Title Page
 Table of Contents
 List of Tables
 List of Figures
 Impact driving of piles, capacity...
 Similitude requirements
 Equipment design, fabrication,...
 Specimen preparation, test results,...
 Conclusions and recommendation...
 Biographical sketch

Title: Development and testing of a device capable of placing model piles by driving and pushing in the centrifuge
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00090203/00001
 Material Information
Title: Development and testing of a device capable of placing model piles by driving and pushing in the centrifuge
Series Title: Development and testing of a device capable of placing model piles by driving and pushing in the centrifuge
Physical Description: Book
Creator: Gill, John Joseph,
 Record Information
Bibliographic ID: UF00090203
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 001124147
oclc - 20063420

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
    List of Tables
        Page viii
    List of Figures
        Page ix
        Page x
        Page xi
        Page xii
        Page xiii
        Page xiv
        Page xv
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
    Impact driving of piles, capacity prediction, and capacity verification
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
    Similitude requirements
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
    Equipment design, fabrication, and operation
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
    Specimen preparation, test results, and discussion
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
        Page 259
        Page 260
        Page 261
        Page 262
        Page 263
        Page 264
        Page 265
        Page 266
        Page 267
        Page 268
        Page 269
        Page 270
        Page 271
        Page 272
        Page 273
        Page 274
        Page 275
    Conclusions and recommendations
        Page 276
        Page 277
        Page 278
        Page 279
        Page 280
        Page 281
        Page 282
        Page 283
        Page 284
        Page 285
        Page 286
        Page 287
        Page 288
        Page 289
        Page 290
        Page 291
        Page 292
        Page 293
        Page 294
        Page 295
        Page 296
        Page 297
        Page 298
        Page 299
        Page 300
        Page 301
        Page 302
    Biographical sketch
        Page 303
        Page 304
        Page 305
Full Text








This dissertation is dedicated to A. Daniel Decker, my good friend

and father-in-law, who lost the final battle but still won the war.


An effort of this scope cannot be accomplished without the cooper-

ation of many people and I would like to thank my many friends who

helped make it possible. I am indeed fortunate to be the person to

receive the credit for and benefit of all of their selfless assistance.

I credit my committee for the conception of and technical assistance

throughout the course of this research. Dr. Frank Townsend consistently

challenged me during my brief stay at the University and thus fed my

desire to pursue higher goals. I appreciate his confidence and trust in

me to achieve the technically difficult project undertaken.

My utmost thanks go to Dr. David Bloomquist for his friendship, time,

and encouragement, all of which were freely given. As my mentor, he

provided the example of what it means to be an engineer and profes-

sional. His attitudes and approach to complex problems will serve me

for years to come. I am sincerely indebted to both Drs. John Davidson

and Michael McVay for their vital assistance in the proper and meaning-

ful interpretation of the data obtained. Lastly, I thank Dr. David

Jenkins for his help in providing important details on the operation of

some of the finer instruments involved in this research.

I am thankful for the monetary assistance of both the United States

Air Force and the Florida Department of Transportation as none of our

achievements would have been possible without such help. I am most

thankful to the Air Force for providing me the opportunity to pursue my

education to its current level.

I wish to thank my family for their constant support, especially my

sister, Dr. Barbara Carnes, for her wise counsel during this effort.

Their individual contributions have helped me in ways they may not even

be aware of.

My co-workers and friends at the University have helped smooth the

often rocky path of these past few years. Niels Kofoed earned my admi-

ration and respect for his finesse with the computerization of the model

pile placement device operations. Jay Gabrielson and Charlie Manzione

deserve much credit for their help and comic relief. I hope I may have

been able to repay them in kind. Likewise, credit goes to Ken Knox and

Gregory Coker for their friendship.

I take this opportunity to thank my many friends working.in the

civil engineering department and the engineering shop. Mr. Ed Dobson

helped tremendously in the initial setup of the centrifuge. Anita

Cataldo and Cheryl Holland guided me pleasantly through the financial

and administrative systems. I thank Candace Leggett for her tolerance

during the final weeks of typing of this dissertation. Charles Simmons,

Marc Link, and Dale Wilcox were of great assistance to me at the EIES

Shop. Danny Richardson, Bill Studstill, and Kirk Waite have all been

willing helpers.

My wife, Carole, and daughter, Sarah, gladly shouldered the burden

of my long hours at school and the many additional hours of "at-home"

absence. Their patience during this process has only increased my love

and respect for them.

Lastly, I thank the many people to whom I am deeply indebted yet

cannot mention for lack of space. Without their assistance, this

research would not have been quite as fruitful. I thank them all.


ACKNOWLEDGMENTS ... ....... ............... .... ....................

LIST OF TABLES.....................................................

LIST OF FIGURES....................................................

ABSTRACT ..................... .............. .... ...... .... ........


1 INTRODUCTION...............................................

1.1 General....................... .........................
1.2 Objectives..............................................
1.3 Scope of Work........................ ..................
1.4 Review of Previous Research...........................

CAPACITY VERIFICATION......... ......... ......... .........

2.1 Introduction...........................................
2.2 Pile Placement by Impact Driving........................
2.2.1 Elastic Wave Propagation in Solids...............
2.2.2 Hammer-Pile Impact...............................
2.2.3 Hammer-Pile-Soil Interaction...................
2.3 Practical Aspects of Pile Driving .....................
2.3.1 Hammer-Cushion-Cap Interaction...................
2.3.2 Pile Forces Developed During Driving
and Loading....................................
2.3.3 Errors in Placement and Driving Affecting
Static Capacity..................................
2.4 Static Capacity Verification...........................
2.5 Pile Group Behavior...................................

3 SIMILITUDE REQUIREMENTS....................................

3.1 Similitude Theory.....................................
3.2 Selection of Dependent and Independent Variables.......
3.3 Development of Scaling Laws............................
3.4 Experimental Requirements..............................















4.1 Model Pile Design and Instrumentation................... 60
4.1.1 Model Material Selection........................ 60
4.1.2 Model Size Determination......................... 61
4.1.3 Strain Gage Placement............................ 64
4.1.4 Pile (Group) Cap Fabrication and
Lead Wire Protection............................. 74
4.2 Model Placement Device Design........................... 74
4.2.1 Specimen Container, Placement Device Protec-
tive Canister, and C-Channel Support Beam........ 81
4.2.2 Stepper Motor......................... ........... 86
4.2.3 Ball Screw Assembly............................. 88
4.2.4 Guide Tube......................................... 96
4.2.5 Electromagnet.......... ..................... 96
4.2.6 Model Pile Hammers............................... 98
4.2.7 Model Pile Cap as an Alignment Tool.............. 101
4.2.8 Proximity Device................................ 104 Charge pump proximity device............ 108 Phase comparator proximity device....... 113
4.2.9 Load Cells........... .................... ...... 117
4.3 Centrifuge Design............................ ....... 117
4.3.1 In-Place Equipment............................... 119
4.3.2 Shroud Construction........................... 120
4.3.3 Main Arm Modification........................... 123
4.3.4 Cover Construction.............................. 123
4.3.5 Slip Ring Placement............................. 127
4.3.6 Rotational Speed Monitoring..................... 129
4.3.7 Specimen Platform Construction.................. 132
4.3.8 Closed Circuit Television....................... 136
4.4 Model Pile (Group) Set-Up Procedure.................... 138
4.5 Computer Control of the Equipment...................... 145
4.5.1 Hardware ............................ ............. 146 Hewlett-Packard 9000 series,
model 216 computer...................... 146 Hewlett-Packard 6940B multiprogram-
mer and 59500A interface unit........... 146
4.5.2 Software......................................... 155 Main Programs........................... 157 Subroutines... .......................... 167
4.6 Data Collection and Recording Equipment................. 173
4.6.1 Hardware......................................... 173 Hewlett-Packard 3497A Data Acquisi-
tion/Control Unit....................... 173 Transducers............................. 173
4.6.2 Software......................................... 195
4.7 Equipment Limitations.............................. .... 195


5.1 Soil Description and Specimen Characterization.......... 212
5.2 PUSHPILE and PUSHGROUP Test Results.................... 217
5.2.1 Qualitative Discussion of Pushed Model Pile
Test Results............ ........................ 217

5.2.2 Quantitative Discussion of Pushed Model Pile
Test Results.................................... 238
5.3 DRIVEPILE Test Results.................................... 254
5.3.1 Qualitative Discussion of Driven Model Pile
Test Results..................................... 254
5.3.2 Quantitative Discussion of Driven Model Pile
Test Results..................................... 264
5.4 Reproducibility of Results.............................. 267
5.5 Modeling of Models..................... ................. 269
5.6 Comparison with a Prototype............................. 270
5.7 Strain Measurement .......... .......... ............. 272

6 CONCLUSIONS AND RECOMMENDATIONS.............................. 276

6.1 Conclusions............................................. 276
6.2 Recommendations......................................... 281



ALGORITHMS ....................... .......................... 293

REFERENCES ..... ...................... 296

BIOGRAPHICAL SKETCH. ....... ............................ ...... ..... 303


Table Page

3-1 Independent and Dependent Variables.......................... 55

3-2 Scaling Relationships................... ..................... 56

4-1 Model Pile Dimensions and Scale Factors...................... 62

4-2 Model Hammer Weights and Dimensions.......................... 100

4-3 Platform Specifications and Factor of Safety Computations.... 137

4-4 Multiprogrammer Operating Mode Codes......................... 151

4-5 Stepper Motor Response to Commanded Rotation
at Test g-Levels........................................... 206

5-1 Pushed Pile (Group) Model Test Series........................ 223

5-2 Pushed Pile (Group) Load Test Results........................ 246

5-3 Pile Group Factors Based on deBeer Capacities................ 249

5-4 Initial Loading Curve Slope, Elastic Deformation Curve
Slope and Resulting Soil Response for Individual and
Group Model Piles............................................ 251

A-1 Pisets Input Matrix.................... .................... 284

A-2 Pi Terms..................................................... 285

A-3 Scaling Law Derivation....................................... 286



Figure Page

2-1 Propagation of an Elastic Wave............................... 20

2-2 Idealized Reflections of an Elastic Wave. a) Reflection
from Stress-Free End Rods; b) Reflection from Fixed End
Rods ......................................................... 24

2-3 Idealized Hammer and Pile Before Impact...................... 25

2-4 Mechanism of Impact of a Cylindrical Hammer on a
Finite Cylindrical Rod....................................... 28

2-5 Effect of Impedance Ratio on the Theoretical Wave Shape
for Same Hammer Mass and Drop................................ 31

2-6 Finite Difference Pile Representation........................ 34

2-7 Finite Element Mesh Pile Representation...................... 36

2-8 Definition of Pile Cap Terminology........................... 38

4-1 Model Piles. a) Individual Model Piles (Left to
Right--49.3, 57.5, 69.8 and 86.0 g models);
b) Model Pile Groups (Left 69.8 g., Right 86.0 g)........... 63

4-2 Micromeasurements EA-06-015LA-120 Strain Gages............... 66

4-3 Strain Gage Application Technique............................ 67

4-4 Alternate Technique for Placement of First Pair of
Strain Gages When Multiple Pairs are Needed.................. 70

4-5 Suggested Gage Posi.tions and Position Designations for
Individual Driven and Pushed Piles (8 Strain Gage Channels
Available). a) Driven Pile, Four Pair of Strain Gages;
b) Pushed Pile, Four Pair of Strain Gages and Load Cell...... 72

4-6 Suggested Gage Positions and Position Designations for
Driven and Pushed Pile Groups (Load Cell Placed on Top
After DRIVEPILE; Load Cell In Place During PUSHPILE)......... 73

4-7 Pile Cap Assembly for Individually Driven Model Piles........ 75

4-8 Pile Cap Assembly for Driven Model Pile Groups............... 76

4-9 Pile Cap Assembly for Individually Pushed Model Piles........ 77

4-10 Pile Cap Assembly for Pushed Model Pile Groups............... 78

4-11 Specimen Container, Placement Device Protective
Canister, and C-Channel Support Beam......................... 80

4-12 Top View of Canister and Placement Device Configuration.
a) Configuration Used for Video Recording; b) Normal
Configuration................................................ 83

4-13 Platform Mechanical Stops.................................... 85

4-14 Stepper Motor Transformer and Variable Power Resistors....... 89

4-15 Bodine Type THD-1830E Adjustable Motion Control.............. 89

4-16 Stepper Motor Control Circuit and Slip Ring Channel
Assignments.................................................. 90

4-17 Ball Screw Assembly, Thrust Bearings, and Reaction Clamp..... 93

4-18 Electromagnet Actuator Circuit and Slip Ring Channel
Assignments.................................................. 99

4-19 Pile Cap, Hammer, and Electromagnet Configuration for
Driving Sequence............................................. 102

4-20 Guide Rod Length Determination (69.8 g's).................... 103

4-21 Charge Pump Proximity Device Schematic Drawing............... 109

4-22 Charge Pump Proximity Device................................. 110

4-23 Charge Pump Proximity Power and Output Signal Circuit
and Slip Ring Channel Assignments............................ 112

4-24 Phase Comparator Device Schematic Drawing.................... 114

4-25 Basic Phase Locked Loop Circuit...................... ..... 115

4-26 Load Cell Power and Output Signal Circuit and Slip
Ring Channel Assignments....................................... 118

