Application of Piezoelectric Sensors in Soil Property Determination

Embed Size (px)

Citation preview

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    1/182

    APPLICATION OF PIEZOELECTRIC SENSORS

    IN SOIL PROPERTY DETERMINATION

    by

    LEI FU

    Submitted in partial fulfillment of the requirements

    for the degree of Doctor of Philosophy

    Dissertation Advisor:

    Prof. XIANGWU ZENG

    Department of Civil Engineering

    CASE WESTERN RESERVE UNIVERSITY

    August, 2004

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    2/182

    CASE WESTERN RESERVE UNIVERSITY

    SCHOOL OF GRADUATE STUDIES

    We hereby approve the dissertation of

    ______________________________________________________

    candidate for the Ph.D. degree *.

    (signed)_______________________________________________

    (chair of the committee)

    ________________________________________________

    ________________________________________________

    ________________________________________________

    ________________________________________________

    ________________________________________________

    (date) _______________________

    *We also certify that written approval has been obtained for any

    proprietary material contained therein.

    Lei Fu

    David Gurarie

    Adel S. Saada

    J. Ludwig Figueroa

    Robert Mullen

    07/09/2004

    Xiangwu Zeng

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    3/182

    DEDICATION

    I love you all

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    4/182

    i

    TABLE OF CONTENTS

    TABLE OF CONTENTS iLIST OF TABLES iv

    LIST OF FIGURES vi

    ACKNOWLEDGEMENTS xi

    ABSTRACT xii

    CHAPTER 1 INTRODUCTION 1

    1.1 Introduction 11.2 Laboratory Measurement of Soil Properties 1

    1.2.1 Introduction 11.2.2 Element Tests 31.2.3 Model Tests 51.2.4 Centrifuge Tests 7

    1.2.4.1 Introduction 71.2.4.2 Scaling Laws 91.2.4.3 Limitations of Centrifuge Modeling 121.2.4.4 CWRU Centrifuge 16

    1.3 Field Determination of Soil Properties 221.3.1 Introduction 221.3.2 Seismic Methods 23

    1.3.2.1 Crosshole Method 241.3.2.2 Downhole Method 261.3.2.3 Shear Wave Refraction Method 271.3.2.4 SASW Method 28

    1.3.3 Cone Penetration Tests (CPT) 291.3.4 Standard Penetration Tests (SPT) 30

    1.4 Piezoelectric Sensors and Their Applications 311.4.1 Introduction 31

    1.4.2 Piezo Motors 331.4.3 Piezo Generators 341.4.4 2-Layer Elements 34

    1.4.4.1 Bender Elements 341.4.4.2 Extender Elements 37

    1.1.1 Series and Parallel Operation 371.1.2 Application of Piezoelectric Sensors in Soil Property

    Determination 39

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    5/182

    ii

    CHAPTER 2 LITERATURE REVIEW 41

    2.1 Review of Piezoelectric Sensors in Soil Property Measurement 412.2 Review of Soil Property Measurement in Centrifuge tests 442.3 Research Objectives and Outline 52

    2.3.1 Research Objectives 522.3.2 Outline 53

    CHAPTER 3 BENDDER ELEMENTS IN CENTRIFUGE MODEL TESTS

    ---DRY SPECIMEN TESTS 54

    3.1 Introduction 54

    3.2 Bender Elements 55

    3.3 Nevada Sand 56

    3.4 Experimental Setup 583.5 Description of the Model 61

    3.6 Model preparation 633.7 Test Procedures 66

    3.8 Experimental Results and Analysis 66

    3.9 Conclusions 79

    CHAPTER 4 BENDER ELEMENTS IN CENTRIFTGUE MODEL TESTS

    ---SATURATED SPECIMEN TESTS 80

    4.1 Introduction 80

    4.1.1 Evaluation Liquefaction Potential of Soils 804.1.2 Content of This Chapter 86

    4.2 Test Equipment 864.3 Input Earthquake Motion 864.4 Transducer Description 87

    4.4.1 Water-proof Bender Elements 874.4.2 Accelerometers 88

    4.4.3 Pore Pressure Transducers 884.5 Preparation of Saturated Models 90

    4.5.1 Description of the Models 90

    4.5.2 Preparation of the Models 92

    4.5.3 Pore Fluid 954.5.4 Saturation of specimen 97

    4.6 Centrifuge Testing Procedure 98

    4.7 Test Results 101

    4.7.1 Shear Wave Velocities 101

    4.7.2 Pore Pressures 1064.7.3 Accelerations 106

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    6/182

    iii

    4.7.4 Examination of Liquefaction Criteria Based on Shear

    Wave Velocities 1074.8 Conclusions 116

    CHAPTER 5 PIEZO CONE PENETROMETER 118

    5.1 Introduction 118

    5.2 Piezo Cone Penetrometer Structure 119

    5.3 Experimental setup 123

    5.4 Principle of the Tests 1255.5 Field Test Procedure 126

    5.6 Typical Laboratory Test Results 127

    5.6.1 Soil Description 127

    5.6.2 Test Procedure 128

    5.6.3 Test Results 1285.7 Conclusions 135

    CHAPTER 6 ODOMETER FOR GRAVELLY MATERIAL STIFFNESS

    MEASUREMENTS 136

    6.1 Introduction 136

    6.2 Equipment Description 140

    6.3 Experimental Setup 143

    6.4 Typical Test Results 1446.4.1 Soil Description 144

    6.4.2 Test Procedures 1456.4.3 Test Results 145

    6.5 Conclusions 148

    CHAPTER 7 SUMMARY OF CONCLUSIONS AND SUGGESTIONSFOR FUTURE STUDY 149

    7.1 Introduction 1497.2 Summary of Observations and Conclusions 151

    7.2.1 Application of Bender Elements in Centrifuge Tests 151

    7.2.2 Piezo Cone Penetration for Pavement Field Tests 153

    7.2.3 Odometer for Gravelly Material Stiffness Measurements 1547.3 Suggestions for Future Study 155

    REFERENCES 159

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    7/182

    iv

    LIST OF TABLES

    Table 1.1 Relative Quality of the Laboratory Technique for Measurement

    of Dynamic Soil Properties (after Silver, 1981) 4

    Table 1.2 Scaling Factors for Centrifuge Tests (after Taylor, 1995) 12

    Table 1.3 Comparison of Methods for Centrifuge Earthquake MotionSimulation (after Whitman, 1988) 18

    Table 3.1 Bender Element Performance (Source: Piezo Systems, Inc.) 56

    Table 3.2 Index Properties of Nevada Sand (after Arulmoli, 1994) 57

    Table 3.3 Locations of the Bender Elements in Models 62

    Table 3.4 Shear Save Velocities Measured during Spin-up of the Centrifuge

    (Dr = 30%) 68

    Table 3.5 Shear Wave Velocities Measured during Spin-up of the Centrifuge

    (Dr = 48%) 69

    Table 4.1 Specifications of Accelerometer 89

    Table 4.2 Calibrations of Accelerometers 89

    Table 4.3 Specifications of Pore Pressure Transducers 90

    Table 4.4 Calibrations of Pore Pressures Transducers 90

    Table 4.5 Locations of Sensors 92

    Table 4.6 Physical Properties of METHOCEL(Source: Dow Chemical Company) 96

    Table 4.7 Shear Wave Traveling Time 102

    Table 4.8 Shear Wave Velocities (Spin-up) 102

    Table 4.9 Shear Wave Velocities (after the First Earthquake) 103

    Table 4.10 Shear Wave Velocities (after the Second Earthquake) 103

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    8/182

    v

    Table 5.1 Types of the Elements 120

    Table 5.2 Index Properties of Two Soils 127

    Table 5.3 Test Results on Nevada Sand 130

    Table 5.4 Test Result on Delaware Clay 130

    Table 5.5 Summary of the Results of CBR Tests 134

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    9/182

    vi

    LIST OF FIGURES

    Figure 1.1 Range and Applicability of Dynamic Laboratory Tests(after Das, 1993) 4

