VIBRATION ANALYSIS FOR ELECTRONIC EQUIPMENT
THIRD EDITION
Dave S. Steinberg
Steinberg & Associates and University of California, Los Angeles
A WILEYINTERSCIENCE PUBLICATION
JOHN WILEY & SONS, INC.
New York
Chichester
Weinheim
Brisbane
Singapore
Toronto
This book is printed on acidfree 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 percopy fee to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923, (978) 7508400, fax (978) 7504744. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 605 Third Avenue, New
101580012, (212) 8506011, fax (212) 8506008, EMail: 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 WileyInterscience publication.” ISBN 047137685X 1. Electronic apparatus and appliancesVibration. I. Title.
TK7870S8218 2000 

621.381dc21 
99056617 
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
xvii
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 threeband technique with much less work. A quick way of estimating the effectiveness of stiffening ribs on plugin 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 throughholes for surface mounted and throughhole components, due to differences in the thermal coefficients of expansion (TCE), can be
_{f}_{o}_{u}_{n}_{d} _{i}_{n} _{m}_{y} _{b}_{o}_{o}_{k} _{e}_{n}_{t}_{i}_{t}_{l}_{e}_{d}
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) _{3}_{8}_{6} in./s2
a length (in.) ASIC Applicationspecific
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 _{.}_{i}_{n}_{.}_{)} 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)
_{A}_{4}
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) _{r}_{n} with subscript: acceleration N in gravity units (dimensionless) 
xix
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 
Fatiguecycle 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 Obin.), 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 SpringMass System Without Damping 
17 
Sample ProblemNatural
Beam
Frequency of a Cantilever
19
2.2 SingleDegreeofFreedom Torsional Systems 
21 
Sample ProblemNatural Frequency of a Torsion System 
22 
2.3 Springs in Series and Parallel 
23 
Sample ProblemResonant Frequency of a Spring System 
25 
2.4 Relation of Frequency and Acceleration to Displacement 
26 
Sample ProblemNatural 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 ProblemRelating the Resonant Frequency to the Dynamic Displacement 
34 
vii
Viii
CONTENTS
_{2}_{.}_{7} 
Multiple SpringMass Systems Without Damping 
36 

Sample ProblemResonant 
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 ProblemVibration Fatigue Life in the Wires of a TO5 Transistor
40
3.3 Vibration Fatigue Life in Solder Joints of a TO5 Transistor 
43 

3.4 Recommendations to Fix the Wire Vibration Problem 
45 

3.5 Dynamic Forces Developed in Transformer Wires During Vibration 
46 

Sample ProblemDynamic Transformer Lead Wires 
Forces and Fatigue Life in 
46 
3.6 Relative Displacements Between PCB and Component Produce Lead Wire Strain 
49 

Sample ProblemEffects Reliability 
of PCB Displacement _{o}_{n} Hybrid 
50 
4. Beam Structures for Electronic Subassemblies
56
4.1 Natural Frequency of a Uniform Beam 
56 

Sample ProblemNatural 
Frequencies of Beams 
60 
4.2 Nonuniform Cross Sections 
64 
Sample ProblemNatural 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 LoadHinged Ends 
77 

5.3 
Strain EnergyBent 
with Hinged Ends 
80 
5.4 
Strain EnergyBent 
with Fixed Ends 
83 
5.5 
Strain EnergyCircular 
Arc with Hinged Ends 
90 
5.6 Strain EnergyCircular 
Arc with Fixed Ends 
92 

5.7 Strain EnergyCircular 
Arcs for Lead Wire Strain 

Relief 
94 

Sample ProblemAdding 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 ProblemResonant 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 ProblemVibration 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 ProblemNatural 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 Plugin 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 ProblemVibration 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 ProblemAdding 
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 
7.12
Problems with PCB Viscoelastic Dampers
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 ProblemQualification 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 ProblemDetermining 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 ProblemResonant 
Frequency of PCB with Side 

Wedge Clamps 
179 

8.7 Effects of Loose PCB Side Edge Guides 
182 

Sample ProblemResonant Edge Guides 
Frequency of PCB with Loose 
185 
8.8 Sine Sweep Through a Resonance 
185 

Sample ProblemFatigue 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 ProblemDetermining the Input RMS 

Acceleration Level 
193 

9.6 Random Vibration Units 
193 

9.7 Shaped Random Vibration Input Curves 
194 

Sample ProblemInput 
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 ProblemFinding PSD Values 
200 

9.11 Using Basic Logarithms to Find Points on the PSD Curve 
201 
9.12
Probability Distribution Functions
202
CONTENTS
xi
9.13 Gaussian or Normal Distribution Curve 
202 