4-27 Centrifuge Main Power Transformer............................ 121

4-28 Parajust AC Motor Speed Control.............................. 121

4-29 Hand-Held Centrifuge Speed Control........................... 122

4-30 Centrifuge Main Circuit Breaker and Resistive Speed Brake.... 122

4-31 Completed Centrifuge Shroud.................................. 124

4-32 Pillow Block Bearing Support Plates Fastened at the
Ends of Each Arm............................................. 124

4-33 Centrifuge Cover, Access Opening and Slip Ring Housing....... 125

4-34 Centrifuge Power Requirements and Performance
Characteristics.............................................. 126

4-35 Electric and Hydraulic Slip Ring Assembly.................... 128

4-36 Slip Ring Assembly In-Place.................................. 130

4-37 Pulse Generator and Slotted Disk................................ 130

4-38 Rotational Speed Monitoring and Control Circuit.............. 131

4-39 Equipment for Monitoring Centrifuge Rotational Speed.
a) Photoelectric Tachometer; b) Display Unit................ 133

4-40 Orthographic Projection of Centrifuge Platform............... 134

4-41 Completed Platform........................................... 135

4-42 Platform In-Place............................................ 135

4-43 Closed Circuit Television Power Supply and Output
Signal Circuit and Slip Ring Channel Assignments............. 139

4-44 Hewlett-Packard 9000 Series, Model 216 Computer.............. 147

4-45 Hewlett-Packard 59500A Interface Unit and 6940B
Multiprogrammer............... ............................. 149

4-46 Computer Variable Designations............................. 158

4-47 Model Pile Calibration Equipment. a) Model Pile Calibration
Accessories; b) In-Place.................................... 166

4-48 Hewlett-Packard 3497A Data Acquisition/Control Unit.......... 174

4-49 Resistance Circuit Element................................... 181

4-50 Wheatstone Bridge Circuit.................................... 182

4-51 Three-Wire Circuit for Strain Measurements................... 186

4-52 On-Board Bridge Completion Circuit for Strain Gages
(8 Channels Available) .................... .............. 188

4-53 On-Board Bridge Completion Unit and Slip Ring
Channel Assignments............... ... ..... ............ ... 190

4-54 Cross Section of a Linear Voltage Differential
Transformer (LVDT) ......................................... 192

4-55 LVDT Power Supply Circuit and Signal Output................ 194

4-56 Maximum Payload Operating Capacities vs. RPM................. 197

4-57 Stepper Motor and RVDT in Test Position...................... 205

4-58 Placement Device and LVDT in Test Position................... 208

4-59 Ball Screw Response to Stepper Motor Drive Shaft
Rotational Displacement ( 1600 Step Displacement,
40-Step Displacement Increment).............................. 209

4-60 Ball Screw Response to Stepper Motor Drive Shaft
Rotational Displacement (t 200 Step Displacement,
1-Step Displacement Increment)............................... 210

5-1 Specimen Preparation "Chimney" and Associated Tools.......... 214

5-2 Specimen Containers...... ........................... 216

5-3 Pile Penetration vs. Load, Single Pile, PG69812072,
PG69812073................................................... 219

5-4 Pile Penetration vs. Load, Single Pile, PG69812151,
PG69812154. ........................................ ..... 221

5-5 Pile Penetration vs. Load, Single Pile, PG86012153,
PG86012181 ................. ............ ............. ........ 222

5-6 Pile Penetration vs. Load, Single Pile, PG69804013,
PG69804011, PG69804021................. ........ ........... 224

5-7 Pile Penetration vs. Load, Single Pile, PG86004014,
PG86004012, PG86004022, PG86012172........................... 226

5-8 Group Penetration vs. Load', Group of Four PG69802251,
PG69802252, PG69802253...................................... 228

5-9 Group Penetration vs. Load, Group of Four, PG86002244,
PG86002243, PG86002242..................... .............. ... 230

5-10 Group Penetration vs. Load, Group of Five, PG69804023,
PG69804051, PG69802254 .... .... .................................... 232

5-11 Group Penetration vs. Load, Group of Five, PG86002231,
PG86002232, PG86002241................ ... .... ...... ......... 234

5-12 Change (%) in Total Frictional Resistance to Pull-Out With
Respect to the Change (%) in Model Pile-Soil Contact Area
(Referenced to Single 86.0 g Model).......................... 239

5-13 Bearing Capacity of Single Pile, deBeer, 04013LOAD,
04011LOAD, 04021LOAD....................................... 240

5-14 Bearing Capacity of Single Pile, deBeer, 04014LOAD,
04012LOAD, 04022LOAD, 12172LOAD.............................. 241.

5-15 Bearing Capacity of Pile Group, deBeer, Group of Four,
02251LOAD, 02252LOAD, 02253LOAD.............................. 242

5-16 Bearing Capacity of Pile Group, deBeer, Group of Four,
02244LOAD, 02243LOAD, 02242LOAD.............................. 243

5-17 Bearing Capacity of Pile Group, deBeer, Group of Five,
04023LOAD, 04051LOAD, 02254LOAD.............................. 244

5-18 Bearing Capacity of Pile Group, deBeer, Group of Five,
02231LOAD, 02232LOAD, 02241LOAD.............................. 245

5-19 deBeer Method Model Pile Capacity vs. Specimen Relative
Density...................................................... 247

5-20 Butt Deflection (inches) at Failure vs. Failure Load
(kips) Using the deBeer Method of Capacity Determination..... 252

5-21 Per Cent Increase in Soil Stiffness vs. Increase in
Number of Piles.............................................. 253

5-22 Pile Penetration vs. Number of Hammer Blows, Single Pile,
DG69803171 ........................................... ...... 255

5-23 Pile Penetration vs. Number-of Hammer Blows, Single Pile,
DG69803282, DG69803181....................................... 257

5-24 Pile Penetration vs. Number of Hammer Blows, Single Pile,
DG69803251, DG69803281 .............. .......... ............... 258

5-25 Pile Penetration vs. Number of Hammer Blows, Single Pile,
DG69803291, DG69803292, DG69832182........................... 260

5-26 Butt Deflection vs. Load, Single Pile, DG69803293............ 261

5-27 Hammer Blows (100% Efficiency, 20 Gram Hammer) for 30
Scale Feet of Penetration vs. Specimen Relatfve Density...... 265


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



John Joseph Gill

August 1988

Chairman: Frank C. Townsend
Major Department: Civil Engineering

Elastic wave propagation and the practical aspects of pile driving

are discussed to develop an understanding of which aspects of pile driv-

ing must be modeled most precisely in the centrifuge. The scaling laws

are developed. Design and construction of a 5 g-ton (110 g) centrifuge

are presented. Construction details of the model pile placement device

are discussed. The placement device has the capability of driving indi-

vidual piles and can model the energy input of any single acting pile

driver. The placement device can also push individual and group piles

through the application of up to 1.5 million scale pounds. Static load

tests can be conducted with the device. All computer software necessary

for the driving, pushing, and load testing of model piles is outlined.

Discussion includes all related software for model pile calibration

(strain gage response to loading) and load test interpretation (deBeer


method). A method of strain gaging the model piles is presented. A

test series is conducted at 69.8 and 86.0 g's involving the pushing of

individual, group of four, and group of five model piles based on a 35-

foot long prototype pile (group) driven 30 feet into saturated fill.

All model tests are conducted on dry granular soil with relative density

varying between 45% and 70%. All models are load tested with several

being subjected to tensile pullout tests. Individual model piles are

driven permitting the comparison between the load-bearing capacity of

drive and pushed models.

The placement device is shown to be precise and accurate in the

measurement of pile displacement and resistance to penetration. Driven

individual model piles are found to have lower initial loading moduli

but similar ultimate capacity to those of the pushed models.

Differences are attributed to disturbance of the model during placement

of load cell on the model butt prior to load testing. Bearing capacity

of the model pile tips are not altered by scale effects; however, skin

friction on the sidewalls decreases exponentially as test g level is

increased. The model group piles demonstrate an efficiency of 1.17 with

a group efficiency of 1.0 being considered conservative for group models

placed in granular soils.


1.1 General
Piles have been used to provide for a suitable foundation for man-

made structures for over 12,000 years. Their suitability as a founda-

tion has been left to chance, sometimes inferred from empirical rela-

tionships, less often determined by some fairly scientific method

relying on known soil properties, and, least of all, determined by full-

scale on-site testing. These approaches have led to outright failure,

unacceptable performance, or overconservative design of many of the

successful pile foundations. While full scale testing may provide the

most accurate information regarding pile performance at the construction

site, it is undeniably the most expensive and time consuming method.

Additionally, only a limited number and type of tests can be performed

on the prototype.

The mechanisms of soil behavior are still subject to interpretation

as evidenced by the wide variety of pile capacity prediction techniques

available. Furthermore, the variability of soil deposits limits the

accuracy with which capacities can be predicted. The nonstandardization

of some sampling techniques and their interpretation contributes to the

uncertainty and sometimes leads to errors in the determination of the

properties on which capacity predictions are based. Lastly, there is a

lack of sufficient data regarding the performance of pile foundations

and the properties of the soil on which they are founded. The high cost

of obtaining such information and the few sites suited to provide useful

results are limiting factors (Harrison, 1983).

Geotechnical engineers rely on theoretical and empirical relation-

ships to aid in the design of pile foundations. Using these relation-

ships requires an accurate knowledge of subsoil conditions. This know-

ledge is becoming more readily available from increasingly accurate in

situ testing; however, the in situ soil properties can change dramati-

cally as a result of placement of the foundation.

Researchers have made use of prototype data, but the lack of infor-

mation has made the investigation of models attractive. Initial inves-

tigations involved pile performance using miniature models in a one-

gravity environment. These efforts have been expanded to include the

investigation of scale model pile performance in a high gravity environ-

ment. This high gravity environment is most frequently generated by the

use of a centrifuge. Recent developments in small-scale testing present

several alternatives to the problems described above. Piles as small as

1/100 the size of the prototype have successfully been tested in high

gravity environments which reproduce the same unit stresses at equiva-

lent locations in the model as are experienced by the prototype. Fur-

thermore, various in situ tests are being adapted for use in the centri-

fuge through innovative research. This combination of modeling and the

ability to characterize the soil adjacent to the model permits the

engineer to conduct parametric studies revealing more about how changing

soil properties affect the bearing capacity, deformation, and load

transfer mechanism of piles subjected to static loads.

Several significant aspects must be considered for one to attempt

strict modeling of pile performance at a reduced scale. First, scaling

laws must be developed allowing the engineer to construct a model which

will react both dynamically and statically to a scale load as would the

prototype. The scaling laws, based on similitude, are needed both for

the design of the model and interpretation of the model response. Sec-

ond, strict modeling of piles requires that a dynamic pile driver be de-

veloped to ensure the piles are placed in the same way as in the proto-

type.' This stipulation lessens the current need to infer results from

load tests on piles that have been inserted statically (pushed) rather

than driven. Lastly, a method must be developed to measure the load

transfer from the pile to the soil during placement as well as loading.

The ability to perform the aforementioned tasks will permit para-

metric studies of a variety of pile types over a wide range of soil

conditions. This is very advantageous as the soil conditions of the

prototype are quite often impossible to reproduce exactly in the model.

Furthermore, a large number of tests can be conducted at an insignifi-

cant cost when compared to prototype testing. Interaction between the

individual piles of a group and group response can also be studied.

This report documents the development and testing of the equipment

designed to accomplish the aforementioned goals. The results of all

tests conducted in association with the development of the model pile

placement device will be presented.

1.2 Objectives

The objectives of this research program are as follows:

1) To design and build a device capable of driving and load testing

model piles (groups) in-flight in the artificially high gravity

environment generated in a centrifuge. The device must also be

capable of pushing model piles (groups) into place and load testing

the pile (group) after placement.

2) To develop the computer software necessary to control the pile

placement device.

3) To instrument the model piles (groups) and develop the means of

measuring butt deflection and residual stress development in the

model subjected to static loading at design gravity level.

4) To drive or push model piles (groups) into homogeneous, dry,

granular soils and compare the differences, if any, between static

capacity, residual stress development during placement, and load

shedding to the soil during subsequent loading.

5) To discern the effects of various gravity levels on the accuracy of

the model by conducting tests on scale models at 69.8, and 86.0

gravities. The validity of the scaling laws used evaluated in this

manner and possible effects on the contribution of relative grain

size to the modeled capacity are explored.

6) To determine the feasibility of using the device to predict proto-

type pile capacity by modeling a well-instrumented pile group.

1.3 Scope of Work
The wave equation used in the development of pile stress, strain,

and particle velocity magnitudes during driving is presented. The

theoretical effects of variation of hammer and pile configurations are

explored. The practical aspects of pile driving are presented to foster

an understanding of the system parameters which must be most closely

modeled. Scaling relationships are developed permitting the determina-

tion of performance capabilities of the placement/loading device.


A complete pile placement (both driving and pushing) and loading

device and associated computer control hardware/software system are

designed, built, and tested. Data measurement, recording, and presen-

tation techniques are developed. A "gravity-level independent" means of

testing and data capture is employed to permit the use of this device

for testing of a variety of scale models.

Model piles (groups) are driven and pushed in the range of 70 to 90

gravities and subsequently loaded statically. Static resistance devel-

oped versus deflection of the butt (cap) is measured and ultimate

capacity is recorded. The validity of the scaling relationships is


1.4 Review of Previous Research

Two major areas of interest exist in the centrifugal modeling of

piles and pile groups. First, the centrifuge must be shown to be a

valid tool for use in the modeling process. Second, the technical

aspects of building, placing, and testing the models must be understood.

