX04-10, and it contains both the square and circular column crosssection solutions.

† The Mathcad solutions to this example are labeled EX06-04 and EX-06-04A. The latter model shows an alternate approach to the solution than that which is shown in the text and in the TK Solver files. ‡ The Mathcad solution to this example is labeled EX06-05.

1080



An alternate approach to the solution of this problem is presented in the Mathcad file EX14-0xA.

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

EX06‑05B

Design of a cantilever bracket for fluctuating bending—part b: a successful design. (See Figure 6-47.)

EX06‑06

Design of a cantilever bracket for multiaxial stresses in fatigue. (See Figure 5-9.)

EX07‑01

Stresses in a ball thrust bearing. Uses SURFSPHR to calculate surface stresses in a spherical-flat contact. (See Figure 10-19.)

EX07‑02

Stresses in cylindrical contact. Uses SURFCYLZ to calculate surface stresses in a wheel-on-rail contact. (See Figure 7-17.)

EX07‑03

Stresses in general contact. Uses SURFGENL to calculate surface stresses in a crowned cam-follower contact. (See Figure 7-11.)

EX07‑04

Stresses in combined rolling and sliding in cylindrical contact. Uses SURFCYLX to calculate surface stresses in a nip-roller contact. (See Figure 7-13.)

EX07‑05

Safety factor in combined rolling and sliding in cylindrical contact problem of Example 7-04. Uses data from Table 7-8. (See Figure 7-13.)

EX08‑01

Deflection of a cantilever beam. (See Figure 8-7.) (TK Solver file only)

EX10‑01

Shaft design for steady torsion and fully reversed bending (parts a to d). (See Figure 9-5.)

EX10‑02

Shaft design for repeated torsion combined with repeated bending (parts a to d). (See Figure 10-5.)

EX10‑03

Designing a stepped shaft to minimize deflection. (See Figure 10-5.)

EX10‑04

Designing shaft keys—parts a to d. (See Figure 10-5.)

EX10‑04A

Designing shaft keys—an alternate approach. (See Figure 10-5.)

EX10‑05

Designing an interference fit—parts a and b. (See Figure 10-18.)

EX10‑07

Designing a solid-disk flywheel—parts a and b. (See Figure 10-21.)

EX10‑08

Determining the critical frequencies of a shaft—parts a and b. (See Figure 10-30.)

EX11‑01

Sleeve-bearing design—parts a and b. (See Figure 11-8.)

EX11‑02

Lubrication in a crowned cam-follower interface. (See Figure 11-4.)

EX11‑03

Selection of ball bearings for a designed shaft.

EX11‑04

Selection of ball bearings for combined radial and thrust loads.

EX12‑01

Determining gear tooth and gearmesh parameters. (See Figure 12-8.)

EX12‑03

Analyzing an epicyclic-gear train. (See Figure 12-16.)

EX12‑04

Load analysis of a spur-gear train. (See Figure 12-21.)

EX12‑05

Bending stress analysis of a spur-gear train. (See Figure 12-20.)

EX12‑06

Surface stress analysis of a spur-gear train. (See Figure 12-20.)

EX12‑07

Material selection and safety factor for spur gears—parts a and b. (See Figure 12-20.)

EX13‑01

Stress analysis of a helical-gear train. (See Figure 13-2.)

EX13‑02

Stress analysis of a bevel-gear train. (See Figure 13-4.)

EX14‑03‡

Design of a helical compression spring for static loading—parts a and b. (See Figure 14-7.)

EX14‑04‡

Design of a helical compression spring for cyclic loading—parts a and b. (See Figure 14-7.)

EX14‑05

Design of a helical extension spring for cyclic loading—parts a and b. (See Figure 14-20.)

EX14‑06

Design of a helical torsion spring for cyclic loading. (See Figure 14-26.)

EX14‑07

Design of a Belleville spring for static loading. (See Figure 14-29.)

