Вы находитесь на странице: 1из 432

VIBRATION ANALYSIS FOR ELECTRONIC EQUIPMENT

THIRD EDITION

Dave S. Steinberg

Steinberg & Associates and University of California, Los Angeles

A WILEY-INTERSCIENCE PUBLICATION

JOHN WILEY & SONS, INC.

New York

A WILEY-INTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York Chichester Weinheim Brisbane Singapore Toronto

Chichester

A WILEY-INTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York Chichester Weinheim Brisbane Singapore Toronto

Weinheim

A WILEY-INTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York Chichester Weinheim Brisbane Singapore Toronto

Brisbane

A WILEY-INTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York Chichester Weinheim Brisbane Singapore Toronto

Singapore

A WILEY-INTERSCIENCE PUBLICATION JOHN WILEY & SONS, INC. New York Chichester Weinheim Brisbane Singapore Toronto

Toronto

This book is printed on acid-free paper. @

Copyright G2000 by John Wiley & Sons, Inc. All rights reserved.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4744. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New

10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: PERMREQ@WILEY.COM.

York, NY

Library of Congress Cataloging in Publication Data:

Steinberg, Dave S. Vibration analysis for electronic equipment/Dave

S. Steinberg.--3rd ed.

p. cm “A Wiley-Interscience publication.” ISBN 0-471-37685-X 1. Electronic apparatus and appliances-Vibration. I. Title.

TK7870S8218 2000

621.381--dc21

99-056617

Printed in the United States of America

109876

To my wife, Annette, and to my two daughters, Con and Stacie

Electronic equipment continues its relentless expansion into virtually every area associated with commercial, industrial, and military applications throughout the entire world. Exotic technology has become commonplace in medicine, entertainment, communication, travel, transportation, manufactur- ing, education, and commerce. The internet has dramatically and forever changed the way many companies run their businesses, without the need for large offices and a large staff of people to perform clerical duties, largely due to the extensive use of the fast, small, powerful personal computers. This has resulted in improved profit margins in companies that embraced the new technologies, and disasters for companies that resisted the tidal wave of changes. Extensive changes were made in this third edition to reflect the changes that are taking place in the world of electronics. One of the big changes is the dramatic reduction in the costs of manufacturing the electronic equipment due to improved automation. This has resulted in the use of more electronics with improved performance over mechanical functions in a wide variety of products such as cameras, automobile controls for braking, ignition. air conditioners, transmission shifting, and combustion control, computers, com- puter printers, FAX machines, and televisions. One of the most dramatic changes has occurred in the military area, where high costs have forced the Department of Defense to scrap many of the military specifications in favor of best commercial practice electronic hardware for their new sophisticated equipment. Lower costs often lead to lower quality, which can result in a reduced reliability. This may only be an inconvenience when your television set fails or your automobile will not start. This can be a disaster in a military vehicle if the guns do not fire, or if the navigation system fails, or if the communications system fails. Therefore a new chapter was added that investigates and evaluates the effects of manufacturing methods and toler- ances on the reliability of the electronic hardware. Another new chapter was added to show methods for improving the ruggedness of commercial hard- ware for improved reliability in harsh military environments. An additional chapter was added that relates to a better understanding of the bending distortion of the circuit board at its resonant condition, and how it effects the forces, stresses, and fatigue life in the component lead wires and solder joints. A new chapter was also added in the area of environmental stress screening, to provide a better understanding of how various combinations of

XViii

PREFACE

thermal cycling and vibration can affect the accumulated damage and the amount of fatigue life used up by various screens. New equations are derived that show how to evaluate the effectiveness of a lead wire strain relief to reduce the dynamic forces and stresses in the lead wires and solder joints, and the resulting increase in the fatigue life. A simpler method for evaluating the fatigue life in random vibration is shown, which has the same accuracy as the three-band technique with much less work. A quick way of estimating the effectiveness of stiffening ribs on plug-in PCBs is shown, which is very simple to use and easy to apply. A new added section discusses a quick and easy method for evaluating the first three natural frequencies for flat plates and PCBs that are rectangular, octagonal, triangular, and round, with 16 different edge and point supports for each shape. A new and more accurate method for calculating the expected transmissibilty Q values for beams, circuit boards, and chassis is presented that relates to the natural frequency and the input acceleration G level. The data for this method are based on extensive vibration tests which showed that very low input acceleration G levels produced very low stresses and damping, which resulted in very high Q values. At the opposite end, high input G levels produced high stresses and damping, which resulted in low Q values. Many detailed sample problems are shown throughout this textbook, to demonstrate methods of analysis and evaluation that are based on 40 years of analysis and testing many different types of electronic systems. A section on vibration testing case histories was added to provide additional knowledge and insights for improving the testing results. Field failures in electronic equipment hardware compiled by the U.S. Air Force over a period of about 20 years show that about 40% of these failures are related to connectors, 30% to interconnects, and 20% to component parts. Failures in these areas can be due to handling, vibration, shock, and thermal cycling. Field failures related to operating environments show that about 55% of the failures are due to high temperatures and temperature cycling, 20% of the failures are related to vibration and shock, and 20% are due to humidity. Since high temperatures and temperature cycling events appear to be the major environmental causes of electronic failures, more information in these areas is very important. These subjects are not ad- dressed in this book in great detail. More information is available on high

temperatures and temperature cycling forces, stresses, and fatigue life effects [54]. Details on electronic components, electrical lead wires, solder joints, and plated through-holes for surface mounted and through-hole components, due to differences in the thermal coefficients of expansion (TCE), can be

found in my book entitled

Cooling Techniques for Electronic Equipment,

Second Edition published by John Wiley & Sons.

Westlake Village, California September, 1999

DAVES. STEINBERG

D LIST OF SYMBOLS

A Area (in.2), amplification

g

Acceleration of gravity,

(dimensionless ratio) 386 in./s2

a length (in.) ASIC Application-specific

H

Horizontal force Ob), drop heightr (in. or ft)

integrated circuit h Height (in.), thickness (in.)

Length (in.) Fatigue exponent, width (in.) Dynamic constant Length (in.) Center of Gravity

Distance from neutral axis J to outer fiber, length (in.), K damping coefficient (lb .s/in.)

B

b

C

C

CG

C

Hz

I

Lri

Frequency (cycles/s) Moment of inertia (area)

(in.4)

Mass moment of inertia (Ib in: s2) Torsional form factor (in.4> Linear spring rate (Ib in.), stiffness ratio (dimensionless), buckling form factor dimensionless) Kinetic energy Theoretical stress- concentration factor (dimensionless) Angular spring rate (Ib .in./rad) Length (in.) Leadless ceramic chip carrier Bending moment (lb .in.) Total moment (lb in.) Bending moment per unit length along the X axis (in: lb/inJ, bending moment at point X(lb. in.) Momentum (lb .s) Margin of safety (dimensionless) mass (lb s2/in.) Number of fatigue cycles to fail

cc

Critical damping (lb .s/in.)

D

Dissipation energy

dB

(lb .in./s) diameter (in.), plate bending stiffness factor (lb .in.) Decibel

D,,

Plate torsional stiffness factor (Ib * in.)

KIN

Kt

KO

L

d Diameter (in.), length (in.)

DIP Dual inline package LCCC

E Modulus of elasticity

(lb/in.2)

A4

e Bolted efficiency factor (%) MT

ESS

F Force (lb)

Environmental stress screen M,

f

f"

fr

Frequency (Hz) Natural frequency (Hz) Rotational natural Mo

MS

frequency (Hz)

FEM

Finite element method

G

Shear modulus (lb/in.2) rn with subscript: acceleration N in gravity units (dimensionless)

XX

LIST OF SYMBOLS

N,+

Number of positive zero

wd

crossings, (Hz)

W

n

Actual number of fatigue

X

cycles

X

NS Number of sweeps through X

 

a resonance

P

Force (lb)

x

z

P. PSD Power spectral density

Dynamic load (lb) Unit load (lb/in.) Displacement along X axis, Xi,E;, Zcoordinate axes First derivative, velocity Second derivative, acceleration Displacement (in.)

(G*/HZ)

PCB

Printed circuit board

d'

Dynamic force

P

Unit load (lb/in.)

Q

Transmissibility (dimensionless ratio)

4

Shear flow (lb/in.), dynamic pressure (lb/in.)

R

Radius (in.), reaction (lb) Stress ratio (dimensionless), sweep rate (octave/min)

Rn

Fatigue-cycle ratio (dimensionless)

RC

Damping ratio (dimensionless)

R,

Frequency ratio (dimensionless)

RMS

Root mean square

RSS

Root sum square

r

radius (in.),

sb

relative position factor Bending stress (lb/in.2>

se

Endurance limit stress (lb/in.*)

S

Stress (lb/in.*)

SMD

Surface mounted device

TCE

Thermal coefficient of ex- pansion (in./in./'C)

T

Kinetic energy Ob-in.), torque (in/. lb)

t

Time (min), temperature, Thickness (in.)

U

Strain energy (lb in.), work (lb in.)

