NPS Thesis - Fitzpatrick

This document was downloaded on July 15, 2015 at 15:22:29
Author(s) Fitzpatrick, Sean M.
Title Source development for a seismo-acoustic sonar
Publisher Monterey, California. Naval Postgraduate School
Issue Date 1998-12-01
URL http://hdl.handle.net/10945/8984

DUDLEY KNOX LIBRARY
NAVAL POSTGRADUATE SCHOOL
MONTEREY, CA 93943-5101

!
MONTEREY, CALIFORNIA
THESIS
SOURCE DEVELOPMENT FOR A SEISMO-
ACOUSTIC SONAR
by
Sean M. Fitzpatrick
December 1998
Thesis Advisors; Thomas G. Muir
Steven R. Baker
Anthony J. Healey
Approved for public release; distribution is unlimited.

REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average ! hour per response, including the time for reviewing instruction, searching existing data
sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other
aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and
Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188)
Washington DC 20503.
1 . AGENCY USE ONLY (Leave blank) 2. REPORT DATE
December 1998
3. REPORT TYPE AND DATES COVERED
Master's Thesis
4. TITLE AND SUBTITLE OF THESIS
SOURCE DEVELOPMENT FOR A SEISMO-ACOUSTIC SONAR
6. AUTHOR(S) Sean M. Fitzpatrick
PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
Naval Postgraduate School
Monterey CA 93943-5000
5. FUNDING NUMBERS
PERFORMING
ORGANIZATION
REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
Office of Naval Research, Code 321
Dr. Douglas Todoroff, Program Manager
800 N. Quincy St.
Arlington, VA
10. SPONSORING/MONITORING
AGENCY REPORT NUMBER
1 1 . SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the
official policy or position of the Department of Defense or the U.S. Government.
12a. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution is unlimited.
12b. DISTRIBUTION CODE
13. ABSTRACT (maximum 200 words)
Buried mines in the beach and surf zone hinder Naval power projection beach assault operations and
endanger humans. In this research, it is demonstrated that seismic interface waves on the surface of a natural
beach can be used to detect buried, mine-like objects. A seismic sonar system, employing guided interface
waves known as Rayleigh or Scholte waves has been proposed for detecting shallow, buried ordnance and is the
subject of an ongoing, ONR-sponsored research program at the Naval Postgraduate School. The primary
purpose of this research was to develop an improved seismic source to evaluate further the concept of using a
seismo-acoustic sonar to detect buried ordnance in the beach and surf zone. The developed source was based
upon a linear magnetic force actuator. Additionally, testing procedures were developed for follow-on research.
Seismic interface waves were generated with two, 25-lb force-controlled linear actuators operated as shakers.
The waves were measured with a two-element horizontal array of three axis seismometers. Buried, mine-like
objects, ranging from 71kg to 290kg, and at ranges of up to 5 meters were echo-located by visual examination
of the raw, recorded seismic signal, and by the use of polarization filtering signal processing.
14. SUBJECT TERMS
Mine warfare, mine detection, buried ordnance detection, seismo-acoustic sonar,
seismic surface waves, Rayleigh waves, Scholte waves
15. NUMBER OF
PAGES
109
16. PRICE CODE
17. SECURITY CLASSIFICA-
TION OF REPORT
Unclassified
SECURITY CLASSIFI-
CATION OF THIS PAGE
Unclassified
19. SECURITY CLASSIFICA-
TION OF ABSTRACT
Unclassified
20. LIMITATION OF
ABSTRACT
UL
NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. 239-18 298-102

DUDLEY KNOX LIBRARY
MONTEREY, CA 93943-5101
Approved for public release; distribution is unlimited
SOURCE DEVELOPMENT FOR A SEISMO-ACOUSTIC SONAR
Sean M. Fitzpatrick
Lieutenant, United States Navy
B.A., University of California at Berkeley, 1990
Submitted in partial fulfillment
of the requirements for the degree of
MASTER OF SCIENCE IN APPLIED PHYSICS
MASTER OF SCIENCE IN MECHANICAL ENGINEERING
from the
December 1998

ABSTRACT
Buried mines in the beach and surf zone hinder Naval power projection beach
assault operations and endanger humans. In this research, it is demonstrated that seismic
interface waves on the surface of a natural beach can be used to detect buried, mine-like
objects. A seismic sonar system, employing guided interface waves known as Rayleigh
or Scholte waves has been proposed for detecting shallow, buried ordnance and is the
subject of an ongoing, ONR-sponsored research program at the Naval Postgraduate
School. The primary purpose of this research was to develop an improved seismic source
to evaluate further the concept of using a seismo-acoustic sonar to detect buried ordnance
in the beach and surf zone. The developed source was based upon a linear magnetic
force actuator. Additionally, testing procedures were developed for follow-on research.
Seismic interface waves were generated with two, 25-lb force-controlled linear actuators
operated as shakers. The waves were measured with a two-element horizontal array of
three axis seismometers. Buried, mine-like objects, ranging from 71kg to 290kg, and at
ranges of up to 5 meters were echo-located by visual examination of the raw, recorded
seismic signal, and by the use of polarization filtering signal processing.

TABLE OF CONTENTS
I. INTRODUCTION 1
A. RESEARCH MOTIVATION 2
B. CURRENT AND EMERGING MINEHUNTING METHODS 4
H. PREVIOUS RESEARCH 9
A. ARL:UT(1996) 9
B. NAVAL POSTGRADUATE SCHOOL (1997) 10
m. THEORY OF SEISMIC WAVES 13
A. ELASTIC BODY WAVES 13
B. ELASTIC SURFACE WAVES 15
1
.
Rayleigh waves 15
2. Scholte waves 20
IV. DESCRIPTION OF EXPERIMENTAL EQUIPMENT 21
A. SEISMIC SOURCE 21
1
.
Actuator Modifications 24
2. Electronic Control Unit (ECU) Modifications 28
3
.
Relative Source Gain 29
B. SEISMIC SENSORS 30
C. DATA ACQUISITION 32
D. SIGNAL PROCESSING 33
E. TRANSPORTATION 34
V. TESTING PROCEDURES 37
A. BEACH TEST SITE 37
B. SEDIMENT CHARACTERIZATION 38
C. AMBIENT NOISE SPECTRA 39
D. SOURCE EMPLOYMENT 40
1
.
Placement and Orientation 41
2. Phasing 44
vn

E. TESTING CHRONOLOGY 46
1
.
Frequency 48
2. Wavespeed 50
3. Background Data 50
4. Target Data 51
5
.
Signal Processing 51
VI. TEST RESULTS (VERTICAL SOURCE) 53
A. ATTENUATION 53
B. VELOCITY AND DISPLACEMENT 54
C. TARGET DETECTION 56
D. UNDERWATER TEST 61
VH. CONCLUSIONS AND RECOMMENDATIONS 63
A. CONCLUSIONS 63
B. RECOMMENDATIONS 63
APPENDIX A. MANUFACTURER SPECIFICATION SHEETS 67
APPENDIX B. UNDERWATER ACTUATOR CABLES AND CONNECTORS 79
APPENDIX C. EXPERIMENTAL SETUPS 85
APPENDIX D. COMPUTER PROGRAMS AND SUBROUTINES 93
LIST OF REFERENCES 95
INITIAL DISTRIBUTION LIST 97
Vlll

ACKNOWLEDGMENT
The author would like to thank George Jaksha & Gary Beck for their insight and
help in developing the seismo-acoustic sonar used in this study. I would also like to
especially thank Professor Tom Muir for his overall guidance and assistance during the
field research phase and Professors Steve Baker and Anthony Healey for their help. Most
importantly, I would like to thank my wife, Laure, for her continuous support. This work
was supported by the Office of Naval Research, Code 321, Dr. Douglas Todoroff,
Program Manager.
IX

I. INTRODUCTION
Buried mines are a hindrance to Naval beach assault operations. An unknown,
buried mine, can disrupt, delay, and in some cases, halt Naval operations. In order to
maintain the relative superiority and capability that U.S. Naval Forces have to project
power ashore, it is critical that our forces have the ability to counter and neutralize buried
mines in stride. This capability begins with the detection of buried mines, both in the
beach and in the surf zone. The proposed concept by the Applied Research Laboratories
of the University of Texas at Austin (ARL:UT) to use a seismo-acoustic sonar and
"guided" seismic interface waves to detect shallow, buried ordnance in the beach and surf
zone offers promise in providing the ability to detect buried mines. This concept will
further enable our Naval Forces to effectively project power ashore.
Buried mines are not only a threat to U.S. military forces, they are also a serious
hazard to civilians and non-combatants. For example, there are more than 100 million
buried landmines worldwide and more than 150 people are killed or maimed each week.
In fact, the casualty rate is so high that one in every 470 Angolans and 1 in every 1,000
Somalis is an amputee. Cambodia alone has an amputee population of more than 30,000.
[Ref. 1] The concept of a seismo-acoustic sonar that can effectively detect buried mines
is encouraging and would help reduce the number of landmines and casualties caused by
buried landmines worldwide.
The research reported in this thesis is one step towards the development and
fielding of a seismo-acoustic sonar for detecting buried ordnance. This research is
presented in seven chapters. The introductory chapter outlines the motivation for
developing a seismo-acoustic sonar including a brief discussion on continuing plans to
develop improved Mine Countermeasure (MCM) capabilities for U.S. Naval Forces.
Also included is a short summary of current and emerging minehunting (MH) methods
and technologies. The second chapter describes previous research efforts, which helped
provide guidance for source development and testing procedures. The third chapter
presents a brief overview on seismic wave theory including a discussion on the
advantages of using seismic interface waves for detecting buried ordnance. The fourth
chapter outlines the design and development of the seismo-acoustic sonar and equipment

used in this experiment. The fifth and sixth chapters describe the field testing procedures
and experimental results, respectively. The final chapter contains concluding remarks
and recommendations for further testing and development.
A. RESEARCH MOTIVATION
Mine warfare (MIW) has been a part of U.S. warfighting since the American
Revolution. [Ref. 2] Since that time, U.S. MIW experiences have been infrequent and
the lessons learned arguably ambiguous. It is partly because of this ambiguity that mine
countermeasures for U.S. Naval Forces are less than ideal, as evidenced by the damage to
the USS PRINCETON and USS TRIPOLI by Iraqi mines during DESERT STORM.
[Ref. 3]
The reality of an increasing, modern mine threat is not surprising considering that
most mines are relatively inexpensive and simple to manufacture. The low cost of
developing an offensive or defensive mining capability combined with its effectiveness
both tactically and psychologically, make MIW particularly attractive to nations with few
resources to devote to more expensive, sophisticated weaponry. As a result, mine
proliferation is becoming a worldwide problem, with more than 49 countries possessing
mining capabilities, and with 30 of these demonstrating mine production capabilities, and
20 of these attempting to export their MrW systems. [Ref. 2]
The increasing mine threat, combined with a downsizing Navy and dwindling
resources, have driven efforts to develop more flexible and robust minehunting systems
and capabilities for U.S. Naval Forces. The modernization plan for the U.S. Navy is
outlined in the United States Naval Mine Warfare Plan [Ref 2] which describes the
requirement for an organic MCM capability consistent with Navy and Marine Corps
doctrine, "Maneuver warfare. . .from the Sea".
Power projection operations, including the effective delivery of U.S. Amphibious
Forces, from the sea, require the avoidance or neutralization of any possible mine threat,
particularly in the vulnerable regions where amphibious forces transition from shallow
water, through the surf zone, and onto the beach. Naval mines like the "Manta" bottom

mine (see Figure 1.1) that damaged the USS PRINCETON typically "scour" below the
surface and in this process are buried, due to water action. Similarly, beach and surf zone
mines are purposely hidden or become buried due to erosion and wind action, making
their detection and neutralization more difficult. These types of buried mines clearly
pose a serious threat to amphibious forces, particularly during an opposed, amphibious
landing. A flexible, dedicated or organic minehunting capability to detect these buried
mines in the shallow water surf zone and in the beach is needed to effectively neutralize
the mine threat to U.S. amphibious landing forces.
J^L^^^ar-"^'
Figure 1.1. "Manta" bottom mine similar to the mine that damaged the USS
PRINCETON during DESERT STORM [Ref. 2]
Currently, U.S. capabilities to detect and localize buried mines in both the shallow
water surf zone and in the beach are limited and no one system is optimally capable of
operating in both environments. The successful development and fielding of a seismo-
acoustic sonar would provide an additional minehunting capability that could cover the

critically vulnerable gap from the surf zone to the beach and help neutralize the mine
threat for our amphibious landing forces.
B. CURRENT AND EMERGING MINEHUNTING METHODS
The critically vulnerable regions for amphibious forces as they transition from sea
to land are the shallow water surf zone and the beach. U.S. amphibious forces are
vulnerable to buried ordnance in these regions partly because there is no MH system that
can seamlessly transition from the sea, through the surf zone, and on to the beach while
continuously hunting for buried mines. While there are a variety of methods and systems
available to U.S. Naval Forces for hunting mines, no single system is capable of detecting
buried mines both in the shallow water surf zone and in the beach. Some of the more
common MH methods are listed below in Tables 1 . 1 and 1 .2 respectively.
MINEHUNTING METHODS - SHALLOW WATER SURF ZONE
Navy EOD and SEAL divers
USMC reconnaissance combat swimmers
Marine mammals (U.S. Navy dolphins)
Towed sweep gear
Remote Minehunting (RMS) Sonar system
Airborne Laser Mine Detection System (ALMDS)
Table 1.1. Current minehunting methods in the shallow water surf zone
MINEHUNTING METHODS - BEACH
Hand-held probes
Manual metal detectors
Ground penetrating radar
Mechanical breaching equipment
Explosive breaching nets
Trained dogs
Table 1.2. Current minehunting methods on the beach

Each of the methods and capabilities listed in Tables 1.1 and 1.2 has distinct
advantages and disadvantages. For example, manual minehunting methods are generally
very accurate and have low false alarm rates. However, they are also relatively slow and
dangerous, as illustrated by the EOD technician probing for buried mines with a hand-
held probe in Figure 1 .2. Additionally, none of the current methods listed in Tables 1 .
1
and 1 .2 are capable of detecting and locating buried ordnance in the shallow water surf
zone and in the beach.
Figure 1.2. EOD technician with a hand-held probe [Ref. 4]
There are a number of emerging technologies and systems being considered to
combat the increasing mine threat to amphibious forces. One of these systems, the
Coastal Battlefield Reconnaissance & Analysis (COBRA) minefield reconnaissance
system (see Figure 1.3), an unmanned aerial vehicle (UAV) optical-based sensor system,
offers promise in that it can detect shallow, underwater mines and is capable of surveying
large areas. While the COBRA system offers relative improvements in minehunting, it is
still unable to detect buried ordnance, both underwater and in the beach.

Figure 1.3. Overview of the COBRA sonar system [Ref. 5]
The concept of a seismo-acoustic sonar, as shown in Figure 1 .4, offers promise as
a practical method to detect buried ordnance both in the beach and the seafloor. A
seismo-acoustic sonar would complement current and emerging minehunting methods
and technologies, as well as provide an additional sensor for multi-sensor data fusion.
Like other systems, the seismo-acoustic sonar is highly dependent on the
propagating characteristics of the medium, which may require the application of
relatively complex signal processing methods to identify the target signal in the seismic
sound field. This requirement will be addressed later.