INDEX OF SOFTWARE ON BOOK’S WEBSITE

EX15‑01

Torque and efficiency of a power screw—parts a and b. (See Figure 15-4.)

EX15‑02

Preloaded fasteners in static loading. (See Figure 15-24.)

EX15‑03

Preloaded fasteners in dynamic loading. (See Figure 15-26.)

EX15‑04

Determining material stiffness and the joint constant. (See Figure 15-31.)

EX15‑05

Determining the torque needed to generate a bolt preload. (See Figure 15-31.)

EX15‑06

Fasteners in eccentric shear. (See Figure 15-42.)

EX16‑01

Design of a statically loaded fillet weld.

EX16‑02

Design of a dynamically loaded fillet weld.

EX16‑03

Design of a statically loaded weldment assembly.

EX16‑04

Design of a dynamically loaded weldment assembly.

EX16‑05

Design of an eccentrically loaded weldment assembly.

EX17‑01

Design of a disk clutch. (See Figure 17-6.)

EX17‑02

Design of a short-shoe drum brake—parts a and b. (See Figure 17-8.)

EX17‑03

Design of a long-shoe drum brake. (See Figure 17-9.)

1081

MODEL FILES - CASE STUDIES Case Studies

Path: Excel Files \ Excel Cases \ CaseNo Path: Mathcad Files \ Mathcad Cases \ CaseNo Path: MATLAB Files \ MATLAB Cases \ CaseNo Path: TKSolver Files \ TKSolver Cases \ CaseNo Path: PDF Files \ Case Studies \ CaseNo

CASE1A†

Case study of the force analysis of a bicycle brake lever under static, 2-D loading. Finds reaction forces. See Chapter 3 and Figure 3-2.

CASE1B

Case study of the stress and deflection analysis of a bicycle brake lever under static, 2-D loading. Finds and plots the beam functions, shear, moment, and deflection and determines stresses at particular locations. See Chapter 4 and Figure 4-47.

CASE1C

Case study of the stress and deflection analysis of a bicycle brake lever under static, 2-D loading. Finds safety factors at particular locations. See Chapter 5 and Figure 4-47.

CASE2A†

Case study of the force analysis of a hand crimping tool under static, 2-D loading. Finds reaction forces. See Chapter 3 and Figure 3-4.

CASE2B‑x

Case study of the stress and deflection analysis of a hand crimping tool under static, 2-D loading. Finds and plots the beam functions, shear, moment, and deflection and determines stresses at particular locations. See Chapter 4 and Figure 4-49.

CASE2C‑x

Case study of the failure analysis of a hand crimping tool under static, 2-D loading. Finds the safety factors at particular locations. See Chapter 5 and Figure 5-22.

CASE3A†

Case study of the force analysis of a scissors jack under static, 2-D loading. Finds reaction forces. See Chapter 3 and Figure 3-8.

CASE3B‑x

Case study of the stress and deflection analysis of a scissors jack under static, 2-D loading. Finds and plots the beam functions, shear, moment, and deflection and determines stresses at particular locations. See Chapter 4 and Figure 4-52.

CASE3C

Case study of the failure analysis of a scissors jack under static, 2-D loading. Finds the safety factors at particular locations. See Chapter 5 and Figure 5-23.

CASE4A†

Case study of the force analysis of a bicycle brake arm under static, 3-D loading. Finds reaction forces. See Chapter 3 and Figure 3-10.



These case studies are also solved with program Matrix. Their files (CASExx.mtr) can be found in the folder PROGRAM FILES \ MATRIX.

1082

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

CASE4B

Case study of the stress and deflection analysis of a bicycle brake arm under static, 3-D loading. Finds and plots the beam functions, shear, moment, and deflection and determines stresses at particular locations. See Chapter 4 and Figure 4-54.

CASE4C

Case study of the failure analysis of a bicycle brake arm under static, 3-D loading. Finds the safety factors at particular locations. See Chapter 5 and Figure 5-24.