9.14 Correlating Random Vibration Failures Using the ThreeBand Technique 
204 

9.15 Rayleigh Distribution Function 
205 

9.16 Response of a SingleDegreeofFreedom System to Random Vibration 
206 

Sample ProblemEstimating 
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 ProblemFinding 
the Desired PCB Resonant 

Frequency 
218 

9.19 Effects of Relative Motion on Component Fatigue Life 
220 

Sample ProblemComponent 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 ProblemDetermining 
Approximate Connector 

Fatigue Life 
224 

9.22 MultipleDegreeofFreedom Systems 
224 

9.23 Octave Rule for Random Vibration 
225 

Sample ProblemResponse 
of Chassis and PCB to 

Random Vibration 
226 

Sample ProblemDynamic 
Analysis of an Electronic 

Chassis 
229 

9.24 Determining the Number of Positive Zero Crossings 
231 

Sample ProblemDetermining Zero Crossings 
the Number of Positive 
233 
10 Acoustic Noise Effects on Electronics
234
10.1 Introduction 
234 

Sample ProblemDetermining 
the Sound Pressure Level 
234 
10.2 Microphonic Effects in Electronic Equipment 
235 

10.3 Methods for Generating Acoustic Noise Tests 
236 

10.4 OneThird Octave Bandwidth 
238 

10.5 Determining the Sound Pressure Spectral Density 
238 

10.6 Sound Pressure Response to Acoustic Noise Excitation 
239 

Sample ProblemFatigue Life of a SheetMetal Panel Exposed to Acoustic Noise 
240 

10.7 Determining the Sound Acceleration Spectral Density 
245 

Sample ProblemAlternate 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 
HalfSine Shock Pulse for Zero Rebound and Full 

Rebound 
252 

Sample ProblemHalfSine ShockPulse 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 ProblemResponse Shock Pulse 
of a PCB to a HalfSine 
262 

11.9 
Response of PCB to Other Shock Pulses 
264 

Sample ProblemShock 
Response of a Transformer 

Mounting Bracket 
265 

11.10 
Equivalent Shock Pulse 
269 

Sample ProblemShipping Crate for an Electronic Box 
269 

11.11 
Low Values of the Frequency Ratio R 
274 

Sample ProblemShock Amplification for Low Frequency Ratio R 
274 

11.12 
Shock Isolators 
275 

Sample ProblemHeat Developed in an Isolator 
276 

11.13 
Information Required for Shock Isolators 
277 

Sample ProblemSelecting a Set of Shock Isolators 
278 

11.14 
Ringing Effects in Systems with Light Damping 
281 

11.1s 
How TwoDegreeofFreedom Systems Respond to Shock 
282 

11.16 
The Octave Rule for Shock 
284 

11.17 
Velocity Shock 
285 

Sample ProblemDesigning Shock a Cabinet for Velocity 
285 

11.18 
Nonlinear Velocity Shock Sample ProblemCushioning 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 ProblemShock Response Spectrum Analysis for Chassis and PCB 
292 

11.21 
How Pyrotechnic Shock Can Affect Electronic Components 
296 

Sample ProblemResonant 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 CenterofGravity 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 ProblemEffects 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 
14.6
Bolt Preload Torque
352
XiV
CONTENTS
Sample ProblemDetermining Desired Bolt Torque 
353 

14.7 Rocking Modes and Overturning Moments 
353 

14.8 OilFilm 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 PushBar Couplings 
360 

14.15 Slider Plate Longitudinal Resonance 
364 

14.16 Acceleration Force Capability of Shaker 
365 

14.17 Positioning the ServoControl Accelerometer 
366 

14.18 More Accurate Method for Estimating the Transmissibility Q in Structures 
367 

Sample ProblemTransmissibility 
Expected for a Plugin 

PCB 
368 

Vibration Testing Case Histories 
369 

14.19 CrossCoupling Effects in Vibration Test Fixtures 
369 

14.20 Progressive Vibration Shear Failures in Bolted Structures 
370 

14.21 Vibration PushBar 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 ProblemFatigue 
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 lowerfrequency 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
1
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 steadystate 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 singledegreeoffreedom 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 singledegreeoffreedom 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 singledegreeoffreedom 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. Singledegreeoffreedomsystems.
X+
(a) f 6)
_{F}
Y
(C)
(di
FIGURE 1.4. Twodegreeoffreedomsystems.
A twodegreeoffreedom 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 rigidbody 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 singledegreeoffreedom system will have only one vibration mode and only one resonant frequency. The sixdegreeoffreedom 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 straightline 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 tradeoffs 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 colddriven shankexpanding 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 highspeed 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.
2.
The screws should be tightened by a torque device, which can be preset to the required value.
FASTENERS
9
TABLE 1.1
~~
Screw 
Torque 
Screw 
Torque 

Size 
(in: 
lb) 
Size 
(in: 
lb) 
264 
33.5 
1224 
4556 

440 
56.5 
1228 
5064 

632 
1012 
$20 
6580 

832 
2024 
_{}_{2}_{8} 
85100 

1024 
2227 
i16 
250320 

1032 
3442 
f 24 
330415 
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 hexhead machine screw seems to be, for general purposes, the most satisfactory. 
5. In all applications involving throughholes, locknuts should be used instead of lockwashers. Most of the standard steel locknuts are satisfac tory.
6. Blindtapped 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 screwlocking inserts should be used with care in elec tronic systems that will be used in the zeroG 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 zeroG environment and create electrical problems. An actual count was made of the metal particles after inserting 24 screws with externalstartype lockwashers, and a total of about 1000 metal particles were counted. The crossedrecess 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 100400 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 0.20in. 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 powerplant 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 resonancefree 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.
Coldplate 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 heattransfer characteristics. The multiplefin heat exchangers are usually dipbrazed 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 coolingair 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 loadcarrying 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. Liquidcooled cold plates are usually used to cool electronic equipment on spacecraft or veryhighflying research airplanes. The cooling liquid can be a mixture of ethylene glycol and water, or some other liquids such as FC75, a fluorocarbon or Coolanol 45, a silicone fluid. The liquidcooled 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, _{i}_{t} 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 400lb force to engage them. The plugin 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 _{e}_{l}_{e}_{c}_{t}_{r}_{o}_{n}_{i}_{c} _{u}_{n}_{i}_{t}_{s}_{.} These modular units are then mounted in a standard air transport rack (ATR), which provides rearlocated dowel pins and connectors and a quickrelease 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 _{H}_{z}_{.} 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 loadcarrying 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, fuelair mixture control, radios, ignition systems, air conditioners, heaters, automatic transmission shifting points, rear view mirrors, door locks, instruments, global positioning systems _{(}_{G}_{P}_{S}_{s}_{)}_{,} windows, _{s}_{u}_{n} 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 underhood operating temperatures of 140"C, high humidity with condensation, highvelocity 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 2.5ton truck instrumented with several accelerometers showed acceleration levels between about 15 and 19 G peak at speeds of 1015 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 6foot 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 timeconsuming 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 SPRINGMASS SYSTEM WITHOUT DAMPING
The natural frequency of many singledegreeoffreedom systems can be determined by evaluating the characteristics of the strain energy and the kinetic energy of each system. Considering a single springmass 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 springmass system with no damping is at the point of maximum velocity. This occurs as the vibrating mass passes through the zerodisplacement 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)
17
_{1}_{8}
VIBRATIONS OF SIMPLE ELECTRONIC SYSTEMS
T
<
<
+Displacement
t
Maximum strain energy
rJ1
, _{;}
_ J
Time
Displacement
' Maximum velocity
FIGURE 2.1. A vibrating springmass 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 SPRINGMASS
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 ProblemNatural
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
1Ayqg
1.4
100
in +
O 50 1rL3
Br
Sect AA
FIGURE 2.3. Cantilever beam with an end mass.
_{2}_{0}
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)
The results are exactly the same as Eq. 2.13.
SINGLEDEGREEOFFREEDOMTORSIONAL SYSTEMS
21
2.2 SINGLEDEGREEOFFREEDOM 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 springmass 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
_{2}_{2}
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 ProblemNatural
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 singledegreeoffreedom torsional
system.
Aluminum
{X7"'i
o&
'R = 6 0 1"
L f = 050 in
SPRINGS IN SERIES AND PARALLEL
_{2}_{3}
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 = 6.0in. 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
_{2}_{4}
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
_{2}_{5}

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 ProblemResonant
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 springsK,, K,, combined using Eq. 2.25:
1 111

K,,
_
K,
++=+
K,
K,
and
1
4000
K,in
1
2000
series, _{s}_{o} 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 springmass
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}_{.}_{2}_{8}_{)}
The negative sign indicates that acceleration acts in the direction opposite to displacement.
RELATION OF FREQUENCY AND ACCELERATION TO DISPLACEMENT
_{2}_{7}
The maximum acceleration will occur when sin Rt is 1:
A,,,
= R2Y0
_{(}_{2}_{.}_{2}_{9}_{)}
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 singleamplitude displacement.
Sample ProblemNatural
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 5.0G 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.
_{2}_{8}
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 (aluminumbeam 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 246Hz 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 beamtype 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}_{.}_{3}_{7}_{)} 
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 bendingmoment 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 bendingstress 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 springmass 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 _{R}_{t} acting on a damped
springmass system. A freebody 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
Гораздо больше, чем просто документы.
Откройте для себя все, что может предложить Scribd, включая книги и аудиокниги от крупных издательств.
Отменить можно в любой момент.