CONCEPT: AM ACTIVE SONAR USING SEfSMO-ACOUSTiC
INTERFACE WAVES AS THE EXCITATION CARRYING
SONAR ENERGY AND INFORMATION
SURFACE-BASED
SONAR TRANSCEIVER
lURIEDTAEGEX
^*h£XC&tf&&!&foflr
COMPLEX
SUBSTRUCTURE WENT LAYE
AMPHIBIOUS
SONAR TRANSCEIVER
.......
Figure 1 .4. Design concept for a seismo-acoustic sonar to detect buried mines in the
shallow water surf zone and in the beach [Ref. 6]

II. PREVIOUS RESEARCH
There has been a long-standing effort to detect buried mines in the beach and surf
zone. The Naval Coastal Systems Station in Panama City, Florida has long been working
on the development of a synthetic aperture sonar that can be "sensor-fused" with
magnetic sensors for detecting buried mines from the surf zone seaward. Two excellent
articles on these sensor approaches, one by G.S. Sammelman, et. al. (synthetic aperture
sonar) and another by T.R. Clem (magnetics), recently appeared in an Office of [Naval
Research publication Naval Research Review, D. Toderoff, Vol XLDC, b. 3, 1997].
Similarly, it is scientifically feasible that seismic interface waves could be used to
detect and locate shallow, buried, mine-like objects. While there have been many studies
on seismic waves, there are two recent research studies of particular interest. The first
study was conducted at the Applied Research Laboratories of the University of Texas at
Austin (ARL:UT), where field tests were conducted to evaluate the use of seismo-
acoustic waves to detect a buried mine-like object in a natural beach. The second study,
conducted at the Naval Postgraduate School, initiated development of a discrete-mode,
seismo-acoustic source, suitable for detecting buried, mine-like objects in the beach.
These two research projects laid the foundation for the research reported in this thesis.
A. ARL:UT (1996)
Feasibility tests were conducted by ARL:UT to study the possibility of using
seismic interface waves to detected buried mine-like objects in a natural beach. The
seismic source consisted of a six-inch by eight-inch exciter foot with protruding nails,
driven by an electro-mechanical transducer, as shown in Figure 2.1. A three-element
array of three-axis seismometers was used to receive the seismic signals. [Ref. 7]
Test results showed that various signal processing techniques were required to
isolate the target signal in the relatively complex, seismic wavefield. [Ref. 7] These
signal processing methods included coherent averaging, background subtraction, and
vector polarization filtering, which will be discussed later.

The research at ARL:UT validated the proof-of-concept use of seismo-acoustic
interface waves to detect buried mine-like objects. However, there was observed a
significant amount of reverberation in the seismic signal. The excessive reverberation
required the coherent subtraction of the reverberant background to detect the target.
These results suggested one obvious method of improving the seismo-acoustic sonar:
reduce the amount of the reverberant background in the seismic, target signals. A logical
step to reduce the amount of reverberant background noise was to consider a different
seismo-acoustic source, one that discretely generates only one type of seismic, interface
wave.
Figure 2.1. ARL:UT seismic target detection beach setup [Ref. 7]
B. NAVAL POSTGRADUATE SCHOOL (1997)
Based on the ARL:UT research results, LT Frederick Gaghan studied a discrete-
mode seismic source designed to selectively excite and generate seismic surface waves.
[Ref. 8] He developed two discrete-mode seismic sources, the second being a larger
scale version of the first. The final, field-tested seismic source is shown in Figure 2.2. It
consists of two, off-the-shelf, 10 pound-force, moving magnet inertial reaction force
10

transducers (bass shakers), mounted at 45 degree angles with respect to an aluminum
base support. Test results showed that, due to medium anisotropy, a drive feedback
system or different type of seismic source needed to be developed to suppress unwanted
wave modes. At relatively short ranges, less than about 15 meters (45 feet), it was found
that multiple body and interface waves were present, resulting in a relatively complex
seismic wave field. At longer ranges, greater than about 25 meters (75 feet), it was found
that the propagating medium naturally filtered the seismic waves so that the wave was
predominantly a single-mode wave. The most important conclusion relevant to this work
was that stronger, more linear sources were needed along with refined field testing
procedures.
Figure 2.2. NPS discrete-mode seismic sonar design [Ref. 8]
Based on the conclusions and recommendations from the two previous research
efforts, the present research focused on the development of a seismic source capable of
being used to detect and locate buried mine-like objects. Significant efforts were made to
11

refine field-testing procedures and successfully detect a buried, mine-like target to
support ongoing target strength studies.
12

III. THEORY OF SEISMIC WAVES
Seismic waves have been used for geological surveying of petroleum and mineral
deposits for many years. While there are actually many different types of seismic waves,
the most common are those generated by an earthquake, which travel through the body of
the earth and along the earth's surface. As a result, seismic waves are generally divided
into two categories, body waves and surface waves. The theory of elasticity predicts two
basic categories of mechanical body waves for an infinite medium, compressional waves
and shear waves. In addition to these body waves, there exist many different types of
surface waves, of which three are of particular interest: the Rayleigh wave, the Scholte
wave, and the Love wave. Rayleigh waves are surface waves, with both radial (direction
of propagation) and vertical motion, that exist on the interface boundary between a solid
and a gas. Scholte waves are similar to Rayleigh waves except that they occur on the
interface boundary between a solid and a liquid. Love waves are "ducted" waves,
existing in the layer between three different media or near the interface, where there is an
increasing shear velocity with depth, and have purely shear motion. This chapter gives a
brief overview of the basic theory of seismic body waves and surface waves, and
describes the advantages of using Rayleigh or Scholte waves for detecting buried
ordnance.
A. ELASTIC BODY WAVES
The stress relations for elastic body waves are found by modeling the propagating
medium as an infinite, homogenous, isotropic and elastic body, and are given by,
o r
=Z&+2Li£ rr , o v =A6+2u£ vv , o„=A6+2u£_,xx * xx y ' }} *-*•
/ o 1
where crand e are the stress and strain, respectively, X and ju are the Lame constants, and
6? is the dilatation, or volume strain. The stress and strain indices in (3.1) represent the
direction of the stress or strain and the plane over which it is applied. [Ref. 9]
13

Neglecting body forces, the equations of motion are given by.
-2-
p—" =(A + //)V0 + //V
2
i7, (3.2)
dt~
where p is the density of the propagating medium, u represents the particle displacement
in vector notation, A. and ju are the Lame constants, and 6 is the dilatation or volume
strain. [Ref. 9]
'
The solutions to the elastic wave equations of motion in (3.2) lead to two types of
mechanical body waves, a compression wave, known as a Primary or P-wave, and a shear
wave, known as a Secondary or S-wave. The P-wave is the fastest travelling wave,
producing particle motion that is co-linear with the direction of propagation. P-waves
can propagate in almost any kind of medium, including rock and water, and travel with a
longitudinal speed given by
(3.3)
where B is the adiabatic bulk modulus, // is the shear modulus, and p is the density (note
that the Lame constant ju and the shear modulus fi are one and the same) [Ref. 9].
Contrary to P-waves, S-waves do not propagate in water, due to the fact that liquids
cannot support static shear stresses. Bulk S-waves produce particle motion perpendicular
to the direction of propagation and travel with a speed given by
where ju is the shear modulus, and p is the density. [Ref. 9] In general, both P-waves and
S-waves are produced when an external force disturbs an elastic medium or when an
elastic surface wave interacts with an interface or a boundary.
14

B. ELASTIC SURFACE WAVES
While there are actually many different types of surface waves, three specific
wave types waves were considered in this research: Rayleigh, Scholte, and Love waves.
As will be shown later, the Love wave was found to result in excessive background
reverberation, and was not considered further. Rayleigh waves, on the other hand, were
found to be ideal for detecting shallow, buried ordnance in the beach. While not
specifically tested, Scholte waves, similar to Rayleigh waves, can also be considered as
ideal wave types for detecting shallow, buried ordnance in the seafloor and surf zone.
1. Rayleigh Waves
In 1885, the "great" English scientist, Lord Rayleigh, demonstrated the existence
of a surface wave that traveled over the interface between a semi-infinite, elastic half
space and a gas; hence these waves became known as Rayleigh waves. [Ref. 10] The
Rayleigh wave equations, obtained by applying the conditions for a stress-free boundary
to the bulk wave equations in (3.2), are given by
£*«c?VV. ^= c?Vy, (3.5)
dt dt
where <p and ^are potential functions representing compression and rotation wave effects
respectively. [Ref. 9] The finite solutions to (3.5) result in a propagating wave along the
boundary that is attenuated with increasing depth provided the following condition is met
[Ref. 9],
re
6
-8k:
4
+ (24
-
I6y
2
)c
2
+ (l6/
2
- 16)= . (3.6)
15

The dimensionless variables /rand yin (3.6) are velocity ratios, given by
K" = ^-and L == ^L_= i_2l 5 (3. 7)
X + 2/u 2-2v
where cr is the Rayleigh wavespeed, cs is the shear wavespeed given by (3.4), c is the
longitudinal wavespeed, given by (3.3), X and jj. are the Lame constants, and v is
Poisson's ratio. [Ref. 9] The solution to (3.6) is found for a single, real, positive root on
the interval k=(0, l), corresponding to a value of Poisson's ratio on the interval
v = (0, 0.5). [Ref. 1 1] For most solid/fluid boundaries, the ratio of Rayleigh wavespeed
to shear wavespeed is typically about k ~ 0.9 . [Ref. 1 1]
As a result of the conditions described by (3.6) and (3.7), Rayleigh waves have
distinctive features that make them more readily identifiable in a complex, seismo-
acoustic wavefield. Probably the most obvious feature of Rayleigh waves is the elliptical
particle motion produced by their passage. This unique characteristic allows Rayleigh
waves to be distinguished from other wave types by identifying those waves that have a
90° phase shift (quarter period delay) between their horizontal and vertical components.
The phase relationship between the horizontal and vertical components resulting in
elliptical motion is shown in Figure 3.1, which shows an idealized seismic wavetrain
generated by a single vertical impulse source. [Ref. 12]
The first two waves (outermost) shown in Figure 3.1a and 3.1b, are the relatively
fast P-wave and slower S-wave, known also as the "Minor Tremor" in seismic earthquake
activity. [Ref. 12] The Rayleigh wave closely follows the S-wave and is known as the
"Major Tremor". In the Rayleigh wave, the horizontal and vertical components differ in
phase by about 90°. Near the surface, this results in motion that is counterclockwise
elliptical, or "retrograde", with respect to the direction of wave propagation, as shown in
Figure 3.1c. Note also that, of the three wave types, for a semi-infinite, elastic medium
with a free boundary, the Rayleigh wave typically has the largest amplitude or
displacement, particularly the vertical component, which is of importance and will be
discussed later.
16

P-Wave
S-Wove R-Wave
(c)
Particle Motion
>- u
Direction of Wave Propagation
|w
Figure 3.1. Seismic wavetrain resulting from a single vertical impulse source [Ref. 12]
Another important and distinctive feature of Rayleigh waves is that the sense of
their elliptical motion changes with depth. This is further illustrated by modeling the
propagating medium, a medium-grain natural beach, as a semi-infinite, uniform (non-
layered) solid with elastic properties. While this model is not perfectly accurate, it will
nevertheless illustrate some of the major features of Rayleigh waves. For this present
project, tests were conducted on a medium-grain, natural beach with a typical Poisson's
ratio of about v = 0.25 , and a measured shear wavespeed of about 90m/s. The seismo-
acoustic sonar was typically operated at frequencies on the order of about 100Hz.
Substituting these values into the required condition (3.6), the resulting horizontal and
vertical Rayleigh wave displacements were determined as a function of depth (see Figure
3.2) with the MATLAB program in Appendix D23.
For depths less than about one-tenth of a wavelength (0.2At), where Xr represents
the Rayleigh wavelength, the horizontal displacement is positive, as shown in Figure 3.2,
17

and the sense of the motion is retrograde with respect to the direction of propagation, as
seen in Figure 3.1c. At a depth of about 0.2^, the horizontal motion is zero, resulting in
purely vertical motion. At depths greater than 0.2^, the relative horizontal displacement
changes direction, becoming negative, and the sense of the motion changes to clockwise
or "prograde" motion with respect to the direction of wave propagation. [Ref. 1 1] Note
also that the relative amplitudes of the horizontal and vertical displacements decrease or
are attenuated with increasing depth into the propagating medium. At a depth of about
two wavelengths, the amplitude of the displacement is less than 1% of the maximum
displacement at the surface and for all practical purposes the Rayleigh wave can be
considered to be localized to within a layer of about two wavelengths deep.
Plot of Horizontal & Vertical Displacement vs. Depth
0.5
.n
c
0)
>
(0
en
re
a.
5.1.5
8
a
-0.5
I
-
/
(
Horizontal
Jr I
I
x i
s
/ Vertical
i /
11/ •
i
/
i | /

f
1 1 1
V J I
!
{
il / 1
0.5
Relative Displacement
1.5
Figure 3.2. Typical Rayleigh wave displacements in a natural beach as a function of
depth (see Appendix D23 for the program listing)
18

As shown in Figure 3.1, Rayleigh waves are the largest disturbance in a seismic
wavetrain generated by a vertical impulse. Additionally, since Rayleigh waves are
localized near the surface, they attenuate due to geometrical spreading less severely than
do other wave types. Neglecting absorption and frictional effects, Rayleigh waves
attenuate approximately as the inverse of the square root of the range (l/Vr J
from the
source or scattering center, whereas body waves attenuate approximately as 1 / r , due to
spherical spreading. As a result, Rayleigh waves contain a significant amount of the total
seismic wave energy, as illustrated in Figure 3.3, which shows the relative energy
partitions for the Rayleigh, shear, and compression wave components resulting from a
vertically oscillating circular plate at the surface of a semi-infinite, uniform, elastic
medium. [Ref. 11]
Circular Footing
r
-2 Geometrical
f
-2
r
~0.5
^ Damping Law _^.
Shear
Window
Relative
Amplitude
Wave Type
Per Cent of
Total Energy
Rayleigh
Shear
Compression
67
26
7
Figure 3.3. Wavefield from a vertically oscillating circular plate at the surface of an
ideal, semi-infinite, uniform, elastic medium [Ref. 11]
19

Note that Figure 3.3 illustrates many of the Rayleigh wave features that have been
discussed, namely, that for an idealized, semi-infinite, uniform, elastic medium, Rayleigh
waves travel along the interface boundary, and contain a significant amount of seismic
wave energy. Rayleigh waves also result in elliptical particle motion near the surface and
have a significant vertical displacement component. These distinctive characteristics
make Rayleigh waves well suited for use in a seismo-acoustic sonar to detect shallow,
buried ordnance.
2. Scholte Waves
Scholte waves are essentially identical to Rayleigh waves except that the interface
boundary is between a solid and a liquid. Scholte waves contain the same essential
features as Rayleigh waves, as shown in Figure 3.4. [Ref. 13] Note that Scholte waves in
the sea floor, similar to Rayleigh waves, are localized near the bottom-water boundary.
The motion is also depth dependent, changing from retrograde elliptical at the surface to
purely vertical at a depth of about one-tenth of a wavelength and then to prograde
elliptical at greater depths. These distinctive characteristics of Scholte waves also make
them well suited for detecting buried naval ordnance in the seafloor and shallow water
surf zone sediments.
PARTICLE
DISPLACEMENTS
O
- t
-o.i
-001
PARTICLE ORBITS
•1
O)
ENERGY FLUX DENSITIES
O -0.5<
~~X
I
wU.O)
jy, *0.0»
BOTTOM +0.1 •
1
0.5
>1
Figure 3.4. Scholte wave characteristics in the seafloor [Ref. 13]
20

IV. DESCRIPTION OF EXPERIMENTAL EQUIPMENT
The most important part of this research were the research tools, particularly the
seismo-acoustic sonar. The experimental apparatus is functionally divided into four
major groups, as illustrated in Figure 4.1: (1) Seismic source(s), (2) Seismic sensor(s),
(3) Data acquisition, and (4) Signal processing. This chapter describes in further detail
the system components that were used in this research, including the modifications that
were made to the seismo-acoustic source, the sensors, and the data acquisition system.
))))
Figure 4.1. Simplified representation of the seismo-acoustic sonar system
A. SEISMIC SOURCE
The seismo-acoustic sonar source was based on two, commercial, off-the-shelf
linear actuator systems, made by Aura Systems Inc, (see Appendix Al for specifications).
[Ref. 14] Depending on the configuration, these actuators are capable of operating as
either positioners or shakers. For this research, the linear actuators were factory-
configured to deliver a voltage-controlled force, capable of generating lOlbf-rms and
251bf-peak output, with 0.5-inches of total stroke. The actuators have a relatively flat
(force) frequency response from 0Hz to 250Hz. The actuators are designed without belts
or gears so that the only moving part is a moving magnet within a solenoid. The result is
a simple, sturdy, direct-drive, linear system, without backlash or hysteresis effects.
21