CASE5A

Case study of the force analysis of a fourbar linkage under dynamic, 2-D loading. Finds theoretical reaction forces. See Chapter 3 and Figure 3-13.

CASE6‑x

Eight files (–0 through –7) for a case study of the fatigue analysis and redesign of a failed power-loom laybar under dynamic, 2-D loading. See Chapter 6 and Figure 6-51.

CASE8A

Design of an engine-powered air compressor. This file sets up the design problem. See Chapter 8 and Figure 8-1.

CASE8B‑x

Design of an engine-powered air compressor. These 3 files (-1, -2, -3) design the transmission shafts connecting the engine and compressor. See Chapter 10 and Figure 9-1.

CASE8C‑x

Design of an engine-powered air compressor. These 2 files (-1, -2) design the spur gears connecting the engine and compressor. See Chapter 12 and Figure 9-1.

CASE8D

Design of an engine-powered air compressor. These 2 files (-1, -2) design the headbolts for the compressor. See Chapter 14 and Figure 9-1.

CASE9A‑x

Design of a hay-bale lifter. These 2 files (-1, -2) set up the design problem. See Chapter 9 and Figure 9-4.

CASE9B‑x

Design of a hay-bale lifter. These 2 files (-1, -2) design a worm and worm gear for the speed reducer. See Chapter 13 and Figure 9-4.

CASE10B

Design of a cam test machine. This file designs the hydrodynamic sleeve bearings for the camshaft. See Chapter 11 and Figures 9-8 and 11-31.

CASE10C‑x

Design of a cam test machine. These 2 files (-1, -2) design the coil spring for the cam follower. See Chapter 14 and Figure 14-36.

MODEL FILES - GENERAL Beams

Path: TKSolver Files \ TKSolver General \ Beams

BEAMFUNC

A collection of rule functions for various beam loadings and supports for use in programs. Can be combined for superposition of loads on any beam with consistent constraints. (See Figure 3-22.)

CANTCONC

Calculates the shear, moment, slope, and deflection functions for a cantilever beam with a concentrated load at any point along its length. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 3-22b.)

CANTCONC3

Calculates the shear, moment, slope, and deflection functions for a cantilever beam with three concentrated loads at any points along its length. Finds reactions, plots the beam functions, and finds their max and min values.

CANTMOMT

Calculates the shear, moment, slope, and deflection functions for a cantilever beam with moment loads at points along its length. Finds reactions, plots the beam functions, and finds their max and min values.

CANTUNIF

Calculates the shear, moment, slope, and deflection functions for a cantilever beam with a uniform load along its length. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 3-22a.)

CURVBEAM

Calculates eccentricity of neutral axis and stresses for curved beams of various cross sections—ellipse, circle, square, rectangular, and trapezoidal. (See Figure 4-16.)

INDTUNIF

Calculates the shear, moment, slope, and deflection functions for an indeterminate beam with a uniformly distributed load over a portion of its length ending at one sup-

INDEX OF SOFTWARE ON BOOK’S WEBSITE

port. Finds reactions, plots the beam functions, and finds max and min values. (See Figure 4-22d.) OVHGCONC

Calculates the shear, moment, slope, and deflection functions for an overhung beam with a concentrated load at any point along its length. Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 4-22c.)

OVHGMOMT

Calculates the shear, moment, slope, and deflection functions for an overhung beam with a moment load at any point along its length. Finds reactions, plots the beam functions, and finds their max and min values.

OVHGUNIF

Calculates the shear, moment, slope, and deflection functions for an overhung beam with a uniformly distributed load over a portion of its length beginning at one support and with an optional concentrated load at any point along its length . Finds reactions, plots the beam functions, and finds their max and min values. (See Figure 4-22c.)

SIMPCONC

Calculates the shear, moment, slope, and deflection functions for a simply supported beam with a concentrated load at any point along its length. Finds reactions, plots the beam functions, and finds their max and min values.