The earliest pile model studies involved placement and load testing of

miniature piles at one gravity. Results from these studies indicated

the need to recreate stresses at the model pile-soil interface that were

similar to the stresses experienced by the prototype. This is most

easily accomplished in a centrifuge. Initial centrifuge studies were

concerned with the placement of model piles at one gravity with subse-

quent load tests being conducted at the design test gravity level.

Several research efforts have since been conducted regarding the in-

flight placement of model piles with subsequent load tests.

Early tests conducted on miniature piles in the United States,

Whitaker (1957), Saffery and Tate (1961), and Sowers et al. (1961),

provided only qualitative results and a cursory understanding of pile-

group load factors. Model piles for these tests were typically nine to

twelve inches long, pushed into the tests specimen, and loaded incre-

mentally to failure. Results of these tests could not be directly

related to prototype performance using available scaling laws (Rocha,

1957) because the unit stresses at the tip and along the side walls of

the piles were not being reproduced. Additionally, the previously

stated irregularities inherent in soil deposits made modeling of proto-

type soils infeasible. Scott (1977) conducted tests on model piles

pushed into a silt at one gravity and laterally loaded after being

accelerated to 50-70 g's. Results were reproducible and internally

consistent in the sense that stiffer soils resulted in a model with

greater resistance to lateral displacement. However, no prototype was

available for comparison. Similitude and the use of scaling laws was

not verified. Most significantly, this research demonstrated the

feasibility of conducting load tests on miniature piles in the centri-

fuge. Hougnon (1980) demonstrated similitude by modeling the response

of individual and group tapered wooden piles subjected to axial and

lateral loads. The wooden piles, 0.2 in. in diameter, were tested at 70

g's. The models were scaled to represent the prototype wooden piles

driven 35 feet into the ground at Lock and Dam #26, near Alton, Illi-

nois. Problems associated with the development of a functional loading

device limited the applicability of the results. Furthermore, the five-

unique soil layers of the prototype were replaced in the model by a

homogeneous, uniform specimen supposedly having strength characteristics

similar to the average strength characteristics of the prototype. Pro-

blems associated with the preparation of uniform soil specimens limited
the applicability of the results; however, useful data was obtained con-

cerning the qualitative effects of pile taper and soil density on the

capacity of the model piles. Centrifugal tests investigating the axial
capacity of modeled steel cylindrical piles, Ryan (1983), United States

Department of Transportation/Federal Highway Administration (USDOT/FHWA,

1984c), and Millan (1985), refined the techniques associated with pile

placement and data retrieval. Results of those studies further esta-

blished the centrifuge as a valid tool for use in the investigation of

prototype pile capacities. Problems associated with the construction of

model piles, application of scale loads, and measurement of resulting

displacements will now be discussed.

Model piles used in the investigation of prototype response to

axial loading have progressed from relatively crude noninstrumented

cylindrical tubes, Whitaker (1957), and Saffery and Tate (1961), to

fairly complex machined aluminum tubes of scale proportions and instru-

mented to measure stresses and strains at various depths while being

statically loaded in the centrifuge, (USDOT/FHWA, 1984a, 1984b, 1984c;

Millan, 1985). That progression will now be outlined.

Although Scott (1977) has been credited with conducting some of the
earliest centrifugal pile capacity tests in the United States, Hougnon

(1980) is among the first to attempt verification of scaling laws.

Models were miniature (1/70th scale) replicas of tapered wooden piles

and were constructed of wood with strength properties similar to those

of the prototype. The models were not individually instrumented to

measure stresses developed during loading. Rather, the pile was loaded

using a hydraulic cylinder with displacement being measured by a linear

variable differential transformer (LVDT). This technique provided

information regarding butt deflection versus applied load but difficul-

ties in modeling the prototype soil conditions made comparison of the

model response with the prototype inappropriate.

Harrison (1983) pioneered a technique which subsequently became the

standard for the manufacture of model piles. His method consisted of

removing half of an aluminum tube exposing the inside which then per-

mitted the placement of strain gages along the shaft. Two halves with

opposing strain gages were then glued together forming an instrumented

model pile. Harrison tested models placed in granular soil at one gra-

vity and loaded after being accelerated to 50 g's, concluding the pre-

sence of the seam along the length of the model influenced both axial

and lateral response to loading. Significant departure from the proto-

type response resulted from the model's splitting during placement (by

hydraulic cylinder) and loading. Harrison cited the smallness of the

available gage sites on the inside of the model pile halves as the

limiting factor in the use of the split-tube method of strain gage

application. Furthermore, he concluded the presence of the epoxied

seams, where the model halves were joined, influenced the model's

response to lateral loading to a greater extent than the response to

axial loading. Ryan (1983), Ko et al. (1984), and the USDOT/FHWA

(1984c), and Millan (1985) conducted further tests making models using

the split-tube technique.

Ryan (1983) conducted the preliminary work for Millan at the Uni-

versity of Florida. Ryan's work involved the construction, calibration,

and preliminary testing of 0.25 inch outside diameter, five-inch long

aluminum tubes which had five pairs of strain gages installed. The

models were not scaled down from a chosen prototype. Rather, the one

size pile was to be tested at 30, 45, and 60 g's. Significant diffi-

culty was encountered in the construction of the instrumented model

piles. Strain gages which appeared to be properly installed did not

provide accurate and reproducible readings. The epoxy used to glue the

halves together was found to stiffen with time precluding the determina-

tion of a repeatable calibration curve. Rather, the piles had to be

calibrated before each test to determine model response to loading.

Lastly, the calibration of such models at one gravity with the model

being unsupported over the five-inch length disclosed the susceptibility

of the model to be influenced by stress concentrations at the butt and

tip. Some reduction in the variation in strain during static loading

was accomplished by the introduction of a load-eccentricity-reducing

support at the tip and butt during calibration. Strain gage readings

during in-flight testing were found to be adversely affected by the need

to transmit the strain gage response through the slip rings prior to

bridge circuit completion. Additionally, vibration and strain of the

gage leads was found to influence the accuracy of the readings obtained.

Ko et al, (1984) achieved greater success with the split tube

technique and tested model piles at 50 and 70 g's in granular soil.

Individual piles were pushed in either at one gravity or the appropriate

test gravity level and results of load tests conducted for both at the

test g-level compared. The relative accuracy of the results indicated

the need to both place and load test the model pile at the design test

g-level. Subsequent tests involved the insertion and loading of an

individual pile in one continuous flight of the centrifuge and comparing

the load test results with a model pile inserted during one flight and

loaded after stopping and restarting the centrifuge. The results indi-

cated that interrupting the insertion and loading cycle by stopping the

centrifuge had no significant effect on pile capacity. Ko concluded

that in granular soils similar to those tested, it is important to con-

duct both insertion and load testing at the appropriate test gravity

level to ensure that self-weight soil stresses on the model are geome-

trically similar to those developed on the prototype. Results of this

study indicated the potential for studying pile groups by sequential

insertion of individual piles and load testing after the piles had been


The FHWA tests were conducted in granular soils using instrumented

piles at 70 g's (split-tube method of strain gage installation) with

several noninstrumented piles being tested at 50 and 100 g's to verify

modeling of models. Specially manufactured miniature "coupons" were

inserted between the pile and hydraulic cylinder to measure the force

required to insert and load the piles. Deflection measurements were

taken by an LVDT attached directly to the pile butt. Strain gages were

placed in the tip and at the butt of the pile permitting separation of

tip capacity from the total force required to push and load the pile.

Side wall frictional forces were inferred by subtracting the measured

tip capacity from the total force required to push the pile. The strain

gages placed at uniform intervals along the shaft indicated the side

wall unit friction increased only slightly, but uniformly as the model

pile depth of penetration increased. Between 70% and 95% of the total

capacity was derived from the tip for the individual piles tested

(USDOT/FHWA, 1984a). Tests by the FHWA involving the insertion of

models at test gravity levels, stopping the centrifuge, and restarting

before load testing, supported the findings of Ko et al. It was con-

cluded that stopping the centrifuge had no effect on the subsequent

performance of the pile embedded in sand when the load test was con-

ducted at the appropriate g-level.

Millan (1985), using piles similar to those tested by Ryan, con-

ducted load tests in granular soils with the same size pile being tested

at 30, 45, and 60 g's. The specimens were created by suspending the

model piles within the centrifuge bucket and raining soil into the con-

tainer. This was done to avoid the potential damage to the model resul-

ting from pushing the pile in at the test g-level. Millan scaled the

results at the three test g-levels to prototype capacity concluding

modeling of models was a valid means of verifying the scaling relation-

ships. Furthermore, the scaling relationships appeared to be valid as

pile capacities were within 20% of the predicted prototype capacity

using the capacity prediction method of Meyerhof (1976). Differences

between the actual and predicted capacities were attributed to the

instrumentation shortcomings outlined by Ryan and the method of

placement of the model.

Harrison (1983), Ko et al. (1984), and the USDOT/FHWA (1984c), and

Millan (1985) all recognized the sensitivity of ultimate model pile

capacity to changes in the relative density of the granular soil

specimen. The increase in bearing capacity resulting from an increase

in the relative density of the specimen was noted regardless of the

method of placement of the pile in the soil as long as the model

capacity was measured at the appropriate test gravity level. Ryan

(1983) and Millan (1985) mention the impracticality of creating a

specimen with the exact values of prototype relative density and

coefficient of friction as limiting factors in the use of centrifugal

models for the prediction of specific prototype pile (group) capaci-

ties. The use of parametric studies over a suitable range of soil

conditions is suggested. The variability of naturally occurring soil

deposits further supports the need for parametric studies.

The insensitivity of ultimate model pile capacities to temporary

pauses in the rotation of the centrifuge makes the study of group piles

possible by the progressive insertion of individual piles in the appro-

priate group pattern with subsequent capping prior to load testing.

This technique was used by Harrison (1983), Ko et al. (1984) and the

USDOT/FHWA (1984c). Each recognized the progressive increase in

resistance to penetration of individual piles due to the presence of the

pile(s) which had already been placed. After all the piles had been

placed, the groups were capped by bolting a multipiece cap in place

which rested on and clamped around the individual butts. The inability

to model the proper connection between butt and cap has been perceived

to alter the results (Harrison, 1983). Tightening of the cap pieces

around the piles inevitably led to the exertion of a lateral load on

some or all of the piles. This resulted in improper or inefficient

transfer of axial load to the piles. Pile groups capped in this manner

were pushed between 6 and 12 scale inches further into the specimen

prior to load testing in order that proper seating could be assumed.

The desired reduction in lateral forces has not been verified. Each

pile in the groups tested was inserted through a spacing template which

rested on the specimen surface, the intent being to ensure the piles

were precisely spaced in all tests of the same group configuration.

Excavation of the soil from around the model group subsequent to load

testing was used to provide a qualitative input to the data obtained.

Ryan (1983) and Millan (1985) attempted to reduce the development

of lateral forces on the individual piles of a group by manufacturing a

one-piece cap and attaching the individual piles to the base of the cap.

As mentioned previously, specimens were then created by raining soil

around the suspended cap and piles. Millan reported relatively low

group efficiencies (less than 1.0) while the USDOT/FHWA, (1984a)

reported higher than expected group efficiencies (greater than 1.0).

The author suggests the lower efficiencies reported by Flillan are due to

the method of placement of the pile group and the higher efficiencies of

the FHWA may be attributed to the incomplete erasure of the lateral

stresses induced by placement of the nultipiece cap. Millan (1985)

reported scouring directly under the cap due to wind turbulence during

testing. The effect of the removal of soil from around the cap base was

not determined.

The most significant remaining variable in the study of pile capac-
ity using the centrifuge as a modeling tool is the degree of realism

achieved by inserting the pile rather than driving as is normally done

with the prototype. Field piles are driven by a variety of weights

falling a specified distance to impart a certain impact energy on the

pile butt. A wide variation is found in the means of imparting this

energy, for example, falling weights, single- and double-acting diesel

hammers, etc.

Researchers suggest the difference in method of model pile place-

ment may have a significant effect on the ultimate capacity of the

model. The variation has been attributed to the creation of stresses at

the pile-soil interface which are dependent on the method of placement.

Ko et al. (1984, p.167) cited the relative density and coefficient

of friction of the soil specimen "in the zone of disturbance of the

pile" as the controlling factors determining ultimate model pile capa-

city. Ryan (1983) and Millan (1985) noted that the soil properties at

the pile-soil interface may not bear any resemblance to the original

properties after placement of the pile and the degree of disturbance due

to the pile placement method may alter the ultimate capacity. The

difference between insertion of a model pile either prior to or during

acceleration of the specimen and driving of the model under the

influence of centrifugal acceleration is speculated to affect the type

and magnitude of soil disturbance and lateral stresses around the pile.

The installation of the pile by the steady jacking force
likely created a zone of disturbed soil adjacent to the
pile. The disturbance produced by the pile installation
probably affected the structure of the sand for a dis-
tance of approximately one diameter around the pile cir-
cumference as shown by Vesic (1977) for dense sand. The
method of pile installation, i.e., a steady jacking
force versus dynamic repeated blows, could affect the
nature and extent of soil disturbance.