V

Velocity (in./s>, force (lb)

W

Weight (lb)

CY

Greek Symbols

Angle (degrees, radians), thermal coefficient of expansion (in./in./"C)

s Displacement (in.)

0 Angular displacement (radians)

P Poisson's ratio (dimensionless)

P Density (lb/in.), mass per unit area (lb * in.^)

4 Phase angle (degrees)

A Difference

R

a"

Angular velocity (rad/s) Natural frequency (rad/s)

Subscripts

av

Average

b

Bending

C

Critical

d

Dynamic, desired

e

Endurance Equivalent

Input

eq

in

max

Maximum

n

Natural

0

Maximum

out

Output or response

st

Static

su

Shear ultimate

t

Tension

tu

Tensile ultimate

tY

Tensile yield

U

Ultimate

Y

Yield

ST

Shear tearout

S

Shear

T

Torsion

- CONTENTS

Preface

 

xvii

List of Symbols

xix

1. Introduction

1

 

1.1 Vibration Sources

1

1.2 Definitions

2

1.3 Vibration Representation

3

1.4 Degrees of Freedom

3

1.5 Vibration Modes

5

1.6 Vibration Nodes

5

1.7 Coupled Modes

6

1.8

Fasteners

7

1.9

Electronic Equipment for Airplanes and Missiles

10

1.10

Electronic Equipment for Ships and Submarines

13

1.11

Electronic Equipment for Automobiles, Trucks, and Trains

15

1.12

Electronics for Oil Drilling Equipment

16

1.13

Electronics for Computers, Communication, and Entertainment

16

2.

Vibrations of Simple Electronic Systems

17

2.1 Single Spring-Mass System Without Damping

17

Sample Problem-Natural

Beam

Frequency of a Cantilever

19

2.2 Single-Degree-of-Freedom Torsional Systems

21

Sample Problem-Natural Frequency of a Torsion System

22

2.3 Springs in Series and Parallel

23

Sample Problem-Resonant Frequency of a Spring System

25

2.4 Relation of Frequency and Acceleration to Displacement

26

Sample Problem-Natural Frequency and Stress in a Beam

27

2.5 Forced Vibrations with Viscous Damping

30

2.6 Transmissibility as a Function of Frequency

34

Sample Problem-Relating the Resonant Frequency to the Dynamic Displacement

34

Viii

CONTENTS

 

2.7

Multiple Spring-Mass

Systems Without Damping

36

 

Sample Problem-Resonant

Frequency of a System

37

3

Component Lead Wire and Solder Joint Vibration Fatigue Life

39

3.1 Introduction

 

39

3.2 Vibration Problems with Components Mounted High Above the PCB

39

Sample Problem-Vibration Fatigue Life in the Wires of a TO-5 Transistor

40

3.3 Vibration Fatigue Life in Solder Joints of a TO-5 Transistor

43

3.4 Recommendations to Fix the Wire Vibration Problem

45

3.5 Dynamic Forces Developed in Transformer Wires During Vibration

46

Sample Problem-Dynamic Transformer Lead Wires

Forces and Fatigue Life in

46

3.6 Relative Displacements Between PCB and Component Produce Lead Wire Strain

49

Sample Problem-Effects Reliability

of PCB Displacement on Hybrid

50

4. Beam Structures for Electronic Subassemblies

56

4.1 Natural Frequency of a Uniform Beam

56

Sample Problem-Natural

Frequencies of Beams

60

4.2 Nonuniform Cross Sections

64

Sample Problem-Natural Nonuniform Sections

Frequency of a Box with

68

4.3 Composite Beams

 

69

5. Component Lead Wires as Bents, Frames, and Arcs

75

5.1

Electronic Components Mounted on Circuit Boards

75

5.2

Bent with a Lateral Load-Hinged

Ends

77

5.3

Strain Energy-Bent

with Hinged Ends

80

5.4

Strain Energy-Bent

with Fixed Ends

83

5.5

Strain Energy-Circular

Arc with Hinged Ends

90

5.6 Strain Energy-Circular

Arc with Fixed Ends

92

5.7 Strain Energy-Circular

Arcs for Lead Wire Strain

 

Relief

94

Sample Problem-Adding an Offset in a Wire to Increase the Fatigue Life

97

CONTENTS

ix

6

Printed Circuit Boards and Flat Plates

103

6.1 Various Types of Printed Circuit Boards

103

6.2 Changes in Circuit Board Edge Conditions

106

6.3 Estimating the Transmissibility of a Printed Circuit Board

108

6.4 Natural Frequency Using a Trigonometric Series

111

6.5 Natural Frequency Using a Polynomial Series

116

Sample Problem-Resonant Frequency of a PCB

120

6.6 Natural Frequency Equations Derived Using the Rayleigh Method

122

6.7 Dynamic Stresses in the Circuit Board

127

Sample Problem-Vibration Stresses in a PCB

131

6.8 Ribs on Printed Circuit Boards

132

6.9 Ribs Fastened to Circuit Boards with Screws

137

6.10 Printed Circuit Boards With Ribs in Two Directions

141

6.11 Proper Use of Ribs to Stiffen Plates and Circuit Boards

141

6.12 Quick Way to Estimate the Required Rib Spacing for Circuit Boards

142

6.13 Natural Frequencies for Different PCB Shapes with Different Supports

144

 

Sample Problem-Natural with Three Point Supports

Frequency of a Triangular PCB

149

7. Octave Rule, Snubbing, and Damping to Increase the PCB Fatigue Life

150

 

7.1 Dynamic Coupling Between the PCBs and Their Support Structures

150

7.2 Effects of Loose Edge Guides on Plug-in Type PCBs

154

7.3 Description of Dynamic Computer Study for the Octave Rule

154

7.4 The Forward Octave Rule Always Works

155

7.5 The Reverse Octave Rule Must Have Lightweight PCBs

155

Sample Problem-Vibration Problems with Relays Mounted on PCBs

156

7.6 Proposed Corrective Action for Relays

157

7.7 Using Snubbers to Reduce PCB Displacements and Stresses

159

 

Sample Problem-Adding

Snubbers to Improve PCB

Reliability

161

 