The actuators are individually controlled by an Electronic Control Unit (ECU),
which converts a user-input voltage to a shaft output force. The ECU converts the input
low-level analog voltage, to a current that is output to the actuator, typically 1.0-amp
output per 1.0-volt input (1.0-amp/volt). The actuator output force is linearly
proportional to the ECU-supplied current. As a result, the force applied at the output shaft
follows the voltage applied to the ECU, regardless of shaft position. The force constants
for both actuators used are listed in Appendix A2 and are nominally about 1.29-lbf/amp
and 1 .49-lbf/amp. The linear actuator system configuration is shown in Figure 4.2.
LINEAR ACTUATOR SYSTEM CONFIGURATION
INTERNAL POSITION
SENSOR
110 VAC -»J POWER
SUPPLY
ACTUATOR DRIVE CURRENT
POWER
AMPLIFIER
POSITION
SENSOR
OUTPUT POSITION
SENSOR
SIGNAL
CONOfnONER
HgHCOMPENSATION
CIRCUITRY
I
OUTPUT
STROKE
CTUATOR
u
IQ
POSITION
OUTPUT
POSITION
COMMAND
FORCE
COMMAND
ELECTRONIC CONTROL UNIT
Figure 4.2. The linear actuator system configuration [Ref. 15]
The actuators can simultaneously generate both DC (static) and AC (dynamic)
shaft output forces. In addition, there is an internal position sensor, connected to the
ECU, that allows for automatic centering of the output shaft with up to a 10-pound static
loading force. The output from the position sensor, which is linearly proportional to the
shaft position, is available via a BNC connector (Position Monitor) on the ECU front
panel, as shown in Figure 4.3. Tests showed that the position monitor was limited to
22

tracking output shaft movements at frequencies on the order of about 20Hz and below.
Since the typical operating frequencies used were about 70Hz to 100Hz, a separate
accelerometer was mounted to the output shaft to track output movements, as shown in
Figure 4.6.
19 Pin CPC Enable
Disable
ON
Power
Lamp Force
Input
Position
Input
Position
Monitor
O © © ©Off
"Spare" BNC
Connector
Figure 4.3. Electronic control unit (ECU) front panel
The ECU requires up to 15 Amps of 1 10VAC, 60Hz electrical power for normal
operation. The actuator was connected to a 19-pin connector (9-pins actually used) on
the ECU front panel, using a 300-foot underwater cable that was constructed for this
research (see Appendix B). The underwater cable allowed remote deployment of the
actuator from the ECU.
The following CAUTION is of particular importance and should be heeded:
Do not disconnect the ECU-actuator cable connection without first
setting the ECU Enable/Disable switch to the "DISABLE" position!
This lesson was learned first-hand. Disconnecting either the actuator or ECU connection,
without first turning the enable/disable switch to "DISABLE", will cause an arc that will
fuse the cable to the connector.
23

1. Actuator Modifications
Numerous modifications were made to protect the actuators from salt-water
intrusion and improve their operation as a seismo-acoustic sonar source. Many of the
modifications were made based on lessons-learned during field tests and by trial and
error. An external view of the linear actuator, before modification, is shown in Figure
4.4.
ACTUATOR MOUNTING HOLES
3X «.203 THRU
ON A 2.50 B.C
1.13 NULL POSITION
(±.25 STROKE)
l/*-28UNF-2B
« .50 UIN DEPTH
f OOOQJ
-.0O0S
.37 BETWEEN FLATS
CONNECTOR
MS3I02A-18-IP
2X 8-32UNC-2B
X .44 UIN DEPTH
ACTUATOR MOUNTING HOLES
Figure 4.4. Schematic of an unmodified linear actuator [Ref. 14]
One of the first modifications was to attach a nylon cylinder and 5-inch square
plate assembly to the output shaft, similar to the aluminum rod and plate shown in Figure
4.7. The actuators were typically placed in an upright position on the beach sand and
driven with a 5-cycle, 100Hz input signal to the ECU. It was found that for ECU-input
voltages greater than about 2.5-volts, the actuator-plate assembly literally "hopped" on
the sand, producing a complex, inconsistent seismic wave at the seismometer, 5-feet
distant. The cylinder-plate assembly was then buried, in the same sediment conditions,
about 3-inches below the sand surface. A noticeable improvement was seen in the
magnitude and the consistency of the seismic wavetrain as seen at the seismometer, 5-feet
distant. The following week, under similar conditions, a 2-kg inertial mass was fastened
to the actuator housing (shown in Figure 4.6), which resulted in yet an additional increase
24

in the amplitude of the seismic wave as measured by the seismometer. In this
configuration, the actuator was operated as an inertial reaction force shaker, with the
moving mass consisting of the actuator housing plus the added mass fastened to the top
of the housing. The vertical amplitude of the seismic wave as measured by the
seismometer, 5-feet distant, is shown in Figure 4.5 for the three modifications.
Plot of Geophone Output vs. Actuator Input
Actuator
y Excursion
ited
4 5 6 7
Actuator Input Voltage (V)
Figure 4.5. Plot of the vertical geophone output voltage vs. ECU input voltage
As can be seen from Figure 4.5, the addition of an inertial mass combined with
improved ground coupling by burying the actuator-plate assembly, resulted in an
approximate gain of 4-dB in maximum achievable seismic wave amplitude. The
apparent amplitude "rolloff' for the buried plate and mass modifications in Figure 4.5 is a
result of the actuator reaching its excursion limit. Note also that the pressure exerted by
the 25-lbf actuator and 5-inch square plate is on the order of about 1-psi, much less than
25

the liquification pressure for medium-grained sands, which is typically greater than about
50-psi. [Ref. 16]
The next series of actuator modifications are shown in Figure 4.6, some of which
allowed the placement and operation of the actuator inside a watertight, 6-inch diameter
PVC pipe, shown in Figure 4.7. The modifications were designed to protect the actuator
from salt-water immersion and allow operation of the actuator system underwater.
Inertial Mass (2kg)
Re-oriented Connector 90c
Roller Bearings (4-sets)
Mechanical Stop
Rubber Membrane
Accelerometer
Figure 4.6. Actuator modifications
The electrical connector was replaced with an aluminum block that re-oriented the
connector 90° toward the back of the actuator and allowed placement of the actuator
26

inside the watertight PVC pipe. A watertight, bulkhead connector (see Appendix B3)
was mated to the PVC housing via a nylon block. Additionally, 4-sets of roller bearings,
2-rollers each, were attached to the actuator body allowing it to roll relatively freely
inside the PVC housing. To protect the actuator from internal damage, a hollowed, nylon
cylinder was mated to the base of the body, over the output shaft, which prevented
excessive force against internal stops when turning the self-centering actuator off.
To seal the output shaft end, a nylon ring was mated to the end of the PVC
housing and 8-holes were tapped into it. Two aluminum plates, with machine screws for
attaching to the actuator output shaft and ground coupling assembly, were mounted to a
black, rubber membrane which was then firmly mated via a metal ring to the end of the
PVC housing. A uniaxial accelerometer (see Appendix A3) was mounted, via a machine
screw, to the external, aluminum plate, as shown in Figure 4.6.
To reduce the "head" mass at the actuator output shaft, the nylon cylinder-plate
assembly, weighting about 1.8-pounds was replaced by a 6-inch long, 0.5-inch diameter
aluminum shaft and 6-inch aluminum square plate assembly, weighing about 0.8-pounds.
The fully assembled, watertight, seismo-acoustic sonar source is shown in Figure 4.7.
Figure 4.7. The fully assembled watertight seismo-acoustic sonar source
27

2. Electronic Control Unit (ECU) Modifications
After a number of field trips, it was noted that the source continued to "ring" after
the electrical signal to the ECU was terminated. Laboratory tests showed that the
actuator continued to oscillate for a few hundred milliseconds at the completion of an
input pulse train. After consulting with engineers from Aura Systems Inc., it was found
that the RC time constant and slewing rate of the automatic shaft centering feature had
been altered by the replacement of the 6-foot, factory-supplied cable with the 300-foot,
underwater cable. The result was an undesirable oscillation until the actuator settled-
down to the static, centered position. The solution was to re-adjust the time constant by
adding a capacitor, in parallel, across the R5 resistor on the main control board. A
speaker-type push connector was hardwired to the R5 resistor and mounted on the back
of the ECU as shown in Figure 4.8, enabling external capacitor loading. Trial and error
lab tests showed that a 2uJF capacitor seemed to work best in reducing the oscillations to
an acceptable level.
Added
Connector
Figure 4.8. ECU rear cover with an added speaker-type push connector
28

The second modification to the ECU included the addition of an extra pin to the
19-pin cable connector and tying it to the "spare" BNC connector on the ECU front
panel, as shown in Figure 4.3. The ECU is factory-delivered with only 8-pins installed in
the 19-pin connector, which are all that is necessary to control the actuator. As shown in
the ECU/Actuator Interconnect Diagram in Appendix A4, the conductors for pins D and
E on the actuator-side are tied together on the ECU-side before connecting to the ECU.
After consulting with Aura engineers, pins D and E were tied together at the actuator-
side, which made one of the 9-conductors in the 300-foot cable available for other uses
and actually resulted in better grounding and shielding properties.
Originally, the "extra" conductor, made available from this modification, was
intended to be used to route the actuator accelerometer signal, which was originally
installed on the waterproof inside of the PVC housing, back to the ECU and from there to
the data acquisition computer. Unfortunately, the electrical interference generated by the
actuator-ECU system added an excessively large 120Hz component to the accelerometer
signal when using this wire, making it virtually unusable, since typical operating
frequencies were on the order of 70-100Hz. As a result, a separate coaxial cable was
used for the accelerometer signal and there is an extra conductor available from the
actuator to the ECU, which may be used for other purposes, i.e. a thermocouple.
It was found that the actuators tended to overheat when constantly operated in the
vertical position. The overheating was caused by the constant ECU current that was
required to maintain the statically centered position with the relatively heavy inertial
mass loaded on top. Therefore the ECU switch was set to "Disable" between
experimental runs.
3. Relative Source Gain
Comparing the vertically operated, linear actuator system with the previous,
discrete-mode, seismo-acoustic sonar source [Ref. 8], both operating at their maximum
capability, the actuator produces a much greater seismic interface wave intensity (see
Figure 4.9). For a source-receiver separation of 70-feet, the actuator source produced a
29

vertical seismometer signal of about 60db greater than the discrete-mode source. Based
on these results, it seems that the linear actuator system looks promising for a seismo-
acoustic sonar with a reasonable standoff range capability.
x10
4 Discrete-Mode Sonar
o
•a
3**
Q.
E
<
><
o
o3
E
<-2
to
E
<
N
-2
0.2
X10"
4
0.2
x10"
4
0.4 0.6
0.4 0.6
0.2 0.4 0.6
Time (sec)
0.5
-0.5
i
0.5
-0.5
Linear Actuator Sonar
0.2
0.2
0.4 0.6
0.4 0.6
0.2 0.4 0.6
Time (sec)
Figure 4.9. Comparison of the seismic interface waves for the discrete-mode and
modified linear actuator sonar at a source-receiver distance of 70-feet
B. SEISMIC SENSORS
For this research, two three-axis seismometers, consisting of three orthogonally-
oriented geophone sensors, were used as receiving elements (see Appendix A5 for
specifications). Each seismic sensor consists of three geophones, mounted in a cylinder
of urethane (see Figure 4.10) to form a triaxial velocimeter. The geophone converts the
velocity of the particle motion caused by the propagating wave into an output voltage.
30

The output voltage is linearly proportional to the particle velocity, which can then be
used to further analyze the seismic wavefield, as will be discussed later. The geophones
have the following the dimensions: height-23 cm, diameter-5 cm.
Reference for the
pointing direction
Figure 4.10. Triaxial geophone seismic sensor
During field tests there were noticeable frequency components in the seismometer
signals on the order of about 150Hz to 175Hz, which could not initially be accounted for
or explained. Similar, unexplainable frequency components were seen during previous
research studies, using the same geophones. [Ref. 17] Closer examination of the
geophone specification sheets in Appendix A5 revealed that the geophones have an
undesirable, "typical spurious frequency" of about 150Hz to 190Hz. Fortunately, for this
research, typical operating frequencies were below these spurious frequencies, typically
70Hz to 100Hz.
The geophone signals are amplified in the deployed unit with a selectable lOdB or
40dB (typically 40dB selected) amplifier, typically powered by two, 12-volt,
rechargeable batteries, wired in series to provide ±12-volts. During field tests it was
noticed that the batteries would typically drain to about 10-volts by the end of the day. In
order to provide more consistent seismic measurements, the batteries were replaced by an
ac/dc power converter which continuously supplied the required ±12-volts.
31

During lab tests it was noticed the actuator-ECU system produced a moderate
20kHz component in the geophone signal. Using the ac/dc power converter had an added
benefit in that it appeared to reduce the 20kHz interference by about 10%. The most
effective reduction in the 20kHz noise interference was found by placing the ECUs in a
separate equipment rack, effectively isolating the receiving elements from the transmit
elements.
C. DATA ACQUISITION
Data acquisition was accomplished with Signal Processing System's SPS390
Signal Analyzer, shown in Figure 4.11. The SPS390 is a 16-bit, 0-40kHz analyzer,
capable of recording up to eight channels of data.
rMrfSJg
MB Screen Display
*j^£*|lgSSlEi^ r '.
™ 'I'JW^^^^"^ H^^^^^55i5SP5J5!?^P!S^5plRlP
Figure 4.1 1. Signal Processing Systems (SPS390) signal
analyzer used for data acquisition
For this research, the SPS390 was upgraded to a 32-bit operating system and the
Random Access Memory (RAM) was increased to 128MB. These upgrades provided
more a more reliable data acquisition system and allowed data file transfers to another
computer for signal processing without terminating the SPS390 signal processing
program. The result was a relatively quick turn-around time (2-3 minutes) from data
acquisition to signal processing and data analysis. To support follow-on research, the
configuration, menu settings, and 8-channel inputs (standardized throughout the research)
are included in Appendix A6.
32

D. SIGNAL PROCESSING
The signal processing of the recorded seismic data during this research has been
an evolutionary process. At the start of this research, seismic data were recorded and
stored in the field, then processed in the lab 2 to 3 days later. This method was found to
be excessively inefficient, considering that it would take a week or more to analyze the
data. As a result, equipment and signal processing capabilities were developed to allow
near real-time (2-3 minutes delayed) signal analysis in the field. This capability allowed
an on-site assessment of the effectiveness of sonar placement and test configurations,
significantly improving testing procedures and results.
Twenty two signal processing subroutines (see Appendix D) were written using
Matlab's Signal Processing Toolbox. [Ref. 18] The data analysis and signal processing
methods that were developed are summarized in 4 basic steps, shown in Table 4.1.
STEP COMPUTER OPERATION
1 SPS390 Record accelerometer and seismic geophone signals
2 SPS390-to-Portable Download the recorded data (via a null-model cable)
3 Portable Convert the data to Matlab 4. 1 format
4 Portable Signal Analysis (MATLAB Subroutines)
Table 4.1. Summary of data acquisition and signal processing steps
The MATLAB subroutines, listed in Appendix D, are sufficiently commented and
do not require further explanation at this point. One particular signal processing method,
vector polarization filtering, deserves further discussion. As discussed in Chapter HI,
Rayleigh waves, by their very nature, have radial and vertical components that differ in
phase, resulting in polarized elliptical particle motion. This feature provides a method
for discriminating between different wave types to identify those interface waves with a
relative phase difference between the radial and vertical components. This method is
known as "Vector Polarization Filtering". [Ref. 7]
33

