Application of Sensors in An Experimental Investigation of Mode D

Rochester Institute of Technology
RIT Scholar Works

Theses Thesis/Dissertation Collections
1993
Application of sensors in an experimental

investigation of mode dampings
Richard Hiram Berg
Follow this and additional works at: http://scholarworks.rit.edu/theses
Recommended Citation
Berg, Richard Hiram, "Application of sensors in an experimental investigation of mode dampings" (1993). Thesis. Rochester Institute
of Technology. Accessed from
This Thesis is brought to you for free and open access by the Thesis/Dissertation Collections at RIT Scholar Works. It has been accepted for inclusion
in Theses by an authorized administrator of RIT Scholar Works. For more information, please contact ritscholarworks@rit.edu.
Application of Sensors in an Experimental
Investigation of Mode Dampings
by
Richard Hiram Berg
A Thesis Submitted
III
Partial Fulfillment
of the
Requirements of the Degree of
Master of Science
ill
Mechanical Engineering
Approved by:
Dr. Hany Ghoneim - Thesis Advisor (R.I.T.)
Dr. Richard G. Budynas, P.E. (R.I.T.)
Dr. Guillermo H. Garzon (Eastman Kodak Co.)
Dr. Wayne W. Walter, P.E. (R.I.T.)
Dr. Charles Haines - Department Head (R.I.T.)
Department of Mechanical Engineering

College of Engineering
Rochester Institute of Technology
Rochester, New York 14623
1993
Application of Sensors in an Experimental
Investigation of Mode Dampings
I, Richard Berg, hereby grant permission to the Wallace Memorial
Library of the Rochester of Technology to reproduce my thesis in

whole or part. Any reproduction will not be for commercial use or
profit.
August 10, 1993

Acknowledgements
I dedicate this work to my parents Richard and Rita Berg, my wife
Laura and my uncle Peter Grimard. Their great encouragement and
devoted support will not be forgotten.
I wish to give special thanks to all who have supported this work, Paul
Daloia of Bruel & Kjaer, Bob Sylvester of Bruel & Kjaer, Nicholas
Miller of Bruel and Kjaer, Dr. Guillermo Garzon of the Eastman
Kodak Company, Mark Hammer of the Eastman Kodak Company,
Mike O'Brien of the Eastman Kodak Company, Dr. Richard Budynas
of the Rochester Institute Technology, Dr. Wayne Walter of the
of
Rochester Institute of Technology, David Hathway of the Rochester

Institute of Technology, Jim Greanier of the Rochester Institute of
Technology, Tom Locke of the Rochester Institute of Technology, and
especially to my advisor Dr. Hany Ghonein of the Rochester Institute
of Technology and to my father Richard Berg of the Eastman Kodak
Company.
ABSTRACT
The subject of this investigation is the method of calculating material damping of a
cantilever beam. Damping is the mechanism by which the vibrational energy is
gradually converted into heat or sound. The loss factor, otherwise known as the
damping factor, is part of the global properties of a structure which can be
identified from its frequency response measurements. Experimental frequency

response measurements require three main components: a source of excitation,
sensors, and analyzer. It is the objective of this thesis to investigate the effect of
different types of exciters and sensors on the frequency response measurement of
the cantilever beam and consequently on the loss factor estimate of the cantilever
beam. Two sources of excitations and four types of sensors are considered.
Discussion of the results, comments and recommendations about the different
techniques are presented.

TABLE OF CONTENTS
Acknowledgments i
Abstract ii
Table of Contents iii
List of Symbols vi
List of Illustrations xi
List of Tables xxiii
/. Introduction 1
2. Damping 4
2.1 Characteristics of Damping 4
2.2 Environmental Effects 6
2.3 Material Selection 10
2.4 Method of Calculating Material Damping 14
2.4. 1 Damping Measurements for a 1-D model
of a Cantilever Beam 15
2.4.2 Decay Constant 20
3. Modal Testing 24
3.1 Modal Testing 24
3.2 Fundamental Functions of the Bruel & Kjaer

2032 Digital Signal Analyzer 27
3.3 ASTM Standard Method for Measuring Damping

Properties of Materials 30
4. Sensors 33
4.1 Laser Vibrometer 33
4.1.1 Description 33
4.1.2 Technical Background 33
Ul
4. 1.3 Experimental Setup and Operation 36
4.1.3.1 Technical Preparation 36

4. 1.3.2 Basic Operation using a Mirror for Signal
Conditioning 44
4.1.3.3 Laser Vibrometer in Application of a
Cantilever Beam 46
4.2 Piezoceramics 49
4.2.3 Experimental Setup and Operation 51

4.3 Piezofilms
4.4 Piezoelectric Accelerometer 60
4.4. 1 Description 60
5. Excitation Techniques 66
5.1 Impact Excitation 66
5. 1.3 Experimental Setup and Operation 66
5.1.4 Experimental Procedure for Calculating the 69

Loss Factor
5.2 Random Excitation 73
5.2.4 Experimental Procedure for Calculating the 76

Loss Factor
5.3 Other Techniques 79
IV
6. Discussion of Results 82
6. 1 Experimental Determination of the Loss Factor 82
6.2 Noise Sources in the Experimental Testing 93
7. Comments and Recommendations 99
References 100
Appendices 103
A: Analytical Determination of Natural Frequencies for

Aluminum 104
B: Impact Excitation Data of Aluminum from Sensors 109
C: Random Excitation Data of Aluminum from Sensors 157
D: Impact Excitation Data of Laminated Composite

Graphite Epoxy 184
E: Identification of Artificial Damping by the Shakertable in

the vicinity of 237 Hz. 187
F-
Case Studies :
Rectangular Window and Pseudo Random Excitation

vs. Hanning Window and Random Excitation
Hanning Window and Pseudo Random Excitation
Random Impact Excitation using a Modal Hammer 216
G: Variable Sine Excitation Data by Piezoceramic Sensor 237
H: Piezofilms with Unidentified Noise Source 240
I: Climatological Data of Rochester New York 244
J: Dantec Laser Doppler Vibrometer Tables 253
K: Mobility Measurements from "A State-Of-The-Art

Assessment ofMobility Measurement Techniques Results
(30-3000Hz)"
for the Mid-Range Structures by D.J. Ewins
and J Griffin 255
LIST OF SYMBOLS
a -
acceleration
A -
area of the detector aperture
A(f) -
frequency spectrum of original time signal
a(t)
-
Enhanced Time Function of channel A
b(t) -
Enhanced Time Function of channel B
C -
capacitance
c -
scalar component, viscous damper
d -
centroidal distance
dir. -
direction
d0 -
spot diameter from the laser beam expansion
dt -
change in time
dx -
change in length
e -
electron charge
E,2 -
elastic modulus
F -
force
fd -
Doppler frequency constant as a function of frequency, cut-off
frequency
flo -
frequency shift in the mixer of the Frequency Shifter
fn -
nth natural frequency
(/: //. ) "
frequency expansion ratio of the laser vibrometer
VI
Oaa -
Autospectrum, 1 -sided, Ch. A
Gab -
Cross spectrum, 1 -sided
Obh -
Autospectrum, 1 -sided, Ch. B
^ -
Hilbert transform function
H -
hysteretic damping constant
h -
Plank's constant
ho -
energy of quantum light of frequency u
H(f) frequency response function
H(s) -
compliance transfer function in the Laplace domain
h(t) -
impulse response function
h,2 -
element heights
HI -
frequency response Oab/Gaa
H2 -
frequency response Gbh /Gha
Hz. -
Hertz
h(t) -
analytical time signal of the impulse response function
1, 2
-
element Inertia
/
-
imaginary part
K,k -
spring constant, Boltzmann's constant, Tracker slew rate constant
kl wave vector going away from the laser

-
k2 -
wave vector reflected towards the laser
/ -
object length, length of cantilever beam
vii
/ -
depth of focus on the laser vibrometer
L, L2,
-
optic lengths in the laser vibrometer
/'
-
new determined length
m -
mass, integer time sample, meters
mm -
millimeters
mph -
miles per hour
N -
Newton
n -
scalar value in the powerlaw, number of time samples
pC -
pico charge
P, P,,
-
Si PIN diodes in light detector circuit
Q -
Quality factor, charge
R -
drag stress, resistor
rms -
root mean square
R,
-
TV lens plate
s -
time in seconds
s. ,
-
complex conjugate poles
signal to noise ratio of the photodetectors

S/Nrb -
t -
time
T -
time record length, temperature
V -
velocity, voltage
V -
fixed velocity offset in the Frequency Shifter
Vlll
Vz -
velocity component in direction of optical axis of the laser
w -
element width
w(t)
-
time weighted function
X -
amplitude, optical path length in laser vibrometer
x -
distance
W(f) -
Fourier Transform of the time weighted time function
Xst -
maximum static deflection
Y -
optical path length in laser vibrometer
i -
velocity
x -
acceleration
5 logarithmic decrement
tsfn -
change in 3db frequency bandwidth of the resonant frequency
Awr change in circular natural frequency
A/ -
change in frequency
AW -
change in potential energy
At -
change in time
s -
stain
ev
-
viscous strain
ev
-
time rate of viscous strain
damping ratio
ieq
viscous damping ratio
IX
e angle
A wave length
n loss factor, quantum efficiency detectors
a axial stress, decay constant
a,me decay constant
cx
-
decay constant of the exponential weighted function
x -
time constant
rw
-
time constant of the exponential weighted function
Tm
-
measured time constant from the impulse response magnitude
function
a>D
-
Doppler shift frequency
cul -
Doppler shifted laser beam
a>'r
-
new natural circular frequency
go,
-
non-Doppler shifted laser beam
(ono
-
resonant circular frequency
Q -
solid angle subtended by the producing illuminated laser light
d max
-
maximum slope of the Tracker slew rate
dt
F -
degrees Fahrenheit
y. -
dynamic compliance
LIST OF ILLUSTRATIONS
Figure 2. 1 Hysteresis Loop 4
Figure 2.2 Effect of

Damping on a Free Response 5
Figure 2.3 Amplitude Decay of the First of Mode of an Aluminum 7

Cantilever Beam
Figure 2.4 Frequency Response 8
Figure 2.5 Variation of Storage Modulus and Loss Factor with
Temperature 8
Figure 2.6 Variation of Loss Factor and Real Part Storage Modulus
with
Frequency 9
Figure 2.7 One Degree of Freedom of a Spring- Viscous Damper System 16
Figure 2.8 One Degree of Freedom of a Spring- Hysteretic Damper

System 16
Figure 2.9 Free Vibration with Hysteretic Damping 17
Figure 2.10 Summary of Lightly Damped Formulas 20
Figure 2. 1 1 Plot of Impulse Response Magnitude and Real 23
Figure 3 1
. Bruel and Kjaer's Flow Chart of Dual Channel Spectrum
Averaging 25
Figure 3.2 Leakage in the Analysis 26
Figure 3.3 Basic Diagram of Experimental Setup for Testing Vibrating

Beams 31
Figure 4. 1 Doppler Shift of a Vibrating Surface by Dantec Electronics 33
Figure 4.2 Doppler Shifted and Non-Shifted Beam by Dantec

Electronics 34
Figure 4.3 Side View of the Laser Vibrometer Component

Identification by Dantec Electronics 34
XI
Figure 4.4 Optics Path of the 55X Laser Doppler Vibrometer by Dantet
Electronics 35
Figure 4.5 55X Laser Doppler Vibrometer with the 55N 10 Frequency
Shifter and 55N20 Doppler Frequency Tracker by Dantec
Electronics 36
Figure 4.6 Basic Laser Setup 37
Figure 4.7 Front Display of the Frequency Tracker and the Frequency
Shifter by Dantec Electronics 40
Figure 4.8 Laser Alignment by Identification of a Square Wave Signal 41
Figure 4.9 Pre-alignment Wave Form Signal 41
Figure 4. 10 Doppler Frequency vs. Amplitude by Dantec Electronics 42
Figure 4.11 Block Diagram of the Laser Vibrometer Equipment by

Dantec Electronics 43
Figure 4. 12 TV Lens Shift Pattern by Dantec Electronics 45
Figure 4.13 Side View of Laser with Adjustment Locations by Dantec

Electronics 45
Figure 4. 14 Vertical Test Fixture 46
Figure 4. 15 General Overview of the Laser Configuration used in Impact

Excitation 47
Figure 4. 16 General Setup of the Laser in Application with
the Shaker Table 48
Figure 4. 1 7 Effects of Poling 49
'
Figure 4.18 Poling of Plates 50
Figure 4. 19 Piezoelectric Voltages from Applied Forces 50
Figure 4.20 Piezoelectric Ceramic Configuration 53
Xll
Figure 4.2 1 General Setup for the Piezoelectric Ceramic 53
Figure 4.22 Top View of the Shaker Table with Piezoelectric Ceramic 53
Figure 4.23 ATOCHEM Piezo Film 55
Figure 4.24 Equivalent Circuit for the Piezo Film 57
Figure 4.25 The Berg Preamplifier 58
Figure 4.26 General Setup for the Piezo Film 59
Figure 4.27 Top view of the Shaker Table with Piezo Film 59
Figure 4.28 Planar Shear and Delta Shear Accelerometer Designs by

Bruel & Kjaer 61
Figure 4.29 Centre Mounted Compression Shear Accelerometer

Design by Bruel & Kjaer 61
Figure 4.30 General Setup of the Impact Excitation for the B&K 4344
Accelerometer 63
Figure 4.3 1 Conditioning Amplifier Type 2626 64
Figure 4.32 Top view of the Shaker Table with B&K 4344 Accelerometer 64
Figure 5. 1 Bruel & Kjaer Impact Hammer Type 8202 66
Figure 5.2 General Overview of an Impact Excitation Setup 67
Figure 5.3 Force Impulse Plot 68
Figure 5.4 Random Noise Signal 73
Figure 5.5 General Setup of the Bruel and Kjaer's Shaker Table 75
Figure 5.6 Pseudo Random Noise Plot 81
Figure 6. 1 Cantilever Beam with an End Load 87
Xlll
Figure B. 1 Accelerometer Impulse Excitation Frequency Response
-
of the first two Modes with the Analyzer 111

Setup
Figure B.2 Accelerometer Impulse Excitation Frequency Response

-
of the first two Modes with the Coherence Function 1 12
Figure B.3 Accelerometer -

Impulse Excitation Frequency Response
of the First Mode with the Analyzer Setup 1 14

of the First Extracted Mode with the Coherence Function 1 15

Impulse Excitation Impulse Response
Function with the Analyzer Setup 116

of the Second Mode with the Analyzer Setup 118

of the Second Extracted Mode with the Coherence Function 1 19


of the Second Mode with the Coherence Function 122
Figure B. 10 Accelerometer -
Figure B. 1 1 Laser -
of the first two Modes with the Analyser Setup 125
Figure B. 12 Laser -
of the First Mode with the Coherence Function 127
Impulse Excitation Frequency Response of the First
Extracted Mode with the Impulse Response Function 128
Impulse Excitation Frequency Response of the Second
Mode with the Coherence Function 130
XIV
Figure B.15 Laser Impulse Excitation Impulse Response Function
-
with the Analyzer Setup 13 1
Figure B. 16 Laser Impulse Excitation Frequency Response

-
of the
Second Mode with the Analyzer Setup 133
Figure B. 17 Laser Impulse Excitation Frequency Response

-
of the Second
Extracted Mode with the Coherence Function 134
Impulse Excitation Impulse Response Function

with the Analyzer Setup 135
Impulse Excitation Decaying Oscillation of the

Cantilever Beam (Top) and Force Pulse from an Impact
Hammer (Bottom) 136
Figure B.20 Piezoceramic -

of the first two Modes with the Analyzer Setup 138
Figure B. 21 Piezoceramic -
of the first two Modes with the Coherence Function 139

of the first Mode with the Analyzer Setup 141

of the First Extracted Mode with the Coherence Function 142


Function of the Second Mode with the Analyzer Setup 145

of the Second Extracted Mode with the Coherence Function 146

Figure B.28 Piezofilm -

XV

Figure B.30 Piezofilm Impulse Excitation Frequency Response

-
of the First Mode with the Coherence Function 1 52
Figure B.3 1 Piezofilm -


of the Second Extracted Mode with the Coherence Function 155
Figure B.3 3 Piezofilm -

Figure C. 1 Accelerometer -
Random Excitation Frequency Response
Figure C.2 Accelerometer -


Random Excitation Impulse Response



Figure C.7 Laser -

Figure C.8 Laser -

Figure C.9 Laser -

XVI
Figure CIO Laser -
Figure C. 1 1 Laser Random Excitation Frequency Response

-
Figure C. 12 Laser -

Figure C. 13 Piezoceramic -
Figure C. 18 Piezofilm -
of the first two Modes with the AnalyzerSetup 1 79
Figured 9 Piezofilm -
Figure C.20 Piezofilm -

Function with the Analyzer Setup 18 1
Figure C. 21 Piezofilm -
Figure C.22 Piezofilm -

Figure D. 1 Accelerometer -
Impact Excitation Frequency Response
of the Undamaged Composite Beam with the
Coherence Function 185
XVll
Figure D.2 Accelerometer Impact Excitation Frequency Response
-
of the Damaged Composite Beam with the
Coherence Function 186
Figure E. 1 Frequency Response of the first three Modes

Artificially Damped Second Resonant Frequency
Beam Length =
10 inches 188
Figure E.2 Frequency Response of the first three Modes with the
Coherence Function
Artificially Damped Second Resonant Frequency
Beam Length =
10 inches 189
Figure E.3 Frequency Response of the First Mode with the Coherence
Function, Beam Length =
10 inches 190
Figure E.4 Impulse Response Function with the Analyzer Setup

Beam Length =
10 inches 191
Figure E.5 Frequency Response of the Second Mode

with the Coherence Function
Beam Length =
10 inches 192

Beam Length =
10 inches 193
Figure E.7 Frequency Response of the Third Mode

Beam Length =
10 inches 194

Beam Length =
10 inches 195
Figure E.9 Frequency Response of the first five Modes

Artificially Damped Third Resonant Frequency
Beam Length =17.5 inches 196
Figure E. 10 Frequency Response of the First Mode with the Coherence
Function, Beam Length =

17.5 inches 197
Figure E. 1 1 Impulse Response Function with the Analyzer Setup

XV Ul
Figure E. 12 Frequency Response of the Second Mode
Figure E. 13 Impulse Response Function with the Analyzer Setup

Beam Length 17.5 inches
=
200
Figure E. 1 4 Frequency Response of the Third Mode

Beam Length =
17.5 inches 201

Figure E. 16 Frequency Response of the Fourth Mode

with Coherence Function
Beam Length =
17.5 inches 203

Beam Length =
17.5 inches 204
Figure E. 18 Frequency Response of the first five Modes

Artificially Damped Fourth Resonant Frequency
Beam Length =
24.625 inches 205
Figure E. 19 Frequency Response of the First Mode with the Coherence

Beam Length =
24.625 inches 206

Beam Length =
24.625 inches 207
Figure E.21 Frequency Response of the Second Mode

Beam Length =
24.625 inches 208

Beam Length =
24.625 inches 209
Figure E.23 Frequency Response of the Third Mode

Beam Length =
24.625 inches 210
XIX
=
21 1
Figure E.25 Frequency Response of the Fourth Mode

withCoherence Function
=
212

Beam Length =
24.625 inches 213
Figure E.27 Frequency Response of the Fifth Mode

Beam Length =
24.625 inches 214

Beam Length =
24.625 inches 215
Figure F. 1 Frequency Response of the first three Modes with
the Analyzer Setup 218
Figure F.2 Frequency Response of the first three Modes with
the Coherence Function 219
Figure F.3 Frequency Response of the First Mode with
Figure F.4 Impulse Response Function with the Analyzer Setup 221
Figure F.5 Frequency Response of the Second Mode with
Figure F.7 Frequency Response of the Third Mode with
Figure F.9 Frequency Response of the First Mode with
XX
Figure F. 10 Impulse Response Function with the Analyzer Setup 229
Figure F. 1 1 Frequency Response of the Second Mode with
Figure F. 13 Frequency Response of the First Mode with
Figure F. 15 Frequency Response of the Second Mode with
Figure G. 1 Frequency Response of the First Mode with
Figure G.2 Impulse Response Function with the Analyzer Setup 239
Figure H. 1 Frequency Response of the first two Modes using a bread

board, combination of resisitors, and without grounding 24 1
Figure H.2 Frequency Response without grounding 243
Figure 1. 1 Climatological Data of Rochester, New York August

1992 Page 1 of 4 245
Figure 1.2 Climatological Data of Rochester, New York August

1992 Page 2 of 4 246

1992 Page 3 of 4 247

1992 Page 4 of 4 248
Figure 1.5 Climatological Data of Rochester, New York September

1992 Page I of 4 249
XXI
1992 Page 2 of 4 250

1992 Page 3 of 4 251

1992 Page 4 of 4 252
Figure J. 1 Operating Window of the Dantec Laser Doppler

Vibrometer (Frequency vs. Amplitude of Harmonic Motion) 254
Figure K. 1 Example 1 . Variation of the Superimposed Frequency

Responses 256
Figure K.2 Example 2. Variation of the Superimposed Frequency

Responses 256
Figure K.3 Example 3. Variation of the Superimposed Frequency

Responses 257
Figure K. 4 Example 4. Variation of the Superimposed Frequency

Responses 257
xxn
LIST OF TABLES
Table 2. 1 Experimental data for Mode 1 of the Undamaged

Composite Beam Date: August 21
-
11
Table 2.2 Experimental Data for Mode 1 of the Undamaged

Composite Beam Date: September 8
-
12

Composite Beam Date: August 21
-
12

Composite Beam -
Date: August 27 12
Table 2.5 Delta Frequency verses Delta Loss Factor per Mode 13
Table 2.6 Analytical Solutions for the Laminated Composite

Graphite Epoxy Beam 13
Table 2.7 Comparison of Analytical and Experimental Natural

Frequencies 14
Table 6. 1 Experimental Determination of the Loss Factor for

the First Natural Frequency 83
Table 6.2 Experimental Determination of the Loss Factor for

the Second Natural Frequency 84
Table 6.3 Determination of Artificial Damping given by the

Shaker Table in the vicinity of 236 Hz. 88
Table 6.4 Rectangular Window & Pseudo Random Excitation vs.
Hanning Window & Random Noise Excitation 90
Table 6.5 Random Impact Excitation using a Modal Hammer 92
Table 6.6 Comparison of Testing Results for the First Mode 95
Table 6.7 Transient Window applied for Impact Testing 96
Table 6.8 Exponential vs. Transient Weighting Windows in the

Loss Factor Calculation 97
XXlll
Table A. 1 Parameters for the Aluminum Cantilever Beam 05
Table A.2 Parameters for the Aluminum Cantilever Beam 106
Table B. 1 Data Sheet # 1 for the Impulse Excitation (Transient Window) 1 13
Table B.2 Data Sheet #2 for the Impulse Excitation (Transient Window) 1 17
Table B.3 Data Sheet #3 for the Impulse Excitation

(Exponential Window) 121
Table B.4 Data Sheet #4 for the Impulse Excitation (Transient Window) 126



(Exponential Window) 15 1
Table B. 10 Data Sheet #10 for the Impulse Excitation

Table J. 1 Tracker Unit and Frequency Shifter Unit Settings 254
Table J.2 Operational Data for the Laser Vibrometer 254
Table J. 3 Tracker Unit Settings 254
XXIV
1. INTRODUCTION
The experimental study of structural vibrations has evolved tremendously

over many years and the measurement equipment continues to grow with the ever
increasing technology of electronics and computers. With the improvement of this

technology, the measurement objectives are made easier and the ability to more
closely control conditions in the experimental test consequently yield more
accurate results as well as detailed information. The experimental testing including

the data acquisition and subsequent analysis is sometimes referred to as "Modal
Analysis"
Testing", "Modal and "System Identification". This is the process of
characterizing the dynamic properties of mechanical structures by identifying their

modes of vibration. The modes of vibration are characterized by the natural
frequency, damping factor, and mode shape [1,20].
Modal testing can be used to compare the corresponding vibration
measurement data to the data produced by a finite element or other theoretical

model. This application is intended to validate the theoretical model prior to
further stages of analysis or to use the acquired modal test data in producing a
mathematical model.
Generally accurate data for each mode of vibration is all that
is required such as its natural frequency, damping and mode shapes to validate the
model. It is generally not possible to predict the damping in each mode of

vibration from a theoretical model so it can be compared to the experimental
damping measurements. However, such information becomes very useful as it can
be incorporated into the theoretical model prior to any predictions. Since damping
plays a crucial role in the dynamic behavior of a system, such attempts are usually
made to adjust or correct the theoretical model in order to bring its modal
properties closer in line with the measured results [20].
The investigation for knowing the damping in composite material was of
interest to Dr. Ghoneim at the Rochester Institute of Technology in Rochester,

New York. Dr. H. Ghoneim, S. Griffin and B. Yost at Phillips Laboratory in
Lancaster, Calfornia, had conducted an analytical/experimental study of mode
dampings in a cantilevered laminated beam. Their goal was to predict analytically
the structural damping of an undamaged and damaged controlled beam and
compare this to the experimental structural damping. In conclusion to their work,
the need for adjustments and of the analytical model as well as the
refining
verification of the experimental data was indicated [10].
The initial objective was to determine experimentally the material damping
of these laminated graphite epoxy beams so it can be used in their theoretical
models to describe the loss factor. The loss factor is another way of simply
describing the damping factor. Later in the investigation, the can ;red
laminated beam was abandoned instead, the aluminum beam was ch(
,and a as a
substitute. The reason for excluding this material will be discussed in C! pter 2,
Section 2.3 of Material Section. In order to calculate the damping in the . iaterial,
the response characteristics had to be created which required the excitation of the
beam. The method of exciting the beam was an impact excitation technique and
for an additional
learning experience, I wanted to incorporate the shaker table. In
using the shaker table, the main focus of excitation was using random excitation,
and in the latter part of my thesis I will discuss the purpose of studying pseudo
random excitation. The sensory part of this investigation was the application of
sensors that would beused in
measuring the vibration to see how they effected the
calculation damping. Some of these sensors are rather new and offer many
of
unique advantages for specific applications such as the piezoceramic, piezofilm
and the laser vibrometer. The fourth sensor, the accelerometer, has been around for
some time and is the most commonly used source of sensing vibration. The
compiled data of the damping results versus each sensor and excitation type are
presented and discussed.

The in calculating the
analyzer used
damping is the Bruel and Kjaer Dual
Signal Analyzer Type 2032 and it became available to the market in 1983. This
horse"
analyzer has many advanced features and is still considered a "work I
found the analyzer was capable of calculating the damping in several manners, and
within the methods investigated, the most accurate found was using the Impulse
Response Function. This function will be discussed in greater detail in the latter
part of this thesis as well as the analyzer itself.
The reference materials damping were very useful in the
that I collected on
understanding dampingof many interesting insights to the effects of

and offered
damping. Not only did I benefit from learning about structural damping but gained
valuable experience on the theory and practice of modal testing. I also learned that
there are other types of excitation techniques which offer many advantages and
disadvantages when used in an application. The most interesting part of this work
was applying the sensors in the mobility tests to see the effect under excitation. A
tremendous amount of work went into getting these sensors to operate properly
such as the laser vibrometer and the piezofilm. Many hours of trouble shooting and
research went into these two particular sensors which, in the end paid, off with
excellant results. Each sensor offers many unique advantages which can provide
great benefits in its application. The other challenge was getting the measurement
system to all work together using the source input with the sensory output to gain
the frequency response data. This came down to a lot of time spent in preparation
and pretesting of the setup in order for it to operate correctly before

experimental
the data could actually be taken. The patience, care, reference materials and the
consultation with experiencied people in the field along with the trial and error in
the experimental testing, all brought together not only the understanding of
concepts, but an intuition as well as creativity in building a successful \ ions
test.
The presentation of the thesis work will proceed with damping; th 3econd
chapter and the third chapter will be about the background of modal test g using
the analyzer withthe proper windowing techniques. Chapter four is ci ncerned
with the background of the four types of sensors used in the investigation, and the
two excitation techniques is found in chapter five. Chapter five will include the
experimental procedure for calculating the damping. Following will be the results
and discussion of the data generated along with the problems encountered during
the setup. Finally, the comments and recommendations of this work will be
presented.
2. DAMPING
2. 1 Characteristics of Damping
The damping is the removal of energy from a vibrational system.

nature of
The energy is lost by some mechanism which is gradually converted into heat or
sound. These damping mechanisms can
generally be classified into three groups:
Viscous Damping, Coulomb
Dry Friction Damping, and Material, Solid or
or
Hysteretic Damping. Viscous damping is the most commonly used model in
representing the damping mechanism of a vibratory system. Viscous damping

involves fluid type systems such as a fluid film sliding between surfaces or even
fluid flowing through an orifice. The amount of energy dissipated in a viscous
model will depend on many factors, but the damping force is related
proportionally to the velocity of the vibrating body. Viscous damping is generally

the model of choice when describing a mechanical system that vibrates in these
fluidic systems. Coulomb or dry friction damping is caused by the friction of two
surfaces rubbing against one another which is dry or lacks lubrication. The
damping force is constant in magnitude but opposite in direction to that of the

motion of the vibrating body. The dry friction force is proportional to the normal
force of the body weight of the system. Lastly, material, solid or hysteretic
damping is theenergy absorbed by the material and dissipated. As the material

such as a cantilever beam is deformed, friction between the internal planes of the
material slip and slide creating what is called hysteresis damping. The hysteresis
loop is the stress-strain relationship created from the loading and unloading of the
vibrating system. This is described in the following sketch where the area of the
loop denotes the energy lost per cycle due to solid damping. Many materials have
shown through
Hysteresis Loop
Stress (force)
Unloading
* Strain (displacement)
A A
Loading
Figure 2. 1
experiments that the energy loss per cycle due to internal friction is approximately
proportional to the square of the displacement amplitude and independent of the
frequency, providing the system is completely linear [16,17].

Some of the precise mechanisms associated with internal changes of the
structure either at the micro or macro level can range from the crystal lattice to the
molecular scale effects. These include magnetic effects creating magnetoelastic
and magnetomechanical hysteresis, thermal effects involving thermal conduction,
thermal diffusion, thermal flow and thermoelastic phenomena and atomic
reconstruction. The atomic reconstruction may involve such effects as dislocations,

concentrated defects of crystal lattices, phonoelectronic effects, stress relaxation at
the grain boundaries, phase processes in solid solutions and blocks in
polycrystalline materials. The list goes on at the atomic level but regardless of the
precise physical mechanisms involved, all materials dissipate energy no matter
how much cyclic deformation occurs [16].

Viscous damping and material damping are similar in such observations as
the decaying amplitudes of a cantilever beam. When the

decaying envelope of the
oscillation is observed to be of an exponential damping is then curve, this nature of
considered unique. As seen below in Figure 2.2, coulomb damping is found to be
constant in its successive oscillations of amplitudes where the viscous decay does
EFFECT OF DAMPING ON A FREE RESPONSE
Displaced position
Time
Equilibrium position
a) Undamped Beam
Envelope IV (M
-3
Time
' ,i }) }/
b) Beam with Viscous Damping
S^
A
*-
Time
c) Beam with Dry Friction Damping

Figure 2. 2
Damping"
NahifT, Jones, & Henderson "Vibration
not. Figure 2.3 on the next page illustrates the decaying amplitude of an aluminum
cantilever beam. This was performed
using the B&K 2032 Digital Signal Analyzer
and sensed Dantac's Laser Vibrometer. As seen by the amplitudes, it is evident
by
exponential decay. With this information on the background of
of an damping, it is
important to identify what type of
damping you have in order to correctly measure
it. In the case of the cantilever beam, the damping would be considered material
damping or hysteretic. For viscous and material damping, the measure of damping
will be the same and is unique for these two types. The measurement of the
damping for these types has been observed to follow this relation.
]- (Eq.
-2
2.1)
(O0
The resonant frequency of the structure is represented by co0. co, and co2 are the
frequencies situated on either side of eo0. The co, and co, frequencies are located
1/V2 times the amplitude of the resonant frequency. The r\ is known as the loss
factor and is by inspection dimensionless. The loss factor indicates the amount of
energy lost from the vibratory system in one cycle of vibration. The dry friction
damping can not be characterized by this measure because it is dependent on the
amplitude. The resonant frequency can be described as the compromise between
stiffness of the structure and its mass. At the low frequency side of the resonant,
the stiffness, will provide the dominant restoring force where as the high
frequency side will be the inertial force. Depending on the precise values of mass
and stiffness, a resonance will occur and the system will have a number of natural
frequencies occurring. At the precise natural frequency along with the absence of
damping there will be no dynamic equilibrium and the system will ever increase in
oscillating amplitudes. In reality, an excited system will attain a final steady state
of dynamic equilibrium because there will always exist some amount of damping
that will attain either a high or low amplitude. Overall damping is responsible for
controlling the vibration response amplitude in conditions of steady-state
resonance and stationary random excitation. One way to identify high damping or
low damping at a resonant frequency is to inspect its response amplitude. This
observation can be seen in Figure 2.4 on page 7. The frequency response suggest
a zero damping, low damping and a high amount of damping at the natural
frequncy coo. [16,17].
2.2 Environmental Effects
The environment plays an important role in some of the variation of
damping properties and an understanding of these interactions can help make a
vibration test successful. The essential effect of the environment on materials can
0 0
(DOZN
o OQ\ E
E ? Q>\ UJ
cn a . E>
3 a .csatM J
a i a. o. cmô . <
N UJ Z UJ ro . . .to >
Q CD u IIhOJ
- 0 u
Hl- (J
Q X 8< UU UJ
O < E ZZ N N a
-
1 UJUIXX
JJ
< _J_I
m e in -d (0(0
. ?< mm o.
u. - o cucm tn
UJ UJ ~ is
.x>
+ .
KX< E E a:
i-i UJ u.
U_l N OOD
o. u.u.aa ui
<Q
JI 0 m Nil N N ^ s*
IUI E
_
.IlfllAII -
D + (D u.o roro cd
< . *
c o _i
I- t t^ < *
J U I Q mi-w
ZH-
< uj a + +Q
Ul UJZ UJ -
tt m < wUJ -I
t uiz m
NSZO <
< .mZ IO<Q-> WCC
Die-' mao:xa>>-<uJ
quh-i CM N I- UJ * Q) D H
UJ
s.
(D 0 <
. CM I- a<tn
I m z .uj . . -<
LID uj u ZCCII Q-
a
E + ^ z - z <u.uu O
uim U1CC " a. t-_I
*
in o aui -u cncci-H<<
M . E 0- DUX UJII ,
,q:h
HCMO D t/)u<a: a ou<omu

r-
r-
<JUJ hizmh . .zuj
f UJ UJIKUJ> otujujuiixujq.
A.' ZHQ< IlUJSUUUUI
(fl
g>-X
FREQUENCY RESPONSE
Zero Damping
Low Damping
High Damping
? oo
Stiffness Inertia
Controlled Controlled
Figure 2.4
cause noise and impede the control of a vibration test especially in critical
designs. The main environmental factors are temperature, frequency, static preload
and dynamic load. Temperature is usually considered to be the single most
important factor affecting the properties in damping materials. Figure 2.5

demonstrates the effect of temperature on the loss factor of a typical (viscoelastic
material) polymer. The figure shows how the loss factor changes with temperature
as it passes through the four different regions. The glassy region has extremely low
Variation of Storage Modulus and Loss Factor withTemperature
"iT-x Tl
Loss Factor n
and
Storage /
Modulus E
// ^^
Glassy Truwxtwn Rubberlike
Region Region Region Flow Region
Temperature
Damping"
Nashiff, Jones & Henderson "Vibration
Figure 2. 5
values for the loss factor where it will increase significantly with increasing
temperature. The transition region contains the maximum value for the loss factor.
The rubberlike region suggests the low values for the loss factor and that changes
slowly with temperature. The loss factor increases as the material tends to soften
with increasing temperature in the flow region [3,16].
The next environmental factor is the frequency. The loss factor will
increase with
increasing frequency in the rubbery region and will take on a
maximum value in the transition The increase in the glassy
region.
frequency will
region as the loss factor decreases. I have also provided a sketch of the loss factor
in the frequency domain as seen in Figure 2.6. The interesting thing to note is that
qualitatively this reveals the inverse behavior of the temperature domain [16].
Variation of Loss Factor and Real Part Storage Modulus with Frequency
E
Loss Factor T)
and
Tl
Real Part
of the
Storage
Modulus E
Frequency (log scale)
" "
Nashiff, Jones & Henderson Vibration Damping
Figure 2. 6
The effect of cyclic dynamic strain and static preload are also another
concern. In the dynamic strain, the amplitude on the damping properties of

case of
material become very difficult to measure. This is due to the high strain amplitudes
in the material which dissipate high amounts of energy causing the temperature to
rise so fast that the effect of the temperature is coupled. Because of this problem,
only the rubbery region has been studied where the loss factor varies more slowly
with temperature becomes secondary to that of the dynamic strain amplitude.
and
The transition region does have the effect of temperature coupling the strain
amplitude because of the high damping which makes any measurements in this
region difficult. The effects of preload on the dynamic properties will forecast
problems in the rubbery region because the loss factor decreases and the modulus
increases with increasing preload. Both the loss factor and the real modulus are
linear over the frequency domain. In general the loss factor r\ is dependent on both
the amplitude and frequency. Other environmental effects that may affect the
damping properties are aging, vacuum, oil exposure, pressure, high temperatures
and even radiation. The rubberlike materials are often affected by these conditions
[3,16].
2.3 Material Selection
In this work, three materials were examined but only one was used
throughout the entire scheme of the thesis. The first material was 1018, cold rolled
steel, which was used as a dummy beam to practice with. The other material was a
cantilevered laminated beam. This material was of interest to Dr. Hany Ghoneim
at the Rochester Institute of Technology in
determining dampings. The the mode
experimental mode dampings were to be compared with the predicted dampings
from the developed analytical model. Mode
damping is the damping that
corresponds to each individual natural
frequency of the structure. The added
challenge to Dr. Ghoneim's analytical model was another beam of the same
material and size and it was made with local controlled

"damage"
The mode
dampings for both these types of beams were to be predicted analytically as well
as experimentally. Unfortunately, the composite beams were excluded from this
investigation because the experimental determination of the damping was unstable
in its testing environment. The daily change in temperature was found to be

affecting the damping property of the material which made
comparing the other
sensors difficult. The third type of material was an aluminum beam selected as a
substitute because its material damping was more stable in the testing
environment. The aluminum material offered two advantages in the modal
testing
over the composite material, namely its conductive and reflective properties [10].
Before I became aware of the sensitivity of damping with respect to
temperature changes, the experimental setup was evaluated to ensure that nothing
in the setup was causing the damping to change. The determination of the loss
factor was confined to the impact excitation method with the same windowing
configuration and accelerometer for each test. The torgue applied to the fixture
holding the cantilever beam was maintained for each test and it was found that
small changes in the amount of torque holding the composite beam affected the
damping. A fixture was built for holding the impact hammer in place so that the
impacts were consistent with each trial. Lastly, a steel beam was tested in the same
manner as a material check and the loss factor was consistent for each test. During
this effort, literature concerning viscoelastic materials was being researched and
concluded that the damping for these types of materials will change with
temperature. After learning about the environmental affects, anything that could
create any air currents was minimized in the laboratory such as closing any open
windows or doors. Dr. Budynas lead me to several individuals with whom to

consult. One was Mark Hammer from Eastman Kodak Company who was working
in the Federal Systems Division. From our discussions, I learned that work was
conducted in determining the loss factor for similar materials. They were even
using an environmental chamber and were still having problems with the stability
of the loss factor. After working on the project for about a year, it was abandoned.
Atmospheric damping was one of the suspicions that may have caused the
10
instability of the loss factor. I also learned that there was a Standard Method for
Measuring Vibration-Damping Properties of Materials given by the ASTM
standards E 756-83. This gave a procedure in how to setup such a test and how to
calculate the loss factor. The of the loss factor
calculation as given by the
standards is the same as equation 2.1, n Aco^ The problem with

following this
co
procedure is that I would need an environmental chamber in order to control such
factors as temperature and pressure [9, 16].

The mode dampings for the laminated composite graphite epoxy beam are
presented in the following tables. All the data in these tables determined
was by
the use of an accelerometer and the impact excitation method. Table 2.1 seen
below describes aday in Rochester, New York that was nice with
plenty of
sunshine. On that day the temperature happened to be as high as 76F and as low
as 49 F where the loss factor of the undamaged laminated composite graphite
beam varied from .09556 to .01458. All the weather related data was obtained
from the monthly summary of local climatological data reports from the Rochester
Airport. The months June, July, August, and September were requested and the
months for August and September can be found in Appendix I. On the third page
of each monthly local climatological data report shows the change in temperature
every three hours within that day, and the first page gives a one line summary of
that day's behavior.
TABLE 2. 1
EXPERIMENTAL DATA for MODE I of the Undamaged Composite Beam
DATE: AUGUST 21
Frequency (Hz) Loss Factor Temperature (F) Wind (mph) Weather Conditions
23 max =
.09556 max =
76 dir =
SW Clear Skies
min =01 458 min =
49 peak gust =
10 100% Sunshine
ave
=
.04993 ave
=
63 ave =
6.3
Table 2. 1 illustrates the point that the temperature did change a lot during that day
and night and effected my results. Table 2.2 shows the data that was taken on
September 8th in which the frequency changed to 22.375 Hz. and the loss factor
only varied from .08058 to .07062. The observation from the two data tables
suggests to me that with the increase in the average temperature for those two
days, the average loss factor rose from .04993 to .07607. The total change in
frequency throughout all the data gathered was 1 Hz. ranging from 22 to 23 Hz.
Also the loss factor had changed a total of .084 throughout the study of the first
mode ranging from .0388 to . 1228.
n
TABLE 2.2
EXPERIMENTAL DATA for MODE 1 of the Undamaged Composite Beam
DATE: SEPTEMBER 8
22.375 max =
.08058
max =
81 dir =
SW FOG
=
min .07062 min =
57 peak gust =
36 62% Sunshine
ave =
.07607
ave =
69 ave=
12.4 Rain Showers
The displayed the same effect with the increasing temperature

second mode
along with increase

the in the loss factor from the dates of August 21 and August
27. Another interesting observation from among all the tables is that the day it had
rained, thus a change in the humidity, may have been a contributing factor to these
changes in the loss factor. Nonetheless, the total change in
frequency from all the
data gathered was 3 Hz. and ranged from 155 to 158 Hz. The change in the loss
Table 2.3
DATE: AUGUST 21
156.5 max =
.004780 max =
76 dir =
SW Clear Skies
min =
.002979 min =
49 peak gust =
10 100% Sunshine
ave =
.00383 ave =
63 ave =
6.3
Table 2.4
DATE: AUGUST 27
157 max =
.006732 max =
84 dir =
NW Fog
min =
.0027348 min =
69 peak gust =
31 20% Sunshine
ave =0043 18 ave =
77 ave =
5.0 Thunder Storms
factor over all the data indicates .003024 and ranged as low as .003840 to a high of
.006864. The third mode ranged in frequency from 452 to 456 Hz., and the loss
factor ranged at a minimum of .00342 to .00420. From all of the three modes of
data that taken, I noticed a trend in the total change in the loss factor over the
was
change in frequency. The change in frequency had increased with the increase in
the mode number while the change of the loss factor became tighter.
12
Table 2.5
Delta Frequency verses Delta Loss Factor per Mode
Mode# A Frequency (Hz.) A Loss Factor

1 1 0.0840
2 3 0.00302
3 4 0.000780
Included the undamaged and damaged composite beams and

are a plot of
they located
are in Appendix D. The frequency response HI and H2 will be
discussed in some detail in the modal testing section, but
Frequency Response H 1
Magnitude was used throughout the
aluminum cantilever beam study as it was for
the undamaged composite beam. The data concerning the damaged composite
beam was very little as this was the only plot I happened to take. In the
early
stages of learning the analyzer's functions, I had examined the damaged beam in
the H2 mode and in the dynamic compliance mode

(l/yco) which is
(displacement/force). When there is no integration symbol
J in this spot as it is
seen in this plot, it means that you are in a mobility mode or velocity/force
receptance. I also placed the analytical solutions for
determining the natural
frequencies of the nondamaged composite beam in Appendix A. The analytical
solutions for the laminated composite graphite epoxy beam are summarized in
Table 2.6.
Table 2.6
Analytical Solutions for the Laminated Composite Graphite Epoxy Beam
Mode of Cantilever Cantilever Beam Cantilever Beam with Cantilever Beam with
Natural Beam using Singiresu S end mass of 2 7 grams end mass of 0.009579
Frequency with no end Rao's Solution grams
mass (Hz.) (Hz.) (Hz.)

(Hz.)
1 31.21 31.21 22.26 31.21
2 195.68 195.60 164.75 195.43
3 547.73 547.7 480.35 547.17
Comparing the 2.7 gram analytical solution, which represents the mass of the
accelerometer, with the experimental data of the natural frequencies, it is observed
by the Table 2.7 that the percent difference is not bad at all. In order to use the
laser vibrometer on the composite beam, a piece of reflective material was needed
to be attached to the end of it since this material was not reflective. The analytical
solution pertaining to the mass of .009579 grams is the mass of a square piece of
13
reflective material.
By comparing the analytical solution of the Cantilever beam
with no end mass to the solution with the mass of the reflective material, it shows
that this amount of mass beam.
will not effect the real natural frequency of the
Table 2. 7
Comparison ofAnalytical and Experimental Natural Frequency
Average Experimental (Hz.) Analytical (Hz.) Percent Difference (%)

22.5 22.26 1.1
156.5 164.75 5 0
454.0 480.35 5.5
Since the steel beam appeared stable in this environment, Dr. Ghoniem
suggested using an aluminum beam instead because it offered a couple of
advantages to some of the Metals
sensors. are not tremendously affected by
temperature. The hysteresis loop for metals is extremely thin and straight, not
curved at the ends. Figure 2. 1 in this section was just for illustration purposes of a
hysteresis loop. The curved ends of the loop would tend to indicate degrees of
nonlinearity. The ASTM standard for measuring damping makes the assumption
that the loss factor is zero when determining the damping properties in composite
bars. They also note that it is an established fact that steel and aluminum materials
have well known loss factors approximately 0.001 or less which is significantly
of
lower than those of the composite bars. With this fact in mind, it was decided to
carry on in the measurement of mode damping and to see how close the expected
value for the loss factor in aluminum will be while exploring the four different
sensors and applying the two types of excitation techniques [9].
2.4 Method of Calculating Material Damping
The determination of loss factor is deduced from the response
characteristics of a vibrating system. In my case, the vibrating system is a
cantilever beam which has an infinite number of degrees of freedom. The

dimensions of the aluminum cantilever beam are 1/8 inch thick by 1 inch wide and
10 inches long. By exciting the beam, the response characteristics produce what is
called a frequency response of the structure or resonances. From these resonances
it is possible to determine the loss factor. There are several ways to calculate
damping and these are just some of the ways: the logarithmic decrement, Nyquist
diagram, half power bandwidth, dynamic stiffness, hysteresis loops, resonant
response amplitude, quadrature bandwidth, and the decay constant. After
investigating some of these options in determining the loss factor, the most
accurate method was the Impulse Response Magnitude Function. The Bruel and
Kjaer Dual Channel Signal Analyzer type 2032 was used to measure the response
of the structure and to calculate the material damping using the Impulse Response
14
Magnitude Function. Limitation on the
accuracy of the loss factor is due to the
analyzer. The B&K Dual Channel Signal Analyzer 2032 is a digital filter analyzer.
For instance, using the half power bandwidth method will become a difficult task
in obtaining the corresponding frequency at the -3db point on either side of the
resonance. Since the analyzer is digital, the frequency at -3db
may have to be
interpolated which
may incur a large error with respect to the true value of the loss
factor. This method requires
knowing exactly what those two frequencies are. The
Impulse Response Method, referred to as the Schroder Method, is the method
suggested by Bruel and Kjaer for this type of analyzer in
determining the loss
factor [5,12,13].
2.4.1 Damping Measurements for a 1-D Model

of a Cantilever Beam
The basic theory of modal analysis for a single degree of freedom

mechanical system begins at the examination of motion in one direction. The
motion can be described by a single coordinate and the equation of motion can be
derived by using Newton's Second Law. In general the system can be idealized as
a spring-viscous damper arrangement as seen in the Figure 2.7. Similarly,
hysteretic damping can be modeled as a spring-hysteretic damper arrangement as
well. The closed loop for both of these models represents the energy dissipated by
the damper in one cycle. The mathematical model for the viscous damper in terms
of dissipated cycle of energy is represented by AW -
yFdx. The force required for
displacement of the viscous system is F = kx + cx where x(t) = Xsin cot. Upon the
substituting the force equation into the dissipated cycle of energy and integrating
in terms of dt the dissipated amount of energy can be expressed as AW kcocX2. -
Due to the similarity between the energy dissipation loops from Figure 2.7 and
Figure 2.8, the hysteresis constant can be defined. Hysteretic damping or material
damping is again caused by the internal planes in the material slipping and sliding
as the deformation occurs. It was found experimentally that the energy loss in a
cycle is independent of due to the internal friction, but it can be

the frequency
approximated as proportional to the square of the amplitude. This observed
behavior was achieved by substituting the coefficient of the viscous damper c by

XJ
the following relationship c =

where H is the hysteresis damping constant and co
co
is the circular frequency. The cycle of dissipated energy for hysteretic damping
X2
can be expressed as AW =
nH [2, 16, 17].
15
One Degree of Freedom of a
Spring-Viscous Damper System
F(t]
M
Mass T c ffi X *-
x(t) loading /
i-
-x:/
C K unloading
Viscous X
Spring c<D
Damper
\\\WA\VA\
Figure 2. 7
One Degree of Freedom of a
Spring-Hysteretic Damper System
F(t)
Stress (force)
M
Mass Loading / Unloading
/
x[tj
J- -* Strain
(displacement)
Hysteretic VLî ^> Spring

Damper
\\\\v\\\\\\\
Figure 2. 8
Because of this known relationship between viscous damping and hysteretic

damping, it is possible to relate the hysteretic constant to the loss factor as well as
to other important relationships. In the idealized hysteretic model as well as for the
viscous model, the spring and the damper being connected in parallel, and for a
general harmonic motion, x =

Xe"D\ the force can be expressed as a complex
stiffness. The force for the viscous case upon substitution of the harmonic motion
is given by F = (k + icoc)x. The hysteretic force-displacement relationship is

LJ
expressed similarly as F = (k+iH)x where c =

was substituted. The loss factor
co
is found from the hysteretic force equation by factoring the spring constant out
TJ
which gives (k + iH) =

k(l + i ) = k(\+iij). The k(\ + ir() is called the complex
k
jj
stiffness of the system and j] =

,
the loss factor, becomes a dimensionless
k
16
constant. The dissipated energy for hysteretic
cycle of damping can be now be
expressed in terms of the loss factor as [16, 17]:
AW=nkriX2
(Eq. 2.2)
The logarithmic decrement is important relationship in relating it to

another
the loss factor. The logarithmic decrement estimates the rate at which the
amplitude of the free damped vibration decays. The logarithmic decrement is
'
8 In
x \ The the jth cycle in the
j
=
expressed as . subscript represents amplitude
X n\
)
of the decay. The response of the system in terms of hysteresis damping can be
expressed using the energy balance equation. The energy balance equation can be
applied in the motion of
lightly damped structures such as the aluminum cantilever
beam when considered to be nearly harmonic in determining the decrease in

amplitude per cycle. By examining the energy loss per cycle from a section of the
decaying amplitude, it is possible to relate this to the log decrement. Figure 2.9
displays a cycle of a decaying amplitude of hysteretic damping and the first peak is
represented by X, [17].
FREE VIBRATION with HYSTERETIC DAMPING
* t
Figure 2. 9
In summing the energies lost from half the cycle between peaks
X, and Xrl,2, it
would give:
W ,
=
W -
AW; -
AW ,
(Eq. 2.3)
W is the
nkvX2
The change in energy lost is AW = and maximum potential energy,
W=VCkX2, stored by the spring during the elastic deformation. Upon substitution
of the above equations, the energy balance is given as:
^kx^^kx^-^knx^-^knx2
(Eq, 2.4)
17
where X -
due to the energy loss per half cycle
2X]
=2X2
nX2
1;2 -k
V2
(Eq. 2.5)
=
A ,h (Eq-2.6)
A,-l,2 \2-kj]
Repeating the energy balance equation for the second half of the cycle.
=
-^kX),vr- ^nkr\x)+Mi ~^KkrlxU (*<! 2-T>
2kXU
and -^-
= 1 (eq. 2.8)
A;+, \2-KT]
By multiplying equation 2.6 and 2.8 together, the amplitude ratio reduces to a
constant in terms of the loss factor.
X 2 + nr\ 2-Kr\' + 2nr\'

; ,
- -
= 1 -t-
7T T7 (Eq. 2.9)
X]+, 2-nr\ 2-nn
The hysteresis logarithmic decrement can be expressed as 5 -

ln(l + k n) = k n and
can be assumedapproximately equal to 7ir) because the loss factor is very small for
lightly damped structures. This analysis assumes the frequency for a structurally
damped system is approximately harmonic [17].
The energy balance equation offers another useful way of evaluating the
damping properties of engineering materials. The energy loss per cycle divided by
the maximum potential energy gives rise to a quantity called the specific damping
is is simply
capacity. The specific damping capacity ^W/w equal to 2 wr\ which
derived from the maximum potential energy and the energy loss per cycle
equations. In solving for the loss factor, it is defined as the ratio of the energy loss
per radian to the total strain energy [2, 1 6, 1 7].
AW
n =
(Eq.2.10)
2kW
18
The resonant response amplitude, also known as the quality factor Q, is
another of in
way measuring the damping a harmonic system. The quality factor
contains two points on either side of the resonance which are called the half power
points. The half power points each have an amplification factor of -j=
which
corresponds to the energy absorbed, AW, through the damping of the system.
These power points each respond at a given that is proportional to the
frequency
square of the amplitude. The half power points in other means are also located -3
db down from the The

resonance peak. frequency difference between the half
power points is known as the bandwidth. The amplification factor or amplitude
ratio is the maximum amplitude divided by the static deflection, , from the
Xst
response of a damped system. For small values of damping, the quality factor can
be expressed in terms of the frequency bandwidth from the power points and the
damping ratio g as:
"
Q = =
(Eq.2.11)
2g co2-co,
The equivalent viscous damping ratio

geq has been found to equal the log
decrement as:
8 r?
= --
m.2.n;
and is related to the loss factor by previously derived relationship of the hysteretic
log decrement. Therefore the quality factor can be resolved into the relationship in
terms of the loss factor as [16, 17];

1 co., -co,
111 = S (Eq. 2. 13)
2Q con
These were some of the basic relationships for calculating material
damping. This was meant to help build a background of how material damping
was developed from the similarities of viscous damping and through the use of the
energy balance methods. On the next page, there is a chart made up of these many
relationships including the ones already discussed that are for systems considered
lightly damped. This chart is from March 1990 article, "Digital Filter vs. FFT
a
Techniques for Damping measurements", from the Journal of Sound and

Vibration. I thought this would be a valuable reference for others who would be
interested in a summary of lightly damped formulas [5].
19
2.4.2 Decay Constant
The decay constant method was chosen to calculate the material

damping of
the aluminum beam utilizing the impulse response and Hilbert transform functions.
Summary of Lightly Damped Formulas
Natural Frequency =
2 f0 (
cj0 K <=
ud fur lightly dumped resonances.)
Unknown
Au =
A/ =
u =
1
=
A <A u T a, -
/.J - 5-
Known
3dB
Au Au Aoj Aoj cou Au; 2 138
Bandwidth au irAu
fir 4.3 I Au
Au [Rad/s]
4t 2u Aoj 7T Au Au cji(
3dB
A/ A/ A/ / 1 6.9
Bundwidth 2ttA/ A/ TA/
2 7TA/ 27.3 A/
A/ [11*1 L 2/ A/ IT Af 7TA/ /
Dumping
frequency A/ / 1 1.1 2*7,
4*7,/ 2/. U 2*7i 516/,
L 1Hz] 2*A 77
Lobs fuctor 1 1 2.2
1)U nL i 4.34 a)7;
n 2 IT/, i; it
2 1 *707? TlL
Fraction of
critical 1 1 1.1
2foj 2f/ n. 2f f 27T/A 869 <jf
damping 2xf
27T/A Kh
r
Quality OJ L L 1 1 2Q 2 2Q 4 3lu, 7T
factor
Q
Q Q 2Q Q 2Q CJ0 / Q
Decay a a 2ct tr 1 6.9 Cf
constant 2(7 a 8 6a a
X 2tT
"77
" 0"
IV'! co
Time
2 1 1 1 1 1
constant T/.-
8 69 i
r 6.9t
T 7TT 2-KT t/t 2jr/r T
T [S] r
Reverbera
13.8 2.2 1.1 2.2 1.1
tion time LTW 6.9 Ao 60 6.9
TA Tw
TM [s] f.TM /Aco /Ace 2.2 *
CO 6.9 a7 LTM
Decay rate D D D D D 4.34u 8 69 60 U

D
[dB/s] 4.34 27.3 54.6 4.34u 8.69 oj0 D 8.69 D 8.69/
The
logarithmic, 5u> if. 5 5
if. 1 6.9
decrement 8.69 / 5 6
2ir X 2tt
5
Figure 2.10
The decay constant a represents the damping coefficient and is the real part of the
complex conjugate poles that comes from the response of the system. The response
of the system is derived applying Newton's second law F=ma and applying the
by
equation of motion to the idealized single degree of freedom model as shown in
Figure 2.8. By summing the forces applied to the mass and equating to a second
20
order differential equation, we have the following equation using the viscous
model.
mx(t) + cx(t) + k(t) =

f(f) (Eq. 2.14)
The next step is to convert the equation

using the Laplace Transform.
ms2X(s) + csX{s) + kX{s) =

F(s) (Eq. 2. 15)
The response of the system is written from the Laplace Transform where the roots
of the denominator are the complex conjugate pair of poles that correspond to a
mode of vibration. The H(s) is known as the compliance transfer function in the
laplace domain and in the frequency domain it is referred to as the frequency
response function of the system.
TTI-
S X(S) ^
H(s) = =
(Eq. 2. 16)
F(s) ms'+cs + k
where s12 =-a+icod (Eq. 2.17) and cod (Eq. 2.18)
The poles can be written in other terms relating them back to the loss factor as
done in the following manner [ 16, 17, 19].
-c ( c V k
=(-Clyl\-C2)cO (Eq.2.19)
2m 2m m
where
0)=J
(Eq. 2.20) & a=
=C,COn (Eq. 2.21)
2m
and the loss factor 17 = (Eq. 2.22)
The decay constant is found through use of the impulse response function.
The impulse response function h(t) is nothing more than the system response in the
time domain. The impulse response function h(t) is a real valued function and is
presented usually on a linear amplitude scale. The Hilbert transform xA is used to
h(t) the time the impulse

compute the imaginary part of real valued signal of
response. The Hilbert transform is defined as the convolution integral and it is

written in the convolute form of the impulse response signal [12,13].
-CO .
h(t) =
Ar{/?(/)} =
\h(t) dx (Eq. 2.23)
21
The analytical time signal of the impulse response function h(t) can be expressed
as
V
h(t) =
h(t)+ih(t) (Eq. 2. 24)
The magnitude of the impulse response function is considered as being the

envelope of h(t) and is computed as
l(t) yjh2(t)+ih2(t) (Eq. 2.25)
The purpose of the magnitude of the impulse response function presented in this
format is that it offers two advantages. The first is the removal of the decaying
oscillations, allowing for a smooth envelope decay as in the top trace ofseen
figure 2.11 on the next page. It also shows, in the bottom trace, the decaying
oscillation from the real part of the impulse response which is also the same if the
signal was viewed under Time Ch. B. Secondly, the magnitude of the analytical
time signal can give a more dynamic range of the impulse response signal by using
a log amplitude scale versus a linear scale. The log amplitude scale permits the
envelope of the signal to become a straight line. The straight line is used to
measure the time constant x (sec) corresponding to the amplitude decay rate of -
A1
8.7db. The time constant is given by the factor or in decibels -201og(e)

= -
8.7db. The time constant is measured from the

selecting a point on the analyzer by
slope and cursoring down the slope a -8.7 db on the y axis. The -8.7db change on
y axis corresponds directly to the time constant found on the x axis. Once I know
the time constant, the loss factor can be computed for a resonance frequency by
[12]:
i
(Eq. 2.26)
fn
The analyzer basically calculates the impulse response function as the
inverse Fourier transform of the weighting function that is applied to the frequency
response function times the 1 -sided spectra of the Enhanced Time Function
d(n) and b (n) extended with zero padding. The Enhanced Time Function of a(n)
is composed of the averaged Time records of each sample across a series of
instantaneous Time Records and it represents the input data from channel A. The
Enhanced Time Function b(n) is the same as a(n) but represents the channel B.
data [20].
22
ft;
J"
.So
ft;
3. MODAL TESTING
3.1 Modal Testing
Modal testing is the application

of experimental methods to determine a
system's dynamic characteristics. Modal
testing can be broken down into three
basic requirements. The first component demands a theoretical background of
vibration in the construction of the mathematical model. The theoretical
basis
starts with an of the The
understanding spatial model. spatial model refers to the
mass, stiffness, and damping properties. Next is the decision for what type of
modal model to represent the structure be
being examined. It can either considered
a single degree of freedom model (SDOF) or a multi degree of freedom model
(MDOF). The modal model represents the occurring natural frequencies with its
corresponding mode shapes and mode dampings. The modeling usually follows a
lumped-parameter model and model theory. The second requirement demands
accurate measurements of the vibration. The main measurement techniques must
be decided and develooed in order to make the measurements. This consists of
mainly three things which are the way of supporting and exciting the structure,
what is to be measured, therefore the force input and output response, and the
appropriate signal processing to be used. The last of the big three is the summary
analysis of the measured data. The summary analysis is referred to as the
experimental modal analysis. It is simply estimating the modal properties from the
structureusing the measured data via computers or analyzers with Fast Fourier
Transforms (FFT). For instance, this involve using software
can packages for
parameter estimation by curvefitting to illustrate the animation of mode shapes
[1,16,18].
Despite the basic guideline given above for doing a complete modal test, it
is most important to remember that no single test or analysis method can be said to
be the best for all cases. It is also important that a clear objective is set before any
tests are conducted. This will ensure an optimum use of all methods or techniques
to be used. To develop a clear objective when
performing a modal test, ask
yourself these two easy questions: what do you expect to achieve or determine
from the modal test? and secondly, what form must the results of the test be in for
maximum use [18]?
The setup for conducting a modal test is called a dual channel
configuration
FFT measurement. One channel is selected for the input and the other for the
output the analyzer takes the data from both of these channels and
whereby
processes the information using the built-in functions on the analyzer. Figure 3.1
on the next page shows the basic diagram of analyzer for Dual Channel Spectrum
Averaging. The input channel is represented by the measured input force where it
is created at a point on the structure, and the output channel measures the response
24
from the excitation at some other point of interest on the structure. The most
difficult part of modal is
testing ensuring that the data collected is accurate. Many
AwlO-
JS,
L3
3f
-,.-_,
^ ConWi
N*eorng . . AjwiyMa Poal ProcvaMng
Figure 3. 1 Bruel and Kjaer 's Flow Chart of Dual Channel Spectrum Averaging
problems can occur from a modal test such as noise from the source of the input
and output devices used, limitations of the Fast Fourier Transform such as leakage,
aliasing, and the "picket-fence", non-linearities of the structure, bad cables, ground
loops, ambient noise and possibly any delays in the analysis not compensated for.
The noise problems arising from the analyzer is due to the way the analyzer
approximates the Fourier Transform [8,16]. The Fourier Transform has the form of
F(f)=]f(t)e-'2*dt (Eq.3.1)
and the analyzer is only capable of calculating it by the truncated Fourier series
At^f{nAt)e"2
i2n{mAf)
F{mAf) = (Eq. 3.2)
where N is the number of samples taken in the time window and
m, n = 0,1,2. ..,N-\ are integers. This is considered to be the discrete Fourier

Transform where its supplied information, such as its amplitudes and phases is
known at evenly spaced intervals in the frequency domain. The Fast Fourier
Transform is simply a rapid method of computing the discrete Fourier Transform.
Therefore the discrete Fourier transform can exist only for periodic functions and
the observed time history is inherently assumed to repeat itself at some period T
for the sampled time intervals of At [12,16].
The limitations arising from the discrete Fourier Transform can be

such errors as aliasing, leakage, and the picket fence
minimized by understanding
so that adequate data can be obtained. The aliasing is
error an error in the digital
where two frequencies can not be distinguished, and this is caused by the
sampling
25
sampling at less than twice the maximum frequency in the signal. This is
eliminated by a lowpass filtering of the signal before sampling. The picket fence
effect is a result of the discrete sampling of the spectrum in the frequency domain,
and these possible errors will depend on the of adjacent filter
overlapping
characteristics. Leakage,
sometimes known as the "time window effect", can give
rise to information in the original signal producing a resolution bias error.
spurious
Leakage may arise in the frequency domain due to the time limitation of the signal
before the FFT calculation is performed. Only a finite time record length or period
of the original signal can be analyzed at a time, and the origianl signal being
analyzed is the original time history multiplied by a time window. These time
windows are called a
weighting function. The time weighting functions in effect
are conditioning the original time signal. This is achieved by the spectrum A(f) of
the original time signal a(t) is convoluted with the Fourier Transform w{f) of the
weighting function w(t) [8,12,16]. This can be shown as
A(f)= ]a{t)e-j2^dt andW(f) = Jw(/)A^c// (Eq. 3.3)
and after convolution:
A{f)*W{f)=
\a{u) W(f-ii)dn= ja{/)-w{t)ej2^dt (Eq. 3 4)
CO \f:
A visible illustration of this is given in Figure 3.2.
" "
Leakage
ww Affl
t\
Mi UU"
Figure 3.2 Leakage in the Analysis
The figure above shows how a weighting function could influence the estimate
from the original spectrum of time history. The power in one frequency region
leaks into the adjacent frequency regions causing the peak amplitude to drop and
the amplitudes in the valleys to rise. The amount of leakage will depend on the
type of weighting function used and therefore the appropriate window should be
investigated for the application being attempted. For example, if a rectangular
window is applied to the original time signal and if this time signal is longer than
26
the length T, an unknown
discontinuity will arise at the loop junction. Therefore a
better choice of a time window should be applied to minimize its effect. The
solution for this example would be to choose a Hanning Window. The Hanning
Window is
matching the record length of the analyzer and
one period of a cosine
lifted so it starts and stops at zero. The sidelobes of a Hanning window will fall off
much faster than the rectangular window and improving the overall characteristics
for this case. Also the Hanning window is normally used as a weighting function
when
analyzing continuous random or sinusoidal signals [12,15].
A way to indicate such errors in the measurement is through the Coherence
Function. The coherence function will measure on a scale from 0 to 1 the degree
of linear
relationship between the two signals a(t) and b(t) at any given frequency.
Anything less than one indicates a quality problem in the measurement. The
coherence function was applied for the impact and random excitation data and is
shown in Appendix B and C. The coherence function was used every time after the
frequency response estimate was made. It is up to the user to figure out why at
times the values are less than one in the coherence function, and many times it is
more than one source creating the problem. This is where I found much time was
spent studying and understanding the effects of different windows and creating
mini experiments to see the effect ithad on the calculation of the loss factor. Also
experienced people in experimental vibrations are very helpful in providing
assistance for this stage of the analysis [12,16,19,20].
3.2 Fundamental Functions of the Bruel & Kjaer 2032 Digital

Signal Analyzer
The B&K 2032 Digital Signal Analyzer was the center of operation for the
measurement and determination of the loss factor. This is a two channel FFT
analyzer which can measure and display 34 different time domains, frequency
domain, and statistical functions. These 34 functions are measured in six different
modes of operation: Single and Dual Channel Spectrum Averaging, Single and
Dual Channel Spectrum Averaging Zero Pad (used to avoid the effects of circular
folding when measuring correlation functions or impulse response functions), Dual

Channel Signal Enhancement and Dual Channel Signal Amplitude Probability
mode. The analyzer has a resolution in the frequency domain of 801 lines with a
frequency span of 1.56 Hz to 25.6 kHz. The frequency span can be selected
anywhere in this frequency range. The analyzer has an array of options to display
the measured data such as real, imaginary, magnitude (Hilbert Transform), phase,
Nyquist plot (imaginary vs. real), and Nichols plot (log magnitude vs. phase). The
analyzer has many weighting functions readily available that can be selected for
each channel and these are Hanning, transient, rectangular, exponential and
Kaiser-Bessel as well as user-definable weighting function. The Hanning window
is commonly used for continuous signals while the rectangular, transient and
27
exponential functions are used for transient signals. The transient window and
exponential window are user defined for the start of a window with respect to the
beginning of the time record and the window width [20,38].
The functions on the analyzer that in loss factor
were used determining the
other than the Coherence, Frequency Response Function HI and the Impulse
Response Function were the functions, Zoom-FFT, transient,
special parameters
Hanning and exponential

weighting functions, and generators functions for
producing excitation signals from the analyzer. The Frequency Response
Functions (output divided by the input) will give the optimum calculation of the
response H(f) in the least squares
sense, and there are two types, HI and H2. The
HI is the averaged Autospectrum Gaa divided
by the averaged Cross Spectrum
Gab where a and b denote the channel input and channel output respectively. H2 is
defined as averaged Autospectrum Gbb divided by the averaged Cross Spectrum
Gba. Each frequency response type offers particular advantages depending on the
analysis. H 1 is best suppressing high noise at the output and will give the most
at
correct results for low levels in the response when the output signal is low. H2 is
best at suppressing high noise at the input and gives the most correct results for
high levels of the response when the input signal is low. For this investigation,
Frequency Response HI was used in estimating the response of the aluminum
beam and in the expectation of noise from the sensors. The Coherence Function by
the way can be calculated by dividing the H2 into HI. Zoom-FFT provides a
higher resolution to examine a specific frequency range more closely but at a cost
to a longer length of time to perform the analysis. This longer length of time is
given by the record length T. The other advantage to zoom analysis is that it
improves the resolution bias error and the coherence in the analysis. The resolution
bias error means the resonant peak is narrower than the resolution of the
measurement which can effect the impulse response. This is why the half power
points can not be calculated directly from the analyzer without it being sent to the
some curve fitting software [7,12,19,20].
is defined is
e~t/T
The exponential time window as w(t)

=
and applied when
the oscillating decay of the signal is longer than the analyzer's record length. The
exponential window as just mentioned before adds artificial damping to reduce the
amplitude near the end of the decay and this type of window introduces leakage.
The signal is therefore fully captured in the time record and minimizes some of the
leakage. accounting for the extra damping in the calculation of the loss factor,
By
this subsequently allows for the correction of the leakage. The correction is made
by subtracting the decay constant aw of the exponential window from the decay
constant a of the impulse response. This produces the true decay constant crtnie that
1 1
0=Om-aw=
(Eq.3.5)
-
28
=
H -r (Eq.3.6)
will be used in the real calculation of the damping. The rectangular window is
simply defined as w(t) = 1 and is not really considered weighting the function at
all. The rectangular window is used to eliminate leakage effects when using
pseudo-random excitation. The psuedo-random excitation is synchronized in
repeating a period of random signals so that the data sampling fits the record
length, therefore achieving good results. The Transient window is a rectangular
window but the user can define the beginning and the end of the window length in
the record length.
Anything outside the window is zero, and this window has the
effect of improving the signal to noise ratio in the measurement. This window was
used in the impact excitation of the force response as well as for the response time
signal. In consideration for long time records, using an exponential window has
the effect of adding extra damping which reduces the amplitude of the resonance
peak and coupling closely spaced modes in the frequency response. This makes
modal analysis very difficult, and it is recommended to use a longer time record to
capture the full response or use zoom analysis. When using zoom analysis, the
response signal may decay well before the time record has finished and therefore
an exponential window is not needed. The transient window length can be set for
longer time records to capture the full response while eliminating any noise
occurring outside the window length. Two examples between an exponential and
transient window in calculating damping can be found in Appendix B for the laser
and the accelerometer at the second resonant. The results and discussion will
explain in more detail at why these choices were made [8, 12, 15].
The special parameters section pertains to the usage in the impact excitation
section. At the beginning of the plots for each resonant, as seen for the data
generated in Appendix B, there is a table created for the special parameter function
numbers. Special parameter number 0 is called the key and allows access to
special parameters by typing in 2032. Number 40 is the frequency weighting

function used when calculating the impulse response function from the frequency
response function. The number placed here represents the type of weighting
function used in the analysis; 2 for the transient window in the impact excitation
and 0 or 1 for the rectangular and Hanning respectively in the continous excitation
via the shakertable. When the transient window is selected special parameters 41
to 44 are used. Number 41 is the left element number when filtering the resonant
and is located under the reference window on the screen. It is determined by

down on each side of the resonance on the
cursoring down a minimum of -lOdb
screen. Number 42 is the difference between the left and right element values. The
right element on the right side of the resonant cursoring down the
is determined by
same amount as the left element was. Number 43 defines the length of the leading
half-cosine taper to smooth the filtering of the resonance, and number 44 similarly
29
defines the trailing half-cosine taper length. Number 76 and 77 allows the user to
view the time function in the upper or the lower trace on the screen respectively.
The number 2 placed in either 76 or 77 means the time weighted function is

multiplied by the normal time function. Numbers 47 and 48 allows the user to
view the filtered frequency response in the upper or the lower trace on the screen
respectively. The number 2 placed in either 47 or 48 means the frequency response
function is multiplied by the frequency weighting function [19,20].
3.3 ASTM Standard Method for Measuring Damping Properties

of Materials
The ASTM Designation: E 756-83 Standard Method for Measuring

Vibration-Damping Properties of Materials addresses the problem of measuring
vibration-damping properties that vary greatly with changes in temperature and
frequency. The method is comprised basically of 6 parts in which to conduct the
test requirements. These six parts consist of the assumptions and precautions that
areinvolved in the test, a description of the test beams used in evaluating the
damping material, the instrumentation and testing setup, analysis of the data
reduced and what the report of the analysis should include [9].
One of the assumptions in this method requires that the damping material
being tested follow the linear viscoelastic theory. This method suggests the use of
an environmental chamber when performing a modal test on these damping

materials in order to study the factors such as temperature and vacuum pressure.
The method gives a guideline for selecting the dimensions of the damping material
as well as for the type of beams being used in order to avoid errors in determining
the true damping of the material. It is suggested to use steel or aluminum in the
composite beam, and the suggested dimensions for these bars are a width of 10
mm, free length of 1 80 to 250 mm, root length of 25mm, and a thickness of 1 to 3
mm. There are essentially three types of composite bars and these are a damped
side beam where one side damping material attached, the second type is
has the
both sides of the beam having the damping material and lastly, the sandwiched
beam with the damping material between. The metal bars alone are often referred
to as Oberst Bars. They have noted that materials such as steel and aluminum have
a well known loss factor of 0.001 or less, and it is sometimes considered to assume
the loss factor of the metal beam to be zero in the calculation of the damping
material [3,9,16].
It is important to select a meaningful force level for the analysis, and the
amplitude of the force signal must be kept constant with frequency, otherwise the
data will not be applicable. This is particularly important because the damping
could be too high when determining the loss factor from the measurement. If the
force level can not be kept constant, then the response of the beam must be divided
the force input in to the normalized transfer function. The current
by order obtain
30
signal applied to the force transducer can be an alternative way to measure the
force since the two are proportional. The first mode of vibration of the cantilever
beam should not be used to calculate the properties of the material being
investigated. This is necessary because nonlinear effects may be
precaution
produced from the high amplitudes given

by the first mode [12,16,24,25].
It is very important to pay close attention at all times when making any
damping measurements. It is usually a very complex and involved process because

it involves the material specimen, the type of excitation, the sensing of the
response, instrumentation and so on. It is necessary to pay attention to all the

details of the experiment from fabrication of the material to the analysis of the data
being measured. The procedure for setting up themount the bar in
modal test is to
a heavy rigid fixture and clamp the root thickness of the beam to simulate a fixed
boundary condition. It is very important in using a non-contacting instrumentation

and to avoid anything else that could influence the damping characteristics of the
test specimen. The method suggests using either a random or sinusoidal signal
which can be applied to a electromagnetic transducer type for creating the

excitation via power amplifier. The sketch below describes a typical laboratory test
setup for the vibrating beam test [3,9, 16, 18].
Figure 3.3 Basic Diagram of Experimental Setup for Testing Vibrating Beams
Once the data has been measured and stored, the suggested method for
dynamic response of each mode is determined
calculating the loss factor from the
Bandwidth (-3db down) [9].
in terms of the using the Half-Power
frequency
Therefore, the loss factor is calculated as:
n =
Ml (Eq. 3. 7)
fn
31
The estimate of precision for individual temperature-frequency evaluations on the
extensional
damping materials should not exceed 25% of the coefficient of
variation for the loss factor [9].

The report basically should
specify all the dimensions concerning all the
materialsinvolved in the test and any special chemical treatments made to the
beam's surface for preparing adhesion of the
damping material. The report shall
include all data such as the damped and undamped frequencies, temperatures, and
pressure. These tests should include a statement that it was done by an accredited
laboratory and the tests were followed in accordance with this procedure. If it is
not true in every respect, list all the exceptions made [9,16].
32
4. SENSORS
4.1 Laser Vibrometer
4.1.1 Description
The Dantec 55X-Laser Vibrometer is an optical system for measuring the

instantaneous velocity of vibrations of solid or liquid surfaces. This sensor permits
non-contact, point measurement of the velocity component in the direction of the

laser beam within a wide range of velocities. The laser system provides wide
bandwidth, real-time digital outputs proportional to the instantaneous

analog and
velocity of the surface includes

which the indication of sign for forward and
reverse direction. These signals can be further analyzed to provide the
displacement or acceleration of the motion. The laser vibrometer is primarily
designed for non-contact optical measurements of vibrations of rough, diffusely
reflecting surfaces. This laser is also capable measuring the velocity
of of a
continuously moving surfaces or single events [32,33,34].

This type of laser vibration thus distinguishes itself from other optical
vibration meters only because of its non-contacting ability to measure

not
specularly reflecting surfaces but it also possesses the capability of measuring

non-harmonic and random motions as well. With such potential, it has found a
wide variety of applications in the measurements of flow-induced vibrations,

vibrations in buildings, biological and audiological research and the list goes on
[32,33,34].
4.1.2 Technical Background
The laser vibrometer measures velocity by means of the detection of the

Doppler shift of light scattered from moving surfaces. In the
or reflected
accompanying figure found in the Dantec Laser Vibrometer Instruction Manual

[33], the Doppler Shift is shown in order to understand the relationship of the
vibrating surface and the laser beam components with respect
VIBRATING SURFACE
Jj j
Figure 4. 1 Doppler Shift of a Vibrating Surface by Dantec Electronics
33
to the optical axis. Figure 4.1 identifies the laser components kj_ and k2.The
kl is the k2 is the
component wave vector
coming from the laser and wave vector
of laser beam. 0 is the angle between the incoming and scattered

the reflected
beams. Vz is the velocity component in the direction of the optical axis of the laser
vibrometer Figure 4. 1 illustrates the Doppler formula which can be expressed by
.
coD
=
(k2-ki)V (Eq.4.i) and fD
=
(2/A)sin(0/2)Vz = 2VZ/A iEq.4.2,
equations 4. 1 and 4.2. The Doppler shift frequency is coq

=
litf^ The laser beam

wave length is represented by A, where A, 27t/|k|
=
. The magnitude of the wave
180
vector |k| is =
|kl| |k2|. In using the
= standard objective, 0 is approximately
which allows sin(0/2) 1 [33].
=
The optical system in the laser vibrometer detects the Doppler shift of the
reflected laser beam by an optical heterodyne process. Heterodyning is the optical
mixing of two different frequencies where in this case the received laser light
beam is mixed with the laser reference beam yielding a signal with a frequency
much smaller than the absolute frequency of the original signals as shown in
Figure 4.2. The new frequency is equal to the difference of the Doppler shifted
(Dl <d2
(YT^
vu<Aj \}y vhH iy~irr
\j~Js *
r
aA-aO.
1 1 ^ ^\Y
/\V1>
U VJ ^
Vy
Figure 4. 2 Doppler Shifted and Non-Shifted Beam by Dantec Electronics
laser beam col and the non-doppler shifted laser beam co2 which has a fixed
40Mhz. optical frequency shift. cc>2 is considered to be the reference laser beam
and the new frequency exits the photo detector as
col-
co2 [34]. By illustrating
Figure 4.3, it attempts to show the optical path through the complete laser
vibrometer.
X19 X25 X29 X23 X66 166
Figure 4.3 Side View of the Laser Vibrometer Component Identification by Dantec Electronics.
34
The 55X19 adapter supports the 5 mWatt coaxial He-Ne laser. The laser beam is
then split by the 55X25 Beam Splitter. In the 55X29 Bragg Cell section, co 1 is
frequency shifted and focused on the object by the fronts lens at L\, L2 and Rj A
better representation of L\, L2 and R\ is found in Figure 4.4. L2 is a "standard TV
lens and can be manually adjusted. The upper beam labeled A is focused on the
object being measured by the transmitter/receiver optics and the part of the
y x
Figure 4. 4 Optics Path of the 55X Laser Doppler I 'ibrometer by Dantec Electronics.
reflected or scattered light which reenters the receiver aperture is collimated to a
beam parallel to the outgoing beam. The beam of reflected light reaching the
detector be orthogonally polarized to the outgoing beam and the reference
must
beam. Notice the dotted lines surrounding the return laser beam is the only
effective amount of light seen in the optical heterodyne detection. In theory, the
optical unit utilizesonly one ray or one speckle of laser light. Since this light is
scattered, it makes a lock loop very difficult at times to maintain frequency feed
back of the Doppler frequency. The lock loop will be discussed with more detail
in the Experimental Setup and Operation section since it plays a major role with
the initial setup. The Frequency Tracker performs the phase-lock loop feature and
the Tracker output is an analog voltage directly proportional to the input
frequency. The solid state photo detector section is located by X and Y in Figure
4.4 which is where the new frequency (co 1- co2) exits [32,33].
On the next page, Figure 4.5 will attempt to give a brief overview of the
Frequency Shifter and Frequency Tracker in operation with the laser vibrometer. It
is seen here that the signals exit the front bnc connector of the laser and enters the
Shifter by the mixer to the preamplifier of the
Frequency which communicates
Tracker. The preamplifier signals in the Tracker unit enters the Frequency Tracker
module where the analog output signal is used by the B&K 2032 Signal Analyzer.
The Frequency Shifter delivers the 40Mhz. signal to the Bragg Cell unit (X29) of
35
the laser vibrometer. Depending on the setting of the Frequency Shifter, the
resulting output from the signal processor is a signal linearly related to the
velocity Vz and fixed velocity offset V0. In

with a combination with the
Frequency Shifter, Frequency Tracker and the Laser Vibrometer, a variable
frequency shift is achieved allowing us to measure vibration [33].
X25 X29 *66 L66
1 '
jr
G LASER
55X
Laser Doppler Vibrometer
A %
i r
1 1
1
Mixer
1 j Preamplifier
Frequency
Tracker
1
i
i
55N20
' Doppler
Loral 1
40 MHz
Oscillator 1 Frequency
| Tracker
I
' 55N10 Frequency Shifter .
I I l_.
Analog Out '
Digital Out
Digital in
Figure 4. 5 55X Laser Doppler Vibrometer with 55NI 0 Frequency Shifter and 5SN20 Doppler Frequency
Tracker bv Dantec Electronics.
4.1.3 Experimental Setup and Operation
4. 1 .3. 1 Technical Preparation
using the laser vibrometer, it is

advised to
Before starting an application
laser itself first. The laser vibrometer is a beautiful tool

learn how to use the by
you have studied your environment in relation to the laser's operating
providing
36
window. Even with the help of the instruction manual, it was difficult to interpret
the alignment signals from the oscilloscope to what was conveyed in manual.
Through trial in adjusting the alignment of the laser and with
and error some
patience, an
understanding of its sensitivity will be helpful in the setup for a real
application.
First, it suggested to start with a basic model to experience the laser. The
basic setup will consist of
using a purely reflective surface such as a mirror. The
only other piece of equipment that would be needed is an oscilloscope. The
oscilloscope should be considered a standard tool whenever using the laser
vibrometer. The purpose of the oscilloscope is to monitor the output signal from
the Doppler Frequency Tracker. As an overview of the basic setup, Figure 4.6 was
constructed as a visual aid. There are many properties that effect the laser in which
BASIC LASER SETUP
EXCITER POWER BOX-
Min Distance
UZ3
OSSCILL OJTCOPE
1 2m
-\
?
MIRROR
V- FREQUENCY
SHIFTER
"?=[ LASER
FREQUENCY
A^ WEAR LASER SAFETY

SA GOGGLES
TRACKER
Figure 4. 6
one must be familiar with when using the laser. These are depth of field (cavity
length of the laser beam), lens speckle effects, electronics, focusing, slew rate,
both the laser
objects'
of the objects surface, alignment of as well as the

reflectivity
surface, wind and temperature [32,33].
The depth of cavity length is determined by the type of laser and the
field or
given optics built into the device. The measurement range is 1.2 to 20 meters but
Another words, measuring distances giving
cavity lengths of n/2 must be avoided.
a path length difference of one half of an integer number of cavity length should
be disregarded. At smaller or greater distances, a special optics adapter is available

for focusing [33].
The speckle effect is the result of the random composition of the phases of
light scattered upon the

objects'
surface and reflecting these rays back towards the

laser. By taking a sheet of paper and placing it vertically behind the TV lens of
the laser, you should notice that part of the paper is covered with a speckle
pattern from the laser light. When using a mirror you may not see much of this
37
because it is almost
perfectly reflected. In an application where the surface is not
reflectiveenough, there are many things you can do to enhance it. Some of these
suggestions to help make the surface more reflective are off
sanding any rust,
or
painting electroplating a reflective coating,
removing any dirt or oil from its
surface, or
applying reflective tape. Some of these reflective tapes are
retroreflecting which means the light hitting the tape will disobey Snell's Law. I
found that the packaging material from Peanut Brittle works just as good or even
better. Other candy wrappers such as York Peppermint Patties, work, but I
recommend not using them. In
applying the peanut brittle packaging, just cut into
circles or squares and stick it on the surface
by using double adhesive tape. The
best tape of this kind is Ludlow #8602. Once you have obtained a good signal
from using any of these reflective tapes, it is now possible to cut away with great
care any unused portion of reflective tape not seen by the laser. This enables the
reduction of any localized mass which may be sensitive enough to change the
natural frequency of the object.
The purpose for the speckle pattern is to gain feedback to the PIN-diode
optical heterodyne detector for signal processing. This detector has a high quantum
efficiency which allows for the suppression of unwanted beat frequencies induced
by the laser or even from the Bragg cell. In combination of the electronics and the
Si PIN diodes, as shown in figure four at
P\ and P2, are combined to act as the
light detector in which the circuit is tuned to the center of the shifted Doppler
frequency. The resulting frequency shift retains all properties of variable optical
frequency shiftalong the range and bandwidth selection made at the

with
Frequency Tracker. With these properties kept in mind, it can now be seen more
clearly how important they become through the relationship in the power
signal-
to-noise ratio at the photodetectors. In general, the equation is as follows:
s/ ^
/^
=
2A/(P1+P2)+hk^
1 2
e2rp.L
where k is Boltzmann's constant, T the temperature of the detector, Rj^ the load
resistor, r\ quantum efficiency of the detector, e the electron charge, hv energy of a
quantum of light of frequency of v and A/ the detector bandwidth. This equation
assumes no back ground light and dark current. The previous expression for S/N
changes if we assume a strong reference beam where Pl is much greater than P2:
r}i?2
S/m =
hl vt
C+4-4>
J
2P,e277RL
38
The above expression can be further if enough beam
simplified reference power is
used such that the denominator can be neglected. The important thing to note
s/ hv ... . .
(hq- 4"J
N-
"
rh livkT
Af(\ + )
2P,e2nR.
from these equations is that temperature could a role in obtaining good S/N
play
depending how strong the reference beam and the adjustment of the optics. I
would recommend paying attention to any periodically strong drafts in the
laboratory. The signal-to-noise ratio becomes obviously sensitive to the chosen
settings at the Tracker. Another
Frequency symptom of poor signal-to-noise ratio
can be found at the coincident cavity length of the laser beam [32].
The focusing of the laser on the object determines the quality of the S/N
which can be observed by using an oscilloscope. The laser beam has been
observed following Guassian characteristics with a
1/e2
intensity diameter of d\.
-
The small spots from the speckle pattern, as seen by holding up a sheet of paper
back behind the TV lens, can tell you a lot about the surface you have targeted.
The spots on the surface of the object are made up spots, these spots are
of more
inversely proportional to the focal spot size. This implies that the spot diameter,
d0, is inversely proportional to the expansion of the beam (/V/i) whereas the
depth of focus, l0, is inversely proportional to the square of the expansion. The
focal length of the optic's objective determines the d0 and by the lens system f\
and f2 respectively. The importance of distance from the object, /, is evident by the
following equations [32]:

j
W'l c ,
32A/,2/
& lo=n-
do=-7T7T (Ri-4-6> (Eci-4J)
nd, f ndj2
This can be viewed when turning the TV lens and watching the expansion grow on
the test object. The selection of distance from the object to the laser is dependent
on the expansion as well as the type of spots reflected. When the speckle pattern is
very coarsed-grained, this can indicate that the focal spot is at a minimal. The
function of the laser's lens system is to control all these properties of the laser
beam such as focal spots, but confined to the limits of the collection aperture. The
optimum size of the collection aperture is found to occur in an optical antenna
condition. This is by AQ 'Kr where A is the area of detector aperture

represented =
and Q is the solid angle subtended by the illuminated spot as seen by the lens.
This collection aperture is satisfied by the antenna condition when the width of the
illuminating beam becomes equal. The downfall to this means that a larger
aperture would not improve the signal-to-noise ratio of the detector current when
sensitivity becomes an issue. Alignment even becomes very critical at the optical
39
axis of the laser because it becomes sensitive to characteristics of the surface and
the amplitude of vibration. It
simply becomes a trade off of properties depending
on the
measuring conditions. Another interesting situation that could occur under
unfortunate situations is that the aperture could be trapped between two bright
speckles giving rise to a poor S/N. The optics could be adjusted

slightly to increase
the power of the beam or make the sacrifice in realigning the laser [32].
Of the two pieces of equipment, the Frequency Tracker Frequency
and
Shifter, any adjustments made to the signal processor on the Frequency Tracker
must be the Frequency Shifter as well. Figure 4.7 is a sketch of both
also made on
The top box is the Tracker with selected ranges which can be chosen by the
units.
RANGE knob and the FILTER knob controls the analog output with a cutoff
frequency as shown in Table 27 of Appendix J. Frequencies of vibration above
the cutoff
frequency under any of the selected BWs or OUT column will be
filtered out [33].
1-lOMHz ft
OUT
NOT 0.3-3.3MHz ft BW3 LOCK NOT
USED 0.11MHz * BW2 O
USED
33-333KHZ ft DETECTOR
BW1 ft
10-lOOKHz ft
lXYEL
3-33KHz ft SCAN REMOTE OUT
55L97
55N24
HV 1-lOkHz ft ? ? ? DISPLAY
SUPPLY
RANGE GAIN FILTER MODULE
TYPE 55N21 TRACKER MAIN UNIT
FREQUENCY SHIFT IN KHz POWER

4*6 >
2 9 g
1 9
MULTIPLIER
(BLANK)
X1 X1
F,lo < 40MHz _ F,i > 40MH*
xioo am X100 55NH
X1000 X1000 FREQUENCY
SHIFTER
MAIN UNIT
FREQUENCY SHIFT CHANNEL
Figure 4. 7 Front Display of the Frequency Tracker and Frequency Shifter

by Dantec Electronics
The green light on the Tracker labeled LOCK DETECTOR will flicker or
may not
flicker step you are at in setup. At this
even at all depending on the what your
point, an oscilloscope is required for tuning in this signal. The signal that we want
to see is a sinusoidal clipped wave which basically appears as a square wave. In

figure 4.8, I have sketched a picture to illustrate this wave form. The square wave
as illustrated means success. By studying the waveform seen on the oscilloscope
while tuning the laser vibrometer, you can begin to recognize features of the signal
that will indicate the closeness of a square wave.
40
OSCILLOSCOPE
o
VOLTS/DIV
O
SEC/DIV
O
CH. 1 CBNC)
Figure 4. 8 Laser Alignment by Identification of a Square Have Signal
For instance increasing or

decreasing the GAIN on the Tracker can help clean up
the signal even more if it looks terrible. I prefer to keep the GAIN somewhere
close to midrange on the Tracker usually at the 12:00 a clock position if possible.
If you find yourself with the gain at the far extremes, you may wish to tweak
some of the other parameters to bring it back midway. This will give you greater
flexibility to probe a distorted signal using the gain knob. The most common cause
in which one may have trouble obtaining a good signal is the alignment of the laser
in respect to object, the distance to the object, the speckle pattern created at the
surface of the object and focusing the beam expansion on the object. I have found
nine times out of ten that the one speckle needed at the photodetectors is the
problem. If this occurs, the only suggestion I can offer is to try and optimize these
four parameters while viewing the oscilloscope untill the ideal wave form appears.
Usually I would start by adjusting the lead screw (one or two turns) which is the
back support leg of the laser. Sometimes creating this angle will match the
curvature or wrinkle in the reflective tape. Another hint that the wave form is
there but just needs to be focused is seen in the following sketch.
Figure 4.9 Pre-alignment Wave Form Signal
Other bits of details can be observed by the square wave form and that is the
random fluctuations of the period. If the period of the square wave appears erratic,
this indicates that there is still noise in the system somewhere. I have found that
this type of noise can be attributed to the ambient vibration from the room. Just by
41
walking hard you can observe the entire square wave oscillate like a sine wave but
exponentially decaying. This may require a vibration isolation table depending on
how sensitive the application. The other possibility could be the object vibrating
much greater than what the laser is set to receive. This can occur if the cantilever
beam is tapped too hard. I found that this laser vibrometer is very sensitive to
vibration, especially amplitude. Though you think the beam has finished vibrating,
in actuality, it has not. The time domain of this could very well be up to 30
all
seconds. To prove this take your fingers and dampen the beam even though the
cantilever beam appears to have stopped vibrating. You should

instantly see the
square wave cleanup with a crisp consistent period. This is because we are dealing
with lightly damped structures that are not heavily damped. This finding was
encountered during the excitation of the beam; one tends to become impatient in
taking the required number of averages to gain a decent frequency response in the
minimum amount of time. Since the Doppler frequency is proportional to
frequency of vibration and amplitude, it can be visualized as a hyperbolic shown
where K is some factor
q
u M f =
. some
constant
'W
Amplitude (A]
Figure 4.10 Doppler Frequency vs. Amplitude by Dantec Electronics.
in the sketch above. When the rate of change of the Doppler frequency exceeds the
Tracker slew rate, the processor is no longer able to keep up with the changing
frequency and the green light goes out. The Tracker only measure up to 10
can
MHz. Figure J. 1 in J and in Table 26. of Appendix J displays

as seen by appendix
the maximum slew rate for the range selected on the Tracker [32,33].
None the less, the other settings such as the SCAN, REMOTE and OUT
buttons should be left in their out most position. These buttons are not used in the
application of the laser vibrometer The LEVEL light on the Tracker will come on
the Frequency Shifter is on and connected properly. The

automatically as long as
proper configuration for these connections among the equipment is seen in figure
4.11. the diagram is self evident but remember to ensure in the back of
Basically
the Tracker that the Diode Detection switch is up.
42
55X Laser Doppler
Vibrometer
to Exciter
40MHz Output
55N10
Frequency Shifter
Mixer Out
CH. B of the
B&K 2032 Analyzer
Up
Diode Detec )
(Photodetec in)
55N20
Doppler Frequency
Tracker
to Oscilloscope
Max 400 mV
Scope 0 1 V/cm
Figure 4.11 Block Diagram of the Laser Vibrometer Equipment by Dantec Electronics.
The settings on Frequency Shifter are quite simple once you have
the
decided the vibration frequency range and Tracker range. The selection for dialing
in the frequency shift value and multiplier is found in the Appendix J of Table 25.
At this point, things tend to simplify because the B&K Analyzer is only capable of
capturing a maximum frequency of 25.6 kHz. This is known by looking at the
screen on the Analyzer for ch. A and ch. B. where the filter should read 25.6 kHz.
You can now ignore the upper half of this table. When making the selection for the
Multiplier, as understood by Dantec, the f\0 < 40 MHz. button is selected.
/j0 is
43
the subsequent electronic frequency shift in the mixer as seen back in figure 4.5
[33].
4.1.3.2 Basic Operation using a Mirror for Signal Conditioning

At this point, all the equipment should be connected and set up in a similar
fashion as specified in figure 4.11. There should be wooded blocks already cut if
needed to help level the laser. Also there is a digital level that can be used to level
the laser and it is small enough to fit on the base of the laser. I had always turned
the equipment on this in particular order (Laser Goggles On).
1 . Power Exciter On (power light goes on)
2. Frequency Shifter (On switch located in back and power light goes on)
Adjust the range knob to . 1-1 MHz. and the filter set to OUT.
3. Frequency Tracker (On switch located in back and power light goes on)
Adjust the frequency shift to 5 and multiplier to XI 00 on the Fi0 <
40Mhz.
4. Oscilloscope on (use ch. A)

Adjust- the vertical to lOOmV/div
-
the horizontal to a vicinity of 100 msec or .5p.sec
by pulling out
the 10X knob

-
Trigger the signal
( If the signal for some reason or another does not appear, scan for
the signal using the horizontal knob.)
5. Ensure the laser is at 1 .2 meters as given in the technical preparation
6. The mirror should be adjusted to ensure the laser beam is reflected back
into laser.
7. the from the TV lens on laser appears as given in the

Verify output pattern
diagram on . Figure 4. 12a is the output pattern with the frequency shift.
In the picture to right of figure 4. 12b, the frequency shift will not appear.
If this does not occur, several adjustments can be made by an alien wrench
T in figure 4.13. Point R will adjust the reference beam

at points R, S, and as seen
to the largest
optimally hit the detector diodes
thus we should expect see
so it can
44
signal. Point S changes the angle of the Bragg cell in order to gain maximum
intensity of the laser beam. Finally point T will adjust the outgoing beam exiting
the TV lens which passes 6 mm from the centerline as sketched out in figure 4. 12.
I would strongly recommend not to touch these settings. I believe I was able to
Shitted
A. B.
Figure 4.12 TV Lens Shift Pattern by Dantec Electronics.
obtain an optimal setting for R, S, T. Another possibility would be to adjust

and
the optics internally by removing the cover plate. As discussed more briefly in the
users manual, I would really strongly suggest not even to consider this option of
adjustment without on site technical supervision. The reasoning behind this is that
if you break a prism or accidentally disturb something, the laser is out of
commission.
X25 X29 X66
>AuU
jr
M
SLV
*
Figure 4.13 Side View ofLaser with Adjustment Locations by Dantec Electronics.
Open the aperture all the way till the dial reads 1.4 and the XIA plate should
45
already be at its marked position.
Focus the TV lens as needed to enhance the signal seen at the oscilloscope
and you should be receiving the desired signal as sketched out in
45
Figure 4.8. If the distorted signal persists, review the technical background
and technical preparation for any possible oversights.
I would now suggest

testing the laser in use with the B&K analyzer while in
this setup. Simply make the connection from the Tracker to channel A of the
Analyzer and input the calibration factor from Table 1. found in appendix K.
Remember the analyzer's units will request the inverse of the calibration factor
Volts/(m/sec). Using the impact hammer setup as discussed in the excitation
section, tap on the top of the mirror's base, and you should be able to obtain a
heavily damped response. None the less, this should now provide you with some
confidence in using the laser.
4.1.3.3 Laser Vibrometer in the Application of a Cantilever

Beam
This section is comprised of applying the laser using two different types of
excitation techniques. The first excitation technique employs the B&K 8202
Impact Hammer and secondly the B&K Vibration Exciter System. The subject of
interest is the classical cantilever beam. The beam is found to be setup in two
different arrangements because of the excitation techniques as well as the
restrictions due to flexibility of the laser setup.
It found necessary to build a special fixture for the cantilever beam in

was
order for the laser beam to sense it. The following sketch shows how the specimen
was clamped to this fixture.
General Area
of Impact
Adjustable
in the
Vertical
Laser Beam
FRONT VIEW SIDE VIEW
Figure 4.14 Vertical Test Fixture
It is also shown that the clamping ability is adjustable in the vertical axis and is
bolted to the vibration isolation block. The adjustability of the cantilever beam on
the fixture gave more flexibility in adjusting the height of the laser table. It was not
46
possible to gain access to a isolation table in which I could set both the
vibration
test fixture and laser on datum. The next sketch describes how the laser
the same
was setup and the distance between the fixture and laser is 1.2 meters. The
only
this arrangement
problem with
obviously is that some of ambient vibration from
General Over View of the Laser Configuration used in Impact Excitation
B8K 2032 Analyzer
Isolation Table
~U^> rv3
PreAmplifier
7 /Impact Swing MotionJ

V
Impact Hammer
Figure -/./_>
the building or from people could be a problem at times. The excitation procedure
for the impact hammer is found in chapter five, Excitation Techniques.

laser setup contains the
The next application of random noise
using the
B&K Vibration Exciter System. In this system it contains the 4801 Vibration
Exciter Body with general purpose Head Type 4812 and the Power Amplifier
Type 1047. In the following the sketch, the output signal from the B&K 2032
Analyzer from the Signal Generator Output is connected by a bnc cable to the DC
Signal Input located in the back of the B&K 1047 Power Amplifier. The Exciter
Body Power (3 phase), Interlock Cables and Signal to Exciter Head should be
already connected as sketched out in the following figure. Greater detail of the
excitation method and setup will be given in the excitation chapter. It is shown that
the shaker table was rotated from its normal upright position to a side position.
With the cantilever beam locked into position by the fixture on the shakertable, the
laser was set on the tiled floor and positioned 1.2 meters away.
In both of the laser setups using the modal hammer and the shaker table, the
settings for the laser should follow the same guidelines as previously discussed.
Overall the oscilloscope proves be the only way to really know if you have
to
obtained alignment when adjusting the parameters of the laser. I hope the
suggestions given will aid in the minimization of time wasted so that the real time
can be spent in analyzing your application. The random vibration application did
not run as smoothly as it did in the application of the impact testing. This problem
was not really understood as to what was causing the poor frequency response.
47
Getting the laser to work was a real challenge, and there will be further discussions
concerning this matter in Results and Discussion of chapter six.
GENERAL SETUP OF THE LASER IN APPLICATION WITH THE SHAKERTABLE
PREAMP
POWER
AMPLIFIER
SHAKERTABLE
\
\
ACCELEROMETER
Figure 4.16
48
4.2 Piezoceramics
4.2.1 Description
The piezoelectric ceramic was purchased from Morgan Matroc Incorporated

of the Electro Ceramic Division. The selected shape for this application exists as a
square plate and is made of a lead zirconate titanate material. The PZT-5A (DOD
Type II) plate can be used as a receiver or generator in accelerometers,
hydrophones, and vibration sensing. This type of material has high sensitivity,
permittivity and time stability. The piezoelectric ceramic has dimensions of 1 inch
by 1 inch and a thickness of inch. The weight of the
.009
piezoelectric ceramic is
1.13 grams [28,29].
Piezoelectricity is property that exists in certain types of crystalline

a
materials and was discovered by Pierre and Jacques Curie in the 1880's.
Piezoelectric ceramics are hard dense materials that are chemically inert and
immune to moisture as well as other atmospheric conditions. These materials will
change shape while under mechanical pressure and produce a voltage according to
this pressure. The voltage is created by the change in the crystalline structure. The
opposite can occur when a voltage is applied across these materials in which are
produced dimensional changes. Piezoelectric ceramics can be made into all sorts
of shapes, and the plate shape ceramic was selected for this application because of
the orientation of the poling axes. Poling is referred to as the polarity of the
ceramic element when it experiences a change in dimension from an applied
voltage or mechanical force. As seen in the Figure 4.17 below [28,29], the poling
axis of a ceramic element can be identified by the y axis where a potential is
created by
EFFECTS OF POLING
AFTER POLING
BEFORE POLING
AB'
1
f-B 1
0 0
49
Figure 4.17 X lorgan X latroc Inc.
the pos and neg. leads. The change in dimension A is greatly exaggerated but the
poling process does change the dimensions of the ceramic element. The electrodes
are created on the opposite ends of the ceramic that is being stretched and the sides
parallel to Y axis constrict. Other planes of the ceramic element can create
electrodes through dimensional changes as seen in the Figure 4.18 below [28,29].
The deformations in these planes can show how the potential is created during a
thickness shear, face shear and thickness shear. It is also not uncommon for two or
more of these actions to occur at the same time. Obviously the dimensional change
will be dictated by the shape of the ceramic as well as the composition, the
POLING OF PLATES
0 0
1
^^ 1
LZ3
A] THICKNESS SHEAR
0~~
J
0
B] FACE SHEAR
o-
0-
C) THICKNESS EXPANSION
Figure 4.18 Morgan Matroc Inc.
orientation of the poling axis and the location of the electrodes. In my case, the
cantilever beam will stretch and contract the piezoelectric plate as it vibrates and
create a voltage which will be fed back into the analyzer. As seen in Figure 4. 19,
PIEZOELECTRIC VOLTAGES FROM APPLIED FORCES
TENSILE FORCE POLING COMPRESSIVE FORCE

AXIS
I
+
+ ' +
COMPRESSIVE
1 >
FORCE TENSIL
'
~~
B FORCE
1
1 i i 1
[ [T
r.
A 0 0
Figure 4.19 Morgan Matroc Inc.
50
the compressive and tensile forces will alternate the polarity of the voltage at the
terminals [28,29].
Another aspect of ceramics that one should be aware of is the time stability
and aging rates. The aging rates of piezoelectic ceramics will change
gradually
over time depending on the composition and the process in which it is
manufactured. The time period between poling apparently effects the stability of
the piezoelectric ceramic. According toMorgan Matroc, it is found that the

material will become more stable if the time period is longer after poling. The
aging process will be accelerated in any ceramic if exposed to one or more of the
following conditions: high mechanical stress, strong electric depoling field and
high temperatures approaching the Curie point. The Curie point for the PZT-5A
plate is around 365 C as listed in the typical values of Lead Zirconate Titanate
Materials. I have enclosed this table of information concerning the Lead Zirconate
Titanate Materials from Morgan Matroc Inc. in Appendix M. The Curie point is
the maximum exposure temperature which the ceramic material can withstand.
When the ceramic element is heated above this temperature, all the piezoelectric
properties are lost, and it becomes a centrosymmetric structure. Below the Curie
temperature, these materials are considered noncentrosymmerric because of its
crystalline structure . This type of crystalline structure provides a net dipole
moment within the crystal unit cell. The voltage limitations of a ceramic is
typically between 500V/mm and 1000 V/mm for a continuous application.
Another important factor is a piezoelectric ceramic which can be depolarized by a
strong electric field with polarity opposite to the original poling voltage. When it
comes to high mechanical stress, the piezoelectric ceramic will depolarize
dependent upon the type of material as well as the duration of the applied stress. It
is interesting to note that for impact applications the material behaves
quasistatically or non-linear for pulse durations of a few milliseconds of more. The
piezoelectric effect becomes linear when the pulse duration reaches the
millisecond time frame. This is due to the short amount of application time in
comparison with the relaxation of the created dipole [28,29].
After receiving the piezoelectric ceramics from Morgan Matroc Inc., the
step was to determine the best adhesive to bond
next the ceramic to the aluminum
beam. Three types of adhesives were investigated and at this time it was thought
that the adhesive playing a major role in affecting the damping of the
was
aluminum beam. Later it was discovered not to be a material bonding issue at all
and I will discuss in the results and conclusion section what had caused the
None the less I thought it would be interesting to discuss in

damping problem.
some detail what I had learned concerning these adhesives. The best adhesive of
the three I had experimented with would be the silver conductive adhesive. This
51
was purchased from the Acme Division of the Allied Products Corporation in
Newhaven Ct.. Acme Division which identifies this as an E-solder No. 3021. The
purpose for selecting this type of adhesive was for its conducting properties, and
its resistance and capacitive effects could be assumed negligible in affecting the
damping. The primary benefit of this adhesive is that it allowed a conducting
pathway through the beam to the fixture where it could be grounded rather than at
the beam. This helped minimize any mass loading effects seen at the beam. The
resistance of the ceramic through the thick section was measured to be about
0.0201 Ohms. The silver conductive adhesive material has superior strength, and it
is very difficult to remove from the beam. The other two adhesives were the
Devcon 5-Minute Epoxy and the Ludlow Technical Paper double sticky tape. The
double sticky tape was used in the attachment of the piezofilm, and I believe this
tape will work well for long periods of testing before detachment. Also the tape
was thin enough to still allow
grounding at the fixture. The Devcon 2 partepoxy
when in application with the shaker table was not able to withstand long durations
ofcycling and at mid to high levels of stress The Devcon epoxy would simply
.
to break away from the aluminum

start beam. It was for reasons due to the
damping problem that I had investigated high levels of gain. High levels of gain
would be considered 5 and up on the amplifier on the shaker table equipment.
Similarly as with the Ludlow tape, the Devcon adhesive allowed grounding at the
fixture as well. These nonconducting adhesives allow a conducting path due to
capacitive coupling. These adhesives act as an insulator when sandwiched between
two metals with an applied AC signal. Another important note to be remembered is
that if any of the conductive material from underneath the ceramic were to leak out
and make contact with the top surface of the ceramic, it will short itself out and no
signal will be present.
The first idea in attempting to bond the ceramic to the beam was a
combination of two adhesives so the prevention of any conductive material could
not leak out and make contact with the top surface of the ceramic. What I did first
was to size a square piece of the double sticky tape to fit the dimensions of the
ceramic and then cut out in the center of this piece another square. This middle
square of sticky tape was removed and filled with conductive adhesive material.
This would then guarantee me a conductive path as well as creating a mechanical
barrier which successfully eliminated any leakage of conductive material. After

further experimentation with these materials, I found it was possible to use the
conductive epoxy alone and if you happen to get some of the conductive material
on the top surface by accident, it is possible to scrape away the excess amount
with a razor blade. The Figure 4.20 is a representation of the ceramic bonded to
the aluminum beam along with a wire solder on to it. The wire was a thin bnc
cable and the shielding that was stripped back and soldered to another wire with a
banana clip soldered at its end. The banana clip could now be fastened to anything
52
on the beam fixture in order to close the loop back to ground. I had found any of
the nearby clamping screws to be a good place to attach the banana clip.
PIEZOELECTRIC CERAMIC CONFIGURATION
CONDUCTIVE ADHESIVE
-
^ "
\ -, SIDE
j VIEW
\
CERAMIC
TOP
VIEW
WIRE SOLDER TO CERAMIC
Figure 4.20
In Figures 4.21 and 4.22, these sketches are the setup for the impact testing
and the shakertable testing. In the impact testing figure, the hammer would
generally be impacted the free side of the beam just behind the ceramic area.
on
The impacts were lightly tapped and the hammer was held in a pendulum format
General Setup for the Piezoelectric Ceramic
Grounded Lead from BNC Cable
Piezoelectric Ceramic
attached by
silver conductive adhesive
from Acme Division
Figure 4.21
ShakerTable Piezoelectric Ceramic

Top view of the with
Fixture Screws
j-
ShakerTable
'
Fixture Plate
BBNC
Cable Taped
Cantilever
Beam with Piezoelectric Ceramic
Figure 4. 22
53
for striking. Each time the beam was inserted in the fixture, the same amount of
torque was applied and a square was always used to ensure the beam was normal
to the edge of the fixture. The beam wasidentified precisely at 10 inches using the
square in order to maintain
repeatability whenever the beam was taken off the
fixture. Any slack of the wire would be taped up in order to minimize any stress
applied to the beam and the same was done for the shaker table setup as well. The
only other thing left to do is to plug the bnc cable into channel B of the analyzer.
Again the shaker table setup is essential the same as in the impact case and a
sketch of this setup is shown in figure 4.22. Ensure the beam is clamped correctly
into the fixture of the shaker table and connect the microdot cable to channelB of
the analyzer. Proceed to fasten the banana clip to a nearby screw and tape the
cable where needed.
54
4.3 Piezofilms
4.3.1 Description
The piezo film components were purchased from ATOCHEM Piezo Film
Sensor Division. The type FDT1-028K was selected and can be used either as a
dynamic strain gage or as a contact microphone. This piezo film is made of a thin
sheet of Kynar with a film of metal on each side. The film elements contain silver
ink patterned electrodes and the circuit lead lines are made of mylar which allows
it to be very flexible. The piezo film weighs .6772 grams and without the blue
female connector it weighs .1759 grams [30,3 1].
The piezo film can come in a variety of patterns fit any application.
to
Many designs have already been optimized for patterned electrodes for convenient
lead attachment, common mode rejection and shielding. These patterns are made
through a process well known in the photolithography field as sputtering. The

custom patterns are sputtered through a mask and photo-etched or screen printed
conductive ink [31].
The FDT1 peizo film is applied as a dynamic strain gage where it performs
mechanical to electrical work or vice versa. This type of piezo film will be used in
the application of the vibration sensing of a cantilever beam. The sensitivity of the
ATOCHEM PIEZOFILM
FDT1-028Kwith protective coating
y
44mm X 20cm
Figure 4.23
piezo film is quite remarkable and requires no external power. The only reason a
charge amplifier would be needed is in the operation of detecting a fraction of a
hertz. The other alternative is to use simple high impedance FET buffer circuits
where the signal levels are relatively high. The extreme sensitivity is largely due to
the format of the piezo film material. Very large stresses are created by the
longitudinal forces because of the very thin thickness of the film. This is why the
piezo film becomes very sensitive to vibration normal to its surface. By
placing the
sensor between two compliant
materials, it can effect the sensitivity in the width
and length directions up to a ratio of 1000:1. The piezo film transducers can be
compared to a strain gage but as long as the area is uniform. The output energy
from the piezo film is proportional to the volume of the film stressed. Thicker
films therefore generate higher voltages but form smaller capacitors. Remember
the area of the film that is not under going stress will act as a capacitive load on
the active area and should be minimized if required. Corrosion can sometimes
occur because of the metallization especially when they are handled. Some
adhesive systems will promote
tarnishing on some of the electrodes. The acrylic
type of adhesives are generally the most active in tarnishing, and conformal
coatings or laminates are frequently applied to maintain good surface quality. If

acrylic adhesives are needed in the procurement of the film, the electrode
thickness should be increased in order to minimize such corrosion effects. In

general the synthetic rubber resins, epoxies and cyanoacrylates are applied in the
lamination and assembly [31].
Other considerations that could effect the piezo film's performance is
driving the film at a high voltage and high frequency. The dissipation factor of the
film may lose a substantial amount of energy in the form of heat. The surface
resistivity of the electrodes may be very high localized currents
effected as well as
which may be encountered. Operating outside the recommended limits may result
in arcing which usually destroys the device. The capacitive nature of the piezo film
may become sensitive to electromagnetic interference such as the fluorescent
lights. The effect from the lights will drown the output signal from the analyzer,
and any strong alternating currents nearby can be a problem as well. Therefore you
may need to ensure the leads are shielded. Another common problem could exist
especially in self-sensing actuating applications when one element is driven while
another is receiving the vibrational signal. Great care must be taken in order to
avoid crosstalk interference. The piezo film is incapable producing large forces
of
or large displacements, but if sufficient gain is applied between two piezo

elements in a structure, it can cause it to self oscillate at its natural frequency.
Figure 4.24 on the next page shows the equivalent circuit for a piezo film. A useful
model for the film which applies in most cases except for ultrasonic applications is
a strain-dependent voltage source in series with a capacitor. Any resistive load will
form a divider network with a simple RC high-pass filter characteristic. The cut
off frequency
y is given
& by
J fn
Jo
= where r = RC is the time constant.
1
2;rRC
Operation below the frequency will give an output signal proportional to
cut-off
the rate of the input parameter. Applying constant stress will generate an initial
level followed by an exponential decay rate (Exp{RCY). The capacitive load will
extend the time constant but decrease the magnitude of the response. Note also
that energy will be lost when transferring charge form one capacitor to another.
56
Large capacitive loads are useful in attenuating the very large signals
arising from
powerful impacts on the order of hundreds of volts. But in this application a using
modal hammer, we need not concern ourselves about this as long as the impact is
Equivalent Circuit for the Piezo Film
Charge Generator Voltage Generator
i o
(Cf -
Piezo Film capacitance]
Figure 4. 24
soft. Tapping the structure with great force would most likely mass load the input
signal causing an overload. When using the power amplifier with the shaker table,
the gain could very well produce a high voltage from the piezo film but I never
saw it get any greater than 100 volts. It could be possible to achieve several
hundreds of volts
depending upon what kind of amplifier you build in line with the
piezo film [31].
At first, I had some difficulties in getting the piezo film to achieve any
discernible results. I started out with the presumption that the piezo film had
be to
shunted to parallel some resistive load. This was prepared on a bread board and
connected by a bnc cable. The frequency response of the cantilever beam
initially
was terrible, and some of these plots in Appendix H. After consulting
can be seen
with several people from Eastman Kodak Company, particularly Mike O'Brien and
Richard A. Berg (my father), a plan of action was developed to try and determine
the source of the problem. It was thought to be noise interference from some
source in the room, or the signal could be just too weak for whatever reason. The
next decision was what type of adhesive to use for bonding the peizo film to the
cantilever beam; later, the bonding material was thought to be the source of our
noise problem.
I was able to conclude after exploring many types of resistors in parallel
with the piezo film on the bread board that this was the wrong thing to do. The
frequency responses were very poor using this circuit arrangement. From the
brainstorming process, I was able to set up a program of experiments in order to

solve this problem of noise. The first suggestion was to build a voltage
preamplifier, thinking that the piezo film signal was too weak. Determining the
type of amplifier to build was discussed with several people, and the diagram is
57
drawn in Figure 4.25 on the next page. The op-amp used in this circuit was an
AD545 where the adjustable pot gave 2.2MQ to 22MQ. The box allowed for
banana jacks as the input but later on, I found the need to add a bnc input
connector as well.
Applying the preamp in conjunction with the bread board, I
tried an
array of resistors and again found no The to
success. next thing was try
eliminating the
r~| T<; 1 The Berg Preamplifier

MVvHlA-vW
2.5 m
|
Figure 4.25
bread board, resistors and preamp. The frequency response looked much better,
but a strong 60 Hz signal was knocking out the first resonance of the beam. The
next but last suggestion was to ground the fixture. The piezo film acting as an
antenna though coupled with the fixture was most likely seeing the 60 Hz signal
from the lights in the room. By grounding the fixture to a nearby outlet, I had
success. The 60 Hz. signal was gone! The fixture was grounded using a thick
wounded copper wire where one end was tied to a welding clamp and the other
end was inserted into the ground section of the wall socket. The ironic thing was
that the double sticky tape had nothing to do with the noise problem at all, and I
did not have to worry about the adhesive having any profound effects. After
finding that grounding the fixture eliminated the noise, I also found the voltage
preamplifier was not needed anymore. The signal strength of the peizo film was
itself. This type tape better than any other type

excellent by of adhesive worked of
material I found, and it is made by Ludlow Technical Paper product #8602. The
advantage of this tape is its strengtrTof bonding to the aluminum and ability to peel
back up for reusing or repositioning.
On the next page, Figure 4.26 represents the general setup for applying the
piezo film when using the modal hammer excitation method. The bnc cable was
adapted with a female connector on the end so the piezo film could be easily
disconnected at any time when changing setups. Taping the lose cabling to the
58
fixture is important because its weight will add stress and could effect the results.
Another important point to make is that the piezo film should be positioned close
to the clamped section of beam. This is because the strain energy is the highest in
this area of the cantilever beam. The hammer will strike softly on the opposite side
of the beam in a pendulum type motion. Figure 4.27 represents the setup when
General Setup for the Piezo Film
Welding Clamp
B&K
2032
Analyzer
Outlet
Input
Piezo Film attached
by Ludlow Double Sided

Sticky Tape
Figure 4. 26
using the piezo film with the shaker table. In this case the only modification
needed is to secure the cantilever beam to the fixture on the shaker table as well as
grounding it. To ground the fixture, use the same copper wire and wrap one end
Top view of the ShakerTable with Piezo Film
Fixture Screws
r
ShakerTable
Fixture Plate
BNC
Cable Taped
Cantilever
Beam with Piezo Film
Figure 4.27
around a pipe located near the All plumbing must be

shaker table.
nearby
standard building code; therefore, less
risk is taken in getting
grounded by
electrocuted by insertion into a wall socket. It was found by bending the other end
into a hook and simply placing it in
an unused threaded hole on
of the copper wire
the fixture would work just as good. Again make sure the bnc cable is taped to the
if needed, to the side of the shaker's exciter body in order to reduce
fixture and
stress on the first taped section.
59
4.4 Piezoelectric Accelerometer
4.4.1 Description
The piezoelectric accelerometer type 4344 is Bruel & Kjaer

made by and
was readily available in the vibration analysis laboratory at the Rochester Institute
of Technology. The accelerometer has a detachable microdot cable and weighs 2.7
grams. The type 4344 was the smallest accelerometer available in the
laboratory
but the lightest accelerometer that Bruel & Kjaer has to offer today weighs .65
grams excluding the permanently attached cable. These accelerometers are

designed to operate over wide
frequency and dynamic ranges. The accelerometer
is considered the standard product in measuring vibration and is tested to meet or
exceed all relevant IEC, MIL, DIN and CISPR standards. The Bruel and Kjaer
products are designed and manufactured to the most stringent standards of quality,
performance and reliability. The 4344 accelerometer calibration is traceable to the
National Bureau of Standards in Washington D.C. [35,38].
The piezoelectric material used in the small type of accelerometers today is

an artificially polarized lead zirconate titanate PZ 27; very similar to the PZ 23
family. The PZ 23 family basically differs in that it has a higher sensitivity. Like
the other piezoceramic sensors, it comes from ferroelectric ceramics where it can
be produced in any desired shape and their composition can be altered in order to
give them special properties. Accelerometers of these type use piezoelectric
monocrystalline materials such as quartz where the composition is fixed and the
shape is restricted by the crystal size in which they are cut. Monocrystalline
elements generally have a lower sensitivity and internal capacitance than those
with ferroelectric ceramic elements [35].
The environment where an accelerometer is exposed can sometimes be very
harsh and must meet many types of industrial demands for successful vibration
measurements. The list of demands placed on the accelerometer must be able to
withstand several effects and is why selecting the appropriate

that accelerometer
can be crucial in order to minimize the sensitivity. Some of these environmental
effects can be humidity, electromagnetic fields, high sounds, transverse vibration,
base bending, temperature fluctuations, high and low operating temperatures, and
even radiation [27,35].
It is quite remarkable that Bruel & Kjaer has an accelerometer to offer for
each of these extreme cases. There are three types of mechanical structures used in
60
the design of the Bruel and Kjaer accelerometer. These three accelerometers are
the Planar Shear, Center Mounted Compression Design, and the superior design
Delta Shear. The Planar Shear and Delta Shear undergo shear deformation of the
piezoelectric material in their designs as seen below in Figure 4.28. Figure 4.29
Planar Shear Delta Shear*
Planar Shear and Delta Shear? designs. M=Seismic Mass, P=Piezoe-
lectric Element, R=Clamping Ring and B=Base
Figure 4. 28 Planar Shear and Delta Shear Accelerometer Designs by Bruel & Kjaer
Figure 4.29 Centre Mounted Compression Shear Accelerometer Design by Bruel & Kjaer
describes the compression type design. The 4344 accelerometer is a compression
type design. The Delta Shear design gives a high sensitivity to mass ratio
compared to the other designs. The Delta Shear type of accelerometer offered
through the Bruel & Kjaer's vast product line is a good example where it may be
deployed with the right piezoelectric composition for resisting acoustic sensitivity,
pyroelectric effect and base strains. In the case of the pyroelectric effect,
piezoelectric crystals and ferroelectric ceramics become charged by temperature
inequalities and changes. The piezoelectric element in the compression type of
accelerometer is more susceptible to erroneous ouput results than the shear type
designs due to these temperature transients. The compression design can
sometimes be affected by base bending which is another disadvantage but in a
control environment, it is used as a standard reference accelerometer for
61
accelerometer calibration. In humid conditions, generally the
housing of the
accelerometer is welded or epoxy sealed in its construction therefore isolating the
piezoelectric materials from its effects [35].
The other concern is how to mount the accelerometer so it does not effect
the results of your measurements. Such factors could be the location of the
accelerometer and the way it is attached to the specimen. In my case with the
aluminum beam, bee's wax was the method
When mounting an
of attachment.
accelerometer, be sure the surface isfree from dirt and grease. The bee's wax was
supplied with the accelerometer kit and its
officially called Wax YJ0216. Needing
only a little wax, the recommended way of applying it is to spread a thin layer on
the location where you wish to mount the accelerometer but ensure the area
covered by the wax is larger than the foot print of accelerometer. You may need to
roll the wax in around in your fingers in order to warm it up so it can flow better
over the test surface. While applying pressure, rotate and slide the accelerometer a
small amount until it feels securely fastened to the beam's surface. The advantage
of using bee's wax as an option over stud
mounting the accelerometer is that it is
quick and easy. The bee's
wax mounted resonance
frequency is slightly lower than
that of stud mounting. Stud mounting is the best of all methods. The disadvantages
of bee's wax mounting is the serious temperature limitation of 40C and a
maximum vibration level of approximately 100 ms2. Having too much bee's wax
will lower the mounted resonance limit the measuring frequency

frequency and
range. The reason for the temperature limitation of bee's wax is simply because it
will melt at these high temperatures and

eventually the accelerometer will fall off.
Though it is fairly obvious, the location of the accelerometer should be
given much thought. The main sensitivity axis of the accelerometer should be
placed in the direction of the desired measured direction. The red dot on
accelerometers should be positioned in the direction of the maximum transverse

vibration. Make sure the accelerometer is never mounted on a nodal point and
understand that the dynamic response of the structure can be changed as well. The
mass of the accelerometer can alter the dynamic properties of a structure. The
accelerometer can influence the mechanical impedance of system but if the mass is
small compared to the mass of the structure, then the changes in vibration will be
negligible. A general rule of thumb is to keep the accelerometer mass less than one
tenth of the mass of the structure. In the case of our aluminum beam, the natural
frequency was affected slightly by the addition of the accelerometer mass [27,35].
A vibration preamplifier model 2626 made by Bruel & Kjaer was used in
this application for several reasons. The vibration preamplifiers essentially
perform the role of converting the high impedance output of the piezoelectric
accelerometer into a low impedance signal which allows for direct transmission.
This preamplifier is designed as a low noise charge amplifier which can be used as
a transducer as well as for accelerometers. It also has the capability of filtering low
and high frequencies to reject unwanted signals. The preamplifier has an overload
62
light to guard for
voltages that may exceed the preset voltage and it can be set
by
the sensitivity knob located in the center of the preamplifier. Charge preamplifiers
are in preference to voltage preamplifiers because the charge preamplifiers are
insensitive to changes in lengths of cables. When using the voltage preamplifiers,
the system is sensitive to cable changes. Charge amplifiers do not charge
amplify
but produce an output voltage that is proportional to the charge. The voltage
preamplifier does amplify the voltage but is proportional to the input voltage. The
voltage preamplifier can also affect the low frequency performance caused by
changes of the input resistance from the In the
cable. use of
very long
accelerometer cables and low gain settings, noise of the charge amplifier will
increase, therefore decreasing the signal to noise ratio of the measurement. Noise
at the input can be caused by a combination of two effects, ground-loops and the
triboelectric effect. The triboelectric effect is the generation of charge in the
coaxial cables due to the movement of the cable during vibration. It is better to
fasten the cable onto the vibrating surface. Ground-loops can occur in large
machines where the accelerometer is not grounded properly. For example,
machinery housing may not be at earth potential, and a voltage drop will exist
along the cable. Lastly, remember that high electromagnetic fields can affect these
cables therefore creating noise as well [35,37].
The setup for the accelerometer is rather straight forward for both impact
excitation and random excitation. The impact excitation method utilizes the same
fixture as used for the other three types of sensors. The microdot cable was
centered along the aluminum beam and taped off in the clamped area of the beam.
The microdot cable inserted into the input jack
was of the Bruel & Kjaer
Conditioning Amplifier Type 2626. The output jack of the amplifier was
connected to Channel B of the Bruel and Kjaer 2032 Analyzer. This requires a
BNC and microdot split cable. The accelerometer was mounted on the end of the
aluminum beam using bee's wax supplied by Bruel & Kjaer. The Figure 4.30
General Setup for the B8K 4344 Accelerometer
B&K
2032
Analyzer
Accelerometer attached
by Bee Wax
B8K. CONDITIONING AMPLIFIER

TYPE 2626
Figure 4.30
illustrates the general setup for impact testing. With each accelerometer kit, a
calibration sheet can be found that contains much information concerning your
63
type of accelerometer. The calibration chart of the accelerometer is basically a data
sheet describing the resistance, polarity, capacitance, cable capacitance, charge
sensitivity, voltage sensitivity, weight, maximum transverse sensitivity and the

undamped natural frequency. It contains environmental information concerning
maximum temperature and shock and magnetic and acoustic sensitivity. The data
sheet even provides physical dimensions of the accelerometer as well. The charge
sensitivity of 4344 accelerometer is 2.97 pC/g which is then dialed in under the
Conditioning Amplifier Type 2626

Bruel 8 Kjaer semitivitypC/Unit
(T ft
Ve'1ad I0dbfi.m
Vol. /Unit Out
Overload
Q RESET
KT ..(-0)
piAUiut .OlV^w
Upper kHz.
Low^rlfa. ->1
30 1
3db JO
10
0 &
Input Output
Q 0
Figure 4.31
Top view of the ShakerTable with B&K 4344 Accelerometer
Fixture Screws
ShakerTable
B&K 2032
Analyzer
CH. B
Fixture Plate
J
B8K CONDITIONING AMPLIFIER
Cantilever TYPE 2626 BNC
Accelerometer Cable Taped
Beam with
Figure. 4.3-
shown in Figure 4.31. The red light will indicate

sensitivity of 2626 amplifier as
turned to 3 Hz. and the
the decimal point. The lower frequency knob should be
upper frequency knob can be set at 3,000 Hz. Similarly, the shakertable setup will
different the accelerometer and preamplifier in this
be no when using
64
arrangement. The only difference could be the setting of the volts per unit knob.
Even in the setup using the shaker table the preamplifier will be the same except
for the Volt/Unit Out may be different and that is dependent on the vibration level
seen by the accelerometer. The units of the charge sensitivity will be matched
under the channel B menu of the B&K 2032 analyzer and the coefficient of these
units will be the volts/unit from the preamplifier. The overload light will come on
depending how much vibration is seen by the accelerometer from the amount of
excitation input to the structure. The overload and 20db from overload will be
discussed in further detail under the excitation section. The cable of the
accelerometer can be seen taped on the fixtures in both setups. The other
precaution with these type of cables is that I found them to be too stiff for this type
of application. As seen in the shaker table sketch, the cable was stiff enough that it
never could touch the aluminum beam. None the less, the cable was positioned in
the matter in which exhibited minimum stress so the best dynamic response could
be measured. Also the excitation energy was intentionally small so that it could
minimize the creation of these noises. The accelerometer was placed at the end of
the beam simply because this is where it would see the most acceleration just as
the case with the laser vibrometer. Also the accelerometer's base can not be placed
at the very edge of the beam because the bees wax needs to come past the base for
secured mounting.
65
5. EXCITATION TECHNIQUES
5.1 Impact Excitation
5.1.1 Description
Impact is the striking of a structure through the use

excitation of an impact
hammer. Impact excitation can be a
veiy fast method of performing a nansient test
for determining the dynamic characteristics in a structure. A Type
8202 impact
hammer from Bruel & Kjaer features a calibrated force transducer, three different
tips, an additional end mass, an optional built in line drive amplifier and a micro
dot cable routed through the shaft for protection as well as
preventing it from
interfering with the impact of a test. The hammer also comes with a very important
specification sheet which gives the calibration factor that is for operating
needed
the analyzer. The figure below is a sketch of what the Type 8202 impact hammer
looks like [12,38].
Bruel and Kjaer Impact Hammer Type 8202
<2
/A
Figure 5. 1
The impact hammer creates a force that transfers energy into the structure.
The energy going into the structure occurs for a small amount of time, and the
shape of the force signal is dependent upon the type of tip on the hammer, the
mass of the hammer, and the type of structure under investigation. The force is
sensed by a sensor called a force transducer located at the tip of the hammer called
the force pulse. The pulse can be viewed by selecting Time Channel A. from the
66
B&K 2032 Dual Channel Analyzer and an example of a such plot can be seen in
the upper trace in Figure 5.3 on the next page. The frequency bandwidth from the
force signal will be dictated by the characteristics of the hammer. The increase in
stiffness of the hammer tip and structure has the effect of decreasing the signal
length and increasing the frequency bandwidth. The effect of additional hammer
mass can widen the force signal and subsequently lower the cut off frequency. The
impact hammer should be recalibrated before any measurements are taken when
making a change in mass of the hammer [6, 12].

Impact excitation has several advantages and disadvantages in its testing.
The advantages of impact testing are that the testing is very fast and usually
requires a low number of averages; there are no fixtures required and it is portable
as compared with the shaker table; lastly, it is easy to use. The trouble areas using
the impact hammer are: the force signal creates high crest factors making the
impact technique inapplicable for nonlinear systems, it requires special
windowing, gives desceptive results, is easy to misinterpret ringing and has limited
control of the excitation bandwidth. The crest factor is defined as the ratio
between the peak and the standard deviation, RMS, in the signal [6,7,1 1, 12,36].
The setup involves the following equipment: the Bruel and Kjaer 2032 Dual
Channel Signal Analyzer, the 8202 Type Impact Hammer, the 2626 preamplifier
for the impact hammer, the isolation table, the fixture for the beam and a selected
sensor. The diagram below gives a general overview of the setup. When
Kjaer's Analyzer, Channel
performing dual channel FFT analysis on the Bruel &
B. is always
General Overview of an Impact Excitation Setup
B8K 2032 Analyzer
1 {Impact Swing Motion)
Figure i 2
dedicated for the sensing input and Channel A. must be the excitation input. The
2626 charge
B&K impact force transducer hammer is connected to the input of the
67
\t+
' j i i * *
Q
1'
: . ,.
^
*
.
|
1
i l
-
0) 1 7 a
"
a 3
'
d (T)
0 i. , ; i
E
in
in i- 1 : ID tO CD
- O 1 E e cn
w .,
o a - r-în
' 1 ?
E : h-CDtO
*
a !
i
B
a>
3
i 1
L IMOl
> . i i
1 1- ts mm
r*
*t
1 i
[ t> B
'O
s
z '
a H -LU
IH L UJ -X_I
K
< . r a:x<w
sx
i. ID
I ! . to
*
f d
d
*
00
1- f
3 - in .sc
a. 1- in
z
'
d
?H r d
1
- *
3- *
-t
d j
r d
J <
< j UJ
r
r "* ^r
r m
- cn 1
*
a
n i
1"
d ^-J~~~- 1
f
i ,
i
^^gp^ *
10
; i a i-
g^^fc^^^. - CM (J)
1- CM +
 ,
' r d 1
3
ihjE ? a.
'
3 ^ _UJ
H +
-
/
-
X CO
'd .
, '
*~
,1
Q
7. i
in
a S
a
in a a
*-4
amplifier, and the output cable from the amplifier goes to channel A. of the
analyzer. The desired sensor is connected B. of the
to the input of channel
analyzer and the
corresponding setup information for each sensor can be found in
chapter four. The beam's fixture is made of
aluminum, and it is designed for the
face of the beam to be set vertically as seen in the above sketch. This is due to the
requirements of the laser vibrometer. The area of impact was located
approximately 1 inch away from the clamped end of the beam. The method of
tapping the beam by hand was investigated; I found it best to swing the hammer as
a pendulum. The thumb and index finger would act like pin joint for pivoting the
hammer which was located at the end the handle.
of
Using my arm to generate the
momentum for the hammer, the hammer would swing slightly back and then
tap
the beam. Upon reflecting back towards myself, I would squeeze my thumb and
finger to stop the hammer from tapping the beam again. The elbow would always
rest on my leg while sitting in the chair. If the beam was fixtured lying
horizontally where the tapping would be in a downward direction, I found the
wrist muscles would often twitch for these type of small taps and cause problems
such as mass overloads and double hits [12,24,25].
5.1.4 Experimental Procedure for Calculating the Loss Factor
The process steps for calculating the material damping using the Bruel &
Kjaer 2032 Signal Analyzer and impact excitation are listed below in five parts.
This written procedure assumes the user has some experience and familiarity with
moving around through the menus on this analyzer [39,40].
1) Choose the type of sensor to be used andfollow the setup instructions given
in chapter four. Connect the cable to the channel A of the analyzer to the output of
the Type 2626 charge amplifier and attach the impact hammer to the input of the
charge amplifier. Set the sensitivity to 1.00 pC/unit on the charge amplifier as
stated on the calibration sheet from the hammer.
2) In the B&K analyzer, go to Measurement Setup and enter number 19 for

the Dual Spectrum Averaging setup using the default transient and exponential
windows for channel A and channel B respectively. The following changes made
at this point will be:
. AVERAGING set to LIN, 100 and MANUAL ACCEPT

. CH. A and CH. B must be 3 Hz. DIR.
. CENTER FREQ. choose BASEBAND
Move to the very bottom of the screen (under the generator function) and
press the number 3 and enter key. The SPECIAL PARAMETERS will
69
activate and enter 2032 for the DEC VALUE at the very end of the line. This
acts as a key to turn on the special parameters.
Set the FREQ. SPAN, to the desired
frequency range and the other settings
will
correspondingly default.
DELAY should be set at less than 10% of the period (T)
3) Go to the DISPLAY SETUP and set to 14 or toggle to TIME CHANNEL

A.. The objective at this point is to set the shift and length of the transient
weighted window in channel A. The shift of the transient weighted window will be
the same as the exponential window or transient window in channel B. The
following steps in determining these time values are:
While tapping the beam with the amount force

of you expect to tap during the
test, monitor the overload light from the charge amplifier. Set the Volts/Unit
Output until this light goes
out and the 20 db from overload light stays lit.
Set this valuefrom the Volts/Unit Output in the measurement setup for
channel A. ( ex. V/Unit was determined from the charge amplifier so key
.001
in lOOmV/N in channel A.)

Note: The same must occur when using the charge amplifier with the
accelerometer sensor.
Set the scale to 80db and change REAL to MAG. (at the top of the screen)
Prepare to tap the beam and press the AUTORANGE button. Then,
immediately begin tapping the beam at the amount of force expected to be
tapped during the test. Continue tapping the beam until the single light turns
off and the continue light goes on. This will preset the voltages on channel A.
and channel B. and create an overload signal that will flash on the screen as
well as on the charge amplifier when you accidentally tap harder than the
expected level of force.
Press the START button when ready to tap the beam. Ensure channel A. is
in the MAG. response mode. Remember to Autoscale the screen when
necessary.
On the screen, move the cursor to the left of the spike where the smooth part
of the curve begins.
The in the X: (under MAIN) is to be inputted in the SHIFT for

value
WEIGHT CH. A: and WEIGHT CH. B:

Cursor to MAIN and toggle this field till the DELT. is seen.
Press the SET SPECIAL CURSORS button and cursor to the right of the
spike. Read the AX: field and input this value for the transient LENGTH in
channel A:
70
4) Cursor to TIME CHANNEL A: and toggle to TIME CHANNEL B:. The
objective of this section is to set the exponential window LENGTH in
channel B. The following steps are:
Set the scale to 80db and change REAL to MAG.

Press the AUTOSCALE button if needed.
Under the TIME CHANNEL B: field, change the max value from the X: 0.0
ms + max to equal the time record T from the
value FREQ. SPAN.
information. This will allow you to view the entire time record.
Go to the SPECIAL PARAMETERS area and input either number 76

(lower) or 77
(upper) for identifying if the screen is in the lower or upper
trace. By pressing the FORMAT button, you are able to identify which
display item is on what trace. Enter a 2 in the DEC VALUE: at the end of the
line of selected trace, the artificial damping created by the exponential
window will be displayed. Refer to appendix J in the special parameters
section for more information if needed.

At this point the objective is to have a fair amount of signal to noise from the
maximum value of the exponential decay in the MAG. mode to about 40 db if
possible. In the case of lightly damped beams, it may not be possible to obtain
this amount of signal to noise ratio, and it would have to be determined for
that particular setup. The way to change the decaying slope is to have the
cursor in channel B. of exponential LENGTH and increase the value till the
slope reaches the appropriate db range. It could be as much as 10 to 15 db as
understood from Bruel and Kjaer. Therefore, the value placed here directly
effects the damping calculation and this is referred to as T as discussed in
chapter 3. According to Bruel & Kjaer, the larger the T the accuracy of the
damping will be improved.

If the exponential window was replaced with a transient window, the
LENGTH for this window would be determined the same way as above but
the damping calculation is now independent of the window length. This
would also involve the consideration of the signal to noise ratio.
5) This part will involve producing the frequency response of the interested
frequency range; zoom in on the resonance frequency of choice and
determine the damping for this resonance.
Toggle the FORMAT button till a dual screen appears.
Select at the top of the screen FREQ. RESP HI and ensure to the right of
this that it is set to MAG. Go to the bottom trace, and at its top, toggle till the
COHERENCE is displayed.
71
Prepare to tap the beam once and then press the SINGLE button and then
PROCEED in order to quickly take the 100 averages.
If the coherence screen looks good and the does not produce tap any
overloads or double taps, proceed to cursor in the FREQ. RESP. HI trace
and determine the resonance frequencies.
Go to CENTER FREQ. choose ZOOM
Select and toggle for the
frequency in FREQ. SPAN, to the desired
frequency range and the other settings will correspondingly default. See the
experimental plots in Appendix B. as reference.
Prepare to tap the beam once and then press the SINGLE button and then
PROCEED in order to quickly take the 100 averages.
Assuming our
tap was good and the beam did not break, proceed
talented to
the cursor in the FREQ RESP. HI. to determine the resonant frequency.
Press SET SPECIAL CURSORS ( this will zero out the X and Y)
Change the MAIN to REF.
Move the cursor from the resonant frequency in FREQ. RESP. HI trace and
cursor to the left of the resonance peak until a minimum of -lOdb change in
AY can be read.
Toggle the AY button till ELEM. (element #) appears and copy this value
into the DEC VALUE of SPECIAL PARAMETERS #41 at the bottom of
the screen. See Appendix B for a typical data sheet with the table of special
parameters and DEC value information needed in the extraction of the

resonance.
Cursor to the right of the resonance and note its element number. Subtract the
left element number from the right element number and input the difference
in the DEC VALUE of SPECIAL PARAMETERS #42
A number 2 should be placed in the DEC VALUE of SPECIAL
PARAMETERS #40 for the transient weighted window. See Appendix J in
the special parameters for more information on #40.
Special Parameters #47 and #48 correspond in viewing the weighted
frequency response from the applied windows by entering a two in the DEC
VALUE.
At this point, the resonance should look filtered as sketched on the data sheets
used in Appendix B. The response will look sharply cut off and now needs to
be tapered to smooth out the resonance. This can be subjective to a small

extent in how much taper should be given but in general, give as little
tapering as possible until satisfied.
The way to taper the left and the right side of the resonant is to input values in
the DEC VALUE in the SPECIAL PARAMETERS #43 and #44
respectively. for symmetry purposes these

Generally values should equal each
other. The tapering could range from 15 up to 25.
72
Go to the upper trace and cursor to the FREQ. RESP. HI and enter 28 to
bring up the IMPULSE RESP. function. Ensure to the right that it
says MA.
for the Hilbert Transform to activate.
Inspect for the straightest portion of the slope from the IMPULSE RESP.
MAG. and cursor to the top of the smoothest portion of the slope. Press the
SET SPECIAL CURSORS to clear the AY in the REF. area. Note: never
start on any portion of the knee of the Impulse Response Magnitude.
Move the cursor to the right on the screen while watching the AY change in
db. When the cursor reaches a db and is still a decent straight
-8.7
containing
slope within this bound, the time constant T is now determined by reading
the AX from above the AY.
The material
damping can now be computed from equations 3.5 and 3.6 in
chapter 3, Modal Testing. Figure 2. 1 1 may be used as well which is found in
chapter 2.
Repeat the process for the next resonance if desired.
5.2 Random Excitation
5.2.1 Description
Random excitation refers to the amplitude and phase of the excitation force
and it is a continuous signal which never repeats itself. This signal is generated
from the Bruel and Kjaer 2032 Analyzer and fed through the power amplifier to
the shakertable. The cantilever beam is rigidly attached to the shakertable which
produces the random excitation and the application of the four types of sensors
will have the opportunity to sense the response of this type of excitation from the
cantilever. Figure 5.4 below illustrates what a random signal may look like in
acceleration versus time [12,26]
Random Noise Signal
Figure 5.4
73
The amplitude of the random signal can only be predicted in terms of

statistical parameters. This can be described in terms of its power spectral
density
called the Autospectrum, its Autocorrelation Function and the amplitude
probability density. The Autocorrelation is related to the Autospectrum via a

Fourier Transform, and the amplitude
probability density is the probability of
containing amplitude values within an interval of Ax which is divided by the size
of that interval Ax . The amplitude probability density can describe how, on the
average, the instantaneous amplitude of the signal is distributed as a function of
the amplitude level but does information concerning the time
not contain
history or
the frequency content [12].

The random signals used in practice will have a normal or Gaussian
probability distribution. The analyzer will have the individual time records
containing the amplitude and phase for each frequency and across the spectrum it
will have, on the average, the same amount of
energy for all frequencies. The
random signal should be a constant spectral
density in the frequency range being
examined, and the Autospectrum should be flat in this frequency range. This signal
is sometimes called a band limited white noise signal. The structure is now excited
over a wide range of force at each given frequency due to the random
characteristics of the signal [6,12,26].
The main purpose for using a random signal is that it randomizes any non
linear effects and averages them to give the best linear approximation. One
advantage of using random noise excitation is that the signal can be shaped to fit a
zoom or baseband measurement for giving more dynamic frequency range in the
analysis. Thus the frequency outside the analysis will not be excited. An advantage
in using a shaker table over impact testing is that it eliminates the skill and error
created by the user [12].

The random signal is considered as noise at the input because of the non-
linearity in the signal, and therefore, the Frequency Response HI is the best
estimate of a linear fit to the system because we are trying to approximate the
response rather than describe the non-linear behavior. Frequency Response HI
relates as much of the output channel (sensor) to the input channel (excitation)
while minimizing the amount of noise at the output. Frequency Repsonse HI is
ideal for expected noise may be encountered when using these sensors.
which
Also, another type of noise that is created at the input is the impedance mismatch
between the shaker and the structure. This will cause the input force signal to drop
at the resonance frequencies of the structure. Frequency Response H2 is usually
insensitive to this the input assuming that there is no output noise coming
effect at
from the sensors. Frequency Response H2 is similiar to HI, but its purpose is to
minimize any noise that come from the input. The weighting function commonly
used in random excitation is a Hanning window because of its smoothness at the
74
entry and exit of the data record. The smoothness of the window will cause
leakage in the spectral estimate. To minimize this effect Frequency Response H2
could also be used unless a zoom analysis is performed. The only problem with
zoom analysis is the extra time needed to conduct the test. In my case, I decided it
was best to go with
Frequency Response HI using zoom analysis when
applying
the Hanning window [6,12].
The setup involves the

following equipment: the Bruel and Kajer 2032 Dual
Channel Signal Analyzer, the Bruel and Kjaer Vibration Exciter System, the 2626
preamplifier for force
sensing using an accelerometer, and a selected sensor. The
vibration exciter system is
up made of the 4801 Vibration Exciter Body with the
general purpose Head Type 4812 and the Power Amplifier Type 1047. The
diagram below When performing a dual
gives a general overview of the setup.
channel FFT Bruel & Kjaer's Analyzer, Channel B. is always

analysis on the
dedicated for the sensing input and Channel A. must be the excitation input. The
B&K Type 4369 accelerometer is connected to the input of the 2626 charge
amplifier and the output cable from the charge amplifier goes to channel A. of the
analyzer. The B&K Type 4369 accelerometer was screwed to the top plate of the
shaker's fixture. The cantilever beam was sandwiched between two parallel plates
that were attached to the head of the shaker table. The desired sensor is connected
to the input of channel B. of the analyzer and the corresponding setup information
for each sensor can be referenced back in chapter four [2 1,22,23,24,25].
GENERAL SETUP OF THE BRUEL AND KJAER'S SHAKERTABLE
DC
Input
PREAMP
POWER
AMPLIFIER
Cantilever Beam
SHAKERTABLE
Figure 5.5
The random signal is sentfrom the Signal Generator on the B&K 2032 Analyzer
via a bnc cable and connected to the DC Signal Input on the Power Amplifier
75
located in the back. The exciter
body 3-phase power cables should already be
connected to the power amplifier. The settings on the Power Amplifier were kept
constant except for the gain control. The standard settings are as followed
[21,22,23,40]:
1) Range A(rms) =
10
2) Head Constant(mm/s)/V =
58.4
3) Displacement Limit (mm) =
12.7
4) Current Limit A(nns) =
22
5) Ouput Impedance =
low
6) Direct Current Output (amps) =
5
7) Amplifier Gain =
operator select
8) Range V rms
=
3
The accelerometer was chosen to sense the force signals from the shaker
because there were no available force transducers that I could find. I did take off
the force transducer on the impact hammer and tested it in comparison with the
accelerometer. I did not see any noticeable differences in the coherence among the
two types in the frequency response. I did use the force transducer from the impact
hammer in determining the damping problem that I ran into with the loss factor for
second natural frequency on the shaker table. Chapter 6 will discuss this problem
in futher detail. This was also done to illustrate the coherence comparison.
5.2.4 Experimental Procedure for Calculating the Loss Factor
The process steps for calculating the material using the Bruel &
damping
Kjaer 2032 Signal Analyzer, the Bruel and Kjaer Vibration Exciter System and
random excitation are listed below in four parts. This written procedure assumes
the user has some experience and familiarity with moving around through the
menus on this analyzer [39,40].
1) Choose the type of sensor follow the setup instructions given

to be used and
in chapter four. Insert and position correctly the cantilever beam between the
parallel plates and tighten all the screws. Connect the cable to channel A of the
analyzer to the output of the 2626 Type charge amplifier and attach the 4369
accelerometer to the input of the charge amplifier. Set the sensitivity to 2.16
pC/unit on the charge amplifier as stated on its calibration sheet and set the units
V/m/s2
on channel A in the analyzer to
2) In the B&K analyzer, using the same Measurement Setup number 19 for
the Dual Spectrum Averaging, make the following changes as listed and some
76
these windows will fall into their default settings. The following changes made at
this point will be:
Ensure MEASUREMENT set to DUAL SPECTRUM AVERAGING

.
AVERAGING set to LIN, 400 and OVERLAP: MAX. (user select for
number of
averages)
CH. A and CH. B must be 3 Hz. DIR.
. CENTER FREQ. choose BASEBAND
Move to the very bottom of the screen (under the generator function) and
press the number 3 and enter key. The SPECIAL PARAMETERS will
activate and enter 2032 for the DEC VALUE at the very end of the line. This
acts a key to turn on the special parameters.
Set the FREQ. SPAN, to the desired frequency range and the other settings
will
correspondingly default.
. DELAY should be 0.0 ms
. TRIGGER set to FREE RUN

. WEIGHTING set to HANNING (hint: Cursor to Weighting instead of
window type and toggle untill just the one Hanning window comes up. This
will be applied to both channels.)
3) This part will introduce the activating of the using the

random noise
generator function and the Bruel and Kjaer Vibration Exciter System. These steps
are:
Follow the instructions given above for connecting the cable from the
analyzer to the power supply and ensure the power box on the wall is turned
on.
Go to the bottom of the screen where it says GENERATOR and cursor to the
right. Toggle the menu till RANDOM NOISE appears. The button to turn
this on from the analyzer is off to the right of the screen and near the button
"on"
that is pressed for printing plots out. The generator should say in the
analyzers'
screen.
Turn the Gain knob to zero on the power supply. Then, turn the SET-UP
CONTROL knob to STAND-BY and listen for the shakertable to turn on. If
it does not turn on, switch back to LEVEL, and again go to STAND-BY.
Once the shakertable is running, turn this switch to OPERATE. (Nothing
should be moving on the shaker table)
While slowly turning up the Gain to 1 in my case, or for whatever setting is
desired, monitor the overload light from the charge amplifier for channel A.
Set the Volts/Unit Output until this light goes out and the 20 db form
overload light stays lit.
77
Set this value from the Volts/Unit Output in the measurement setup for
channelA. ( ex. V/Unit determined from the
.001
was charge amplifier so key
in lOOmV/N in channel A.)
Note: The same must occur when using the charge amplifier with the
accelerometer sensor.
Set the scale to 80db

Press the AUTORANGE button while the shaker is still on. This will be the
expected amount of force to be excited during the test. Turn the analyzer's
generator off when the single light turns off and the continue light goes on.
This will preset the voltages on channel A. and channel B. and create an
overload signal that will flash on the screen as well as on the charge amplifier
when you
accidentally excite the system more than the expected level of
force.
Turn the power amplifier SET-UP CONTROL knob to STAND-BY or to
LEVEL depending upon when the test is going to be started.
4) This involve producing the frequency response of the interested

part will
frequency range, zoom in on the resonance frequency of choice and determine

the damping for this resonance.
Toggle the FORMAT button till a dual screen appears.
Select at the top of the screen FREQ. RESP. HI and ensure to the right of
this that it is set to MAG. Go to the bottom trace and at its top toggle till the
COHERENCE is displayed.
Turn the random generator on and the power amplifier to level while at the
selected Gain level. Press the SINGLE button to start the number of selected
averages.
Once the averaging has stopped, power off the generator and set the power
amplifier to level. If the coherence screen looks good and no overloads have
occurred, then proceed to cursor in the FREQ. RESP. HI trace and
determine the resonance frequencies. Remember to monitor the accelerometer
sensor and the green light and oscilloscope for the laser sensor for overloads
conditioning problems.
or signal
Go to CENTER FREQ. choose ZOOM

Select and toggle for the frequency in FREQ. SPAN, to the desired
frequency range and the other- settings will correspondingly default. See the
experimental plots in Appendix C. as reference.
Proceed to turn on the generator and the power amplifier to the operate mode.
Press the SINGLE button to take the number of desired averages.
Proceed to the cursor in the FREQ RESP. HI. to determine the resonant
frequency.
Change the MAIN to REF.
78
A number 1 be
placed in the DEC VALUE of SPECIAL
should
PARAMETERS #40 for the Hanning weighted window. See Appendix J in

the special parameters for more information on #40.
Go to the upper trace and cursor to the FREQ. RESP. HI and enter 28 to
bring up the IMPULSE RESP. function. Ensure to the right that it says
MAG. for the Hilbert Transform to activate.
Inspect for the straightest portion of the slope from the IMPULSE RESP.
MAG. and cursor to the top of the smoothest portion of the slope. Press the
SET SPECIAL CURSORS to clear the AY in the REF. area. Note: never
start on
any knee of the Impulse Response Magnitude.
portion of the
Move the cursor to the right on the screen while
watching the AY change in
db. When the cursor reaches a -8.7 db and is still
containing a decent straight
slope within this bound, the time constant T is now determined by reading
the AX from above the AY.
The material
damping can now be computed from equation 2.26 found in the
chapter 2, Damping. Figure 2.11 may be used as well which is found in
chapter 2.
Repeat the process for the next resonance if desired.
5.3 Other Excitation Techniques
The other excitation techniques that were briefly examined was the pseudo
random excitation and the sinewave excitation. Both these methods were applied
using the B&K 2032 Analyzer and they are easily selected by toggling the window
to the right of the generator window. This section is to become familiar with the
technical background of these methods for purposes of discussions that will be
presented in the next chapter [6, 1 1, 12].
Pseudo-random excitation is a random signal of length T which repeats
every period of time T. The spectrum becomes discrete where the energy only
exist at the frequencies being sampled in the analysis. The period length is
matched with the record length of the analyzer so that the frequency components
of the pseudo-random signals will coincide with the computed frequency lines in
the analyzer. Rectangular weighting should be used with pseudo-random
excitation because of its leakage in the

periodic nature; as a result there will be no
spectral estimates as long as there is This is the

no noise at the input or output.
main advantage of using pseudo-random excitation because the signal is designed
so that each frequency component has the same amplitude in the frequency range
of interest, but the phase angle between the different components will be random.
This is good for low frequency work, and only a low number of averages are
79
needed. Pseudo-random excitation and rectangular weighting is only valid for a
lineai-
approximation of a system; it is not valid for a non-linear system because
the non-linear effects will not be periodically averaged out. On the page 81, Figure
5.6 is a plot of time channel A. that was measured from the force transducer from
the impact hammer [6, 1 2] .
Sinewave testing is a traditional method for creating an excitation source
and the sinewave is applied usually by stepping or sweeping through the frequency
range of interest. This type of
testing controlls the input signalvery well, the crest
factor is low, and the signal to noise ratio is good. It is also best for studying non
linear systems. The only disadvantage with this type of testing is that it is the
slowest to conduct [1 1,26].
80
w Q
z
0 Z3
E \N UJ
>>
O a e _i
oa <
.a >
u
UJ
N N a
II
XX
(OD
mm u
(M(M Z
0 a 3
O r-h-
U. to
X
 UJ in aaoir
< > N zu.
a N N
i o . N in .
3 E U.I tnmaa
a: a <QJK ?
i- (T)< z*
u ao W_J <
uz a + ta:
0.3 .(M u
tnam z o
? N < a
III-
_1UJ<
3
<UJ .z QQU > ujo:
DOUJ axnuj
QU.U-J -NIX CJCTULH
UJ
I
<
r-
a IT
z -LU . .<
a UJ 13 ZKU CO.
2 . Z <IlZ a
d * iiia: H 0. " hj
OiUJ -u tfllZH <<
a. DUX UJI . .Kin
3 wu<o: at-u <muiu

I- <w_IUJ uiz . .zki
UJ ujq:uj> KlUUJ IIUJQ.

Zl-Q< u.us UUUU)
DISCUSSION OF RESULTS
6.1 Experimental Determination of the Loss Factor
The loss factor was determined for a 10 inch aluminum cantilever beam
using four different types of sensors and two types of excitation techniques. The
four types of sensors are the laser vibrometer, piezoceramic, piezofilm, and
piezoelectric accelerometer, and the two types of excitation methods are the
impact method and the application of random noise through the use of a
shakertable. The determination of the loss factor for the first and second modes of
the beam to be investigated. The analytical solution of the natural

were chosen
frequencies for these modes were determined by two representative models of the
beam. The first analytical model was a free cantilever beam and the second
analytical model was a cantilever beam with an end mass. These models are
assumed to represent the general physical characteristics of the beam. The free
cantilever beam model of the predicted natural frequencies was compared with the
beam used in the application of the laser vibrometer, piezofilm, and the
piezoceramic. In this application it was assumed that these sensors would not
effect the dynamic characteristics of the beam. The accelerometer was simulated
as an end mass in the second model for predicting the natural frequencies of the
cantilever beam. This model can be compared with the model of the free cantilever
beam to determine whether the mass of the accelerometer was affecting the
dynamic characteristics of the beam. I assumed that any partial amount of weight
from the accelerometer cable was very small, therefore negligible. To test this
assumption I incremented the end mass in the model to determine what amount
was needed to match the measured natural frequency. I found this only to be
several grams more than the 2.7 grams of the accelerometer. Learning this, I felt it
was not worth trying to measure the partial amount of cable mass since the
measured natural frequencies of the free cantilever beam did not even match the
free cantilever beam model. Since the predicted resonant frequencies of the two
models were different, I should expect the damping to be different from the other
sensors due to their weight being much smaller than that of the accelerometer and
also because the damping is dependent upon the resonant frequency.

The analytical and experimental natural frequencies of the ten inch
aluminum cantilever beam are listed in the summary Tables 6. 1 and 6.2 on pages
81 and 82 respectively. The experimental and frequencies of

the analytical natural
the first and second mode are well within a difference. The measured
ten percent
natural frequencies by the piezoceramic sensor using the impact excitation was the
closest to the natural frequencies predicted from the model. The natural frequency
measured from the laser vibrometer and piezofilm using the impact excitation was
The measured natural
essentially the same for both the first
and second modes.
frequencies from the piezoceramic sensor varied the most between the two
82
+ 4 ?
u O CN cn
in vO
r-
CN
cn r-
CN
-? vO
3
a*
SSJ
T ^ o=
OS * m a\
r-~ <N
03 C o VO
3N r-~ r-
She- c-i
V-*
\J O
o O
^ -
O o
<: <:
k O d
a T3
5 a a
* 4) O
Ov O Q. CX
^^ Vi oc ^^
II a, o,
>/-) 00 00 00 < <
9 ^-^ ^?
X! ^ 00 Ofi
p B S
5 p a. a
S S
.O
-3-3
Ss o o
B B
") !v *
r- -a t3
CB
s|i oo
00
VO
r~
vi
r-
m
a
es
va
to
a
cs
co
CO
eil
ro m m CS CS
JD 1 k
6 6
T3 13
I-"-]
00
Ml a a
O CJ CJ
o a
as 3 1 00 CN r^ B cs
o 2 JS -E
r-
00
^* 3--
p p o p S 6
^5 w ~^ ~^ ~^
cs
CJ
cs
<D
JO >
a Vm
 >
1 O
00
o
as
VI
rt
CJ
a
a
CJ
-a
a
I
o G CS CS
SI o o
o 01 O CJ
o o
s2 d d
1 cs cs
s Cw
O
Cm
O
3
s
"5 "3
1
11
-.
T3 T3
II O O
o j 3
X! rs s s
a ~
>n <N <
So
ill
ii CS CS
CN VO
-) ^ II B CJ CJ
ill
^ Ol r- r-
00
QJ p p
g m m m
-t
r*-,
"E
hJ Z
Is cs
s CU s < a a
k CJ
< <
P9 * ?
'E
os
a w
u H
ta
o
1
&3
J
PS OS
w
fc u J
w
o o lid
Sl Sl
w
3
J S a. <
5 1 * + ?
u
C
CT\
S 5* r Fi
VO
Os
VO
ON
vO
ro
a*
VI V Vi ro
CN CN CN CN
u
ta 5^5
< O >
T 03 ^
"5
*
ON ro
CN -t
V,
si CN
o\
OV C\
OS
O
< <
^s
fl TJ TJ
a e
o CJ CJ
UJ % VO
ro
o
VO
^^ V,
II
ex
ex
ex
ex
o Os V r-
< <
z
9 ro 0
^ O
5 00
e
Ofl
a
0
s 3 ex ex
S 6
J- CS CS
TJ TJ
Sn 0 0
B B
s- Vl 0 O O a -a
r-
o B B
00 O cs cs
w ro VO ro d CO
CO
CO
CO
ro ro ro CN
VO ss CS CS
JD ta 1 S3Bfc CN CN CN CN
S B
TJ TJ
5 B B
CJ CJ
mi ^
o s; 0 a
H CN -t B cs
ta P Co 0
35 ro
rt
00
V
rt
VO
CN
VO j= .a
r< VO OS 0
sV
fc! 0 0*^ -
3 ? ^
O
S 6
fc ~^
cs cs
CJ CJ
,0 .>
I- >-
<D <D
fl VO ro CN > >
Os CN JU
*
^P r^
CN 00 00
%*
1 1^
CN ro rt
3
a a
V) CS
C8 S ^ O CJ
CS
0 cj
S ^ ^59
*
O
d d
s
u
cs
CW
cs
Cm
0 0
s 3
e
"3 "S
s- *h
>N "i II
TJ
O
TJ
O
3
V CN J3 fS B 6
^3
CN sO
V V
S) TH CS cs
-1. II a
V CJ
111
ro
<N
ro
CN
-+
CN
CN
CN
1
''-
z
a
Zl
^1 fc
OJ
CS
E
CJ
B
< <
B
OS * t ?
s
"C ei
Zl w
a u H
H-1
OS
ta 0
O
1
co E u
os
os 0 0
J/J
Sl Sl u
w w
lJ & <<
methods of excitation. The measured natural frequencies using the accelerometer
and piezofilm did not vary much with the method of excitation.
The analytical models for determining the natural frequencies are in

Appendix A and the plots of the experimental frequencies are in Appendix B for
the Impact Excitation Appendix C for the Random Noise Excitation. The
and
analytical solutions for these natural frequencies were solved by the classical
approach using the method of continuous systems.
The results for the loss factor were very interesting in the comparison
between the applied sensors and the excitation methods. As shown in Table 6.2 of
the second mode, the difference in the loss factor immediately stands out between
the two excitation techniques. It was not apparent to me at first what was
causing
the extra amount of
damping in the second natural frequency. The outcome of the
investigation as to why there was such a difference proved to be caused by the
shaker table. By inspecting the frequency response of the second mode, the peak
of the resonance was blunt in comparison with the response from the impact test.
Several ideas were attempted in order to determine the cause for this difference
such as
examining the adhesives used in the bonding of the sensors and varying
the gain control on the power amplifier of the shakertable. Shifting the frequency
of the second mode to other resonant modes was attained by changing the length
of the beam and applying alternative methods of excitation such as the random
impact excitation using the modal hammer and pseudo random excitation on the
shaker table.
The determination of the loss factor began with the use of the random
excitation equipment. After determining the loss factor for the modes using the
piezoceramic, the response from the second mode was observed to have a blunt
shape at the resonance. In examining the response of the second mode from the
piezofilm, the same effect had occurred as observed using the piezoceramic sensor.
An investigation of the bonding material was thought to be affecting the response

with artificial damping. It was also observed that the level of gain would effect the
sharpness of the resonance's peak. Because the damping was found to decrease
when increasing the level of gain for either the piezofilm or piezoceramic, the
impact test using these two sensors was conducted. When the loss factor was
obtained using the piezoceramic and piezofilm sensors for each of the two modes
in the impact test, the comparison of these results between the two methods of
excitation were quite different. The peaks from the frequency response between
the excitations techniques was obviously different. The peak from the impact
excitation looked sharper where the results from the shaker table were basically
blunt. The differences in these peaks resembled Figure 2.4 back in Chapter 2. The
gain on the shaker table was increased in the attempt to match the results of the
loss factor with those from the impact test and overcome whatever was causing the
high This idea failed cracked the piezoceramic sensor.
damping problem. and only
The accelerometer was tested using the random noise on the shaker table and I
85
found the response of the second mode to be effected as well. In light of the
frequency response from the accelerometer via the shaker table, I concluded that it
could not be the adhesives used to bond the piezoceramic nor the piezofilm
because the bee's wax is a common material used in vibration testing. Secondly,
the results of theimpact test were assumed to be correct because the loss factor for
the second mode was less than 0.001 as described in the A.S.T.M. Standard
Method for Measuring Vibration-Damping Properties of Materials. The bonding

theory causing the artificial damping must be false.
problem
By increasing the gain on the power amplifier, the material damping would
decrease but the peak on the frequency response would appear to be less blunt.
The piezoceramic would crack quite
easily at veiy high amounts (Gain >
of gain
9)
or for short periods of time at lower levels of gain (Gain =
5). During the
experimental testing of the beam using these sensors, the idea of using a certain
amount of the gain as a variable in controlling the test setup failed. Upon
reassuring the rest of the setup was in order, the level of gain was selected for
testing the sensors. The loss factor data for the second mode was chosen with a
gain set to 9 on the power amplifier for the random excitation method. The reason
for choosing nine is because it was the best approximation of the loss factor as
compared with the impact case. Secondly, the frequency response was very poor at
a gain level of one, and trying to determine the time constant from the Impulse
Response Magnitude became quitedifficult. This setting was applied to all the
sensors for the second mode of vibration except for the laser vibrometer. The laser
vibrometer would not work for any level of gain higher than about 0.5 on the
power amplifier. The laser vibrometer presented the same problems for the first
mode of vibration and the gain was set to 0.5 as well. The gain was chosen at 1 for
the rest of the sensors in the first mode.
The next thought was the possibility that the shaker table might be causing
the problem at the second resonance of the aluminum cantilever beam. The second
resonant using the shakertable definitely described a blunt frequency response

which could be only causedby the influence of damping. The bonding material as
the cause of external damping was eliminated and the only other possibility left
was the shakertable system. The proposed method in proving the shakertable was
the problem causing the artificial damping in the vicinity of 236 Hz. was to shift
the second frequency by

natural changing the length for the cantilever. I decided to
force the frequency of 236 Hz. to be a third mode at another beam
second natural
length and then force this frequency again to be a forth mode at another beam
length. These new lengths of beam were estimated by assuming the following:
Cantilever Beam with an End Load
3FI &M'=J
Figure 6. 1 K = ^~
(Eq.6.1) (Eq.6.2)
86
(Eq. 6.3)
l*{wj*
&l'*(w'nfi (Eq.6.4,
( . V?
v_ (Eq. 6.5)
I \
"
n J
The stiffness, K, of the cantilever beam using an end load inversely

was assumed
proportional to the length cubed and the frequency equation was
natural
approximated as the square roof of the stiffness for the lightly damped structure.
The new natural frequency, w'n, would equal the natural

frequency of the second
mode and now the new length, /', can be estimated. The estimated length for
shifting the 236 Hz to a third and forth resonance was 15 and 30 inches. The true
lengths were found to be 17.5 and 24.625 inches and this is listed in Table 10 on
the next page. The modes were then determined using the random noise excitation
and at a level of 3 on the gain. The Hanning weighting window and the piezofilm
sensor were used. The plots at the three different lengths show how the mode
damping at the frequency of 236 Hz was always effected with the greatest amount
of damping. The general trend that can be seen in the three beams is how the
damping decreases as the resonant mode increases but
only until that mode of 236
Hz shows up. The only small exception is with the longest beam at the fifth mode
that did not succeed in being less than the third mode. The fifth mode may be
seeing some artificial damping from the shaker table as well. It is intersesting that
the shaker table turned out to be the cause of this problem. This is something that
I thought would not be a problem in the testing environment. The plots of the
frequency response magnitude and impulse response magnitude for all the three
beams are found in Appendix E. The shaker table also sits on three isolation pads
and has a wooden plate fastened a the bottom. The three isolation pads were
removed and response of the cantilever beam was still affected by the artificial
damping. The wooden platform was part of the assembly for the shakertable. In the
future, the shakertable should not be used for the determination of mode dampings.
In the search for understanding where the damping was coming artificial
from, I have learned that the application of different weighting windows and
87
Table 6.3
Determination of Artificial Damping

given by the Shaker Table
in the vicinity of 236Hz.
Material =
Aluminum
Gain =
3
Weighting =
Hanning
Excitation =
Random Noise
Sensor =
Piezofilm
Beam Length =
1 0 inches
Mode 1 Mode 2 Mode 3

Natural Frequency 37.375 237 659.375
Loss Factor 0.005 0.0214 0.00167
H Mode 1
H Mode 2
0.025-1 B Mode 3
0.02-
0.015
0.01-
0.005-
0-
Loss Factor
Mode 1
Mode 3
88
Beam Length =17.5 inches
Mode 1 Mode 2 Mode 3 Mode 4

Natural 13.437 84.355 237.750 465.000
Frequency
Loss Factor 0.0051 0.001916 0.00858 0.001718
I Mode 1
) Mode 2
I Mode 3
0.01 1 Mode 4
0.005-
Loss Factor
Model Mode2
Mode3 Mode 4
Beam Length =
24.625 inches
Mode 1 Mode 2 Mode 3 Mode 4 Mode 5
Natural 6.812 42.937 120.125 236.75 390.250
Frequency
0.00454 0.00304 0.00204 0.00718 0.0047
Loss Factor
1 Mode 1
! Mode 2
9 Mode 3
I Mode 4
3 Mode 5
Loss Factor
Mode Mode
Mode Mode Mode
1 2 3
5
89
Table 6.4
Rectangular Window & Psuedo Random Excitation
vs.
Hanning Window & Random Noise Excitation
Material =
Aluminum
Gain =
3
Sensor =
Piezofilm
Beam Length =
1 0 inches
? Effected by Artificial Damping
Hanning Window and Random Noise
Mode 1 Mode 2 ? Mode 3

Natural Frequency 37.375 237.000 659.375
Loss Factor 0.005 0.0214 0.00167
Coherence .36 1 .98
Rectangular Window & Psuedo Random Excitation
Mode 1 Mode 2 ? Mode 3
Natural Frequency 37.500 236.750 659.250
Loss Factor 0.00318 0.0202 0.00161
Coherence 1 1 1
90
excitation may affect the true value of the loss factor. It was suggested that a
may be
rectangular window and better
pseudo random noise a choice in
determining the damping rather that a Hanning window using random noise. I
investigated this option and found that for the first mode from the two setups gave
a 36.4% difference in the calculated loss factor. These results are shown in the
summary table 6.4 on page 88. The second mode and third mode gave
relatively
similar results between the two setups. The other observation to note was that the
coherence was weak in the first mode using the random noise and Hanning
window but the coherence was unity for the pseudo random noise and rectangular
window. The coherence for the third mode in the random excitation appeared to be
less than one. I also built a fixture that could grasp the impact hammer. The
purpose of this fixture was to have total control of the area where the impact from
the hammer going to be repeated. The arm of the hammer was free only to
was
pivot in the fixture and a string was attached to the neck of the hammer where it
was bobbed up and down by my wrist. The random impact excitation was created
using this setup with the Hanning window applied. Note: the cantilever beam was
lying horizontally in another fixture. The impact hammer was struck downward in
a random bobbing manner using the string of the hammer and the piezofilm sensor
was applied. The results of this test can be seen in Table 6.5 on the next page. The
loss factor in the first mode and the second mode was similar to the measurement
of the laser vibrometer using the impact excitation. Lastly, I did not have much
time to pursue the sine testing setup using a feed back loop but this is for those
who wish to know a short cut in bypassing the feed back loop. Simply change the
excitation signal source in the generator window in the analyzer and set to sine.
This setup can be found in Appendix G. Next, cursor to the right where one can
manually change the frequency of the sine wave. Activate the shaker table system
and notice by turning the knob you can scan the frequency. Before collecting any
data and while the system is running, tape the knob off so that the rate of the
frequency is at suitable level. The plots were made of the first mode only and
located in Appendix G. The results are also shown in Table 6.5 and the
piezoceramic sensor was chosen for no special reason. This produced results of
loss factor that were much greater that the random noise excitation.
The loss factors for the first and second mode using the impact testing
method turned out to be very interesting. The piezoceramic and accelerometer
gave very close results. The results of the laser and the piezofilm were similiar yet
different from those of the piezoelectric and accelerometer. I was delighted that
the piezofilm sensor compared quite well with the laser in the impact testing
because the piezofilm is really an excellent sensor to work with. The loss factor in
the first in the random excitation turned out to be
mode using the accelerometer
the highest value even compared with the data from impact test. The other
observation concerning the accelerometer used in the random excitation is how the
loss factor compared quite well with the piezoceramic and the piezofilm in the
91
Table 6.5
Random Impact Excitation using a Modal Hammer
Material =
Aluminum
Sensor =
Piezofilm
Beam Length =
1 0 inches
? Different Fixture
Hanning Window and Random Impact Noise
Mode 1 ? Mode 2 ?
Natural Frequency 36.625 229.750

Loss Factor 0.00183 0.000529
Coherence .8 .68
Variable Sine Excitation using the Shakertable System
Material =
Aluminum
Sensor =
Piezofilm
Gain =
3
Beam Length =
10 inches
Hanning Window & Sine Excitation Excitation
Mode 1
Natural Frequency 37.500
Loss Factor 0.00318
Coherence 1
92
second mode. Another interesting
is the loss factor from the laser
result
vibrometer. Despite all the technical difficulties I had
operating the laser
vibrometer, the loss factor determined from laser vibrometer was the only sensor
that gave similar results for the two excitation methods.
Overall, the values of the loss factor for the first mode were above the
suggested value of 0.001 given in the A.S.T.M. Standard Method for Measuring
Vibration-Damping Properties of Materials. The second mode in the impact testing
gave results of the loss factor that agreed with the standard, being less than 0.001.
6.2 Noise Sources in the Experimental Testing
The most important part of collecting and taking experimental data is to

have an
understanding of all the parameters that could effect the results of your
experiment. I found the measurement of material damping or rather the loss factor
to be a complex and involved process due to all the necessary components used in
determining it. It is necessary to pay careful attention to all the details of the
experiment from the materials tested to the actual data reduction and analysis.
The assumptions that I thought would not affect the loss factor could just be
part of the problem as well as not
knowing what wasreally involved in trying to
access the optimum setup for detennining the true mode damping for aluminum. In
review of the from the experiments, I decided it was best to list all the items
results
that could have affected the loss factor and in this way, it helped me to understand
why the loss factor varied with the applied sensors and the excitation methods.
The list of these items were broken down into six different categories and they are
as followed.
Governing Parameters of the Experiment

Sensors
Excitation Equipment
Analyzer Setup
Environmental Conditions
Testing Materials
Testing Fixtures
The possibility of the noise from the sensors was of the utmost concern
because of the problems I had in the beginning with the piezofilm and the laser
vibrometer. The laser found to be very sensitive to the vibration. If

vibrometer was
the amount of force tapped in exciting the beam was deflected too much, the
speckles in the speckle pattern would shift just enough to cause noise. This was
oscilloscope become distorted. The
observed watching the squarewave from the
by
piezofilm worked well after the fixture was grounded, so the 60 Hz signal could be
removed. It also helped to eliminate all the other circuitry thought to be needed to
93
condition the signal. Since the potential of noise existed from laser vibrometer and
possibly the piezofilm, I thought it was best to use the Frequency Response HI as
it was recommended to help minimize noise at the output. The accelerometer was
checked to ensure that it Hand-
was calibrated
correctly using Bruel and Kjaer's
Held Exciter in conjunction with the B&K 2032 analyzer. The only other
noticeable problem through the course of doing the experiments was the charge
amplifiers. One of the two, or
possibly both the charge amplifiers were
acting
irregularly from time to time. It would sometimes keep overloading, but this would
occuronly intermittently. The other problem I had that was more frequent was the
impact hammer. I am almost certain that the microdot connector at the bottom of
the handle was the intermittent problem. The overload signal on the analyzer
would frequently trigger and I knew then before collecting the frequency response
data that contamination was occurring at the input. The stiffness of the
piezoceramic may suggest why the natural frequency was much higher than the
other sensors and therefore effected the loss factor. The first mode of the beam
may imply high amplitudes indicating near non-linear effects. The A.S.T.M.
standard for measuring the damping properties has suggested skipping the first
mode because of non-linear effects. Maybe this is why the loss factor for the beam
was higher than the suggested value of 0.001, and also the fact that the test was not
conducted in an environmental chamber nor was the aluminum beam at the
suggested dimensions for the testing. The other possibility is the selected level of
gain was not exactly the same.
The change in the experimental setup could be enough to account for

differences in the determination of the mode damping. The stiffness and the mass
from the type of fixture used to clamp the beam could have an effect. In Appendix
K, there are four plots taken from a paper; "A State-Of-The-Art Assessment of
Measurement Techniques Results for the Mid-Range Structures (30-

Mobility -
3000Hz)"
by D.J. Ewins and J Griffin that display a superposition of frequency

responses [4]. The goal of this paper was the assessment of the accuracy and
consistency of mobility measurements; it occurred in Europe during the period of

1979 to 1980. Four test specimens were passed from facility to facility to be
measured, and this participation takes into account many different techniques and
various types of instrumentation. These six superimposed plots display the
individual results of the 15 to 20 participants from the United Kingdom, and
France who tested these structures. These results are obviouly very noisy. The type
of excitation that was applied to these structures were sinusoidal excitation,
random excitation, and transient excitation. This paper included data concerning
the average time spent producing each mobility plot, the testing time per plot, and
the lost time for these besides collecting the data for the loss factor and
structures
the resonant frequency. It is very interesting to note how time consuming

experimental testing can be and this is an aspect I did not fully realize. The
average time spent each mobility plot, the testing time per plot, and the
producing
94
lost time for the simplest structure was 150 min ,
27 min and 100 min respectively.
Basically this is
saying that about 79 % of the time required to conduct a test is
lost. In my opinion, I believe
my required testing time was much longer due to the
setup time for the laser vibrometer and also for the setup of the impact
testing due
to the intermittent overloading. The results from the first mode of the natural
frequency and the loss factor for the simplest structure used in the testing from all
the participants and
including my results are shown below in Table 6.6. The
purpose of this table is to illustrate the statistics of how the data was affected from
the techniques and the equipment used to obtain the loss factor and the resonant
frequency. The column of data from my results without the accelerometer shows a
definite improvement in the spread of the loss factor between Ewins and Griffins
and the results for all the sensors.
Table 6.6
Comparison of Testing Results for the First Mode
D.J. Ewins and Berg R. R.

Berg
J. Griffins Summary Results Summary Results
Summary Results of the Cantilever of the Cantilever
of the Simple Beam for all Beam for all
Structure IIB Sensors Sensors but w/o
from All the Accelerometer

Minimum 41.67 33.75 37.19
Frequency Mean 43.60 36.72 37.67
(Hz.) Maximum 45.70 38.562 38.562

Spread 4.03 (9.25%) 4.81 (13.11%) 3.64(9.67%)
Minimum 0.00055 0.0017 0.0017
Loss Mean 0.0074 0.0038 0.0027
Factor Maximum 0.0220 0.012 0.0038
Spread 0.0215(290%) 0.0103 (268%) 0.00210(78%)
The last comment I wish to discuss is the analyzer setup. I found that there
are several ways to set the different weighting windows and excitations using the
impulse response function as the means of calculating the loss factor. The three
typical setups are the impact excitation using the transient window, random
excitation via a shakertable using the Hanning window, and pseudo random
excitation using the rectangular window. These setups require an

understanding of
In my case, I
the optimum
setting for the time record or the frequency resolution.
used the default settings that were provided by the analyzer when selecting the
95
frequency span for the test. The other option provided in the setup is to either use a
baseband or a zoom analysis setting.
Either one can be used to determine the loss
factor provided the other settings used are correct. I have experimented with
changing the windows and the
frequency resolution as well. I have found that
trying to accommodate both worlds of noise is impossible to satisfy. I decided the
optimum setting was to be
going a
frequency span of 50 Hz. surrounding the single
resonant peak and by default from the analyzer, the frequency resolution of
62.5mHz. or a time record of 16s. I tried to increase the frequency resolution in the
zoom analysis using the impact hammer
and found very little difference in the
calculation of loss factor. The time record is calculated as the inverse of the
the
frequency resolution. The problem I ran into for the impact testing was the choice
of
weighting window for the output response. This is also another reason why the
Frequency Response Mag. HI was used rather than Frequency Response Mag. H2.
The Frequency Response Mag. HI would minimize the leakage from the output
signal. The choice was either to use an exponential
weighting which is the most
common choice, but an alternative choice would be a transient window. The
response signal may decay well before the end of the time record when
using
zoom analysis, and an exponential not be needed. The transient
may window can
therefore any noise beyond the end of the response signal. I have used the
cut out
transient window for the output signal on the

following sensors as given in the
table below:
Table 6.7
Transient Window applied for Impact Testing
Sensor Modes
Accelerometer 1,2
Laser 2
Piezoceramic 1
Piezofilm none
The reason for doing this was the effect it had on the value of the loss factor that
was calculated in the impact testing. The other modes not listed in the table above
used the exponential window. Table 6.8 shows two cases comparing the loss factor
using the exponential window and the transient window. The accelerometer clearly
shows a difference in the value of the loss factor as well a change in frequency for
the two window types. The laser vibrometer displays how close the lost factor and
the coherence are within the two windows. The laser vibrometer was
intentionally
displayed to show that either window can work out as long as the window length is
altered properly. In the laser vibrometer section, the trace of the decay response
shows the transient windowing. These plots can be found in Appendix B under
96
 <n
1- r- 00

ON ri
c =
s S o> r~- ^^
J
ri
^"
s
t-l
,_ >.
- o N in
% N
ft ^ o X in
n
m
ri
in
Cl
g*^ 2 S P-H ri
w N
s
Cs t^
3*H C4 rt
</5 O SO m
o
o o
o o
o
^H
<u
*-*
<u
O E
o CO
CO
<u
OO "u J
o
o
<
1
the respective sensor. The piezoceramic turned out be very noisy in the impact
to
testing and the only reason I can explain this is the possibility of microcrack in the
ceramic. The piezoceramic may have fatigued from being used on the shakertable
because the plots shown in the random testing do not appear to be contaminated
very much as impact testing did. The plots for the random excitation data are
the
shown in Appendix C. The rest of the analysis for the impact
testing was
comprised of both windows
making the analysis to some extent more difficult. I
found it very interesting to study the effect of the windows because no one
analysis is correct. In far

my opinion, the random excitation was more easier to
setup than the impact excitation. In another paper from the Journal of Sound and
Measurements"
Vibration called "Digital Filter vs. FFT Techniques for Damping ,
(March 1990), a comparison was shown between the weighting windows and the
type of excitation
using the FFT and Digital Filter [5]. The paper was concerned
with the calculation of damping for lightly damped structures of single mode
resonances using these different techniques. The end result of this paper discusses
the relative differences in the standard deviation of loss factor. The first mode of
this freely hung plate were 15% and 5-7% for the other resonances. They
concluded that the deviation was random rather than systematic. This also may be
the reason why the first mode of vibration is skipped in the A.S.T.M. method. Also
the results from the first mode shown in Table 6.4 showed a great difference in the
loss factor using the pseudo random and random noise excitation with the different
windows. For extremely small changes in the test setup and in the environment, it
is shown how the values of the loss factor can be influenced.
98
7. Comments and Recommendations
The application of sensors in the experimental determination of the mode

dampings was
very challenging. Not only haved I learned how to determine the
mode damping, but I was also exposed to a new and intriguing subject area
concerning modal analysis. It was interesting to have explored the different

weighting windows, excitation techniques and sensors. The sensitivity of the
viscoelastic materials with respect to temperature was not expected but it would
have been interesting to see the results of the mode dampings providing that the
temperature was controlled. The interference with the shaker table concerning the
damping with the second resonant of this cantilever beam was definitely an
unexpected problem. This turned out to be a learning experience. One lesson
learned most of all is the appreciation of time it takes to perform modal testing.
Learning how to use
many the especially sensors was the laser vibrometer
challenging as well. The piezofilms are the best to work with simply because they
involve little setup time and are just as accurate as the laser vibrometer at least in
this application. The impact testing using the modal hammer was time consuming
and the level of complexity was greater than
using the shaker table system. The
values of the lost factor were the best I could achieve based upon experience with
the testing equipment and knowledge of modal testing. I am confident that the
results of the loss factor from the second resonances using the impact testing for
this cantilever beam are correct [4].
I would recommend learning more about the laser vibrometer finding out
why it is so sensitive to small amounts of vibration. Another type of excitation
technique that would have been nice to investigate in detemining the mode
dampings of the cantilever beam is the impulse response option found in the
generator window on the analyzer. Another interesting comparison would be to
determine the mode dampings in the baseband mode rather than using zoom
analysis. A lighter hammer for the user would prove invaluable for the impact
testing.
I was thankful to have met some great people who were willing to lend a
helping hand and to offer their experience in the field of experimental vibrations. I
was happy to have made this choice as my thesis topic because I was able to gain
invaluable experience which can be applied to other areas in the modal testing
field.
99
REFERENCES
1. D.J. Ewins, "Why and wherefores of modal

testing,"
SEE Jl, 18(3) (1979)
2. S. H. Crandall, "The theory,"
role of
damping in vibration J.Sound Vib.,
13(4) (1970)
3 D. I.G. Jones
.
materials,"
"Temperature-frequency of dynamic properties of damping
J. Sound Vib., 33(4), 451-470 (1974).
4. D. J.Ewins and J. Griffin "A State-of-the-Art Assessment of Mobility

Structures"
Measurement Techniques-Results for the Mid-Range J. Sound
and Vibration 78(2), 197-222 1981
5. S. Gade and H. Herlufsen "Digital Filter vs FFT Techniques for Damping

Measurements"
J. Sound and Vibration pp. 24-32, March 1990
6. S. Gade and H. Herlufsen "A Hand-Held Exciter for Field Mobility

Measurements"
J. Sound and Vibration pp. 16-21, March 1993
7. W. G. Halvorsen and D.L. Brown "Impulse Technique for Structural

Testing"
Frequency Response Sound and Vibration pp. 8-21, Nov. 1977
Measurements"
8. J. T. Broch "Mechanical Vibration and Shock Bruel &
Kjaer, 1984
materials"
9. "Standard for measuring vibration damping properties of

American Society for Testing and Materials, E756-80, 1980
10. H. Ghoneim, S. Griffin, and B. Yost "Analytical and Experimental

Beams"
Investigation of the Mode Dampings in Laminated Composite R.I.T
Dept. of Mechanical Engineering, Dr. Ghoneim
11. R. C. Stroud "Excitation, Measurement, and Analysis Methods for Modal

Testing"
Sound and Vibration pp. 12-27, Aug. 1987
12. "Digital Signal Analysis using Digital Filters and FFT Techniques-Selected
Review"
Reprints from Technical Bruel & Kjaer, Jan. 1985
transform"
13. N. Thrane and J.S. Bendat, et. al: "Practical use of the "Hilbert
B&K Appl. Notes
100
14. Et. al: "An introduction to modal
testing"
B&K Appl. Notes
Windows"
15. Et. al: "Time B&K Appl. Notes
16. A. D.Nashif, D.I.G Jones, and J. P. Henderson "Vibration

Damping"
John
Wiley & Sons, Inc. 1985
17. S. S. Rao "Mechanical Vibrations 2nd Ed."

Addison-
Wesley Publishing
Co., 1990
Practice"
18. D.J. Ewins "Modal Testing: Theory and Bruel & Kjaer, 1986
19. "Instruction Manual 2032 Vol. 1 Familiarization for Dual Channel Signal
2032"
Analyzer Type Bruel & Kjaer, Revision June, 1987
20. "Instruction Manual 2032 Vol.2 Operation for Dual Channel Signal
2032"
Analyzer Type Bruel & Kjaer, Revision Oct., 1987
1047"
21 . "Instruction Manual 1 047 Exciter Control Type Bruel & Kjaer,
Revision Feb., 1979
2707"
22. "Instruction Manual Power Amplifier Type Bruel & Kjaer,
Revision Aug., 1982
V"
23. "Instruction Manual Vibration Exciter System Bruel & Kjaer,
Revision Apr., 1982
Measurements"
24. O. Dossing "Structural Testing Part 1: Mechanical Mobility
Bruel & Kjaer, Revision Apr., 1988
Simulation"
25. O. Dossing "Structural Testing Part 2: Modal Analysis and
Bruel & Kjaer, Revision Mar., 1988
Booklet"
26. Et. al: "Vibration Testing Bruel & Kjaer, 1983
Booklet"
27. Et. al: "Measuring Vibration Bruel & Kjaer, 1983
Designers"
28. "Piezoelectric Technology Data for Morgan Matroc Inc., 1993
Ceramics"
29. "Guide to Modern Piezoelectric Morgan Matroc Inc., 1993
List"
30. "Application Notes: Product Summary and Price ATOCHEM Piezo

Film Sensor Division, 1993
101
Components"
31 "Piezo Film ATOCHEM Piezo Film Sensor Division,
Number 11 Rev.(7/91), 1993
"
32. P Buchhave Laser Doppler Vibration Measurements using Variable
Shift"
Frequency DISA Information pp. 15-20 No. 18 Feb. 1975
33. "Instruction and Service Manual for Dantec 55X Laser Vibrometer"
Dantec
Electronics
Vibrometer"
34. "The Laser Doppler Appl. Notes Dantec Electronics
35. M. Serridge and T.R. Licht "Piezoelectric Accelerometer and Vibration

Handbook"
Preamplifier Bruel & Kjaer, Revision Nov., 1987
Considerations"
36. D. Corelli and D.L. Brown "Impact Testing Proc. 1st
International Modal Analysis Conf, Orlando, Fla. pp. 735-742, 1982
2626"
37. "Instruction Manual Conditioning Amplifier Type Bruel & Kjaer
Instruments"
38. "Master Catalog Electronics Bruel & Kjaer, 1980
"
39 Dr. Richard Budynas Testing using the B&K 2032
Modal Analyzer and
Software"
SMS Star Vibrations Laboratory #3, Feb. 1992
Methods"
40 Dr. Richard Budynas "Vibrations Testing Vibrations Laboratory

#2, Jan. 1992
Engineering"
41. W.D. Callister, Jr. "Materials Science and John and Wiley &
Sons, Inc., 1985
102
APPENDICES
103
Appendix A
Analytical Determination
of
Natural Frequencies
for
Materials : Aluminum and
Laminated Graphite Epoxy
104
Analytical Solution for determining
the first three Natural Frequencies
of an Aluminum Cantilever Beam
Governing Equation
cv
cfv . d'Y n
FI ^ + pA r
=
0
dx4
dr
:4r =
o
Y(x)e"" z4
( pA\
co:
where y =
and =
Fh
General Solution
Y(x) =
a cosh(zx) + b sinh(rx) +c cos(rx) -r d sin(zx)
Boundary Conditions
Location Physical Condition
x =
0 Displacement is 0
x =
0 Slope is 0
x
=
L Moment is 0
x
=
L Shear is 0
Table A. 1
Parameters for the Aluminum Cantilever Beam

10x10
Elasticity (lb/ in2) [41]

Width (in) 1
Thickness (in) 1/8
Length (in) 10
(lbs-
0.09729/386
Density /in*) [41]
1.63xl0"4
[in4
Inertia )
Results
Mode I Mode 2 Mode 3
z .1875 4695 .7855
Natural Frequency 40 2 252 2 705.8
105
of an Aluminum Cantilever Beam with an End Mass
Governing Equation
CT
^V A
dA rs
dx dr
Y"-z4Y =
0
Y(x)e,c"
where y =
and z4=co2\^-)
FI { )
General Solution
Y(x) =
acosh{zx) + bsmh(zx)+ccoizx) + dsm(zx)
Boundary Conditions
x =
0 Displacement is 0
x =
0 Slope is 0
x =
L Moment is 0
EIY'"
x
=
L + o):MY =
0, End Mass
Table A.2

10x10
Elasticity {lb/ in :) [41]

Width (in) 1
Thickness (in) 1/8
Length (in) 10
(lbs2 "5
End Mass /in) 1.54x10

(lbs2
/in4) [41] 0.09729/386

Density
"4
{in4 1.63 \10
Inertia )
Results
Mode I Mode 2 Mode 3
z .1793 .4515 .7591
Natural Frequency 36.8 233 2 659 2
106
of a Composite Cantilever Beam
Governing Equation
cTy .d2v
r.T
EI pA~ = n
-
+ 0
dx df
-z4Y = 0
where v =
Y(x)e""
and
z4
=
or
( pA
FI
General Solution
Y(x) =
acosh(zx) + bsinh{zx) +
ccos(oc) + d sin(zx)
Boundary Conditions
x =
0 Displacement is 0
x
=
0 Slope is 0
x
=
L Moment is 0
EIY'"
x =
L + arMY =
0, End Mass
Table A3

Equilivant Elasticity^ Inertia 295.4
(lb in2) [10]

Width (in) 1
Thickness (in) 0.063
Length (in) 10
1.507xl0"4
(lbs2
Density /in4)
Results
Mode I Mode 2 McxJe 3
~
.4695 .7855
.1875
Natural 31.2 195.7 547.7

Frequency
107
of a Composite Cantilever Beam with an End Mass
Governing Equation
cf\'
.... .d2y .
FI -
dx4
+-
pA
F ^
=
0
dr
Y"'-:dY =
0
Y(x)e'01
where v =
and =
of
FI i
General Solution
Y(x) = a cosh{zx) -r b smY\(zx) + c cos(zx) + d s\n(zx)
Boundary Conditions
x =
0 Displacement is 0
x
=
0 Slope is 0
x
=
L Moment is 0
EIY'"
x =
L + of MY =
0, End Mass
Table A.4

Equilivant Elasticity& Inertia 295.4
(lb in1) [10]

Width (in) 1
Thickness (in) 0.063
Length (/'//) 10
1.54x10-'
/in)
(lbs2
End Mass
(lbs2
4
/in4) 1 507x10
Density
Results
Mode 1 Mode 2 Mode 3
z .1653
4308 7356
Natural 24 3 164 7 4803

Frequency
108
Appendix B
Impact Excitation Data
from
Accelerometer. Laser. Piezoceramics and Piezofilms
Material: Aluminum
109
<?
/A
ACCELEROMETER DATA
no
' ' " ' ' ' ' ' ' ' '" '
f ; : i i ! 1 !
j
1 :! !
li. 1 I
1
i ' i '
E E (M
j i
1 ! . 1 -a
' '
->aizu
i 1
0
TJ
i !
-
: L a
i "> 3
.
Q0)>>
.\N UJ
D
J ! i i H
a m ! i \i i ' n r-. CM01O E J
-a i a. hSDO <
tn a.
'
i i I i >
tNN , 1 ! ! !
z o
~> E
.
ii
- .a
ntn
ininrj
i : ! i ! J : .
xa
a
<
l-h U
UJ
>)
< .
so
ZZ N N
tUUJII
a
l
_I_I.E.
10(0
J 5
mm u s
c\i<\i z
(M Ll
CD I 1
I
10U.U. a
C\J
LU
(\j
t^
N -oca: a tg X .CO
I UJ
in HDD a:
-i u.
INN
ain .
u. mm a
-Si
<
< * s
i-
az k+ + a a
ZUJ z uj -
Nmu z
ILUZ D
UJ J
m
 (fltE
xm>LU
toat-
lu
LU
<
a< m tr
z .uj . .<
LU LD zci I CD.
2 - Z <LlU u a
UJQ. M Q. i-_i
orui .u Ul KI r-
<<
DUX WI I . .a>-<
u)o<or ? l-U u<muju

<-. j_i WZ - . .ZUJ
UjQrrUJ> qcujuj luiill; a.

2i-a< L.U2 2UUUU1
H3>-xui _____
.
c
->
^ c *
s.
"O
<.
-^-i Sv
^ ^
il
-a
a s
>_ -5
;o
1
i> a-
p ^
^ ^
5u
'o
X 5-
.is.
Table B. 1
Data Sheet # 1 for the Impulse Excitation (Transient Window)
Sensor: Accelerometer
Materiakaluminum
Natural Frequency:34.0
Mode: 1
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 392
42 14
43 15
44 15
47 2
48 0
76 7
77 0
FREQUECNY RESPONSE
-? CO
LEFT RIGHT
Left Element # =
392
Right Element # =
406
From the Impulse Response Magnitude

Time Constant T (sec) =
3.023
=
Damping Ratio .001548
Loss Factor 7] = .003097
113
-5
IS
^
4J
T
^ o-
-3
^ Is
a
^^
a i> Q s>
s --a
5 ^S
^
M
^
5j
5 ft: -5
ii 5
il ^
o ^
sj
^ ?J
W .
J
m
<
ld tr
>>wUJ
mmai-
<
r-
a:
z .w . . .<
w u ZEII CEO.
s - z <u.uu o
wo: - Q. h-J
irm -LD ma: i-h <<
DUX wn . .{-
uiuxa QhDU <muju

<-_lLU LUZ1-"-1 . .zw
UJKW> imiiiiiii IIUJQ.

IhQ< llOsj UUUU1
1=
>* "Vi
^
-"
**
l_,
3
e, * a in
*"
5 -s
^ -a
-Si
* ft;
^ a
i>
CD N
-Dim
hQQ
imm .
<
Ul: N
Iffll
oooi
C3INQ>-h
LU: .OS
.U-jPiaii
t-
-1
- .UJ
S>XU1
Ul
:
i
j
!
;
I
! ! I
ill:'
M : 0
0(0
m
rt
i
] i ' i ' '
i 0 O01ZO
OtsNN UJ
.
CO
<
1 i 1 ! 1 - ^ E
TJ : j 1
'
i ! i ; ! i H
t-
io a .>> D
J
oi m 3 ,NEE
1^, [S
I i ! 1
! i i -
i
a.
Z
o.
UJ
m
-i
hhOQ
. .ao
<
>
! i 1 i 1 1 !
>- u
o
.
ii^-
U
/
r- r-r-
0 00 i 1 ; i i .
i
X ffl< < OO UJ
torn i o
i
'
1 I ! ' i r< < E ZZ N N
-Q(\J
1 ! SO J UJUIII
>0)O .
! | i ;
i
r \
;
i I >
! t
i
i
0 (\I(M h-
!i + 10
OCX<< 1 I ! ' .
i-i-
.J-;
a
4 1 i rH QC .*>
' I 1 ( 1 1
JJ UJ ?i)
rs 03
-* .u
/ 1 j 1 r | i 1 1 a. F
| 1 3
j -2
1
' UUJ 0 OLLLL < a
1 f 1 j I ! :
i-
/ ' 1 : 1 i z> E E
OO VO
1-
j 1 r i-iUJ ,1
! N
*<*
z
o
1 i
1 ] ': f UJ
< x - - <
X
1
ex. a
Q- : i !"? e QC
\ QC . .oca:
Z ! , ' )-(--"-' h-
i 1 : U1U1 m 2j
*H
*
! J i i >a. . itao ij î â
: <o PJ MM U.
j / i 1 '
a -a
: 1 j ! | ; | j o (0 II N N
I'1
'
i | 11)1)111
/ SU1 E
r
m
\ 1 i .
! ,
! 3 + (\l u. n mm * -Si
i / >
!
i
j i
i
rm
QC
t-
-
m
>'-'UJ
., /
Q-O i 1 QUI-_J mNt t-ni(Dnr-
U10D . 4 i r i i UJ
LU < i / 11 i i. 2
j
QC % t / ... <
!
0 / r H ? <CD QC
w o i if i r z .LU . . .<
mm \ \ ! 01 UJ o zacii QCQ.
JT3 + / ! 1 i
r-
rH
X z <LLUU a
or- i
i
F 1i
i i .
i 3 UJQC "-
.
a. (-_!
CL . 0 / tn tr t t <<
zm e / i
!
'
i i
1 '
i i a.
QCUJ .O
DUX uu . . ItH
mo 7 I I ! :
a uio<o: ? i-oo<a)UJU
# i
! '
' l
1
i
1 j' ! ': r-
<_;,jj LUZ"-"-" . Z'J
x ui uj uj 1 1 uj a.
j :
'
i .
LU ujqcuj>
^ zhg< ltU-.3UUUul
JjVx
. \ 3
, i i i 1 1 -lÂ-q
Ul
3 3 t3 3
a a 3 (3
M
(M I 1 1
1
Table B.2
Data Sheet #2 for the Impulse Excitation (Transient Window)
Materiahaluminum
Mode: 2
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 378
42 23
43 15
44 15
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
-><Q
LEFT RIGHT
Left Element # =
378
Right Element # =
401

1.5
Damping Ratio
= .0004812
117
0(0
o O01ZO
CD E ONNS UJ
D h 10 O .>> D
in m D h .NEE _1
. N NTJ Q. Q. Ul HiÔQ <

-IIPJ Z LU . .ao >
I0OO . (J ii
oom U i-t-
U
mint X 0< oo UJ
< E ZZ N N a
vo'tri . ZQJ LUUJII
0J(\J < j_ijcj:
CM I m -d (0(0 5 -S
rHCOZ
U. . . t< iriiri a.
UJ -x> 0<2. 0 c\ic\i ui a"
QCX<< + . (0 LU * *>
i-i- Z\ ii ^-*
I cr I .'4>
.u _i_i
_l
HH *H a
OLU 0U.U. LU
z> E QC
11 UJ o U. 00
N

^
>Q- U.U.QQ LU a
<o
0
-1
i\i
to I INN
3 H ^
a A -a
21/1 E NU1 Kill c ft; **-
D + (\J U.I mm
QC <^
5; ?\
-a o
s
m C\l 3" a
Nr- t- O ii
U i a >> a
LU a Z z + +a
Q. UJ LU UJ .
^ o
%i
Ul < 1-1 ,v
t Ul ui m -s
_i<o NIZ z <
< .>-z IO< < mac
DICh QOQC ac>>'-'LU
U1NI- HmODDI-
aui-j
LU
s
<
a< CD QC
z .LU . .<
0) UJ a zaci i aco.
z . z <u.u CJ o
i-
UJQC "-" a. j
U1QCI- i-
QC'JJ -O <<
CL DUX UII I . -QC-i
D U)U<QC ai-u u<muju

1- <-. JUJ uiz>- - . .ZUJ
UJ luq:uj> QCLULU UJIIUJQ.

Ul Zl-Q< LlU* 3UUOU1
!"
- - ! -
in
1 ! *
1 rj !
! 1 :
' ''
i i 1i 1
1
!1 1!
1 : I i 1 1
O 1 i : 1
m A ! '
! -
a
s i
o
in
m
"?
nh
.
A
i
i
;I
I
|
:
P
i
j
i
; j
:N
e e
m N NO
"II .
MI .
*
! j
1
cooom ,
! raaom
QQCO -.:
| m \ i ; in ooim
mm-" A i . mm i
1 ;
' ! I 1
>-qm - A i
'
! .
(MM , 1
i h (\J(\I ^ O
1
PJ 1 i
, \ :
(M 1 J-
i * **.
^
l
LU -x>-
-
ii ! |
<
j a
ro
w
;
,
!
'
i
i 1
(i)
(M
U. -
LU -X>
Q
*. *-^
^
ocx<< : ; ' QCX<<3 v Ia î
: !
1 1 i '
\
i
A ! 1 tn
ra
i
- in 3 st: t\
(M
in. '
I'll'J 1
a 5a ?-
Si
Os
z ! "t L !
j
j*5^2'^!-
'
i
[
l
:s (M
i
f
! !
1
'
- a
(V
1S SJ X
1
1 ; :
| 00
g -a
j Z
~
H !
'
i ' ; .
1 i 1 11 ^
^*
Cl -a
J '
nn -i
i. 'a
i "
2 *
55
MM ft:
1
i i f
4
U T , I <y w
!! '
i
i :l
:
1 N ij
N ; i 
CQQ 1 a l 1 rQ
a
"Dinm i 1
, | (\i
'
i .1.
"O ! ><,
: +
IO+
<
.
'
:
!
t
i
|
<
.
.
I i
Q. N* A UJ N*
, "
(fl I H
' i in I in u i
' <
lu o : i ? >
9 2 a
'
' <\i (M LU O
ocmo
*
1 ! ! or ooi
"0O01 : !
1 i LUO
? ID rH . ; ! .-*
UJ .(02 IO(0
. | 1
ocmoi
LLCO^Q.
"
1
. a ; 1 j 1 .
"
a a .01
U""fL
a
D A rj I'll CM D
1-
i- A : 1 i tf
! " - 'll
'
'-"
Ui
i
.
i.
.
i
"
3>XJ1
JjVxui
| 1 I i :i , .... .... ...
a a a a c) (D (0 * IM C3 I
ID * (\1
: d a cj d
S a
3 a $,
X ri ^
9> r? ?
43 =0
a su
o
(N
'
^
3
^
^
a
r
Is
is
i. ^
"O
"o
Table B.3
Data Sheet #3 for the Impulse Excitation (Exponential Window)
Materiakaluminum
Mode:2
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 367
42 69
43 25
44 25
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
LEFT RIGHT
Left Element # =
367
Right Element # =
467
Weight Ch. B: Exponential Length => T^, (sec) 10

Time Constant T (sec) =.812
Ratio =
Damping .0008144
Loss Factor 7] =
.001629
121
I
5^
~ a
3*.
'j-
L)
-N -0 a
i) >*,
ii
a s a?'
a
*
il il I"
1 >*
-a
* > o,
a. Cl
i.
5 ii -a
o
<
> s
4J
ii a *
o ii
o
<
^
>
N
I
mo
uinra
-o
I0D+ -
<
0. N*
tn i
LU Q
qccdo
"DO01
ao .rt ujo .
LU .103: loioa
KCMCD a o .CD
u.mrtQ. g Urt-Q,
D N D
(M I- * H
rt . .UJ
.3.- .UJ
LU>-xcn
d
m
a (\l
oo
ID rôzo
E ODONN Ul
(0 .>> D
OOO _l
in rt rt Q Q <
.... >
Ilrtrt
t- l-l-
u
< uu UJ
ZZ N N Q
LUUJII
-1-1XX
(0(0 CM
<
mm
o (M(M Li 1-
CO m m
QC i>
LU o
i)
.a
HM Q. -L,
a ou.u. < 42
E E a >
05
N
mm
Z 1
-" CN
I mm < cl
E . -QCQC QC ii ^
m HI-mh r-
ii ft: i;
U.U.QQ -H
t\3 il i. s
(0 II N N 5 a
a
NIfllflll . 0 a
u.1 mm * L. I- 5
<3rt _J -Si
(M < % il -^
(Ml- - O
O
zi-+ +a
^
UJZ LU .
"UJ _I
uiz a
NJZO <
ia<Q.>>wir
OQQCXQO'-UJ
inNi-uic\imai-
tu
<
I-
? <m QC
z -uj . . .<
LU U ZQCII oca.
X . z <U-UU a
ujo: 0.
qcuj -o wach-f- <<
DUX IUII -QC
U1U<QC a(-uu<muju
<_JUJ UJZ>-"-> . .ziu
UJ LUQCLU> QCLULULUIILUQ.
H-Q< IlUJUU(D
r\
LASER DATA
124
a o
E E 0 rj
-a ZN
rt(Q S E
a >S UJ
DO) E> D
t\10J aw _l
rtin <
a did >
m ii r-CD
hh U
< uu UJ
zz N N a
UJLU II
_1J XJC
I
(0(0
mm
(VI (M
0 LU 2:
?* >
.
5?
Ihl-
ra
a ejj
EI\|HH
a . a
3 s
ir SS.
gj
-1
a
.a LU <j~i
CO-* i-
CN
i
-2
a s Si
N u
I
. .QCQC
HI- <--
i r-l
ii
m u-lloq UJ
II N N
Br
il
Ull/l II
Ll mm
.
!> as
s
< J
< *
ii
azi-
zuiz u .
<<-UJ _j
ii
CD
.L.
NCQUJZ
iujzo > <
auxo. in uxr
a<trx
rtfflQI-
ui
<
a<m QC
-LU . . .<
ztrn oca.
<u.uu a
a. l-J
U1QCHH <<
mil mK
ai-ou <muiu
UIZ*-"-" . .ZUJ
QCU1UJUJ IIUJCL
ILU>J UUUU)
Table B.4
Sensor:Laser
Materiahaluminum
Mode: 1
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 401
42 27
43 15
44 15
47 0
48 2
76 0
i-7
2
FREQUECNY RESPONSE
-
oo
<,.
LEFT RIGHT
Left Element # =
403
Right Element # =
430

4.0
Ratio =
Damping .0008506
Loss Factor 701

7] =
.001
126
v-,t,"|'
r '
4 1 \ 1 L
1
i
41 1 i -
Jr :
i! if! i- o -| .
! ! -
a
! '
| ; i - 4^ , *
i 1
t j1
1
? 1 r
f 1
i
<t i- O
<
3J
|I -a
|>! r >?
[
j > "*
E
O CD r i 1 t- a
in n tj -, ? 1 1
I
oi n in
II NO t ! f I"
i Oil N .
OI . ->
,)
(2 1 1 r * QIO
OOO > F* i "*
OOD)
r
.ma tl J r
r ! I y -mo
>-
.a
r-
.
ti
1
r '
l > .o . *
c
. !
r N .
i- w
mo -ii
4 ( "*
.
1
'
! 1
'
-
,
N
*
mo
^ ~
' 1
LL. il L LL
1
B
'
_ 1 1 .
LU .XV
QCX<<
4
J j (
t
i
3
'
J
'
i
l
. UJ
QCX<<
-X>
a **.
-1
u
c
1 I h a ! 1 a Fu
, t
1 1 / !
j <N
r-
4
4
4
\J
>^
1
t
L (D
1
i
1
a
e
of
d j: m J 1 m
*^\
5 s fc
__
t
Q. S!*---^ i! ! [ 3 '
-
j1
z ^ : ; j
?" J X
t SfL,
4
i
- a y- ;
a
5 h,
'
m fl) 51
j (
, .
i S -s;
i> -i. S
3 i !
\1 :*
cs ri
<
5 1
1 i
-
,
*
m
1
1
1
1 . m
z
*>
_i
a
5>
s
2 I !
4
i ! (\l ji
N 4 T m a m
mi i ,
! i
uinm ] i mm
! i\i
*]
rtOPJ
I0D L
-
h a 1 8 +<
! _ i
a.
(0 N
* i
4
i i U N
] Z I
iucdi :: 1
CTTJO 1!
a
tan
.uirt
-
i
1
. CM i S
w
QC
ujairirt
IO
OOJ
turn .3 il f i
.3
ocmt
LL 1 NO.
1
1
N_ i ! '
o .*
Urt(\|Q.
- r O O 3
D
t-
(\l 4^
rt * I rt . .UJ
.
g>-X0)
.UJ
| 1 1 JVXI/1
t ....
a a a a O o cu
to ID a N
i I a d a
ID 4
TJ
CD
CD .
(0 "?
t^ (n
13
a (DO -ii
*a i
.ma
>-r-a "51
. .
.
]
< ri
=
^
ri '-S
~ >< ^j
LL -
Ul -x>
j
S J
o:x<<
<n
^ ri 00
OS CN
5,*c^ ?
I
1
s: &.
5.
Table B.5
Sensor: Laser
Materiahaluminum
Mode: 2
SPECIAL PARAMETERS
# Dec Value
0 20? 2
40 2
41 355
42 28
4? 15
44 15
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
co
CO,
LEFT RIGHT
Left Element # =
355
Right Element # =
383
Weight Ch. B: Exponential Length => T^(sec) -

10

1.984
Ratio =
Damping .0002735
Loss Factor 7] =
.000547
129
" " " "
I 'i
1!
J
% 1-s
L
a
o
1 N '
CU
j r
m
TJ
J
F
fLin :
>
in
Ul in
jJ
e
r w N
.NO. rtN a
mi 13 oi m
'
i m no / oiin n .
(MI . ff i
(Mia
rtOO .
.
} o .
Q rtQ(l) .Su
. .a { in in . .o "^
>mo . (\j w >ma
/
.
m
(MO
. 1 i
i
m .
(MO
1
UJ -X>
J / i
:! ui.x>
5 S
2
8
s
*3
QCX<<
j
t
N N QCX<<
s
^ rS
>
a a
^
-
1- ?
3
o
3
Q.
(U
1^ 5> ro
Z
?-
PI 1
* * in
m .
">
cu
z r Z
-
-J _J
a 5
.
a - a
">
a
i m .
< ni
N
I
a
in*
\ \
\ in
1
in
N
I
a
int
-t,
rt(D
IT)+
a <
. \ l
N
n
CU
' + -
<
CLOD N*
U) I
\\ (
i
,
.
UJ
U
N*
I
uj
QC
a
O *
\\ a i
1
.q
n
Z
ui
O
a *
ooo \ (U cu a. ao
L ma
? *D -
'
I .rt
UJ(M(M3
QC .-<
.
\
i i
. ION*
a -
U.OCMQ.
(M
3
1- \\ h
.in
(\l
*
3
"2
'
,
UrtCUO.
*
3
t 1
Jl
-
SUJ
gjyxoi SVXD1 1
a a o cu
o o
cu o
I i i a ?
r
0 0
o s
D oaz E
E EQNS
V- (0 (0 .>>
D r-
aiaarj
Q. 0. in tvrtQ .
Z LU . . .ID
U IlrtO)
U
X 0< OU
<J E ZZ N N
SO J UJU1II
< j_ij:jc
m .a (0(0
rtffiZ
. ?< mm a.
o<s: e (M(M (fl
ST
4-
. (0 . . UJ 'o
i
-J-i
-u
_I
a
OLU e au.u. UJ 1*3
Z>
1-1 UJ
E E
oo
QC
U.
î 3&
m
N (M(M is.
< r mm I
I
QC E . -QCQC o 5^
S.
LULU |J h-H-"rt
-^ OS
>(L . U.U.OO UJ 5u
-s;
<a S
_i o (0 II N N
ZOl E NU11/1II
.
D + (D u.i mm
a <ts
a: _J
m ma:
aiac-1 aaocxo>rtUJ
inNhUlHODQt-
oui-j
UJ
I
<
H a<o QC
.<
Z .ui . .
a uj a ZQCII oca.
rt z . z <LlUU a
* woe ?-< a. H_J
qcui .a inQCHH <<
UJII new
a. :u><
ino<ac ? hoo<ouju
i- <rt_lUl UJZ1-11-1 -ZUJ
UJ LUQCLU> QC UJ UJ UJ 1 1 UJ ft.
zh-o< U.UJs*UUUJ)
j)
Table B.6
Sensor: Laser
Materiahaluminum
Mode: 2
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 358
42 22
43 20
44 20
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
-?CO
LEFT RIGHT
Left Element # =
358
Right Element # =
380

Time Constant T (sec) = 2.492
=
132
IB
ea
a r-az
E oms UJ
(0 -> .> a
.(MO j
m rt .r.Q <
. . . >
II-
l-l-
(j
< oo UJ
ZZ N a "5
UJUJI Ci
_1_IJC
(0
2
- "3
s:
mm cl 5
o (M
(Mf in i S*
a
(0 m UJ CS .5* -Hi
f-
QC CS >3 a.
_l
> C/->
rH
a
o ou. u. UJ ^
<3 g -i
E E QC
mm u. of
3
i mm I 2i
E . .QC QC a
m n - e
U.U.O a UJ 7 & k
(M *
(0 UN fc3^
Ncncni
u.i m *5
-J
s
m *j
(Mh-H a
ZZ + ta o
UJUJ uj .
 01 QC
oaociica >iu
IflNhhH
Ul
s
<
H ? <m QC
Z .uj . . .<
UJ zocn QCQ.
2 . <u.uu a
UJK - a. i-_j
OIQCHI-
qcuj .a <<
aux U1II . -QCrt
U)U<QC ai-uu< muu

<>H_1UJ UJZrtrt .ZUJ
uiacuj> QCUJ UJUJI XUJQ.

Zr-a< U-USJtU UUU)
LljVx
I
4 .
\
.
m
a o
\ 1
i
,
N
. n Nm I m n n
Oil . ]1 Oil E
iwinm
(Mrô *
1
! ainitiH
IMMO
I
rtO rt .
hOh
at. 1 "^3
vinio . , m<ri .
1
rnrj ! m<M
|
CJ 1 1
(M I
.
u.
Ul -x>
-
j1 5 ul-x> <
QCX<<
QCX<<
j 1
N
CS "CJ .
'
t t\
*
K j
3 Si *>
D is -s ^
S ^r
Q.
Z 1
M
^ &^
z
rt
_J
O
< &
s c
N
I
J
8.
ma
hQ
IO+
ouio
.
J
,
I
< ]
a. n* \
01 I j
uj a j
ocma
"Daa <
(3[S ,r1
UJ .NS .
rrort
U. 1 (MQ.
a
(M H i
rt . .UJ ,
HI>-XU1
a a
cu co
i i
* 3^ aj
St?"0
00
--I
i*3 <*> &
eq
$
</ Io Is1
cs <
l/~)
<
- ?
1
-3 -s;
iu to
Q.IO
01O 0 .
UJI o<
'OC *
+ <
uj! ? <o QC
IDD1 z .UJ . . .<
E
_lP5 UJ O ZQCXI QCQ.
oino s . z <U-UU O
nX.in uitr -< CL hj
z,tMm QCUJ U1QCHI-
-O <<
aux UJII .Km
U)U<QC ? HUU<ObJU
<rtJUJ UJZ-<rt . .ZUJ
UJQCUJ> QCUJUJUJIIUJQ.
2hD< ILU3XJUUU)
Q 0 ? a
rg m in
i i i i
it
*^ ** ij
S 5
^ a* S SO
^
S
1
8 ^
an
O 0
. (Q
I
ua
E + a
u/m
so os
rt .E
Kmaa.
a
(M H
rt . .UJ
J>Xll)
/\
PIEZOCERAMIC DATA
137
IS
I
1 1
=
^
CL
S^
>J *
S "i -S
!^
*5 rS
7"
C5
^> s
^ ft i .a .
OCJ
m : is
I
<=
uj .x>
t;
O -r~
"o
QCX<< K *! <~
^
<< ?
Os
D
a. 5.
?j
i
z
O 5 -^
U 7.
< -
2J
3 Nm
rtQX
o<
t/1
Ui +
oca
"u NO) "
Qtlrt
^J .OS -
^
acai .
ii. : cc -
a .
sr>-.
Table B.7
Sensor:Piezoceramic
Materiahaluminum
Mode:l
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 417
42 13
43 10
44 10
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
-> co
co
LEFT RIGHT
Left Element # =
417
Right Element # =
430

2.687
Ratio C=
Damping .001536
Loss Factor 7] =
.003072
140
H
OO z
0 (Mi^za
E rtinw UJ
(0 m .>> a
i-
a .m eo j
o. a. in hhQQ <
z ui . .a . >
- QC
rtUJ u.
OJ
<
N
i I
u
1 s*s
QC E 1
2 ^
.QCQC
LULU m 1-1
>(L iLQD Ul Si ^Ji s;

<o CM *
0 (0 I N N l
-1
21 01 E 0111 2 X
a + rj U. N mm *
x *-> <
uom < .wz ia< <> EWQC
aiocrt OOQC ocmortui
t-rt(I)OH-
xo aut-j DNh
+ < UJ
a. * I
0! N <
t- a< m cc
uimi
trvo z .uj . .<
mo a UJ o ZQCI i aco.
a .a z . z <u.u u a
LU CD . UJQC rt Q. H_l
acmrj QCUJ .O OIQCH r-
<<
u. i - a. aox UJI i . .QCrt
o 01O<QC ? HU u<ouju
H <rtJUJ UJZ-1 - . .ZUJ
UJ UJQCUI> QCUJUJ ujiiujcl

j*>x 0) ZhQ< U.U3 SJUUI/I
5
13
S ]3
5
Ci
CS
C
5
c
a S >
M < =
-s <; >
as g
O r,.
r^*
I
sit:
2: ?s
a. a
5j
\0
Q rt CM
ao z
B (Mt^za
E H|/|\\ UJ
r-
(0 m .>> a
3 r-
.m eo j
a. a. in rlHQD <
z UJ . .a . >
-. u XXHH
u
HH- u
X 0< UO Ul
< E ZZ N N a
SO-J UJUJII
< _I_IJCJC
m -a (0(0
rtflDZ
. t< mm a. 3 ?}
0 (M(M 01
+ . to _ - Ui ** a ~
I l-H a:
^
.U
_ij
.a >3
i-i i-i a il kS- .
>
_J
U.U. UJ
Otil
z> 0 B CC "a % S3
i-iLU E E u. m
a a cs rt
U_J
i
n mm
?5 Si, 5 Si
<
2 ^
QC E .
-(
.occc
rt rt
u F?c
UJU1 m
>a. .U.OQ UJ
S
s;
<a (M 5{
(0 :i n n 5 <
j o *- a
zoi E 101II j" 5. 5
a + cm U. N mm
QC <
< .rtZ ia< <> eoiq:
aiocrt aooc ocmortuj
i-rtmat-
aui-_i U1NH
ui
<
?<m ac
-<
z .LU .
01 O ZQCI i oca.
jh 2 . Z <LLU u o
ysi ujqc CL H_l
I- <<
: QCUJ -U 01QCH
:o.i au>< UJI I . -art
D 01O<QC ai-u u<duju

UJZ-
,|_ <t-iJUJ . .zui
ir ni a. m > QCUJUJ UJIIUJfL

ZHD< 3UUUJI
JlJ
Table B.8
Sensor: Piezoceramic
Materiahaluminum
Mode:2
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 402
42 26
43 15
44 15
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
? oo
LEFT RIGHT
Left Element # =
402
Right Element # =
428
Tw(sec) =
12.5
Weight Ch. B: Exponential Length =

1.234
=
144
\0
BO Z
rjoza
-nnxN
m .>>
.IMOO
rirtQQ <
>
II rtrt
\~r-
CJ
LU
zz :: n Q 8
UJUJI I
J-JJCJC
(0(0
sl
mm cl cs c
rj(M 01
~
?
. . UJ 3
r-h-
oc
in
1-41-1 ?
LU
o e QC Jv N
LL
5 sr.
E E
*
mm
.-1 ,-*
It
g <
IQ<1> E01QC
OQQCXmOrtUJ
inMH'JiClQh
11
s
_ . . <
? <m QC
. . .<
.ui
ZQCII oca.
<iuu a
C (-_;
OlQCI-t- <<
LUII -
.QCrt
? (-Ut3<OUI(J
LU Z >- i-i . . Z 'll ~
OrC 'i! UJ u: I Xc
Ll Ll 3: S U Li O 01
s
cs ^ ^
*" r*<
.>
tS"a =
*>
Cr5 s
Vj
<
0
^ -a 5u
a 03 y
I"
Si 5
~> *~
5,
Si
I
N 3^> ^.
.*
~. -5: a.
>L> "> Sf
m a
T r~
-O
S S S
ii
g
=<
^
Si
1
* -I
5.
A\
PIEZOFILM DATA
148
, ,
y , . p. -
ir
1 i
1 !
'
1 \ 1 1 1 i
[ J ' ' ! B B
.
i
| '
1
'
1 S1 f EO
rtQ
O
1 J
CD J ; I
rtOZZ -
"0 O ; i / !
'
a
1-
B .OSN UJ
(D Tl
; A . =i oa>> a
a h t- C\l . E*
1 J i 1 J
1 \ r i cl a. [s rtrtO , <
ON.
! / ! Z BUJ ?i . .at >
1 1(0 /' - EU iirtrt
Nlf)
I.I;
J ,
i : I xo<
ou
<
-
l- l-l-
OU
u
UJ
1/ ! '
l 1
-
.mo
1 <
>.a. ,
| 1 3
?
. ZZ N n a
rsrt 1 -tt
f 1i ! 1
'
JDJ LUUJII
l ^
<
m i i 1 )\ if m .a
-l-JJCJC
(oio a
j !
IL .
UJ .X> .
1
1
i if j i r
rtfflZ
.
0<2
t< mm
CMCM
o.
01
1 Sit
QCX<< 1
+ 8
i ^
. . . UJ
I (M I-I-
QC 1
4 -u . B B_l_J 1
_i
1-
E Erti-i Q 11 *2
' ; OUJ minu.lL uj 1I
1- * i f ^^. : : z> oo or 1 a, Jg ^>
3
i-i LU u. 1
o_i M-> 1
a. i < ft I 1
z S ^ ^
a In
X N . .QCQC U 11
! i
1
l \\ i ! i- CM * UJUJ I 1 H-rtrt i_
1 **
>CL B m u.u.aa uj 1I '
6 J
"
\ <a e . ^ i ^
:
.
i 1
i
i
I
1
\
t
1
t
Jh
soia
O
.
II N N
010)11 .
1
1
4 a + . u. mm in 1
1
1 1
1i V
j -
a
* cr a < _i * 1 ^
s
o z ; ! 1 i \ in >
H
u
O
i a i-rt
< * 1l
? 1 ; Ul o azi-t + a 11 "-
S -J i
;
,
'
.
S CL .-^
ZUJZ UJ . 1 3f>
! 01 < <rtUJ 1
& _l
j"
1 1 t nooiz m 1
! i
i I r IN _i<o IUJZO < 1
Nm -
i
< .rtZ oiixi oiac I
rtDI | OIQCrt o<o:x>>rtuj
ITJO .
OQ<
, ! i
'
1
us 5: QUhJ toi-ujmiMai-
|
ILOt* i
1
1.
1 H
'.
UJ I
; i
z 1
tn . ^^) 1 m ... < 1
UJ +
i
^*
i -A 1-
? <ffl QC I
X ^-
ft Z -Ul . . .<
ON: ; !l >
-
! i \^^s~ 1 a
LU O ZQCII QCQ.
atii
ujmo
i ! ;
-
3
z -
U1QC
z
-i
<u.uu
a.
a I
<
-
i. u
.
h_i
"*
QC ..
"^ l QCUJ -U OIQCK-H <<
U.01O -
^v^ r o- aox UJII . .QC-
|
i
1
1
i
^\V.
"\^ i
i 'r
Ir
Z
'-U
01U<QC
<-_;uj
ujq:ui>
? i-uu<auju
UJZrtrt .
X UJ UJ LU 1 1 LU a.
.ZUJ
IJ^x i 1 i 1 V*-<-. 1 -Q
si-a< U.U33UUL301
oaoaaoaa
CM IT> m
l 1 1 1 1 i i
3
S ^ -2
5 &! y
o ^>
=5 o
wn
2!
-5
<H r?1
"^
^
^'2
.2;
Table B.9
Sensor: Piezofilm
Materiahaluminum
Mode:l
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 401
42 17
43 20
44 20
47 2
48 0
76 0
77 0
FREQUECNY RESPONSE
CO
LEFT RIGHT
Left Element # =
401
Right Element # =
428
Weight Ch. B: Exponential Length T (sec) =

13.0
W

3.359
Damping Ratio = .0009339
151
- ' ' '" llJl
J I I C^t ] Lr
' -'-'' '
M 1
! 1 >
-
?
o
i .
-^=~^C~\ I 3
O
1 ^
f : '
^ ! 1
ao -
i i / .in -<f U)
in
DO'l ; | j ;
i !
f !
: l
E
0> N
.mm i | J I 1 1 O NI E
Min i '
inrô
.
i / ! S i 1
'
oiimts
pjmo
J ' c '
mt^rj
>
-(D .
4
1
'
/
/ o
in
f
: '
L
.
q
in .(0
(\im*
.
J
.* . ; > .* .
/ 1 1
[-.CM
m, 4
m i
1
i i
11 -
i / in c: f- c
i
*
r
!
UJ -x>
QCX<<
J
. j
1
F*
r
I
3
: I ! -5 ui-x>
QCX<<
* Si
i
-a
1 ^ -s ,
-i
4 j / 1 1
i
j
' -
a
*2
1-
: t ^ c
.1
| ;
'1
1
1
,
5 ?-> S"
n. -*. 1 i ;
1 ! I
Z 1 "*>. 1
;
w .
1 i
1
l
s:
1 \\
^ -
in 1 I -
in
m ! m
4 .
i 1
.
! \ ; 1 1
1
| \ *
; !
1 \
2 i z
4 ) 1 p i i
rt I 1 \ rt
o
f \
-
-a >
U J
< -
i ro -1
Z i
1 i 1
1
N
x
\
* 1-
N
I
i J
'
\
am -. \ L m I -in am
Hirnn j .
; |\ 3 1 !
l <m m
It ! |( 1 l m
a+< i 1 r
'
+ <
(LO * l\ i Ul %
01 N .
1 ! l I i i U N
'il
uc
x
a *
-
:
i
i ! (1 i '< o
r Cd
-
: i .a
cj
z
u.
I
a *
!
maai ; i 1 ri.
'
i < OC OO)
a-DQrt .
I i i
!
i ; jiaort
UJrt ,} . i ! 1 L t i
|
IO .
X ,(\J " ) i 1 j a .ra
U. ts rt Q. ; \ r ': -t : Urt-HQ.
; f in in -3
a -
j i
i- < } *
rt
rt
. .LU \ _ .UJ
laiVxoi i ! V , I *>X01
a a a a ID IM
ID
I i i a 6 d Q
fi 6 ,
r*1
rS ^
=4i QL> ?J
lis in
1
15 ? ^
,'5
P^
Table B. 10
Sensor: Piezofilm
Materiahaluminum
Mode: 2
SPECIAL PARAMETERS
# Dec Value
0 2032
40 2
41 362
42 28
43 20
44 20
47 2
48 0
76 2
77 0
FREQUECNY RESPONSE
go
to-
LEFT RIGHT
Left Element # =
362
Right Element # = 390
Weight Ch. B: Exponential Length =

T^, (sec) 13

1.773
C=
Loss Factor 7] =
.0006582
154
m
TJ
r
OI
rCJ NO
T3 1
i
il
OI . 1
moo h
- ^ 4
"3
vino . t ^
m .
NO
s s
u. . ft .H
.2
Ul -x>
QCX<< s $
tti fc^4 ^4
J ^ y
ft
Si
3 S-'S 8.
C.
*v -ft
-. 3 Si
I
I |
'
;
-
% 5 -s
I
1 Si ** -^
1 | j
*
O 3

N*
I ! |
ui
(ECHO
a j Q
N
i i CJ
"DQOI i i
ON .rt
i i i
UJ .(MS
acort ; i i
1
i i i
U-rtTJO.
i >
in
D
H
i i
j j (M
-
. _LU i
J*>xoi A i i i
? a a
CJ c
I i i
a
a CM
BO
B oazz
E rtOSS UJ
r-
O w .>> 3
3 r-
.not J
0- 0. m hhQ . <
Z UJ . . .* >
- u Ilrtrt
u I-H (J
X 0< UU Ul
< E ZZ N N a
20J UJUJII
< -J-IJCJC
m .3 (0(D
rtfflZ
g
o c
. +< mm
0 CJCJ
+ . (0
-r
a: ^ 5A
.0 _1_1 UJ ri
rS
^
_1
a. f-i
fn
ULU B BU.U. < ^ u S B.
z> E E H
i-iUJ mm as S >0
g O Ci
N (\l(\l Z
Q. U.U.OO
<o
J a O II N N
2 0) E . NO) 0)11
3 + C\i u-i mm
QC <r^. _i
1-
m m o
_IOrtUi
DUhJ inNi-uwrai-
ui
X
<
r-
X
z .uj . . .<
Ul O ZQCII oca.
X . z <u.uu a
U1QC - a. i-_i
QCUJ .O OIQCHI-
<<
3UX UJII . .QCrt
0)U<QC a(-uu<muju
< JUJ UJZrtrt .ZUJ
UJQCU1> QCUJUJUIIIUJCL
x\-n< U.U33UUUU)
Appendix C
Random Noise Excitation Data
from
Accelerometer, Laser, Piezoceramics and Piezofilms
Material: Aluminum
157
ACCELEROMETER DATA
158
II
0
\
EU
a \N UJ
>> 3
N ?a J
aa <
0) >
U
< UJ
n n
II
JCJC
a
4
2^
(OO
mm u
2 |
c
rjrj z '5 CL
S"
(0 3 M "ft
CM U.
X _I_J
 -ft1
<44
ON
IT)
u CL fc.
< <
QC J N acoc a .1)
Ul QC I l-l 1-1 3^J ^4

> UJ m aa x
< > u.
OO N N
X E II .
8 i ~
3 O LL mmujo
tr O < ow
H -*
u
LUZ
oo
O a
a
+ + z
s
x
i&
CL3 -PJ 2U
oiocm <Z
t NUlrt a
JUJ< IU1Z > a
<UJ .z O01Z E>ZQC
3QCIrt o<< DO<UJ
DU.UJ -mi mmocH
UJ
X
<
H a X
Z .ui . m<
UJ u ZQCU QCQ.
2 . Z <LlZ a
I-
UJQC - a. -> J
QCUJ -U OJQCH <<
3U>< U1I .
.QCrt
0)U<QC ? HU <chuju
<- JLU LUZrt . .ZLU
LUQCUJ> QCLUUJ IIllitL

x\~a< LLUS CJUUOl
E
a> e
O N"vt
(C ID
Nmm
I ! .
.mtc
> .o
m--i
. 4
m i S>
a s
Ll
LU -X> ^
q:x<<
*
a
"*. fc^ fc
5" <N
5
C **"
O
a; S S SO
fc.
!*.
c o
O
I
;
ft > -ft
*-
c Qc;
fc. ^ i^
?J 5 S
a Si
^ 5-
J
L _
r
4 .J
L
IC
Q a
EC
3
[-
CD
t N3
DNIQ
hID .
Qlflt
mtM^t
.t-. .
~i?
> .r*
.
mrj
m i 5^ -.
LL .
LU .X> ti -ft
>
o:x<< 1 -ft ii
t * ft
1 ^S <- v,
u
H
1^8 VD
3
H
Z
^ >?"A
*> Ai
^"
^
5 -ft
fc. O .^
^.
>u .
U _J
< X .5:
X
^
NO
cdiq
ruorj
rtQin
IO .
+ <
Q.: *
0) N
mm i
am a
0)Q^
?: .Ort
U1(0 .3
QJVHrH
LuU-i .-
o.
3
H
3) .LU_
*>XLf)
a
(fl 3>M
\
0 EU .
E \\ UJ
CO >> 3
OO J
rn DO <
. . >
11 T-*
CJ
LU
N N O
II
XX
(0(0
mm u
ft ;>
(0
CMN
Hr^-
3 $
*i*
ii
ft
ft
r^>
1-1
u. r* SO
X _1_J
s ^ fc.
< H rt rt I
U X LlLl o Si N
z ?-
UJ a 3 *
Q
Q. 3e ->-
S- = S -2
< 
QC m LU *4 3 54
LU rt l-l
^
> LU aa oc ^4 l
< > c\i U.

N N s; th
O to
II .
Ifc- X
3 LL N mmuio -Si
QC 0 a
<LU .Z iaz E>ZQC
3o:irt oo< QO<UJ
QLlUJ IDNI mrtQiH
LU
X
<
H a X
z .uj . .<
Ul u ZQCU XQ.
X . z <LlZ a
UJQC rt CL "- H_J
QCU1 -U 01QCH <<
CL 3UX LUI .
-acrt
3 0)U<QC ? HO cmuju
H <- JUJ LUZ>-i . .ZUJ
LU li!QCLU> QCU1LU II'iltL

01 ZhO< !LU3 OUL0O1
E E
QNNt
(Oil .
oiaoDi
oam
aa i
^
>da .
ruin 2 ^
cm i
3 ;
ll . ft ft .2
O >! -S
ui .x>
.*
Si '^
QCX<<
"*> t*l
J fc
_ 3
^ -5; o
t
of
1
8?
1"
3
-
2- 54 ^4
ft Si -ft
rt ?s -ft
3"
5
s -
10 w
i
; ,
\ ! f ~ 0 ^
o 'j
:'
: 1
j \ '
j ^ 0
E
EU
\S
.
UJ
3 1 V
L
H m
y i r^
>> 3
v "; |1
1
f 1 -
(0 o EE J
1 i-
in is
*'
!
'
v. i i> 1 QQ <
rH
r> oa >
0) o .
-
j ; "
' .- r-l
E OO . { f H
.
o e : s.
*
(J
.rj |
'
^""i^ . < UJ
^~*~P~ n n a
>(M
ew
. . *,
-
i 1 1
1
1
'
! r"
*
II
rt[S i ^ XX
V^
i (0(0
u.
UJ -XV
-
,
j j |
| r
5 .
s-
H
QCX<< ! a ft
-.,
i CJ a
\
-
I L^ ii ft
*
o ^"
1 fc
fc
-^
C^
>3 X .
i ! < H -. fc. vo
i i u X
1**'***^'
'
*
z 1 "i ^ |u
! >
S] fd
rt m UJ
3 ' SO
:
'
\ '
^
u a. 3
< < N
1 ! r*
V '
-! i QC I oca: ?
1 >L^
1
1 r .J
i **_ " UJ a: E i-i i-i iii

ii
1 > Ui m aa x ft . -ft
r
< > OJ u.
' !" s
"; o rH
N N
"
; 1 2 IS N II
j j jj d . .
"' 3 E U.I mmiua

! o: Q ow
i
""*
:'
ri
U i
i I | . .
!
i-
, (M ?-*
54
u
.
1
~^~
J aa rj o Li
< ! 1 ! ! -
(0 LUZ a t + z
X '
i ! '
1 , "
.
0-3 -IM u "
1 1 a oiocm
r-
I-, l', ' <-f z X
i r
:m S i
^v^ j t N i-i a
ma-i j I ! i J*" i
4
_1U1< 122 > a
3 (M L <UJ aaz E>ZQC
.;
^1 .z
o.a o i
. ^r
3irirt ao< oo<uj
oiorj . ~, ; '
i-5 i i r -
QtLUJ hNI mmacH

UJ: < -
1 ' >"
d UJ
QC +* . 1 i L X
s.
<
UJO B i 1 ?
"
H ? x
wrn E i 1 i1 1 Z M .UJ . .<
j(\j(\i 1 ^r i ! 1 "N ts UJ u ZQCO oca.

1 !
1
_*
Q.'rtt
fl .
->'
1
-
^q.
j
1 '
\o i-H
S
z
uioc
. z
1-1
CU.Z
0. -
a
HJ
/^
Zrt^. qcuj -u 01QCH <<

- '
1
rt.rt X '
.
L 3UX UJI . .QCrt
; t 3 01U<QC ? HU <tnuiu
H <i- JUJ WZ"
-
a . .ZUJ
3) . .
1 LU UI(Ui> QCLULU IIUJQ.
; 1 ^^ ?*
*>X "i i 1 i 01 2HD< LUJ UUU01
a ? a a a ? a a
Q at m rs io in m
LASER DATA
165
*
ID 10
\\
E E
ti \\ UJ
>> 3
H EM J
(L a . <
LU CD 0(0 >
U rt 0)
U U
< LU
N N D
? II
H XX -S
3 toco
<
mm U
(M(M z
a 3
PJ HH Ll
ft fc;
t"^
t;
-**
X _J_J
,X 54 v, |^
< X
U
^ -ft 5 ^ r
n
t t
UJ
VO
u Q- 3 - -
< < ^ ^
QC J N QCQC ?
s STifSP
LU QC I i-i t-t LU 5 ^ ^
> UJ m aa oc
< > LL
fc 54 -5
(BO N N 54 s*
E II -
3 a u. mmuio
QC a < oit
-I*
u do! a
UJZ ,a: a + + z
0.3 4mj zu u
oiocm <z Z X
t NOrt rt> Q
o
JUJ< IU1Z z e a
m a zoc
<LU .Z aoiz za
3 Nm 3QCI- ? << <oxui
hQI
QLlU-I fCOI I^fTf QCH
ioo . UJ
Q< X
CL *
<
0) ? <t
:m X
H
UJCD +
QC3 *
z .ui .<
LU U :i
zqci: XX
ts NO)
2 - Z IU
<U-U( a
? .Irt
>-i H_J
.UJQC
UJ"f03 QCLU _U 01QC1 -H <<

qccm .
DUX uii::i
a. .QCrt
u. i an. 01U<(Z ? h-ULKHl'JJU

3
3 H <i-_JUJ UlZrtrt . .ZUJ
H LUQCUJ> QC UJ UJ UJ 1 1 LU 0.
rt
LU
. .Ul
0) ZHQ< [LU53UUUID
>X01
CO
Tl
rt
00
!-- 3
m no
i i .
NO<*
I . I
.00)
>-
CDrt
.0) . 1
m i
LL .
LU -X>
QCX<< P^ >5 ^
ft fcC 5
y ii
'^ "5
*"*
>"
r \
_ sj \J
VO
3
a.
z
1 tsO
>r 54
fc- 54 -^
54 cy ^5
o z 54
< rt
X _1
^ a
'
; a :m .
Nm ,3
'rtjQX 1
TOO . ,
: a<
;o.i t*
(fl!
uiirn +
QC3 * .
^ NO)
1
? '
.Irt
uj'mas J
ac:r\i .
"*
i]i ao. *
'
H
- -
. .UJ
SVX01
a:
C If) 3
"*.
x:-
-:
ft 54
On
lw
s ? 54 00
O 54 54gj VO
3-V^
^J
v3 fc -5
r 54
.fc.
O
IIO
r.(\i
TJ
<
UJ *
r; (J
ii z _,
LU _J
-
1 v>
QC
UJ :i :.~i -
.1
Ii 1 . _--
71 U
(J- - r j n.
!*.-,. L'l
I I II
a . : e .,
E .
71
*'
.-.
3 .
'*
m .(0
. r-. fl]
i-
L)
r-i lJ
c:
-t-
:x
tc tn
LJ
,i;n
ft c
g-
c TjTJ ft ft
rj M M
*
5 i ft
m t r-
.1.
.-fc
ft .54
a 05
fc.
1.1 54
u. u
ON
VO
(J r_ ;;
X
t
QC
*
fc arm a ix
U.
'.-
X CT) v-. i-h n; = s ^ s;
I'. , :: 3 n; 1 54 ft:
< ..
,
fc. $ -ft
3 m : ; i
.c a. j. , .,
5" s
3 !.. Ll m mu o cs
rt o < . if-.
1-
E O M
^,
li aa r r;
L*
x'z a j r
. ^
ru3 .rj ,,
v
U)XX
* i .<
Cl
r^
3 LU < '.'> Z1
L*f>'
".- !:
*. ui . :: in::
Li--..-
3QCl-> loo -: r,
r-
c ll u _j rj : Z IrtlKl-
.J
JV
3
**>
m .'j. - *
'jl U 'ZX'.Z kcl

i. - <-.,; 3
luo: -' a. - (-_;
ocu; -ti inaci -
< <
DUX !m ZJ . .Krt
01 u <x ? HO < O UJ U
'
N .
o-rn OrtfJ
. . m i
3
>mco . >-m<o .
mm mm
CM 1 a 1M 1
10
CJ
Ll .
Ul -x> UJ -XV
CkCX<<] QCX<< CS .54
ft
**
Co *
o ft]
n
in rS "ft
CJ
54 :
^
t-
5 O
r-
3 "5 S" Si 52
54 fc
CL
Z 5 '" f&P
'
S 54
r>
54 eg
^^
-ft
-^
s; 5y ft:
J 2
5j
0
<
2 N .
OO
nao
<?
IIO+ .
<
CL I*
0)
LUt
X"
io
oiaai
? .
.r-
uimas
:-ai
JlJI rtO.
>-
5 X 01
a
CJ
I I
11
3 E
UJ
>>
CC OCJ
a .
. (C
-<
o
H
!IW
toco
mm CL
cm cm 01
3 UJ ft
O tie
2 ii
ft .54
.55
CO
3 i-; -ft fc. <N
LU
QC
.* n s^
5 -^
c Ci c a 54
< 
a
ft-
3 m m uj o C s
fc
QC <? ow
H >-*
U ao rj
luz a f x
(L3 .-* C3 "
oiaca 2
t >a
_'UJ<
<1L'
.Z adz DZQC
3QCIrt ca< >o<lu
CU.33 OO.TCH
.U
<
. <
Ui U ZQCU QCC.
-z . z <LlZ a
.u,ac
>-i CL -i H_J
qclu -u IflQCH <<
3UX LUI . -QCrt
tfltr<ac ana < cn uj a

<rt_>LU UlZrt
LU QC 3 > a: lulu ii Luc!
2r-C< Ll3jC u a u oi
PIEZOCERAMIC DATA
in
Ll -
>* ^
-5
<ft .54
1s u
>.-ft"
?! til
ft; ^ -v ft
i
a ^ &
54 J5
54 oi -ft
O -
rs
?4 --
54
fc- 54
r a
I- B
r "
f-l
[in
m E
r-- N3 O N E
LO NIOl > Nim
1 IN . oiwrj
Mor-- -
a MOrt
mow in
.
mm i
.? . -* .
> .m . > .m .
r-(M f-C\J
m i m i
r m
u. .
i ^ Ll .
UJ -x> UJ -X>
QCX<< QCX<< ^ c
ft
Si rS
? ^ >
-~1
i ! >- a
i i
H ! 54 H-
J_ 4.. m 54 54
3 .
>J "! fc- fc. ^h
L ~
Q. : i Si fc
i 5 -S Be
I- in L>
r3 w
1 54
i ! t 1
S.
Ct
^
>.
54
-ft
's
1 ; r
'4J
I a -ft
| i
i -
;
a
ci
i-
a
m
1
-
CO 3
< ^ 54
X
r"
NO NO
-
tn I-" IO
DIC 3
IM w
30W otM
'<flll! m
13 .
T < + <
a. * UJ *
01 N -
o U N
uicai CO Z I
0C3O uj a
QC OlN
moN
UJOOrt
? .O-l
LUrt .3
IO .3
acuirt a .rt
mi! -'0.
(JrtrtQ.
3 3
H "* H
rt
01 _ .UJ
. .Ul
.<> X 01 !*>X0)
io
d d d d
IP
""!"
I
N H
Q) rt
XZ
0 E3
E XX UJ
CO >> 3
E E J
ui OO <
aa >
r-
u
UJ
n n a
II
JCJC
com
v3 fc
mm u ft -ft ft.
at (MCM Z J-; J ft
to . . 3 '-> fc r>
HH li
X _l_l
fc >*, fc. <o
< MH X *
a
z
X U.U. u
t-H
- in ST)
UJ
i-i
a a. N ^ ft. 3 a
 UJ ?a qc | 54
< > (\j LL Ifc-

-3
=
-ft
o to N N -ft
2 II . U?
3 U. N mmuja
QC 0 a
<UI .z iaz EZQC
3QCIrt oo< >a<ui
OU.U3 IflNI ^mocH
UJ
2
<
? QC
z .UJ . .<
UJ u ZQCU QCQ.
2 - Z <u.z O
U1QC - 0- rt HJ
QCU1 -U 01QCH <<
3UX UJI .
-acrt
01U<QC ? HU <muju
<rt_|Ul UJZrt . .ZUJ
UIQCUJ> QCU1UJ IIUJCL

0) . .
UUUO)
JhO< U-US
>-X
I i
E E
O N NW
N oil .
:
r oioow
OOrt
OO I
>mm .
mm
c\j 1
| 5
uj .x>
o ocx<< .* .54 si
CJ
54
vo
S-5
vo
.r-
* fc
w <;
i
ft. A t
3> >r
5 54 54
2 ft; -S
G>-S
54
N
i
aa
? a
rtt\i
+ .
<
LU N*
U I
4r_ a
LU a
QC ON
LUO .rt
1003
a .o
UrtrtQ.
3
"* H
rt _ .UJ
3>X01
r-
Bm t-l
XI
0 E3 .
E \N UJ
>> 3
a E 5 3
ac <
oa >
u
< UJ
n n a
ii
XX
(0(0
in in u
(\J(M z M
o .
(9 . . 3
si ft
O l-H U. ^
ft
x 33
ft; 3-^
.S
>J
< i-i i-i
x -ft fc. ^
a U.U. u
z
1-1
UJ
8 r-
fc
u o.
< <
^ C Q ?4 r-
N
QC I
5f" "*
S
QCQC ? 1
54 "^ bo
UJ QC E rt i-i UJ "O
>
<
UJ
>
m aa QC ^-
w U.
?-
o h
N N cs ft "ft
2 (9 . N II .
3^.
54
3 E U.I
o: a <to
m
mmujo
tn-t
-<*
o
N
M
u aa w
' 54
a
wz a t +z
0.3 -w u "
01QCO z 2
t N >a
JUJ< I2Z eo
<w .z aaz >azor
3QCI- ao< oa<uj
QIlUJ rtNI miooci-
UJ
2
<
r-
? QC
Z .UJ . .<
UJ U ZQCU oca.
2 . z <U.Z o
LUttC Q. - H_l
QCLU .U 01QCH <<
3UV< UJI . .QCrt
0)U<QC Ol-U <muju

<rt_JUJ UJZrt . .ZUJ
U1QCUJ> QCUJUJ IIUJCL

2HO< U.U* UUU01
PIEZOFILM DATA
178
cu
o
\
EZ
tt \\ UJ
a >> 3
E^r J
o . <
0) >
U
UJ
II
N N ?
1
XX
10(0
5 ?
iiivi u
0
ww
. .
z
3
*^, 5
"^ ft.
.^
IP
w r-\-
U.
33
i-i i-i
x
U.U. u
1-4
UJ Ov
1
O5
:* 5>
N QCQC ? s ^
I rtM
LU I
m ?a qc
u.
N N 5: 54 -s:
II .
LL mmujo
< oin 54 5y
>->*
"** 54
a
a
? + + z ft-
ZO "
<z 2
a
iujz a
O01Z > ZQC
o<< mxuj
<*mi rtô:n
UJ
2
<
H ? QC
Z .UJ . .<
UJ u zocu oca.
2 . Z <u.z a
UJQC >-> Q. rt Hj
QCU1 .U 01OCH <<
3U>-< UJI . -ttCrt
0)U<tr QHU <ouju

<i-.jijj ujzh-i . .ZUJ
UJ UJ QC Ul > QCLULU IIUJQ.

ILUS uuuoi
ft >:. O
fc.
fc-
ft :ft
-ft
.i,
jJ
vj
r*
Si r
^ "S -ft On
O
1Q 2!
54 fc
00
fc
^ !" 5)
1
I?
ftC 54
~~
^
ft: -ft
ft s "S
Si 5 'fc
ft
ft s
54
xma
HON
?itOrt uiaort
ui. . .a
IO .3
ocimw a .w
Ort wQ.
iLil rtQ.
3 3
H t H
rt
0) . .UJ . .UJ
3>X0) S>X0)
n
o
\
EZ
E XX UJ
(0 >> 3
a* 3
m a . <
>
H U
< UJ
n n a
II
XX
men
mm u
a ww z ^ -- *^
(0 . . 3 T- O fc|
HH li
X 33 r3 -fc >0 _
< H *4 LC, ^ C>

1-11-1 x
U 2 lilL U
z i-i
UJ
u Or N
3 O Q 54 00
<
QC
UJ
<
3
QC
I
E
m
QCQC
rt rt
?
LjJ
I HI
'
> UJ aa x ^ -ft ^
< > w u.
O (0 N N
2 II .
3
QC CD
li N
a
3LU< N2Z EQ
<UJ .Z IOZ azo:
3QCIrt oo< >a<ui
nixuj wni (D^QCH
UJ
2
<
H ? QC
Z .UJ . .<
UJ u ZQCU QCQ.
2 . Z <U.Z O
ujq: m Q_ rt HJ
QCLU .U 01QCH <<
3UX 111 I . .QCrt
0)U<QC ? HU <muiu
<HJU UJZ-< . .ZUJ
0) . . UJQCUJ> QCLULU i iiu a.

SXX 2HD< U.UC* uuuoi
"
r ....,,
1
|'| r
,
1 ! J
I
i
*
1- a ! i j
4 1 L ! :
1 '
1 w i
o i !
[
,
3
m
4
4 i Ii I- i
1
' 1 :
CD 4
( 1 I j- a i 1 .
a E E
hNNTJ / r n j i
t-
(D N NO
tilt i ;
/ ' N w aiiir^
1 oo .
DON j
i
/
/
i
i
r
r
i
iii
i i : ooo .
oat
OOW 4 I oo I
i i
.
i i ^8 ]
g
mm . i i
:
r
(D
n ! i i : cj mm
-
W 1 t
i w i
i I
j ! i 5 "5 ft
u. . i i : U. .
UJ .x> ' ! 1 UJ -x>
QCX<< a ; -
a ! O KX<<
4 j in
CJ .*
^ r3
i ! I
i i i '
*
'
f i ; : S >i fc ^
/ L
3 1 \ rs S 54 ^ 00
0. t I / f N I ' [ >"
fe fc-
z
t-(
-i
-
!
1
4
\ 1
i \\ 1 I J
z
-
i j i- o
I
1 a
2
i 1 (M
t-H
4 L
1 54 ft
3 L
\ | \
i
5 *
3
<
*
-1 i - a
1
j
1 '
a
2 N !
\ CO i l CJ N
1 ! CJ 1
i
CJ
I
H OO
] 1 OO
!
1
moo 1 aa
i 3rt(\j rt(\j
rtQ ?
j
i
!
-
I.O+ .
i + .
4
** "
 i
a t
H i 1 i r a a
0)
SVX0)
. .Ui " !
i i
11 1 r
a
rH
i , . . .
^
S>XO)
.UJ
o c
. a a o CO CJ
<
CO <c a]
; i d d d
....
1! ' '
>
J
iZ_ L O
' i
*
0 ^
0
" X
3 " 0 EZ .
E XX UJ
i^ >> 3
m .
O 4
O 3 1 ^ a
w i> j rH
IB 1 1
E OO
H
WEI .
<
1
. .m
-
vm t
rt o
. .
i i '
rH
*
IL
Ul .x>
.
i mm
WW
u
z
Q
5:
-2
*-
CL
^
QCX<< J
1 i
ftl 0 . . 3 >
H ft
vj ft
54
^ . o r-\-
li ^
i X. i x JJ
r
i .
ft) ft, ^ CN,
'. j
H i-h rt
x
!
i l
i ^s. I f U 2 U.U. u
rt -2 s: <S^
rn
UJ
i O oo
i 1 N
i i
I > if QC 3 I QCQC ?
1*^. I UJ QC E WW UJ
> UJ
41 ,
L
< >
m
w
aa qc
u. 5
11 rH
N N
1 1 5= In 2 8 . N ii .
4 i [5- -J E U.I mmujo

r 3^
I QC O <o wt
J
1 H m rt*
2

i
3UJ<
t N "< a
s
AT) ?,
W 4
i f
I <U1 .z
I2Z
aoz >
a
zoc
10.0 0 ( 3 1 I, OQCIrt oo< OXUJ
mow
IUJ'
. i L-Â
ir*
1 Tk . DllUJ rtNl rttQCH
< 1 a UJ
'QCI + * i 1 c 2
L * y r . <
.i i
,UJ! 0 H ? QC
wm e 1; i Z .UJ . .<
:_irn(M 4 i ftl ts UJ U ZQCU QCQ.

Jl
"^
o;t .
j ! d
a
^ 2 . Z <u.z a
i
|Q-I .t J ! * UJttC - Q. - HJ
1 QCUJ -U oqch <<
ta i i CL 3UX UJI
i|/
i/^ 3 0)U<QC ? HU
.
<OUIU
.QCrt
J a H <rtjuj UJZ-i
0) . .
-i *^ =
- i UJ UJQCUJ> QCUJUJ
. .ZUJ
IIU1Q.
I 0) 2HO<
SXX I U.U* UUU01
cj o o o a o o a
c-
'
j ai * IT c a
i i i 1 i i
Appendix D
Impact Excitation Data
from
Accelerometer
Material: Laminated Composite Graphite Epoxy
184
fc
8
ft.
"vS
fc
SC5
00
54
ft S-SP
a. as
54
5 54
!
-Si
3
I
54 ft
-ft
54
O
L 54
ft:
c
54
.2!
i
54
54
I
ft"
O 54 fc
6ra
<x
Ii *
VO
oo
'
1 54 a
"ft
-
-fc Sc
54
5
S-
??> v.
A
CP^
54
54
Is
!? S
5
D 54
1
Appendix E
Identification of Artificial Damping

by the ShakerTable
in the Vicinity of 237 Hz.
for Aluminum
(Gain =
3)
Sensor: Piezofilm
187
~! I
11 II TT
ii./;
ir
.
a
a
1
i
!
;
! ;
/
/ !i '
H
W rt
; I /i z
za -
m
r
j
3
'
i
i J /;
J/ i
a
a
II
3
XN
>>
LU
a
i ^r-T
O aa _i
CD i 1
o aa <
(M TJ ! f
^^ 1 ir t . . >
i a I 1'
1 H U
n a < Ul
i
1
-I
>r-
-a
n n a
n . a
II
mi - o
i XX
WO 4
i (DO
u. . i1 I
'
iriiri u 51
LU 1 1
-Xli
i> fc
-X>
4 1 / (UN Z
KX< 1 i < 
<
UJ
>
N
I
aa x
U.
I i
1 / a H
N N
1 2 E
3 J g
'
!
. IX .
1
3 a u. mmujo
t
3
i 1 cr a < wt
: 1 o H -*
1 G
U
<
J 1
>
ra u
uiz
aa
a a
a
t +z
2 Z i 1
Q.3 -w ZU H
it oiocm <z 2 Si *,
JO 4 i i ? NIDrt a
m a , iI i 3
3UJ< IUJZ a
TJ w 1 <UJ .Z awz > ZQC
rtQ
j L 1
-
a OQCIrt o<< OXUJ
,
a
IO N . ;
'
dl QH.UJ OOI rtCMQCH
i< ! I UJ
o. a* 4 '
2
o a ' ~~
M m <
1 1
LU O "*""
H a QC
j^-"
oca J ;i Z . .UJ . .<
"0 + y\ Ul u ZQCU oca.

j
llo J^L
2
am
UI . N
T
4
/ | . i n 2 .
U1QC
Z
<h
<u.z
a.
a
hj
QCOI i ill QCU1 -O OQCH <<
LLrtO 4 .-^
ir q-i aux UII . .KM
i
I :ai 0)U<ttC C3HU <auiu
ts 3 <wJUJ UIZ-
n . .ZUJ
J U1QCUJ> QCUJU1 IIU10.

lH
s>x
m m
j
i
*
-J-
i
'. 2hQ< U-U uuuo
aaaooaaa
. * pJ 0i *r u
1
m
O E
3 a
O o
m ^
o
fj
;m
54
-N-
>-xw
oo
m i
.
.
>I
N
o
m
1
54
54 fc
-,
z ft
u. M is &5
.
fc 54
LU .x>
QCX<<
< .
2X
"z '$ S
5
-ft
U
th
e 3 s*. E.
=
H
of
3
CL se
Z
%^Tk&
i -ft
1*
e <hu
^> q as
* -&>
(J 54
< a 2
2 Z !> o
2>
JQ
fe
-^
JO ft.
o a a
3 (M w ^
rtQ
IO N . N .
I< I<
o. a* a*
tn a
a
ui o a
o:o
34- QJ + CD
? rt r4
a
LU . ,-i L* a N*
.1
QCrtJ
rtOQ.
U.WOQ.
3
3
C H
H
. .UJ
rt . .Ul
r-X0)
BP-XO)
a
CM
E
m
^ N
0)1
m
r--
.m
>-
r-
.
I
z
m 5..
i-i
ft O <
< . C
;ft
ft Si
D -ft
2X , fc Sj
s
*> .a
s t:
S1
fc 11
^ s. -fc 54 o
as
fc -s; be a
S 5 ft So
&u A 5;
a .54
in
3* I
G^-ft 54
z
rt
a
*
J
ft
NO
IO
a aw
ra a
rt .
<
UJ t*
u
N z N
:toi (M LU I
"OOO) QC aa
lOOrt UJOOrt
'
.as I003
ocln . a i
luwto. L)Hd
U 3
a
[^ h m H
rt
_ U|
w . .Ul
*>xu> J>XUI
-
i [ H
i l i "
z
:
0 a Z3 .
"D 1 j [ i.
E
tH
\\
>>
UI
a
CM O
. "D
f i '
t OD ?a 3
4
in i> ! l a i aa <
Is* >
a a
! f-
- . .
.
. >H rt
(Mrt(S .
r H u
am i i
-ta ! r < Ul
>-aiN n n a
.
ii r
("
. II
XX
[
i rt (0(0
1
U. .
LU -X> H ! L uiiri u
(MN Z
O.X<< ; l a ft
. - 3 & G
[" o HH U.
X m 33 a &Q
< H i-ii-i I
U 2 '
Z
U.U. u
1-4
S>
gj
rt
u a.
Ul
*
I "fc1 "
-
a Os
3
**
<
a:
UI
<
3
0.
N
I
E
oca:
i-i i-i
a
UJ
54
Oi,
>
54
g*
54
I
L -S > UJ m aa qc 54 -S O
! < > w u.
'
a .*
N N
t^l
;:1 2 8 s ii .
a
qc
e
a
U. N
h a
d a <
* JUJ< I2Z a
v w N <U1 .z aaz > ZQC
CL ? a *
ao;irt oo< OXU1
ID DIM . sl au.uj rtNI rt(\io:H
Ul < * - *
Ul
X + * 4
| d 2
<
Ul u H a QC
U) fflE
3 TJrt
1 1 z .UJ . .<
1 1 ' a ui a ZQCU oca.

3 ID . J (M 2 . Z <u.z a
0. .10 J j i 3 UJ0. - a. m t-j
2 inm a qcuj .a WQCH
i < QCUIUJ IZU1Q.

s>x i i A \ SI 2HO< U.U* UUU0)
oooaaoaa
CM ~ i ai * If
i i i i I
rr i,,,a i
i
OD
: l
i
!
i
1
N
i
!
1
1/
i
1. UB
f
j
TJ
o o
1 '
/
/
t
'
N NTJ 1
rtHO)
.
.
'
1
'
[o
I OO
oats
.
i ! i 1"
oaw i ; /
'i r
.
i i / i
I
i
>tNrt .
mm
W 1
1 i
ii
'
lo
r
tN
i i
lL . > / ii
1O *
!
/
.
LU -XV
QCX<
54
-fc
1 :
/ .fa
,54
^
fc
,a
Sj
-5
t*
' /! (
1 1 -fc
h \
3
0-
! ! / ! :
"]
2 / i ! rs 5i 54 f I2
4
rt ^ g"
\ 1 i i
' J-*
a: ^ I
Z
-< \\ ' i1 -
a
m 5k -fc 54
3 \ i ra
\ i i
\\ 1 fc S
U i
i T
<
2 N .
\ -
fc-
_ I '
\ 1 Al
Aaa i
i \
"
(M
ooo i
TJrt(\| .
rtD .
ID+ .
1 1 |
< 1 -
a
3-.
* '
ftUl O
!j 3 '
1
1
(M
ocna i
'
l :
<
rcaa i ' *
an
I
.-
, 1 1 1 -
a
Ul .o j 1 1 a
Qcno 1 ' CM
ERrtQ. i I j
a 1 \ i
N H 4
-v. .ui j
a
1
-
s>xcn a
rH
a a a a
(M nJ *r
i 1
z
0 Z3 .
E XX Ul
-H
>> a
a aa j
aa <
. . >
H u
< Ul
N N a
II
JCJC
(DO
iriiri u
WW z
a & ? G
o hh u.
X 33 ,a
03 -
< H Mrt I ^ . <=>

u U.U. u
z 1-4
I-. Ul m
u Q. -fc
Os
< < N fc, fc
QC 3 I KttC ?
UJ
>
<
QC
UJ
>
E
m
w
-i-i
aa
uj
qc
u. %
54
r
2 9
a r^
. N
N N
II .
I
3 E lil mmuja
x o <a wt
h . m
u ao w a
uiz a + tz
0.3 -w u
0)QCO z 2
N - a
3UI< I2Z a
<U1 .Z aoz > ZQC
aocirt oo< axui
DU.U3 rtNI rtWOCH
ui
2
<
a QC
Z .uj . .<
Ul u ZQCU QCO.
2 . Z <ILZ a
U1QC - a. w HJ
Q.UI -U OIQCH <<
3UX UIZ . -QCrt
0)U<QC QHU <muiu

<rtJUJ UJZ-i . .zui
uiirui> acuiui IIU1Q.

2HO< U.U3 UUUO)
T2
a
a
ID
TJ
CD O E
N N3
.
t H
QUO)
'
M
mm .
am
N(MO
mom m
M
>d6 .
vol
mm m
o o
z
fc fc: Si
< . .fc VJ -fc
2X K fc ^
TT ,
-5
ID
^8 ii ^
^54 54
54 C -fc 2. Os
a s^ & s>
fc O fc
t o
to
fSf
z 54
.O-fc
rt
ST
N
I
aa
aa
rt(\i
f .
<
N*
I
a
a
qc aa
uia .^
I- a ico
rj a .rt
luwoo. UrttDIL
3
m h
rt
*>X0)
. .UJ
jt
a
L
a o
lla
W . .UJ
*>X01
rJ d a a
41
i ;
ll
"
>:
Z
ll 0 Z3
ID 4| E XX UJ
3 CD 4'
-*
>> a
H 13 a ?a 3
4!
1^ ao <
0) 0 0 .
.1 >
E EO
IMO 1
a t
i!
m
u
UJ
>I00) . fi N N a
oo ii ii
tIM 4l
XX
(OO
LL . 1 t<5
UJ 4 mm u
.x>
WW z fci ^ Si
EX 4 :
0 3
O hh li
X 33
 Ul m aa QC -J
< > w lL 8
2 0
a H
.N
N N
II .
-fc
a ^s i
54
3 E U.I mmuia
QC o <a wt
H . o
u aa o a
uiz a + +z
0.3 -W u "
01QCO Z 2
? N ii a
3UI< I2Z a
<UJ .Z ooz > zoe
aocirt oa< axui
au.U3 rtNI rtwacH
ui
2
<
H a QC
Z .ui . .<
UJ u ZQCU QCQl
2 . z <U.Z o
U1QC 0. rt H3
QCU1 .U 0)QCH <<
3UX UJI . -QC-
0)U<QC ? HU <muiu
<rtJUI UIZ" . .ZUJ
UJQCUJ> QCUIUI IIUIO,

2na< ILU3 UUUO)
Z
Z3 .
a NX UJ
a >> 3
o OO J
o ao <
t . . >
rH .-I
H u CrJ
< Ul 54
n n
II
XX
a
!
,8.
(OO
2 54
mm (J
WW z
3
S L>
\-r-
u.
S
- -fc
-1-1
MM I
U.U. u
Ul VO
Os
I
efc -fc
&4
oca: a
MM Ul
54
N aa tt
I ii
N N 54 5 fc
II .
li mmuia
< oit
?-*
a
a T +Z
zu m
<z 2 .ft
NO" a
IUIZ a
OtflZ > zotc
o<< axui
omi rt(\|KH
ui
2
<
H a QC
Z .ui . .<
UI u ZttCU oca.
z <u.z a
Zj
uior m a. HJ
qcuj -u W0.H <<
3UX UIX .
-QCrt
0)U<QC OHO <muiu

<mJUI UJZ-i . .ZUJ
_ UIQCU1> QCUJU1 IIU1Q.

2na< U.U3 UUUO)
ual
'
' '
i
o
(M
t
t
It- B
It (M
il
E
't m
^ N
t
In N
OI
m
.t
> . a
(M (M m
l IM (M .-i 5
z "3
h
< . B $ .8.
u
O TX ** <s<
(M u
54 fc
"1
<fc>
1
u ^ ^
1
^ r-
54 fc 54 OS
-
o I
54 V
Re
i
54
z 05
i-i
3 fc-
54
..
(M p (M
a
NO
IW
t- o m
w .
<
UJ + *
u
z N
UJ I
QC oa
uiaort
iaas
a
a . .
u-ita.
a
m h
W . .Ul
SVXU1
.-U-
a o o t
o
i d
'
f
1
tn
3 o
i:
i
E
m
Z3
XX
>> 3
.
UJ
oa j
a ti r ao <
in i
m . . >
ID
a ao H u
oa i Ul
.rt*
N N a
>wo .
II
JCJC
int DO
fc "1
IL .
mm u
Ul -x> a NW Z
S fc,^
1
ocx<< w
m r-i-
3
U.
si g
X -1-1
< MM I
u 2 U.U. u
z rH
rt Ul it11 ^ 00
u 0. N * * OS
< LU aa x TT *: 54
< > u.
O m n N N
2 ii
^
.1 .
3 u.in mmujo
ra
QC < . wt -5 Q
H E o h*
<
2
$
,.
U
U1Z
OO
0.3 -W
O
u
o
+ +z
fi wxm z
a t M
ID a K 3U1< N2Z
N
TJ w > <UI .z IOZ > zoc
Q.Q aocirt mo< axui
? U.UJ WNI rtCMQCH
m o< Ul
QC *
+
UJ H a
mo a Z .ui . .<
3TJE O Ul u ZQCU oca.

300 2 z
.. <u.z a
a. .t UJQC n. m H_l
20rt /- -f Kill -u tnocH <<
-rt| 1 1 OOX UJI . -QCrt
0)U<QC ? HU <muiu
svx
l 1 !
IflJ
<mJUJ
U10CUJ>
2HO<
UJZm
QCU1UJ
U.U
. .ZUJ
IIUIQ.
UUUIfl
a a a a a a
w4
(U ra t in eg
1 i i i i
-
1
i
j
1
14 i
3
m 1
TJ
rt O E
. NT) , W
NNIW '
j /1 O N
Nllf) . M
ijkmo If ! l
I l m
MCt .
.m
>-
. l -m
n
.o - ! > .
t(M t
O 1 o
in ' z
U. . a l-l fc 54
i
Ul -X> 1 < .
5
VJ
0CX<< 1
I ;
IX
i ^
H
rS
a i "*I
a
! t*5
1-
'. 1
Os
Os
3
0.
J
I
Z fc 54
M
-9
in
a> hii-=i i I ;
2 -S
CL.O 54
i 54
_
5 -5
'
Z z 3
n
w
fc
M a 54
U J J J
<
^ \!
NO
* j NO
.2.
in za
bio -
;
'
TIOW 4 1 aw
r*
i in
I 8". i *<
+ < *
\ *
0. * i \ UI
U
in
Ul
N -
\ o
ts Z
N
1
\
x Ul ?
non a: aa
a llQrt 3
UJOOrt
za
UJ
x
* 4 \ i
a f*
.*
S,< J I i in
E NfflQ. 1* i t
a UrtOQ.
a
3
N I m h
H 4
\ w . .Ul
..Ul
>-xw
J
. i
O
X
O
uJ
O
.
i ri
O
t i i *
a
i i ^ *
t
i ' * *
N
' ' ' *
S>XU)
N CM t I
I I o d a 6
"I"
i
i
! i ! 1 \\
i
t I1 ! !t
i
1 ! i
1
1
,
j
\: t
f
.. z
H
rt fl
f- in
0 Z3 .
0 Ul
o
3
1
fni E XX
o >> 3
a o i
r aa 3
TJ i in oa <
f | .
H . . >
-"
0 BO
4
1 / ! W l-l*^
(J
OO 1 -
! / < LU
WO
>-ma .
J
| ! ifra n n
II
a
.
c\jrt
.
1 i ! XX
:
1:
jr
1 i
oo
fc "3
U. . in'iri u 3 -
a
UI .x>
QCX<<
* / j 1 '
'
.- IB
(0
WW z
3 --S
. . fc
1 iH r-h-
U. fc -fc
|i r N .54
-jj / i 1 1 X m 33
/ i i :
'
< H rtM X
1 ! | ] I a 2 lili u "-*i
h /
! | z l-l
li ^ o
D
i
!
i !! i
!
i if
1 <
_
-
u
.
a. N
UJ
* lt "
54 O
0. / -u
< l
!
(M QC
Ul
J
QC
E
m
QCQC
HH
?
UJ
54 60
/ | ,
r -K >
<
LU
> w
?a qc
u.
O o N N a
S
1 2 II .
i^S 8
3 li N mmuia
u
/
fl
1 1
|t- X
H
u
0
E
ao
 <UJ .z
3QCIrt
IOZ > za:
Q. ? 4 I j oo< OXUI
U) D / 1 [\ sj QU.U3 mNZ rtWQCH
1|
.
/
Ul < i Ul
i i
QC t* 3
2
1 1
m <
LU t / j H a QC
tn m Z .Ul . .<
1 1 UJ u ZQCU oca.
3 3d 1 ; ; '! a
3m e / i
! 2 . Z <U-Z a
a. .a ' t * uitr - a. - H3
1 i
2 Ms ' acui -u OOCH <-_iUJ UIZ"-" . .ZUJ
1 i
1 ! UJQCUJ> QCU1U1 IIUJQ.
x "j I 1
-
Hi 2Ha< U.U* UUUO)
aaaaooao
pt 4 " ' PJ 0i -r ui
1
i i
1
i
| i
a
i
4 i a
i (M
i
0
3
j\ i
i
t o A i
E
.N NTJ
J Q
miia i / i DZ
/
'
i oa . w pj oia
initio / r m
m\ih .
/ 1
>-rs.o
min
W I
. 1
-i
r o
r m
>-fs
z
m
w ! *
U. . -I r c
UJ -x>
*
/
/
< .
o.x<< IX K
-
a
5 -fc
; / in ">
N ^
^
.1
tt, v
h g H ^ ^
3 % 5> o
a. CN
Lr t
I5
Z
w
-J N
g*"s5 54 l
H
i f Ai ^ ^
^5
.
l r
La a
5
z -
z
M i m m y * i
(M
3
r 54
fc
^"
i
u
< 4, ^
2
X
t-R
IK
a
N
(M
N
Z I
aa 1 aa
oaa ; , r aa
rt(M
| \ j
io+< ii i
i a a
LU N*
<
a. n* j \ IM
'
(M
U Z
in z , v i
z o
ui a
QCfflO
] i
1 Ul o
\ i <
QC ?a
tjoo
?ro .m
-
'
1
,
\ f
a LUO .rt
[a u iao*
ui ,o . V N
acta < a .a
,
iLlMrtQ. '. (JrtrtQ.

j i
N
3 ? \\ i
m
3
H
h -
1 ; < r n w . .Ul
A.
Eivxo)
.UJ .
| ; i
?
\ -
'2
O
s O
o t pj
a >XU)
o a a a
*t
ni ra
d a o a
1 r
i 1 j
i
H
4
I | -
o M rt
j I
1
z
1 i 0 23 .
m i E XX Ul
4 1
i
i k
3 o J rt
>> 3
1 i
0) Tl j r o OO 3
rs i aa <
in o o
I
1
1
1
| Ll
rs . . >
\
.
E EO ; 1 ~ rHr*
J
mw i ? r H <->
f
< Ul
VOID .
n n a
QUI
mrt u II
JCJC
\ oo
IL .
J
\ 1 i \
win u
1
1
s ?
Ul -x>
i
xx WW z 1
1 0 a 1I
. .
ii fc
!
t
S
j
1
LN o HH li I1 s>*
^
,
X -J-1
1I 1 L?
^ "1
"-a < H Mil X I *< *s ^
U 2 U.U. u 11 54 S ^
Z rt 1
j. -
w 1 It"
u a. Cs]
<
*
I o
t3 < N
11
I
Cs)
'
QC 3 Z ttCQC ? 8 ^ 3c
V Ul QC E MM U 1 0^' 54 Sf
'
> Ul m ?a qc 11 * 5i
< > w u. 1
a N N 1 a "S
2 0 N II 1 S
1
1
3
qc
E
a
.
U.I
<o
mmuio
.
cat
11
1
III
S 54
1J / H
u
.
oa
m
w
~*
?
1
1
1 j uiz a + +z 1
2
] k)
0.3 -W
oiocm
u
z 2
"
1
f |
~L i 6 N m a 1
m a 4
ii //
i
K 3U1< I2Z a 1
i j <U1 .Z
V aaz > ZQC
<lS.W
7 /
1
i 30.1m aa< axui 1
.t
s au.U3 rtNI rtlMQCH 1
K +* 1 | d
Ul
2 1
hi
UJ n 4l
^
/
/
\
\
< 1
fl H a
OlffiE 1 Z .Ul .
3TJrt 4
3m .
a Ul u ZQCU oca. I
4 1 V (VJ 2 . Z <IiZ o 1
i .m 3 UJQC
j m 0. -
2tm t 1
i d
dw i t
'T QCUJ -U OlttCH << 1
i
1 i r
f
o.
D
DUX
0)U<QC
UIZ
? HU
.
<muiu
-QCrt
I
^
J/ i
1 f
"
<MJUJ UlZ-i . .ZUJ I
i>X f i L 1 ffi
U1QCU1>
2HO<
QCUJUJ
U.U*
IIUJO.
CJUUO) I
oooooooo
PJ r*
(VI m t in
1 1 i i i
A
3
^ O -fc
2 3 "">
5 i< 'O
, -^
S. 54
^ !l c*5 m
r
Si 54 -fc
? o
^ ^ ^5 S>- fs|
5 S> So
5 -fc
fc. fc
S
54
fi)
Zy ^
fc 05
54
a
z
Z3 .
E XX Ul
rH
>> 3
O ?a 3
aa <
rs . . >
MM
H a
ui
n n a
zz
XX
oa
fc <*
mm u S 54
a
o
WW
HH
Z
3
U.
15 8""S
,54 -fc
X
< MM Z
U 2 U.U. u
z M
w Ul it" *5
U Q. o
< < N
X
8 CN
tt j z QCQC a 54 tr
o-
UJ QC E MM Ul 54 s:
> UJ m aa tt ~r ^ 54
< > w lL % ~
^
O N N "3 "S
2 I - N zz .
5,'s
3 E U.Z mmuia
QC a <(0 tot ?5 05
H . O >-*
u aa t a
uiz a t +z
0.3 -W u "
WQCO z 2
t N i-i a
3LU< I2Z a
<LU .Z aaz > zoc
aocz oo< oxui
QQ.UJ rtNI rt(\|KH
Ul
2
<
H ? X
z .Ul . .<
a ui u ZQ.U KQ.
ni a
"^
z z <u.z ?
uik - Q. - KJ
QCUJ -U ojoch <<
3UX UIZ . mtt-
0)U<QC ? HU <ouiu
<MJUI UIZ . .ZUI
U1QCUI> acuiui ZIUJQ.

*>x UflJ 2HO< U.U* UUUO)
-
a
i in
r m
E
W
r r^
f O) N
I
If)
-^ i-H
>o
1 fc-
' i
z
m
w *l
54
^ 65
L
i
-
a
rt
< .
2X
I
> m <N
N Bj 54
r
s * <*
^^ 'I 5> O
g ^s sr I
* I
Mi
*. J
-3Li
54
a
m
Nrt
i
a .
a<
UJ t*
u
z
Ul
ffi NO) QC NO)
?rairt uia Irt
za OS
o .
jjoido. Urt a.
3 3
rs h m H
. .uj
w . Ul
XO)
E
t
t
O N
z
w
>m I
o *5
54
-fc
<
2 5J
. K
2X
54
8 *">
fc
SI vo
^ 54 >
?> o
5 54
I)
z
w
fc 54
3 54
05
54
N
ZO
mo
.w
w
H
m
<
Ul + *
m U
z N
UJ Z*
oc aa
uiauirt
n iats3
a . .
UrtOQ.
3
m h
- N - -UJ
*>XU)
0
E xl ui
m >> 3
OO 3
w OO <
o ^
H u
< Ul
N N a
II
JCJC
OO >1
t
a
mm
ww
. .
u
z
3
1" 1
c
t-r-
o u. a:
X *->
-i-i
<N
< z VO
u 2 U-li u
z
rt
UJ 1o t*5
u 0. N S II
< < S fc o
Z ><
54 X
i
QC J
UJ QC
E
O
QCQC
rt rt
?
UI
?
LU
> iri
?? QC
U.
1 5)
O rt N N N a -fc o
2 -I Zl .
a
3
rr o
U.O
<o
mmujo
wt 1 '? i
s 54
H E -*
(J OO o
uiz m + +z
0.3 -W u m
UlQCffl N Z 2
? I - O
3UJ< OT2Z ?
<U1 .z .oz > ZQC
3EIH wo< OXUI
OU.UJ rtNZ rt(MKH
UJ
2
<
H ? QC
Z .UJ . .<
UJ u ZttCU oca.
2 . z <u.z o
uiac - 0. M H_J
QCUJ -O tnocH <<
3UX uiz . -QCrt
0)U<QC ? HU <OUJU
<mJUJ UIZ" . .ZUJ
UJ0CUJ> Q.U1U1 IZU10.

2HO< liU UUUO)
i m
pin
E
'a o
t N
OI
rs
m
j-
l
i.
in
-a
>
w
t
.
I
"^5
.
54
z S
l-l R ?
o 5 a
< . o -~ fc
lo 2X
L m
00
-in
Si11 8 o
cs
Z ~ 54
ih
3
-in
m
NO
I
ia
ON
- a m
: m
.
N I i
1
N*=
-
in o 33 -
D ni E \x ui
TJ (0 >> 3
O m ao 3
3 m oa <
r- . . >
oo L)
or i a u;
-*
M N a
>-nt . II
XX
NW oo
i . mm u I
Ul .XV WW
in IC 3
orx<<
U.
 I
> Ui 4 -S
< > w
a o
3 M
I* i
QC 0 <I -fc 54
r-
E m 03
u aa t
< u-z a
2 0.3 -w u
010CD z
a t l-l
o a 3UJ< N2Z
3 w <UJ .3 iaz
o
3QCIrt ao<
(OO DtiUJ mm
Ul 0<
Q. t*
Ul + a
-ui .
01.0 * in
33 QOi a Ui u ZQCO
30 E- 2 - i; <'x-
L .as
UJQC m 0. rt
2tnr. QCU) -U oqch

cl 3UX Uil
w|pj ; il
3 0)U<QC ai-u
3 UiZrt
H < m j U!
S H
_
rt . Ul U1QCUJ> QCUIUJ
. Ul
0) 2HO< U.U-*
*>X0)
O a ? a a
r* pj m in
I i i i
: T O '!"--p-
j
j
V4
1
n
I
J
fr
j
1
-a
'
o
.H
a
a
"J j ; rt
IS o j
. '. NTJ J E
irio
1
j i O N
-.nm . -
a i i r,
M
r;t\rt , I in i
.
i
i
"Din .
f P4 ,
2 W
J
m
VUO . j i 1 i .
-i
u.
wt
.-
1
j
j
j
J
-a
'
i
i
i
1
O
>-a
2
w
rt !
2
I
. m
"
f
' .
ui -x> M
trx 1 f < .
S
2X Sl:s
f i
-
o i a
m
1 1 2 fcl [
^
Q. I c*5
z - --._-
| I
I 5I
__ . ..
o - r-"
r- cs
-1- a
1-4
i
i
l
31
r4
I-*
54
z
rt Z
\ o '
'
a m
3 r*
13
-
j 1 \ a o ii
mac: i za
r*
Daw m aa
\
-Ort
3D
] ii 1 r^
. i
*
1
t < 1 , rl
1
a.
'-"
* f o O .; +.y
N a O i .
3 UN
1
aroo Z
oaa u a
cimort
r*
tt aa
r.
.
~
.*
l ao-
.
a a
acta f
o
-a .2
'
cj
^jmisq. > .o
A
i lIhML
-J 1.
3
1 m h
A"
-W
J
*>-X0) jj 1 o -<>-X01
. .Ul
*
8
N P
PJ *
o a a t ni a
1 i w4
a a o a
r
O
c
o
TJ
in IX
- a a
WINS
oo I -t
.oo
>tm .
IL .
ui.x>
ocx S,54 |~
"L - vo >
8i 1^ <N
tcj
I-
o a ii
54 r-
3 Pl s
Q. 54 >
Z QCQC a >S
MM UI
jsl ?a qc 54
-fc
^s c
Ss 1*,
I u.
"3 -fc <1
N N a .
Si
ZZ
mmuia
.
Ills
5
wt
-* Q
a
o
< * +z
2
r^
O
o a
Tl W
CLOa
tnoiM .
Ul <
CK ? *
ui H ?
a
Z .Ul .
ooe u ZQCU
3TJrt a Ul
CM - 2 . Z <U.Z
on .
Q. -<
a. .in * UlttC ->
d QCUJ -U wqch
2t(0 UJZ
o. 3UX
iNI
3 0)U<QC ? HU
<mJUJ UIZ-
H
uiituj> QCUIUJ
D
0) 2HO< U.U*
*>-X !>
o a o a o a
a t in
N (VI ra
I i i
' "
1
ax
n aa
in
*
5 m i *5
54
1
fc
1
M I.
< . a o
fc
2X ,
a k, VJ
fc ">
in 54 CN
pj
,a
VO
^5 k,
^
54
1
54
<*> !-.
oi s
fc
o
8-
54
&
"fc
v3
I
54 I
3
O
n
(M
s:
54
? I
05
5-
N O N fc
N Z
Z N oa
?o ?a
moo rt(M
T|rt(M
f .
I|D+ .
< 1 Ul N*
N* U Z
I z a
a ui a
OCfflO qc oa
tioa UJO irt
?in .rt
IOO*
.ta*
a .m
QCtoO UrtrtQ.
dWrtQ. 3
m f-
w . .ui
-1
.Ul >-XUl
*>-X(fl
t
a d
z
a Z3 .
E XX UI
m
O ?a
3
3
?a <
rs . . >
rt-4
H U
< UI
n n a
zz
XX
*5
DO
mm u I J!
ww z
a . . a
U
X
<
2
o r-h-
jj
MM
U.U.
u.
Z
u
I e5S "O
N
vo
<N
I.
z M vN
Ul
U
<
0.
< N
rt
fc cs
tt
Ul
>
<
-1
tt
UJ
>
Z
E
in
w
oca:
MM
??
?
Ul
QC
tl
54 ^s I
54
-2
a r* NN
2 a . N ZZ .
8.1
3
or
H
U
E
a
OO
U.Z
<o
ro
w
mmuia
u>t
*
a
1
05
UIZ ? + +z
0.3 -W u
wocrn z 2
? N a
3UI< ZZZ a
<UI .z ooz > ZQC
3KIrt oo< axui
hNKI-
DU.UJ hNI
Ul
2
<
H ? QC
Z .Ul . .<
UJ u ZQCU QCQ.
2 . Z <liZ a
UJQC - 0. -
QCU1 -U WQCH <<

3UX UIZ . .(:-
0iu<ac ? HU <auiu
<MJUJ UIZw . .zui
IUQCU1> QCUIUI zzuia.

2HD< U.U* UUUtfl
ID
3
[S
.N NTJ
ffl
,
J .
'
mno "
1 OO .
intnw .
NMD , ,
W
/
>oo . J
0)t
m i
/ '
54
'
Il .
UI -x> f
'
&
r- J* si
^
irx<. g 10
/
'
VO
'
h
/ 5s
8 <N
3 '
I i
II*
54 -h
\ - "
M 4
,
i^
A Sf 5
z
&! i
,
m . .
J .
5*1
a 5
u
< .
fc
2 N ,
x "
n ?a
'
ooo
T>rt(\J
rt O
I o+ .
0.
U) Z
N*
<
-
\\ \
:
a
Ul a
\
o: oa
\
\
'
TJOO
a in .rt .
\
Ul .aa
QC at
iu-(no.
.
r \ *
a
a *
N H 4
ri
..Ul ,
* >-xw
3 I I I . 1 I . I
a a 85
t
i
z
a Z3 .
E XX Ul
-
>> 3
D ?a j
?a <
. . >
m
H u
< UJ
N N a
zz
JCJC
OD
in in u rt M
WW z
a 3
o t-t-
u.
X -1-1
< MM X **. J. vo
U 2 U.U. u 54 54
z M
UJ t*5 ST)
U 0. !
I
< < N
fc CS
QC 3 z oca: ? IN *
Ul tt E MM |]]
> Ul m ?a a:
< > w
5> 3
u.
o N N
2
3
oc
a
E
a
. N
U.I
<o
zz .
mmuio
wt
I '5
S I
s?
H . a ?-*
(j aa m a
uiz a + +z
0.3 .w u
tn oca z 2
? N m O
3U1< Z2Z ?
<UI .z ooz > ZK
oo< OXUI
OU.U3 rtNZ rt(\|KH
Ul
2
<
H ? QC
Z .Ul . .<
UJ u ZQCU 0.0.
2 . Z <liZ a
uior m a. -h HJ
qclu -u OIKH <<
3UX UIZ . Ch
0)U<QC OHO <OUIU

<m_JUI UJZ-< . .ZU1
.a
UJ UIQCUJ> QCU1UI IIUJQ.
*a ,n
2HO< U.U3. UUU0I
Appendix F
Case Study
of
Rectangular Window and Psuedo Random Noise

vs.
Hanning Window and Random Noise
Hanning Window and Psuedo Random
Random Excitation Generation using a modal hammer
Material: Aluminum
216
Rectangular Window
and
Psuedo Random Noise
Length =
1 0 inches
(Gain =
3)
217
- O
z
Z3
0 XX ui
fc >> 3
o OO 3
o oa <
t >
m
H U
< UJ
n n a
II
JCJC
oo 54
5i
itiui u
WW z ^5 9,
. . 3 ^-- "^-
Hh- 'f- 54
U.
X
-1-
to
-1.1
<fcL
< i-i rt x
54 54 ,
U 2 U.U. u t>
-fc
rt
UJ
^
00
U fl. v. s %,
< < UJ
54 ^
QC J QCQCOIO be
UJ QC HH -H - IjJ
> UJ N aaooc ^
fc -L
< > I zu.
BO N N
fr-s
5. -fc
X E 112 .
a; ==
3 O u. mmoa
QC o < QC at r->
H L>
< z*
u OO 3 < 54
UJZ o D3 + + QC
0.3 -W ZU "
fc"
010CO <z O 54
+ Nm< ?
,V
3UJ< IUJH 3
<U1 .z D0)U > ujq:
3QCIrt 0<U1 O>0)Ul
OU.U3 omoc rtNQ.|-
ui
2
<
H ? X
Z .UJ . .<
UJ O ZQCU QCQ.
2 . z <u.z O
UJQC Q. rt r--l
QCUJ .U 01QCH <<
3UX UJI . -QCrt
tnu<x ? HU <ouju
<M_JUJ UlZrt . .ZUJ
UJ0CUJ> QCU1UJ IIUJO.

2HO< 0.U3 (JUUO)
ID
TJ
01
fl)
I
NO
NIC)
.101
*
Q
? MO .
min -5
(\i i
54 .
u;
fc 5
.x>
5
~"
'*
rtx<<j o
^<
54 54 <S
i- i*;
Os
Q. *>. fc fc
CS
Z *>
* "rX
.
Re
-S
"s-* k,
fc .
C 54
>-fc
*-
5.
-o -
a
o
a
r m
m
fc-
:
rt 54
_.- _. . . r
*
30
-
a
rt
r O w
-
Q
o O
N .
I<
L UJ
U
z
o*
o
o
-
UJ
-
o X + 0)
- Q
o U1C rt
*-*
i a ns
o I
U-l oa.
3
m
f L w . .UJ
*>XO) !
v "i
w-L.
Q J- X>
D o o D m o PJ
* a
(NJ PJ t
1 1 -
6 d Q d
a
?
N
1 Zl
a
.a
m
3
5 SX
-fc <
*"" <s-l
54
fc*
fc k
54 O
14 ^
(N
* -5
!
-
X
m
MO
ra
::w
3
-
<
UJ *
U
z
Ul ?
QC aai
UJ30rt
la .*
a .ei
Urtr-,0.
3
i-
-mm
-
, W - . 3
> x oi
ID j
d d d d
H
Sl Z3 .
E Vs UJ
Tl to >> 3
0) m oa 3
, V iri aa <
m r-- . . >
rH ,
0 u CD u
St 1 UJ
rt(Q N N Q
> -O . I j-
XX
mw tote
mm u
uj -x> CD ww z
q:x<< CO . . 3 - .54
l-l-
u. .a
^r.
J3 "-N i-
< H hi i-i X >
x liU. u
I-H k
H '3 2 S
3 3 0. 3
fc (N
a CS
1
Z N
<
QC
 Ul aaaoc 54 ~
SI < > w u. 5 -5
S a to N N
k z 1X2 .
3 1 N mmaa
in
.
*-'
S
s
ac to
E
<1IQC
o<
at
z*
s1 u
uiz
ao
a 3
nj
t
<
+ QC
0.3 -w u "
5 oiaco) z a
t < a
u
Q 3UJ< N2H 3
s. <UJ .Z IOU > UJQC
3 1 OQCIrt oouj a>wui
sj ? U.UJ mNQC rtPJCLI-
UJ
X
<
X
Z .<
Ul u ZQCU oca.
2 . 2 <u.z a
UJQC -i
0. rt H3
QCUJ -U 01QCH <<
3UX 31 .
-QC-"
0)LD<QC ai-u < a uj u

<rt 3 Ui UIZ" . .ZUJ
'3 UI3C3> QCUJUJ IIUJQ.

0) 2H3< U.U* UUU01
Mil
a a O o a
(V PJ m in
i !
T T
1i 1
1
T j
4
1
- a
o
'
PJ
m 1 1
"D 1 1 a
a o 4 1 on
. N NTJ . a a
i oa . 1
1 rs
pj PJ rtO
o
inmo
P-WW
i
4
/ a
ytdrt
mm
.
*
/ a
>a
o
ID w
PJ 1 pj z "3 5
l-l
u. ' < . L. O
Ul
QCX<<
-X>
^
/ -
a
2X .-4 fc
y in
pj
i
4 /
/ k
54 CS
CS
I-
i 54 54
/ - a S -fc
CS
^ pj
A fcC *=
,
\\
-
a
Z "
\
" CM
\
1 \
3
<
|\ \ 1
-
a
o N
S CJ I
I
1 CM .'
PJ aa
\\
'
PJ
Rboa a I+ oa
rtf\l
rortrj 4
-IO 4 t -
(
x:o + ~ .
a <
n
'
UJ
l\l
CLi N% \ I
10)1 I \ Z a
ilill o 1 UJ a
locima
f
uaa
\ ?
QC
LUO
aa
.-<
a a
? IO .rt
-
? 1003
1 pj
!UJ! .OS .
pj a .o
nemo ! \ j. (JrtrtQ.
[LyWrttL
3
1' \\ L
m
3
h
s h ; w . -UJ
-
- -UJ 3. 1
1 1 1 Fs 3>-X0)
3>XU) -*
2 ID pj
Q o o o
PJ PJ *
1 d d d d
1
ZD
o
3
0) aa
a a
a r-.
EOO
0) e :
XO '. .
wtc x
rtic 3U vD
mm
t\: w
OCX-
i-i
k
54 m
h t-J
3 r, C.
u cs
0. < -.
J
:i
-
CK2CI
5 T- Sf,
z X
_
2 54
*
î
^
X
'3
E
LO
7
54
-fc
-^
Nl <; ,> CM "i -s:

a rH
X
.
-7
3
a
c i
. K
I mm -a
1-1
u <OQC
<Y a
h-
ro<
a D aa W3
< Li- IT D 3
lL "3 .w a
^ (/.arc ._
*
n , <s
a
n ce a
w
sj
"> < Uj
'
< :
^_
ZSH
aau
> 3QC X *"* aaui a > oi :
OIKDCY Nl 3 '.- 33 ->nx
<
+ *
Q
one E
3rt rn n zacu ici.
-
<\i rH
Z m < u. ^_
|301
, "iS UJQC - CL -<
3 u 01QCJ-
si mci -
n
QCU1 -U
3UX Uil - - 1
w ; Cll-U
-
"^ tnn<x
h-
r-.
<-i_,L;J J. !
*'
'
.
i QC '3 > ocuiu:
.
-
a LJU'DLT
01 zi-a< U.U3
o
rH
rs
Q
a
rs
E
m n
ai
u aa
a
03 a
a
>d
f-
a t>
a z
ID i-i
< .
2X
Fi
a
[s
in.
(0
k
54
CM
CS
a
ID
D
1
$
z
l-l
a
in 1
D
N
a
I
oa
ID aa
rtf\l
<
a III N*
( ) I
CO
7 a
111 a
X aa
UJC .rt
n IOOS
pj n rH
ID u- (DQ.
3
m H
w . .UJ
SXXO)
a> ID pj
D o o
ID d
PJ
1
<r
1 1 d d a
' '
f v
i
:
H
1
.
a) >- a
<J
z
c "*
a Z3 -
E XX UJ
CD rH >> 3
TJ O o aa 3
CD 3 - ao <
n .
(Q
[s! . . >
t tn o .
i w
*
v-l rH
E EO "^ (J
trs i
< Ul
.
.. .
n n a
xmo .
ii
ao i- 1 JCJC
mm
t- (OO
u.
Ul -XX
.
" [ mm
WW
u
z
QCX<< -
HI . - 3
\-r-
-
CJ o li
X H -1-i
< H oc
- 54
U 2 lili u k
Z rt ~fc' in
1- 1t ft 54
m UJ (N
3 u a. 3 5- * (N
-
a
a. < < N UJ
z ^ QC 3 I QCQCOIO
5 ^ .1
H- / UJ QC E i-i i-i rt LU 2 54 i-JL,

*=
i
* > UJ m aaooc 54
N < > w 211
i
i
s a -i
N N
; 2 0 N 112 .
j .
3 E lil mmaa
a \, oc a <OQC ot
/ s H o< z%
a / it:
u aa 03 <
< / ujz a 3 + + QC
'/
X -ID
0.3 -w U
^ 01QCO z a
a d t N < o
o a Q 3UJ< I2H 3
3 CM "> <UJ .Z OQU > UJQC
a. a o ll 3QCIrt oaui O>0)UJ
j
01 ow - i OU.U3 rtNQC -IMQ.H
UJ < - Ul
QC t* d 2
<
Ul o H ? QC
01 tn e Z .UJ . .<
3!"0rt a UJ u zacu QCQ.

3rs . ;
-
(\J rt 2 . Z <U.Z a
Q-i m / 3 UJQC -I Q. >-i 1-3
//
r a
zjom 1 QCUJ -U 01QCH <<
'
rtjrt 1 I a. 3UX UIZ .
-QCrt
/ 1
3 0)U<QC OHU <fflUJU
N \l i
H <-i_;iu UJZrt . .ZUJ
"--
sxx
J__ hi. t
Lo w
UJ
ujq:uj>
2HQ<
QCUJUJ
U.U3
IIUJQ.
UCJU01
? oaaaaog
' 4 "<
0J Pi <r u1 113
Hanning Window
and
Random Noise
see Appendix E
Length =
10 inches
(Gain =
3)
226
Hanning Window
and
Psuedo Random Noise
Length =
1 0 inches
(Gain -3)
227
' ' ' '
T 1 1* * * ~
,.,..
1 ;
'f
r
!
i
rg
a r o
?
i 1
1
H
[
i |
'
i
-
o a
nil
} | j I 1
a
w o 4 i 1 ;
J
3
E
r>
in Nit 1 ! fs N
(MIO . i i
-
a r,
amo
tnwm
1
j
i .
*3
i m
.[s
. f j 3
>-
.w . 1 > .
MO i 1 r>
m I
-j
-
o
ts
1 !
a m ^ =
ts
4
1/ z
"" "
4
rt
UJ .X> 1 1
< m
QCX<< 1
1 '
j 2X
54 r~
4
I
-
a CN
4
1
4 sV fc 00
I-
->
i i | 54 5"
Si cs
- 54 cs
!
Z f
4
-
.
a
m
i
]
!
-
a
in
54
I
i
1 / 1
;
i
j
j 3-S 54
j
1
/ ! l
-t -
a -
a
2
i ,
t ^ t z 54 S
-il -1 ^crî i
W
-J \ HH
3 3 1 "**TV 1 3 54
< 1 r\ '
2 J \\ i
-
a
A NO 1 m i "m NO
ttza <
\\ |1 ! la
30rt Qrt
1 1 1 1
-IOO
IlOrt .
] j \ 1 1
a
rH M
| < I ) <
!Q-l a i uj + *
+* -| f g
CM|
u
lOll
uj n
4
: r i Z N
UJ I
octal
fnoa
aimart
1
\
i
!
' [
fc
qc
UJOOrt
aa
iaa3
1 1 a * .
-a
a . .
*
^mao. UrtOO.
3 i : 3
*v I ll 1 |=^ m h
rt . .UJ
hi! i
^*-
f^^-
'
W . .UJ
.
sxxoi , I 1 -jÂ3
a
' .TT-1
?
3>X0)
CD D *t (\1 c
oooo C)
PJ PJ t 1 < *
1 1 C) cJ cJ c
' * ' '
""1 i
,
'"I i
1
1
4 i
i f a h
01-1
f r-t
1
xz
in 9 E3 .
'
d E XX Ul
-i
m
r4
>> 3
3
a EE 3
OO <
N 1 ro OO >
(\1 0 0 . J
t>rt(D
- H u
ON 1 - a
-a-
< UJ
ai n n a
>oo .
II
a
' i XX
/
r4 rH
1
i oo
i
IL .
i
i trim u
UJ -x> 1 WW z
QCX<< in ^H
-j | i
0
o
.
r-r-
. 3
li
54
c\i ô
X
j < H
-1-1
HH X
sl. <
54 54
4 U 2 U.U. u >>
Z l-l
fc Os
h rt UJ
. fc
3
Q.
f _ U 0. 3 CL **
CS
cs
1 ra
< < N UJ -*J
X
2 4 X 3 I QCQCWa
i-i i
3
rd
UJ QC E rt ^H HH LJ oi 54
,
j "
"T
i
> UJ m oaooc il
-fc
i < > w zu. -3 -fc
4
1 r
? 2 0
a H N N
112 . fc- s
"
i .. 3 E li mmoa ,fc
-
Ul
QC O < at
3.
'
r. h ; z*
Ul u aa <
< UIZ o a t +K
2 0.3 .~
zu m
i oiocm <z a
Pm o
a
Ii 3UI<
? Nffl-i
IUJZ
a
3
!TJ- - -
a
rt
i <UJ .z QUIZ > UJQC
0. O 0 . rH 3QCIrt o<< O>0)UJ
:0) at . OU.U3 rtQjI rtmo.H
lui < Ul
jQC + * . 2
1
l m <
!UJJ_B / H a QC
itntaE Z .UJ .
urot -! : a Ul u ZQCU oca.

3t) . Q r^ 2 . Z <u.z a
tLi .to i 3 UJQC >- 0. - h--i
2|rt(0 J QCUl .U UltEH <<
itsjm i j i 0. DUX U1I . .QCrt
'
1 3 <na<x anu <muju
N . : H <rt_iui UIZ" . .ZUJ
"
J 1 UJQCLU> QCUJUJ IIUJ0.
*>-X
. .
j
~
1 ! Si 2HD< U.U3 UUUO)
c} a a o o a o a
p1 c\1 c\ I <
7 i i i
J
; |
i
1
i
1
I
o
T3 O
IT) N NTI
3
\
.Ilts
.
1 ts ai
'
NQQ . n aa
minh m
NWW
M I
,
1?
ytD* . >o
mm m
W I w 5 ?
z
ih
u. .
Ul .X> i < .
QCX<< 1 2X ^ .5
ki"
-fc
IS 54
j
1 1 VS O
l-
fc 54
54 fl
/ ?> 54 133-
CN
3
Q. -5
f
' \ fc
\
1
^V
Z
ih s ?
3 55 *
fc
fc-
U
< 54
2 N O N
I PJ I
N oa
oa
aa
ooo
TJrt rt
a 4-
.
I o+ . o
<
<
N*
CL
0) I
N*
b I
z O
UJ a UJ o
QC oo
QC aa
"Do a O UJO .*
aO i- -
a 1003
UJ .03 N
a .o
mo UrtrtQ.
ilumrtQ.
3
m h
N H w . -UJ
rt . .13 . \ a
a 3>xui
s >-X0) J~^ _,.
pj
o o a
CM t
PJ
I i d d d d
'
BE
i
i i
j ~~
i
J Q
H
rt rH
*Â^ \z
. Ul a E3 .
D ^ ! E XX Ul
en
g - >> 3
tÊ_ 3
E E
r-
CD O O
3 aa <
-<X-> | ao >
(M ts t>
W-t 9-t
(MB 1 m
E OO H U
t E 1
'
^*~^ _ i
-
a <
n n
UJ
a
r m
. .t
>U) . . ^. II
ao XX
*-i (Q
oo
mm u
lL .
Ul -x> WW z
m a a Si -51
QCX<<
" . .
fc .54
O r-i-
ti 05
CM .a
X 33
< H rti-i I
U 2 U.U. U sL
i-i
z
en
rt . Ul
3
Si
3 U 0.
- Q < < N Ul
1
.a
X 3 i acacwa
Z _. '
CM
UJ QC E i-ii-ii-i iu at
$ k,
> UJ m aaaoc
< > W Zli
fc *
a ^ N N
2 9 .
u.i
N 112
mmaa
.
.'
3 E fc
in oc a <m at
H m z*
u oo w <
uiz a + +QC
 ujqc
m -H
3QCIrt ao< axnuj
Q-P 0 au.u3 rtNI rt(T)0.H
tnpt . Ul
Ul < 2
QCl +* <
H a oc
UJ 0 z .
tn HE -
in U ZQCU QCQ.
a LU
3 3t Z <u.z a
o w-t 2 .
3O UJQC "- a. ih H3
2
a. .(0 QCUl .U oiqch <<
2 mo 0. 3UX U1I .
-Krt
i-i m i
3 01U<QC ? hu <muiu
<rt_JUI UJZrt
H . .ZUJ
Is -* IIUJQ.
UJ UJQCUJ> QCUJUJ
rt -
a U.U3 UUU0)
m m
0) 2HO<
?-*. [
? a a 3 a
a a
pj rH pj TJ
m t4
i
I i 1
Random Excitation
gerenated by a Modal Hammer
Length =
1 0 inches
(Gain =
3)
232
m
1
m
rH N3
rt Nits
i un .
inM/i
cumin
.to s
>- .m .
too
A
u.
UJ .x>
-
iI
1
* fc:
acx< H
-
. .UJ
SXXO)
-
T ! t
t
ri
in a zz .
CD
E XX UJ
V Cl ri >> 3
(T)
oi a-^ J
O '
a . <
0) 3 >
rn rs m .t
1 1
I to o . .
(OOO
a
mm i
.mix ni
vtm a
mo
WW
i mm u
li. .
WW z
uj -x>
i *o ^;
-
in 0 . . 3 54
QCX<<
"
i i-h u.
"*.
-5 C\j
a s ^
X . -1-1
>*, Is.
t < H rti-i
_
X J" s*
U 2 U.U. U <*> y
-fc-
z |-1
5
-. UJ fc n 54 fi
1- L. (N
U 0. 3 !^S
D -
a < N X
< 
Q. 3 i qcqc a Qr-
pj cr 54
z UJ QC E rtM UJ -sr
> uj m aa x sU
>*
< > w u. "*** -fc
a rt n n
2 8 - II -
*
zi e u. mm a ,1
in o: a < "*
ri
H . *
(j ao
L3 uiz a a + + a
< 0.3 -w ZU UJ .
Z oiocrn <z 3
' ? NO1-! a
I
3UJ< IU1Z <
m oi Q <ui .z aoiz oio:
3 LO
-
i **
3QCIrt 0<< >>"UJ
a.o a
mint .
I .
OU.U3 rtOJI tWOH
UJ
.'uj:
< 2
X + 3: <
h a oc
LU 0 Z - -Ul - -<
oi m e -
in ZQCU QCQ.
0) UJ U
_J3t r-l 2 . Z <liZ O
'
O
3'rt . 5 UJQC rt 0. rt i--l
CL. .O 3
QCUJ -U 01QCH <<
ZOO a. 3UX< U1I --P^"
ill I 3 0)U<QC OHU <OUJU

<rtJUJ UJZrt
'
H . .ZUJ
Ul U1QCUJ> QCUIUJ IIUIQ-

a
j UUU0)
-
0) 2HO< U.U3
1-
3XX 1
o a o o
o O a o rs IB
PJ fl) t in <D
%-4
i i 1
1 I i i i
' ' '
! ,xy_ !
|
3
3
-
o -
a
a o
PJ CM
)
CD
3 O 3 : E
NNNTJ . . n .a
*N
.lit rs N tsl
'
moo . pj i CM NO
mina m
NWt ts
J
.
1
'
>-0)O . | r,
^
a
win
W I '
o
pj
! rt
"3 5
u. .
'
Ul -X> < .
QCX<< 2X
-
a -
a
in in 54 k.
cu CM
^ 54
H
^
O fc
54 54 en
t~
fc Oh
3 -
o - a
Q. t ,
t .So
Z pj CM
i-i
s-
* o z
z ->
8 r
,
Cl w
rt CM
CM _j
U S.
< 54
2 N O ÔCM N
I
I PJ
N CM aw
aw
am
oom : w -i
"Urin
\
rtQ \ i
IO+ .
\ O
ri
2 +<
'
CM Ul N4
CM -
L) I
\
z a
ui a
oc aa
9 r
o UJO .r*
1
.
a
a 1003
\
,
UJ .03 '
pj N
a .o
QCOO
iLrtrtQ.
^\ UrtrtQ.
3
3 j m h
rs
JH.
h
J
^ -
o
.a
w . .uj
3>xoi
\
. -Ul
a
LSp-xoi a j
a io * w a
? a o a 3
M t 1D 3 ?
1 a ?
1 1
z
in H) Z3 .
OD E
V m
CD o o
3
D ts (s
10 0.
ininm C
to i a < UJ
-to
m D
>tom .
u. . O rv
UJ -XX
o:x<< in 0 . . 3
o r-h-
U. C ,54
ft!
X . -1-1
< H HM
^ ^ v. --H
u 2 LLlL U
i-
z
l-l
i-i
UJ
s * 54
O
en
u 0. 3 CS
a. a < N
<
z cm QC
UJ
>
3
QC
UJ
I
E
m
QCQC
i-i rt
aa
a
LU
oc
C<
54
54
'S
I
< > w ll
a *-t
N N -fc -5
2 0 .N II .
3 E U.I mm a
in
QC a <o t
j^ H , m *
to u oa w
< UJZ a + +a
0.3 -w u UJ .
mo: cd z 3
W t N i-i O
o m 3U1< I2Z <
Q
3 n <U1 .z oaz 01QC
CLIO ID rH
3QCIrt oo< >>rtUJ
tnlot . OliUJ -NI tWOH
UJ < UJ
QC f * 2
m <
U o H O QC
int
1IO E z .UJ . .<
in
33t a UJ u ZQCU oca.
ap) . a rH 2 . z <liZ o
A-'ID 3 UJQC t 1
0. - r--l
sb co QCLU .u WQCH <<

Sjrt | 0. 3UX UJI .
-ac-i
3 o)u<ac OHU <muiu

H <rt 3UJ UIZ*-" . .ZUJ
UJ UJQCUJ> QCUJUJ IIUJQ.

o
sxx 0) 2HO< U.U3 UUUO)
n o O o a
pj CD in ID
i 1 1 i i
Appendix G
Variable Sine Excitation Data
Sensor: Piezoceramics
237
I 0
....
O T , ,,,.,,,,,
?
7 1
1 / 1-
2 \r a
rH
i
i t 1 I ; 1 ';
i
i I ^ i
r ; | o
i / a i
!
a
1
tn I
fi 1 !
3 i
11
i
w o .
I i E
. N3 / :
a n e
t NIrt O Nit
rtIO .
-
1
a
o
! S
B
oio .
DOO aao
omm . |i / i , amm 5)
i -m . fc
/
.in .
> .w .
'
i i i r x .w . O
MD i / i I r NO
5
m i i -
o i [a mi
I i / rs | i
!
N ^-
C^3 5
IL fcj
"
Ul .XX
-
J
' !1 i |
.
UJ -XX -L y
acx<<
! 1
i
!i QCX<<
54
>
s sL,
-
a .
Q
i -': i i in
~"
[ ^_.j_ ID 54
0
i
i
i '
i- i
'
\ ^ C
00
i ; ; 54
h r
54 2! fc. fl
' 54 CN
3 1
-fc
a.
z J
fa -
1 -
a
in
c
SJ
M
\ 54
i
1i oi
54
-fc
1 ! 2s
u
1 -ST
fc
J i O
-
r
a
* Z
i4
a s
i
_sl
3 P
f"
fc"
rt
-aj -s^ J
U
<
3 j j 54
fc.
1 1 ^L
s
o
Nm 1
! i
-
.
a
m
IN
oits *
! am
3am 1 v, o
hOD x1 ri .
IOrt . I <
tt.
< L .
o
"
q
w
ui + *
+* J CM i . u
cn Z N
UJ N i UJ I
QCOI 1 i j aa
j oc
HQOl I UJOOrt
? NOrt i
i -a 1003
UJ .03 I -
O
i i rtQ.,
\
**
ccm . .
l i UrtOQ.
UNDO. i \\ L i
|1 3
,
3
1 ^ t H
(M H .
"'"
! "
i
rt . .13
n . .UJ i Â ig-^
*^xw
1
B^XM 1. i , , , , i ^ . . lo .a
CD ID t ftl 3
3 o o 3
pj M C3 3 ?
1 1 rJ 3
I $t
IU H
Q) i-l
\z
ffl E3
E XX
rH >>
0) oo
oo
m . .
m
rlrl
h-
<
N N
II
JC JC
OO
mm
WW
0 M
O r-V-
. =o
X . -1-1
*, fc. <N
 UJ m DO I 54 -5
< > w o
a N N ts
2 0 IIUJQ
3 E u. mmz .
QC OO < -io
H .o 0)
U oo
UIZ w a+ + U1
0.3 m zuu 3 .
OIQCU) <zz O .
t NOrtrt <a
i-i LU
3UJ< IUJZZ
<UJ .Z O01ZZ >QCQC
3ttCIrt a<> 0<li
DU.U3 rtOJIIt
QC
a
H
H a<a <
Z .uj . . .QC
Ul u ZQCII QCUJ
2 . Z <u.uu OZ
UJQC a. HUJ
QCUJ .U OlOCHH QCUJ UJUJI lUJrt

2HO< U-U33U UUO)
Appendix H
Piezofilms
with
Unidentified Noise Source
Material: Aluminum
240
^ft
r"
a
0 <-i
X
EZ .
0 XX Ul
3 fc >> 3
to a r> Et 3
in i Cs a . <
. NUI a ot >
rH Irt
*
NKlh H u
.ON
>
i .a
.a .
1 <
N N
II
UJ
a
-Si
^
QJ rt XX
m ; s^ oo
i . mm u
LU -XX WW z fc c
QCX<< r^
3
k
M
7
X
< i-i i-i I
All 54 fc a
. ~
fc 1
JO 2 U.U.U.U th
H.
U-rt
h <0rt QUI of
3 3
U Q. 3
a. <
^QC 3 N QCQC a
buj QC i i-i rt Ixl
UJ m aa x
> u.
N N 3 "
"5
>2 E II . oi -g 5 "
S3 a u. mmuia & S
o < 0)t
? -
a z u ao a
% W
< ii ûiz a a + +z
2 3 0.3 -t zuu "
:? &
""
oiocm <zz 2
a t N fflurt a
a 3UJ< IUJZZ >o
Nt <U1 .Z amzz EZK
3QCIrt o<<< >o<uj
'OI
13 O . OiUJ tan mmocH
aa< UJ
o.ota 2
in <
3 + H a<a IK
Y Z .uj . . .<
N Ul u ZQCII QCO.
3 fcl 2 . <U.UU a
30 0 UJQC rt a. H3
ra .
QCUJ .U OlftCHH <<
-4/-0
3UX 311 . -QC-i
01U<QC OHUU <muju

<rt_JUJ LUZrti-i . .ZUJ
ujq:uj> X Ul UJ '3 IIUJQ.

i 2Ha< U.U33 UUU01
3VX
a o
a
Peizofilms
with
Identified 60 Hz Noise Source
Material: Aluminum
242
0 ffl
E E
rtO
rtfflZZ
o . .XX
aa>>
r> wa Et
IN r-irsa .
0) . -at
iirtrt
HH
UU
ZZ N N
U1UJZZ
-1-1XX
OO
mm
WW
o
w r-i~ 2 tc
0 033
H E
(DOU.U.
Erti-i ^3 <N
WW 5. a =C
fl
^ gi
S"
WW (N
!
rsrs
n . -acoc
X HHi-irt
m U.H.QQ
,
I I I
a ii n n . 5
. 0)0)11
u. mm
<J 3
<
Hrt
ozh + +a
zujz ui
<rtUJ J
Namz>>m
iuizo E E<
o 00X0-000)
3 NW o<ocxaort
rtOI tmnuiooa
IOO .
a<
a. t*
0) ? <m
'3 + .ui . .
xm zqcii x
3 N <u.uu a
? WI o. h
3 .a U 0)0CHH <
QCO . U1II - -QC
U. i a 0)U<QC OHUU<mUJ
<rtJliJ UJZrtw . ,z
UJQCUJ> QCU1UJUJIIUJ
U.U33UUU
a a a a
a a a o
in (D p- m
pj m *
i
i i i i
i i i
Appendix I
Climatological Data
of
Rochester New York
244
U6 HI 2
ISSN 0118-StSO
ROCHESTER. HI
NAT'L HE* SER ore .MOM
SREATER ROCHESTER IHTER'L MRP0RT

LOCAL
CLIMATOLOGICAL DATA
ROCHESTER-MONROE COUNT" AP
Monthly Summary
t EXTREME FOB THE MONTH LAST OCCURRENCE DATA IN COLS b ANO 12-15 ARE BASED ON 21 OR "ORE OBSERVATIONS
' TRACE AHOUNT AT HOURLY INTERVALS RESULTANT HIND IS THE VECTOR SUH OF HINO
ALSO ON EARLIER DATE :S I SPEEDS AND DIRECTIONS DIVIDED BY THE NUMBER OF OBSERVATIONS
HEAVY i-OG- VIS I BiL i T' '/4 HILE OR LESS COLS lb S I': PEAK GUST HIGHEST INSTANTANEOUS HIND SPEEO
BLANK ENTRIES OENOTE MISSING OR UNREPORTED OATA ONE OF TUO HINO SPEEDS IS GIVEN UNOER COLS 18 i IS: FASTEST
hIGHES-
KILE -
RECORDED SPEEO FOR HH1CH A HILE OF HIND PASSES
SIATIDN IDISECTION IN COMPASS POINTS I FASTEST OBSERVEO ONE
MINUTE HIND HIGHEST ONE HINUTE SPEEO
-
(DIRECTION IN TENS OF
DEGREESI ERRORS ul LL BE CDRRECTEO IN SUBSEQUENT PUBLICATIONS
I CE'!CT THAT TH! S iS AN OFFICIAL PUBLICATION OF THE NATIONAL GCEANI AND ATMOSPHERIC ADHI Nl STRUT I ON. AND IS COMPILED FROH
RECORDS ON F!<.E A; THE NATIONAL CLIMATIC DATA CENTER
NATIONAL NATIONAL NATIONAL

OCEANIC ANO ENVIRONMENTAL SATELLITE, DATA CLIMATIC DT CENTER DIRECTOR
noaa ATMOSPHERIC ADMINISTRATION ANO INFORMATION SERVICE ASHEVlLLE NORTH CAROLINA NATIONAL CLIMATIC OATA CENTER
Climatological Data of Rochester, New York

August 1992
Page I of 4
Figure I. J
245
OBSERVAT IONS A' AUG 1112 14768
3-HOUR INTERVALS ROCHESTER. Nl
iflf I5I-
KISI-
I_ 911 it
l[HP[RIIUR( HI NC
BI111T iniPIRAIURT. HIND
8IL IT KAPtRAIURE NINO
1I -z z E
MAIMER
'
| | I
I
u Z
UC IHCR
s
I 5 2 | . -5
s
HEAIHI B
.li
3jj
i s |
1 1 * *
-
'
5 =i 1 5 ; i I S sj=! * SIS
~ z
5 VI i X
s
AUG ll AUG 2nd AUG 3rd
0' 13 ILf
IO 5 it 55 55 ii ?i 1 8| 10 I5| 1 112 ,57 il 70 11 1 1 ii 60 li1
S3 i
60 57 B 1 1 20 ' 5 I
Dl IC 55 7 jl 57 55 54 10 26 12 21 UNL 15 ! 60 5b 52 75 21 6 10. 10 IS'
| 62 51 56 61.22! 7 !
IT D UNL 15 58 51 55 10 27 11 01 UNL 20 63 53 54
10'
| 73 11 1 10 10: 3 1 51 5b I8i 20-1

10 0 II III 20 1 18 10 54 n 21 U 0 UNL 20 !" li 51 2) ' 8
12 n 12 li UNI a, is i bi bl III
t) 1 UNL 20 75 13 54 (8 28 10 5 UNL 20 I 78 ll 53 50 23 10 1. 31 7! : 75 ] 1,1 b3 ll 25 ! 3
Ib 3 UNL 20 II 15 53 54121
IS 0 UNL 20
13 1 UNL 25 71 ll 54 42 21 i 10 10! 25 l, BUT 66 I
65 b5 17 02 | 1
10 12 51 III 24 1 1 UNL 25 ,? bl 51 ll 21 1 1 10! 12 ?l 8! T 65 | ll bl 13 I4 I 3 I
|
Ul _a Vl 15 14 58 53 18121 ! 1 3 UNI 15 ! w ! .0 51 701 21 ! 1 101 25 li li our bl : el 63 17 ill 1 1
AUG 4th AUG 5th AUG bth
01 10 70 b r 13 162 12 !17 03 i I 11 UNL 56 55 55 17 24 i 51 151 64 21
"1 ! | D! UNL I5| 1 52 6
04 10 Ib 5 r 12 ll ll 1 17 11 3 OlUNL 54 54 53 11 25 i 1 S3 52 10 23 5
15| Oj UNL 15 55
DT K 5 1 F 13 12 12 117 32 1 10 21 UNL 57 155 54 10 21 6 0 UNL ! 58 151 51 67 21 5
12l 20 i
10 H 35 12 10 55 58 13 23 1 5 li UNL 20! 1 b> 58 50 55 30 1 Ol UNL 20
1
12 III 58 12 23 I
13 10 no 15 12 51 57 ! 84 24 i 6 IOI 250 25i 72 158 41 41 30 1 ! UHL 76 13 53 12 23 7
lb 1 250 15 18 12 57 i 18 25 1 10
| 20
I0| UNL 25! 72 10 50 46 21 1 8 1 UNL 25 18 |12 50 38 20 5
IS 1 SO 15 13 58 54 j
73 23 1 101 UNL 18 III 55 63 24 a 1 UNL 25 13 III 5! 50 12 1
Ll 3 UNI 15 57 151 55 IS] 22 1 4 8i UNL 15'
! 10 ! 56 52 75 23 7 0 UNI 15 (.2 151 56 Jl 00 0
AUG 7th AUG Bth AUG llh
Dl 1 UNL 15 58 51 55 10 20 3 10 250 17 60 55 61 17 Ibl Ufa 64 64 21 5
12! 5 10 34 8 1
D4
0T
3 UNL
7
15 51 55 55 17 00 0 10 250 15 ll 58 52 11 ll 5 10 11 7 ib? U 64 10 IB 5
10 20 10 5! 51 81 21 4 10 150 15 67 bO 55 ll 13 7 7 250 6 F 18 lib 65 10 25 I
10 0 UNL 15 71 14 58 58 21 a 10 15 10 R 70 13 51 16 16 1 10 70 7 1 70 17 61 87 23 7
'
13 3; uni 15 81 13 51 35 20 6 10 40 2 BlRUr 66 15 15 17 13 8 10 11 7 1 72 Il7 14 76 23 10
16 31 UHL 15 81 ll 51 42 16 a 10! 41 10 1 13 ie 15 76 11 8 10 30 7 111 117 15 87 30 I
IS ll UHL 12 75 ll 56 52 12 4 10! 130 7 7! n 71 20 68 115 13 84 31 1
i ll 12 10 75 7
fez 101 250 -11
18 bl 51 lb ll 7 i 5 81 250 7 ;; 17 15 81 11 10 ID 21 10 i 66 114 63 SO 27 I
AUG 1 Oik AUG 11th AUG 12th
01 2 UNL ll IF 12 III ll 17 00 0 81 250 8 1 117 61 15 13121 8 1 UNL 15 56 54 53 ISO 11 1
01 10 33 5 f 13 112 62 17 1) 3 81 15 10 68 11 bl 13 24 7 0 UNL 15 53 51 50 10 23 5
Dl 10 3b 2 llF 15 114 13 13 24 3 7i 250 15 65 63 12 10 28 1 6 25 25 51 57 55 87 35 8
10 7 30 8 1 72 17 14 II 27 1 ll UNI 25 71 61 bO 68 27 16 6 27 25 63 51 52 68 36 5
13 10 UNL 1 [ 78 lis 61 12 21 1 3 UNL 25 73 1 ll 51 12 31 14 6 28 25 66 5) 50 57 01 5
16 1 UNL 10 83 i 72 li 51 27 10 3l UNL 25 !7! I 12 155 65 55 46 50 33 5
53| 30 12 a 65 25 |
ISl 1 250 S r 71 72 68 11 17 < II UNL 25 117 51 5.3 ll|.3l 8 7 UNL 20 1 ll 54 46 63 35 6
U. 10 31 7 I TRN 70 lib 61 81 24 112 01 UN'. iSi 1 1 10 1 57 154 81 26 1 b 2 UNL 15 1 55 52 5.0 12 OO 0
AUG 13th AUG 14th AUG 15th
Dl kj 70 151 i 52 51 150 S3 17 5 101 21 I5i 158 55 153 81 03 7 2 UNL 15 51 50 45 13 21 3
Dl 10 130 15 53 51 150 10 18 3 10! 21 15 58 56 |5I 81 02 10 1 UNL 15 41 41 46 16 23 1
DT 10 43 201 51 53 51 8] 01 5 5! UNL 10 10 56 53 76 02 1 10 UNL 1 52 52 51 16 21 5
10 10 36 sl R 51 55 53 81 10 1 1 25 IS 14 58 53 68 01 I 1 20 20 16 61 57 73 01 6
13 101 II R 5! 53 5' 81 07 1 13 5 62 58 18 0! 10
'1 UNL 20 68 51 51 55 03 8 10 24 15 61
It 10 27 '5 57 55 53 87 10 ! 5 UNL 2S ie 56 11 51 02 1 10 60 20 66 61 57 73 03 7
11 1 14 'A 58 51 54 187 06 I 2 UNL 25 13 55 111 56 02 6 10 70 12 64 61 58 Sl 05 5
!?j 101 26 is! I 58 151 54 187 01 3 01 UNL 15 51 53 52 13 00 1 0 10 100 ^2 63 H 60 10 01 5
AUG Ibth AUG 17th AUG 18th
01 1 0j 75 a R 13 II 10 10 01 1 01 UNI IOI 1 11 60 IG | 17121 6 7 250 1 T 51 58 5B 17 23 5
1'
D4i IC! <1 7 12 I'i 60 13 11 5 11 15 I 1 1 55 58 58 21 1 6 :oo 7 57 56 51 17 22 4
17 tOi 55 2 8 LF ll 10 10 17 10 5 IOI 24 1 F 61 10 5S S3 11 3 10 UNL 6 F 58 57 56 S3 21 5
ioi 10 b 7 14 12 11 10 13 I 10 21 12 ! 67 12 58 hi! 24 a 3 UNL 15 70 63 56 66 28 4
nho u 15 15 15 62 71 12 5 10 250 15 { 72 13 57 5SI 11 i 3 UNL 15 75 62 53 4k 24 10
IbllO! 45 15 71 15 12 73 11 5 10 250 15 1 71 65 62 73! 03 i 3 UNL 15 75 11 50 42 27 10
7
11, 3|UNL 18 ll 12 31 13 1 10 250 12 67 63 60 1 78, ll 4 ; UNL 15 70 51 50 41 21
*l
?;! n ioo 10. ll 13 12 iSH '.6 4 5! UNL IOI 13 t? 61 : ill oo 0 11 15 Lis 15 S' 54 M IT 5
MAX ! MU 1 SHORT DURATION PRE : : p i t at i ]N

TIME PERIOD IHiNUTESl 5 10 | 15 20 30 45 bO 80 100 | 120 150 180
PRECIPITATION ! INCHES! 0 44 0 67 0 70 0 72 0 78 0.S8 1.07 i OS '. 1.11 1 11 1 ll

0?|
ENDED: DATE 27 27 27 27 27 27 1 27 p 27 1 27 27 27
ENDED: T IME 1441 1441 1447 1451 150' 1517 i 1525 1534 1534 I 1555 155S I55S
THE PRECIPITATION AMOUNTS FOR IHE 1N0ICAIED HUE INTERVALS MAY

OCCUR AT ANT IIME OURING THE MONTH IHE TIME INDICATED IS IH!
ENOiNG TINE OF THE INIERVAL DATE ANO IIME ARE NOI ENTERED
FOR iRACE AMOUNTS.