7.8 Controlling the PCB Transmissibility with Damping

162

7.9

Properties of Material Damping

162

7.10 Constrained Layer Damping with Viscoelastic Materials

163

7.11 Why Stiffening Ribs on PCBs are Often Better than Damping

164

X

CONTENTS

8. Preventing Sinusoidal Vibration Failures in Electronic Equipment

166

8.1 Introduction

166

8.2 Estimating the Vibration Fatigue Life

167

Sample Problem-Qualification System

Test for an Electronic

168

8.3 Electronic Component Lead Wire Strain Relief

169

8.4 Designing PCBs for Sinusoidal Vibration Environments

171

Sample Problem-Determining Frequency

Desired PCB Resonant

174

8.5 How Location and Orientation of Component on PCB Affect Life

175

8.6 How Wedge Clamps Affect the PCB Resonant Frequency

177

Sample Problem-Resonant

Frequency of PCB with Side

Wedge Clamps

179

8.7 Effects of Loose PCB Side Edge Guides

182

Sample Problem-Resonant Edge Guides

Frequency of PCB with Loose

185

8.8 Sine Sweep Through a Resonance

185

Sample Problem-Fatigue Sine Sweep

Cycles Accumulated During a

187

9. Designing Electronics for Random Vibration

188

9.1 Introduction

188

9.2 Basic Failure Modes in Random Vibration

188

9.3 Characteristics of Random Vibration

189

9.4 Differences Between Sinusoidal and Random Vibrations

190

9.5 Random Vibration Input Curves

192

Sample Problem-Determining

the Input RMS

Acceleration Level

193

9.6 Random Vibration Units

193

9.7 Shaped Random Vibration Input Curves

194

Sample Problem-Input

RMS Accelerations for Sloped

PSD Curves

195

9.8 Relation Between Decibels and Slope

197

9.9 Integration Method for Obtaining the Area Under a PSD Curve

198

9.10 Finding Points on the PSD Curve

200

Sample Problem-Finding PSD Values

200

9.11 Using Basic Logarithms to Find Points on the PSD Curve

201

CONTENTS

xi

9.13 Gaussian or Normal Distribution Curve

202

9.14 Correlating Random Vibration Failures Using the Three-Band Technique

204

9.15 Rayleigh Distribution Function

205

9.16 Response of a Single-Degree-of-Freedom System to Random Vibration

206

Sample Problem-Estimating

the Random Vibration

Fatigue Life

208

9.17 How PCBs Respond to Random Vibration

214

9.18 Designing PCBs for Random Vibration Environments

215

Sample Problem-Finding

the Desired PCB Resonant

Frequency

218

9.19 Effects of Relative Motion on Component Fatigue Life

220

Sample Problem-Component Fatigue Life

221

9.20 It’s the Input PSD that Counts, Not the Input RMS Acceleration

222

9.21 Connector Wear and Surface Fretting Corrosion

223

Sample Problem-Determining

Approximate Connector

Fatigue Life

224

9.22 Multiple-Degree-of-Freedom Systems

224

9.23 Octave Rule for Random Vibration

225

Sample Problem-Response

of Chassis and PCB to

Random Vibration

226

Sample Problem-Dynamic

Analysis of an Electronic

Chassis

229

9.24 Determining the Number of Positive Zero Crossings

231

Sample Problem-Determining Zero Crossings

the Number of Positive

233

10 Acoustic Noise Effects on Electronics

234

10.1 Introduction

234

Sample Problem-Determining

the Sound Pressure Level

234

10.2 Microphonic Effects in Electronic Equipment

235

10.3 Methods for Generating Acoustic Noise Tests

236

10.4 One-Third Octave Bandwidth

238

10.5 Determining the Sound Pressure Spectral Density

238

10.6 Sound Pressure Response to Acoustic Noise Excitation

239

Sample Problem-Fatigue Life of a Sheet-Metal Panel Exposed to Acoustic Noise

240

10.7 Determining the Sound Acceleration Spectral Density

245

Sample Problem-Alternate Analysis

Method of Acoustic Noise

246

Xii

CONTENTS

11. Designing Electronics for Shock Environments

248

11.1

Introduction

248

11.2

Specifying the Shock Environment

249

11.3

Pulse Shock

25 1

11.4

Half-Sine Shock Pulse for Zero Rebound and Full

Rebound

252

Sample Problem-Half-Sine Shock-Pulse Drop Test

253

11.5

Response of Electronic Structures to Shock Pulses

257

11.6

Response of a Simple System to Various Shock Pulses

258

11.7

How PCBs Respond to Shock Pulses

260

11.8

Determining the Desired PCB Resonant Frequency for Shock

260

Sample Problem-Response Shock Pulse

of a PCB to a Half-Sine

262

11.9

Response of PCB to Other Shock Pulses

264

Sample Problem-Shock

Response of a Transformer

Mounting Bracket

265

11.10

Equivalent Shock Pulse

269

Sample Problem-Shipping Crate for an Electronic Box

269

11.11

Low Values of the Frequency Ratio R

274

Sample Problem-Shock Amplification for Low Frequency Ratio R

274

11.12

Shock Isolators

275

Sample Problem-Heat Developed in an Isolator

276

11.13

Information Required for Shock Isolators

277

Sample Problem-Selecting a Set of Shock Isolators

278

11.14

Ringing Effects in Systems with Light Damping

281

11.1s

How Two-Degree-of-Freedom Systems Respond to Shock

282

11.16

The Octave Rule for Shock

284

11.17

Velocity Shock

285

Sample Problem-Designing Shock

a Cabinet for Velocity

285

11.18

Nonlinear Velocity Shock Sample Problem-Cushioning

Material for a Sensitive

286

Electronic Box

288

11.19

Shock Response Spectrum

288

11.20

How Chassis and PCBs Respond to Shock

291

Sample Problem-Shock Response Spectrum Analysis for Chassis and PCB

292

11.21

How Pyrotechnic Shock Can Affect Electronic Components

296

Sample Problem-Resonant Bond Wire

Frequency of a Hybrid Die

298

CONTENTS

Xiii

12. Design and Analysis of Electronic Boxes

300

12.1 Introduction

300

12.2 Different Types of Mounts

300

12.3 Preliminary Dynamic Analysis

303

12.4 Bolted Covers

305

12.5 Coupled Modes

308

12.6 Dynamic Loads in a Chassis

311

12.7 Bending Stresses in the Chassis

316

12.8 Buckling Stress Ratio for Bending

318

12.9 Torsional Stresses in the Chassis

320

12.10 Buckling Stress Ratio for Shear

324

12.11 Margin of Safety for Buckling

325

12.12 Center-of-Gravity Mount

326

12.13 Simpler Method for Obtaining Dynamic Forces and Stresses on a Chassis

328

13. Effects of Manufacturing Methods on the Reliability of Electronics

330

13.1 Introduction

330

13.2 Typical Tolerances in Electronic Components and Lead Wires

331

Sample Problem-Effects of PCB Tolerances on Frequency and Fatigue Life

332

13.3 Problems Associated with Tolerances on PCB Thickness

333

13.4 Effects of Poor Bonding Methods on Structural Stiffness

334

13.5 Soldering Small Axial Leaded Components on Through- Hole PCBs

335

13.6 Areas Where Poor Manufacturing Methods Have Been Known to Cause Problems

336

13.7 Avionic Integrity Program and Automotive Integrity Program (AVIP)

338

13.8 The Basic Philosophy for Performing an AVIP Analysis

340

13.9 Different Perspectives of Reliability

343

14. Vibration Fixtures and Vibration Testing

346

14.1 Vibration Simulation Equipment

346

14.2 Mounting the Vibration Machine

347

14.3 Vibration Test Fixtures

347

14.4 Basic Fixture Design Considerations

348

14.5 Effective Spring Rates for Bolts

350

XiV

CONTENTS

 

Sample Problem-Determining Desired Bolt Torque

353

14.7 Rocking Modes and Overturning Moments

353

14.8 Oil-Film Slider Tables

355

14.9 Vibration Fixture Counterweights

356

14.10 A Summary for Good Fixture Design

357

14.11 Suspension Systems

357

14.12 Mechanical Fuses

358

14.13 Distinguishing Bending Modes from Rocking Modes

359

14.14 Push-Bar Couplings

360

14.15 Slider Plate Longitudinal Resonance

364

14.16 Acceleration Force Capability of Shaker

365

14.17 Positioning the Servo-Control Accelerometer

366

14.18 More Accurate Method for Estimating the Transmissibility Q in Structures

367

Sample Problem-Transmissibility

Expected for a Plug-in

PCB

368

Vibration Testing Case Histories

369

14.19 Cross-Coupling Effects in Vibration Test Fixtures

369

14.20 Progressive Vibration Shear Failures in Bolted Structures

370

14.21 Vibration Push-Bar Couplers with Bolts Loaded in Shear

371

14.22 Bolting PCB Centers Together to Improve Their Vibration Fatigue Life

373

14.23 Vibration Failures Caused by Careless Manufacturing Methods

375

14.24 Alleged Vibration Failure that was Really Caused by Dropping a Large Chassis

376

14.25 Methods for Increasing the Vibration and Shock Capability on Existing Systems

377

15.

Environmental Stress Screening for Electronic Equipment (ESSEE)

379

15.1 Introduction

379

15.2 Environmental Stress Screening Philosophy

379

15.3 Screening Environments

381

15.4 Things an Acceptable Screen Are Expected to Do

383

15.5 Things an Acceptable Screen Are Not Expected to Do

383

15.6 To Screen or Not to Screen, That is the Problem

384

15.7 Preparations Prior to the Start of a Screening Program

384

15.8 Combined Thermal Cycling, Random Vibration, and Electrical Operation

387

15.9 Separate Thermal Cycling, Random Vibration, and Electrical Operation

389

CONTENTS

XV

15.10 Importance of the Screening Environment Sequence

389

15.11 How Damage Can Be Developed in a Thermal Cycling Screen

390

15.12 Estimating the Amount of Fatigue Life Used Up in a Random Vibration Screen

392

Sample Problem-Fatigue

Life Used Up in Vibration

and Thermal Cycling Screen

395

Bibliography

401

Index

405

- CHAPTER 1

Introduction

1.I

VIBRATION SOURCES

Electronic equipment can be subjected to many different forms of vibration over wide frequency ranges and acceleration levels. It is probably safe to say that all electronic equipment will be subjected to some type of vibration at some time in its life. If the vibration is not due to an active association with some sort of a machine or a moving vehicle, then it may be due to transporting the equipment from the manufacturer to the customer. Vibra- tions encountered during transportation and handling can produce many different types of failures in electronic equipment unless the proper consider- ation are given to the mechanical design of the electronic structure and the shipping containers. Vibration is usually considered to be an undesirable condition, and in most cases it is. However, there are many applications where vibration is deliberately imposed to improve a function. Some sophisticated applications of vibrations are in the use of ultrasonics to clean medical instruments, measure wall thickness, and find flaws in castings. Vibration is also used in sorting rocks of different sizes by passing them over vibrating screens, which have groups of graduated holes. When the first jet airplanes were introduced, standard instruments from airplanes with piston engines and propellers were used. Theres instruments had a tendency to stick when they were used on the jet airplanes. The difference was due to the lower-frequency vibration levels developed in the piston engines. In order to make these same instruments work in the first jet airplanes, small vibrators had to be mounted on the instrument panels. Mechanical vibrations can have many different sources. In household products such as blenders and washing machines, the vibrations are due to the unbalance created by rotating and tumbling masses. In vehicles such as automobiles, trucks, and trains most of the vibration is due to rough surfaces over which these vehicles travel. In ships and submarines the vibration is due to the engines and to buffeting by the water. In airplanes, missiles, and rockets the vibration is due to jet and rocket engines and to aerodynamic buffeting. Most of the vibration in a missile, during subsonic flight, is due to

2

INTRODUCTION

+ Displacement

~

Period

- Displacement

FIGURE 1.1. Simple harmonic motion.

the sound field developed by the rocket engines [1,6].* This is due to the extreme turbulence of the jet exhaust downstream from the rocket engine.