(USDOT/FHWA, 1984c, p. 66)

Researchers agree the effect on pile behavior from different meth-

ods of placement should be investigated (Ko et al., 1984; USDOT/FHWA,

1984a, 1984c; Millan, 1985). Such an investigation requires the deve-

lopment of a device to dynamically drive the pile in the centrifuge

until the desired penetration is achieved. Model load tests equivalent

to those performed on prototype individual and groups of model piles

could then be interpreted more closely in accordance with established

methods. Parametric studies of model piles driven in-flight may prove


to be a significant step in the determination of prototype pile (group)

capacities by modeling. Additionally, the ability to drive model piles

in-flight will provide a greater understanding of the sensitivity of

pile performance to changes in soil properties at the pile-soil inter-



2.1 Introduction

The transfer of energy from hammer to pile during driving is a

complex occurrence involving elastic and inelastic deformations, energy

losses, nonhomogeneity in the soil medium, deformation and deformation

rate-dependent soil response, and transient phenomena, including soil

consolidation and dissipation of pore pressure (Holloway, 1975). Devel-

oping a mathematical model which incorporates all of the important

variables of the driving process is extremely complex as evidenced by

the computer programs which model wave propagation. Extending that

model to predict the capacity of the driven pile introduces yet another

order of complication. It is necessary to understand the concepts of

how individual events occur before an investigation of the complete

process can be conducted. Simplifying assumptions are necessary even if

the process is broken down into smaller events. This chapter will

present wave mechanics theory as it applies to impact driving, practical

considerations of that theory concerning the impact driving of piles,

and a discussion of how those forces result in the development of

residual stresses during driving.

2.2 Pile Placement by Impact Driving
2.2.1 Elastic Wave Propagation in Solids

The initial assumption regarding the transfer of energy from the

hammer to the pile is that only one dimension (length) be considered in

the analysis. This assumption simplifies the proposed mechanism of

energy transfer in that energy to the pile is transmitted in the form of

a planar wave. In other words, the planar cross sections of the hammer

and pile remain planar during the transmission of the strain pulse and

the resulting stress over the section is uniform. These assumptions

permit the simplest of solutions, commonly referred to as the one-

dimensional wave equation, (Kolsky, 1963; Richart et al., 1970;

Palacios, 1977). That equation is expressed by the following partial

differential equation:

d2u 2 d2 Eq. 2-1
Sc Eq. 2-1
dt2 dx2


u = displacement in the direction of the wave front

t = time
c = /- = longitudinal wave propagation velocity

x = distance along the rod


E = Young's Modulus of elasticity

P = Y/g = mass density

y = unit weight

g = acceleration due to gravity

The assumption that lateral inertial forces are negligible in the

derivation of the wave equation implies the longitudinal wave must be

long with respect to the cross-sectional dimension of the transmitting

medium. Kolsky (1963) showed the wave equation to be valid when the

wave length is at least five times the diameter of the medium (rod)

through which the wave is propagating. This condition is always met in

the practical situation of driving piles (Krapps, 1977).

The solution of the differential equation has the following form:

u = f(x + ct) + h(x ct) Eq. 2-2

The letters f and h represent arbitrary functions such as sin w,

ew, w" etc., where w is either (x + ct) or (x ct).

The general form of the arguments of Equation 2-2 indicates the x-t

plane is divided into regions of constant compression and velocity by

disturbance lines of constant slope. The slope of these lines is

defined as the characteristics. The left-hand term in the above equa-

tion (f(x + ct)) represents a wave traveling in the negative x direction

with velocity c. The remaining term denotes the wave traveling in the

positive x direction with velocity c as shown below.

Let f(x + ct) = 0


u = f(x ct)

Eq. 2-3

The passage of a nonattenuating elastic wave is depicted in Figure
2-1. Since u = s when t = ti and x = xl, and also when t = t2 and

x = x2, the rate of wave propagation with respect to the x axis is as

shown below.

c = (X2 Xl)/(t2 t1) Eq. 2-4


s = f(x1 ct ) = f(x2 ct2) Eq. 2-5

then (x1 ct ) = (x2 ct2) Eq. 2-6

With the understanding that t2 > tI and c is constant, x2 must be

greater than xi indicating the term h(x ct) refers to the wave

traveling in the positive x direction. A similar manipulation of the
left-hand term of Equation 2-2 indicates that term refers to the wave

traveling in the reverse direction.

When considered independently, each term represents a valid solu-
tion to the wave equation. When the terms are used together, a valid
solution is still obtained, indicating Equation 2-2 is a simple partial

linear differential equation. The significance is that the separate

effects of the two solutions may be added together at any instant to
determine the net effect of the passage of the waves.

Kolsky (1963) and Richart et al. (1970) have shown that the speed,
c, at which a stress wave propagates in a medium is dependent on the

material properties of the medium, namely, the Young's modulus of elas-

ticity and density of the medium.

t = t2

0 x2

Figure 2-1

Propagation of an Elastic Wave

c = /P Eq. 2-7

Several other equally important wave parameters are dependent on

the magnitude and type of input wave. Particle velocity and stress are

two such parameters which are directly proportional to one another.

Particle velocity (Vp) is the partial derivative of displacement with

respect to time. In a compression wave, the particle velocity and wave

propagation are in the same direction. In tension waves, the particle

velocity is opposite the direction of wave propagation. The particle

velocity is always less than c in the propagation of elastic waves.

Krapps (1977) and Richart et al. (1970) have shown the particle velocity

to be proportional to longitudinal wave speed and stress, and inversely

proportional to Young's modulus of elasticity as follows:

ac a
V Eq. 2-8

Likewise, stress can be determined if the particle velocity is


a = pc V Eq. 2-9

Since dynamic force (F) equals the stress times the cross-sectional

area of the medium,

F = Vp = pcA V Eq. 2-10
cp p

The product pc is referred to as the "characteristic impedance"

(Zo) and is dependent on the material properties of the medium. A

scalar value of impedance (Z) results when the characteristic impedance

is multiplied by the cross-sectional area of the medium. The impedance

of each component in the pile driving system affects the transfer of

energy from the hammer to the pile and thus the total amount of energy

available to the pile to be used for penetration. System impedance will

be discussed following a brief summary of boundary condition limita-


The preceding presentation has been limited to the propagation of

elastic waves in a uniform, geometrically regular medium of infinite

length. These assumptions were necessary to simplify the solution of

the wave propagation equation and develop a general understanding of

this form of energy transfer. The necessity of driving piles of finite

length and using system components of complex geometry requires that the

boundary conditions'and their effect on the dynamic system also be

understood. The influence of end conditions on the magnitude and stress

sense of the reflected wave will be considered before the effects of

system geometry.

The elastic wave discussed in previous examples propagates at a

constant velocity in a freely suspended geometrically uniform rod

without changing shape. Assume the wave reaches the boundary of the

medium (i.e., the end of a rod) and is reflected. If the end of the rod

is unsupported and free to vibrate in the direction of the propagating

wave, the incident stress wave is reflected from the free end with the

same magnitude but having the opposite stress sense. A compression wave

reflects as a tension wave and vice versa. The free end implies a zero

stress state at the boundary as shown in Figure 2-2(a). The opposite

occurs when the wave impinges on a fixed boundary (Figure 2-2(b)). In

this situation, the stress wave is reflected as a wave of equal magni-

tude and stress sense. This causes the stress wave to double in magni-

tude, either compressive or tensile, at the fixed boundary. Most prac-
tical applications of these two limiting boundary conditions require the

determination of intermediate boundary conditions based on empirical


2.2.2 Hammer-Pile Impact

The impedance of two bodies plays a significant role in the effi-

ciency of the transfer of energy as one impacts the other. Consider the

hammer-pile system depicted in Figure 2-3. The variables A and a refer

to the cross-sectional areas of the hammer and pile, L and 1 to the

length of the hammer and pile, and V to the impact velocity of the

hammer. The contacting ends of the hammer and pile are assumed flat and

contact is made across the full cross-sectional area of the rod during

impact. Two initial assumptions are needed for this generalization to

be valid. The force in the hammer equals the force in the pile. Like-

wise, the velocities of the contacting hammer and pile faces are equal

while the two are in contact. Let ah denote the stress in the hammer

and a the pile stress. Regarding the initial assumption that the

forces in the hammer and pile are equal, the relationship between the

impedance of each and the stress in each is readily determined. Using

Equation 2-10,

Force = Stress (Area)

Eq. 2-11


Ix-L -a



a) Reflection from Stress-Free End Rods




b) Reflection From Fixed End Rods

Figure 2-2

Idealized Reflections of an Elastic Wave.
a) Reflection from Stress-Free End Rods;
b) Reflection from Fixed End Rods

a +
\ l-




Figure 2-3 Idealized Hammer and Pile Before Impact

pVpo P = PhchVhoAh Eq. 2-12

From Equation 2-10, the impedance of each component is pcA. Thus,

ZVp Z Vho Eq. 2-13
p po h ho

The quantity termed the characteristic impedance ratio (r) is

derived by dividing the characteristic impedance of the struck body by

that of the striking body as presented below.

Z p c A
r = - = Eq. 2-14
Zh h ch Ah

The characteristic impedance ratio is simply a function of the

areas of the hammer and pile if the two are made of the same material.

Consider the second assumption which requires the velocities of the

striking end of the hammer and the struck end of the pile to be equal.

Let V be the hammer velocity. Thus,

V Vho = Vp Eq. 2-15


Vho = the velocity with which the hammer face particles

are compressed backwards relative to the unstrained

portion of the hammer yet to recognize that impact

has occurred. The hammer velocity (V) is uniform.


Vpo = the velocity with which the pile butt particles

are compressed relative to the unstrained portion

of the pile.

Palacios (1977) solves Equations 2-13, 2-14, and 2-15 for Vho and

Vpo in terms of the hammer velocity and impedance ratio.

Vho = ) Eq. 2-16

Vo= ( r) Eq. 2-17

Palacios continues his development of the impact based on the

findings of Fairhurst (1961):

As impact continues, the strained region in both mem-
bers extends away from the interface at the propaga-
tion velocity c so that, at a given instant, the
strain wave has covered the same length of hammer and
rod (Figure [2-4b]). The particle velocity within
this region is constant at values Vho and Vp respec-
tively. Upon reaching the free end of the hammer af-
ter time t = L/c, the compression wave will be reflec-
ted as an equal tension pulse which combines with and
cancels the outgoing compression wave, giving the par-
ticles a total velocity of 2Vho away from the inter-
Thus, as the tension wave returns, the hammer is pro-
gressively released from strain, such that the abso-
lute spatial velocity (V ) of the unstrained portion
(Figure [2-4d]) is given by:

V = V 2Vho [Eq. 2-18]

At the instant t = 2L/c that the reflected wave
reaches the hammer [pile] interface, no strain exists
in the hammer (Figure [2-4e]) and it is traveling with
uniform velocity, V. The hammer velocity at the
interface is thus abruptly changed at time t = 2L/c as
V changes suddenly to V1.

(Palacios, 1977, p. 46)


W Er



2~p J~pb'
L+LI --f



L 2L
SI (e)

2L+L3 --
(1) After Fairhurst (1961)

Figure 2-4 Mechanism of Impact of a Cylindrical Hammer
on a Finite Cylindrical Rod. After Palacios (1977)







The change in hammer velocity from V to V1 implies new boundary

conditions and the process is repeated using the new boundary conditions

for each complete cycle of compression and reflected tension wave tra-

versing the hammer. Palacios concludes


V pq ( r)n

2nL 2(n + 1) L
c c

Note that Vp


Eq. 2-19

Eq. 2-20

Eq. 2-21


1 r0n
= apo (i T- r

Note also that ab


When the hammer and pile are made of the same material and

same cross-sectional area, r = 1 and

abo = PcV

Eq. 2-23

have the

Eq. 2-24

The energy is thus transferred from the hammer to the pile in a

stepped waveform with the hammer decelerating incrementally and the pile

butt accelerating likewise. The hammer and butt separate when the


Eq. 2-22

velocity of the butt exceeds that of the hammer. The energy transmitted

to the pile is used productively for penetration and/or dissipated as

heat, vibration, etc. This separation typically occurs when the wave

reflected from the pile-soil interface returns to the contact face.

Palacios concluded the theoretical maximum stress is uniform for the

duration of contact when the impedance ratio is 1.0. An incremental

decrease in the impedance ratio (with hammer mass and drop height held

constant) results in increasingly higher theoretical maximum stresses

acting over increasingly smaller time periods as shown in Figure 2-5.

The total time required for the complete transfer of hammer energy

increases as the impedance ratio decreases with total time required

being in excess of twice that which is necessary when the impedance

ratio equals 1.0.

2.2.3 Hammer-Pile-Soil Interaction

The proceeding discussion has been concerned with the propagation

of elastic waves which were not altered in their passage through a uni-

form medium. This situation is possible in an experimental setup but is

too limited for accurate representation of the transmission of energy

from hammer to pile. Several factors complicate the simple model pre-

sented thus far. Dispersion, interaction between the hammer, cushion,

and pile, and soil-pile interaction all affect the transmission of


Dispersion results from two frequency dependent aspects of wave

propagation. Both wave velocity and amplitude attenuation occur more

rapidly for higher frequency waveforms. A waveform may appear uniform

over a short time; however, monitoring the progression over a suitable

distance reveals the composite waveform is made up of various high and


Time, 2L/c

. 44L


Sr.25 .5LJ r .0625
-- --t i =

Figure 2-5

Effect of Impedance Ratio on the Theoretical Wave Shape
for Same Hammer Mass and Drop. After Palacios (1977)

- .0625

r .25
--___ "

SI 1 .44

-- o--! ~ ---

S-r -l
___ "' .a

L-- ..

r-1 L





low frequency components. The high frequency components proceed rela-

tively quickly through the medium followed by the waveforms of succeed-

ingly lower frequency components. It is also apparent that the higher

frequency waveforms attenuate and slow down more quickly than the lower

frequency components. The low frequency waveforms never "catch up."