In this research, polarization filtering is the process of extracting from a radial-
vertical signal pair, the component of the crossed signal power that comes from the
elliptical particle motion of the interface wave. [Ref. 7] This is accomplished by
applying the Hilbert Transform to the radial-vertical signal pair (see Ref. 7 for further
discussion) and calculating the time domain, crossed power function as given by,
*n> ~ rhilbert * Vhilbert ' V*. 1)
where Pn is the crossed power function, rmbert is the complex conjugate of the hilbert
transform of the radial component of motion, and vMbert is the hilbert transform of the
vertical component of motion, consistent with the notation used in Ref. 7. The in-phase
radial and vertical components are contained in the real parts of Pn. while the out-of-
phase components are contained in the imaginary parts. [Ref. 7] As a result, the
imaginary part of Pn,
is approximately proportional to the power of the Rayleigh wave
and deflects positively or negatively for polarized signals, depending on whether the
motion is prograde or retrograde, respectively. [Ref. 7] This technique is used to reject
other signals, particularly the bulk P and S waves, and will be illustrated later.
E. TRANSPORTATION
During this research a John Deere, 4x6 "TrailGator" (see Appendix A7) was
acquired to move and position naval mine-like targets in the beach, see Figure 4.12.
Additionally, since part of the focus of this research was to refine testing procedures,
considerable efforts were made to minimize and simplify equipment requirements. As a
result, the transmit, receive, and signal processing equipment components described
previously were condensed down to two equipment racks that were transported and
staged from the bed of the TrailGator, as shown in Figure 4.13. The TrailGator proved to
be an invaluable addition to the research project and presents opportunities for expanded
development (i.e. deploying a combination seismo-acoustic sonar and sensor array),
which will be discussed later.
34

Figure 4. 12. Front view of the TrailGator and beach equipment
Figure 4.13. Transmit and receive equipment racks in the TrailGator bed
35

V. TESTING PROCEDURES
One of the major goals of this research was to refine field-testing procedures for
the seismo-acoustic sonar development. As a result, a significant amount of time and
effort, more than seven months, was put into developing streamlined and efficient beach
setup and testing procedures. Rather than describing the evolution of the developing test
procedures, this chapter will outline the procedures that were developed and successfully
used at the later latter part of this research. Before describing the testing procedures, this
chapter begins with a brief description of the beach test site, characterization of the beach
sand, and a spectral analysis of the ambient ground vibration.
A. BEACH TEST SITE
The beach test site was located on the Monterey Bay coastline, adjacent to the
Naval Postgraduate School, at the NPS Beach Research Facilities. See Figure 5.1.
—*o
EUNE
FISHEftUAffS
WHARF
, CUSTOU
HOUS£ HAZA
tf" UONTEREY
Underwater
Test Site
MONTEREY
VU . ^^*X_ u)_.t J—— *I—^^^^^^ _^—' _a* J --A
Figure 5.1. Beach and underwater test sites
Map copyrighted 1994 by the California State Automobile Association. Reproduced by permission.
37

The beach site proved to be well suited for this research, with easy access to the
beach via a 20-foot wide, dirt road. The width of the beach and available test area varied
with the tides, ranging from a width of about 120-feet at low tide to less than 10-feet at
high tide. See Figure 5.2.
„™_^„_„— ,
—._—^^^-^^^j^^^. ..
T" -•• -w '
—
m v.;
'
Figure 5.2. Beach test site at high tide
B. SEDIMENT CHARACTERIZATION
In general, beach sand is composed of a variety of elements, including minerals,
rocks, organic and inorganic compounds, water, and air. The relative amounts of these
different elements vary, depending on the geographical location and the depth of the
sediment. [Ref. 8] In addition to its composition, sand is characterized by a variety of
features, including grain size, density, and moisture content. Each of these characteristics
affects the propagation of elastic wave energy along the beach-air interface. While the
relationship between sand characteristics and interface wave propagation are relatively
38

complex, it is generally accepted that interface wave energy is less attenuated in high
moisture content and densely compacted soil environments. [Ref. 16]
Although the Rayleigh wave equations, for a uniform, semi-infinite, elastic half
space (3.5), are frequency-independent (dispersionless), real beach sediments are not
uniform, meaning that they are not necessarily dispersionless. Hence the speed of
Rayleigh waves may depend on the frequency. In a natural beach or seafloor sediment,
the overburden weight of the surface sediments causes an increase in shear modulus with
increasing depth into the medium. Consequently, longer wavelength, lower frequency
acoustic waves propagate deeper into the sediment, where the shear velocity is greater.
As a result, long wavelength, low frequency signals travel faster than short wavelength,
high frequency signals. At relatively higher frequencies, greater than 100Hz, there is little
dispersion because the wave remains near the surface, where the sediment is relatively
uniform. To complicate matters further, a core sample of the beach sediment revealed,
unsurprisingly, that the moisture content of the sand varied with depth. At a depth of
about 1 -meter, it was found that the sand was completely saturated, corresponding to the
depth of the tide water table at that time.
C. AMBIENT NOISE SPECTRA
Using the geophones as receivers, a measurement of the ambient ground vibration
at the beach test site was made. The measurement allowed an assessment of the influence
that the ambient noise would have on identifying the target signal, and provided
qualitative information on which operational frequencies to avoid. See Figure 5.3. The
test setup is outlined in Appendix CI and the MATLAB program used for the spectral
analysis is outlined in Appendix D16.
Note that the spectral content of the ambient background noise is greatest from
about 10Hz to 40Hz, consistent with previous measurements made at this site. [Ref. 8]
The predominant source of the ambient noise is undoubtedly the breaking of waves
similar to those shown in Figure 5.2. The measurement was useful in providing guidance
on selecting the high-pass filtering frequency and allowed a relative sense of confidence
39

that the sonar operational frequencies, typically 70Hz to 100Hz, are not within the strong
ambient noise band.
Geophone #1
>> ~
en
C
a
a
is
-
—
a
-
C
2-
Figure 5.3. Omnidirectional velocity power spectral density for the ambient ground
vibrations at the beach test site (see Appendix C 1 for the test setup)
D. SOURCE EMPLOYMENT
Two seismic sources were developed to independently generate horizontal and
vertical motion. The versatility of the sources allowed for creative testing of many
different actuator orientations and operating combinations. More than 100 different tests
were conducted over 21 days of field testing, each with different drive signals, actuator
orientations, geophone locations and target positions. The seismic interface waves
resulting from these many tests were closely studied and evaluated for further use in
buried target detection.
40

The placement and operation of the actuators can be categorized into two different
configurations. In the first configuration, the actuators are operated either horizontally or
vertically, while in the second, they are operated simultaneously, one horizontal and one
vertical, with a 90° relative phase difference between their drive signals. With the reader
as the receiver, Figure 5.4 shows a typical vertical and horizontal actuator placement,
resulting in vertical and shear wave excitation, respectively.
Figure 5.4. Typical actuator placement for vertical and shear wave excitation.
1. Placement and Orientation
Different actuator orientations and operating combinations resulted in the
generation of various seismic interface waves, each with distinctive features and
characteristics. These waves were evaluated for their suitability for use in detecting
buried targets. Based on previous research [Refs. 7 & 8], the evaluation centered on
finding seismic waves that had both a single-mode polarization and minimal
reverberation.
The first, and simplest, actuator placement that was tested was to orient the source
vertically, as shown in Figure 5.4. The test setup is shown in Appendix C2, which shows
two actuators separated by 46-cm (18-inches) and 1.8-meters (6-feet) from the receiver.
The actuators are driven with a single cycle, 10-volt, 90Hz signal. The geophone signals
for the relative x, y, and z-motion are shown in Figure 5.5, which was generated using the
MATLAB program in Appendix D4. The vertical line at about 0.01 -seconds indicates
the start time for the input actuator pulse. Note that the x-radial and z-vertical motions
41

are the dominant wave components and that the reverberant wavefield has decayed to less
than 5% of the maximum amplitude at about 0.1-seconds.
n 1
-
o
£o3
E-1 -
B 1
o
>^
S nD3
E-1
N
0.05 0.1 0.15 0.2 0.25
Time (sec)
0.25
Time (sec)
0.05 0.1 0.15 0.2 0.25
Time (sec)
Figure 5.5. Relative seismic motion (velocity) from a vertically-oriented source
As discussed in Chapter HI, computation of the crossed power function (4.1) is a
practical means of separating the Rayleigh wave signal and rejecting the bulk P and S
waves. A plot of the imaginary component of the crossed power function provides an
indication of the polarization, or direction, of the elliptical wave motion, prograde or
retrograde. With the seismometer reference post (see Figure 4.10) pointing in the
direction of the incident wave, a positive deflection of the crossed power from (4.1)
indicates that the motion of the Rayleigh wave is prograde, whereas a negative deflection
indicates a retrograde wave. The observed imaginary crossed power for a vertically-
oriented source is shown in Figure 5.6, which 'was generated using the MATLAB
42

program in Appendix D12. Note that the deflection is negative, corresponding to an
interface wave with prograde motion for a geophone that is pointed away from the
incident wave source as shown in the test setup in Appendix C2.
Similar tests were conducted for a horizontally-oriented actuator first pointing
directly at the receiver at a 0° angle, resulting in a P-wave, and then transversely at a 90°
angle, resulting in a Love wave. In both cases it was found that reverberations were
excessively large and that there were at least two excited modes, as the polarization of the
elliptical particle motion was observed to change from one sense to the other within the
pulse train. These two actuator orientations were not considered further for use in
detecting buried targets.
1
-
0.5-
o
>^
•v
a.
E
<
-0.5-
-1 -
^Tfx—0.05 0.1 0.15
Time (sec)
0.2 0.25
Figure 5.6. A plot of the imaginary crossed power for a vertically-oriented source
with the receiver directed away from the source
43

2. Phasing
In the second type of configuration, the actuators were oriented side-by-side, one
horizontal and the other vertical. The actuators were driven with a relative 90° phase
difference to simulate and excite either retrograde or prograde Rayleigh wave motion.
The results were ambiguous; in all of the tests the polarization of the elliptical particle
motion changed during the pulse wavetrain. Similar phenomena had also been observed
during previous studies [Ref. 8]. It is unclear whether there were actually two different
modes, one prograde and one retrograde wave, or whether the same wave was being
reflected from the saturated layer about 1 -meter in depth. It seems unlikely that the
change in polarization was due to reflection from the underlying layer, as the wave
polarization for the vertically oriented source should have been just as likely to generate
waves that were reflected from the underlying layer.
The test configuration for generating retrograde and prograde motion with the
actuators is shown in Appendix C3. The test configuration shows two actuators, one
horizontal and the other vertical, separated by 46-cm (18-inches) and driven with a two
cycle, 6-volt, 100Hz signal with a relative phase difference between the two, for
simulating retrograde or prograde elliptical motion. Figure 5.7 shows the crossed power
calculated from (4.1), which indicates the polarization or direction of motion for a
retrograde input and a source-receiver distance of 5-feet (1.5-meters) and 10-feet (3.0-
meters). Note that the polarization initially starts out as prograde at 5-feet (1.5-meters)
and changes to a mixed retrograde/prograde wave at 10-feet (3.0-meters).
Using the same geometry and configuration as for the retrograde test, the phase of
the input signal was changed to simulate and excite prograde motion. The resulting
motion is indicated in the plot of the imaginary crossed power in Figure 5.8. Note that
the polarization is a mixed retrograde-prograde motion at 5-feet and changes to purely
prograde at 10-feet. Although there appears to be a single-mode, prograde polarization at
10-feet, similar to the polarization for the vertically oriented source, this type of wave
was not used because of the desire for a short-range sonar system and the fact that the
mechanism for mixed polarization was not fully understood.
44

Source-Receiver Distance (5 feet)
o
3
1.5
1
0.5
"5.-0.5
E
< -1
-
-1.5-
0.05 0.1 0.15
Time (sec)
T
0.2 0.25
1.5
•^ 1
o
•a
3
0.5
O.-0.5I-
E
< .
-1.5-
0.05 0.1 0.15
Time (sec)
0.2 0.25
Figure 5.7. The imaginary crossed power resulting from a retrograde-phased
actuator pair with the receiver oriented towards the source
In all tests of this type, orienting the two actuators, one horizontal and one
vertical, and driving them with a relative phase lag resulted in a combination of modes or
wave types that included both retrograde and prograde motion. After extensive testing, it
was determined that the vertically oriented source resulted in a simpler, single-mode
wave that was more suitable for detecting buried targets. Compared with other wave
types, the vertical source produced waves that had relatively less reverberation and a
consistent, single mode or polarization. The following testing procedures are based on
operating the seismo-acoustic sonar as a vertical source.
45

32-0.5-
o
,>,
Q)
"O
3
"5.
<-0.5
£ 0.5
o
«
Q.
1-0.5
-JH—b-
0.05
0.05
0.1 0.15
Time (sec)
0.1 0.15
Time (sec)
0.2 0.25
s_X - .... . V<rv . L..^. — ,
— j
!
0.2 0.25
Figure 5.8. The imaginary crossed power resulting from a prograde-phased
actuator pair with the receiver oriented towards the source
E. TESTING CHRONOLOGY
The bulk of the equipment was transported from NPS to the beach site in a large,
unfurnished mobile home. The mobile home was stationed on the bluff overlooking the
beach and served as a staging area as well as supplying electrical power for the beach
equipment. The equipment, including mine-like targets, was transported from the bluff to
the beach using the TrailGator 4x6 utility vehicle. Once on the beach, the equipment was
unloaded and set up according to pre-planned test configurations as shown in Figure 5.9.
After the initial equipment setup, optimization measurements were made to find the best
operating frequency and the associated wavespeed at that frequency.
46

Figure 5.9. Equipment setup at the NPS beach research site
47

1. Frequency
Based on previous research, the driving frequency for the source was initially set
at 100Hz. Typical velocities for this frequency were found to be on the order of about 80
m/s ±10, resulting in a wavelength of about 0.8 m. After many unsuccessful target
detection attempts at this frequency, consideration was given to testing other frequencies.
In general, it was found that frequencies below 100Hz worked best, consistent with
previous "ground truthing" measurements made to determine the frequencies and
wavespeeds of the dominant interface wave modes. [Ref. 8]
Generally, it was found that at frequencies of about 55Hz and below, the motion
of the drive plate, as measured by the accelerometer, became noticeably different from
the drive waveform. The cause of this apparent non-linearity is not known. At higher
frequencies, greater than about 150Hz, the interface wave was found to quickly attenuate
with range. Based on trial and error, various frequencies, starting from about 60Hz and
increasing in 10Hz increments, were tested on-site to determine the frequency that
resulted in the wave with the largest vertical component.
The evolving waveform as a function of distance for two, different operating
frequencies, 100Hz and 70Hz, are shown in Figures 5.10 and 5.11, respectively. Note
that the wavespeed for the 70Hz wave is slightly higher, as it penetrates deeper into the
sediment, where the shear velocity is higher. More importantly, the 70Hz wave is
attenuated less than the 100Hz wave, making 70Hz the optimal source drive frequency
for the sediment on this particular day. In general, the optimal frequency was determined
by placing the geophones at a fixed distance from the source, typically 40 feet (see
Appendix C4), and adjusting the frequency to find the wave with the largest vertical
amplitude.
48

25
20
El5 r
c
10
i i i
y i
Velocity = 70 m/s
Deviation = 9 m/s
/ i i i
0.1 0.2 0.3 0.4
Time (s)
0.5 0.6
Figure 5.10. Surface wave evolution for a 100Hz input drive signal
0.3 0.4
Time (s)
Figure 5.11. Surface wave evolution for a 70Hz input drive signal
49

2. Wavespeed
After determining the optimal operating frequency (test setup in Appendix C4), the
wavespeed was measured by cross correlating the vertical geophone signal with the
source accelerometer signal (see the MATLAB program listing in Appendix D7). The
cross correlation provided the time delay for maximum correlation between the two
signals, providing the time that it took for the wave to travel from the source to the
receiver. Based on the known source-receiver distance and the measured time delay, the
wavespeed was. quickly determined, within 2-3 minutes. The measured wavespeed was
then used for echo-ranging in subsequent background measurement and target detection
tests.
3. Background Data
After determining the optimal drive frequency and its associated wavespeed, the
geophones were repositioned at an appropriate distance from the source, typically 6 to 10
feet (1.8 to 3.0 meters), as shown in Appendix C5. A series of 10 background (i.e. no
target) measurements were made using a single cycle drive cycle at the optimal
frequency, and the geophone signals coherently averaged. This averaged "background"
data was then analyzed using the vector polarization filtering algorithm in Appendix D12
to determine the time that the wavefield, as seen at the geophones, had decayed back to a
nearly quiescent state. Based on the known wavespeed, the minimum range at which a
buried target might be detected was estimated. Using this information, the target was
buried at an appropriate distance from the receiving geophones.
4. Target Data
Two differently shaped and sized targets were tested in this research. The mass of
these two targets was varied, and will be discussed later. The target data collection test
phase was conducted by burying a target, typically 5 cm below the surface, at an
50

appropriate distance, and collecting 10 data sets, using the same drive conditions as for
the background data. As necessary, the data were coherently averaged and vector
polarization filtered (see Appendices D2 and D12 for MATLAB subroutines) for study
and target detection analysis.
5. Signal Processing
Numerous signal processing routines, more than 23 total, were developed, using
MATLAB's signal processing toolbox. [Ref. 18] These subroutines were invaluable in
allowing near real-time analysis of the measured seismic signals. As previously
discussed, the background and target signals were coherently averaged and polarization
filtered. Also available for target signal analysis was a coherent subtraction subroutine
(see Appendix D19), which coherently subtracted the background, reverberation field
data from the target signal. This capability allowed a comparison of before and after
seismic wavefield characteristics to determine and assess the detection of the target
signal. Also available was an incoherent subtraction subroutine (see Appendix D18)
which subtracted the background imaginary power from the target imaginary power.
Similar to the coherent background subtraction, this method offered an additional
capability to assess detection of the target signal.
51

VI. TEST RESULTS (VERTICAL SOURCE)
Summarized in this chapter are some of the important test results from a vertically
oriented seismic source. As previously mentioned, because of the versatility of the
seismo-acoustic actuator source, it was possible to conduct numerous tests studying
seismic interface wave generation and propagation. The results presented in this chapter
include wave attenuation, source velocity, energy, and displacement measurements,
target detection, and underwater testing.
A. ATTENUATION
Although it was not studied in depth, limited attenuation measurements were
made to help evaluate the types of seismic waves that were being generated. Based on
the theoretical results presented in Chapter HI, neglecting absorption, Rayleigh interface
waves can be expected to attenuate due to cylindrical spreading. As a result, the
acoustical power of the cylindrically spreading interface wave can be expected to
decrease as the inverse (1/r) of the radial distance from the source. Since the complex
power (4.1) gives the relative power of the seismic pulse [Ref. 7], determining the
attenuation of the seismic wave is relatively straightforward. Using the 70Hz seismic
source and evolving seismic pulse shown in Figure 5.11, the cross power function was
calculated and the maximum observed values were plotted as a function of range in
Figure 6.1. Curves for cylindrical (1/r) and spherical spreading (1/r ) are also included
for comparison.
The measured attenuation curve in Figure 6.1 indicates that the seismic wave is
attenuated less severely than for spherical spreading. In addition to spreading effects,
absorption from the friction of the sand grains rubbing together results in wave
attenuation. The results shown in Figure 6.1 qualitatively suggest that the vertical source,
operating at 70Hz, is generating an interface wave that attenuates due to cylindrical
spreading.
53

10
10
10
Measured
R"
2
Dependence
R"
1
Dependence
Cylindrical
Spreading
Spherical
Spreading
10 12 14 16
Range (m)
18 20 22 24
Figure 6. 1 . Seismic wave attenuation using two vertical actuators operating at 70Hz.
B. VELOCITY AND DISPLACEMENT
The velocities and displacements of the motion produced by the vertically
oriented seismic source were determined for the test setup shown in Appendix C2. The
recorded geophone signals are shown in Figure 5.5. The geophone velocities are given
directly by the output signal and the source velocity was found by integrating the actuator
accelerometer signal. The displacements were obtained by integrating the velocities.
The relative velocities and displacements for the seismic source and geophone (computed
sing the MATLAB subroutines in Appendices D20 and D8, respectively), 6-feet (1.8-
meters) distant, are shown in Figures 6.2 and 6.3, respectively.
54

2 100
3
O
if)
a
E
<2 -1 00 -
0) J I I L j L J L
05
£
E
o
O
0)
>
E
E,
«*
c
o
E
o
«
a.
Q
a i0
-a 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
4l=
. 2-
(0
a
a 1Q
-a 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
--•-•i i i i i
- - - - i - - - t - - • - f - - - i- = = =
4f
S 2h
•^
>1
^ 10
-3 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
- - -j - - i
'"rV "ri
i
- - - - i
-
A^/A 'J I I X /* t
•
.
i i ** i i i i i i
0.02 0.04 -0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (sec)
Figure 6.2. Relative velocities of the linear actuator shaft and geophones
t
—7 1 1 1 1
r
j L J L
-6 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
10
- 5-
(S
Jj
O^g-s 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
i i i i i i-
j [_ J L j i
J L
10
^^-6 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
10
ro 5 -
r
> -5(-
----- i i ' '
a"
' '
'
" " - - -i - - - 1 - - - - r - - - n i
,...._
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (sec)
Figure 6.3. Relative displacements of the linear actuator shaft and geophones
55

The average power transmitted from the actuator to the sediment was estimated
by calculating the product of the actuator force times the drive shaft velocity. It was
determined that approximately 0.6-watts of seismic power was generated at the drive
plate of a single, linear actuator system.
C. TARGET DETECTION
After months of testing and equipment modifications, two buried mine-like
targets were unambiguously detected using the linear actuator seismo-acoustic sonar
system. The first target was a 156-lb (71 -kg) helium gas tank, shown in Figure 6.4. The
gas tank was modified so that the top could be removed to allow for the placement of 26-
lb (1 1.8-kg) lead blocks inside, thus increasing its mass.
Figure 6.4. Buried helium gas tank target
- .
w&*^^&^sk-f**Zi
The second target, shown in Figure 6.5, was a 1.5-feet (0.68-meter) diameter, 1-
foot (0.45-meter) height gunpowder keg weighing about 16-lbs (7-kg) empty. The
removable lid allowed placement of 26-lb blocks inside, shown in Figure 6.5.
56

Figure 6.5. Buried gunpowder keg target with 25-lb lead blocks alongside
A test was conducted to detect the buried helium tank in Figure 6.4. The test
setup, outlined in Appendix C5, consisted of two linear actuators, oriented vertically, and
driven with a single cycle pulse at 70Hz. The geophones are 5-feet (1.5-meters) from the
source and the target is buried 11-feet (3.35-meters) from the source. The wavespeed
was determined to be about 290-ft/s (88-m/s). The background and target signals, using
10 coherently averaged records (subroutine in Appendix D2), for the source, and the
relative the radial, shear, and vertical components of the geophone motion are shown in
Figure 6.6. Note that the target signal, which should be seen at about 0.08 seconds, is not
readily identifiable in the raw, recorded data.
The crossed power for the background data, target data, and target data with
background coherently subtracted out was calculated using (4.1) and the results are
presented in Figure 6.7. The measured wavespeed of the day (67m/s) was used to
transform the time-based record to a range-based record.
57

^^ co
o o
> cc
a
o•*-H
*—
>
o
s
u
> CO
4—1 o
C3 .c
CO
1h
-1
T
, n A
w r-
u
v
T
0.02 0.04 0.06 0.08 0.1 0.12
Time (sec)
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
0.14 0.16 0.18 0.2
Figure 6.6. Relative source and geophone motions for the background and target data.
20
Range (ft)
Figure 6.7. Imaginary crossed power for the background and target data.
58

Note that the imaginary crossed power gives a clear indication of the presence of
the buried tank at about 17-feet (5.2-meters), as expected for a round trip range of 1 1-feet
(3.3-meters) to the target and 6-feet (1.8-meters) back to the geophone. The negative
deflection of the imaginary power indicates that the motion of the target echo is
retrograde.
Similar tests were conducted two weeks later. The tests used both the helium tank
and the gunpowder keg, which were buried a few days prior to the tests. In these tests,
the geophones were positioned at 10-feet (3.0-meters) and the targets at 1 6-feet (4.9-
meters). The test setup is outlined in Appendix C6. The mass of the targets was also
increased in 104-lb (47-kg) increments using the lead blocks shown in Figure 6.5. The
RMS values of the relative motion for the buried, 5721b gas tank are shown in Figure 6.8.
Note the apparent evidence of the target at about 23-feet (7-meters).
0.5 -'
ifC^T" ": : : :
'"
o (0
>v—
'
cs
c IX
o
4-1
o
2
<u
>
*->
es n
0.5 -
1 -
0.5 -
~ 0.5 -
10 15
15
Range (ft)
20 25
Figure 6.8. Relative RMS source and geophone motions for the buried 5721b gas tank.
59

The crossed power for the gas tank with increasing mass loading was calculated
using (4.1) and the results are presented in Figure 6.9. Note that the imaginary crossed
power gives a clear indication of the presence of the buried tank at about 23-feet (7-
meters) as expected for a round trip range of 16-feet (3.3-meters) to the target and 6-feet
(1.8-meters) back to the geophone. The relative target strength appears to be related to
the target mass, as seen by the increasing imaginary power amplitude with increasing
target mass. The positive deflection of the crossed power indicates that the motion of the
target signal is prograde, which is surprising. It should be noted that, both targets had
scoured themselves to a depth of about 0.5-feet (0.15-meters) due to water action from
relatively high tides. It may be that this results in target scattered signal motion that is
prograde.
0.02 -
0.015 -
22 23 24
Range (ft)
Figure 6.9. Imaginary crossed power for the helium gas tank target
Similar results were seen for the gunpowder keg, which was left buried in the
beach for 4-days after the helium gas tank measurements. The crossed power for the
60

gunpowder keg with increasing mass loading was calculated and the results are presented
in Figure 6.10. The results are similar to the buried gas tank, which was detected four
days earlier. The presence of the keg is clearly indicated by the positive deflection of the
crossed power at a range of about 24 feet (7.3-meters). As with the gas tank, the relative
target strength increases with increasing target mass.
0.04
0.03
0.02
= 0.01
o
>
5
o
a.
•a
0)
-0.01
o
-0.02
-0.03
-0.04
20 22 24
Range (ft)
26 28
I

120lbs
224lbs
328lbs
432lbs
-/xl —^^— bSblDS
^Qr^Vr
Sl^
—
i
30
Figure 6. 10. Imaginary crossed power for the gunpowder keg target
Relatively unambiguous target detection using the vertically oriented seismo-
acoustic source was seen for both the helium gas tank and the gunpowder keg.
D. UNDERWATER TEST
Near the completion of this study, a single field trip was made to test the seismo-
acoustic sonar system underwater. Unfortunately, high tides and storm surge prevented
61

successful deployment and underwater testing. The lessons learned from the test are
included to help support further testing in this area.
The underwater testing initially started at the beach site. It was found that the
300-feet underwater actuator and geophone cables were not long enough to allow
deployment just beyond the surf zone. Although not available for that particular day, a
small boat would allow deployment of the system near the surf zone, from the water side.
Since deployment from the beach was not possible, the equipment was loaded
into the mobile home and transported to the municipal wharf at the Monterey Marina.
See Figure 5.1. Unfortunately, the relatively strong storm surge did not allow underwater
placement of the actuators or geophones. Typically, this site would be ideal for
underwater tests, as it is known for its relatively good visibility and flat, sandy beach
bottom. The location is also ideal because it allows use of the mobile home to stage the
electrical equipment and has easy beach access for water entry, as shown in Figure 6.11.
Vy^^fe
Figure 6.1 1. Underwater test site showing easy water access (the author and his spouse)
62

VII. CONCLUSIONS AND RECOMMENDATIONS
A. CONCLUSIONS
A uniaxial seismo-acoustic sonar source based on a linear force actuator was
developed for this research. This source resulted in an impressive 60db improvement
over the previous seismic source that was developed at NPS. [Ref. 8] Previous efforts at
ARL:UT resulted in the detection of a buried object at a range of 3-meters after
coherently subtracting the reverberant background data from the target data. [Ref. 7]
Clearly it is not practical to rely on background measurements to detect buried mines in
an operational environment. This source detected buried mine-like targets at ranges up to
5 -meters and did not require coherent background subtraction.
The linear force actuators worked exceptionally well to generate seismic interface
waves with minimal reverberation. The complexity of the seismic wavefield was reduced
and buried target detection was enhanced. The results of this thesis support the use of
seismic interface waves to detect ordnance.
B. RECOMMENDATIONS
While the linear force actuator source resulted in successful target detection, it is
limited in the amount of force that it can apply to the sediment (25-lb peak force).
Therefore, the intensity of the seismic interface wave and the effective operating range of
seismo-acoustic sonar system is limited by the force capability of the actuator-based
system. To increase the intensity and range of the system, a single, waterproof actuator,
capable of 100+lb peak force should be acquired and used in subsequent studies.
In addition to a more intense source, an array of receivers should be developed to
determine the azimuth or bearing to the target. The combination of an intense seismic
source combined with an array of geophone sensors will allow for bearing and range to
target measuremencs at greater standoff ranges.
63

Along with the development of a seismic receiver array, a MATLAB or other
appropriate subroutine should be written to make practical use of the range and azimuth
data provided by the array. Conceivably, within 2-3 minutes after recording the data, a
polar plot display, similar to the display presented in Figure 7.1, would allow for a near
real-time assessment of buried mine-like objects in the seismic wavefield. The display
could also be updated as more data was collected and processed.
Figure 7.1. Polar plot display showing bearing and range to a buried target [Ref. 7]
Bearing
-15 -30
105
• source location
o receiver locations
6 7 8
Range
-1 o 1
x10
The SPS390 Data Acquisition and Signal Processing System has the capability to
store and average data based on an external trigger signal. To support increased and
more effective data collection, the averaging feature available with the SPS390 should be
used. The SPS390 owners manual [Ref. 19] contains an "Averager" conversion program
that would allow for the conversion of the .avg file format into an ASCII file format
compatible for analysis using MATLAB's Signal Processing tools.
64

Finally, underwater tests should be initiated as soon as possible. In theory,
underwater Scholte waves are nearly identical to Rayleigh waves. The seismo-acoustic
source developed in this research should work equally as well underwater as it did on the
beach. However, this needs confirmation. In fact, since naval mines are typically more
massive (5001bs to 20001bs) than land-based mines, it should relatively easier to detect
them using a seismic-based sonar. Conceivably, an autonomous underwater vehicle
could be used as the supporting platform for a seismo-acoustic sonar system to detect
buried naval mines in the shallow water near the surf zone and seaward.
65

APPENDIX A. EQUIPMENT SPECIFICATIONS AND SPS390 SETTINGS
This appendix contains the equipment specification sheets and SPS390 data
acquisition settings that were used during this research.
Appendix Al - Aura linear actuator and ECU specifications
Appendix A2 - Linear actuator force constants
Appendix A3 - Actuator accelerometer specifications
Appendix A4 - ECU/Actuator interconnect diagram
Appendix A5 - Geophone specifications
Appendix A6 - SPS390 setup and menu settings
Appendix A7 - John Deere TrailGator specifications
67

APPENDIX Al. LINEAR ACTUATOR AND ECU SPECIFICATIONS
I—
I
o
C <
« «
C/3 .2
« <
a »>
< m
O W) Onm On e
2^
1
4 2 5?
^g §
uj o .a
c<-i 5
"? §en »
SO *-
o —
K
<
0)
LU
o
<
DC
oLL
DC
LU
Q.
DC
o
D
h-
O
<
LU
a
X
a
3 o X
CM
£2
OID
o ocm in
CO CO
ooco
in
CJ
co
CM
in
in
V
O
2_
5-s
? CO
r^ '—
-
CM
CD
o
><
q
V
in
*
CM
<
2
CM o oCD
Oo
H
Q
>_I
LU
z
o
z
LU
<
X
in oin
CM
CO CO
3
CO
d
o
o oCO CO
oin
CD
d
CO
co
in
V
cm in
NT" CM
S o
CO »-
CO ^~-
CO
CD
o
><
CM
m
V
co
d
CD
CM
CM
•<
2
o o oCD
oo
h-
Q
>
LU
z
o
z
X
LU
<
X
oCO
omCM
lO CO
01
co o o in CO CO in cm in
o ->
S oq t^
CO NT-
CO
—
CO
CD
o in
co CD <: o o oCO o
1-
n
LU
7
X
LU
V <
o
55
CO CD CM o "3- V «- CM
X V co CM z i^ CO > o
z
<
X
CO Z
CM CO
3
in
d
oCM
in m»- CM
IfCM
co
o
TT m
V
CM CO
CO n—
o —O oOnCM n—
in —
CO
CD
o
Xin
in
CM
d
V
q to
z d
o oCD
oo
H
Q
>_1
LU
Z
o
z
X
LU
<
X
CO
oCO
oin
CM
09
in co O m Tt in CM CO
o —O oo rC
CM *"
in ^--
CO
CD
o
in
CM CO in .<
p~
o oCO o n
LU
z
X
LU
V n <
go
«- CJ " o *"" V CO n—
X oV
CO z d t-. CO > o
z
<
I
co Z
*" t/J
a. n X
CM
CO
OOCO
O OO CO
o o
oin
co
CM
CO
in
CO
in
V
CM O
CM ^
o _.
«c co
CO CO
CNJ **
CD
CD
o
Xco
oCM
V
oco
o00
CD
ooco
CM o oCD oCO
Q_i
2
LU
X
CO
LX
m
or-
<
31-
O
<
in <
z
a. o _, r»
a
LU
X
o
CO
ooCO
o oo in
co f
f-
CM
CM
CO
CO
in in
V
CD
^~"
CMS g-8
CM
^^
CD
o
Xco
o
V
o
dCM
o
in o co
o oco oco
r—
Q_1
2
I
CO
3
tr
m
<
3
O
<
m •<
z
0. o _, f~
r-
o_l
2
a
in
X
O
CO
o
o
o oo mCM m-
oo CM
CO
in mV
CD
*"*
CMS
8o
g
-8
CM
~^
CD
o
XCO
o
r--"
V
dCM
CO
CM
CO
o CO
O1^
oCO oco
X
CO
3
X
m
h-
<
3
r-
o
<
m •<
z
X
a.
CM
o o f-
in
CM m U!»
o _
g s
CD
CD
O -I o w
in
in o oCD o n
LU
z
o1-
< o <
CM
o
NT- CO
CM
•" V
X V CM l-~ *~ r- CO > u
z
3
H
CO z
U
<
X
a.
CM
o
o o<— CO ^ o
CM
CO
CO in
V
CO
o —O o
O T-
CM ^-~
CO
CD
O
X
dV
en
CD o
d in
CD o oCD oCO
i-
Q
>_l
LU
z
o
z
u
<
3
Fo
<
oco z
X
CL
o o»- in CO q <r in
V
o in
c CO
o —O o
o »-
CD
o
X
CM
dV
CM o
d in q o OCD oCO
1-
Q
>
LU
z
o
o
r-
<
3
oCO zo —
-
in _i ^ H
A
<
X
a
in m o o CM in m in co
o —
.
o o
CO
CD
o in
dV
O <j-
in
in
o OCD o Q
LU
Z
o
< o <
go
n— CO
d y~ V CM n—
in
•"
Xin
*~
d d t-- CO >_i
u
z
J
P
O
CO z
<
'c*
o
o
S?
CO
ID
CO
£l"
O o CO
LL ^ a.'
^-*
O V ft
CD
b
<
CD
CL
CD
O
O
LL
l/>
"O
C
O
CL_
C
ra
c
o
c> 10
o
Q
CO &
^ "D
^3 CO
_1 10
•o
c
3 CD
_o
CO
"O
c
o
0.
o
CO
13
To
IT
a.
E
<
CD
10
ID
O
w
CNJ
u "c/T
CO
3
LL
<
CO
DC 'co
LL?
CD
13
3
2
CD
Q.
E
CD
1-
LL
o
<
E
3
I
CD
CO
JC
CO
o
CD
CO
E
cd"
'co
a.
CD
3
CO
CO
BE
LU
CD
3
3
z
a>
o
c
a
E
o Q
£*
j? TO
O T3
m cu
m 2
CU CD
CD
O
o
ll
CO
3
o
3
o
"co
CD
.c
o
c
CD
>
o
crj
CD
c
_l
c
o
B1
Id
5 S
CD t
"CD 3
O
"• CD
F <5
CO 10
IsCO co
8|
LL
WLL oj
F S
rM
I
"o
CD
E
3
CD
(0
u
CD
o
LL
CO
oc
3
O
IX
"53
>
CD
_l
CO
"D
c
3
O
0^
CO
CO
CO
CO
"D
c
3
O
CL
EOl
CD
5
3
o
CL
"
CO
O
_1
ID
"D
CO
F
E
.c
O
CD
O
c
ra
CO
CD
LX
c
CD
E
o
">
c
LU
O)
c
CD
a.
E
CD
r-
O)
c
as
CD
CL
O
co
CO
O
O)
5
Q.
O
F
CD
&
CD
Q.
>.
-
c
o
o
o
15
3
O
<
CD
&
c
CO
ra
I
^T
u
o
c
CO
F
CL
c
o
CD
E
n3
CO
F
_i
U- LL
o (
1
7 3 ." 3 ." b
r-
CD
c
3 3 >
3 CD
U
>
3 3
LU
Q
o
o
r
•o
CD
CD
o
CD
o
CD
O II ii
CD
CO
O E
X O
CO
CD CD
b
'x
u
o
10
c
b
F
b
X
b
X
CO 03 ,—
.
CO —
I
c o CO C> a. CO CD CD o CD
S a a a. cr co ll LL •^^ _l v CO 3 *, :> a V a i a & CO V a i >
68

= C <N —
a <
<
3 O g?
a
££@=s * c
9 "3
«n g
E
<
(0
UJ
o
DO
oLi-
CC
LU
Q.
o
CC
H
Z
o
o
o
z
o
CC
oUJ
-I
LU
n
CO
oCM
OCM
OCO
oCM
ooCO
co
CO
oCM
oCO
o o o o CM
CD
d
CM
CD
d
o
cr
O
CO
c
<
o oCO X X X X
3
O
LU
n
CO CO CD
o CO o o o o o
o
o
J?
d
cn
o
CO
c
<
CM
oCD
oCM OCD
ooCO
CO
oCM
oCO
+ ~+
^P
"+
o o o X X X X
3 CM CM
CJ
UJ
to
CO
o CD CD
CM
Oin
OCM O ooCO
CO
CO
o
oCO
o o o o o o
CO
o
re
o oCO X X X X
3 CM
~+
+ ^F
"+
o d aO CM CM
UJ
d
.-
CO a> ©
o
in
OCD
in
•3-
ooCO
CO
oCM
oCO
o o o o o
CO
o
o
to
d
o
ao
re
c
<
o oCO X X X
3 CM CM
OUJ
,-
CO
o o o o o cn
cm
mCM
3
in
a-
in
CM
ooCO
CO
OCM
in
"? ~+ "+ ~+
CM
CM
CO
CM
CM
CO
o
CO
o
CO
c
<
o oCO X X X
o
UJ
o CD <E
y. ,_ o o o o O o
cn
re
c
25 OCM CO
in in
CM
o1^ s o o
CD
o oCO X X
3 *" ~+
# + "+
o o <
o CM CM
UJ
CO
o CD ©
s CD CO
o s o
o o o o o o o
cr
o
re
c
o oCO X X
CM *™ + "+ "+
+ o o <3 CM CM
O
UJ
<
CD ©
o
CO
3
O
CD CO
or- s N
o o o o
+
O
oCM
o
"to
dCM
o
CM
o
CO
cz
<
o oCO X X
UJ
5
CO
jd.
E
3
17r
2 Hi JJ
a
E
<
<
CO
*co
i2~
"o
>
CO 5CO c
SB
= 1
§>°
TO <->
— CD
Sis
c ©
r- <°
"o
5^ i?
"a?
o
u
c:
co
c
o
> u?
CC
UJ
m
s
Z
_l
©
o
c
TO
E
CO
a
E
<
c
E
o
"D
CO
a.
Q.
CO
re
s
o
</>
3
o
3
c
c
o
O
E
3
o
>
CD
CT
s
o
>
co
co
3
CD
>.
aa
"o
>,
CI
Dl
3
o
>
>.
Q.
Q.
o.
co
CD
5 -«
E
<
£
2
(J
>
a
a
00
CO
5
©
o
TO
r
o
CD
o>
—
c eg
re -x
CC 2
§,55
re CO
> 00
HyJ a)
£ 3
o
> ^~
CO c
O) C3
S|1 1
S,o
.=: o
> o
_ u-
3 CO
9-ro
J= CJ
>
cn S
eg
rr S
cn a?
co CC
!i3 LL
S- 00
O 0)
CD
a.
o
>
c
"re
CD
3
ac
c
©
CL
CO
O
>
'co
CD
3
o.
3
o
CD
CL
to
CL
E
<
c
'co
a
o
c
o
2
©
a
©
©
E
c
o
~>
c
UJ
Dl
c
©
3
re
©
a
E
©
H
cr
c
©
a
O
LL
b
<
a
E
3
X
© to
©
J3
CD
g
o
CL
>.
o.
Q.
©
3
o
0.
c
3
o
c5
2r
3
Q
UJ
Q
O
o
r
E
'x
O CO
Q. co
o
0-
.9 oo
fife
.2 w.-= CO
CO o
<u CO
o =
£fe
<D W
CO =
o
'co
o
CO
c
©
o re
©
"D
©
>
CO
3
00 o
o
CO
c
>
re
© 0) CC 3 O x: o o ° ^ iffe
o o 5 o a. re © 10 o o © ©
E Q. a. 2 00 < 0. < CJ a. a. O O O cr CC C_> < 5 00 I
69

APPENDIX A2. LINEAR ACTUATOR FORCE CONSTANTS
20
18
16
14
12
10
8
6
4
2
00 F
00
00
00
00
00
00
00 F
00
03
A101A ACTUATOR FORCE CONSTANT SN 161
00
j
_,^:
R 8
CO
R
CO
R
CO
R
CO
CURRENT IN ANPS
8
CO
s
10
FORCE CONSTANT: 4 to 16 + = 1.49
COfTlENT: UUT MOUNTED VERT 1.7H ADDED TO Y AXIS, MID-STROKE
FILE: A101F161.DAT SETUP: A129-FC.DSU 10-24-1997 11:03
20.00
18.00
16.00
w 14.00
§ 12.00
S 10.00
- 8.00
$ 6.00
u. 4,00
2.00
0.03
A101A ACTUATOR FORCE CONSTANT SN 162
|
_^^_
Lj l_i J--1-1 1 11 1.1 I. I..I..1 .1_1- X. J..I 1-1-1.1 1 L. 1_ I I 1 1 I I I I. 1 .J.-JL 1 1 1..1. i.-1-L-l.l 1 , 1
8 8 8 R
G> «H CO ^T
§S> CD SO © CD OK S io S K S
cd r-' co s> oj co in
«-4 «H v4 ^H
CURRENT IN AMPS
FORCE CONSTANT: 4 io 16 + = 1.2
COMMENT: UUT MOUNTED VERT 1.7H ADDED TO Y AXIS, MID-STROKE
FILE: A101F162.DAT SETUP: A129-FC.DSU 10-24-1997 10:54
70

APPENDIX A3. ACTUATOR ACCELEROMETER SPECIFICATIONS
with built-in electronics
Dirensions shown in inches except where noted.
ID
English Selection Guide
5-44 Thd Coax
5/1 g Connector
Hex^V -^.250K)
10-32 Thd Coaxial
Connector-7
k j§§ 9/32
.65
.10
*£},2
hH I
Di
5-40 Thd *
Inlegral Mtg Stud
351 B11
10
V i—-275
5-40 Thd
DiaTyP
Integral Mtg Stud
10-32 Thd Coax
Connector
Dia
/ -I.25 U
Hex 10-32 Thd
Mtg Hole
351B14
(351 B1 2) Tod connectc model numbers listed in parentheses
351 B03
(351 B04)
•75 Dia
V^(3
351 B31
351 B32)
HIGH FREQUENCY / LOW MASS GENERAL HIGH SENSITIVITY
MODEL NUMBER Unit 351 B11 351 B1
4
351 B03 351631 351 B41
Voltage Sensitivity (@ 4 mA); (1.2) mV/g 5 5 10 50 100
Frequency Range (±5%) Hz 1 to 10 000 1 to 8 000 1 10 6 000 1 to 4 000 1 to 2 000
(±10%) Hz 0.7 to 15 000 0.7IO 10 000 0.7 to 9 000 0.7 to 7 000 0.7 to 3 500
Resonant Frequency kHz >40 >40 >35 >22 2 14
Amplitude Range (3) ±gpk 300 300 150 30 15
Resolution (broadband) gpk 0.01 0.01 0.005 0.001 0.0005
Mechanical Shock Limits ±gpk 5 000 5 000 5 000 2 000 1 0O0
Temperature Range (4) »F -320 to +250 -320 to +250 -320 to +250 -320 to +250 -320 to +250
Temperature Coefficient (5) %PF see graph see graph see graph see graph see graph
Amplitude Linearity % ± 1 + 1 ± 1 ± 1 ± 1
Transverse Sensitivity % <5 <5 S5 £5 <5
Base Strain Sensitivity g'Mt < 0.002 < 0.002 < 0.0005 < 0.0002 < 0.0002
Excitation Voltage VDC 1810 30 18 to 30 18 to 30 18 to 30 18 to 30
Constant Current Excitation (6) mA 4 4 4 4 4
Output Impedance ohm <1200 <1200 £1200 < 1200 <1200
Output Bias Voltage VDC 3 to 10 3 to 10 3 to 10 3 to 10 3 to 10
Discharge Time Constant second £0.5 >0.5 i 0.5 >0.5 2 0.5
Sensing Element quartz tri-shear tri-shear tri-shear tri-shear tri-shear
Connector type 5-44 coax 1 0-32 coax 1 0-32 coax 10-32 coax 10-32 coax
Sealing type weld/herm weld/herm weld/herm weld/herm weld/herm
Mounting Thread size 5-40 5-40 10-32 10-32 10-32
Housing material titanium titanium titanium titanium titanium
Weight gram (oz) 2 (0.07) 1.8(0.06) 10(0.35) 25 (0.88) 32(1.1)
OPTIONAL MODELS: (7)
Top Connector 351B12 standard 351 B04 351 B32 351 B42
Ground Isolated (>l08 ohms) J351B11 J351B14 J351B03 J351B31 J351B41
Adhesive Mount A351B11 A351B14 A351B03 A351B31 A351B41
SUPPLIED ACCESSORIES: (8. 9)
Mounting Stud (10) N/A N/A 081 B05 081 B05 081 B05
OPTIONAL ACCESSORIES: (8)
Adhesive Mounting Base 080A15 080A15 oeoA 080A1
2
080A12
Magnetic Mounting Base 080A30 080A30 080A27 080A27 080A27
Triaxial Mounting Adaptor 080A16 080A16 080B10 080A1
1
080A11
CABLING: (8)
Mating Cable Connectors type G A A A A
Recommended Stock Cable series 002 002 002 002 002
NOTES:
".
. All models nave a sensitivity tolerance ol +10%.
2 Sensitivity determined with 4 mA constant current excitation.
(Consult PCB lor calibration at other current levels.)
3 Amplitude range is tor 3*1.5 volt output tun scale.
4. In-house testing capaDilities limited to -320 F (-196'C). Sensors
nave been successfully operated in liqu'd helium environments.
5 Within 5% of typical graph
6 Operational range: 2 12 mA Sensitivity drops 1% for each mA increase
7 See page 8 for a descnption rt the optional models (Specifications may
ditler slightly Consult factory before ordering.)
8 See cables/accessories section beginning on page 78 for a complete
description of the supplied and optional accessories.
of ,„
Typical Sensitivity Deviation vs. Temperature
- 1 +20
i®%0
II "10
* ; -20
.
_ I
" - - _ 1
1
1
•300 -200 •100
Temperature ( F)
+200
9 NtST traceable calibration certificate from 10 Hz to +5% frequency point
supplied. Exact temperature coefficient at -320 °F (-1 96 'C) also provided
10 Supplied mounting stud is threaded 10-32 to 10-32.
71