1.2

DEFINITIONS

Vibration, in a broad sense, is taken to mean an oscillating motion where some structure or body moves back and forth. If the motion repeats itself, with all of the individual characteristics, after a certain period of time, it is called periodic motion. This motion can be quite complex, but as long as it repeats itself, it is still periodic. If continuous motion never seems to repeat itself, it is called random motion. Simple harmonic motion is the simplest form of periodic motion, and it is usually represented by a continuous sine wave on a plot of displacement versus time, as shown in Fig. 1.1. The reciprocal of the period is known as the frequency of the vibration and is measured in cycles per second, or hertz (Hz), in honor of the German who first experimented with radiowaves. The maximum displacement is called the amplitude of the vibration.

condition where the equilibrium

of a system is disrupted by a sudden applied force or increment of force, or by a sudden change in the direction or magnitude of a velocity vector. Shock is not transmitted easily in the light airframe structures generally used in airplanes and missiles. Impact forces often result in a transient type of vibration, which is influenced by the natural frequencies of the airframe. Only steady-state linear vibrations are considered in this book. Linear vibrations occur with linear elastic media where the displacements are directly proportional to the applied force. If the force is doubled, the displacement is doubled. Only stresses up to the elastic limit of any given material are considered. Higher stresses generally result in permanent defor- mations, which extend into the plastic deformation region and are not

Crede [2,3] defines shock as a transient

considered here.

*Numbers in brackets [ ] refer to bibliography entries in the back of the book.

DEGREES OF FREEDOM

3

FIGURE 1.2. Rotating vector simulating a single-degree-of-freedom system.

1.3 VIBRATION REPRESENTATION

A rotating vector can be used to describe the simple harmonic vibration of a single mass suspended on a coil spring (Fig. 1.2). The vector Yo rotates counterclockwise with a uniform angular velocity of R rad/s. The projection of the vector on the vertical axis represents the instantaneous displacement Y of the mass as it vibrates up and down. This can be written as

When the vector Yo rotates through one revolution, it rotates through an angle of 360°, which is 277- radians, for one complete cycle. The angular velocity is measured in radians per second, and the frequency f is measured in cycles per second. This leads to the relation

1.4 DEGREES OF FREEDOM

A vibrating system requires some coordinates to describe the positions of the elements in the system. If there is only one element in the system that is restricted to move along only one axis, and only one dimension is required to locate the position of the element at any instant of time with respect to some initial starting point, then it is a single-degree-of-freedom system. The same is true for a torsional system. If there is only one element that is restricted to rotate about only one axis so that only one dimension is required to locate the position of the element at any instant of time with respect to some initial starting point, it is a single-degree-of-freedom system. Some samples of systems with a single degree of freedom are shown in Fig. 1.3.

4

INTRODUCTION

Spring and mass

(a)

Rod and disk

(b)

Pendulum

(Ci

FIGURE 1.3. Single-degree-of-freedomsystems.

X+

(a) f 6)

F

Y

(C)

(di

FIGURE 1.4. Two-degree-of-freedomsystems.

A two-degree-of-freedom system requires two coordinates to describe the positions of the elements. Some samples of systems with two degrees of freedom are shown in Fig. 1.4. In rigid-body mechanics, a single body can have six degrees of freedom. These are translation along each of its three mutually perpendicular X, Y, and Z axes, and rotation about each of them, as shown in Fig. 1.5. A typical beam can have an infinite number of degrees of freedom since it can bend in an infinite number of shapes, or modes, as shown in Fig. 1.6.

FIGURE 1.5. A single mass with six degrees

of freedom.

+Z

-Y

VIBRATION NODES

5

-----

FIGURE 1.6. A simply supported beam showing

several degrees of freedom.

k

,/-------- Node point

.

First harmonic mode

(a)

‘.

--

Second harmonic mode

(bi

FIGURE 1.7. First and second harmonic modes for a simply supported beam.

1.5 VIBRATION MODES

A standard manner in which a particular system can vibrate is known as a vibration mode. Each vibration mode is associated with a particular natural frequency and represents a degree of freedom. A single-degree-of-freedom system will have only one vibration mode and only one resonant frequency. The six-degree-of-freedom system shown in Fig. 1.5 can have six vibration modes and six resonant frequencies. The simply supported beam shown in Fig. 1.6 can have an infinite number of vibration modes; this is the same as saying this beam can have an infinite number of different shapes for each of its resonances, which are also infinite. The fundamental resonant mode of a vibrating system is usually called the natural frequency or the resonant frequency of the system. Sometimes it is called the first harmonic mode of the system. For example, a simply sup- ported uniform beam vibrating at its fundamental resonant frequency has the mode shape of a half sine wave as shown in Fig. 1.7~.When this beam is vibrating at its second natural frequency, in its second harmonic mode, it has the mode shape of the full sine wave shown in Fig. 1.7b. The first harmonic mode of a system, with the lowest natural frequency, is the fundamental resonant mode; this often has the greatest displacement amplitudes and usually the greatest stresses. The second harmonic mode, or second resonance, usually has a smaller displacement than the first harmonic mode, so the stresses are usually smaller. The displacements continue to decrease for the higher resonant modes.

1.6 VIBRATION NODES

Vibration nodes are unsupported points on a vibrating body that have zero displacements. Nodes are generally associated with bending or torsion modes. At the first harmonic bending resonant mode in a beam there are no node

6

INTRODUCTION

First

Second

Third

Fourth

mode

mode

mode

mode

FIGURE 1.8. First four harmonic modes of a circular membrane.

First

Second

Third

mode

mode

mode

FIGURE 1.9. First three harmonic modes of a square plate.

points. At the second harmonic mode there is one node point, the third harmonic mode has two node points, and so on. Figure 1.7~shows the bending mode of a vibrating beam with no node points, and Fig. 1.7b shows a beam with one node point. Vibrating plates can have straight-line bending nodes and circular nodes. The first four harmonic modes of a circular membrane are shown in Fig. 1.8. The plus (+) signs show positive displacements and the minus (-> signs show negative displacements. The dashed lines represent nodal locations of zero displacement. The first three harmonic bending modes of a square plate with free edges are shown in Fig. 1.9.

1.7 COUPLED MODES

In a system with two or more degrees of freedom, the vibration mode of one degree often influences the vibration mode of the other degree. For example, in Fig. 1.4~if mass 2 is held rigidly while mass 1 is displaced in the vertical direction, mass 1 will oscillate up and down. Now if mass 2 is released, the motion of mass 1 will act upon mass 2 so mass 2 will begin to oscillate up and down. Because the motion of mass 1 has a direct effect on the motion of mass 2, these two vibration modes are defined as being coupled. In coupled modes, the vibration in one mode cannot occur independently of the vibra- tion in the other mode. Coupled modes can occur in translation, rotation, and combinations of translation and rotation for systems with more than one degree of freedom. For coupled modes in translation and rotation, it is often possible to

FASTENERS

7

L

3

Y

A

I

I

1

I

r

>

L+L+

FIGURE 1.10. A single mass with two springs.

determine whether the system is coupled or uncoupled by making a simple test. Apply a steady load to the body of the system at its center of gravity (CG), in a specific direction. If the body moves in the direction of the applied load without rotation, then the translational mode is not coupled with the rotational mode for motion along the direction of the applied load. Consider, for example, the mass with two springs as shown in Fig. 1.10. If

a steady load is applied to the CG along the X axis, the mass will not only translate along the X axis, but it will also tend to rotate at the same time. This test shows that for vibration along the X axis, the translational mode is coupled with the rotational mode. If a steady load is applied to the CG along the Y axis, the mass will translate along the Y axis without rotation. Also, if a moment is applied to the mass about the CG, then rotation will take place without translation. These tests show that for vibration along the Y axis the translational mode is decoupled from the rotational mode.

1.8

FASTENERS

Many different types of fasteners are used in electronic equipment. These include screws, nuts, rivets, and clips. Fasteners are responsible for a very large percentage of field failures and, in most shock tests, they are the largest single source of failure. Although fasteners have been used in very large quantities in electronic equipment for many years, most of the applications are based on static installation considerations. The fastening techniques, which are generally chosen for ease of installation and low cost, are usually not satisfactory for severe shock and vibration environments. A contributing factor is that fasteners are such small items and their use so universal that their application tends to be semiautomatic without regard to their strength. This is particularly the case for machine screws. When large screws are used,

it is often because someone has had some association with the automotive or

aircraft industry, where very small screws are seldom used.

8

INTRODUCTION

Because fasteners play an important part in the overall reliability of the electronic equipment, extra consideration should be given in their selection as follows [5]:

1. Select the proper type of fastener (screws, rivets, etc.), considering such trade-offs as environment, strength, maintenance, and cost.