Rather, the high frequency portions dissipate at some point while the

low frequency portion continues to propagate for a relatively large

distance. Thus, a waveform changes during propagation and has a dif-

ferent energy content dependent on the instant at which the wave is

observed. Dispersion is not usually a problem in the practical appli-

cation of driving piles (Palacios, 1977). However, it is an important

consideration if waveform analysis is to be performed. After several

cycles of reflection of the initial wave from the ends of the pile, the

now separate high and low frequency portions interfere with one an-

other's passage and can dramatically alter the predicted energy content.

The aspect which has more practical bearing on the transmission of

available energy from the hammer to the pile is interaction between the

hammer, cap/cushion, and pile itself. When an elastic wave impinges on

a boundary between two media, reflection and refraction occur. In the

most general case, two distinct waves are generated each time an indivi-

dual wave passes the boundary; one reflected and one refracted (Kolsky,

1963). As the two new waves can have only as much energy as the initial

wave, the refracted wave (moving in the same general direction as the

initial wave) has lost the amount of energy contained in the reflected

wave. Reflection, refraction, and energy loss occur at every interface

between the hammer face and the pile and can be significant.

The interaction between the pile and soil changes with depth of

penetration, soil conditions, and the remaining energy which is avail-

able to be productively used for penetration as the stress wave pro-

gresses. Attempts to understand this aspect and its contribution to the

development of residual stress and static capacity have resulted in

several mathematical models. The most frequently used models are the

finite difference and finite element methods.

The finite difference method discretizes the physical problem into

small segments as shown in Figure 2-6 (Smith, 1960; Holloway, 1975).

The pile is approximated by a series of short longitudinal segments.

Each segment is a discrete mass connected by springs of known deforma-

tion constants which approximate similar deformation characteristics

between the mass centroids of the segments and the equivalent portion of

the pile. This model, developed by Smith (1960), can be solved by

integrating element displacement and velocity with respect to time.

Total displacement can be determined by integration over small time

increments from known initial conditions. This method has been modeled

on digital computers with moderate success. Particular care must be

taken in the selection of the time increment of integration. Too small

an interval results in an inordinately large number of calculations'

being performed and possible magnification of small numerical errors in

the model. A large time increment could permit the bypass of an element

and subsequent instability in the solution (Holloway, 1975; Smith,

1960). The time increment is sometimes determined simply by dividing

the length of the pile segment by the stress wave velocity (E/p) in a

freely suspended rod made of the same material as the pile. Inelasti-

city of the system components, most notably the pile cushion, has a


Capblock -- K1
Cap ---^ tW2
K3 R3
K4 t R4
K5 > R5
K6 R6
K7 > R7
1 w8
K8 t R8
K9 t R9
K10 R1i
K11 ) t R11
K12 R12
r W13

Point Resistance

Actual Representation

Figure 2-6 Finite Difference Pile Representation.
Smith's Lumped Parameter Representation.
After Smith (1960).


profound effect on the transmission of energy in the prototype and

model. Permanent deformation of the cushion introduces energy losses

which reduce the efficiency of the transfer. Soil strength and resis-

tance were highly simplified in this model as less was known about in

situ conditions contributing to pile strength.

The finite element method was the successor to the finite differ-

ence solution. This method relies on an axisymmetric idealization of

the three-dimensional pile-soil model. The elements suggested by Desai

and Abel (1972), project as wedge segments when the model pile represen-

tation (Figure 2-7) is viewed from above. The first analysis of pile

soil interaction using the finite element method was conducted by

Ellison (1969). This method has since become the subject of several

texts (Zienkiewicz, 1971; Desai and Abel, 1972). Deformation is con-

centrated at the nodes and is equal for the pile and soil at the-inter-

face. The pile material strength parameters are usually well known and

the soil parameters conform to a tri-linear approximation based on

either lab tests or in situ test results. The soil strength is mobi-

lized until it exceeds the Mohr-Coulomb strength parameters at which

time it is considered to have developed ultimate strength. The soil

continues to deform if the stress is increased; however, no additional

strength is mobilized. The deformation "envelope" simply grows larger.

This method provides acceptable results in static loading predictions as

the relative motion between the pile and soil is small. It is less

applicable for analysis of dynamic situations because of the difficulty

of modeling the interaction (and differential displacement) between the

pile and soil. A knowledge of the propagation mode of both compression

and shear waves is very helpful in the initial determination of the




Figure 2-7 Finite Element Mesh Pile Representation

nodes and elements, yet this is the type of information the finite

element investigation is supposed to provide. Holloway (1975) correctly

predicted this method would continue to receive emphasis as a research

tool for quite some time.

2.3 Practical Aspects of Pile Driving

The practical considerations of prototype pile installation will be

presented as a guide for determining the aspects which must be modeled

most accurately. Pile installation considerations include the inter-

action of the hammer, cushion, and cap (or helmet), as equally important

members of the system needed to drive the pile. Pile forces experienced

during driving will be discussed. Errors in the placement and driving

of piles affecting capacity will be detailed.

2.3.1 Hammer-Cushion-Cap Interaction

Similarities between the energy transmission from the hammer to the

Standard Penetration Test (SPT) sampler and the hammer to the pile

permit several conclusions to be drawn from research conducted on SPT

rigs. The need to have the relatively simple SPT rigs energy calibrated

(Schmertmann, 1977) prior to interpreting and comparing results under-

scores the necessity of viewing the hammer, cushion, and pile cap system

(Figure 2-8) as a whole when determining the amount of energy being

transmitted to the pile.

The ram weight and impact velocities are generally
the most important variables with respect to pile
penetration for a given pile-soil system. Heavier
rams generally give more penetration than lighter
rams with the same kinetic energy at impact. Note
that the heavier ram has more momentum at impact
than the lighter ram in this case and the trans-
mitted stress wave generally has a longer wave-
length. Heavier rams are generally more efficient

Striker Plate

Pile Cushion -
(Concrete Piles

Prototype Model

Guide Rod


Definition of Pile Cap Terminology

Figure 2-8

with respect to utilization of kinetic energy than
lighter rams.

(Krapps, 1977, p. 49)

The dramatic effect of these variables on energy transmission are under-

scored by the efficiency of some pile drivers, which can range from 30

to 95% (American Society of Civil Engineers (ASCE) Committee on Deep

Foundations, 1984). The hammer weight and lift height are simply a

measure of the energy available to be transmitted to the pile during an

individual impact.

Energy transfer from hammer to pile is less dependent on the degree

of tilt (or list) of the driver than if the driving components are mis-

aligned. Palacios (1977) determined that the concentrically aligned

hammer and rods could be tilted as much as three degrees before causing

a decrease in transmitted energy.

Full understanding of the hammer, cushion, and cap interaction is

vitally necessary prior to modeling of the driving process. The charac-

teristics of these components should be chosen to satisfy the following

two criteria (Peck, Hanson, and Thornburn, 1974). The components must

be able to transfer the amount of peak driving force at least equal to

the desired ultimate capacity of the pile being driven. Unless this

criterion is satisfied, the pile cannot penetrate far enough to develop

the desired capacity. Refusal will occur first. The components must

also transmit as much of the available energy from the hammer as pos-

sible. This criterion is more flexible than the first. It is important

only for economy in driving. The lift height or hammer weight can, and

must, be modified if their standard configuration would overstress the

pile. Stresses in the pile and determination of the amount of energy

transferred from an impact can be directly measured by the appropriate

instrumentation (i.e., strain gages and accelerometers), however, this

method is not used on a widespread basis. Computerized wave equation

analysis is more frequently used to find the best combination of hammer,

cushion, and cap for a given hammer and pile combination (ASCE Committee

on Deep Foundations, 1984). These methods suffer from the need to make

assumptions regarding the chosen components.

The cap distributes the hammer blow to the butt of the pile and can

also serve to hold the butt in place during the initial stages of

driving. A close fit should be maintained between the cap and butt to

prevent buckling of the pile or bulging of the butt itself. The bearing

surface of the butt or cap should be machined to assure proper fit. The

cap should be sufficiently massive to provide efficient transfer of

energy to the butt. An insufficiently massive cap may separate pre-

maturely from the butt effectively stopping the flow of energy from the


Historically, hammer cushions were made of hardwoods cut to fit

snugly within the cap, but they are now usually made of aluminum or

micarta. The relatively soft hardwoods transmit an initial compressive

wave of lower frequency and magnitude to the pile for a given hammer and

cap combination. Additionally, the hardwoods have revealed their ten-

dency to catch on fire as the amount of energy transferred from the

hammer has increased. The coefficient of restitution (COR) of hardwood

is only 0.5. Aluminum and micarta have the advantage of a higher COR

(0.8), more nearly linear elastic properties, and greater and more con-

sistent energy transmission characteristics than the hardwoods. The

significant improvement in the use of aluminum and micarta lies in their

more predictable elastic properties and reduction of energy losses.

Excessively high compressive and tensile forces in the pile are

controlled by proper selection of these components. Pile penetration

per impact may be controlled somewhat by selection of cushions with dif-

ferent elastic properties.

2.3.2 Pile Forces Developed During Driving and Loading

Tubular steel shell piles were the only type of piles considered in

this research. This type of pile can develop high capacities, but the

limiting criteria for pile selection is usually the pile's ability to

withstand the significant stresses experienced during driving. The pile

must be of adequate cross-sectional area to provide the necessary drive-

ability characteristics and thus be able to achieve proper penetration.

The driving force a pile can withstand is dictated by the impedance of

the pile (refer to Section 2.2.1), the limitation being that the pile

material should remain within the boundaries of elastic deformations

during driving. Increasing the impedance, and thus the driveability,

can be accomplished by changing the strength characteristics of the pile

material, increasing the cross-sectional area, or both. A reasonable

balance must be maintained between the resulting pile size and available

driver. The increase in impedance will provide the added benefit of

greater pile capacity given adequate soil conditions.

The driving stresses in a pile are usually relatively insensitive

to driving resistance (dependent on soil properties) for a given hammer,

cushion, and cap. The stresses then would be equal in a given pile

regardless of the soil type in which it is driven. The variation would

be in the amount of penetration per impact. The amount of tip resis-

tance experienced by the pile also affects the driving stresses

developed. Excessive resistance at the tip contributes to crushing of

the tip if driving is continued. If the pile tip rests in a soft soil,

a tensile wave is reflected from the free end which may exceed the

tensile strength of the pile. Fracture can result. This is usually a

concern only in concrete piles. Thus, forces developed at the tip

depend on the stress transmission characteristics of the hammer and cap

assembly, the impedance of the pile, and the general pattern of distri-

bution of soil resistance at the tip and along the side. A better

understanding of the relative contributions of the tip and side resis-

tances will provide valuable information regarding ultimate pile


The aforementioned stresses of both driving and static loading can

be measured by a variety of methods. Displacement of the tip and side

walls can be measured directly or inferred from the readings of strain

gages appropriately located on the interior of the pile. The strain

gages will also provide information concerning the unit pressures and

total loads felt by the walls and tip. Tell-tales can be mounted along

the inside wall of the pile to monitor movement of the tip and walls

during static loading. An inclinometer will permit determination of the

accuracy of placement of the tip as well as deviations from the intended

placement along the pile length.

Strain gages are the most frequently used instruments to measure
stress distributions along the pile shaft. They are easily mounted on

the interior of the pile, durable if suitably protected, easy to inter-

pret, and accurate. Strain gages measure strain directly; however

displacement, unit stress, and thus total load can be inferred from the

readings. Properly attached gages provide information regarding the

passage of elastic compression and tension waves during the driving

process. Recording the variation in these waves at regular intervals

during driving can reveal where the input driving energy is spent during

penetration and how much energy is stored in the pile as residual

stress. Likewise, recording the magnitude of static stresses developed

in the pile at regular intervals during driving provides a record of the

development of residual stresses. Continued recording of the pile

stresses during loading aids in determination of the load shedding

characteristics of the pile-soil system and the mechanism by which the

soil mobilizes its strength to support the pile. Measuring the residual

stresses over time will reveal how the residual stresses are redistri-

buted either due to changes in the soil (movement of the water table or

changes in lateral effective stresses) or movement of the pile itself.

Tell-tales, or indicator rods with one end attached to specific

points along the interior wall and the other brought to the surface,

directly measure the displacement of the pile at the attached point.

Movement of the pile relative to the butt is measured. Total displace-

ment of a pile section can be determined if the movement of the butt is

measured from a separate reference. The tell-tales can be used in

conjunction with strain gages, as verification of the readings of one

another, or independently. The tell-tales cannot give information

concerning the dynamic response of the pile.