    Figure 1.2 Sketch of the Mechanics Relating to the Two Types of Centrifuges 6

    Figure 1.3 Takenaka Corporation Centrifuge (Source: Takenaka Corporation) 8

    Figure 1.4 Stresses in Model and Prototype 10

    Figure 1.5 Stress Variation with Depth in a Centrifuge Model

    (after Taylor, 1995) 14

    Figure 1.6 CWRU Laminar Box (after Dief, 2000) 21

    Figure 1.7 Crosshole Method 25

    Figure 1.8 Downhole Method 26

    Figure 1.9 Shear Wave Refraction Method 27

    Figure 1.10 Spectral-Analysis-of-Surface-Waves Method 29

    Figure 1.11 Sketch of a Cone Penetrometer 30

    Figure 1.12 SPT Test (after Kovacs et al., 1981) 31

    Figure 1.13 Peizoceremic Material 32

    Figure 1.14 Sketch of a Motor 33

    Figure 1.15 Sketch of a Generator 34

    Figure 1.16 Structure of a 2-Layer Piezo Element 35

    Figure 1.17 Bender Element Transmitter 36

    Figure 1.18 Bender Element Generator 36

    Figure 1.19 A 2-Layer Bender Element Poled for Series Operation 38

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    10/182

    vii

    Figure 1.20 A 2-Layer Bender Element Poled for Parallel Operation 38

    Figure 1.21 Setup for Incorporating Bender Elements in the Resonant

    Column Apparatus (after Dyvik, 1985) 40

    Figure 2.1 Experimental System (after Shibata et al., 1991) 48

    Figure 2.2 Layout of Bender Source and Receiver (after Gohal and Finn, 1991) 49

    Figure 2.3 In-Flight Shear Wave Measurement (after Arulnathan et al., 2000) 51

    Figure 3.1 Dimension of the Bender Element 55

    Figure 3.2 Grain Size Distribution of Nevada Sand 57

    Figure 3.3 Experimental Setup 60

    Figure 3.4 Test Equipment in Laminar Box 60

    Figure 3.5 Bender Elements in Soil Model 61

    Figure 3.6 Traveling Pluviation Apparatus 65

    Figure 3.7 Typical Recorded Signal 67

    Figure 3.8 Effect of Relative Density on Shear Wave Velocity Measuredduring Spin-up of the Centrifuge 70

    Figure 3.9 Repeatability of Shear Modulus Determined during Spin-up of

    Two Different Centrifuge Models 71

    Figure 3.10 Change of Shear Save Velocity with Acceleration during Spin-up

    of the Centrifuge 71

    Figure 3.11 Comparison between Test Results and Results from

    Empirical Equation 74

    Figure 3.12 Shear Modulus during Spin-up and Spin-down of the Centrifuge 75

    Figure 3.13 Comparison of Gmax by Resonant Column Tests and Bender

    Element Tests during Spin-up of the Centrifuge

    (Dr= 30%) 77

    Figure 3.14 Time History of Base Acceleration Scaled to Prototype Values 77

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    11/182

    viii

    Figure 3.15 Effect of Earthquake Shaking on Shear Modulus Measured

    during Spin-down of the Centrifuge 78

    Figure 4.1 Relationship between Cyclic Stress Ratios Causing Liquefaction

    and (N1)60 Values (after Seed et al., 1975) 84

    Figure 4.2 Liquefaction Relationship Recommended for Clean, Uncemented

    Soils (after Andrus and Stoloe, 2000) 85

    Figure 4.3 Prototype Acceleration of Base Input Motion 87

    Figure 4.4 Locations of Sensors in Centrifuge Model 91

    Figure 4.5 Schematic of the Chamber with an Air Tight Lid (after Dief, 2000) 94

    Figure 4.6 Effect of Pore Fluid Viscosity on Pore Pressures in Centrifuge

    Model Tests (after Dief, 2000) 96

    Figure 4.7 Model Saturation System (after Dief, 2000) 98

    Figure 4.8 Shear Wave Velocity Vs Acceleration

    (Spin-up, before Earthquake) 104

    Figure 4.9 Variations of Shear Wave Velocity with Depth in the Model(Spin-up, before Earthquake) 104

    Figure 4.10 Shear Wave Velocities before and after the First Earthquake 105

    Figure 4.11 Comparison of Shear Wave Velocities after the First and

    Second Earthquakes 105

    Figure 4.12 Pore Pressure History of the Top Point during the

    First Earthquake 109

    Figure 4.13 Pore Pressure History of the Middle Point during the

    First Earthquake 109

    Figure 4.14 Pore Pressure History of the Bottom Point during theFirst Earthquake 110

    Figure 4.15 Pore Pressure History of the Top Point during the

    Repeated Earthquake 110

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    12/182

    ix

    Figure 4.16 Pore Pressure History of the Middle Point during the

    Repeated Earthquake 111

    Figure 4.17 Pore Pressure History of the Bottom Point during the

    Repeated Earthquake 111

    Figure 4.18 Acceleration History of the Top Point during the First Earthquake 112

    Figure 4.19 Acceleration History of the Middle Point during the

    First Earthquake 112

    Figure 4.20 Acceleration History of the Bottom Point during the

    First Earthquake 113

    Figure 4.21 Acceleration History of the Top Point during theRepeated Earthquake 113

    Figure 4.22 Acceleration History of the Middle Point during the

    Repeated Earthquake 114

    Figure 4.23 Acceleration History of the Bottom Point during the

    Repeated Earthquake 114

    Figure 4.24 Case history of Liquefaction with Centrifuge Model Test DataAdded Based on Shear Wave Velocity 115

    Figure 4.25 CRR Values and Overburden Stress-Corrected Shear WaveVelocities before and after the First Earthquake 116

    Figure 5.1 Detailed Schematic of the Piezo Penetrometer 121

    Figure 5.2 Photo of the Piezo Cone Pentrometer 122

    Figure 5.3 Experimental Setup for Laboratory Tests 123

    Figure 5.4 Experimental Setup for Field Tests 124

    Figure 5.5 Experimental Principal 125

    Figure 5.6 Typical Bender Element Test Results 129

    Figure 5.7 Typical Extender Element Test Result 129

    Figure 5.8 Test Results on Nevada Sand (Dry Density=1600 kg/m3) 131

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    13/182

    x

    Figure 5.9 Test Results on Delaware Clay (Dry Density = 1590 kg/m3) 131

    Figure 5.10 CBR Test Results on Delaware Clay (Dry Density =1557 kg/m3) 133

    Figure 5.11 CBR Test Results on Nevada Sand (Dry Density =1610 kg/m3) 134

    Figure 6.1 Shear Modulus in Different Plane (after Zeng and Ni, 1999) 137

    Figure 6.2 Relationship between Shear Moduli in Different Shear Planes:

    (a) Theory; (b) Experimental Data (Vertical Stress Is 60 kPafor Inner Ellipse and Is Increased by 60 kPa for each Ellipse,

    Pool Filter Sand) (after Zeng and Ni, 1999) 139

    Figure 6.3 Detailed Schematic of the Odometer 141

    Figure 6.4 Photo of the Odometer 142

    Figure 6.5 Loading and Displacement Measurement System 143

    Figure 6.6 Particle Size Distribution Curves of the Two Soils 144

    Figure 6.7 Shear Moduli during One Cycle of Loading (Soil with 25%

    Gravel, Cycle No. 1) (by Zeng, Wolfe, and Fu) 146

    Figure 6.8 Settlement of the Soil Sample under Repeated Loading

    (Soil with 50% Gravel) (by Zeng, Wolfe, and Fu) 147

    Figure 6.9 Comparison of Test Results of Two Soils

    (by Zeng, Wolfe, and Fu) 147

    Figure 7.1 An Automatic Piezo Cone Penetrometer System 157

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    14/182

    xi

    ACKNOWLEDGEMENTS

    The author would like to express his sincere and heartfelt appreciation to

    his advisor, Professor Xiangwu Zeng, for his valuable guidance and

    encouragement in academic and personal life. It was his idea which originated

    this research.

    The author is thankful to Professor Adel S. Saada, Professor J. Ludwig

    Figueroa, and Professor Robert L. Mullen for serving on the graduate committee

    and for their invaluable instructions.

    The author is also thankful to Professor David Gurarie for being a member

    of the graduate committee.

    The author is indebted to Judy Wang for her proof reading of this

    dissertation, Gang Liu for his help in centrifuge tests, and Yunyi Zou for his

    friendly help in many aspects.

    Many thanks are extended to Kathleen Ballou, Sheila Campbell, and

    Bernie Strong for their support.

    Finally, the study reported here was founded under NSF Grant No.

    01960166. The author is deeply grateful to Dr. Clifford Astill (NSF Program

    Manager) for his support.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    15/182

    xii

    APPLICATION OF PIEZOELECTRIC SENSORS

    IN SOIL PROPERTY DETERMINATION

    Abstract

    by

    LEI FU

    Piezoceramic as a smart or adaptive material has been widely used in

    engineering measurement and control. When used in soil testing, the material can

    work as a wave transmitter or a wave receiver. Among the piezoelectric sensors,

    the bender element and the extender element are the most widely used. While the

    bender element is used to generate and receive shear waves, the extender element

    is used to produce and detect P-waves.

    One application of piezoelectric sensors is to measure shear wave

    velocities in centrifuge tests. Measurement techniques for both dry and saturated

    specimens were developed. A series of centrifuge model tests were performed.

    Bender element tests were carried out during the spin-up, spin-down, and in-

    flight stages of a centrifuge. Dynamic centrifuge tests were conducted on both

    dry and saturated models to investigate the influence of earthquake motions.

    Shear wave traveling times, accelerations, and pore water pressures were

    monitored during the tests. Based on the test results presented in this study, the

    current shear wave velocity-based liquefaction evaluation criteria were examined.

    Piezoelectric sensors were also used to develop a new piezo cone

    penetrometer. The piezo cone penetrometer is equipped with one set of bender

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    16/182

    xiii

    elements and one set of extender elements. The equipment is designed for the

    purposes of measuring stiffness of the subgrade and the sublayer of a pavement

    in the field. Compared to the conventional CBR test, this method is quick and

    simple as well as being theoretically sound. The results in the laboratory have

    shown it to be a promising tool.

    Finally, a large odometer was developed for testing gravelly materials.

    The advantages of the device include that it can measure the stiffness of a soil

    with large grain sizes and can measure moduli in different planes.

    In conclusion, measurement of soil properties using piezoelectric sensors

    is direct and accurate. The technique is theoretically sound. The application of

    the technique is limited to low strain levels, which are in the elastic range of soil

    deformations.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    17/182

    1

    CHAPTER ONE

    INTRODUCTION

    1.1 Introduction

    Geotechnical tests are used to investigate soil properties, provide

    parameters for analyses, and evaluate the performance of a particular prototype.

    Measurements of stiffness of soils are an important part of soil tests.Generally,

    the test methods fall into two categories, field testing and laboratory testing. In

    laboratory tests, either a specimen is tested to represent a point with a given initial

    stress state in the field, or a model is tested to study the behavior of a prototype

    under specific working condition. The ability of laboratory tests to provide

    accurate measurements of soil properties depends on their ability to replicate the

    initial stress conditions and loading conditions of the problems investigated. In

    field tests, all the measurements are carried out in the existing state of the soil.

    The influences of stress history, structural conditions of the soil body, thermal

    conditions, and chemical conditions are reflected in the test results.

    1.2 Laboratory Tests

    1.2.1 Introduction

    Currently lab tests provide material properties that are difficult to

    measure by means of in-situ tests, such as high-strain modulus and high-strain

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    18/182

    2

    damping. Nagarai (1993) cites the advantages of determining the mechanical

    properties of soils in the laboratory, including:

    (1)Total control over test conditions, including boundary conditions can be

    attained;

    (2)Control can be exercised over the choice of material to be tested;

    (3)Tests can be carried out under simulated field conditions or can be very

    different from the in-situ conditions, e.g., environmental factors such as

    temperature and humidity can be simulated;

    (4)Laboratory testing permits the examination of a great variety of other

    conditions which may be relevant to further changes in environmental and

    other stress conditions;

    (5)Tests can be carried out on reconstituted (destructured or remolded) and

    processed materials under different stress conditions;

    (6) Laboratory test data enable the understanding or the basic mechanisms of

    material behavior.

    There are two types of laboratory tests, element tests and model tests.

    Element tests are usually performed on relatively small specimens that are

    assumed to be representative of elements which are subjected to uniform initial

    stresses and undergo uniform changes in stress and strain conditions. Model tests

    may be used to evaluate the performance of a particular prototype or verify

    predictive theories.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    19/182

    3

    1.2.2 Element Tests

    Static and dynamic tests are performed to measure soil properties with

    respect to different controlled conditions. For measuring the of stiffness of soils, it

    is argued (Woods,1994), however, that we should no longer distinguish between

    dynamic and static properties as they are indeed a continuum, and we should,

    rather, distinguish properties on the basis of strain level. Soil stiffness

    measurements methods include seismic wave tests (ultrasonic pulse tests), cyclic

    triaxial tests, resonant column tests, and cyclic simple shear tests, etc. Saada and

    Townsend (1981) pointed out that the thin long hollow cylinder apparatus

    complemented by the true triaxial apparatus with rigid boundaries, driven by the

    proper controls, can simulate most of the conditions met in the field. The

    magnitudes of modulus and damping are functions of the shear strain amplitude.

    Figure 1.1 shows the amplitude of shear strain levels, types of applicable dynamic

    tests, and the areas of applicability of the test results. Table 1.1 gives a summary

    of the parameters measured in dynamic laboratory tests.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    20/182

    4

    Figure 1.1 Range and Applicability of Dynamic Laboratory Tests (after Das,

    1993)

    Table 1.1 Relative Quality of the Laboratory Technique for Measurement of

    Dynamic Soil Properties (after Silver, 1981)

    Relative Quality of Test Results

    Shear

    Modulus

    Youngs

    Modulus

    Material

    Damping

    Effects of number

    of Cycles

    Attenuations

    Resonant Column Good Good Good Good

    With adaptation Fair

    Ultrasonic pulse Fair Fair Poor

    Cyclic triaxial Good Good Good

    Cyclic simple shear Good Good Good

    Cyclic torsional shear Good Good Good

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    21/182

    5

    1.2.3 Model Tests

    In other laboratory tests, specimens are tested as models. A model is a

    reduced scale simulation of the prototype. Model tests may be used to evaluate the

    performance of a particular prototype or to study the effects of different

    parameters on a general problem. Model tests usually attempt to reproduce the

    boundary conditions of a particular problem by subjecting a small-scale physical

    model of a full-scale prototype structure. Commonly used model tests include the

    centrifuge test and the shake table test.

    The Geotechnical centrifuge test benefits from the additional centripetal

    force acting on a model while the centrifuge is rotating. As it increases the self-

    weight of the soil and thus creates a stress distribution in the soil sample that is

    comparable to prototype. It is well known that soil behavior is a function of stress

    level and stress history. It is this reason that centrifuge modeling of major use to

    the geotechnical engineering. In principle, the stress conditions at any point in a

    model should be identical to those at the corresponding point in a prototype. Then,

    it is assumed that the overall behavior of the model, such as displacements and

    failure, should also be identical to the prototype.

    There are two types of centrifuges: the beam centrifuge with a swinging

    basket and the drum centrifuge with a fixed channel (Figure 1.2).

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    22/182

    6

    Figure 1.2 Sketch of the Mechanics Relating to the Two Types of Centrifuges

    A shake table test models the response of field structures during an

    earthquake in 1 g environment. Shake tables of many sizes have been used in civil

    engineering research. Some are quite large, allowing models with dimensions of

    several meters to be tested. In geotechnical earthquake engineering, the shake

    table test has been used with great success in modeling the dynamic effect of

    structures. But, in the shake table model, with densities kept constant while linear

    dimensions reduced, the same the stress field as that of the prototype can not be

    produced. For this reason, the shaking table test is less successful for soil

    mechanics problems.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    23/182

    7

    Model tests share certain drawbacks, among the most important of which

    are similitude and boundary effects. Boundary effects are usually associated with

    the metallic boxes in which shake table and centrifuge models are usually

    constructed. The side walls can restrain soil movement and reflect energy that

    would radiate away in the prototype problem. The industrial filler material

    Duxseal has been used as an absorbent wall lining with some success (Steedman,

    1991).

    1.2.4 Centrifuge Tests

    1.2.4.1 Introduction

    The centrifuge test is one type of model test that is used to study

    prototype problems. By spinning the soil package and testing the model at a high

    speed, an artificial gravity is induced. The increase in gravity allows the stress,

    strain, and strength to be modeled in a scaled soil model. The conclusions

    regarding the prototypes behavior can be made by observing the model in similar

    circumstances. In order to replicate the gravity-induced stresses of a prototype in

    a 1/n reduced model, it is necessary to test the model in a gravitational field n

    times larger than that of prototype. This idea was applied for the first time in

    1930s, in the field of geotechnical engineering by Bucky (1931) and Pokrovsky

    (1932). Since then, a number of geotechnical centrifuges have been developed all

    over the world as an important research tool. The centrifuge has proven valuable

    both in the soil property investigation and the modeling of prototype.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    24/182

    8

    The main features of a centrifuge include payload capacity, arm radius,

    maximum acceleration, and payload size. Figure 1.3 shows one of the most

    powerful centrifuges in the world, the Takenaka centrifuge. It has the following

    specifications:

    Centrifuge radius: the platform radius is 7.0m;

    Usable payload dimensions: 2.0 m 2.0 m 1.1 m (width depth

    height);

    Performance: maximum payload is 5000 kg; acceleration at

    maximum payload is 100 g; payload at maximum acceleration is

    2000 kg; and acceleration range is 10 g and 200 g.

    Electrical slip-rings: 214 electrical slip-rings.

    Figure 1.3 Takenaka Corporation Centrifuge (Source: Takenaka

    Corporation)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    25/182

    9

    Possible geotechnical centrifuge studies include: 1) modeling of a

    prototype. A model being constructed to be geometrically similar to the prototype

    using the same material, the prototype behavior can be simulated in the centrifuge;

    2) investigation of new phenomena; 3) parametric studies; 4) validation of

    numerical models (Takemura et al., 1998).

    Two key issues in centrifuge tests are scaling laws and scaling errors.

    Scaling laws can be used by dimensional analysis. Centrifuge modeling is often

    criticized as having some scaling errors due to the non-uniform acceleration field

    and the difficulty of representing sufficient detail of the prototype in a small-scale

    model. In physical modeling studies, it is seldom possible to replicate precisely all

    the details of the prototype and some approximations have to be made.

    1.2.4.2 Scaling laws

    The basic scaling law derives from the need to ensure stress similarity

    between the model and the corresponding prototype. The following scaling law

    analyses were given by Taylor (1995).