2. Select the correct fastener size and location based on dynamic loads and geometry of the structure.

3. Select the correct locking device for screws and nuts.

4. Select the right installation technique.

Most electronics manufacturers choose screws and rivets on a production basis. If the designer uses screws, the size and locking device are selected according to the tolerances as well as for ease in assembly. In installation, the designer depends on production personnel to use their own judgment in the proper installation of fasteners. The results may be that the wrong fasteners are used, the wrong sizes are used, the wrong lockwashers are used, and the wrong installation torques are used. It has been found, generally speaking, that cold-driven shank-expanding rivets are very satisfactory and should be used more frequently in electronic assemblies. It has been found, too, that when screws are used, they should be of larger sizes than those customarily considered and they should be made of better materials. Many locking devices commonly used are unsatisfactory. Some are often the source of many severe problems. It has also been found that a mechanic’s judgment in tightening a machine screw is usually faulty. Many failures in electronic equipment have resulted because bolts have come loose. Consider what can happen to a sensitive electronic chassis when

a large transformer comes loose and rattles around during vibration. Although many investigations have been made with high-speed films and strain gages to study the loosening action of screws and nuts, the mechanism by which this occurs is still not well established. It appears, however, that the bolt stretches slightly under the action of a dynamic load, so that the interface friction forces in the area of the bolt are suddenly and sharply reduced. Since the threads in the screw or nut tend to return to their original shape during the small time increments, the small geometric changes become

a driving force, which tends to loosen the screw and nut. Some specific recommendations can be made to improve the quality of the fasteners in electronic equipment:

1. Steel screws should be used in all screw fastenings. The steel should have the minimum properties of SAE 1010.

FASTENERS

9

TABLE 1.1

~~

Screw

Torque

Screw

Torque

Size

(in:

lb)

Size

(in:

lb)

2-64

3-3.5

12-24

45-56

4-40

5-6.5

12-28

50-64

6-32

10-12

$20

65-80

8-32

20-24

-28

85-100

10-24

22-27

i-16

250-320

10-32

34-42

f -24

330-415

3.

The tightening torque should be 6040% of the torque required to

twist the head off the screw. These torque values for steel screws are

given in Table

1.1 [lll.

4.

The screw head should permit a positive nonslipping grip for the driving device and should also withstand the driving forces. A slotted- hex-head machine screw seems to be, for general purposes, the most satisfactory.

5. In all applications involving through-holes, locknuts should be used instead of lockwashers. Most of the standard steel locknuts are satisfac- tory.

6. Blind-tapped holes should be avoided, if possible. When they are necessary, locking devices such as lockwashers should be used under the screw head in order to prevent the screw from backing out during vibration. The tightening torque should be increased by an amount equal to the resisting frictional torque of the locking device.

7. The fasteners for a unit should be distributed so that the failure of one fastener will not free the unit or cause it to malfunction. Even for very small components there should be a minimum of two fasteners.

Lockwashers and screw-locking inserts should be used with care in elec- tronic systems that will be used in the zero-G environment of outer space. These two devices create a binding friction in the screw by biting into the metal during installation. This action will very often shave metal particles from the screw. If these particles are not removed, they can float around in a zero-G environment and create electrical problems. An actual count was made of the metal particles after inserting 24 screws with external-star-type lockwashers, and a total of about 1000 metal particles were counted. The crossed-recess screwdriver slot (Phillips head) will also tend to be cut by the action of the screwdriver. When the screw is seated and the screw- driver is twisted, the screwdriver will very often twist out of the slot and shave small bits of metal out of the screw head.

10

INTRODUCTION

In order to avoid shaving small bits of metal from the screws with devices that bit into the metal, many electronic firms use liquids such as Loctite and Glyptal, which bind the screws. Some firms use nylon inserts in the screws. Nylon inserts are usually good for about a dozen insertion and removal cycles before the nylon cold flows and reduces the binding torque. NASA will not permit the use of volatile products in the area of displays and optical devices for their spacecraft. These products tend to outgas in a vacuum and then deposit themselves in a thin film; this can coat optical lenses and interfere with the operation of sensitive optical units such as cameras and telescopes.

1.9 ELECTRONIC EQUIPMENT FOR AIRPLANES AND MISSILES

Electronic boxes used in airplanes and missiles often have odd shapes that permit them to make maximum use of the volume available in tight spaces. Since volume and weight are generally quite critical, the electronic boxes usually have a high packaging density. This value normally ranges from about 0.03 to about 0.04 lb/in.3, depending on the severity of the environmental requirements. The weight of a typical electronic box will range from about 10 to about 80 lb. The vibration frequency spectrum for airplanes will vary from about 3 to about 1000 Hz, with acceleration levels that can range from about 1 G to about 5 G peak. The highest accelerations appear to occur in the vertical direction in the frequency range of about 100-400 Hz. The lowest accelera- tions appear to occur in the longitudinal direction, with maximum levels of about 1 G in the same frequency range. For helicopters, the frequency spectrum will vary from about 3 to about 500 Hz and acceleration will range from about 0.5 to about 4 G. The highest accelerations appear to occur in the vertical direction at frequencies near 500 Hz. The displacement at low frequency are very large, with values of about 0.20-in. double amplitude at about 10 Hz. Missiles have the highest frequency range in this group, with values that generally go up to 5000 Hz [7]. The lower frequency limit is about 3 Hz, and this appears to be due to bending modes in the airframe structure. Accelera- tion levels range from about 5 to about 30 G peak, with the maximum levels occurring during power-plant ignition at frequencies above 1000 Hz. The vibration environment in supersonic airplanes and missiles is actually more random in nature than it is periodic. However, sinusoidal vibration tests are still being used to evaluate and to qualify electronic equipment that will be used in these vehicles. Because the forcing frequencies in airplanes and missiles are so high, it is virtually impossible to design resonance-free electronic systems for these environments. Of course, it is always possible to completely encapsulate an entire electronic box with some expanding rigid type of foam, which could

ELECTRONIC EQUIPMENT FOR AIRPLANES AND MISSILES

11

drive the resonant frequency well above 1000 Hz (possibly to 2000 Hz) for a small box. This is generally considered to be impractical, however, because it becomes too expensive to maintain, troubleshoot, and repair such a system. The obvious conclusion is that the forcing frequencies present in airplanes and missiles will excite many resonant modes in every electronic box. What becomes equally obvious is that extra care must be taken in the design and analysis of an electronic system, or it can literally shake itself to pieces. The electronic support structure must be dynamically tuned with respect to the electronic components to prevent coincident resonances that can lead to rapid fatigue failures. The first thought that comes to the mind of an experienced mechanical design engineer, when confronted with a severe vibration specification, is to mount the electronic equipment on vibration isolators. There is no doubt that a set of isolators, properly designed, can control shock and vibration. There are four major factors that must be considered when isolation mounts are being discussed.

1. Sway space must be provided all around the electronic equipment to keep it from colliding with other objects. If volume is scarce, it might be more practical to pack more electronics into the same volume by using a larger electronic box with hard mounts. 2. Cold plates are being used more and more in electronic structures to remove the heat dissipated by the electronic equipment. If isolators are used, flexible couplings must be provided between the airframe struc- ture and the electronic box to take care of the large displacements developed by the isolators. Reliable flexible couplings must be used because cooling effectiveness may be sharply reduced if a coupling fails. 3. Electrical wire cables and harnesses must be used to connect the typical electronic box to the main electrical system in the airplane or missile. If isolators are used, these cables and harnesses will be forced through large amplitudes because of the sway space required by the isolators. Special precautions must be taken to prevent fatigue failures in the cables and harnesses. 4. A good vibration isolator is often a poor shock isolator, and a good shock isolator is often a poor vibration isolator. The proper design must be incorporated into the isolator to satisfy both the vibration and the shock requirements.