SIMPUNIF

Calculates the shear, moment, slope, and deflection functions for a simply supported beam with a uniformly distributed load over a portion of its length ending at one support. Finds reactions, plots the beam functions, and finds max and min values. (See Figure 4-22a.)

Bearings

Path: TKSolver Files \ TKSolver General \ Bearings

BALL6200

A ball-bearing selection program that calculates the L10 life for 6200 series ball bearings under specified loads. Based on data from the SKF bearing catalog. (See Figure 11-17.)

BALL6300

A ball-bearing selection program that calculates the L10 life for 6300 series ball bearings under specified loads. Based on data from the SKF bearing catalog. (See Figure 11-17.)

EHD_BRNG

Solves for film pressure in general elastohydrodynamic (EHD) contact between lubricated, nonconforming surfaces. Also finds the minimum oil-film thickness. (See Figure 11-4.)

SLEEVBRG

Calculates the film thickness, eccentricity, and oil pressure in a short (Ocvirk) sleeve bearing under hydrodynamic lubrication conditions. (See Figure 11-8.)

Clutch/Brake

Path: TKSolver Files \ TKSolver General \ ClchBrak

DISKCLCH

Designs a disk clutch for uniform wear. Allows single or multiple disks. (See Figure 15-6.)

LONGDRUM

Designs a long-shoe drum brake. (See Figure 15-9.)

SHRTDRUM

Designs a short-shoe drum brake. (See Figure 15-8.)

Columns

Path: TKSolver Files \ TKSolver General \ Columns

COLMNDES

A column design program that handles round, square, or rectangular concentric columns and uses both Johnson and Euler criteria to find critical load, weight and safety factor. (See Figure 4-42.)

SECANT

An eccentric column design program that handles round, square, or rectangular concentric columns and uses the secant method and Johnson and Euler criteria to find critical load and safety factor. Plots critical-load curves. (See Figure 4-44.)

Fastener

Path: TKSolver Files \ TKSolver General \ Fastener

BLTFATIG

Calculates the safety factors for preloaded bolts with fluctuating tensile loads. Determines the necessary tightening torque and plots the load-sharing, safety-factor, and modified-Goodman diagrams. (See Figure 15-26.)

1083

1084

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

BOLTSTAT

Calculates the safety factors for preloaded bolts with static tensile loads. Determines necessary tightening torque and plots the load-sharing and safety-factor diagrams. (See Figure 15-24.)

PWRSCREW

Calculates the torque and efficiency of an Acme-thread power screw. (See Figure 15-4.)

Fatigue

Path: TKSolver Files \ TKSolver General \ Fatigue

GDMNPLTR

A plotting utility that creates and plots a modified-Goodman diagram for any set of supplied stresses and strengths. No calculations are done on the data. (See Figure 6-44.)

GOODMAN

Calculates the corrected endurance strength based on supplied data about finish, size, strength, etc. and draws a modified-Goodman diagram for supplied levels of alternating and mean stresses and material strengths. Also calculates safety factors. (See Figure 6-46.)

S_NALUM

Calculates the corrected endurance strength based on supplied data about finish, size, strength, etc. and draws an S‑N diagram for supplied levels of alternating and mean stresses for nonferrous material. Draws both log-log and semilog plots. Based on the file S_NDIAGM. (See Figure 6-33.)

S_NDIAGM

Calculates the corrected endurance strength based on supplied data about finish, size, strength, etc. and draws an S‑N diagram for supplied levels of alternating and mean stresses. Draws both log-log and semilog plots. (See Figure 6-33.)

S_NFCTRS

Calculates the coefficient and exponent of the S‑N line for a material. (See Figure 6-33.)

SESAFTIG

Calculates and plots the variation on stress with phase in multiaxial fatigue based on the SESA algorithm. (See Figure 6-49.)

Flywheels

Path: TKSolver Files \ TKSolver General \ Flywheel

FWDESIGN

Program to find the best combination of flywheel diameters and thickness to balance its weight against size, stress and safety factor. Calculates the maximum stress, outside diameter, weight, and safety factor as a function of the thickness of a flywheel. (See Figure 10-21.)

FWRATIO

Optimizes flywheel mass versus the ratio of radius to thickness and plots that function. (See Figure 10-25.)

FWSTDIST

Program to find the flywheel stress distribution over its radius. Calculates and plots the stresses across the radius of a flywheel. (See Figure 10-24.)

Fract Mech

Path: TKSolver Files \ TKSolver General \ Frctmech

STRSINTS

Plots the stress intensity around a crack tip. (See Figure 5-19.)

Gearing

Path: TKSolver Files \ TKSolver General \ Gearing

BVLGRDES

Program for straight-bevel gearset design. Finds bending and surface stresses in gear teeth and safety factors using AGMA methods. Requires I and J factors be manually looked up in AGMA Tables. (See Figure 13-4.)

HELGRDES

Program for helical gearset design. Finds bending and surface stresses in gear teeth and safety factors using AGMA methods. Requires I and J factors be manually looked up in AGMA Tables. (See Figure 13-2.)

SPRGRDES

Calculates bending and surface stresses for a single-spur gearset (with or without an idler) based on AGMA formulas and determines safety factors for supplied material strengths. (See Figure 12-2.)

WORMGEAR

Worm and wormgear design based on AGMA formulas. (See Figure 13-9.)

INDEX OF SOFTWARE ON BOOK’S WEBSITE

Impact

Path: TKSolver Files \ TKSolver General \ Impact

IMPCTHRZ

Calculates the impact force on a horizontal rod struck by a mass. (See Figure 3-18.)

IMPCTVRT

Calculates the impact force on a vertical rod struck by a mass. (See Figure 3-18.)

Linkages

Path: TKSolver Files \ TKSolver General \ Linkages

4BARSTAT

Calculates the joint forces for a static fourbar linkage subjected to a known force applied to the coupler. (See Figure 3-13.)

4BAR_NEW

Calculates kinematics of a fourbar linkage.

DYNAFOUR

Calculates the kinematics and inverse dynamics of the fourbar linkage. Plots various linkage parameters like accelerations, forces, and torques. (See Figure 11-3 in ref. 1.)

ENGINE

Calculates a slider crank’s kinematics and the gas force and gas torque due to a specified explosion pressure at any position of the crank. It is a static force analysis. (See Figure 13-3 in ref. 1.)

ENGNBLNC

Calculates the dynamic balance condition of an IC engine. Plots shaking forces, torques, and moments. (See Figure 13-12 in ref. 1.)

FOURBAR

Calculates the kinematics and dynamics of a fourbar linkage, position, velocity, and acceleration, of various points for any range of motion. (See Figure 11-3 in ref. 1.)

SLIDER

Calculates the offset slider crank’s kinematics for any one position and input omega and alpha. Lists can be added for multiple-position analysis. No forces are calculated. (See Figure 7-6 in ref. 1.)

Shafts

Path: TKSolver Files \ TKSolver General \ Shafts

HOLTZER

Find first natural frequency of a shaft with lumped masses using Holtzer’s method. (See Figure 10-27.)

SHFTCONC

Program to design a simply supported shaft with concentrated load at any point along its length. Left support must be at x = 0, but right support can be anywhere. A fluctuating torque may be applied to the shaft. The moment is assumed to be fully reversed. Calculates and plots the shear, moment, and deflection functions. (See Figure P10-1.)

SHFTDESN

Program to design a simply supported shaft for fatigue in combined bending and torsion. A fluctuating torque may be applied to the shaft as well as a fluctuating moment. Also calculates stresses in a standard square key for the shaft diameter. (See Figure P10-3.)

SHFTUNIF

Program to design a simply supported shaft with uniform load over any portion of its length. Left support must be at x = 0 and right support is at length R2x. A fluctuating torque may be applied to the shaft, and the moment is assumed to be fully reversed. Calculates and plots the shear, moment, and deflection functions. (See Figure P10-2.)

STATSHFT

Calculates the shear stress in a shaft subjected to a constant torque with no transverse loads or moments. (See Figure 4-28.)

STEPSHFT

Program to design a simply supported stepped shaft with uniform load over any portion. Calculates deflection for stepped shaft. (See Figure 10-5.)

Springs

Path: TKSolver Files \ TKSolver General \ Springs

BELLEVIL

Calculates the load, deflection, and spring rate for a Belleville spring. Plots the nonlinear force-deflection curves for a family of springs. (See Figure 14-29.)

COMPRESS

Designs a helical coil compression spring for fatigue or static loading. Plots curves to allow an optimization of spring design. (See Figure 14-7.)

EXTENSN

Designs a helical coil extension spring for fatigue or static loading. Plots curves to allow an optimization of spring design. (See Figure 14-20.)

TORSION

Designs a helical coil torsion spring for fatigue or static loading. Plots curves to allow an optimization of spring design. (See Figure 14-26.)

1085

1086

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

Stress

Path: TKSolver Files \ TKSolver General \ Stress

COULMOHR

Calculates factors for the Coulomb-Mohr diagram for brittle, uneven materials. (See Figure 5-11.)

ELLIPSE

Draws the distortion-energy ellipse for demonstration purposes. (See Figure 5-3.)

MOD_MOHR

Modified-Mohr theory calculator for brittle materials. Uses Dowling’s method to find an effective stress for combined loading in brittle, uneven materials. (See Figure 5-13.)

STRES_2D

Calculates the principal stresses, maximum shear stress, and Von Mises stress for any two-dimensional applied stress state specified. (See Figure 4-3.)

STRES_3D

Calculates the principal stresses and maximum shear stress for any three-dimensional applied stress state specified. It also plots the stress cubic function. (See Figure 4-1.)

STRSFUNC

Two rule functions, one for the calculation of 2-D principal and one for Von Mises stresses. Use for merging into other programs that need these functions.

VONMISES

Uses the distortion-energy method to find an effective stress for combined loading in ductile, even materials under static loading. (See Figure 5-3.)

Stress Conc.

Path: TKSolver Files \ TKSolver General \ StrsConc

APP_E‑01

Calculates stress-concentration factor for a shaft with shoulder fillet in tension. (See Appendix E, Figure E-1.)

APP_E‑02

Calculates stress-concentration factor for a shaft with shoulder fillet in bending. (See Appendix E, Figure E-2.)

APP_E‑03

Calculates stress-concentration factor for a shaft with shoulder fillet in torsion. (See Appendix E, Figure E-3.)

APP_E‑04

Calculates stress-concentration factor for a shaft with a U groove in axial tension. (See Appendix E, Figure E-4.)

APP_E‑05

Calculates stress-concentration factor for a shaft with a U groove in bending. (See Appendix E, Figure E-5.)

APP_E‑06

Calculates stress-concentration factor for a shaft with a U groove in torsion. (See Appendix E, Figure E-6.)

APP_E‑07

Calculates stress-concentration factor for a shaft with transverse hole in bending. (See Appendix E, Figure E-7.)

APP_E‑08

Calculates stress-concentration factor for a shaft with transverse hole in torsion. (See Appendix E, Figure E-8.)

APP_E‑09

Calculates stress-concentration factor for a flat bar with shoulder fillet in tension. (See Appendix E, Figure E-9.)

APP_E‑10

Calculates stress-concentration factor for a flat bar with shoulder fillet in bending. (See Appendix E, Figure E-10.)

APP_E‑11

Calculates stress-concentration factor for a flat bar with notch in axial tension. (See Appendix E, Figure E-11.)

APP_E‑12

Calculates stress-concentration factor for a flat bar with notch in bending. (See Appendix E, Figure E-12.)