The tip load can be inferred from placement of strain gages in

close proximity to the tip or measured directly using a specially

installed load cell. A ring of strain gages at the tip will reveal the

tip load if the elastic modulus of the pile material and tip dimensions

are known. Measurement of the tip pressure by a load cell dictates

separation of the tip from the pile walls to ensure the load cell is

subjected entirely and only to the load at the tip. This determination

requires a specially manufactured pile and extra precaution to ensure

the integrity of the tip is maintained during driving.

Inclinometers provide information regarding the truenesss" of the

pile with regard to its intended placement. Typically, the channel or

track for the inclinometer is placed in the pile during manufacture and

the inclinometer readings taken after driving. Such measurements reveal

the degree of tilt of the pile either from vertical or with respect to

the intended degree of slope of a batter pile. Furthermore, departure

of the pile tip from its desired position can be determined if a com-

plete record of the pile inclination is obtained. Such information is

important especially if there is a question of exceeding the bearing

capacity of the soil at the tip of an end bearing pile.

The effects of residual stress on static capacity became important

to researchers only ten years ago (Davisson, 1978). These effects had

not been considered in pile load capability estimates and may have

resulted in significantly higher actual tip and friction stresses.-

Neglecting these effects may also have contributed to unknowingly uncon-

servative designs. Early attempts (1960-65) to investigate the develop-

ment of-residual stresses and their contribution to static capacity

indicated unit tip bearing capacity and skin friction increased with

depth until some critical depth was attained (ten to twenty pile dia-

meters). Further penetration did not increase either unit stress

capacity (Vesic, 1970). Subsequent field testing with 16- to 18-inch

diameter piles driven in sand indicated the tip and side wall unit

capacities increased linearly until a penetration of ten diameters had

been achieved. The unit capacities then became constant after tip

penetration of twenty diameters with the zone between ten and twenty

diameters being a zone of smooth transition between the unit capacities

(Vesic, 1970). The three studies previously mentioned indicated the

final values of tip and skin capacities appeared to be dependent on the

initial relative density (Dr) of the sand. Vesic suggested the

following ultimate tip and side wall unit capacities.

q = 4(1024)D r3 Eq. 2-25

q = 0.08(101.5)Dr4 Eq. 2-26

The subscripts p and s refer to point and side wall, respectively.

The suggestion that residual stresses might influence the ultimate

capacity of statically loaded piles led to further research. The time

dependence of residual stress relaxation was also investigated using the

measurement techniques discussed in the proceeding paragraphs. Such

efforts have provided the current understanding of pile behavior. The

distribution of friction along the pile wall appears to be parabolic

(Vesic, 1970). Other researchers have also recognized this distribution

of pressures along the shaft. The overall participation of the shaft in

carrying the total pile load is proportionately greater in the early

stages of loading for the end bearing pile as well as the friction

pile. The side walls make some initial contribution to supporting the

applied load before the tip makes its initial contribution. Application

of additional load results in mobilization of friction capacity along

the walls until the ultimate capacity is progressively achieved at

increasing depth along the shaft. The total load at the tip increases

during mobilization of the skin friction and increases linearly in con-

junction with the applied load after skin the friction component has

been fully mobilized (Marcuson and Bieganousky, 1977a). The skin fric-

tion component is not necessarily fully mobilized in all cases. Either

the pile is conservatively designed and will never reach its full capa-

city in service or end bearing failure occurs prematurely and the pile

is rejected.

Cyclical loading of piles can have a significant effect on the

distribution of stresses along the shaft. Initially the pile may

distribute residual stresses from driving resulting in latent movement

of the tip even before the design loading is applied to the butt. Load

transfer in piles has been found to be sensitive to small changes in

soil strain and pile compression (Lundgren, 1978). Loading and

unloading can cause irreversible changes in the distribution of effec-

tive lateral pressures along the shaft. What was originally positive

skin friction contributing to the ultimate capacity of the pile has, in

limited cases, been found to reverse itself over time thus contributing

to the ultimate load of the pile (Brierly, Thompson, and Eller, 1978).

2.3.3 Errors in Placement and Driving Affecting Static Capacity

The static capacity of an individual pile or group can be

influenced by several factors during initial placement and subsequent

driving. Each factor and its effect on static capacity will be con-

sidered individually; however, it is not uncommon to have more than one

error in a driven pile with the effects being either compounded or can-


Axial misalignment results from an initial misalignment of the pile

and driver components, field layout errors, trying to drive a flexible

pile, driving in the proximity of a subsurface obstruction, uneven

ground compaction, excessive surcharge placement after driving, or

penetration of a sloping hard strata. Pile misalignment can result in

problems at the tip as well as the butt. Butt misalignment problems are

found most often in slender piles carrying large loads and least often

in mat foundations supported by piles as the loads are relatively

lighter and carried by many piles. Proper axial alignment is most

important where the butt enters or is encased by the pile cap. Mis-

alignment leads to stress concentrations which may quickly exceed the

design stresses. This error is easily checked and can be corrected by

modifying either the pile or the cap, or both, to reduce stress concen-

trations. The most critical applications may-require redesign of the

foundation once the actual butt locations are known. Likewise, mis-

alignment at the tips of group piles can overstress the soil causing

local failure (ASCE Committee on Deep Foundations, 1984).
Significant overstress can result even when the pile is placed

within the normal design tolerances, usually 3.0 inches. Davisson

(1978) found one instance in which the load of an individual pile had

been increased 24% although all piles in the group had been placed

within design tolerances.

Severe axial misalignment, which sometimes results in bending of

the pile, is not necessarily cause for rejection. Several analytical

capacity prediction methods are available (ASCE Committee on Deep Foun-

dations, 1984) and load testing is a viable alternative.

Static capacity can be reduced in two ways due to misalignment of
the hammer and pile cap elements. Visible damage and failure can result

from the concentric overloading of thin-walled pipe piles such as those

modeled in this research. High hammer impact velocity contributes to

"rolling" of the butt (Dismuke, 1978). Damage may be severe enough to

warrant rejection of the pile. Should the pile be judged suitable for

service, its capacity may have been reduced as the stiffness of the

shaft usually decreases due to deformation. Piles adjacent to the

damaged one must carry a larger share of the load.

2.4 Static Capacity Verification
Static load tests are conducted to verify the design capacity of a

placed pile and determine the suitability of the pile type selected.

Although many variations exist in the conduct of the load test, all

involve the static loading of the pile in increasing increments with

various measurements being made of the settlement of the pile at the

butt. Appropriately instrumented piles will render data concerning

displacement and load at the instrumentation points. Consult ASTM

D-1143 for specific criteria regarding pile capacity verification.

After placement of the test pile (group), a framework is con-

structed above the pile. This framework serves to provide the reaction

load against which the pile is jacked. A hydraulic jack is usually

placed between the pile and reaction framework and load is applied by

increasing the fluid pressure in the jack. Load can also be applied

simply by placing iron ingots or even soil in a suitable box mounted

directly on the butt.

An independent framework is also constructed which serves as the

reference against which settlement readings are made. In most

instances, a dial gage is installed on the reference beam with the indi-

cator stem resting on the test pile butt. Redundancy is important and

gages are usually installed on opposite sides of the butt to determine

average settlement. ASTH D-1143 stipulates two independent systems be

used to determine settlement. The backup system sometimes consists of a

taut wire stretched between two posts placed in the ground beyond the

zone of influence of the pile or its reaction load. Readings are taken

visually from a ruler and mirror placed on the butt.

The pile is loaded incrementally up to as much as twice the design

load and settlement of the pile butt is measured for a specified period

of time after placement of each load increment. The time period between

placements of the test load is determined by the size of the load and

the type of soil in which the pile is placed. Maximum load can be held

for up to two days before the load test is considered complete.

The test load is removed incrementally and rebound readings are

made. The readings taken after the load has been removed give an indi-

cation of the total settlement of the pile and its ultimate capacity.

Rebound readings should be made after the removal of each increment for

a sufficient amount of time to ensure rebound has stopped before the

next increment is removed.

2.5 Pile Group Behavior

The behavior of statically loaded individual piles has often been

the basis for predicting the capacity of pile groups given the same type

pile and soil stratification. This approach has led to the development

of efficiency formulas whereby the capacity of a pile group equals the

capacity of an individual pile times the number of piles in the group

multiplied by an efficiency factor (ASCE Committee on Deep Foundations,

1984). This approach most closely approximates the true bearing capa-

city of pile groups spaced at 2 to 3 diameters and deriving their

principal support from granular media. The efficiency formulas do not

account for the time effects which can manifest themselves as excessive

settlements. ASTM D-1143 recognizes that capacities and settlements of

pile groups cannot usually be inferred from the test of an individual

pile in a like mass of soil. Due to the relatively short time span over

which the load test is conducted on the individual pile, the positive

skin friction resulting from driving may not have time to redistribute

itself and can lead to a higher perceived capacity than the pile can

maintain over its lifespan. If the pile group is underlain by a com-

pressible layer, even at a significant depth, the entire foundation may

settle subsequently to the consolidation of that layer. This type of

failure will not be apparent from measurements taken from nearby

reference piles as the reference piles are settling with the pile

group. The piles may not be settling with respect to the soil in which

they have been driven.

Group pile response to loading may vary significantly from that of

individual piles due to the different zones of influence created by each

foundation. It is important to consider the relative contribution of

skin friction and point bearing to the overall capacity. A single pile

can derive a large part of its capacity from skin friction with a pro-

portionately smaller part being derived from end bearing. The pile

group stresses the soil in such a way that the entire block of soil'

contained within the group boundary transfers a relatively larger

portion of the load to a lower stratum. Additionally, a rigid cap

serves to transfer a greater proportion of the total load to the outer

piles. This may contribute to the total mobilization of the friction

capacity of the outer piles with the inner piles being only slightly

stressed. In a group configuration, the piles act more as end bearing

than friction piles (Peck, Hanson, and Thornburn, 1974). Thus, the zone

of influence created by individual piles contributes to the overall

influence of the group when a number of piles are driven as a group.

The interaction between piles of a group becomes more pronounced as the

Poisson's ratio of the surrounding soil decreases (Butterfield and

Banerjee, 1971).

The behavior of the group may vary significantly from that of the

individual pile. This variation makes the prediction of group behavior

dangerous and difficult when based on only individual pile load tests

even if the bearing soil is the same. Regarding group settlement in

granular soils, the major portion of settlement occurs immediately. The

shape of the load-deflection curve is similar to that for individual

piles;.however, the proportion of immediate settlement is generally

smaller for a group than for the individual pile (Butterfield and

Banerjee, 1971).


3.1 Similitude Theory

Driving model piles in an artificially induced high gravity envi-

ronment is a complex task which does not lend itself easily to analy-

tical solutions based on mathematical models. As discussed previously,

conducting tests with miniature piles at one gravity does not accurately

model the prototype stresses experienced at the soil-pile interface.

Uncertainties are introduced if the results are extrapolated to proto-

type depth as soil has stress-dependent mechanical properties that vary

with depth (e.g., strength, moduli, wave propogation velocity, etc.).

Whitaker (1957), Saffery and Tate (1961), and Sowers et al. (1961)

produced qualitative results in their studies of miniature pile (group)

response to loading at 1-g. Collectively, these efforts served to

verify the findings of Rocha (1957) that the results of 1-g tests cannot

be scaled up to re-create or predict pile performance. The model must

be subjected to stresses equivalent to those experienced by the proto-

type. Scott (1977) demonstrated the ability to conduct pile load tests

in a high gravity environment by cyclically loading piles inserted in

silt. There was no prototype and thus no standard with which to compare

the model performance.

It is necessary to create equivalent self-weight-induced stresses

at similar points of interest in the model as in the prototype. This is

accomplished by subjecting a properly scaled model and soil mass to a

gravity level sufficiently high to produce the desired stress. This

acceleration is provided most simply by a centrifuge.

Scaling relationships permit the design of the pile and driver for

the proposed gravity levels and interpretation of the data obtained by

operation of the device at its design gravity level. Testing in a high-

gravity environment requires the energy input of the miniature pile

driver to be scaled down in some manner to insure equivalency with the

prototype. Likewise, the pile capacities and stresses developed during

placement and loading must be scaled up for comparison with the proto-

type. The scaling relationships are derived by dimensional analysis.

Dimensional analysis is a method by which we
deduce information about a phenomenon from the
single premise that the phenomenon can be des-
cribed by a dimensionally correct equation among
certain variables.

(Langhaar, 1951, p. 17)

As the number of variables affecting a process increases, it becomes

progressively more difficult to determine the function which relates the

experimentally controlled inputs (independent parameters) to the

response being investigated (dependent parameters). By creating

dimensionless parameters relating the most easily controlled experi-

mental parameters to the ones being measured, ratios can be defined

which relate model response to prototype response. These dimensionless

ratios, called Pi (r) terms, were first developed by Buckingham and

presented in 1914.

3.2 Selection of Dependent and Independent Variables

The choice of independent and dependent variables is of most criti-

cal importance in the development of the Pi terms.

If variables are introduced that really do not
affect the phenomenon, too many terms may appear
in the final equation. If variables are omitted
that logically may influence the phenomenon, the
calculations may reach an impasse, but, more
often, they lead to incomplete or erroneous

(Langhaar, 1951, p. 19)

The selection of basic units for the Pi terms is also very important.

The mass, length, time (MLT) system was selected for this research as

unit mass does not change in multiple gravity environments and the MLT

system appears to be favored in previous references (Nielsen, 1983;

Bradley et al., 1984; and Tabatabai, 1987).