Cold-plate designs for airborne electronic equipment have become more and more sophisticated with the use of air and liquids. Air heat exchangers make extensive use of multiple fins, wavy fins, split fins, and pin fins in order to improve the heat-transfer characteristics. The multiple-fin heat exchangers are usually dip-brazed aluminum with as many as 22 fins per inch. These fins may be only 0.006 in. thick, but the large number of fins in a typical heat

12

INTRODUCTION

exchanger makes it quite rigid for its weight. Airplanes make extensive use of electronic equipment where a cooling-air heat exchanger is built right into the electronic support structure. The heat exchanger is riveted, brazed, or cemented to the major structural members in the electronic box, so that the heat exchanger itself becomes a major load-carrying member of the system. The cooling air for the cold plate is usually taken from the compressor on the jet engine that powers the airplane. This air must be conditioned before it can be used for cooling, because the air temperature from the first stage is usually greater than 300°F. Liquid-cooled cold plates are usually used to cool electronic equipment on spacecraft or very-high-flying research airplanes. The cooling liquid can be a mixture of ethylene glycol and water, or some other liquids such as FC-75, a fluorocarbon or Coolanol 45, a silicone fluid. The liquid-cooled cold plates are usually made part of the spacecraft airframe structure instead of the electronic box structure. Then, when the electronic box must be removed from the spacecraft, it is not necessary to disconnect fluid lines, which can become quite messy. Since the cold plate stays with the airframe, the heat dissipated by the electronic box is usually transferred to the cold plate through a flat interface on the mounting surface of the electronic box that makes intimate contact with the cold plate. The trend in commercial and military electronics is toward the line replaceable unit (LRU), with which it is possible to replace a defective electronic box right on the flight line in a matter of minutes. This is accomplished by providing all of the required interface connections, both mechanical and electrical, at the back end of the electronic box. The box becomes similar to a printed circuit board (PCB) that can be plugged into its receptacle. If there are several large electrical connectors on the back end of the electronic box, it may be quite difficult to insert the box and engage the electrical connectors properly. Some connectors may require a force of 0.50 lb/pin for proper engagement. When there are 8 connectors, each with 100 pins, there is a total of 800 pins, which will require a 400-lb force to engage them. The plug-in electronic box is usually engaged and locked into position by some mechanism at the front of the box. Since the connectors are at the rear of the box, this means the force required to engage the connectors must pass through the box. When this type of electronic box is subjected to vibration, in many instances the vibration loads must be added to the installation loads to determine the total load acting on the structure. There has been an attempt to standardize electronic equipment used in military and commercial airplanes by establishing certain sizes for modular electronic units. These modular units are then mounted in a standard air transport rack (ATR), which provides rear-located dowel pins and connectors and a quick-release fastener at the front.

ELECTRONIC EQUIPMENT FOR SHIPS AND SUBMARINES

13

1.lo

ELECTRONIC EQUIPMENT FOR SHIPS AND SUBMARINES

Ships and submarines will generally make use of console cabinets to support their electronic equipment, since there is usually more room available and weight is not very critical. The electronic components are usually mounted on panels and in sliding drawers. Panels are generally used to support dials, gages, manual controls, and test points. Only small masses are mounted on panels, because they are fastened to a frame or rack in the cabinet and they cannot withstand large dynamic loads. Drawers are often used to support the more massive electronic compo- nents such as those normally used in power supplies. The drawers are mounted on telescoping slides to provide access to the equipment. For safety, the drawers usually lock in the open and closed positions. For convenience, the drawers may also tilt to improve access in tight spaces. The vibration frequency spectrum for ships and submarines varies from about 1 to about 50 Hz, but the most common range is from about 12 to about 33 Hz. The maximum acceleration level in this range is about 1 G and appears to be due to vibrations set up by the engines and propellers. In military ships, shock is an important factor, due generally to various explosions, which can do extensive damage to electronic equipment, unless proper consideration is given to the design and installation. For example, it is not desirable to have a very rigid structure supporting the electronics, because a very rigid structure may not deform enough to absorb much stain energy. Theoretically, any structure that does not deform when it is subjected to an impact load will receive an infinite acceleration. A large displacement is desirable, since it can substantially reduce acceleration levels. Either this displacement must be confined to the structure or shock mounts must be used. In either case, provisions must be made in the design and installation to make sure parts will not collide and equipment will not break loose [36]. If shock isolators are used, they should be designed to deflect enough to absorb the shock energy without transmitting excessive loads to the electronic equipment. The shock mounts should have a minimum resonant frequency of about 25 Hz [2,51. Ideally, the resonant frequency of the electronic compo- nents should be at least twice that of the shock mounts, but never below 60 Hz. If the electronic components have resonances substantially below 60 Hz, it might bring them into the range of the most common vibration forcing frequencies, which, as previously mentioned, are as high as 33 Hz. If this should happen, the electronic components would be driven continuously near their resonance and could suffer fatigue failures. When the forcing frequency of the ship’s structure is near its higher frequency limit (around 25 Hz), the resonant frequency of the electronic equipment cabinet will be excited, since the shock isolators also have their resonance at 25 Hz. This condition should not impose high stresses on the electronic components mounted in the cabinet if the component resonance is

14

INTRODUCTION

twice that of the isolators. The cabinet support structure will, however, have to withstand the dynamic vibration loads. These loads will be determined by the amplification characteristics of the shock isolators during vibration. Shock isolators are available that provide a vibration amplification of about 3 for the conditions described above. Since the general vibration acceleration input level in these frequency ranges is normally quite low, an amplification factor of 3 for the isolators does not result in high stresses in the equipment cabinet.

It is generally not desirable to increase the resonant frequency of the electronic equipment as high as possible. If the equipment is very stiff, it may be good for the vibration condition but poor for the shock condition. A very high spring rate may result in very high shock stresses because of the high

a maximum resonant frequency of

acceleration loads. Lee [5] recommends

about 100 Hz and a maximum acceleration of 200 G on the electronic components. On tall narrow cabinets the load-carrying isolators should be at the base and stabilizing isolators should be at the top. A rigid structure must be used to support the stabilizing isolators at the top. If there is excessive deflection in the top support structure, it can change the characteristics of the entire

system because of severe rocking modes. If shock isolators are not used, the shock energy must be absorbed by deflections in the electronic equipment cabinet and in the structure of the ship supporting the cabinet. In this case the natural frequency of the assembly, which consists of the cabinet and the ship’s structure, should be about 60 Hz. The natural frequency of the electronic components mounted on the equipment cabinet should be twice that of the assembly, or about 120 Hz. The ship’s structure must provide a good part of the deflection required to attenuate the shock force, or the dynamic stresses in the equipment cabinet may be high enough to cause structural failures. When the vibration forcing frequency in the ship’s structure is near its normal maximum limit of 33 Hz, the dynamic loads in the equipment cabinet will not be amplified to any great extent since the resonant frequency of the cabinet, at 60 Hz, is almost twice the forcing frequency. Furthermore, the electronic components mounted in the equipment cabinet have a still higher resonant frequency (120 Hz), so their dynamic vibration loads should be relatively small. The materials that are best suited for shock are ductile materials with a high yield point, a high ultimate strength, and a high percentage elongation. In general, metals that are mechanically formed are more desirable than cast metals, which have a relatively low percentage elongation. Since acceleration forces become progressively smaller as they propagate into the interior of the equipment cabinet, the electronic components that can withstand the highest G forces should be mounted near the exterior of the cabinet. Electronic components that have their own rigid structures

ELECTRONIC EQUIPMENT FOR AUTOMOBILES, TRUCKS, AND TRAINS

15

should be used to lend additional strength to the outer structural elements in the cabinet. Electronic components that cannot withstand high G forces should be mounted at the maximum elastic distance from the application points of the shock load. This will usually be at the center of the cabinet. The load path for each mass element in the system should be examined closely to determine the path the load will take as it passes through the structure. For example, a large transformer should be mounted close to a major structural support in order to reduce the length of the load path. This will result in smaller deflections and stresses.

1.11 ELECTRONIC EQUIPMENT FOR AUTOMOBILES,

TRUCKS, AND TRAINS

Electronic equipment for automobiles and trucks has grown very rapidly in the past few years. Electronics are being used in their antilocking brake systems, fuel-air mixture control, radios, ignition systems, air conditioners, heaters, automatic transmission shifting points, rear view mirrors, door locks, instruments, global positioning systems (GPSs), windows, sun roofs, cruise control, air bags, television, telephones, and many other features, with more being added every year. Anticollision systems are being developed that will automatically engage brakes when vehicles are dangerously close to each other at high speeds. There are plans for computerized car trains that will link several automobiles together going to similar areas. Antitheft systems such as Lojack are available that can be activated by radio to send out a silent alarm to police when an automobile is stolen. Automobile and truck electronics must operate in harsh environments that include wide temperature swings, high under-hood operating temperatures of 140"C, high humidity with condensation, high-velocity splashing water, and subfreezing temperatures with only moderately high vibration and shock levels for the automobiles. Trucks are often required to travel over unpaved country roads for deliveries in outlying areas. Test data on a 2.5-ton truck instrumented with several accelerometers showed acceleration levels between about 15 and 19 G peak at speeds of 10-15 mph, at frequencies between about 15 and 40 Hz. At truck speeds above about 35 mph over rough roads the frequency appears to be random in nature. Diesel electric trains have been making use of electronics on their drive wheels to sense when the wheels are just reaching the slipping point on the tracks. A large engine may have six pairs of driving wheels. The electronics can sense when any one pair of drive wheels will slip. Sand is automatically dispensed to increase friction and the electric motor driving the one set of drive wheels is automatically slowed slightly to prevent slopping on the

16

INTRODUCTION

tracks. A single diesel electric engine with this new feature can go up steeper grades and pull three times as many cars as previous diesel electric trains.