APP_E‑13

Calculates stress-concentration factor for a flat bar with transverse hole in tension. (See Appendix E, Figure E-13.)

APP_E‑14

Calculates stress-concentration factor for a flat bar with transverse hole in bending. (See Appendix E, Figure E-14.)

INDEX OF SOFTWARE ON BOOK’S WEBSITE

NTCHSENS

Plots the notch-sensitivity curves for steels. (See Figure 6-36 part 1.)

Q_CALC

Calculates the notch sensitivity q of a material. (See Figure 6-36 part 2.)

SC_HOLE

Calculates and plots stress concentration at an elliptical hole in a semi-infinite plate. (See Figure 4-35.)

Surface Stress

Path: TKSolver Files \ TKSolver General \ SurfStre

ROLLERS

Solves for subsurface stresses in cylindrical contact with sliding for the plane strain case (long cylinders).

SURFCYLX

Calculates the surface stresses for Hertzian contact of two cylinders, with or without a sliding component. Plots subsurface stress distributions across the contact-patch X-width at surface or at any Z -depth into material. (See Figure 7-17.)

SURFCYLZ

Calculates the surface stresses for Hertzian contact of two cylinders with or without a sliding component. Plots subsurface stress distributions from surface to any Z depth at any X-width across contact patch. (See Figure 7-17.)

SURFGENL

Calculates the surface stresses for Hertzian contact of two bodies of general shape. (See Figure 7-11.)

SURFSPHR

Calculates the surface stresses for Hertzian contact of two spheres, sphere-on-plane or sphere-in-bowl. Plots subsurface stress distributions. (See Figure 7-13.)

THCK_CYL

Calculates the stresses in walls of thick-cylinder pressure vessels. (See Figure 4-46.)

MODEL FILES - MASTERS Masters

Path: TKSolver Files \ TKSolver Masters

FORMATS

This file contains only a format sheet that can be added to (merged into) any other TK file without disturbing its other contents. The format sheet enables formatting of variables to any desired number of decimal places.

MDUNITS

This file is blank except for the format sheet from FORMATS and the units sheet from UNITMAST. It is intended to be merged into any file to add a units sheet and format sheet without disturbing the file contents. It is a combination of the files UNITMAST and FORMATS.

PROWEBSITEURS

This file contains a large number of rule, list, and procedure functions that are used in many of the other TK files provided. These functions can be imported and used in new models. See the functions’ listings for documentation.

STUDENT

This file is blank except for the format sheet from FORMATS and the units sheet from UNITMAST. It is intended to be used by the student as a starter file for a new model to which rules, functions, variables, etc., can be added. Starting each model with this file eliminates the need to merge the UNITMAST or FORMATS files into your models and provides their advantages with minimal effort. Be sure to save the file with a new name to avoid overwriting the master file STUDENT each time it is used. Use save as from the file menu to provide a new filename.

UNITMAST

This file contains only a units sheet that can be added to (merged into) any other model without disturbing its other contents. The units sheet enables units conversion of variables.

EXECUTABLE FILES (PROGRAMS) Path: Program Files \ (Programname) DYNACAM

Calculates kinematics and dynamics of cam-follower systems. Data files mentioned in the text (SPRAY.CAM, CASE10A.CAM) are in the folder PROGRAM FILES \ DYNACAM.

1087

1088

MACHINE DESIGN



An Integrated Approach

-

Sixth Edition

LINKAGES

Finds position, velocity, acceleration, forces, and torques of any fourbar, fivebar, sixbar, or slider linkage. Calculates kinematics and dynamics of any single- or multicylinder internal-combustion engine (or compressor) of inline, vee, opposed, or W configuration.

MATRIX

Solves any linear system of up to 16 equations in 16 unknowns. Files CASE1A.MTR, CASE2A.MTR, CASE3A.MTR, and CASE4A.MTR are also included in the folder PROGRAM FILES \ MATRIX.