    By rotating a package in the centrifuge at high speeds, an acceleration

    normal to the package is achieved (see Figure 1.4). The relationship is given by:

    a=Re2=Ng (1.1)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    26/182

    10

    Model

    Prototype

    m=p

    p

    where, a = centrifugal acceleration

    Re = effective length of the centrifuge arm

    = angular velocity

    N = scaling factor

    g = earths gravity (9.81 m/sec2).

    Figure 1.4 Stresses in Model and Prototype (after Taylor, 1995)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    27/182

    11

    Then the vertical stress, vm at depth hm in the model is:

    vm

    = Nghm

    (1.2)

    where, = mass density of the soil

    hm = depth of the point below soil surface in the model.

    In the corresponding prototype, the stress p is

    p = ghp (1.3)

    where, hm = Depth of the point below soil surface in the prototype

    Thus for

    m = p

    or

    ghp = Nghm

    then

    hm =N

    1 hp

    The scaling factor for linear dimensions isN

    1. Since the model is a linear scale

    representation of the prototype, then the displacement will also have a scaling

    factor ofN

    1. It follows therefore that the strain has a scaling factor of

    1

    1, and the

    stress-strain curve mobilized in the model will be identical to the prototype. More

    scaling factors of parameters are listed in Table 1.2.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    28/182

    12

    Table 1.2 Scaling Factors for Centrifuge Tests (after, Taylor, 1995)

    Parameter Model Prototype

    Acceleration N 1

    Density 1 1

    Stress 1 1

    Strain 1 1

    Velocity 1 1

    Length 1 N

    Area 1 N2

    Volume 1 N3

    Force 1 N2

    Energy 1 N

    3

    Time (static) 1 1

    Time (dynamic) 1/N 1

    Time (dissipation) 1/N2

    1

    1.2.4.3 Limitations of Centrifuge Modeling

    In physical modeling studies, it is seldom possible to replicate precisely all

    of the details of the prototype, and some approximations have to be made. The

    limitations involved in the modeling are: 1) errors in the centrifugal acceleration

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    29/182

    13

    field; 2) grain size effects; 3) difficulty in modeling the actual site, i.e., aging,

    sophisticated subsoil conditions; 4) inconsistency of scaling factors with time

    ((Takemura et al., 1998).

    (1) Particle Size Effects

    In the centrifuge test, the dimensions of a prototype is scaled down by a

    factor of N, but generally, the soil particles can not be scaled down at the same

    scale. This will produce grain size effects. An index used to show the effects is

    the ratio of the representative length of the model (B) to the average grain

    diameter (D50). For circular foundations, the critical ratio is 15 (Oveson, 1979).

    Some researcher suggested that it is more appropriate to examine the particle size

    effect by considering the ratio of particle size to shear band width (Tatsuoka et al.,

    1991)

    (2) Acceleration Field Scaling Errors

    1) Variation in Vertical Direction

    The Earths gravity is uniform for soil deposits in prototypes. When using

    a centrifuge to generate the high acceleration field, there is a slight variation in

    acceleration through the model (Taylor, 1995). Figure 1.5 shows there is exact

    correspondence in stress between model and prototype two-thirds of the model

    depth, and the maximum under-stress and the maximum over-stress exist at one-

    third of the height of the model and at the bottom of the model, respectively. For

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    30/182

    14

    most centrifuge models, hm/Re is less than 0.2, and therefore the maximum error

    in the stress profile is generally less than 3% of the prototype stress.

    Figure 1.5 Stress Variation with Depth in a Centrifuge Model (after Taylor,

    1995)

    2) Lateral Acceleration Component

    In the centrifuge model, the acceleration is directed towards the center of

    rotation and hence in the horizontal plane. There is a change in its direction

    relative to the vertical across the width of the model. There is, therefore, a lateral

    component of acceleration. To minimize this effect, it is good practice to ensure

    that the major events occur in the central region of the model where the error due

    to the radial nature of the acceleration field is small.

    Maximum Over-stress

    Depth

    h

    2h/3

    h/3

    Prototype

    Centrifuge Model

    Maximum Under-stress

    Stress

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    31/182

    15

    3) Coriolis Acceleration

    Coriolis acceleration occurs when changing the reference system of

    observation. The Coriolis effect may be observed in a centrifuge model, such as in

    the modeling of a explosion, when the movement of the mass is in the plane of

    rotation.

    The Coriolis acceleration, ac can be expressed as:

    ac = 2V (1.4)

    where, = centrifuge angular velocity

    V = velocity of moving mass.

    The centrifugal acceleration is:

    A = v (1.5)

    where, v = velocity of centrifuge.

    It is assumed that for low velocity of moving mass, Coriolis effects become

    neglectable if ac/a < 10%. This implies V2v (Taylor, 1995).

    (3) Boundary Conditions

    In a centrifuge test, a model is built inside a container. For dynamic tests,

    the excitation has to be transmitted through this container. The model container

    imposes artificial boundaries and leads to potential boundary effects. Three major

    boundary effects may be caused by a model container, namely the effects on the

    stress field, on the strain field, and on the seismic waves (Taylor, 1995). The basic

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    32/182

    16

    requirements for the containers of dynamic tests are that the end walls function as

    shear beams with the same shear stiffness as the adjacent soil, and that the end

    walls should have the same friction as the adjacent soil. Of the many containers,

    the equivalent shear beam container (Zeng and Schofield, 1996) is the one close

    to satisfying the requirements. This container is built from rectangular frames of

    dural separated by rubber layers as to achieve the same dynamic stiffness as the

    soil. To sustain the complementary shear stresses induced by base-shaking, a

    flexible and in-extensible friction sheet was attached to the end walls.

    1.2.4.4 CWRU Centrifuge

    The Case Western Reserve University geotechnical centrifuge has a dual

    platform with an effective radius of 1.37m. The centrifuge payload capacity is 20

    g-ton with a maximum acceleration of 200g for static tests and 100g for dynamic

    tests. The centrifuge is equipped with a hydraulic shaker designed by the TEAM

    Corporation. The laminar box used in this test consists of 13 rectangular

    aluminum rings separated by linear bearings. The internal dimensions of the box

    are 53.3 cm (length) 24.1 cm (width) 17.7 cm (height). The following details

    of the centrifuge were reported by Figueroa et al. (1998).

    The centrifuge driving system consists of a 15 HP premium efficiency AC

    motor and torque control inverter which powers the centrifuge arm through a belt

    drive. The support structure consists of the main shaft, rotational bearings,

    bearings housings, a triangular shaped support skirt, and three footings. The

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    33/182

    17

    imbalance force as well as dead loads is transmitted to the foundation through this

    structure. The centrifuge arm is balanced by adjusting the counterbalance weights

    on the swing platform which opposes the testing platform. The dynamic

    imbalance force is monitored at one of three support footings with a sensitive

    LVDT which is connected to the centrifuge control computer.

    The centrifuge is controlled using a program. The computer is configured

    with two data acquisition boards which send and receive analog and digital

    signals to and form the centrifuge.

    The centrifuge data acquisition system is designed to accommodate a wide

    variety of static and dynamic tests. The signal conditioning is performed by the

    signal conditioning chassis which rides on the centrifuge arm in the

    instrumentation rack mounted over the center of rotation. The digital analog lines

    for the signal conditioning chassis are connected to a multifunction data

    acquisition and control board. The conditioned analog signal is then passed

    through the slip rings and on to the AT M10 16E.

    (1) Electro-Hydraulic Shaker

    In the centrifuge test, earthquake motions are simulated by the shake table

    installed on the arm. This is an ideal method for modeling the responses of field

    structures during the earthquakes.

    Many different techniques have been introduced to generate motions of

    the shake table. Table 1.3 is a comparison of the techniques. It is clear that the

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    34/182

    18

    electro-hydraulic method is considered superior to other methods because of the

    large forces, which can be generated by the hydraulic actuators as well as its

    versatility in producing pre-programmed motions (Ko, 1994). The centrifuge

    shaker of CWRU is the electro-hydraulic type.

    Table 1.3 Comparison of Methods for Centrifuge Earthquake Motion

    Simulation (after Whitman, 1988)

    Cost Simplicity Adjustability Frequency Range

    Low High

    Cocked

    Springs

    Very

    Low

    Very Simple Poor --------------

    Piezoelectric Low Simple Good ------

    Explosive Low Simple Moderate ------------

    Bumpy Road High Complex Moderate ------------

    Hydraulic Very

    High

    Very Complex Very Good ------------------------

    (2) Model Container

    In a dynamic centrifuge test, the soil model in the container should

    simulate the behavior of free field of the continuous soil horizon in the field

    during an earthquake. The container may affect stress field, strain field, and

    seismic waves. The boundary effects of the model container are considered a

    major potential source of error. The design criteria for an ideal container as

    following (Taylor, 1995):

    1) The end walls function as shear beams with the same dynamic

    stiffness as the adjacent soil, so as to achieve strain similarity and

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    35/182

    19

    to minimize the interaction between the soil and the end walls and

    hence minimize the generation of compression waves.