August 1992
Page 2 of 4
Figure 1.2
246
AUG ISS2 14768
3BSERVA1I0NS AT 3-HOUR INTERVALS ROC HESTER. NT
VIS V 1ST -
UsT
'T
IE*P[RAIUR[ RIND IENPERAIUR-
NINO TENPERAIURE NINO
.11 Bit ITT BIL IT '
1 ' 1
* | " |
I *
i I-.
I
= "11 .
j"!
H
~
~
= >t
j f !"
-
_
- _
HEATHER -
i - ! NEA'HER s- ! ^.
NEATHEBI 7 1= 1 .
3
; 1
'
=
|~| =
lS '
= 3= = i i ._ 3 lo =3
e , m * ^ 9 0 la. % 1 u <=>
K 1 2 -.
-
\SW
aus nth AUG 20th AUG 21st
Dt 10 10 S 01 RN ' ' : ' 10' 20' 5
12 61 160 13123 | I 5| UNL 15! '56 54 !52 1 BH 27 5 .
0| UNL 15'
'52 50 141
D4 1 100 51 158
'
5' 1
" 57 111! 24 , 6 10! 48 15! 1 56 1 54 1 53
'
101 27 i 5 71 UNL I si :50 I 46 3
10j 21 .
17 3 UNL 6 F 51 '51 56 I 101 33 1 6 ll 55 I5| 1 5' |55 1 54 1 101 25 6 01 UNL 25. : 53 '51 11 1861 23 >
'
6
10 3 UNL 151 ll | 60 56 i 10! 22 , 1 ai 21 20l 1 |64 1 56 ; 50 '
61! 31 5 . 0! UNL 20! ,
16 1 61 i 55 63124 1
13 5 UNL 20 'I 61 53 :53l 26 |' 1 bl 18 201 , |65 | 56 ! 41 5! 31 1! OlUNL l 71 1 61 |53 1 50', 22 5
16 5 UNL 20 66 158 51 '55, 21 8 2 UNL 20| 1 168 I 51 ; 47 1 41: 28 i 8 01 UNL 251 76 1 6! I 53 1451 27 1 5
IS 5 UNL 20 ;62 i 55 ! 11 , 63l 2' ! '54 155 26 5
65 | 58 53 !l5i 31 6 " ill
: 3| UNI 20 : 6 0! UNL 251 1
20'
58 1 54
p ' '
22 6 51 ' 154 i 51 1 11 lB3l 22 151 156 154 !64l 4
55 _li
i 78! 34 B 01 UNL! I5i 5 Ol UNL 25! .
AUG 22nd AUG 23m AUG 24th

Dl 01 UNI 20 55 53 |52 101 22 !
'
5 DIUNLI 121 1 1 57156 55 in: ib 3 31 UNI 121 | 67 ill bO | 78 11! 5 !
D4 0 UNL 20 i 53 | 52 ill 13! 20 ; , I 0. UNI 10! 1 153 ! 53 52 161 20 , 1 5 UNL 10 I 15 1 12 10 81 22 j 5
07 Ol UNL 25 55 153 I5I-I87' 23 I i 55 67! 17 | IF | 84 23 1 7
! 4 OlUNLi 12! |51 1 57 6 10 UNL ll 16 1 63 ll
10 21 UNL 15 10 | b2 3 51 14! 23 '
1 0 UNl IOI | 'lb! 63 141 11 ! 7 10 UNL I H 76 161 15 111 11 I
13 2|UNL 20 18 ; ll | 51 I II, 21 5 0'
UNL !0| , 182 IB 51 46 11 ! 8 10 250 I H 94 1 73 17 157 18! 5
16 01 UNL 20 77 j 66 51 | 511 03 8 01 UNLI 12! : ; 84 |67 56 38 1! I 4 10 UNL 5 H 34 1 71 61 III 22! 5
'
11 01 UNL 20 70 62 : 51 ll! 0! 3 6
2'
UNL! 12 1 i 7) 61 51 54 14 4 8 250 1 H 71 1 72 bl 1 72 00! 0
'
il Ol UNL _!i
i 12 151 !5I 81100 , 0 3! UNL! '1 168 : 64 61 78 22 3 10 65 1 FH ! 75 1 72 I 71 187 181 4 3
AUG 25t AUG 26th AUG 27th

01 I0| 41 6 16 17 ! 67 i 17, 20 ' 1 '
73 1 70 68 1841 22 1 6
| 5 81 '201 5! FH 1 12 ! 70 61 1 10 21 7 101 no B!
04 IOI 85 3 f 70 , 61 | 61 ! 17| 21 5 b: UNL 4| IFH bl : 63 61 | 13 21 1 6 10 II 3 ! F 72 71 70 |14 20 J
07 ll 100 1 8 F 71 70 1 70 I ll! 20 . 5 Ol UNL 1
'
IFH ; 70 68 66 87 18 i 6 10 5 2 1 F 7'
70 70 17 03 5
10 5 UNL 5 H 71 ] , 71 111 23 74 | '
1 01 UNL
6'
H 1 81 | 74 71 1
72 25 8 10 24 < F 75 72 70 85 15 3
13 7| 30 a 84 ! 75 1 II 1 65i 22 : 8 6 11 71 3 | 86 I 74 61 I 51 25 1 10 250 1 ! H 84 76 72 I 67 10 3
16 4. UNL j 82 1 75 1 72 I 72:21 6 61 UNL 8l ' 66 1 75 1 70 1 51 25 | 13 1 80 5 IRNF 71 70 61 113 08 7
:
111 11 UNI 3 H 77 i 73 71 82 21 i 1 81 UNL '2: I 77 | 70 1 61 I 71 23 I 6 6 80 1 IF 71 70 70 117 24 3
22 101 120 4 1 FH 73 71 ! 70 : 10l 18
' 3 1 IOI 250 10 1 ! 73 70 168 184 17 1 1 10 4 2 Bl r 70 1 6S 61 117 031 4
AUG 28lh AUG 21th AUG 30th

Dl 10 1 21 81 f 61 168 168 !17I05 ! 5 SI 47| 10 | 51 ! 56 51 . 81. 23 8 1 UNL 151 1 56 5! 51 63 22 1
04 10 3 1! ,1 68 66 68 00 13 , 5 ll 70! 15 i 5754 | 52 84123 I 10 1 UNL
l5i 5! 53 41 75 21 7
D7 10 60 21 |F 70 111 | 18 11 17 I 51 UNLI 20 1 5754 1 51! 81 23 | 12 8 UNL 20 6! 57 53 7] 18 1
10 10 140 31 |* 78 1 73 1 70 1 11 16 ] 1 81 85; 20 62 1 55 ! 50 65 24 , 11 10 250 20l ! I! 61 56 57 23 15
13 10 100 6: Ih 77 3 7!
'
68 14 18 1 '
10: 12! 25 162 1 55 ,50 65l 26 ! 22 10 50 61 1 TRN 16 63 61 84 23 1
16 10 70 |RF 13 | 11 i 70 1 101 11 3 6: 15! 20! i 65 I 57 50 511 24 : 20 8 55 15l 1 71 65 61 71 23 11
6| .
11 10 80 b, RHf 65 ll i 14 1 17| 23 , 5 51 UNi: 201 ! 160 |54 48 j 65| 24 i 1 2 UNL 121 61 64 61 76 22 14

1 !
15'
ii 10 16 61 If 12 ll !60 ! 131 27 '5 2! UNL , 56 153 51 1 83! 16 : 7 0 UNL 17; 3 66 10 55 I! 24 12

AUG 3lsl
bi 0 UNL 15 60 57 151 I8tl 2! ; ,
D4 0 UNL 15 60 57 |54 181 22 1 .
07 4 UNL 25 61 51 52 J72 25 12
10 7 31 25 62 55 1 41 1 13 26 11
13 8 40 25 61 51 : 41 1 58 21 ! 15
16 6 65 25 65 57 | 50 51 21 ! 13
11 3 UNL 25 | 10 51 148 | 15 28 ! 10
il 0 UNL 15 53 51 1 41 ! 8I1 21 i 5
SUMMARY BY HOURS
AVERAGES RESULTANT
NINO
TEMPERATURE
WEATHER CODES
2
TORNADO 5U SNOH SHOHERS GROUND FOG
THUNOERSTORM SG SNOH GRAINS BLOWING DUST
8L0UING SAND | K
SQUALL SP SNOH ElLETS
ICE'
CRYSTALS BLOWING SNOH z X in
|
RAIN
RAIN SHOWERS
IC
IP ICE PELLETS
IPH ICE PELLET SHOHERS
BLOHING SPRAY
SMOKE
~
! *
FREEZING RAIN
DRIZZLE A HAIL HAZE
F OUST 31 6 JS 46o 61 5$ 57 88 6.4 21 i.i
FREEZING DRIZZLE FOG
04 7 2S.4BQ 60 58 57 so 6.3 22 4,7
SNOH IF ICE FOG 4.1
07 6 2S.50O 61 5S 57 87 7.8 23
10 I 2S.5IO 65 63 5S 70 S3 24 5.5
CEILING: INDICATES UNLIMITED
UNL
HHICh THE HIND BLOHS, INDICATEO
,....,.
13 2S.4SO 73 64 58 61 S.S 25 4.S
HINO DIRECTION. DIRECTIONS ARE THOSE FROM 62 5.8 26 4.5
SOUTH Ib 6 2S.470 73 64 58
IN TENS OF DEGREES FROM TRUE NORTH: I E , OS FOR EAST, 18 FOR 7 2 24 3,2
s 5 25 470 6S 62 58 70
27 FOP HES! AN ENTRY OF 00 INDICATES CAlM 3,7
21 485 63 60 SB 83 i.o ii-
SPEED: THE OBSERVED AVERAGE ONE-MINUTE VALUE, EXP RESSED IN KNOTS 22
'MPH=KNOTS ' 1 15
'Hi 1
New York
Climatological Data of Rochester,
August 1992
Page 3 of 4
Figure 1.3
247
NAllONAl ClIMMIC DMA CtklfJ
TtDERAi BUILDING FIRST CLASS
37 BAIURT PARK AVE
ISHEVILLE, NODIK CAROLINA 28801-2733 DQC
TACE ANO FEES PAID
IT: j
OFFICIAL BUSINESS
PENALIT FOR PR1VAIE US[ HOC
AUG 111 i 14768
HOURLY PRECIPITATION >,a: I R E 9 'J I VAL ENT IN I NC HE Ei ! USC0Hmi -

n'oaV -
asneville. nc 375
A M HOUR ENDING A! P "

HOUR fNDING AT J
! 1 U
'
!
'
3 1 1
i-hi-
1 2 3 * 5 b 7 C
'
1 2 3 & 5 7 8 3 10 11 12 =3
01 0.04 C 02 1 i
1 i ! i 01
1 02
02 i '
1 0 03 0 08 1 0]
03 i I 0 06 0.O2 0 20 0.14 0.04
| i
04 0.0310 01 0 03 0 Oi O.Ol 04
05 1 1 1 1 05
01 '
01
! 1 01
07
08 I 1 : U <?4| 0.11 1 1 1 I 1 I 0 02 0.01 oa
01 c ot 0 oil 1 ; 01
10 1 0.08 0.O4 0 01 IO
! J
II n
12 12
13 ; C 0! 0 0310 08 0 17 0.02 I 1 1 n
i 1 ; 1 11
14
15 i i ' 1 1 1 1 15
16 I I 0.01 0 03 1 1 ; ! 16
1 11
I'I
1 I 1 18
IS | 11
11
20
0.05| ! 0.03 I
1 i j 20
. .
J 21
22
22
21
23
0.15 0.46 24
24
25
25 0.02
1
26
26 i
27 I 0 04 0 8110 30 1 |0 02 21
1 0 03 0 OliO 06 0 24i0 01 0.08 0.22 0.06 1 28
28
' 21
21
T 0,03 1 1 10
30
31
1 31

August 1992
Page 4 of 4
Figure 1.4
248
SEP 1112 ISSN 0118-3660
ROCHESTER. NY
MT'l DEU SEN OFC..MM .1 Of
BREATER ROCHESTER L AIRPORT

INTER'
LOCAL
CLIMATOLOGICAL DATA
ROCHESTER-MONROE COUNTY AP Monthly Summary 'rs
0<
" 77
LAII1 00 43 07 N IONOITU0E 40 H Elevation i.bouno 54' < [I
IINE I0( ASTERN 14766
HEATHER IYPES SNOH l!(IA(..

OF. ORE DAIS Y COYER
TEMPERATURE F ICE PPECIPIIAI10N SIIIIM
HIND
SUNSHINE
BASE 15"T IH P H 1 TENTHS 1
< FOG PELLEIS PIESSUH
x _ x z 1 HtAVT FOG OS <~
IH C3
PEAK FASTEST
-
: THUHOERSIORH cr ON l>C HIS GUST I-H1N
2 5
-
~2 i i |l ICt PELLEIS GROUND a _
3 O
Al
I. 1 ElEV
I I 1 i o
2e
0700 555 Ln
X
13
| I S
5
li
3
INCHES il '
FEET
ABOVE
1 2
~
X
X * * O ^ - ll BL0HING SNOH 5- -
! n s l K Of ; a C3 Ln o E ct 5.
2 3 5 b 7A '3 ! 8 S 10 1'
12 13 14 1 15 16 17 IB 11 20 21 22 71
01 68 47 58 41
-8
i 0 0 00 0 0 25 580 30 4 4 6 n 20 NU 12 31 743 14 4 4
48 62 -4 53 3 o ! 0 T 0 0 21.570 15 4 6 12 F 1 13 712 10 1 S
61 3 65 3 4 I 1 0 0 62 0 0 21 370 20 6 5 24 SH 16 24 160 20 1 1
53 64 56 i 0
05 81 50 66
-2
l 61 0
j 1 0 0 00 0 0 21 660 11 1 1 15 N 1? 14 745 15 I 6
'1 0 0 00 0 0 21.710 13 5 4 15 S 12 17 674 87 2 3
06 77 65 71 6 62 0 6 | 8 0 1 0 0 21 630 17 8 5 2? S 1? 16 315 51 B
81 64 73 8 65 0 1 3 8 0 0 II 0 0 21 560 11 8 1 18 s 11 IS 447 SB 8 '
ua 81 57 61 4 63 0 1 0 0 02 24 477
c\? 0 0 21 415 22 8 5 12 4 16 SH 25 62 8 b
us 78 52 65 3 6/ 0 3 c T 0 0 2S.480 16 4 1 20 s 14 18 673 8R 5 6
IU 80 56 68 4 61 0 3 8 0 1 0 0 21 310 22 8.1 10.2 32 NH 21 31 554 73 1 7
2>- 11
12
63
67
45
42
54
55
-10
-B
4'
4b
"
10
0
0
0
0
0 00
0 30
0 0 21 630 ?R
0 0 21.780 17
7
? n
1 21
14
NU
N
15
B
21
01
485 64
740 18
6
4
5
4
IJ 74
z 80
45 60 -3 47 5 0 0 0 00 0.0 21.720 IH 4 6 11 S 13 20 756 100 8 7
52 66 3 55 0 1 0 0 00 0 0 21.640 73 5 5 15 SH 12 11 741 11 7 6
CL 15 82 51 71 1 .62 0 6 1 8 0 0 00 0.0 21.610 22 6.1 17 Sri 12 27 652 S7< 3 5
16 83 64 74 12 64 0 1 1 8 0 0 00 3 0 21.550 22 6 6 21 SH IS 28 681 12 1 B
\l 83i 66 751 n 63 0 10 8 0 2S SH 23 23 716 16 1 6
8 7 10 2
i/jh IB 81 58 TO 8 62 0 5 1 3 8 0
0
0
00
56
3 0 21.310
0 0 21 225
22
21 10 4 11 6 40 SH 21 .25 612 83 1 8
E- IS 62 47 55 46 10 0 IS 21 728 11 4 2
-6 0 0 00 0.0 21 420 27 6 5 IB H
20 61 45 57 47 6 0 0 0 00 0 0 21 500 IB 3 3 IB NE 10 04 717 IB 1 4
Ul -4
21 68 56 62 1 60 3 0 1 0 0 11 0 3 25.300 17 1 1 21 S 17 16 0 0 10 10
22 71 41 60 0 51 5 0 1 0 0 34 0.0 2S.21C 25 7 1 11 4 26 NH II 31 0 0 10 10
23 51 38 45 -15 .15 20 0 0 0.00 0 0 21 73C 12 7.4 21 N 17 36 725 100 i 3
24 51 35i 47 -13 37 IB 0 0 0 00 0.0 21. IOC OB 3 5 16 NE 12 02 724 100 0 3
O 25 65 36 51 -B 31 14 0 0 0 00 0 0 21.800 10 7 6 18 E 14 01 720 100 3 4
O 26 65 53 5S 0 55 6 0 1 0 0 27 0.0 2S.5T0 11 6 1 15 F 12 01 27 4 10 10
a. 21 15 52 61 2 56 4 0 1 0 0 11 3 0 21.37C 23 1 5 11 6 41 SH 30 25 211 42 6 6
20 bl 45 57 -l 47 8 G 1 0 0.00 0.0 21.481 24 8 0 11 N 23 21 702 11 2 3
21 50 43 47 -11 17 18 0 0 T 0.0 21.615 28 11 6 14.3 32 U 20 26 157 22 to 1
30 50 37 441 21 0 0 1 0.0 21 640 21 7 8 8.1 25 16 33 281 40 1 B
-13 35 H
*
SUN Suit TOTAL 101AL TOTAL 101AI (0 IHE 1 IHIH: TOTAL X SUN SUN
HUH8ER OF AYS
2131 1522 172 57 3 02 0 0 21 55C ?? 4.0 8 1 41 1 SH 33 1 25 1605S IH 181 iSo
AVE. 116 AIE OfP AVG OEP SEP PRECIPI1III0N 0t IAlt:f7 1A1E: 27 mum Mill Alt AIE.
71 0 50 7 60.1 -1
'
5,- '
40 a .<
INCH 01 8 3 36 22414 7i I 3 It
SEASON 10 DATE SNOH. ICE PEL CIS MEAICS1 OERTH ON SNOUNO OF
NUMBER OF DAYS GAF TES1 IN 21 HOURS A ID OAKS
'0IAL 10'Al i ', 0 INCH 0
SNON. IC PELLEIS OR ICE AND 0A1
mil UN IEHP IN NUN I HP 244 308 THUHDEHS10RH5 PREC1P! IA1I0H 1 SNOH. IC PELLEIS
<. tl'
' 0
so 32
V | OE" HE1YT TOG 0 20 21-22 3.0
i
[ -. < OCP i
0 i 0 ( I 0 71 -221 CLEAR ' BAfitLI 10UDY 1 CLOUDY 14
EXTREME FOR THE MONTH LAST OCCURRENCE IF MOBE THAN ONE DATA IN COLS 6 ANO 12-15 ARE BA5E0 ON 21 OR HOSE OBSERVATIONS
I TRACE AHOUNT. AT HOURLY INTERVALS. RESULTANT HINO IS THE VECTOR SUM OF HINO
>
ALSO ON EARLIER DATE (SI .
SPEEDS ANO DIRECTIONS DIVIDED 8Y IHE NUMBER OF OBSERVATIONS
HEAVY FOG: VISIBILITY 1/4 HILE OR LESS COLS 16 I 17: PEAK GUST HIGHEST INSTANTANEOUS HINO SPEED -
BLANK ENTRIES DENOTE HISSING OR UNREPORTED OATA ONE OF THO HINO SPEEDS IS GIVEN UNOER COLS 18 I 11: FASTEST
NILE HIGHEST RECOROED SPEED FOR HH1CH A HILE OF HINO PASSES
-
STATION iDlRECTION IN COMPASS POINTS] FASTEST OBSERVED ONE .
MINUTE HIND HIGHEST ONE MINUTE SPEEO

-
IDlRECTION IN TEHS OF
DEGREESI. ERRORS HILL BE CORRECTED IN SUBSEQUENT PUBLICATIONS
FROM
I CERTIFY THAT THIS IS AN OFFICIAL PUBLICATION OF THE NATIONAL OCEANIC ANO ATMOSPHERIC ADMINISTRATION ANO IS COAPILEO
RECORDS ON FILE AT THE NATIONAL CLIMATIC OATA CENTER
Alw^C ) 'Z}/lullA~'
/jr^w~~~
_ _ NATIONAL NATIONAL NATIONAL

DIRECTOR
#% ^"V O O OCEANIC AMD ENVIROHCNTAL SATELLITE. DATA CLIMATIC OATA CENTER
NAT I0NAL CL I HAT I C OATA CENTER
I W\P4\ K\ ATNOSMPIC AOH1HISTRAIION A* [HFORWTION SERVICE ASKVILlE NORTH CAflaiNA
J
September 1992
Page I of 4
Figure 1.5
249
SEP 1112
OBSERVATIONS Al 3-HOUR INTERVALS ROE HESIFR Nl
14 76!
V1SI-
YISI-
Sill IT TEHPERAIURE HINO IERPERAIUR! HINO ItNPERAIURE HINO
8ILIIY BH 1Y
-1
| HCIIHIR
-
=; z z ~ = 1 1 1 ;=
I i
HEATHER
1 =
i
| I
HEAIhER
1 m
o
| |
=
-
s 1 2 s 3 ft 5 =i
5 "
~
5 z
=
5 5 |=l t 5
-
ft
SEP l>i SEP ?,nl SEP 3rd
Dl OlUNL IS 11 11 46 16 12 5 t ONI 16 17 41 60 1 00 i
11| 0 IOI 110 121 ! 64 62 1 60 1 87 lib
114 1 UNL 15 11 46 41 11 DO 0 i m 16 41 11 46 1 It! 00 0
Rh.'
17 1 1' 13 7
10 60 7 8| 63 12
3' 31 UNL 20 52 60 41 10 24 i UNI 17 Sl 17 51 |
111 20 1 10 31 21 8. RHF 17 66 1 65
| 11 111 1 I
ll 10 25 62 61 42 10 13 10 UNI 70 17 10 55 It 11 6 >1 68 bl 6 7 17 11 1
10 ! RHF
13
Ib
51 UNL 25 66 67 tl 44 Jl 10 UNI 76 14 10 41 | 41 16 1 10 II : !rh 71 161 66 10 70 1
21 UNL 20 66 56 8 41 11 10 1 740 76 14 53 148 17
11 6l 80 20 61 51 46 63 12 10 740 16 bl
b?
11 61 18 1?
b 01 110 17 | ll 1 68 66 14 '1
10 71
10
b 71 UNL in 61 17 16 4
51 UNL 16 V 52 51 11 00 10 no 16 RH 14 : 10 ! 57 l 18 14 b ll UNL 6 66 1 64 14 1! 71 4
SEP 4th SEP 5th SEP tth
Dl 81 27 1 8|F 12 62 12 POO 24 4 1
| 0 UNI 17 53 153 52 I Iti 14 1 1 0| 260 fl 6) 64 67 84 14 1
Si UNL 12 10 58 41 1 10 31 6 0 UNI >0 51 51 60 16 00 0 14 67 64 1? 7 :
'
1 0| 240 I)
16 61 57 66 ! 10 26 4 0 UNI 70 ! 65 53 52 10 00 n IOi 65 6 68 64 bl 64 14 1
26 ll 61 51 68 21 0 UNI 26 i 61 !62 67 16 11 4 10 70 14 70 64 61 11 IR 1
I UNL 26 1 73 66 41 62 36 0 UNI 14 11 18 62 46 10 R 10 100 70 77 th 61 11 17 8
26 i 70 61 44 41 01 10 3 UNI 17 81 j 71 64 18 II 1 1 UNI 74 74 17 tl 66 11 10
ll UNL 20 62 61 42 10 13 10 UNI a | 73 67 64 11 11 fl 1 UNI 70 61 14 I? 71 16 10
tl SlUNL 10 66 56 44 17 II 10 UNI 7 67 15 61 RI 14 6 12 ! 66 hi tl 14 17 7
SEP 7lh SEP 8th SEP 1th
gi 4 UNL 1? 65 62 to 84 18 7 0 UNI 7 44 11 00
17 61 17 RI 18 8 0 IIHI 151 66 64 0
2 UNL 12 65 62 to 84 18 S 0 UNI 7 17 64 62 84 11 1 6 14 15 47 47 61 16 71 1
07 6 60 10 61 64 12 84 11 1 10 80 1 70 61 71 11 44 44 41 Ik 00 0
61 1 4 UN 16
UNL 8 16 61 61 74 20 10 a 10 IC RH 11 70 66 71 IR 11 1 UNI 70 16 41 44 66 14 6
11 111 75 7 ao 72 tl 6' 11 11 10 35 1 60 71 67 65 70 16 1 UNI 70 71 64 48 10 14 6
ll 8 80 6 IH 11 71 tl 61 II II 3 UNI 14 17 61 64 64 74 16 6 UNI 75 7R 66 41 47 18 4
11 1 86 2 8 RNF 71 70 tl 1] 14 4 a 37 14 66 64 61 10 71 1 in 70 14 71 66 63 16 11 6
li 1\ F 61 ll 64 81 18 6 4 UNI 14 62 kO 48 R7 11 1 1 1 fl 8 RH 71 67 64 11 II 17
_LS
SEP 10th SEP 11th SEP I2lh

Dl 7| 120 IC 11 tt 14 64 16 10 li 250 14 46 41 52 67 77 4 0 UNI 14 44 143 47 11 74 6
D4 ! UNL f 70 II tt 81 18 11 4 UNL 14 51 11 60 10 71 4 7 UNI 14 41 41 47 16 II 4
D) 8 250 8 71 11 tt 71 20 10 7 31 74 61 47 5i 10 77 k 7 UNI 15 44 41 47 13 1! 4
10 S 31 to IB 70 tt tl 21 1 1 3C 24 sa 57 t7 77 4 UNI 44 60 46 72 16 1
47 1 70
IJ 1 121 2( 12 tt 13 7i 24 11 ti 32 24 61 41 lb 6! 11 11 1 IIHI 70 61 61 13 41 77 6
ll II 11 26 ll to 41 7! 23 1 li URL 74 60 61 16 61 17 1 1 UNI 74 64 66 44 41 00 0
11 I UNL 2C 13 to 46 SI 22 1 01 UNL 74 61 46 II 71 76 6 7 UNI 74 46 40 44 67 07 4
tl i UNL 16 58 51 44 8) 11 1 Oi UNL 15 11 46 II l<
76 5 1 UNI 15 41 41 46 61 00 0
SEP 13th SEP 14th SEP 15th
11 10 UNL 12 46 48 47 16 11 4 1 UNI 14 64 57 11 80 18 4 0 UNI 6 F 60 41 68 13 17 4
Dl 1 UNL IC 41 47 41 16 20 4 1 UNI 14 61 51 11 Ri 71 6 7 ll( 6 F 61 51 Si 11 18 4
Di 2 UNL 25 1! 47 41 li 20 3 UNI 71 54 67 60 R:
70 8 II UNI 6 F 61 41 41 11 18 4
II !< UNL 25 IJ 61 40 6i 18 4 l( UNI 14 61 60 54 6'
1) 7 t UNI i II 66 62 li 11 7
li 6 UNI 26 11 41 41 46 14 4 11 UNI 1? 11 66 41 li 77 1 7 UNI i 80 10 66 61 21 10
11 II UNI 25 11 51 4] 74 11 4 1 UNI 12 77 64 5H 67 71 4 1 UNI 1 81 70 64 61 24 8
1'
: UNL 26 14 56 48 66 17 4 II UNI 1 67 67 46 7! 06 6 i 11 ; 74 61 65 71 22 1
I?? 6 UNI 16 66 62 SO 6'. 11 4 i UNI 61 41 46 11 II 4 II 71 i 71 68 66 P ll 4
SEP 1 b IK SEP 1 7 Ih SEP lilh
II IOI 140 6 M tl tt 64 84 21 4 10 750 10 61 66 61 84 74 8 0 UNI 7 67 65 63 67 11 7
8'
11 5 UNI 6 F 66 64 63 11 70 4 II 761 l( 68 64 17 16 8 7 UNI 6 F tt 64 62 1. 11 1
fl' 8'
li II 251 6 F 61 15 61 8i 70 4 II 741 4 H 67 61 61 11 1 1
741 6 H 18 64 62 20 1
7'
1( 1 UNI 5 H 76 70 66 11 11 h 1 UNI 1 H 76 68 61 I'
71 1 1 111 7 75 61 64 18 1
6'
6'
r '
111 i 81 17 61 21 10 1 11 1 60 70 61 41 71 16 1( 41 1 R 77 70 61 20 14
'
n UN! 1 60 70 64 41 74 17 7 UNI l 87 17 67 6 76 6 II 11 11 80 61 6! 51 21 13
i<
; UNI 6 H 74 66 61 6i 70 1 1
61 71 ta 64 7 78 11 11 44 11 RH 73 61 64 74 24 IS
7' 11
ii i_J UNI II 61 13 41 71 II 5 1 UN^ 1 70 tt 13 ii I II 41 1 RN .Ji. id tt 3? 11)
MAXIMUM SHORT DURATION PRECIPITATION

IIME PERIOD IHINUTESI 5 10 IS 20 30 45 fed 80 ! 100 120 150 180
PRECIPITATION ( INCHES ' 0.08 0 16 0 24 0.32 0 38 0 40 0 43 0 47 0 50 0.51 0.54 0.57
ENDED: OATE 2'

21 i 21 21 21 21 21 21 21 21 21 21
ENDED: HUE 1038 1038 1038 1038 1038 1036 1038 1038 ! 1038 1114 1l2fe 1202
IHE PRECIPITAIION AMOUNIS FOR THE INOICAIED IIME INTERVALS Ml)

OCCUR AI ANT TIME DURING IHE MONTH IHE HUE INOICAIED IS IH!
ENOING IIME OF IHE INTERVAL OAIE AND IIME ARE NOI ENTERED .
FOR IRACE AMOUNTS

September 1992
Page 2 of 4
Figure 1. 6
250
IUG 1112 14716
OBSERVATIONS AT 3-HOUR INTERVALS R0CHES1ER. Nl
VISI-
| HI!
'Bill l",!
lERPERIiURT. A U I NO
-1
BH II' jttRPERHuAE | NIKS
1 .1
VIS
SH -t|
TEMPERATURE ! ''"
!
1 '
J | '
1 L !IHUIHEB
j
-. _ 1
i
1 =
HE4IHER
!
as
ill!
'- 1
1
- ,
, 0
I -*
!-!
1 m
1-
*
- 1 It ! - 1 -
3 a 1 = I '
m la. 1 x
>
w 0 a , <n \a. =
L 1
.
i 1j S
-
- ^s s
11 -
^ .- '^2 is!*:
'
a 1 '-= 1 ',
-
i | AUG llth AUG 20lh AUG :'l

tail 10 10 10 | RH 62 161 60 1131 23 | h IIUNll 151 ; 46 41 152 187-2' 5 0: UNl 1 L 42 40 i 41 101 20 5
' 100 12 51 |5! 57 1 13| 24 t 101 48! 15! 46 44 I 43 i 10 27 1' 4'
50 43 ': 1
fiil ] UNI 6
| F 51 57 56 |10! 23
3
'
6 11 55i '4i
,
41 55 54 10: 25
'
1 UNL '5
|
5! ! 51 11 it: 21
r.\ : ,
6 11 UNI
1Ol 3 UNl 14 61 160 66 | 70 22 1 1 21 20l ! 64 66 1 50 i tn 31 '
5 13 11 55 bl, 74 <
, , Ol UNl 20
5 UNL 20 ! 71 bl 53 1 53! 26 1 tl Ifli 201 56 ! 41 '
73 H 53 501 22 5
13|
161 6 UNL 20 68 58 51 I 55121 ! B 2 UNLI
.
ill
b5
51 14?
:5b. 11 '1 01 UNL 25
't 13 ! 53 15177
20! .
47| 28 ; 8 3. UNL 25
11! 1 UNL 20 65 58 53 1 651 31 ! 551 7b
8 31 UNl 201 i 162 55 1 11 itll 27 t Ol UNL 25 M ;
ll .54 6
^21 hi 44 15 i 48 I 44 ! 4! i 761 34 ' 8 01 UNL: 151 '
'54 51 1 11
'
111 22 . 4 Oi UNL 25 51 51 1 54 84:70, 1
AUG 22nd AUG 23m AUG 21 th

bi o UNL| 20 54 '53 i 52 | 101 22 1 5 0! UNLI 12! ; |57 41 i 45 | HI 18 3 3 UNL !2 b7 | 13 oO 781 1115
Cl 0 UNL 20 i 41 | 42 | 41 | 11 20 1 6 01 JN! 10! i5i 53 :52 111 20 1 1 5 UNL 10 66 112 to Hi 22 1 5
D7 54 43 4i 1 87! 23 D 1 UMLl
5'
0 UNLI 24 I ! 4 12 151 155 : 671 17 | 6 10 UNl t F tt It] H 841 23 1
70 ! 12 |67 1 64| 23 ! 1'
10 2 UNLI 14 66 1 13 bl 11 1 11
UNL! 20
; 78 ' 64 44 ,441 21
. 4
4
Oi JN! 10!
i 162
76
68 ' 51 j lb 11
'
7 10 UNL 6 H 76
84
'
bl
73
15
67 157' 181
t
5 I
,
'? 10
:I6 UNLI 20 I 77 |66 41 54103 ! 8
01 UNL
01 UNL'
10;
12 184 bl St 161 11 3
8 IOI 250 b
5
H
64 74 11 bl! 22 5 I
4 : 01 UNL 1" |
111 0 UNL 20 70 162 56 I 611 07 1 b 2luNl! 12! 1 17'
tt 151 54! u ! 4 8 250 1 H 71 I 72 11 72100! 0 {
il 0 UNl! 15 i 62 151 1 57 1 84100 .
0 12'
3, UNL! ! !68 tl I tl 3 Ifll 2? '
3 10! 15 4 :FH 75 ! 72 7'
871 '.8 4
AUG 24th AUG 26lh AUG 27th
tail ioi H| 8 1 68 1 67 67 1171 20 6 8|I20| 51 IFh ! 72 70 Itl '10121 1 7 10 110 S 71 70 168 84 221 t |
jr
'
114! 101 85 3 f ! 70 161 I

61 1?l 21 4 bl UNLI 1 |FH ill 18 I 17 111 21 | t 10 11 1 72 71 | 70 14 20 i 3 :
iv il ion 1 |f 71 i 70 70 171 20 4 18 1 tt I 87 18 ! t ! t 11 10 70 17 03 4 I
, 0, UNl, 4 FH | 70 10 5 2
4 771 23 ! 12 1 70
'
101 5|)l UNL ! X 71 74 71 4 0! UNL bl H |8I 74 ; 71 1 72W5 1 6 10 24 4 ,

T '5 85 14 3
In; 30 6 1 84 j 75 1 71 66! 23 1 8 bl 411 71 ! 'It 74 itl 157125
'
1 10 240 4 H 84 76 1 72 67 '0 1
ll 4. uNl 1 62 1 75 : 7! 72l 21
6 tl UNL! 81
,
et 75 70 151121 1 13 1 SO 5 jRHF 71 70 61 13 08'
7
111 1|UNlj 3 IH 77 1 73 1 71 B2i 24 ! 1 al UNLI 121 '77 70 | 17 71 23
1 6 8 80 1 F 71 70 j 70 17 74 3 .
22! 'Ol 170! 41 IFH 1 73 I 71 1 70 JlOilfl L 4 1012501 IOI ! I 13 I!

H:II,l'
4 10 _L
bIf 70 61 HI SL 03 1 4
AUG 26th AUG 21th AUG 10tn
Oil 101 ll 21 81 F 61 i 68 163 117 01 1 4 11 47! |51 56 153 81123 ! 1 3 UNL 14 56 51 51 83 22 1
I F
101 |
041 10! 3 : 68 | 66 168 n 00 13 ! 4 7! 701 151 , 57 54 57 84 71 10 1 UNL 14 57 53 41 75 21 7
07! IO1
601 2! ! F 70 61 168 I 13 17 1 7 51 UNl! 20l 1 '57 1 54 I 51
81 23 12 a UNL 20 62 51 51 73 18 1
101 101 MCI ' I !h 76 i 73 | 70 ! 77 16 ! 1 81 85 20i i 12 1 55 | 40 65 24 11 10 240 20 '
72 63 51 57 23 15
13,!0 100| 61 I H 77 1 71
'
66 I 71 16 i 1 IOI 321 25| i 12 155 ,50 65 2t 2! 10 40 t IRN 66 63 tl 84 23 1

16! 10! 70 61 : RI 73 j 71 I 70 110 11 3 6 111 20! j |66 67 ! 50 51 24 1 20 8 44 14 1 71 66 tl 71 23 It
111 I0{ 60! 6. 1 RHF 65 ! 61 ! hi 1 17 23 ! 5 4! UNl 20l !
'
> 60 1 54 1 48 15 24 I 1 2 UNL 12 61 64 tl 16 22 14
22; 10! 161 6 : 62 1 61 i 60 131 2' , : 15 2! UNL. '41 i 5o 1 13 ! 51 183! 18 I 7 0 UNL 12 tt 1 tO 155 Li! 24 12
AUG 31s
Sill 01 UNI 14 1 ' 60 157 !5l 61 2' i 1
4| 01 UNL 1'
1 | 60 ! 57 154 1 81 22 ! 1
6;i 41 UNL 2'
! 156 152 | 72 25 i 12
11
61
ilO! 7! 31 2'
62 55 1 48 63 26 1 11
1], 8! 40 2'
I I 64 I 56 i 41 |56l21 : 15
2'
lb'
6! 66 (65 ! 57 | 50 1511 21 13 '
'
|11i 31 UNL 2'
j 60 : 54 46 651 26 10
]
22: 0: UNLI
1'
i 152 5! 141 '86124 ; 5
SUMMARY BY HOURS
AVERAGES RESUL1ANT
WEATHER ijU l. 0
il TEMPERATURE
S 5
1 TORNADO SH SNOH SHOHERS GF GROUND FOG
T THUNOERSTORM SG SNOH GRAINS BL 3l.uu.nG OUST
IJ SQUALL SP SNOH PELLETS BN BLOWING SAND
R RAIN IC ICE-
CRYSTALS as BLOWING SNOH
au RAIN SHOHERS IP ICE PELLETS BY BLOHING SPRAY
7R FREEZING RAIN IPH ICE PELLET SHOHERS K SnOKE 1 " - " 0 " = "*
DRIZZLE A HAH. H HAZE

OUST
fl'
6 51 4 ft. hi S<! 5J 46 6.4 21 4.4
71 FREEZING ORIZZ'.E F FOG u
04 7 25.481 HO 58 57 10 6 3 22 4.7
SNOH IF ICE FOG
07 fi ?S 401 hi 5S 57 87 7.8 23 4.1
111 25.511 hS hi 45 70 S 3 24 5 S
il 7 2S 451 71 64 58 hi 5 1 25 4.S
HIND DIRECTION: DIRECTIONS ARE THOSE FROH UHICH THE HIND BLOHS. INOICATED
'6 fi 25 471 71 M SB 62 5.8 26 4.5
IN TENS OF DEGREES FROH TRUE NORTH: i.E OS FOP EAST. 18 FOR SOUTH .
IS 5 25 4 7C 61 62 58 70 1 2 24 1.2
21 FOR HEST AN ENTPf OF 00 INDICATES CALH 3.7
ONE-HINUTE VALUE. EXP RE5SED IN KNO.S 72 ! 5 125.485 JtL 60 58 81 i.o 22
SPEED: the OBSERVED AVERAGE
:HPH=KNOTS < I '5:
lt( 1
Climatological Data of Rochester,

New York
September 1992
Page 3 of 4
Figure 1. 7
251
NAIIONAL CLIHAHE DA1A CENIES
FEDERAL BUILDING
17 8A11ERT PARK AVE FIRST CLASS
SHV I LLE NORTH CAROL IMA .
28801-2733
POSTAGE ANO FEES PAID
OFFICIAL BUSINESS NOAA
PEHAUT EOS PRUAlf USE 1300
PFRMIi G-11 i
SEP 1112 14768

HOURLY PRECIPITATION '" mATEfi
innii.ii EQUIVALENT
luumulliii IN INCHES!
I iy
1 11 i. n L 0 1
""Chester, nt
I mi-nun -
utUA -
itur v 1 1 i r iir 7H
^3 A.M HOUR ENDING AT PM HOUR ENDING AT J
cz i 1 2 ; 3 4 5 fc 7 6 <i ; 0 ! : ! .2 1 c 3 4 5 fc 7 8 9 10 ll 12
a:
=2
01
i
01
02
i I 1 02
03 i 0.03 0 04 0 03 0 05 0.01 0 02 0 07 0 10 0 07 0 17 0 01 i T 0.02 03
04
04
05
01
06 I 1 1 I
07
06
| i i 1
0 01 0 06 0 02
06
07
0 02 08
01
I i T I 01
IO i
T I i 10
II
11
12
12
13
tl
14
14
15
15
It
16
1 1
17
0 06 0 17 0 71 0 1? 1 IR
11
11
20 20
21 I 0.02 0 10 0.40 o.ot 0 04 0 05 0 10 0 01 1 0 01 0 0? 0 04 0 04 71
12 i i 1 1 0 20 0 OS 1 1 0 01 0,03 I 0 01 72
23 71
24 74
24 24
21 I 0 04 0.08 0.08 0,02 0 03 0,01 I 76
27 1 0 02 0 06 T I 0 06 0 04 0 01 0.01 77
28 28
21 1 1 21
30
1 1 I 30

September 1992
Page 4 of 4
Figure 1.8
252
Appendix J
Tables and Figures
of
DANTEC LASER DOPPLER VIBROMETER
253
Tracker Freq. MulL
TABLE J. 1 Vibration DC Calibration
Range Shift
Frequency Offset Factor
Range [volts]
~ " "
Tracker Unit -
1-10 MHz 5 XI 000 0.74 MHz ci-
and
5 316- i ra/sec/volt
MHz
.3-3.3 2 X1000 240 kHz io-
6 01 104*
m/sec/volt
Frequency .1-1 MHz 5 XI 00 74 kHz 5 31.6 10 3m/sec/volt
Shifter Unit 33-333 kHz 2 X100 24 kHz 601 10.4 10 Wsec/volt
10-100 kHz 5 X10 7.4 kHz
Settings 5 3.16 10 Jm/sec/volt
3-33 kHz 2 X10 2.4 kHz 6.01 1.04-
10 "'m/sec/vol
1-10 kHz not used
Vibralion Max. Doppler Max. Calibraiion

Frequency Frequency Slev.'
Rale Factor
TABLE J.2 Range
0.74 kHz 10 kHz 210 kHz/sec 0.3164- 10

m/scc/Voli
2.4 kHz
Operational Data 33kHz 630 kHz/sec 1.04 -
10 -J
m/sec/Voli
7.4 kHz 100 kHz 6.3 kHz/msec 3.164 A 10 -3 m/sec/vol!
for the 24 kHz 333 kHz 63 kHz/msec 10.4 A 10 -3 m/sec/volt
74 kHz 1 MHz
Laser Vibrometer 630 kHz/msec 31.64A10-3mAec/voll
240 kHz 3.3 MHz 6.3 MHz/msec 104 A 10-3mAecA'oll
0 74 MHz 10 MHz 63 MHz/msec 3I6.4A10-3m/secAolt
Range Filter
TABLE J.3 BW1 BW2 BW3 OUT
1-10 kHz 19 Hz 59 Hz 194 Hz 27 kHz
3-33 kHz 19 Hz 59 Hz 194 Hz 27 kHz
Tracker Unit 10-100 kHz 19 Hz 59 Hz 194 Hz 27 kHz
Settings 33-333 kHz 59 Hz 194 Hz 509 Hz 27 kHz
.1-1.0 MHz 194 Hz 509 Hz 1.9 kHz 27 kHz
.3-3.3 MHz 590 Hz 1.9 kHz 5.8 kHz 27 kHz
1-10 MHz 1.9 kHz 5.8 kHz 19.4 kHz 27 kHz
FIGURE J. 1
Operating Window
Maximum Doppler
of the frequency
Dantec Laser Doppler Vibrometer
(Frequency vs. Amplitude of

Harmonic Motion)
utmum laser
coherence
254
I0"*10-*IO-,IO-*10-M0""n0~1l0"1l0-1
I a (m)
Appendix K
Mobility Measurements
from
"A State-Of-The-Art Assessment of Mobility Measurement Techniques

"
-
Results for the Mid-Range Structures (30-3000Hz)
by D.J. Ewins and J Griffin
255
Figure K. 1
Example I. Variation of the Superimposed

Frequency Repsonses
-
1 1
1 1
-kk>
"
1 1 1
(d)
-JOO
-
-
-400
mi JfflWl
-500
-*oo
7-î^ /jRv
-
-TO-0
->0
\l
-0-0
-IOO-0 -
-
-IIO-0 -
-120-0 1 1 1 1 1 11 1 1
t-OOOilO' *OOOilO'
JOOOilO* lOCOim1 10' 10' K>'
-4-000i MOO i 6-000.
Frequency (Hz)
Figure K. 2
Example 2. Variation of the Superimposed

Frequency Repsonses
O00il0' ÔOOilO' -000iK>

10' l-OOOilO1 JO00ilO
tOOO I B-000 l 10
Frequncy (H;)
256
Figure K. 3
Example 3. Variation of the Superimposed Frequency Repsonses
400 -
Frequency (Mi)
Figure K.4
Example 4. I 'ariation of the Superimposed Frequency Repsonses
2 -90-0 -
t-0O0il0*
Fr*4uncy (Hz)
257

Application of Sensors in An Experimental Investigation of Mode D

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Application of Sensors in An Experimental Investigation of Mode D

Загружено:

Авторское право:

Доступные форматы

Rochester Institute of Technology

RIT Scholar Works

Application of sensors in an experimental

Follow this and additional works at: http://scholarworks.rit.edu/theses

Richard Hiram Berg

Dr. Richard G. Budynas, P.E. (R.I.T.)

Dr. Guillermo H. Garzon (Eastman Kodak Co.)

Dr. Wayne W. Walter, P.E. (R.I.T.)

Dr. Charles Haines - Department Head (R.I.T.)

Department of Mechanical Engineering

I, Richard Berg, hereby grant permission to the Wallace Memorial

Library of the Rochester of Technology to reproduce my thesis in

August 10, 1993

I dedicate this work to my parents Richard and Rita Berg, my wife

Laura and my uncle Peter Grimard. Their great encouragement and

devoted support will not be forgotten.

Rochester Institute of Technology, David Hathway of the Rochester

The subject of this investigation is the method of calculating material damping of a

cantilever beam. Damping is the mechanism by which the vibrational energy is

damping factor, is part of the global properties of a structure which can be

identified from its frequency response measurements. Experimental frequency

different types of exciters and sensors on the frequency response measurement of

Discussion of the results, comments and recommendations about the different

techniques are presented.

Table of Contents iii

List of Tables xxiii

2.1 Characteristics of Damping 4

2.2 Environmental Effects 6

2.3 Material Selection 10

2.4 Method of Calculating Material Damping 14

2.4. 1 Damping Measurements for a 1-D model

2.4.2 Decay Constant 20

3.1 Modal Testing 24

3.2 Fundamental Functions of the Bruel & Kjaer

3.3 ASTM Standard Method for Measuring Damping

4.1 Laser Vibrometer 33

4.1.2 Technical Background 33

4.1.3.1 Technical Preparation 36

4.2.2 Technical Background 49

4.2.3 Experimental Setup and Operation 51

4.3.2 Technical Background 55

4.3.3 Experimental Setup and Operation 55

4.4 Piezoelectric Accelerometer 60

4.4.2 Technical Background 60

4.4.3 Experimental Setup and Operation 63

5.1 Impact Excitation 66

5.1.2 Technical Background 66

5. 1.3 Experimental Setup and Operation 66

5.1.4 Experimental Procedure for Calculating the 69

5.2 Random Excitation 73

5.2.2 Technical Background 73

5.2.3 Experimental Setup and Operation 75

5.2.4 Experimental Procedure for Calculating the 76

5.3 Other Techniques 79

6. 1 Experimental Determination of the Loss Factor 82

6.2 Noise Sources in the Experimental Testing 93

7. Comments and Recommendations 99

A: Analytical Determination of Natural Frequencies for

B: Impact Excitation Data of Aluminum from Sensors 109

C: Random Excitation Data of Aluminum from Sensors 157

D: Impact Excitation Data of Laminated Composite

E: Identification of Artificial Damping by the Shakertable in

Rectangular Window and Pseudo Random Excitation