MOHR

Computes the cubic stress function and plots the Mohr’s circles for any 2-D or 3-D stress state. Files EX04-01.MOH, EX04-02.MOH, and EX04-03.MOH are also included in the folder PROGRAM FILES \ MOHR.

CAD MODEL FILES Path: CAD Model Files \ Problem Files \ Figure_No These files provide Solidworks 2008 CAD models of figures for various problems. If the Solidworks program is not available, a free viewer for these files, eDrawings, can be downloaded from: http://www.edrawingsviewer.com Fig_P03‑01 FIG_P03‑02 FIG_P03‑03 FIG_P03‑16 FIG_P03‑17 FIG_P03‑18 FIG_P04‑01 FIG_P04‑02 FIG_P04‑03 FIG_P04‑12 FIG_P04‑14 FIG_P04‑15 FIG_P04‑16 FIG_P04‑20 FIG_P04‑21 FIG_P04‑22 FIG_P04‑26B FIG_P04‑26C FIG_P04‑26D FIG_P04‑26E FIG_P05‑01 FIG_P05‑02 FIG_P05‑03 FIG_P05‑14 FIG_P05‑15 FIG_P05‑16 FIG_P05‑20 FIG_P05‑24B FIG_P05‑24C FIG_P05‑24D FIG_P05‑24E FIG_P06‑01 FIG_P06‑02 FIG_P06‑03 FIG_P06‑14

For Problem 3-3 For Problems 3-4, 3-5, 3-6 For Problem 3-9 For Problems 3-36 and 3-37 For Problems 3-38 and 3-39 For Problems 3-40 and 3-41 For Problem 4-3 For Problems 4-4, 4-5, 4-6 For Problem 4-9 For Problem 4-27 For Problems 4-33 to 4-36 For Problem 4-37 For Problem 4-40 For Problems 4-57 and 4-58 For Problem 4-59 For Problems 4-60 and 4-61 For Problem 4-69 For Problem 4-70 For Problem 4-71 For Problem 4-72 For Problem 5-3 For Problems 5-4, 5-5, 5-6 For Problem 5-9 For Problems 5-33 to 5-36 For Problem 5-37 For Problem 5-40 For Problems 5-58 and 5-59 For Problem 5-69 For Problem 5-70 For Problem 5-71 For Problem 5-72 For Problem 6-3 For Problem 6-6 For Problem 6-9 For Problems 6-33 to 6-36

INDEX OF SOFTWARE ON BOOK’S WEBSITE

FIG_P06‑15 FIG_P06‑16 FIG_P07‑01 FIG_P07‑02 FIG_P07‑03 FIG_P10‑02 FIG_P10‑03 FIG_P10‑04 FIG_P10‑05 FIG_P10‑06 FIG_P15‑01 FIG_P15‑02

For Problem 6-37 For Problem 6-40 For Problem 7-3 For Problems 7-4, 7-5, 7-6 For Problem 7-9 For Problems 10-2, 10-5, and 10-16 For Problems 10-6, 10-9, 10-11, and 10-12 For Problems 10-7, 10-10, and 10-14 For Problems 10-17 and 10-18 For Problem 10-19 For Problems 15-4 to 15-6 For Problems 15-23 to 15-25

FEA MODEL FILES Path: FEA Model Files \ Case Study Models \ Case_No These files provide Solidworks 2008 CAD and FEA models of various Case Studies. CASE STUDY 1 CASE STUDY 2 CASE STUDY 4 CASE STUDY 7

Bicycle Brake Lever Crimping Tool Bicycle Brake Arm Trailer Hitch

DERIVATIONS OF EQUATIONS Path: Derivations Fourbar Acceleration Derivation.pdf Fourbar Position Derivation.pdf Fourbar Velocity Derivation.pdf Slider Acceleration Derivation.pdf Slider Velocity Derivation.pdf

REFERENCES 1

R. L. Norton, Design of Machinery, 6ed., McGraw-Hill, New York, 2020.

1089

THE END