    2) Each end wall should have the same friction as the adjacent soil so

    that it can sustain the complementary shear stresses induced by

    base-shaking and thus the same stress distribution as in the

    prototype shear beam can be achieved.

    3) The side walls should be frictionless so that no shear stress is

    induced between the side walls and soil during base-shaking to

    create the same two-dimensional condition as in the prototype.

    4) The model containment should be rigid statically to achieve a zero

    lateral strain condition, and after shaking to maintain its initial size.

    5) The fictional end walls should have the same vertical settlement as

    the soil layer contained during the spin-up of a centrifuge to avoid

    initial shear stresses on the boundary.

    Containers with different types of boundaries have been developed over

    the years. By far, the most successfully device is a stack of ring or laminar box

    (Whitman et al., 1981; Hushmand, et al., 1988; Zeng and Schofield, 1996). The

    laminar box consists of rectangular frames stacked together with low friction

    bearings between the frames. The idea is that the box would be essentially

    massless and frictionless; the response of the soil to input motion at its base would

    be controlled by the soil rather than the box properties. Energy absorbing

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    36/182

    20

    materials have also been applied to the inner wall of the container to reduce the

    influence of incident waves (Taylor, 1995).

    CWRUs laminar box consists of 13 rectangular aluminum rings separated

    from one another by linear motion anti-friction bearings. The internal dimensions

    of the box are 53.3 cm 24.1 cm 17.7 cm (length width height). The

    laminar box is shown in Figure 1.6. Detailed performance of the box under

    dynamic loads can be found in the thesis of Dief (2000).

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    37/182

    21

    Figure 1.6 CWRU Laminar Box (after Dief, 2000)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    38/182

    22

    1.3 Geotechnical Field Tests

    1.3.1 Introduction

    The conventional approach to characterize the engineering properties of

    soils consists of a chain of events involving the overall appraisal of the problem,

    sampling, testing, and analysis. Where these materials are too difficult to sample,

    likely to be severely disturbed by the sampling process, or could not be sampled

    due to structural safety, in-situ methods are used for engineering property

    evaluation.

    Geophysical techniques are normally used to determine the Youngs

    modulus or shear modulus of the soil through measurement of the compression

    wave and shear wave velocities, respectively. Penetration tests, such as standard

    cone penetration test (SPT) and cone penetration test (CPT), are also widely used

    in the field to evaluate soil properties, including stiffness, density, etc.

    The measurement of soil properties by field tests has a number of

    advantages:

    1) Field tests do not require sampling, which can alter the stress, and

    structural conditions in soil specimens. The problem becomes very acute for soils

    such as very loose sand and peat which are difficult to sample without

    considerable disturbance. Intensifying the need for in-situ determination of

    dynamic properties is that these soils which are difficult to sample are also the

    type of soils which are often of the most concern in dynamic site response

    analysis.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    39/182

    23

    2) Many field tests measure the response of relatively large volumes of

    soil, thereby minimizing the potential for basing property evaluation upon small,

    unrepresentative specimens.

    3) Furthermore, it is often not possible to accurately simulate the in-situ

    stress conditions in the laboratory due to equipment limitations.

    Field tests have some shortcomings however, such as the fact that they

    do not allow the effects of conditions other than the in-situ conditions to be

    investigated easily. In many field tests, the specific soil property is not measured

    but is determined indirectly by empirical equations.

    Field methods range from relatively simple penetration test to small-scale

    loading tests to extensive nondestructive tests. Some field tests can be performed

    on the ground surface, while others require the drilling of boreholes or the

    advancement of a probe into the soil. The followings are some field measurement

    methods of soil elastic modulus and other parameters.

    1.3.2 Seismic methods

    In situ measurements of the propagation behavior of low-amplitude stress

    (seismic) waves within geologic media provides valuable geotechnical site-

    characterization information. Seismic wave methods have proven to be both

    effective and theoretically-sound in the determination of the low-strain material

    stiffnesses of soil layers. The material properties measured from seismic methods

    can be directly applied to low strain problems, such as vibrations of machine

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    40/182

    24

    foundations. Seismic methods combined with other high strain tests, such as

    triaxial tests can provide stiffness and damping values for high strain applications,

    such as earthquake response analysis of earth structures.

    There are four primary field methods routinely utilized to measure the

    shear wave velocity: (1) downhole method, (2) crosshole method, (3) surface

    refraction method, and (4) SASW method. In general, all of these techniques have

    a common limitation: they provide velocity measurements only at low strain

    levels (generally less than 10-4 percent).

    Soil properties that influence wave propagation and other phenomena

    include stiffness, damping, Poissons ratio, and density. Of these, stiffness and

    damping are the most important.

    1.3.2.1 Crosshole Method

    The crosshole method involves generating a shear wave or compression

    wave in one borehole, and at the same depth in one or more adjacent boreholes

    measuring the average travel time for the waves, as shown in Figure 1.7. Based on

    the seismic velocities, shear modulus, Young modulus and even Poissons ratio

    can be determined. By testing at various depths, a velocity profile of the soil layer

    can be obtained. The depth of test can reach 30 to 60 m.

    The major shortcomings are that more than one borehole is needed. The

    measured velocities may not be equal to the actual velocities when the high

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    41/182

    25

    velocity layers exist nearby. In some situations, thin low-velocity layers may be

    missed.

    Figure 1.7 The Crosshole Method (Source: Olson Instruments, Inc.)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    42/182

    26

    1.3.2.2 Downhole method

    The downhole method involves the generation of shear waves or

    compressive with an impulse source at the ground surface adjacent to borehole.

    The travel time of the down-propagating shear wave is measured at one or more

    multi-axis geophones clamped in the borehole at various elevations (see Figure

    1.8). The profile of shear wave or compression wave can be obtained. This test

    requires one borehole, so it is a low cost test compared with the corsshole method.

    Figure 1.8 The Downhole Method

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    43/182

    27

    1.3.2.3 Shear Wave Refraction Method

    In shear wave refraction test, horizontal geophones are employed with the

    sensitive axis of the sensors aligned in a horizontal plane transverse to the

    direction of wave travel (see Figure 1.9). The energy source may be sledge

    hammer blows in extremely shallow search surveys (less that 10 meters), a

    shotgun source when overburden conditions allow, or explosives where depth

    and/or energy attenuation is a deciding factor. This is a non-intrusive method.

    Measurements are performed on the ground surface without the use of boreholes,

    so it is also economic. The test results can provide information for preliminary

    planning purposes and feasibility studies. But, the accuracy of the method is

    restrained in complicated multilayered deposits. It is accurate only under

    conditions where velocities increase with depth;

    Figure 1.9 The Shear Wave Refraction Method (Source: Frontier

    Geosciences, Inc.)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    44/182

    28

    1.3.2.4 SASW (Spectral-analysis-of-surface-waves method )

    The SASW method is a nonintrusive technique that employs surface waves

    of the Rayleigh type to determine the layer thicknesses and stiffnesses of

    subsurface profiles. The SASW technique is based on the dispersive property of

    surface waves propagating in a layered system. In a layered system, the velocity

    of Rayleigh waves varies with the wavelength of the surface waves. The variation

    of velocity with wavelength depends on the subsurface stiffness. In the

    application of the SASW technique, the stiffness profile is determined by first

    measuring the dispersion curve (variation in surface wave velocity with

    wavelength or frequency) and then through either an inversion or forward

    modeling process, determining the layer thickness and their stiffness properties

    (Stephen et al., 1994) (Figure 10). This method is relatively rapid and

    inexpensive. The disadvantages arise due to difficulty of interpretation and lack of

    specific stratigraphic information. Currently it is considered as secondary site

    investigation tools for determining dynamic geotechnical properties. Like the

    shear save refraction method, the SASW has particular application advantages to

    evaluate material properties in sites that are difficulties to penetrate.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    45/182

    29

    Figure 1.10 Spectral-Analysis-of-Surface-Waves Method

    (from http://www.geovision.com/SASW.htm)

    1.3.3 Cone Penetration Test (CPT)

    A cone penetrometer consists of the cone, friction sleeve and measuring

    system as shown in Figure 1.11. The cone penetrometer is pushed into the ground

    at a constant rate and continuous measurements are made. During the penetration,

    the cone resistance and the friction are measured. The results from a cone

    penetration test can be used to evaluate: soil type, soil density, stiffness, and shear

    strength parameters. It permits the incorporation of other sensors, such as pore

    water pressure transducer and temperature gage.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    46/182

    30

    Figure 1.11 Sketch of a CPT (after Baldi et al., 1988)

    1.3.4 Standard Penetration Test (SPT)

    The Standard Penetration Test (SPT) provides a measure of the resistance

    of the soil to penetration, in term of number of blows, N (Figure 1.12). During the

    test, a disturbed soil sample can be obtained. The sample can be used for

    classification and index tests. The SPT N value, which gives an indication of the

    soil stiffness, can be empirically related to many engineering properties, such as

    relative density, stiffness, shear strength, compressibility, and liquefaction

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    47/182

    31

    potential. The test is easy to execute and convenient both above and below the

    water table.