APPENDIX A4. ECU/ACTUATOR INTERCONNECT DIAGRAM
ui
1
* 1
« -
>- g 1
_J X
CD
2UJ
</l
<n
r 3
1
r
1 r V
~1
1 u.
m
t—
z
3
{/I
o
t
<
^
u
•a £O) CJ
te
o
3
o<
O i
S
i i 3
<
1-
i 1 Zii UJ
1 ' CC
o
z
oC£
t-
uLJ
UJ
J|
-
i §
PS U
8 §i
i
a cc u. u.
bJ ce
o o
i zi cc
d o UJ
3 •x. z
rr t)
a
7
LU
3! i
| o o > > > UJ UJ
t- i- o o o ce or
! ! ° ° T *~ *T
1 5 o
z
o
z
a i_
CD| >- ce CD CO U Z C CO *
J ! L
1- -H 1 T 1
1
cf <1= 1CD<C_>OUJU_OX — "? — CM — • CM ro •* in = J 3D
Tl
< <-I1
h M )
0.
<
o
— i c )
1
UJ
en |
<C_)
y H )(
1 1^-5
^T^
g
s
_J
&l-CMlDin K) » ° ^ ^
^
)
1
-J VCI
e
< i i
z
oX
o
X
o
a
3
CJ
UJ
<
z ifi
.a
5
2
<
a:
o
o
a<.
o
o o
z
CO
o
z
a
ite St q: < —
z
1 l
°I §1 03 p Qo Q.5 llS a o
2 cjz
> rf^ ^* @ >V| ) • • I Wr-
z r~ ii i
ou O ^
^1
UJ
s
0.
1/) r
-^r-i*^<oi-oocn°^ {v'^'«- ,n to
^»2!°^SSS m id
Csl CnJ
1 3 °
UJ p^.1 J I J
O S » " » ' LJ
oQ-
1tr i.
I qmvd Nid no 1—
-z.O O
m O ~ N * 5 5; O O ^
1
O>
§ -«g gl 1 1 1 i 1
1
t
"53 z,
2,uJca.a.a.o_cL z z <->*-> o o _.
3c-
3
J 33Op-~ c_ Q.a_0_a.C-OOOCi OO UJ _ O
= = H M,2, 333333ZZZZ CC^^q^Qz Z z
2
>>>>>>>>>>mPQOO<<5 >
« o
> e
lb
5
Z
'
Boa! oo<niOinoommm<Q.Q.u.u.jJS.'ioo
33333+ I++I 1
-H-h-h-m5+ I + 1 <C§-H+I
m i/>
- O 1
H 0.
;
L_ _J 1
j
72

APPENDIX A5. GEOPHONE SPECIFICATIONS
(Made by Halliburton Geophysical Services, Inc.)
SPECIFICATIONS FOR SM-6 GEOPHONES
Parameters Symbol Unit Geophone Model
Frequency
Natural Frequency
Frequency Tolerance
Maximum tilt angle for specified
Fn (vertical units only)
Typical spurious frequency
Fn
A .B B B B
Hz
DEG
Hz
4.5
* 0.5 Hz
4.5
± 0.5 Hz
8.0
± 0.5 Hz
20
150
10.0
±5%
25
170
14.0
±5%
25
190
Distortion
Distortion with 0.7 ia/s pp coil
to case movement
Distortion measured at
frequency of
Maximum tilt angle for distortion
specification (vertical units only)
Hz
DEG
< 0.3%
12
< 0.3%
12
< 0.2%
12
15
< 0.2%
12
20
< 0.2%
14
20
Damping
Open circuit damping for
375 ohm coil
Shunt resistance used for
damping calibration with
375 ohm coil
Damping with shunt for
375 ohm coil
Damping tolerance
Bo
Rs
Bt
Ohm
0.265
±5%
0.560
= 5%
0.315
2257
0.60
±5%
0.250
1339
0.60
±5%
0.180
645
0.60
±5%
Standard coil resistance
(other values available
Re 375 ± 5% Ohm
Sensitivity for Model 'A'
Sensitivity for Model 'B'
28.0 ± 5%
0.71
28.8 ± 5%
0.73
V/m/s
V/in/s
V/m/s
V/in/s
RtBcFnforModel'A'
RtBcFnforModel'B'
3875
6000
Ohm Hz
Ohm Hz
Physical Specifications Basic unit (element) PE-4 case PE-5 case HPE-case
Length - 40.0 mm (1.57 in) 40.0 mm (1.57 in) 57 mm (2.24 in)
Width - 35.0 mm (1.38 in) 35.0 mm (1.38 in) 45 mm (1.77 in)
Diameter 25.4 mm (1.00 in) - - -
Height 36.0 mm (1.42 in) 56.0 mm (2.20 in) 62.0 mm (2.44 in) 35.0 mm (1.38 in)
Weight 81.0 g (2.86 oz) 1 83 g (6.46 oz) 141 g(4.97oz) 1 92.5 g (6.79 oz)
Guarantee with normal use* 2 years 1 'A years 1 'A years 1 Vi years
Suspended mtst Moo* 'A
-
76 1
g? mode! guaranteed tor one year
ai use excludes damage caused by high voltage and physical damage (o case _ yeM'
tf
}}
,
l 9
,rt ,« „
Toot cod ID uu mcumart 4 mm (U.16 m
Sensor reserves the right to alter specifications to offer the best possible product.
NEV
SmMriry
N«jra( kaqunc-ji
Optmng temp**w» rar-g* -40*C B * 100*C Opefi C-i-CUff Dvnpng
|-«C-F 10 +?l*f-l Cod Curw« DvT-eng : S C *
aj pt/wruMo x>*ci*«d » 20*C Cod fte»-nnct
Shunt RMoanc*
TMil Damping
Tot* tktmmanm
PlSc-Fi Uod* 'A = 10 S Re
Mod* B- - 14. Ft
Trvuducaon oorvatnt
Mod* A' C = OJ3367 /"Re V'ni/»c tSX I.44S n/"Ac Vim/**:. £5%
Mod* ff G = 0.0377 s/flc V/H/HC tSX. 1-*»* ,/5c V'mtmc. ±5%
73

The SM-6 is a digital grade, miniature geophone based on the popular SM-4 design.
Engineered for consistent performance over a long life.
Advanced computerized testing, manufacturing techniques and quality control are used in the production process
to provide both the uniform parameters and the rugged qualities necessary in modern geophones.
Features of the SM-6
• Available coil movement of 4 mm (0.16 in) which
makes it suitable in a vertical mode and high tilt for
lower natural frequencies and for industrial purposes
where the force of one G is exceeded.
• In the horizontal mode the large coil excursion of the
10 Hz model makes it suitable for shear wave
detection with planting faults of up to 20 degrees
without resorting to cumbersome leveling devices.
• Symmetrical rotating dual coil construction minimizing
the force on the spring arms.
• Use of precious metals ensures optimum electrical
contact and long operating life.
• The unbreakable and lightweight PE-4, PE-5 and HPE
case provide for easy access to the element in the
field.
• Sensor's UV stabilized rubber sleeved gasket
minimizes cable fatigue and has excellent watersealing
capabilities.
SM-6 FREQUENCY RESPONSE CURVES
30
10
1 1 !
-I:"! 1 i
ttame o*rtcro« OOMI cu«v€
~r*V*t MOOO. a
CA.WS3TMCI inomts
$P«»rTnrtTy a ti vi i »****
»V»T CI ommno %
f ^l
-r i
1
~«
>3J«I !->Q
2 os
o
/ / J .'S«L—m T
-i
/ ^
If
/
'
02 ///
1
J// .
:
;
i
1
"
1
Mr
J
i
E
J7» »
I
|
1
Tv»esM fl UOOCB
1
colmcsstmcx ncHa
i
? ~
A !
sonsmvTTYan
muma
/' m .
S- '
1 1
'
oc
515
1QK )
685
N,
vJ*—
!
,
If
r j
——
i
.
1 ! 1 ' |
'/
s | |
J
*
1 |
Jv
'/
/
iii
in 1
i
[
i
«
ll!1
'
1 '
: 1
!
/ <
i
;
1
| I
1 1
1
1
II!!. i ' ; I
•arouse a*vt
Tvn sm» wooat
JO"*T"<TTY
OAWXMOX
J» 300 ^00 80D
20
Ot/rwr * nwoue*cr
_L N
K.I*Wi»««M"D UHZ
10
t>
MK3rTtvrrr QM vn-1 NV*t
.1
1
^S[
If rw'_. : 1
ll
y/ / ^"P^fiJt'
/
Z i
-
1/ U' . !
1
02
/
//(I !
h ' i
n
W ' i ! . i
1 1
h
>
i I'll. 1
1 1
J '
i
' 1
i ii -
R '
1
.
i/4 ! ! ! i Mill: 1 .
oum/r -» r«iQufio
HAT**i <HTOufO> »*«Z
Ctx *WT*»Cl no-i
700 300 *» 600 -XL) 300 *X 600
ALL SALES AND DELIVERIES ARE SUBJECT TO SENSOR STANDARD TERMS AND CONDITIONS INCLUDING THE GUARANTEE PROVISION
STATED THEREIN.
74

APPENDIX A6. SPS390 SETUP AND MENU SETTINGS
File Menu
Open Configuration -> std8chan (configuration settings are listed below)
Save Extended Record -> used to save .xrc data files
Setup Menu
Acquisition
Coupling -> AC
FS rms -> user selectable, typically 1-5 volts
mV/EU -> 100
EU Name -> Units
db Ref ->
V Ref ->
Source -> External
Repetitive -> as desired
Source Channel Delay -> -100 samples
Sampling Method -> Internal
Frequency Span -> 4fc#z
RPM Readout -> deselected
Analysis
Channel... Active -> select all 8 channels
Channel... Reference -> select channels 1 &2
Block Size -> 4096 samples, 1600 lines
Overlap -> 0%
Process Weighting -> Manning
Memory -> 8192 samples
Domain -> Cross Properties
Method -> Linear
Stop Criterion -> Single
Preferences
DDE Server Capability -> selected
Display Buttons -> selected
Controls on Top -> selected
Overload Audio Tone -> 500Hz (as desired)
Shorthand Notation -> deselected
Peak Interpolation -> deselected
75

Setup Menu(cont)
Display
Function -> Time Trace
Source -> Live
X-Axis Units -> Seconds (Linear)
Y-Axis Units -> V (Linear)
Active Reference -> channel 2
Tics -> deselected
Grids -> deselected
• Shorthand Notation -> deselected
Mode Menu
Mode -> Cross Properties
Waterfall -> deselected
Total Input Samples -> 65536
* It is necessary to open a compressed time trace window for one of the reference
channels under the "Window" Menu for proper data formatting during save operations.
76

APPENDIX A7. JOHN DEERE TRAILGATOR SPECIFICATIONS
Specifications.
4x2
Engine and electrical
Make KM
Type 4-cycle gas
Cylinders 1
Valving Overhead valves
Displacement 17.5 cu. in. (286 cc)
Horsepower (SAE J1940) 10
Maximum torque 14.5 Ib-ft @ 2,500 rpm
Ignition Capacitive discharge
Lubrication Full pressure
Oil filter Replaceable (optional)
Rpm, idle (no load) ,. 1,175
Rpm, fast (no load) 3,750
Cooling system Air.
Air cleaner Dry, replaceable, single element, with remote intake .
Muffler Spark arresting
Battery 32-amp/hr, 295 cold cranking amps
Alternator 13-amp regulated
Headlights Two 37.5-watt Halogen
Fuel system capacity 5.3 U.S. gal. (20 L)
Transmission
Type Continuously variable transmission (CVT)
Differential lock Standard: hand operated
Drive chain N/A
Ground speed 0-15.5 mph (0-25 km/h)
Transaxle Fully enclosed; oil bath ,
Gear selection Forward, neutral, reverse
Overall reduction ratio:
Low 73.8:1
High 16.5:1
Brakes Wet disk in transaxle ,
Steering Rack and pinion with Ackerman-type geometry
Ground pressure (max.)
With 200-lb. operator only 5.6 psi (.39/kg/cm2)
Fully loaded vehicle 7.5 psi (.53/kg/cm')
Dimensions
Length (without bumper) 100.5 in. (255.2 cm)
Length (with bumper) 101 .6 in. (258 cm)
Width 60.0 in. (152.5 cm)
Front-tread centers 50.0 in. (127 cm)
Rear-tread centers 48.0 in. (122 cm)
Height (overall) 43.6 in. (110.7 cm)
Wheelbase 69.5 in. (176.6 cm)
Weight (including fuel/fluids) 879 lb. (398.6 kg)
Ground clearance:
Under transaxle 6.7 in. (170 mm)
Under foot platform 8.5 in. (215 mm)
Seating capacity 2
Seat type Standard medium back ,
Turn clearance circle 22 ft. (6.7 m)
Towing capacity 900 lb. (408 kg) ,
Payload capacity' 900 lb. (408 kg)
Cargo Box
Material 16-gauge steel
Capacity:
Volume 10.1 cu. ft. (0.29 m3
)
Weight 500 lb. (227 kg)
6x4
KHI
4-cyclegas.
2
Overhead valves
37.7 cu. in. (617 cc).
18
32.5 Ib-ft @ 2.200 rpm
Transistor controlled
Full pressure
Replaceable (standard)
1.175
3.650
Liquid
Dry, replaceable, single element, with remote intake
Spark arresting
32-amp/hr, 295 cold cranking amps
20-amp regulated
Two 37.5-watt Halogen
5.3 U.S. gal. (20 L)
Continuously variable transmission (CVT)
.
Standard; hand operated
No. 50 industrial roller chain
0-15.5 mph (0-25 km/h)
Fully enclosed; oil bath
Forward, neutral, reverse
73.8:1
16.5:1
Wet disk in transaxle
Rack and pinion with Ackerman-type geometry
.
6.9 psi (,49/Vg/cm2
)
.
7.1 psi (.50/kg/cm?)
.
104.9 in. (266.5 cm)
106.0 in. (269.3 cm)
60.0 in. (152.5 cm)
50.0 in. (127 cm)
48.0 in. (122 cm)
43.6 in. (110.7 cm)
79.0 in. (200.6 cm) front/rear.
1.057 lb. (479.4 kg)
6.7 in. (170 mm)
8.5 in. (215 mm)
2
Professional high back
.
24.8 ft. (7.6 m)
1.2001b. (544 kg)
1.2001b. (544 kg)
16-gauge steel
11.2cu.ft. (0.32 m3
).
800 lb. (364 kg)
"Includes 200-pound operator. 200-pound passenger, and maximum box capacity (Specifications and design subject to change without notice.)
77

APPENDIX B. UNDERWATER ACTUATOR CABLES AND CONNECTORS
This appendix contains the specifications and wiring diagrams for the underwater
actuator cables and connectors that were used during this research.
Appendix B 1 - Belden 9775 cable specifications
Appendix B2 - Seacon rubber-molded connector specifications
Appendix B3 - Seacon bulkhead connector specifications
Appendix B4 - Underwater cable wiring diagram
79

APPENDIX Bl. ACTUATOR UNDERWATER CABLE SPECIFICATIONS
Paired
Individually Shielded
[Beldeii
Audio and Computer Cables
Description L
,
OL AWU Style '
c
1
ncc No. No :
NEC Tyoc of
SACcrt. Pairs
Standard
Length r
11 m
Stcl
Unit
1
Lbs
'
ea.
Noniin.-.I O C-R.
;ond Shield
Nominal
O.D. N
1
Inch mm '°
J . _
Nominal
»•" XT j
Cmmmuua
ap | i
• • j " "
lms>. Peon pF: fTl fp! " F!
1
1 « .
- 1
18 Gage
Stranded Conduclors (1 9x30)
Polyethylene Insulated
Product Description
Tinned copper, polyethylene insulated, twisted pairs. Each pair individually shielded with
Beldl oil* alum inum-potyester shield and 20 AWG stranded tinned copper drain wire. Overall
chrome PVC jacket. Color code chart No. 3, Technical Information Section.
2-FoW*
/
c
fcs=
•773
NEC CM
CSA
PCCFT1
3 100
soo
1000
30.5
152.4
304.8
9.4
48.2
94.3
6.40/M'
2in/km 27.2fWm
.404 10.26 50 66% 30 98 55 180
p-*g^j
Beictloil
1 00% Shield
Coverage
*12919
3OV80'C
See Attenuation,
Rise Time and
9774
NEC CM
CSA
pep ^t t
e 100
500
1000
30.5
152.4
3048
19.3
95.6
192.5
6.4Q/M'
21Q/km
6.3n/M'
27.2£l/fcm
.560 14.22 SO 66% 30 98 S5 180
f »773 NEC CM 1
CSA 1
(PCCFTy
9 100
500
10O0
30.5
152.4
304.8
25.7
131.8
256.6
6.4n/M"
21£i/km
8.30/M-
27.20/Vm
.655 16.64 SO 66% 30 98 55 180
Bit Rats data tor
this series on
pegs eg. NEC CM^
CSA
PCCFT1 '
.
1 2 100
500
1000
30.5
152.4
304.8
32.4
206.4
420.8
6.4Q/M'
2in/km 2720JVxn
.735 18.67 50 66% 30 98 55 180
•777
NEC CM
CSA
PCCFT1
Vs 100
500
1000
30.5
152.4
304.8
45.0
213.7
460.0
6.4fl/M'
21£}/km
6.30/M'
27.2«/km
.819 20.8 50 66% 30 98 55 180
Belden 9775
Cable used
102
1
Capacitance between conductors
'
Capacitance between one conductor and other conductors connected to shield.
80