1.12 ELECTRONICS FOR OIL DRILLING EQUIPMENT

Oil drilling equipment probably has the most severe thermal and vibration requirements for electronic systems. Deep drilling at 30,000 feet can en- counter temperatures as high as 200°C. The electronics are usually located in a 6-foot long tube section just above the cutters. The coolant is the 200°C mud washing over the electronics tube section. Sometimes the cutters get stuck in a rock section while drilling at 30,000 feet. The turntable at the surface keeps turning while the cutters are stationary. The long drilling pipe winds up and when it breaks loose it generates random vibration acceleration levels as high as 30 G RMS. This means that the electronics section in an oil drilling rig must be very rugged to withstand the high vibration levels and the high temperatures and still provide a high reliability. When an electronic failure occurs, it is a very expensive and time-consuming task to remove 30,000 feet of pipe, replace the electronics section, and drop 30,000 feet of pipe down the hole again.

1.I3

ELECTRONICS FOR COMPUTERS, COMMUNICATION,

AND ENTERTAINMENT

According to many people in the three fields of computers, communication, and entertainment, they are rapidly converging into one giant industry with the internet. Sometimes this is called bandwidth, because fiber optic cables are being added, which can carry a very wide range of frequencies that are required to provide services for television, radio, data processing, stock trading, banking, internet, telephone, security, and more energy efficient homes and offices.

- CHAPTER 2

Vibrations of Simple Electronic Systems

2.1 SINGLE SPRING-MASS SYSTEM WITHOUT DAMPING

The natural frequency of many single-degree-of-freedom systems can be determined by evaluating the characteristics of the strain energy and the kinetic energy of each system. Considering a single spring-mass system, for example, if there is no energy lost, then the maximum kinetic energy of the mass must be equal to the maximum strain energy in the spring, if the spring mass is negligible. All real systems have some damping. If a mass is suspended on a real coil spring and the spring is stretched and released, the mass will vibrate up and down. This free vibration may continue for a long time, but eventually all free vibrations die out and the mass stops vibrating. This is because damping in the spring dissipates a little energy with each cycle and eventually the mass stops moving. If there is no damping, the mass will theoretically keep on vibrating up and down with the same amplitude and frequency forever. In many systems, the damping is so small that it has very little effect on the natural frequency. Under these circumstances the natural frequency of the damped system can conveniently be approximated by the natural fre- quency of the undamped system. The maximum kinetic energy of a vibrating spring-mass system with no damping is at the point of maximum velocity. This occurs as the vibrating mass passes through the zero-displacement point, as shown in Fig. 2.1. From elementary physics the maximum kinetic energy T of the vibrating mass is

(2.1)

The instantaneous tangential velocity I/ can be expressed in terms of rota- tional velocity as shown in Fig. 1.3:

(2.2)

To= imV2

v=Yon

Substituting Eq. 2.2 into Eq. 2.1, the kinetic energy of the system becomes

To= ;my,Zn*

(2.3)

18

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

T

<

-<

+Displacement

t

Maximum strain energy

rJ-1

, ;

--_------- J

Time

-Displacement

' Maximum velocity

FIGURE 2.1. A vibrating spring-mass system.

The maximum strain energy can be determined from the work done on the spring by the mass as it stretches and compresses the spring during vibration. Since the spring is linear, the deflection is directly proportional to the force as shown in Fig. 2.2. The area under the curve represents the work done on the spring. Since the work is equal to the strain energy, the strain energy becomes

U" = +P,Y"

The spring rate K can be defined in terms of the maximum load P, and the maximum deflection Yo as K = P,/Y,, so

Po = KY,,

(2.5)

Substituting Eq. 2.5 into Eq. 2.4, the maximum strain energy becomes

(2.4)

uo= ;KY;

If damping is assumed to be zero, then at resonance the kinetic energy will be equal to the strain energy:

i,y2flz

2

(I

=

'Ky2

0

2

Q" = [E)

1/2

SINGLE SPRING-MASS

SYSTEM WITHOUT DAMPING

I9

This is the natural frequency in radians per second. Using Eq. 1.2, the natural frequency in cycles per second (Hz) is

The natural frequency equation can be written in a slightly different form by considering that the static deflection a,, of the spring is due to the action of the weight W acting on the spring. The spring rate can then be written as

Expressing the mass in terms of weight W and acceleration of gravity g, we have

Substituting

natural frequency:

Eqs. 2.8 and 2.9 into Eq. 2.7 gives another relation for the

(2.10)

Sample Problem-Natural

Frequency of a Cantilever Beam

Determine the natural frequency of the cantilevered beam with an end mass, as shown in Fig. 2.3, in the vertical direction.

Solution. When the weight of the aluminum beam is small compared to the weight of the end mass, it can be ignored without too much error. The cantilever beam can then be analyzed as a concentrated load on a massless beam, which retains its elastic properties. The static deflection at the end of

1Ayq-g

1.4

-100

in ---+

O 50 -1rL3

B-r

Sect AA

FIGURE 2.3. Cantilever beam with an end mass.

20

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

this beam can be determined with the use of a structural handbook:

where W= 10.0 Ib (end weight)

(2.11)

L = 10.0 in. (length)

E = 10.5 x lo6 lb/in.’ (modulus of elasticity, aluminum) I = bh3/12 = (0.50)(3.0)3/12 = 1.125 in.4 (moment of inertia)

(10.0)(10.0)~

= 3(10.5 X 106)(1.125)

= 0.282 X

in.

(2.12)

Substitute Eq. 2.12 into Eq. 2.10 and note that the acceleration of gravity, g, is 386 in./s2. The natural frequency becomes

1

=

187 HZ

(2.13)

The natural frequency of the cantilever beam can also be determined from the spring rate, K, of the beam using Eq. 2.11:

(2.14)

Using the values previously calculated, the spring rate becomes

Also,

K=

3( 10.5 x lo6)(1.125)

(10.0)3

= 3.54 X lo4 lb/in.

w

10

m=-=-= g

386

0.0259 Ib . s’/in.

Substituting into Eq. 2.7, the natural frequency becomes

”’-[-)27r m

1

K

1/2

f, = 187 HZ

=z(0.0259 1

1

3.54 x lo4

1/?

(2.15)

SINGLE-DEGREE-OF-FREEDOMTORSIONAL SYSTEMS

21

2.2 SINGLE-DEGREE-OF-FREEDOM TORSIONAL SYSTEMS

The energy method is convenient for determining the natural frequency of a torsional system with one degree of freedom, as shown in Fig. 2.4. The torsional system is similar to the spring-mass system shown in Fig. 2.1, except that the spring action is due to the twisting of the rod and the inertia is due to the moment of inertia of the disk about an axis perpendicu- lar to the plane of the disk. Assume the rod mass is small. The maximum kinetic energy of the oscillating disk is

KIN = ;zme2

(2.16)

where

I

= mass moment of inertia of disk 6 = rotational angular velocity

The maximum angular velocity for the oscillating system moving through the angle Bo is

(2.17)

Substituting Eq. 2.17 into Eq. 2.16, the maximum kinetic energy becomes (assuming a small rod mass)

(2.18)

The maximum strain energy can be determined from the work done on the twisting rod by the disk as it oscillates. Since the spring rate of the rod is linear, the angular deflection will be directly proportional to the torque T applied, as shown in Fig. 2.5. The area under the curve represents the work

io= eon

22

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

done on the rod. Since the work is equal to the strain energy, the strain energy becomes

u, = +T,O,

(2.19)

The torsional spring rate K, can be defined in terms of the torque To and

the angular displacement

so

On as follows:

K

0

TO

=-

On

To = K, Bo

(2.20a)

(2.20b)

Substituting Eq. 2.20 into Eq. 2.19, the maximum strain energy becomes

U, = iK,Oi

(2.21)

If damping is assumed to be zero, then at resonance the kinetic energy (Eq. 2.18) must equal the strain energy (Eq. 2.21):

The natural frequency (in Hz) becomes

j""f('=2T I,,

1/2

Sample Problem-Natural

Frequency of a Torsion System

(2.22)

(2.23)

Determine the natural frequency of the torsional system shown in Fig. 2.6.

FIGURE 2.6. A single-degree-of-freedom torsional

system.

Aluminum

{-X7"'i

o&

-'R = 6 0 1"

L- f = 050 in

SPRINGS IN SERIES AND PARALLEL

23

Solution. The torsional spring rate K, of the rod can be determined using its angular displacement 8 under the action of an external torque To:

so

where

TOL

e= -

GJ

and

TO

KO=-

e

L = 10.0 in. (length of rod)

4.0 X lo6 1b/ine2 (shear modulus of aluminum) J = rd4/32 = ~(1.0)~/32= 0.0981 in.4 (polar moment of inertia)

G =

K, =

(4.0 X 106)(0.0981)

10.0

= 3.92 X IO4 in:lb/rad

The mass moment of inertia of the aluminum disk must be taken about the axis perpendicular to the plane of the disk:

I,

WR

= -

2g

where W = ~(6.0)~(0.50)/(0.10lb/ir~.~)= 5.65 lb (disk weight) R = 6.0-in. disk radius g = 386 in./s2 (gravity)

I,

=

(5.65)(6.0)2

2( 386)

= 0.264 1b.in:s’

The torsional resonant frequency can now be determined from Eq. 2.23:

’’=%[

I

3.92 x 104

0.264

1

1,/2

= 61.5 HZ

(2.24)

2.3 SPRINGS IN SERIES AND PARALLEL

If a mass is suspended on two different springs in such a manner that the load path is directly from the mass through one spring, and then through the second spring before it reaches the support, the springs are said to be in series. The series spring, therefore, implies a series load path where the load must first pass through one of the springs before it can pass through the

24

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

Tension and

compression

-

Torsion

Bending

FIGURE 2.7. Springs in series.

other spring. Cutting either spring would completely destroy the system. Some samples of springs in series are shown in Fig. 2.7. Springs in series can be combined into one equivalent spring using the relation

1

Ke,

-

111

+-+-+

K,

K?

K3

(2.25)

If a mass is suspended on two different springs in such a manner that the load path from the mass to the support is split between the springs, then the springs are said to be in parallel. The parallel path then permits the load to pass through one spring without passing through the other. Cutting one spring would still permit the other spring to carry the load. Some samples of springs in parallel are shown in Fig. 2.8. Springs in parallel can be combined into one equivalent spring using the relation

K,,

= K, +K, +K, +

(2.26)

SPRINGSIN SERIES AND PARALLEL

25

-

1

Kz = 10001b/in. <

5

<

<

6

K3 = 1000 Ib/in.

3

> K4 = 800 Ib/in.

>

FIGURE

and parallel springs.

2.9. A

combination of

Sample Problem-Resonant

Frequency of a Spring System

series

Determine the natural frequency of a mass supported by several springs in series and parallel combinations as shown in Fig. 2.9, by obtaining the equivalent spring rate for the system.

Solution. Springs K, and K, are in parallel, so they can be combined using Eq. 2.26:

K, = K, +K, = 1000+ 1000 = 2000 lb/in.

Now there are three springs-K,, K,, combined using Eq. 2.25:

1 111

--

K,,

_-

K,

+-+-=-+-

K,

K,

and

1

4000

K,-in

1

2000

series, so they can be

+-

1

800

The system now consists of one spring and one mass (Fig. 2.10). The natural frequency can be determined from Eqs. 2.7 and 2.9 as follows:

f,,= 49.5 HZ

3

;K,,= 500Ib/in.

>

1 W= 2 Ib

FIGURE 2.10. A single spring-mass

system.

(2.27)

26

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

2.4 RELATION OF FREQUENCY AND ACCELERATION

TO DISPLACEMENT

Many important approximations can be related to the dynamic displacements developed during resonant conditions. For example, if the geometry of a specific structure can be defined, then dynamic bending moments and bending stresses can be calculated from the dynamic displacements. Since displacements are very seldom measured during vibration tests, because optical measurements are often difficult to make, accelerometers are gener- ally used. This permits test data to be taken in terms of frequency and acceleration G forces. The relation of frequency and acceleration to the displacement must therefore be determined to use the test data. Consider a rotating vector that is used to describe the simple harmonic motion of a single mass suspended by a spring. The vertical displacement Y of the mass can be determined by the projection of the vector Yo on the vertical axis as shown in Fig. 2.11. The vertical displacement can be represented by the equation

Y = Yosin nt

The velocity is the first derivative,

dY

I/= Y= -

dt

= Szy, cos nr

The acceleration is the second derivative.

d2Y

A = Y= ~

dt’

= -a2yosin fit

(2.28)

The negative sign indicates that acceleration acts in the direction opposite to displacement.

RELATION OF FREQUENCY AND ACCELERATION TO DISPLACEMENT

27

The maximum acceleration will occur when sin Rt is 1:

A,,,

= R2Y0

(2.29)

The acceleration in gravity units (G) can be determined by dividing the maximum acceleration by the acceleration of gravity:

g = 32.2 ft/s2 = 386 in./s2

Also, changing radiants to cycles per second, we have

fl=2.rrf

Substituting into Eq. 2.29,

G=-- A,,,

g

- 4X2f2YJ

386

(2.30)

Note that the displacement Yo is the single-amplitude displacement.

Sample Problem-Natural

Frequency and Stress in a Beam

Determine the resonant frequency, dynamic displacement, and maximum dynamic stress in the bracket with a transformer, as shown in Fig. 2.12, for a 5.0-G peak sinusoidal vibration input.

Solution. Consider the transformer as a concentrated mass at the center of a massless beam. Since small screws are used to support the beam, the ends are considered to be simply supported. The natural frequency of a simply supported beam with a concentrated center mass can be determined from the static displacement as shown in Fig. 2.13, using Eq. 2.10. The static displacement can be obtained from a hand- book:

W Transformer

Screw

WL3

6,, = -

48EI

Bracket

(2.31)

L ---

0503Sect AA

FIGURE 2.12. A transformer mounted on a bracket.

28

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

FIGURE 2.13. A simply supported beam with a concen- trated load.

where

W = 2.0 Ib (transformer weight)

--

--

W=201b

~ -8O---.--.

L

= 8.0 in. (length)

E

= 10.5 x lo6 (aluminum-beam modulus of elasticity)

I = bh3/12 = (1.2)(0.50)'/12 g = 386 in./s2

= 0.0125 in4 (moment of inertia)

(2.0) (8.0)3

= 1.62 X

in.

= 48(10.5 X 106)(0.0125)

Substitute into Eq. 2.10 for the natural frequency:

f, = 246 Hz

(simply supported ends)

(2.32)

Equation 2.30 can be used to determine the dynamic deflection at the center of the transformer bracket, using the 246-Hz resonance:

9.8G

yo=--

f2

9,8Gi,Q

-

f2

(2.33)

where G,, = 5.0 G (input acceleration) Q = 30 (transmissibility at resonance) f, = 246 Hz (simply supported beam)

Yo=

9.8(5.O) (30)

(246)'

= 0.0243 in.

(2.34)

The approximate transmissibility for a beam-type structure can be ob- tained from the resonant frequency. For an electronic subassembly, experi- ence has shown that a value equal to two times the square root of the resonant frequency, or about 30, gives reasonable results.

RELATION OF FREQUENCY AND ACCELERATION TO DISPLACEMENT

29

The dynamic load acting on the bracket can be approximated by letting Eq. 2.34 equal Eq. 2.31 and solving for the dynamic load wd:

w, = ~

48EW,

L3

48(10.5 X 106)(0.0125)(0.0243)

w,=

W, = 300 lb

(8.0)3

(dynamic load)

(2.35)

If the acceleration loads were assumed to act directly on the transformer, the dynamic load would be

?t$ = WG,,Q

(2.36)

where W = 2.0 lb (transformer weight) Gin= 5.0 G (peak input) Q = 30 (transmissibility at resonance)

Thus

W, = (2.0) (5.O) (30) = 300 Ib

(2.37)

This result is exactly the same as that shown in Eq. 2.35, which was obtained from the dynamic deflection of a simply supported beam with a concentrated load. The dynamic bending stress in the simply supported transformer bracket can be determined from the geometry of the structure, using the dynamic loading as shown in Fig. 2.14. The maximum dynamic bending moment

Lr4.0 7

Wd = 300 Ib

R= 150

8.0

Jlransfcyy

R= 150

bracket

_L50_t.--150-~-Shear diagram

FIGURE 2.14. Shear and bending-moment diagram for a beam

30

VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS

occurs at the center of the bracket:

=150(4) =600in:lb

(2.38)

The maximum dynamic bending stress can be determined from the standard bending-stress equation:

MC

s, = r

where M = 600 in: lb (bending moment)

Hence

c = h/2 = 0.50/2 = 0.25 in. (ref. Sect. A.A. Fig. 2.12)

I = bh3/12 = (1.2)(0.50)'/12

= 0.0125 ine4

s, =

(600)(0.25)

0.0125 = 12,000 lb/in.2

(2.39)

(2.40)

A stress concentration factor must be used when several million stress reversals are expected. When only a few thousand stress cycles are expected in a ductile material, then a stress concentration factor is not critical.

2.5 FORCED VIBRATIONS WITH VISCOUS DAMPING

When a harmonic shaking force acts on a spring-mass system with damping, the resulting forced vibration will also be harmonic. The final frequency of the mass will be the same as that of the shaking force, because the initial transient vibrations will eventually be dissipated by the damping.

the harmonic shaking force Po cos Rt acting on a damped

spring-mass system. A free-body force diagram is shown in Fig. 2.15. The differential equation of motion with an external force exciting the system is

Consider

mY+ CY+ Ky= P" cos at

(2.41)

The solution of this equation consists of a complementary function plus a particular function. The complementary solution is the free vibrations. These will die out because of the damping. The particular solution can be taken in the form

(2.42)

Y=Y,cos(Rt-

e)

FORCED VIBRATIONS WITH VISCOUS DAMPING

31

t

Po cos Ltt

1

Po cos nt

FIGURE 2.15. Forced vibration forces acting on a simple system.

The maximum displacement, Yo,can be expressed in terms of the maxi- mum impressed force, Po, as follows:

PO

Yo= [(K - mQ2)' +c2R2]'''

(2.43)

Dividing the numerator and denominator of the abo