    Figure 1.12 SPT Test (after Kovacs et al., 1981)

    1.4 Piezoelectric Sensors and Their Applications

    1.4.1 Introduction

    Piezocelectric sensors are made of piezo ceramic material. When a

    piezoelectric sensor is electrically stressed by a voltage, its dimensions change.

    When it is mechanically stressed by a force, it generates an electric charge. If the

    electrodes are not short-circuited, a voltage associated with the charge appears. A

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    48/182

    32

    +

    -

    piezoelectric sensor is therefore capable of acting as either a sensing or

    transmitting element or both.

    Figure 1.13 shows a piece of piezo ceramic material, in which P is the

    original polarization field within the ceramic established during manufacture by a

    high DC voltage that is applied between the electroded faces to activate the

    material. The polarization vector P is represented by an arrow pointing from the

    positive to the negative poling electrodes.

    The relationship between the applied forces and the resultant responses

    depend upon the piezoelectric properties of the ceramic, the size and shape of the

    piece, and the direction of the electrical and mechanical excitation. Most

    information about piezoelectric sensors of this section is from the website of

    Piezo System, Inc.

    Figure 1.13 Peizoceremic Material

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    49/182

    33

    1.4.2 Piezo Motors

    Piezo motors convert voltage and charge to force and motion. When an

    electrical field which has the same polarity and orientation as the original

    polarization field is placed across the thickness of a single sheet of piezoceramic,

    the piece expands in the thickness or "longitudinal" direction (along the axis of

    polarization) and contracts in the transverse direction (perpendicular to the axis of

    polarization) (see Figure 1.14). When the field is reversed, the motions are

    reversed. However, the motion of a sheet in the thickness direction is extremely

    small (on the order of tens of nanometers). On the other hand, the transverse

    motion along the length is generally larger (on the order of microns to tens of

    microns) since the length dimension is often substantially greater than the

    thickness.

    Figure 1.14 Sketch of a Motor (Source: Piezo System, Inc)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    50/182

    34

    1.4.3 Piezo Generators

    When a mechanical stress is applied to a single sheet of piezoceramic in

    the longitudinal direction (parallel to polarization), a voltage is generated which

    tries to return the piece to its original thickness (see Figure 1.15).

    Figure 1.15 Sketch of Generator (Source: Piezo System, Inc)

    1.4.4 2-Layer Elements

    Two-layer elements can be made to elongate, bend, or twist depending on

    the polarization and wiring configuration of the layers. A center shim laminated

    between the two piezo layers adds mechanical strength and stiffness but reduces

    motion. "2-layer" refers to the number of piezo layers as shown in Figure 1.16.

    The "2-layer" element actually has nine layers, consisting of four electrode layers,

    two piezoceramic layers, two adhesive layers, and a center shim.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    51/182

    35

    Figure 1.16 Structure of a 2-Layer Piezo Element

    1.4.4.1 Bender Elements

    Figure 1.17 illustrations common bending configurations. A 2-layer

    element produces curvature when one layer expands (the top layer in Figure 1.17)

    while the other layer contracts (the bottom layer in Figure 1.17). These

    transducers are often referred to as bender elements. Bender motion on the order

    of hundreds to thousands of microns and bender force from tens to hundreds of

    grams are typical. When the direction of the electric field is changed continuously,

    the direction of the curvature of the element deformation will change

    continuously. Thus, the element will vibrate and work as a wave generator.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    52/182

    36

    Figure 1.17 Bender Element Transmitter (Source: Piezo System, Inc)

    On the other hand, when a mechanical force causes a suitably polarized 2-

    layer element to bend, one layer is compressed, and the other is stretched. Charge

    develops across each layer in an effort to counteract this deformation. This

    arrangement is good for a wave receiver (see Figure 1.18).

    Figure 1.18 Bender Element Generator (Source: Piezo System, Inc)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    53/182

    37

    1.4.4.2 Extender Elements

    A 2-layer element behaves like a single layer when both layers expand or

    contract together. If an electric field is applied which makes the element thinner,

    extension along the length and width results. Extender motion on the order of

    microns to tens of microns and forces from tens to hundreds of Newtons are

    typical. When a mechanical stress causes both layers of a suitably polarized 2-

    layer element to stretch or compress, a voltage is generated which tries to return

    the piece to its original dimension.

    1.4.5 Series and Parallel Operation

    The element arranged in series as shown in Figure 1.19 can generate a

    total output voltage two times the voltage generated by an individual layer. This

    arrangement is good for a receiver. On the other hand, for the same motion, a 2-

    layer element arranged for parallel operation (Figure 1.20) needs only half the

    voltage required for series operation. An applied electrical field causes maximum

    deformation, making this arrangement suitable for a transmitter.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    54/182

    38

    Figure 1.19 A 2-Layer Bender Element Poled for Series Operation

    Figure 1.20 A 2-Layer Bender Element Poled for Parallel Operation

    Shape after Deformation

    Vout

    Input Electric Field

    Shape before Deformation

    Original Polarization Field

    Output Force

    Piezo Layers

    V out

    Output Electric Field

    Input Force

    Piezo Layers

    Original Polarization Field

    Shape before Deformation

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    55/182

    39

    1.4.6 Application of Piezoelectric Sensors in Soil Property Determination

    The piezoelectric sensor with its smart or adaptive structure has many

    engineering applications, such as control transducers for aerospace systems and

    health monitorings of structures.

    The bender element technique enables geotechnical engineers to

    determine Gmax of a soil by measuring wave velocity through a porous media. An

    advantage of the test is the measurements and computations are more direct and

    simpler than those in a resonant column test. The bender element may be

    incorporated into other laboratory tests, such as resonant column, triaxial or

    odometer tests. One experimental setup used by Dyvik and Madshus (1985) is

    shown in Fig. 1.21. The shear wave velocity (vs) in the soil specimen can be

    determined by:

    vs = L/t (1.6)

    where, L = the distance between the tip of the transmitter and the tip of the

    receiver

    t = the travel time of the shear wave through the distance L.

    The elastic shear modulus (Gmax) is determined by:

    Gmax = (vs)2

    (1.7)

    where, = the mass density of the soil.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    56/182

    40

    In Figure 1.21, the bender elements can be replaced by a set of extender

    elements. Then P wave velocity vp can be measured as:

    vp = L/t (1.8)

    and elastic modulus can be calculated by:

    E = (vp)2

    (1.9)

    Figure 1.21 Setup for Incorporating Bender Elements in the Resonant

    Column Apparatus (after Dyvik, 1985)

    Pedestal

    Receiver

    Receiver Signal

    Driving Signal

    Digital Oscilloscope

    Top Cap

    Soil Specimen

    Transmitter

    Wave Generator

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    57/182

    41

    CHAPTER TWO

    LITERATURE REVIEW

    2.1 Review of Piezoelectric Sensors in Soil Property Measurements

    Piezoelectric crystals were first introduced to soil tests by Lawrence (1963,

    1965). He used the crystals to send and receive shear and compression waves in

    sand and clay. Latter, Bender elements were used by some researchers in

    measuring shear wave velocities (Shirley and Hampton, 1978; Shirley and

    Anderson, 1978a & b). Horn (1980) used piezoelectric transducers comprised of a

    stack of six bender elements connected together in saturated soil tests. The

    elements coated by epoxy resin were used both to generate and receive shear

    waves. Howarth (1985) used piezoelectric transducers in measuring compression

    and shear wave velocities through rock specimens in a triaxial cell. Tarun et al.

    (1991) used a triaxial-piezoelectric device to measure compression and shear

    wave velocities of a granular material, a glass sphere assembly. They used very

    high voltages to generate enough energy in granular media to have measurable

    amplitude signals at the receiving transducers. The P-wave was generated by

    applying 400 to 600 volts of a single pulse to compression wave transmitters. The

    shear wave transmitters were actuated with a voltage of 80 to150 volts.

    Lings and Greening (2001) introduced a single hybrid element termed

    bender/extender, capable of transmitting and receiving both S and P waves

    using a single pair of elements mounted across a dry sand sample. A 10 kHz

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    58/182

    42

    single sinusoidal pulse with peak-to-peak amplitude of 20 Volts was used as

    source signal. The received signals were connected directly to a digital

    oscilloscope without using a charge amplifier. The test results showed that the

    bender/extender could provide clear signal that are easy to interpret.

    Bender elements have been incorporated into conventional geotechnical

    test devices, such as triaxial, odometers, and direct simple shear devices. Since

    specimens are not disturbed during bender elements tests, the specimens can be

    subsequently tested for other soil properties. Dyvik and Madshus (1985)

    combined piezoelectric tests into triaxial tests. A parallel connected bender

    element was used as transmitter, and a series connected element was used as the

    receiver of shear wave. For the series connected element, when used as a receiver,

    it is twice as effective as a parallels connected element. On the other hand, a

    parallels connect bender element is twice as effective as a series connected

    element when used as a motor. Square waves of amplitude 10 Volts were used

    as input signals. The maximum shear strain for the tests was estimated to be on

    the order of 10-3

    %. This is in the elastic range of soil deformation. The bender

    element test results compared with those of resonant column tests showed general

    agreement.

    Thomann and Hryciw (1990) performed bender element tests in a large

    odometer to measure Gmax under Ko condition. Zeng and Ni (1999) applied the

    bender element technique to investigate the stress-induced anisotropy in Gmax of

    sands. In their tests, bender element receivers were located at different directions

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    59/182

    43

    in an odometer with respect to transmitters. Thus, shear wave velocities in

    different planes of a specimen were measured and shear moduli were calculated

    from the shear wave velocities. Zeng et al. (2002) developed a new test device

    called the bender element cone penetrometer to measure the stiffness of base and

    subgrade layers in a pavement.

    Besides the measurement of stiffnesses of soils, the bender element

    technique has also been used in investigating the liquefaction properties of the

    soils based on the measured shear wave velocities. Shirley and Anderson (1978a

    & b) used single bender element crystals as shear wave transmitter and receiver.

    They performed shear wave propagation experiments on a saturated specimen

    under undrained condition. The liquefaction was introduced by applying a sudden

    increase in chamber pressure. They noticed that the disappearance of shear wave

    was observed near liquefaction stage. De Alba et al. (1984) used piezoelectric

    transducers in a triaxial specimen to investigate the liquefaction property of sands

    and its relationship to shear wave velocities. The transducers were an array of four

    cantilevered bender elements.

    The factors that affect the bender element test results include the first

    arrival time of the received signal, frequency of source signal, and rising time of

    source signal etc. The first arrive time of the received signal is a main factor that

    may affect test results. It is common practice to locate the first arrival of the shear

    wave at the point of first deflection of the received signal. Reversal of the polarity

    of the received signal as the polarity of the input signal is reversed is often taken

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    60/182

    44

    as demonstration the arrival of the shear wave (Abbiss, 1981). However,

    theoretical studies by Salinero et al. (1986) showed that the first deflection of the

    signal may not corresponding to the arrival of the shear wave but to the arrival of

    the so-called near-field component which travels with the velocity of a

    compression wave. The near-field effect may mask the arrival of the shear wave

    when the distance between the source and the receiver is in the range of ~ 4

    wavelengths, which can be estimated from = Vs/f where f is the mean frequency

    of the received signal (Viggiani and Atkinson, 1995). Regards to the length of the

    wave, Thill et al. (1968) reported that a wavelength approximating the average

    grain size resulted in almost complete signal attenuation and therefore made the

    detection of wave arrival almost impossible. They recommended a wavelength

    approximately 10 times the average grain size for low attenuation of propagated

    waves. Their tests were conducted on rock samples. Zeng and Ni (1999) studied

    the effect of the rising time of source signal on the test results. Square waves of

    different rising time were used. They observed that the change in rising time of

    the electric pulse did not affect the travel time for shear waves. But if the rising

    time is too large, the vibration of the transmitter will be too weak to make a well-

    defined first arrival of the wave.

    2.2 Review of Soil Property Measurements in Centrifuge Tests

    Centrifuge testing has gradually become a routine experimental method in

    geotechnical engineering throughout the world given the fact that it can be used to

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    61/182

    45

    study a wide range of problems. Centrifuge modeling provides engineers with a

    cost effective experimental tool to investigate phenomena of concern at full-scale

    stress conditions on a small-scale model. Data from centrifuge tests can be used to

    study the mechanisms of complex problems, to validate numerical models, and to

    verify new design methods. However, centrifuge modeling is also a complex

    experimental method. Like many other experimental methods, centrifuge

    modeling can have inherent inaccuracies and difficulties arising from factors such

    as boundary conditions imposed by the model container; inability to satisfy all the

    scaling laws in certain situations; system limitations of equipment, transducers;

    and data acquisition, and the measurement of soil properties under centrifugal

    acceleration during in-flight conditions. The usefulness and accuracy of testing

    data depend critically on how well the effects of these problems are understood

    and addressed. In recent years, a significant amount of research has been

    conducted to study problems of transducer response (Kutter et al., 1990; and Lee,

    1990), boundary effects in earthquake centrifuge tests (Hushmand, et al., 1988;

    and Zeng and Schofield, 1996), the use of viscous pore fluids (Zeng et al., 1998;

    Ko and Dewoolkar, 1998), and the influence of variation of centrifugal

    acceleration and model container size on the accuracy of centrifuge test (Zeng and

    Lim, 2002).

    One of the challenges in centrifuge modeling is the measurement of soil

    properties during the centrifuge flight. In most cases, soil properties in a model

    are measured before and/or after a test under a 1g condition. However, as

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    62/182

    46

    mechanical properties of soils such as strength and stiffness are highly stress

    dependent, one can expect that these properties can be significantly different

    during centrifuge flight. Without direct measurements, some important soil

    properties, such as Gmax and K0, in the centrifuge model are usually assumed or

    obtained by other test methods, including monotonic or cyclic triaxial tests,

    monotonic or cyclic direct simple shear tests, or resonant column tests. This

    creates a difficult situation for numerical modelers in the simulation of

    experimental results. For example, in the VELACS project (Arulanandan and

    Scott 1993), the values of Gmax and K0 for the same models used by different

    predictors for numerical simulation of liquefaction problems varied significantly

    for the same sand (Zeng and Arulanandan 1994). The differences in soil

    parameters used by different predictors could have contributed significantly to the

    reported differences in the numerical simulation results. Therefore, a reliable

    measurement of soil properties during the flight of a model is necessary in order

    to get a good numerical simulation of the test phenomena and to have a clear

    understanding of the test results.

    Due to the uncertainties and errors exit in centrifuge tests as previously

    discussed, soil property measurements in in-flight centrifuge tests are important.

    The investigation of samples can be conducted before, during, and after the test.

    This research will focus on an investigation carried out in-flight. The

    investigations before and after a centrifuge test can follow the same procedures as

    for full scale field tests. In-situ investigation tools have been scaled down to

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    63/182

    47

    conduct comparable tests in-flight, such as vane tests, T-bar tests, cone

    penetration tests and seismic method tests.

    In recent years, in-flight measuring devices have been developed to

    provide more reliable soil property measurements in centrifuge tests. For example,

    cone penetrometer tests are performed to determine soil strength or to check

    homogeneity of the soil in the model. A test is carried out by measuring the

    penetration resistance on the cone and, if possible, including the measurement of

    pore water pressure. To use the penetration test data, empirical relationships for

    their correlation with fundamental soil properties must be developed (Ferguson,

    1985). The inter-laboratory variability of test results was assessed by five

    European laboratories (Renzi et al., 1994). A comparison showed that the

    difference in test results for sand was of the order on 20%. Katagiri and Okamura

    (1998) proposed a cone penetration test manual for centrifuge tests.

    Shibata et al. (1991) measured shear wave velocities in-flight. A

    piezoelectric oscillator is used for generating seismic waves. The accelerometers

    of piezoelectric type are placed at two different points to detect seismic motions

    propagated in the soil. The accelerometers are embedded at the same depth as that

    of the oscillator and are aligned with the oscillator (see Figure 2.1). The driving

    signal of a rectangular waveform is produced by a function generator. The

    frequency of the signal is 10 KHz, and double amplitude is 10 V. Before it is sent

    to the oscillator, the signal is amplified 10 times. The signals from the

    accelerometers are processed through high-pass filter, with a cut-off frequency at

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    64/182

    48

    1.5 KHz. The phase shifts brought by the filter are assumed to be the same for the

    two receivers. From the distance between the generator and the receiver, and the

    traveling time difference between them, the shear wave velocity is determined.

    Figure 2.1 Experimental System (after Shibata et al., 1991)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    65/182

    49

    Gohal and Finn (1991) reported measurements of the shear wave velocity

    in centrifuge tests. Their experimental setup is shown in Figure 2.2. The elements

    were attached to bearing plates and lined up vertically. A 10 volt single

    amplitude, 30 Hz square wave pulse was used as the source signal. The received

    signal was first amplified, then passed via the centrifuge slop rings to a data

    acquisition system. The difficulty of the setup was to accurately determine the

    location of the receivers during the test. A loose and a dense sample of one kind

    of sand were tested and the shear wave velocities at different depths of the model

    were obtained.

    Figure 2.2 Layout of Bender Source and Receiver (after Gohal and Finn,

    1991)

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    66/182

    50

    Arulnathan et al. (2000) used a mini-air hammer to generate shear waves.

    The propagation of shear waves in the model was recorded in-flight by a vertical

    array of four accelerometers placed at different depths (Figure 2.3). The hammer

    consisted of a hollow aluminum cylinder 47 mm long. A 25 mm long piston fit

    inside the cylinder. The piston was fired towards the end of the cylinder by

    applied air pressure, and the motion of the cylinder produced the shear wave.

    Models of Nevada sand at 80 g were tested. The shear wave measured were

    compared with those from piezoceramic bender element tests in a triaxia device

    and got excellent agreement. Models of Nevada sand at 80 g were tested. The

    shear waves measured were compared with those from bender element tests in a

    triaxial device and were in excellent agreement.