Table 3-1 presents the researcher's choice of independent and

dependent variables considered pertinent to this investigation. The

first three variables are the independent parameters which are con-

sidered easily controlled experimentally. The choice of which three

variables were used was subject to the restrictions that, among the

three, each of the basic units had to be used at least once, and the

variables had to be independent among themselves to ensure independent

solutions. The remaining variables were chosen to represent the

material and physical properties considered to be important in the

process being modeled. Certain properties of each of the materials used

are important, such as modulus, compression wave speed, and strength.

However, as the units for any one of those parameters are the same for

each material under consideration, the collection of like parameters is

represented by one generic term. Thus, Young's modulus of elasticity,

E, is included only once in the collection of variables and not once for

each material. Material properties relating to the presence of water in

the soil were included as the model piles were designed to be capable of

being driven into saturated soils.

Table 3-1

Independent and Dependent Variables











w Stress
7r Acceleration
t Impact Energy
r Impulse
r Cohesion
iT Dynamic Time
r Wave Speed
7r Yield Strength
T Displacement
r Area
iT Permeability
i Hydrodynamic Time













The Pi terms were developed using the program PISETS written by

Theodore Self at the University of Florida in 1983. PISETS was avail-

able on the Commodore SuperPET microcomputer as well as the Northeast

Regional Data Center (NERDC). Bradley et al. (1984) and Bradley (1983)

present the method of setting up the input matrix for use with PISETS.

Appendix A presents the derived Pi terms and verification that each is


3.3 Development of Scaling Laws

Scaling laws were formulated by taking the derived Pi terms and

equating the model Pi term to its equivalent prototype Pi term. The

scale length factor was used to change geometric properties by an appro-

priate factor of n. Like materials in the model and prototype permitted

the cancellation of material properties in the equivalent Pi terms. The

scaling laws were derived as presented in Appendix A and are summarized

below in Table 3-2.

Table 3-2 Scaling Relationships

Property Prototype Model

Length 1 1/n
Density 1 1
Modulus 1 1

Stress 1 1
Acceleration 1 n
Impact Energy 1 1/n3
Impulse 1 1/n3
Cohesion 1 1
Dynamic Time 1 1/n
Wave Speed 1 1
Yield Strength 1 1
Displacement 1 1/n
Area 1 1/n2

Dimensionless quantities such as strain and Poisson's ratio remain
dimensionless and scale 1:1. Quantities not specifically derived above

can be determined by combining the appropriate Pi terms. The two

remaining pertinent scaling laws, permeability and hydrodynamic time,

will be derived as examples.

Permeability (k) has units of length/time. Model permeability

equals prototype length times the scaling factor 1/n divided by the

prototype time multiplied by its scaling factor, 1/n. The scaling

factors (in this example) cancel one another indicating that model

permeability equals the permeability of the prototype. This is a pre-

dictable result considering the requirement to utilize like materials in

the prototype and model.

Hydrodynamic time (Th) is determined by a similar method. The

equation for hydrodynamic time is derived from Terzaghi's one-

dimensional consolidation differential equation which provides a

nondimensional time T called the time factor (Lambe and Whitman,

1969). The time factor equals the coefficient of consolidation, Cy,
multiplied by real hydrodynamicc) time, Th, and divided by the square of

the drainage path, H. The equation for hydrodynamic time and its units

are provided below.

T H2
Th = C Eq. 3-1

T L2
Th = T T2 Eq. 3-2

The units of hydrodynamic time are T2. Thus, model Th equals pro-

totype Th multiplied by the square of the dynamic time scale factor

(1/n), or 1/n2. The two scaling laws derived are summarized below. The

subscripts m and p designate the model and prototype variables, respec-


km = kp Eq. 3-3

Th = Thp 1/n2 Eq. 3-4

3.4 Experimental Requirements

Care must be exercised in the application of the scaling laws to

the development of the model pile driver. Consider the impact energy

imparted by the hammer striking the pile. The scaling factor which

relates the prototype energy to the amount of energy input by the model

driver is 1/n3. Modeling the 50,000 ft-lbs of work needed to lift the

prototype hammer (5,000 pounds with a drop height of 10 ft) at 70 g's

requires the model pile driver input 0.146 ft-lbs. The model driver

input can likewise be determined by considering the hammer weight and

drop height independently. The ram weight of 5000 Ibs divided by n2

equals 1.02 Ibs and the fall height is reduced to 0.143 ft. The product

of these two properly scaled values is the required 0.146 ft-lbs. How-

ever, the hammer weight required during driving is the weight of the

hammer mass at 1 g multiplied by the gravity level at which the test is

conducted. The model hammer weight is then 1.02 Ibs divided once more

by the gravity scale factor, n. Thus, a model hammer weighing 0.233

ounces at 1 g is sufficient to model the 5,000-lb prototype hammer when

subjected to 70 g's.


This chapter documents the design of the system developed to place

the model piles. Computer control of the system and data recording

techniques are presented. Pertinent fabrication methods are discussed.

The model pile sizes were determined prior to the system design as

all critical dimensions of the placement device are predicated on the

model dimensions. Designing the piles to be used as models required an

understanding of the type of pile to be modeled, the loading regime

under which the prototype pile was tested, and the modeling techniques

used by previous researchers. The higher stresses experienced by the

model pile during driving predicated a significant departure from the

existing techniques.

The placement device was designed to permit the pushing and driving

of model piles and groups permitting the comparison of the residual

stresses developed in each model resulting from the different methods of

placement. Design required an understanding of the prototype pile

driver performance capabilities.

The centrifuge was designed to permit positioning of the specimen

and placement device with sufficient clearance to allow all tests to be

safely performed. The uses of the Hewlett-Packard 6940B Multiprogrammer

and 3497A Data Acquisition/Control Unit are also detailed. Several

limitations in the use of the pile placement device are presented in the

final section of this chapter.

4.1 Model Pile Design and Instrumentation

The pile system modeled resembled the Hunter's Point group of five

piles (Oneill, 1986). This system was designated as the prototype

although there were limitations in modeling of the axial stiffness and

soil properties. Each pile was 35-ft long tubular steel with a wall

thickness of 0.375 in. The outside diameter of the piles was 10.75 in.

When driven in a group, the piles were arranged in an 'x' pattern with a

pile in the center and one at each corner. The corner piles were 3 ft

from the center pile. The cap was made of concrete having a thickness

of 5.0 ft. The pile butts were embedded 2.0 ft into the cap leaving 3.0

ft of cover. The bottom of the cap was 3.0 ft above the ground surface.

Nominally, the cap was 6.0-ft square. Total driven depth was 30.0 ft.

4.1.1 Model Material Selection

Similitude theory is simplified by the understanding that models

will be made of the same material as the prototype to ensure the same

stresses are developed in the model as in the prototype during testing.

Use of the same material in the model was important in this application

because the transfer of forces from the pile to the soil was investi-

gated. As soil strength is mobilized by the slight movement of the pile

surface relative to the surrounding soil, a variation in this relative

movement would alter the load bearing capacity of the model pile as well

as the load-deformation curve. However, strict adherence to similitude

would have necessitated the use of a thin-walled steel tube as the pile

model. The wall thickness would have varied between 6.3 and 4.7 thou-

sandths of an inch for tests at 60 and 80 g's. Model piles with this

wall thickness would have been crushed during driving due to lack of

axial stiffness or the inability of the model driving device to strike

the pile perfectly in flight. Furthermore, the available steel tubes

which could reasonably be considered as potential models had wall

thicknesses far in excess of the desired amount. This would have

resulted in an axial stiffness too great to permit accurate modeling of

the stress transfer (load shedding) between the pile and soil. There-

fore, various materials were considered which had lower strength moduli

and were available in small tubular shapes. The axial stiffness of the

prototype could be accurately modeled if the scale cross-sectional area

of the model pile was increased by a factor equal to the modulus of the

steel prototype (30,000,000 psi) divided by the modulus of the model

material being considered. Aluminum was chosen as a suitable modeling

material and has been used successfully in several previous studies

(Millan, 1985; Ko et al., 1984; Harrison, 1983). Aluminum has a Young's

modulus of 10,000,000 psi requiring that the scale cross-sectional area

of the model pile be three times that of the prototype. Theoretically,

the model pile would strain the same amount under the same scale load as

the prototype thereby having the most potential of accurately mobilizing

the same soil strength.

4.1.2 Model Size Determination

Limitations in the availability of tubular aluminum and the diffi-

culty of manufacturing models with accurately scaled dimensions forced

the need for a compromise between the modeling of pile diameter (and

thus scale surface area) and axial stiffness. All tests were conducted

with the models having the correct scale outside diameter and length.

Because the available tubular aluminum had thinner walls than required

for correct modeling, the scale axial stiffness of these models was

lower than that of the prototype. This approach correctly modeled

the scale surface area of the model pile as well as the scale area of

the tip; however, the model piles were less stiff (axially) when com-

pared to the prototype stiffness.

The gravity level of testing for pile models with scale geometric

proportions (outside diameter and length) was determined by dividing the

outside diameter of the prototype by that of the available model mate-

rial. The length of the model was then the prototype length divided by

the gravity factor. Table 4-1 provides the model dimensions and scale


Table 4-1 Model Pile Dimensions and Scale Factors

Controlled by Geometry (Outside Diameter)

Prot. A B C 0

O.D. (in.) 10.75 0.125 0.154 0.187 0.218
g-level (g's) -1 86.0 69.8 57.5 49.3
Length (in.) 420 4.88 6.02 7.39 8.52
Stiffness (%) 100 95.4 73.8 66.6 57.5

NOTE: The axial stiffness of the model piles is given as a percentage of
that of the prototype.

In summary, four suitable aluminum tubes were found for the manu-

facture of model piles. Scaling factors based on the outer diameter of

the tubes yielded four pile sizes to be tested at g-levels between 49.3

and 86.0 g's. There was no manufactured aluminum tube which accurately

modeled both the outer diameter and axial stiffness. Figure 4-1(a) de-

picts the individual model piles and associated gravity levels of test-

ing. Figure 4-1(b) depicts the model pile groups and test gravity


a) Individual Model Piles (Left to Right--49.3, 57.5, 69.8

and 86.0 g models)

b) Model Pile Groups (Left 69.8 g., Right 86.0 g)

Figure 4-1

Model Piles. a) Individual Model Piles (Left to
Right--49.3, 57.5, 69.8 and 86.0 g models)
b) Model Pile Groups (Left 69.8 g., Right 86.0 g)



, I




4.1.3 Strain Gage Placement

Research has been conducted using aluminum tubes which were split

in half to permit the placement of strain gages, then glued together

prior to pushing them into the modeled soil (Millan, 1985; Harrison,

1983). The difficulty of obtaining a strong bond between the two pile

halves frequently resulted in the model piles splitting and exhibiting

irregular butt deflection patterns during loading. The significantly

greater stresses required for driving the pile prevented the reuse of

this technique as only continuous tubular stock could be expected to

withstand the driving process. The most significant drawback in the use

of continuous stock is the difficulty of placing strain gages inside the

model pile.

The method of placing the strain gages had to satisfy three cri-

teria; the bond between gage and model had to be complete and permanent,

the gages had to be positioned precisely, and the gages had to be able

to perform after completion of driving. In addition to the above

criteria, the gages had to be placed with leads already soldered on.

Connection of the leads to the gages after placement was not possible

without exceptionally sophisticated equipment.

Each strain gage, regardless of its working environment, must be

placed on a clean, chemically prepared site. Proper cleaning ensures

the gage can be bonded as intimately as possible with the material being

measured. Preparing the sites inside the models was accomplished in

accordance with Micro-Measurements Instruction Bulletin B-130-2. Fine

grade emery paper was rolled into a small tube and inserted in the

model. Being a soft metal, the aluminum required little sanding to

smooth the interior and expose clean metal at the sites. An acid

solution (Micromeasurements M-Prep 5) was used to etch the metal during

cleansing. The site was neutralized with the appropriate solution,

Micromeasurements Neutralizer. Preparation was completed by pushing

small wads of lint-free cotton material down the pile shaft in one

direction only to prevent recontamination of the cleaned sites. The

cleansing process was repeated until the cotton wads emerged dirt-free

from the model pile interior.

Suitably sized strain gages (Micromeasurements EA-06-015LA-120,

Figure 4-2) were then prepared for placement. These gages were chosen

because the lead terminals were large with respect to the field and were

placed above the field in such a way that they presented a narrow plan

view. This was important as the gage had to conform to rather tight

radii on the insides of the models. The minute gage size required the

leads be soldered on while being viewed under magnification. Conse-

quently, the gages were sent to a laboratory where suitable equipment

was available.

The strain gages with leads attached were prepared using the fol-

lowing procedure. Excess gage backing material was removed as indicated

in Figure 4-3(a) to make the gage as tall and slender as possible. Gage

resistance and continuity were checked to determine if damage had

occurred during the trimming process. A piece of Micromeasurements

nonstretch tape was cut to the width of the length of the strain gage

field. This tape was affixed to the two strain gages holding them in

the same position relative to one another (Figure 4-3(c)). Special care

was taken to ensure the principal strain measuring axes of both gages

were parallel and the gages were the correct distance from one another.

The distance between the two gages was predicated on the inner diameter

Figure 4-2

Micromeasurements EA-06-015LA-120 Strain Gages


f) Z

g) ..

i) p.