APPENDIX B2. UNDERWATER RUBBER-MOLDED CONNECTORS
RMK-FS AND ASSOCIATED PARTS
STANDARD RUBBER MOLDED
IN-UNE SERIES
Note: Available with grippers, must specily.
|
Material: Special Neoprene, Molded
Contacts: Copper Alloy. Hard Gold Plated per MIL-G-45204
Voltage Rating: 600 VDC-contact factory for higher voltage requirements
Pressure Rating, Mated: up to 20,000 psi
RMK-9-FS/MP
Connectors
RMK-i/o-FS/MP (HK-l-FS/UP RMX.SVFSJVP PWK-2-fVMP RMK 3-FS/IJP RM(-4.F5/VP
i tn i *a in 1 no 1 f10 l »10
iC-Wr. 1-CW»ttngJ(CJ*)to;anSO> IM 1 !« 2#14
(C«B* WSO) [C»btelW2SOI (CaMKlMSO] (C*M«: IIMSO) tC^b*: 1 araSO) F*fc£«*M9 (Cafeta. W7SO) (Cabta 18/SSO, CjW«: 1&H2SO]
LOCKING SLEEVES DUMMXWJONNECTORS
Note: k-fls-X Is recommended when mating
RMK-FS to a bulkhead connector. Must specify cable OD.
Note: specify number ol conductors when ordering
K-MLS-D (AVAILABLE FOR RMK-MP)
K-FLS-D (AVAILABLE FOR RMK-FS. RMK-FS-R/A. WOC & BCL'S)
RMK-FSD MATES TO RMK-MP
RMK-MPD MATES TO RMK-FS & RMK-FS-R/A
Pressure Rating, Mated:
RMK-FSD (FOR HIGH OR LOW PRESSURE. NO CONTACT REQUIRED)
RMK-MPD-LP (LOW PRESSURE ^"4,000 PSI, NO CONTACT REQUIRED)
RMK-MPD-HP (HIGH PRESSURE > 4.000 PSI CONTACT IS REQUIRED)
WOC CONNECTORS-
Note: WOC stands for "without cable." The connectors are supplied as shown below for installation in the field
and/or to non-neoprene jacketed cables. Bonding can be achieved by molding to the connectors shown using
polyurethane with appropriate primers. Inexpensive plastic syringes can be used as molds.
Part Names: RMK-FS WOC, RMK-MP WOC
Contacts: Brass, Hard Gold Plated per MIL-G-45204
Voltage Rating: 600 VDC-contact factory for higher voltage requirements
-
m a-
U..J
50-28
81

APPENDIX B3. UNDERWATER BULKHEAD CONNECTORS
RMK-FS AND ASSOCIATED PARTS
MATEABLE BULKHEAD CONNECTORS
BULKHEAD CONNECTORS
(MATES TO RMK-FS & RMK-FS-R/A)
Part Name: RMK-BCL
Material: 316 Stainless St.; Internal Insulation: Special Neoprene
Contacts: Copper Alloy. Hard Gold Plated per MIL-G-45204
Pressure rating, Mated: up to 20,000 psi
Part Name: XSK-BCL
Material: Glass Reinforced Epoxy
Contacts: Copper Alloy, Hard Gold Plated per MIL-G-45204
VISUM. WOtX
I (M-UNC WUM COKTACT |l)
Description: Special insulating neoprene is
permanently molded to the metal body. These
types of connectors are used where physical
abuse would be too great for our glass reinforced
(GRE) epoxy series.
Description: A non-metallic connector
homogeneously molded of glass reinforced epoxy
(GRE). One of the oldest and most reliable
connectors in our product line. For the same
connector body, but with female sockets refer to
the VSK (page 50-31).
VISUAL
INDEX
RMK-1/0-8CL RMK-1-BCL RMK-KMCL
XSK-I/O-BCL XSK-I-SCL XSK-6/8-8CI.
1 110 1 «3 1 16
1 <8
RMK-2-3CL
xsK-2-ecc
1 f 14
1>10
Part Name: XSK-DEBCL
Material: Glass Reinforced Epoxy
Contacts: Copper Alloy, Hard Gold Plated per MIL-G-45204
Pressure Rating, Mated: up to 20,000 psi from front side.
0-«HC 0-131)
RMK-9-BCL
Connector
Used
<t>
50-29
82

APPENDIX B4. WIRING DIAGRAM FOR UNDERWATER ACTUATOR
CABLES AND CONNECTORS
O
b
w
O
O
rf ft
o <J
p
O
o
o
o
B
o
o
o
^ (Q <
ffi cS i£
o% ^ CN
£ U, ffi
OS n <N
Uh (E <£
0 ; vd
I'
4
en
vo c^ oo
U n, S
o C^ 00
k u u.
£rl
SU _, •->
c2 c£ c2
;-*>':. ;S"
x '
'
£
s
5#£-&is
V ':i '
''
1 „
:
s
io n i-
U, i_L, £.
vi n tf
i-L, cE U,
I 1
!
00
Q
83

APPENDIX C. EXPERIMENTAL SETUPS
This appendix contains the experimental setups for selected tests presented in this
thesis.
Appendix CI - Ambient Noise Analysis
Appendix C2 - Vertically Oriented Seismic Source Setup
Appendix C3 - Rayleigh Wave Seismic Source Setup
Appendix C4 - Frequency Optimization and Wavespeed Measurement Setup (Vertical)
Appendix C5 - Test setup for helium gas tank target tests
Appendix C6 - Test setup for helium gas tank and gunpowder keg target measurements
(increasing mass)
85

APPENDIX CI. AMBIENT NOISE ANALYSIS
This appendix contains the experimental setup for the ambient acoustical noise
analysis conducted on September 4, 1998.
Comments: Geophone gain = 40db
Sunny, low tide, 3-4ft waves.
Unfiltered signals were recorded as shown below
I
45'
Waterline
G2Q—
*
10'
6 Gl
86

APPENDIX C2. TEST SETUP FOR VERTICALLY ORIENTED SEISMIC SOURCES
This appendix contains the experimental setup for vertically oriented seismo-acoustic
sources on October 20, 1998.
Source #1
Orientation:
Voltage:
Frequency:
Cycles:
Vertical
10V (20Vpk-pk)
90Hz
1
Geophone #1
Filter: High Pass 40Hz
Gain: 40db
Source #2
Orientation:
Voltage:
Frequency:
Cycles:
Vertical
10V (20Vpk-pk)
90Hz
1
Geophone #2
Gain: 40db
Comments: Sunny, low tide, 3-4ft waves.
87

APPENDIX C3. RAYLEIGH WAVE SEISMIC SOURCE SETUP (PHASING)
This appendix contains the experimental setup for Rayleigh wave excitation on
October 2, 1998.
Source #1 Source #2
Orientation: Horizontal Orientation: Vertical
Voltage: 6V(12Vpk-pk) Voltage: 6V (12Vpk-pk)
Frequency: 100Hz Frequency: 100Hz
Cycles: 2 Cycles: 2
Phase: 0° Phase: -90°
Geophone #1 Geophone #2
Filter: High Pass 40Hz Filter: High Pass 40Hz
Gain: 40db Gain: 40db
Comments: Sunny, low-medium tide, 2-3ft waves.
88

APPENDIX C4. FREQUENCY OPTIMIZATION AND WAVESPEED SETUP
This appendix contains the experimental setup for determining the optimal drive
frequency and its associated wavespeed on October 20, 1998.
Source #1
Orientation: Vertical
Voltage: 10V (20Vpk-pk)
Frequency: 90Hz
Cycles: 1
Geophone #1
Gain: 40db
Source #2
Frequency: 90Hz
Cycles: 1
Geophone #2
Gain: 40db
Comments: Sunny, low tide, 3-4ft waves.
1
18"
r
1
60'
S2 g:
Waterline
-
o L' * A
V
r
•
A
oi k
C
G
>j *. *
SI
^ .
1
t
Not to Scale
89

APPENDIX C5. TEST SETUP FOR HELIUM GAS TANK TARGET TESTS
This appendix contains the experimental setup for target detection with increasing
mass tests conducted on October 23, 1998.
Source #1
Frequency: 80Hz
Cycles: 1
Geophone #1
Gain: 40db
Source #2
Frequency: 80Hz
Cycles: 1
Geophone #2
Gain: 40db
Comments: Overcast, med-high tide, 4-5 ft waves.
90

APPENDIX C6. TEST SETUP FOR HELIUM GAS TANK AND GUNPOWDER KEG
TARGET TESTS (INCREASING MASS)
This appendix contains the experimental setup for target detection with increasing
mass tests conducted on November 6, 1998 and November 10, 1998.
Source #1
Frequency: 80Hz
Cycles: 1
Geophone #1
Gain: 40db
Source #2
Frequency: 80Hz
Cycles: 1
Geophone #2
Gain: 40db
Comments: Overcast, med-high tide, 4-5ft waves.
SI
4
18"
+
S2
Helium Tank (Nov 6
th
)
thx
Gunpowder Keg (Nov 10
in
)
Gl
10'
G2
16'
91

APPENDIX D. COMPUTER PROGRAMS AND SUBROUTINES
This appendix contains the Matlab subroutines and programs that were used
during this research for signal processing and data analysis of the recorded beach data.
To obtain a copy of the subroutines contact either Professor Thomas G. Muir, Code
PH/Mt, or Professor Steven R. Baker, Code PH/Ba at the Naval Postgraduate School,
Monterey, California 93943-5000
Appendix Dl - Attenuate.
m
Appendix D2 - Avgsum.
Appendix D3 - Chanl.m
Appendix D4 - Chan8.m
Appendix D5 - Coherence.m
Appendix D6 - Convert.m
Appendix D7 - Correlate.
m
Appendix D8 - Displace.m
Appendix D9 - Filtr.m
Appendix D10 - Frequency.
m
Appendix Dl 1 - Hankel.m
Appendix D12 - Hilbrt.m
Appendix D 1 3 - Hodogram.m
Appendix D14 - Modification.
m
Appendix D15 - Patch8.m
Appendix D16 - Powersd.m
Appendix D17 - Rms8.m
Appendix D18 - Subhilbrt.m
Appendix D19 - Subtract.m
Appendix D20 - Velocity.
m
Appendix D2 1 - Wavespd.m
Appendix D22 - Xzplot.m
Appendix D23 - Theory_Displace.m
93

LIST OF REFERENCES
1
.
"Landmines - The World Federation of Public Health Associations",
[http://www.alpha.org/wfpha/landmines.html], Nov. 1998.
2. U.S. Navy, Chief of Naval Operations, "United States Naval Mine Warfare Plan" (3
rd
ed), Washington, D.C., 20350, 1996-1997.
3 . "Chapter VII - The Maritime Campaign",
[http://es.rice.edu/projects/Poli378/Gulf/gwtxt_ch7.html], Nov. 1998.
4. "A Moldovan Soldier Probes for Mines at Camp Lejeune",
[http://www.defenselink.mil/photos/Augl996/960818-A-8072J-003.html], Nov.
1998.
5. "Coastal Battlefield Reconnaissance and Analysis (COBRA)",
[http://www.om.nav.mil/sci_tech/ocean/jcrn/cobra.htrnl], Nov. 1998.
6. "Concept: An Active Sonar Using Seismo-Acoustic Interface Waves as the Excitation
Carrying Sonar Energy and Information", ONR Slide #Ga-95-23 1 IDES/NLIGA4.
7. Smith, E., Wilson, P., Bacon, F., Manning, J., Behrens, J., Muir, T., "Measurement
and Localization of Interface Wave Reflections from a Buried Target", Journal
Acoustical Society of America, 103 (5), Pt. 1, May 1998.
8. Gaghan, F.E., Discrete-Mode Source Development and Testing For a New Seismo-
Acoustic Sonar, Master's Thesis, Naval Postgraduate School, Monterey, CA, Mar.
1998.
9. Kolsky, H., Stress Waves in Solids, pp. 1-25, Dover Publications, Inc., 1963.
10. Rayleigh, Lord (J.W. Strutt), "On Waves Propagated along the Plane Surface of an
Elastic Solid," Proceedings London Mathematical Society, 1885, vol. 17, no. 4.
11. Graff, F.G., Wave Motion in Elastic Solids, pp.273-380, Dover Publications, Inc.,
1975.
12. Richart, F.E., Hall, J.R., Woods, R.D., Vibration of Soils and Foundations, pp. 88-92,
Prentice-Hall, Inc., 1970.
13. Slide on Scholte Wave Characteristics from D. Rauch, SACLANT ASW Research
Center, La Spezia, Italy.
14. Operator's Manual - Electromagnetic Linear Actuator, Aura Systems Inc., El
Segundo, CA.
95

15. Printed with Permission, Aura Systems Inc., El Segundo, CA.
16. Personal communication between D. True, Naval Facilities Engineering Service
Center, Naval Construction Battalion Center in Port Hueneme, CA and S.M.
Fitzpatrick, Naval Postgraduate School, Jul. 1998.
17. Applied Research Laboratories, University of Texas at Austin, Characterization of
Marine Sedimentsfor Seismic Surveillance and Minehunting, by Lemair, C, Sep.
1997.
18. Matlab Version 5.2.0.3084 (Signal Processing Toolbox), Jan. 17, 1998, The
Mathworks Inc., Natick, MA.
19. ProSeries Model SPS390 Dynamic Signal Analyzer Installation and Operation
Manual (Ver. 3.8/Rev. C), Signal Processing Systems, San Diego, CA, 1998.
20. BBN Systems and Technologies Final Report 7677, Feasibility ofAcoustic Landmine
Detection, May 1992.
96

INITIAL DISTRIBUTION LIST
1
.
Defense Technical Information Center 2
8725 John J. Kingman Rd:, STE 0944
Ft. Belvoir, Virginia 22060-6218
2. Dudley Knox Library 2
411 DyerRd.
3. Dr. Jeffrey Simmen 1
Office of Naval Research, Code 3210A
800 N. Quincy St.
Arlington, Virginia 22217-5660
4. Dr. Douglas Todoroff 1
Office of Naval Research, Code 321
800 N. Quincy St.
Arlington, Virginia 22217-5660
5. Professor Thomas G. Muir, Code PH/Mt 2
Department of Physics
6. Professor Steven R. Baker, Code PH/Ba ...2
Department of Physics
7. Professor Anthony J. Healey, Code ME/Hy 2
Department of Mechanical Engineering
8. Dr. Raymond Lim 1
Coastal Systems Station, Code R22
6703 W. Highway 98
Panama City, Florida 32407-7001
9. Library, Coastal Systems Station 1
6703 W. Highway 98
Panama City, Florida 32407-7001
97

10. Dr. Richard Gramann
Applied Research Laboratories
University of Texas at Austin
P.O. Box 8029
Austin, Texas 78713
11. LT SeanM. Fitzpatrick
12362 Caminito Festivo
San Diego, California 92131
12. Bill Taylor, Senior Scientist
Automotive/Industrial Division
Aura Systems, Inc.
2335 Alaska Avenue
El Segundo, California 90245
98

18 « 355e#
10/99 72527-1DU » i

NPS Thesis - Fitzpatrick

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to NPS Thesis - Fitzpatrick (20)

NPS Thesis - Fitzpatrick