    More recently, in-flight operating robots have been developed to conduct

    sophisticated soil property measurements (Gaudin et al., 2002 and Ng et al., 2002).

    In recent years, as micro machine-work and measuring technologies

    advance, instrumentation applicable to centrifuge model tests is developing. A

    wide variety of instrumentation is employed in centrifuge tests. The instruments

    should be tested and calibrated over their work load using the model ground,

    which is performed in a centrifuge or at the Earths gravity.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    67/182

    51

    Figure 2.3 In-Flight Shear Wave Measurement (after Arulnathan et al., 2000)

    Mini-air Piston

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    68/182

    52

    2.3 Research Objectives and Outline

    2.3.1 Research Objectives

    The primary objective of this research is to utilize piezoelectric sensors

    in the determination of soil properties in the field and laboratory. The first

    objective of the research is to develop a technique for soil property measurements

    in the centrifuge. Then the technique will be used to measure shear wave

    velocities in dry and saturated models of Nevada sand before and after

    earthquakes during the spin-up, spin-down, and in-flight stage of a centrifuge. The

    effects of repeated earthquakes on the liquefaction property of the soil will also be

    investigated. Then, based on the test results, shear wave-based liquefaction

    criteria will be examined.

    Another objective of the research is to develop a new cone penetrometer

    equipped with piezoelectric sensors. The equipment is aimed at in-situ stiffness

    measurement of sublayers of existing pavements or construction projects. A series

    of laboratory tests will be performed using the equipment, and the test results will

    be compared with conventional CBR test results.

    Finally, a large odometer equipped with piezoelectric sensors for

    laboratory measurements of properties of gravelly materials will be developed.

    Typical test results of the odometer will be presented

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    69/182

    53

    2.3.2 Outline

    Chapter 3 presents the use of bender elements in dry specimen centrifuge

    tests. The bender element technique used is presented in detail. Sample

    preparation methods and test procedures are described. The test results are

    presented and compared with the results obtained by other methods.

    Chapter 4 presents the use of bender elements in saturated specimen tests.

    The bender element technique for wet soil specimen tests, sample preparation

    methods and test procedures are described. The dynamic centrifuge test results on

    Nevada sand will be presented. Shear wave-based liquefaction criteria will be

    checked and discussed in this chapter.

    Chapter 5 introduces a new piezo cone penetrometer. The device is

    described in detail. Typical test results are presented and compared with the

    results obtained by CBR tests.

    Chapter 6 introduces a new odometers for gravel material tests. The

    devices are described in details. Typical test results are presented.

    Chapter 7 summarizes the significant findings of this research.

    Suggestions for future research are also presented.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    70/182

    54

    CHAPTER THREE

    BENDER ELEMENTS IN CENTRIFUGE TESTS---DRY SPECIMEN TESTS

    3.1 Introduction

    Due to the uncertainties and errors that exist in centrifuge tests as

    discussed in section 1.2.4.3, the measurements in in-flight centrifuge tests are

    important. The investigation of samples can be conducted before, during, and

    after tests. This chapter presents and discusses in detail the bender element

    technique in measuring soil properties during centrifuge tests. Models of Nevada

    sand were tested in dry conditions.

    The CWRU centrifuge was used in all of the tests. The centrifuge was

    swung up in steps of 10g, 20g, 40g, and 50g. At each step, a total of nine bender

    element tests were performed. At 50g, an earthquake was applied to the soil

    model. After the earthquake, the centrifuge was spun-down at the step of 50g, 40g,

    20g, 10g, and 1 g. Again, at each step, bender element tests were performed.

    Parameters, such as accelerations and settlement were monitored.

    From the test results, the shear wave velocities of the soil in different

    stages of the tests were calculated. By further calculations, the profiles of the

    stiffness of the models were obtained. The test results were discussed and were

    compared with the results of the empirical equation proposed by Hardin and

    Richart (1963), as well as the results of resonant column tests.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    71/182

    55

    3.2 Bender Elements

    The bender element types Q220-A4-303Y (transmitter) and Q220-A4-

    303X (receiver) used in these tests are produced by Piezo System Inc. with

    dimensions of 34.7 12.7 0.5 mm (length width thickness) as shown in

    Figure 3.1. The performance of the bender element transmitter Q220-A4-303Y is

    listed in Table 3.1.

    Figure 3.1 Dimensions of the Bender Element

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    72/182

    56

    Table 3.1 Bender Element Performance (Source: Piezo Systems, Inc.)

    Part Number Q220-A4-303Y

    Piezo Material 5A4E

    Weight (grams) 2.5

    Stiffness (N/m) 843

    Capacitance (nF) 50

    Maximum Voltage ( Vp) 90

    Resonant Frequency (Hz) 250

    Free Deflection (m) 325

    Blocked Force (N) 0.27

    3.3 Nevada Sand

    Nevada sand is used in the model tests. It was purchased from the Gordon

    Sand Company of Compton, California. The sand has been widely used in

    laboratories around US for liquefaction studies. It was used in the VELACS

    project (Arulanandan and Scott, 1993) for dynamic centrifuge model tests.

    Classification tests determining material properties for Nevada sand such as

    specific gravity, grain size distribution, and the maximum and minimum void

    ratios were previously performed by other researchers. The grain-size distribution

    is plotted in Figure 3.2 and the corresponding index properties are shown in Table

    3.2.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    73/182

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    74/182

    58

    3.4 Experimental Setup

    The tests described in this paper incorporate bender elements into

    centrifuge tests to measure shear wave velocities during the spin-up, in-flight, and

    spin-down of the centrifuge. The experimental setup is shown in Figure 3.3. The

    elements, which are fixed to the supporting columns which are placed in a laminar

    box, are shown in Figure 3.4. There are six bender elements, three transmitters,

    named S1, S2, and S3, respectively, and three receivers, designated R1, R2, and

    R3, respectively. They are located approximately at one-third and two-thirds of

    the model depth and near the bottom of the model. The distance between the

    transmitting and receiving elements at each depth level is 10cm. In the setup, the

    tips of the elements are 2.8 cm away from the columns, so the influence of the

    columns on the stress states of the soil between the tips of the elements is small.

    This can be seen from the comparison of the bender element tests results with

    those by resonant column tests shown in Fig 3.14. A large model box, a all

    column size, and large distance between the transmitters and receivers can reduce

    the effects induced by columns.

    The bender element types Q220-A4-303Y (transmitter) and Q220-A4-

    303X (receiver) were used. The triggering signal, a square pulse, which was sent

    to the transmitters in the laminar box, was produced by an Agilent 54624A wave

    generator located in the control room. Vibrations of the transmitters produced

    shear waves that propagate through the soil and were recorded by the receivers.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    75/182

    59

    The recorded signals were first amplified by Keithley MB38-02 signal-

    conditioning modules and then transferred through the slip rings to an Agilent

    33120A oscilloscope located in the control room. The traveling time of the shear

    waves could be read from the oscilloscope directly, or the data could be

    downloaded on disk for later processing. A multiple switch was devised in the

    control room to schedule tests in the sequence of S1 to R1, S1 to R2, S1 to R3,

    and then from S2 to R1, etc. For each test, the traveling time was measured based

    on one pulse test. The test usually had good repeatability. The wave traveling

    distances for the horizontal travel directions were measured from tip to tip, and

    for non-horizontal travel directions, they were measured between the closest

    corners of the elements.

    Square waves of amplitude 10 volts were used as triggering signals in the

    tests. Some researchers have used signals of the same amplitude, such as Dyvik

    and Madshus (1984) and Gohl and Finn (1991). Dyvik and Madshus (1984) also

    showed in their tests that the maximum shear strain was on the order of 10-3

    %.

    This is well within the elastic range of soils. Later, the Gmax values obtained by

    bender element tests were compared with those by resonant column tests (see

    Figure 3.13). The frequencies of the triggering signals used in the tests varied

    from about 100 to 3000 Hz. The frequencies have no obvious effect on the

    traveling time measured, but they can be changed to receive clear signals.

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    76/182

    60

    Foundation

    Main Shaft

    Agilent 33120AOscilloscope

    ArmCounterweight

    Skirt

    Agilent 54624A

    Wave Generator

    Wires

    Slip Rings

    Keithley MB38-02

    Signal-Conditioning Modules

    Receivers

    Transmitters

    Laminar Box

    Figure 3.3 Experimental Setup

    Figure 3.4 Test Equipment in the Laminar Box

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    77/182

  • 7/31/2019 Application of Piezoelectric Sensors in Soil Property Determination

    78/182

    62

    Table 3.3 Locations of the Bender Elements in Models

    Model Transducer X(mm) Y(mm) Z(mm)

    S1 0 -50 107

    S2 0 -50 57

    S3 0 -50 8

    R1 0 50 107

    R2 0 50 57

    Model 1

    R3 0 50 8

    S1 0 -50 107

    S2 0 -50 53

    S3 0 -50 9

    R1 0 50 107

    R2 0 50 53

    Model 2