** -9 7

Strain Gage Application Technique

I I I -

1 --L-------~

Figure 4-3

of the pile model in which the gages would be placed. The correct dis-

tance equaled one-half of the inner circumference of the model. A small

amount of rapid-curing epoxy was applied as a sealer to the leads,

gages, and tape as shown in Figure 4-3(d). This assembly was set aside

to dry while the insertion device was prepared.

A Coronary Balloon Dilitation Catheter served as the insertion

device. The catheter is fitted with a balloon which expands to a

specific diameter upon pressurization. The balloon was wrapped with

Teflon tape as shown in Figures 4-3(e) and 4-3(f) while the epoxy was

drying. The gages and tape assembly was then wrapped 'gages out' around

the catheter balloon (Figure 4-3(h)). Gage resistance was rechecked at

this time to ensure only functional gages were inserted into the model.

The small size of the deflated balloon permitted the catheter and gages

to be inserted as a unit down the shaft of the pile (Figures 4-3(i) and

4-3(j)). The lead wires were inserted first, into what was to eventu-

ally become the pile tip. They were followed by the gages and balloon

assembly which was advanced by pushing the pressure line leading to the

balloon. Proper positioning was accomplished by marking the pressure

line at the correct depth of insertion and inserting the assembly up to

the mark. Each pair of gages became successively more difficult to

insert as the leads from the previously placed gages interfered with

feeding the new leads through. When in the proper position, the balloon

was inflated, pressing the gages into the replaced bonding material at

the site. The balloon remained inflated until the bonding material had

cured at which time the balloon was deflated and removed. The wrap of

Teflon tape prevented the balloon from being bonded to the gages.

Placement of the gages had to be accomplished in a specific order.

Models requiring strain gages only at the tip were produced using the

previously described technique of inserting the lead wires into the tip

and following with the balloon and gage assembly. Models requiring the

placement of multiple pairs of gages underwent a slightly modified

placement procedure to simplify the placement of the first pair. The

balloon was first inserted through the model pile before being wrapped

with teflon tape and having the gage assembly attached (Figure 4-4).

This method improved the placement success rate of the butt gages by

requiring they be pulled back into the model a short distance rather

than being pushed through the length of the model before bonding. The

balloon was deflated and withdrawn after placement in the same manner as

when the gage pair is placed by being pushed through the model. The

resistance of each gage was measured after being placed in the model.

Variation of the gage resistance by more than 5.0% of the specified

factory value of 120 ohms was cause for rejection of the gage and noted

in the model pile construction log. There was no way to replace the

gages if they were rejected. After placement of the final pair, the

butt was sealed with a silicone cap reducing the chance of the leads

breaking off of the gage during driving and testing.

The model was completed by fitting the tip with an individually

machined cap which was inserted into the model plugging the tip. The

outside diameter of the plug was equal to the model diameter. Each cap

was held in place by pressure fit and sealed on the inside with silicone

sealer to prevent the intrusion of water. This technique and the afore-

mentioned procedures yield a fully instrumented waterproof pile capable

of withstanding the stress of driving.



Figure 4-4 Alternate Technique for Placement of First Pair of
Strain Gages When Multiple Pairs are Needed

Strain gages were positioned in the instrumented models at the tip

and butt only. Initial tests were conducted on pushed piles with one

pair of strain gages at the tip (0.97d with d being the distance the

model was pushed into the soil). Instrumented driven models (Figure

4-5(a)) should have pairs of strain gages placed at the one-third points

and tip. The uppermost pair (gages 4A and 4B) serve to measure the

total resistance of the pile to penetration due to both tip capacity and

skin friction. The pushed models use a load cell to measure the total

resistance to penetration and can therefore have four pair of gages

placed at the quarter-points and tip (Figure 4-5(b)). Individual piles

fitted like the one in Figure 4-5(a) should be used for measurement of

the tip and side friction capacity in driven models. Again, the gage

pair at the top (butt) of the model is required because the load cell

could not remain in-line when models were driven. Positioning of the

gage pairs at the butt in individually instrumented driven piles permits

the immediate conduct of loading after the pile is driven.

Figure 4-6 depicts the suggested placement of gage pairs when

instrumented model groups are driven and pushed. After being driven,

the centrifuge must be stopped to fit the top of the group with a load

cell. Load tests are then conducted after the centrifuge has been spun

up to the required test rpm. When the group has been pushed in, loading

is conducted immediately as the load cell is already in place.

Provision was made to protect the leads during driving and loading

as will now be discussed. This method of protection was also used to

prepare all models pushed into the soil with only slight modifications

of the caps depending on whether the model was driven or pushed.

4A and 48

0.33d I I 3A and 3B

0.67d- I 2Aand2B

0.97d -


0.25d I 4A and 4B

0.50d I 3Aand3B

0.75d -

2A and 2B

1Aand1B 0.97d-U 1AandlB

a) Driven Pile, Four
Pair of Strain

Figure 4-5

b) Pushed Pile, Four
Pair of Strain
Gages and Load Cell

Suggested Gage Positions and Position Designations
for Individual Driven and Pushed Piles (8 Strain
Gage Channels Available), a) Driven Pile, Four Pair
of Strain Gages; b) Pushed Pile, Four Pair of Strain
Gages and Load Cell


0.97d -

-2A and 2B

*4A and 4B

1A and 18


Figure 4-6 Suggested Gage Positions and Position Designations
for Driven and Pushed Pile Groups (Load Cell Placed
on Top After DRIVEPILE; Load Cell In Place During

4.1.4 Pile (Group) Cap Fabrication and Lead Wire Protection

An important consideration in pile driving is the protection of the

butt of the pile from the potentially damaging blows of the hammer.

This was accomplished using a pile cap as discussed in Section 2.3.1,

Hammer-Cushion-Cap Interaction. A similar method was used in the

driving of the models. Figures 4-7 through 4-10 show the types of caps

developed for driving and pushing the model piles (groups). The caps

served to protect the model as well as the strain gage lead wires while

permitting the models to be interchanged quickly between tests. The

lead wires were terminated in miniature connectors mounted on the sup-

port arm shown in Figures 4-7 through 4-10. When only one connector was

necessary, the weight of the connector and shielded leads to the side of

the specimen container was counterbalanced by the placement of a suit-

able weight at the end of the support arm extension through the other

side of the slotted cast acrylic guide tube. This precluded the devel-

opment of excessive moment at the top of the model when under the

influence of the design gravity level. The model pile, cap, lead wires,

and connectors are thus a complete unit capable of being plugged into

the placement device as necessary for testing. This permits the models

to be reused and interchanged quickly between tests.

4.2 Model Placement Device Design

The placement device needed for this investigation was required to

lift a considerable variety of weights, provide accurate feedback re-

garding position of the lifted weight, and accurately raise the lifted

weight a predetermined height above an object whose position changed

with every impact. All of these activities were to be performed in an


Cast Acrylic
Guide Tube

- Hammer

- Support Arm

* I

Counterweight or
Additional Connectors
(as needed)

Figure 4-7 Pile Cap Assembly for Individually Driven Model Piles


Cast Acrylic
Guide Tube

Group Cap Assembly

Lead Wires

Figure 4-8 Pile Cap Assembly for Driven Model Pile Groups

Slotted Cast
Acrylic Guide

Load Cell Adaptor
Set Screws
Support Arm

r 4.

or Additional
Connector (as

Figure 4-9 Pile Cap Assembly for Individually Pushed Model Piles

Lead Wires

Cast Acrylic
Guide Tube

Load Cell Adaptor
Set Screws

Lead Wires

Model Pile

Figure 4-10 Pile Cap Assembly for Pushed Model Pile Groups

environment whose gravity level was high and subject to variation during

the course of an experiment. The system developed was chosen after

exhaustive elimination of many other mechanical methods, several gas/

fluid hydraulic systems, and even a purely electromagnetic device. The

main objection to some of the proceeding systems was their inability to

"follow" the model pile (group) as it advanced through the soil and

their reliance on awkward position monitoring devices which would have

affected the dynamic response of the system. Compounding the difficulty

of building the device was the limitation imposed by the lack of clear-

ance above the platform. A suitably sized soil specimen with a model

pile on the surface and the placement device above requires a relatively

deeper centrifuge than that needed for experiments with compact speci-

mens. Furthermore, the system must be designed with a low center of

gravity. An alternative design considered the driving device being

mounted statically on the arms of the centrifuge with the specimen

swinging into position for driving. However, pile placement accuracy

could not be guaranteed with this design. The chosen configuration con-

sisted of a soil specimen container with the driving device mounted

above in a protective canister.

The system chosen (Figure 4-11) consists of a stepper motor geared

through a ball bearing screw assembly (ball screw) to a support shaft

which lifts an electromagnet and pile hammer when individual piles and

groups are driven. Rotation of the ball nut (the housing around the

ball screw which contains the ball bearings) either raises or lowers the

ball screw to which the electromagnet is attached. During driving, the

electromagnet is advanced (downward) until it comes within 0.005 in. of

the ram. The closeness of approach is determined by a proximity device.

Figure 4-11

Specimen Container, Placement Device Protective
Canister, and C-Channel Support Beam

The electromagnet is then energized. Reverse rotation of the stepper

motor retracts (upward) the magnet and ram to the predetermined drop

height. The magnet is then de-energized allowing the weight to fall and

impact on the pile cap.

The predetermined lift height is achieved by rotating the stepper

motor a known number of "steps". Counting the number of steps which the

magnet is lowered to regain contact with the ram permits the determina-

tion of the number of steps the device has penetrated the soil. That

number of steps is easily and accurately converted to a vertical dis-

tance. The process then repeats itself with the energizing of the

magnet and the retraction of the magnet and ram assembly to a newly

determined lift height. This system offers the advantages of accuracy

and control of the height to which the ram is lifted, repeatability, a

free falling weight, the capability of following the device as it

penetrates the soil, and no bulky displacement measuring devices to

alter the dynamic performance of the system.

When piles (groups) are pushed, the electromagnet and hammer are

replaced by an in-line load cell which fastens directly to the base of

the ball screw. During pushing, the pile (group) is advanced into the

soil with readings being taken from the load cell at frequent intervals.

Load tests are conducted as will be presented.in Section 4.5.2,

Software. The major components of the system will now be presented.

4.2.1 Specimen Container, Placement Device Protective
Canister, and C-Channel Support Beam

Support for the device consists of a circular specimen container

machined from high-strength aluminum and welded to a square base plate.

The square base plate is bolted to the swinging platform on the end of

the centrifuge arm (Figure 4-11). The base plate can be mounted in four

configurations, with any given side being able to be positioned at 0,

90, and 180 degrees with respect to the center line of the platform.

The top of the soil specimen container has a groove machined around the

rim. This groove serves as the male end which fits snugly into the

female groove machined into the base of the placement device canister.

The grooves hold the canister in place above the specimen container

while permitting rotational flexibility. The top of the canister sup-

ports the aluminum c-channel which is the main frame of the placement

device components. The c-channel has limited positioning capability

which permits the placement of the ball screw directly above the center

of the canister centerline as well as in any position within approxi-

mately 2 in. of the centerline. This flexibility permits the driving of

pile groups by individual piles as with the prototype. The combination

of the back and forth freedom of movement of the c-channel with the

rotational freedom of the canister top provides substantial positioning

freedom within the specimen container.

The capability of mounting the specimen container base plate in

four unique positions was included in the design to ensure the least

stress was exerted on the platform support arms during testing. Figure

4-12(b) shows how the canister and placement device should be config-

ured. The intent is to decrease the eccentricity of the load on the

platform due to the placement of the stepper motor. Figure 4-12(a)

shows the configuration of the canister and stepper motor when video

recording of the placement process is necessary. This position exposes

the placement mechanism to the camera through a port in the side of the

canister when the platform and canister have rotated to horizontal. A

mirror is needed for the video camera to view the placement device.

a) Configuration Used for Video Recording

b) Normal Configuration

Figure 4-12 Top View of Canister and Placement Device Configuration.
a) Configuration Used for Video Recording;
b) Normal Configuration.

However, this is the least desirable configuration and was used only

when necessary. In this position, the concentrated load of the motor

(close to the vertical centerline of the platform) will force overrota-

tion of the platform. This configuration would ultimately cause the

payload to fall off into the bottom of the centrifuge housing if the

payload were not properly fastened to the platform or if the platform

were not fixed against overrotation. Rotation of the canister until the

stepper motor is almost in line with the horizontal centerline of the

platform (Figure 4-12(b)) will decrease the tendency of the platform to

overrotate. However, video recording of the placement process is not

possible because the line of sight is blocked by the canister side.

Rotation of the canister until the stepper motor is above the center

line of the platform will induce underrotation, another obviously unac-

ceptable occurrence. The combination of stepper motor positioning which

caused the least stress on the platform bearings was the position which

developed a slight tendency to overrotate just pushing the platform arms

against mechanical stops installed on the centrifuge arms (Figure 4-13).

This was the configuration used most often. Once this position was

established, the driving of second and subsequent piles in a group could

be accomplished by pushing the stepper motor slightly closer to the

central axis of the specimen container. The shift in placement would be

equal to the scale distance between multiple piles. Shifting the step-

per motor towards the central axis of the specimen container would then

serve to decrease the eccentric loading induced by the motor. For

subsequent piles, the specimen container would be rotated ninety degrees

underneath the stationary canister.

Figure 4-13 Platform Mechanical Stops

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs