Академический Документы
Профессиональный Документы
Культура Документы
FOUNDATION
ANALYSIS
AND
DESIGN
SERIES IN GEOTECHNICAL ENGINEERING
Edited by
T. William Lambe
Robert V. Whitman
Professors of Cil'il Engineering
Massachusetts lnstítute of Technology
BOOKS IN SERIES:
Soíl Testing for Engineers by T. William Lambe, 1951
Soil Mechanics by T. William Lambe and Robert V.
Whitman,, 1968
Soil Dynamics by Robert V. Whitman (in progress)
Fundamentals of Soil Behal'ior by James K. Mitchell, 1976
Elastic Solutions for Soil and Roe k Mechanics by H. G.
Poulos andE. H. Davis, 1974
Soíl Mechanics, SI Version by T. William La mbe and
Robert V. Whitman, 1978
( f
ti5lt~~.fZ ;ttJT N ~~~t
PI LE
FOUNDATION
ANALYSIS
AND
DESIGN
H. G. POULOS
E. H. DA VIS
The University of Sy~._;r·:>¡>·:-'73:'.';_:;'"~~
. .
~··.4
~ainhnfu-~rHr.s~ ~nnk Qin. ·
554786
Copyright 1980
All rights reserved. Published simultaneous1y in Ca nada.
Rcproduction or translation of any part of this work beyond that
permitted by Sections 107 and 108 of the 1976 United StatesCopy·
right Act without thc permission of the copyright owner is un1aw-
ful. Requests for pcrmission or further information should be
addressed to the Permissions Department.
~52~~~~~~
rp ~ 1:: + if- + J3 + /"' S f!!#- Jíti !n
lil:J~~iU'!;Cf:~ _.iJ1lO-~
A: J~ ~
{í tJl: : ~[: íjí Jfi, A5 .#t 1 o7
~ :fr ?fr :
!4:rrf1fT;ttg :W:: :
ffJ Xllil ?JT
* * "1\.ra 11
;¡r_ f,f¡
~t
1t "ñ rll
{lf
1? j!. 45 @.. ¡;p JP'J "ñ r!l .~;::;- i5J
~
1
.~;::;-
o7
i5J
it
This book deals with methods of analysis that may be use- the interaction between two or more piles and, therefore,
fui in design of pile foundations. Many excellent text- to examine the behavior of groups of piles .
. books are concerned with the more practica! aspects of pile The material contained in this book is organized as
foundations, such as the factors influencing the selection of follows:
the type of pile, the techniques of installation, and practica!
details of construction and maintenance of piles. No l. The behavior of piles under vertical loads (Chapters
attempt has been made to duplicate this type of inform- 2 to 6).
tion. The aims of the present book are to: 2. The behavior of piles under lateral loading (Chapters 7
and 8) and under combined vertical and lateral loading
l. Present a consistent theoretical approach to the predic- (Chapter 9).
tion of pile: deformation and load capacity. 3. The behavior ofpiled rafts (Chapter 10).
2. Present parametric solutions for a wide range of cases. 4. Piles subjected to vertical or lateral soil movements
3. Demonstrate how such solutions can be used for design (Chapters 11 to 13).
purposes. 5. Miscellaneous topics such as pile buckling, dynamic
4. Review the applicability of these approaches to practi- loading, and pile load tests (Chapters 14 to 16).
ca! problems.
Although the text deals with a relatively wide range of
In any theory, a certain amount of idealization is neces- topics, it is by no means exhaustive. Furthermore, since
sary to obtain a tractable mathematical solution; this is geotechnical analysis is advancing ata very rapid rate, there
especially so when dealing with problems involving soil. In may well be cases in which the analytical techniques we
dealing with the deformations of pile foundations irr this describe may have been superseded by more versatile
book, we have generally considered the soil as an elastic methods capable of modeling real soil behavior more
material, with allowances made for pile-soil slip and soil realistically. Nevertheless, we feel that the techniques and
yield where appropriate. Although real soils possess few, if solutions presented in this book can be usefully applied to
any, of the attractive attributes of an ideal homogeneous most practica! problems and provide a basic series of results
isotropic elastic material, they nevertheless can often be against which the results of more sophisticated analyses
treated as elastic over a limited range of stress, provided may be checked.
that the "elastic" parameters are determined for this stress Sorne worked examples are given to illustrate the appli-
range. When u sed in this manner, with due discretion and a cation of the solutions to practica! problems. Beca use units
measure of engineering judgment, e!ashc-based theory has are by no means standardized as yet, sorne of the examples
had considerable success in predicting the deformation of are worked in •SI units, sorne in British units, and a few in
both shallow and deep foundations. Although other simple the Continental metric system.
soil models have also been successfully used for various We thank the· many people who have contributed to
. aspects of pile analysis (for example, the theory of subgrade th:is book and in particular Dr. N. S. Mattes, of the Elec-
reaction as applied to laterally loaded piles), elastic theory tricity Commission of New South Wales, who obtained a
provides a unified basis for the analysis of all types of ¡;onsiderable number of the elastic solutions presented, Dr.
foundation; it also makes possible identification of the J. R. Booker and Dr. P. T. Brown of the University of
parameters that exercise a significant influence on pile Sydney, who provided a great deal of assistance with
performance. Since elastic theory allows consideration of various aspects of the theoretical analyses, Mr. P. J. N. Pells
stress transmission through a mass, it can be used to analyze who provided valuable information on the subject of piles
V
vi PREFACE
to rock, and Dr. T. J. Wiesner, who obtained sorne of the made the facilities of the Department available to us, to
solutions presented in Chapter 10. The Civil Engineering C. J. Peiti, B. Crook, J. Kilpatrick, S. Picken, J. Knight and
Graduates Association of the University of Sydney gave B. Rocke who undertook the typing and assembly of the
financial support for the post-graduate .course on pile manuscript, and R. Brew and H. Papallo who prepared
foundations that formed the basis of this book. Grateful many of the diagrams.
ac.knowledgement is given to Professor J. W. Roderick, H. G. Poulos
former Head of the Department of Civil Engineering, who E. H. Davis
CONTENTS
9.3 .2 Determination of Equivalen! Bent 235 11.4.6 PiJe in Soil Subjected to Variable
9.3 .3 Torsional Re'sponse of Piles 237 Loads 284
9.4 Elastic Analysis of Pile Behavior 237 11.4 .7 Data on Pile-Soil Parameters 285
9.4 .1 Analysis of Single Battered Pi! e 237 11.5 Pile Groups 288
9.4 .2 Analysis of Pi! e Groups 242 11.6 Comparisons Between Measured and Pre-
9.4 .3 Pa:rametríc Studies of Pi! e Groups 243 dicted PiJe Behavior 289
9.5 Comparison of Methods of Píle-Group
Analysis 248
12 PILES IN SWELLING AND SHRINKING SOILS 294
12.1 Introduction 294
12.2 Existing Methods of Analysis 295
10 PILE-RAFT SYSTEMS 250 12.3 Analysis Based on Elastic Theory 296
12.3 .1 Basic Analysis 296
10.1 Introduction 250
12.3 .2 Pile-Soil Slip 297
10.2 Analysis 250
12.3 .3 Compression Failure of Pi! e 297
10.3 Elastic Solutions for Square Groups 253
12.3 .4 Tension Failure of PiJe 297
10.3 .1 Influence ofPoisson's Ratios Vs 257
12.3.5 Nonuniform Soil 297
10.3 .2 Influence of Pile Arrangement 257
12.3.6 Variation with Time 298
10.3 .3 Systems Containing Large Numbers
12.4 Typical Solutions for Pile Movement and
of Píles 257
Load 298
1O.3 .4 Effect of Pi! e Compressibility and
12.4 .1 Purely Elastic Pile-Soil Interface 298
Raft Flexibility 258
12.4.2 Solutions IncorporatingPile-Soil
10.4 Simplifíed Analysis for Load.Settlement
Slip 299
Curve to Failure 259
12.4.3 Effect of Tensile Failure of the Pile 303
10.5 Other Analytical Approaches 262
12.4.4 Differences Between Piles in Swelling
and Consolidating Soils 304
12.5 Design Curves 304
11 NEGATIYE FRICTION ON 12.6 Application of Theoretical Analysis to
END-BEARING PILES 265 Practica! Problems 306
12.6 .1 Prediction of Soíi-Movement Pro file 306
11.1 Introduction 265
12.6 .2 Pile-Soil Interface Strength 307
11.2 Field Stúdies on Instrumentcd Piles 269
12.63 Soil Modulus 30?
11.2.1 .Observed Downdrag Forces 269
12.7 Observations of Pile Behavior and Comparí-
11.2 .2 Deve1opment of Downdrag with
sons with Theory 309
Time 269
11.2 .3 Effect of PiJe Drivíng on Negative
Fríctíon 269 13 PILES IN SOIL UNDERGOING
LATERAL MOVEMENT 311
11 .2 .4 Methods of Reducing Negative
Fríctíon 269 13.1 Introduction 311
11.3 Analysis of Downdrag Forces 271 13.2 Analysis 312
11 .3 .1 Introduction 271 13.3 Typical Results 314
11.3.2 Analysis of Final Downdrag Forces 272 13.3 .1 Effect of Relative PiJe F1exibility 315
11.3.3 Development of Downdrag with 13.3.2 Effect of Boundary Conditions 316
Time 273 13.3 .3 Effect of Soil-Movement Distribu-
11.3 .4 Modiflcations to Elastic Analysis 274 tion 316
11.4 Theoretical Solutions for Single Pile 274 13.3 .4 Effect of Magnitude of Soil Move-
11.4.1 Final Maximum Downdrag Force 274 ment 316
11 .4 .2 Rate of Development of Downdrag 13.3 .5 Effect of Pi! e Diameter 318
Force 278 13.3 .6 Effect of Es and Py Distributions 318
11.4 .3 Pile Settlement 279 13.4 Application of Ana1ysis to Practica!
11.4.4 Rate of Development of Settlement 282 Problems 319
11.4.5 Effect ofPile Crushing 283 13.5 Comparisons with Field Measurements 319
X CONTENTS
14 BUCKLING OF SLENDER PILES 323 15.5 Pile Response to Earthquake Forces 353
14.1 Introduction 323
14.2 Fully Embedded Pi1es 323 16 PILE LOAD TESTS 354
14.2 .1 Basic Subgrade Reaction Theory 323
16.1 Introduction 354
14.2.2 So1utions for Constant kh 324
16.2 Maintained Loading Test 355
14.2.3 Solutions for Linearly Varying kh 325
16.2.1 Procedure 355
14.3 Partially Embedded Piles 327
16.2.2 Interpretation of Load Tests 356
14.3.1 Theoretical Approach 327
16.3 Constant-Rate-of-Penetration Test 358
14.3.2 So1utions for Constant kh 327
16.4 Method of Equilibrium 358
14.3.3 Solutions for Linearly Varying kh 327
16.5 Sources of Error in Settlé"inent Measure-
14.4 Effect of Practica! Complications 328
ments in Pile Load Tests 359
14.4 .1 Axial Load Transfer Along Pile 328
16.5 .1 Errors Resulting from Use of Re-
14.4.2 Initial Imperfections 329
ference Beam 359
14.4 .3 Inelastic Buckling 330
16.5.2 Errors Resulting from Jacking
14.4.4 Group Effects 330
Against Anchor Piles 360
14.5 Analysis Using Elastic Theory 330
16.5.3 Errors Resulting from Jacking
14.5.1 Analysis 331
Against Ground Anchors 363
14.5.2 Typical Solutions 332
16.6 Lateral Load Tests 365
14.5 .3 Comparison with Subgrade-Reaction
16.7 Torsional Testing 365
Solutions 335
APPENDIX A
15 DYNAMIC LOADS ON PILES 336
INTEGRATION OF MINDLIN'S
15.1 In troduction 336 EQUA TI ONS FOR
15.2 Estimation of Dynamic Loads 337 PILE-SETTLEMENT ANAL YSIS 366
15.2.1 Machine Loads 338
15.2.2 Wave Forces 338
APPENDIX B
15 .2 .3 Earthquake F orces 338 ELASTIC EQUATIONS USE:>
15.3 PiJe Response to Axial Loads 339 FOR LATERALLY-LOADED
15.3 .1 End-Bearing Piles 339 PILE ANAL YSIS 369
15.3.2 Floating Piles or End-Bearing Piles
w.ith Load Transfer 341
15.3 .3 PiJe Gro u ps 345
REFERENCES 371
15.4 PiJe Response to Lateral Loading 347
15.4.1 Equivalent Cantilever Systems 348
AUTHORINDEX 383
15.4 .2 Finite-Difference Analysis 348
15.4.3 Novak's Analysis 351
15 .4.4 PiJe Groups 351 SUBJECT INDEX 389
PI LE
FOUNDAT;ON
ANALYSIS.
AND
DESIGN
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
j
GENERAL PRINCIPLES
reaso11s alone is assumed to produce certain column loads, the pile base, ordinary bearing-capacity theories may be
and L1e foundations are merely required to carry these applicable. Thus for vertical failure, the shear stress at the
predete:míned loads. In reality, for complicated structures, .shaft-soil interface attains a limiting val ue (possibly varying
the load.' on the foundations determine their movement, wíth depth and soil type), and for horizontal failure result-
but this n;ovement af(ects the loads imposed by the struc- ing from lateral load or moment, the normal stress at the
ture; there is ~~:':~-~?jy interaction between structure and
1
foundation. 1rl fact; the whole complex of structural frame,
interface attains a limiting value (again, possíbly varying
with depth). In such a simplified approach, any reduction
foundation components (footings, piles, piJe caps, raft, in failure load for a particular pile because of the presence
etc.), and soil or rock forming the founding materíal, of a nearby piJe cannot be taken into account, except
tog,ether comprise one interacting structural system. The that the failure load for a group of closely spaced piles
interaction between a pile ánd its embedding soil, and that can sometimes be calculated from bearíng-capacity theory
between one pile and another piJe, provide subsets of the for buried footings on the assumption that the piles and soil
larger set of al! interacting structural components. between them act as one solid block. This load can be
If an overall structural approach is to be successful, taken as the answer if it is less than the sum of the failure
we need to know much more about a particular piJe than loads for the piles, calculated individually.
that it can be classified as, say, a 50-ton pile. We need to
know its load-settlement behavior up to failure, possibly
its behavior under lateral load and moment, and how its 1.3 .2 Elastic Theory
behavior is modified by adjacent piles. This is analogous
to saying that we need the complete load-deformation Soil and rock are not ideal elastic materials in that stress
characterístics of beams and columns, not just their load and strain are not linearly related, strains are not fully
capacities, before we can analyze complete structural recoverable on reduction in stress, and strains are not
frameworks. independent of time. However, at least i.t can be said that
Most of this book is concerned with bringing the ana- strains in soil increase as stresses increase. Furthermore,
lytical treatment of the load-deformation and the failure the assumption of anything more complicated than a
behavior of pile foundation systems to the same level of Iinearly elastic material for the soil in the pile-soil contin-
sophistication as similar analytical treatments available for uum situation would generally lead to unduly complicated
systems of structural frames. With this aclúeved, it is a theory lacking useful generality. The use of linear elastic
relatively simple matter with modern compu ter programs theory is therefore expedient and should be sufficiently
to combine the structural and foundatíon systems ínto accurate for engineering purposes, provided that elastic
. one-but thar matter is outside the scope of this book. "constan ts" are employed that are appropriate to the
particular problem. That is, they have either been back-
figured from field tests on piles in similar situations, or
determined from laboratory tests employing stress changes
1.3 BASIC THEORY REQUIRE:D
similar to the average changes in the soil mass in the partic-
ular case.
Piles embeded ín soil províde a reinforcement to the soil,
The basic elastic response of the soil from which the
increasing íts load capacity and modifying its deformation
solutions for elastic piles in elastic soil can be derived is
behavior in rnuch the same way as the steel reinforces the
given by Mindlin 's set of equations for the stresses and
concrete in reinforced or prestressed concrete members.
displacements throughou t an elastic half-space resul ting
Unfortunately, although a sufficiently accurate analysis
from horizontal or vertical point load applied at a point
of the effects of reinforcement in concrete members can
beneath the surface. As will be explaíned in subsequent
usually be obtained by adaptation of the simple theory
chapters, this bas!c ·response can be integrated to give the
of bending, the extended-continuum nature of the embed-
pile-soil interface stresses in such a way that the displace-
ding soil around piles makes the corresponding analysis
ments of the pile and soil are compatible. Modifications
of the reinforcement effect of piles much more difficult.
to take account of failure at sorne parts of the interface
are the:n relatively easy to make.
1.3.1 Failure Theory Alternatively, the elastic response can be assumed to
be that of a series of unconnected springs, that is, a Winkler
In the present state of knowledge, it is generally only medium or the subgrade reaction assumption. In spite of
possible to consider failure as something that o~curs mainly what is said by sorne of the protagonists of this approach,
at the interface between the sides of the pile and the soil, it must fundrunentally be inferior to the elastic continuum
ignoring the details of failure within the soil, although for approach of the Mindlin equations, since it ignores the
GENERAL PRINCIPLES 3
1.3.4 The Role of ldealization FIGURE 1.1 Example L Foundatíons on deep clay.
Table 1.1, from which it can be seen that the surface pad 340 kips
footing, an unlikely choice for a comparatively heavy load
on a soft clay, settles what would probab!y be an excessive
amount. Furthermore, a high proportion of the settlement
is irrecoverable, so that variatíons in load might produce
further settlement. The single large-diameter pile and the
four-pile group have similar behaviors and may even involve 2' diO
5oft Cloy
settlements that are more than satisfactorily small. The
e = 05 k S f
case of a pad with one small-diameter pile is unusual but
represents ari interesting intermediate case between pad Eu 100 k 5 f
E' 75
80 k s f
only and pUes only. At the working load, the pi! e is V 0·3
carrying its full failure-load but nevertheless succeeds in
reducing the settlement well below that of the pad on íts
own.
(e) Pad Footír>g (b) 'End
1.4.2 Vertically Loaded Foundations on Clay over Grave! plle
unaffected by the grave!, since the clay has a depth of FIGURE 1.2 Example 2. Foundations on clay over grave!.
five times the footing breadth. The pile, being "end-
bearing," can be of smaller diameter than before; in fact,
it is the strength of the concrete of the pile which deter-
mines íts diameter, rather than soil properties. The settle-
ment of the pile is now even smaller, as would be expected,
although still greater than the straight column compression.
TABLE 1.2 BEHAVIOR OF EXAMPLE
It is interesting to note that although the pile is classified
FOUNDATIONS ON CLA Y OVER GRAVELa
as "end-bearing," a third of the load is in fact taken by
2a
side shear on the shaft. For more slender piles, the pro-
2b
Example Pad footing "End·bearing" Pile portian of the load taken by the shaft can be even higher.
The fact that "end-bearing" piles are far from 100 percent
Factor of Safety 2 2 (concrete strength
end-bearing has been verified in the field in severa! instances.
governs)
Percent Pad lOO For the same vertical lcrad as in the previous examples,
working Shaft 33 but also with a horizontal load and á moment applied,
load taken Píle base 67 a pad and pier are compared in the third example (Fig.
by: 1.3). The results of calculations are given in Table 1.3.
Settlements Immediate 4.1 in. in In order to carry even a relatively small moment, the pad
at Percent ímmed. 56% o footing has had to be enlarged to an inordinate size, and
working resulting in view of the movements and rotations at the working
load from yield loads, it wouh1 unlike1y to be considered a satisfactory
Consol. 1.2 in. o solution. The length and diameter of the pier have been
Total final 5.3 in. 0.5 in.
selected so that the factor of safety against failure resulting
a See Fig. 1.2. from the vertical load is the same as that against failure
b Elastic shortening of piJe as a column ·:. 0.25 in. resulting from the horizontal load and moment.
GENERAL PRINC!PLES S
rt. 3a 3b
Example Pad Footing Single Pier
Factor of Safety 2 2
IJ
EFFECTS OF INSTALLATION OF PILES
6
EFFECTS OF INSTALLATION OF PILES 7
where
1.5 -
!lum maximum excess pore pressure
Ka in-situ coefficient of earth pressure at r<:st
Su undrained shear strength
A¡ = pore-pressure coefficient A at failure
, = ini tia! vertical effective stress in soil
Ovo
0.5 -
(a) The Lo and Stermac expression (Eq. 2.1) is used to near the top and tip of the pile being ignored. The relevant
obtain the maximum pore pressure C:.u 111 , from the face equation of consolidation is then
of the pile to a dístanc~ R from the face. On the basis
of Fig. 2.2 and also the analxsis 'of Nishida (1962), R
(2.3)
varies from 3l:1 to 4a for insensitive clays, to 8a for sensitive
clays.
(b) Beyond the distan ce R, the excess por e pressure is where
assumed to vary inversely as the square of the distance r
from the pile, thatis, eh is the two-dimensional coefficient of consolidation
for horizontal drainage
u is excess pore pressure
(2.2)
The above equation may readily be written in finite
The in verse vanat10n is predicted from elastic theory, difference form (e.g., see Gibson and Lumb, 1953), and
as utilízed by Ladanyi (1963) and Nishida ( 1962). solved for. the appropriate drainage condition at the pile
(e) For pile groups, the pore pressure distributions around and initial pore-pressure distribution. Solutions for the
individual piles may be superposed, except that the pore excess pore-pressure dissipation at the pile face, for an
pressure cannot exceed C:.um, as found by Lo and Stermac impermeable pile, were obtained by Soderberg, who found
(1965). that the form of the initial pore-pressure distribution had
a relatively small influence on this solution.
The excess pore pressures around a pile in sens1t1ve A reasonable measure of the rate of strength or adhe-
clay as calcuhited by the above procedure, agree well with sion-regain after driving appears to be to consider the rate
the average observed curve in Fig. 2.2. of consolidation within a limited are a in the vicínity of the
piJe. Such solutions are shown in Fig. 2.3, assuming an
initial excess pore-pressure distribution as suggested in
2.2.3 Dissipation of Excess Pore Pressures Sectíon 2.2.2 and a failure zone having ratios of radius
R to piJe radius a of 3 and 5. The degree of consolidation
A relatively simple solution for the rate of dissipation within a radius "f R is shown for both a permeable andan
of excess pore pressures around a driven pile has been pro- impermeable piJe.
posed by Soderberg (1962a). It is assumed that dissipation It is interesting to compare these theoretical solutions
occurs radially only, the vertical dissipation that may occur with an empírica! relationship suggested by Radugin ( 1969).
Assumad limit
0·2 --...._ of farlur\2
¡'' \ zone
,~¡
0-4
\,_
R¡
Üp
with1n
a ~r ~R R
- - - 0=3
0·6 - - - B a. :5
0·8
FIGURE_ i.J. Theoretical sotutions fo-r rate of consolidation near a: driven pile.
10 EFFECTS OF INSTALLATION OF PILES
Tírne ( days)
0·2
.5 ~
o
§ ~
15 ::! 0·4
H
"1::)
o e
~ .Q
OL"'
u o 'O
.2! 0·6
o ~
"' o
"'g' o 0·8
L 01
o n::
1·0~----------~------------~------~--~--~
FIGURE 2.4 Comparison between empirical and theoretical solutions for rale of adhesion increase.
Assuming R/a 5, c¡2 0.04 sq in./min (typical of a 2.2.4 Displacement Caused by Driving
medium clay), and a permeable concrete piJe of 6-in.
radius, Fig. 2.4 compares the theoretical degree of consoli-
dation versus time curve and Radugin's empírica! curve, PiJe driving generally causes a he ave of the el ay surrounding
assuming the rate of consolidation is the same as the rate the pi! e, followed by consolidation of the clay. This move-
of adhesion increase in Radugin's relationship. There is ment caused by pile driving may have a significant effect
sorne difference between the shape of the curves, but on adjacent structures and may also cause the piles driven
they are generally in sufficient agreement to suggest that earlier in a multiple-píle installation to rise during the
the simple consolidation analysis provides a reasonable driving of the later piles. Under these circumstances,
estima te of the rate of increase of load capacity. redriving of the earlier piles is often considered necessary,
From a practica! point of view, solutions such as those or may lead to a decision to use bored rather than driven
in Fig. 2.3 are of most use in giving an estímate of the time piles. The limited data available on the magnitude of the
that should elapse after driving before a load test is carried heave is rather conflicting, although much of the conflíct
out, if a reliable estímate of t~e ultimate undrained load may arise from differences in soil types in the various
capacity and load-settlement behavior is to be obtained. investigations. The ratio of the total volume of initíal
A more rigorous analysis of the stress changes, excess heave to the total volume of driven piles within a founda-
pore pressures, and subsequent consolídation around a tion has been found to be a.bout 100% by Adams and
driven pile in clay has been presented by Wroth et al. Hanna (1970) for steel H-piles in a firm till, 50% for piles
( 1979). The pile-driving process is modeled as the creation in clay by Hagerty and Peck (1971), 60% by Avery and
of a long cylindrical cavity by radial soil movement. Values Wilson (1950), and 30% by Orrje, and Broms (1967) for
of stress and pore-pressure change have been obtained precast concrete piles in a soft, sensitive, silty el ay. The
using a finite-element analysis incorporating a work· latter investigators found that the heave near the edge of
hardening soil model (the Cam-clay model). It is concluded the foundation was a&out 40% of the value at the center.
· that the total and effective stresses a:djacent to the pile Outside the edge of the group, only very small heaves
just after dríving may be related directly to the original were noted by Adams and Hanna, and Orrje and Broms.
undrained strength of the soil, and are essentially indepen- Adams and Hanna measured radial and tangential move-
dent of the overconsolídation ratio: The final stress state ments as well as vertical heave, and found that the maxi-
after consolidation is similar to that in an oedometer mum radial movement was about 1.5 in., and the maximum
(K 0 ) test, except that the radial stress is now the major tangential displacement about 0.4 in.-both these values
principal stress. being considerably less than the average vertical heave of
EFFECTS OF INSTALLATION OF PILES 11
about 4.5 in. As with vertical heave, very small lateral sensitíve clays than for corresponding piles in insensitive
movements occurred beyond the edge of the group. clays, and that when piles penetrate alternating strata of
Measurements of the- movement of an exísting building fme-graíned soil and granular materials, the observed
as a result of driving of piles for the foundations of a surface-heave may be much less than that which would
new building were reported by Lambe and Horn (1965). have occurred in insensitive clay soils. It was also found
It was found that, at the near corners of the existing that if the sequence of piJe driving involved driving piÍes
building, a heave of about 0.3 in. occurred duríng driving, first along the perimeter of the foundation, the heave of
but that by the end of construction, a net settlement of the soil surface in the central area of the foundation is
about 0.35 in. had occurred. Despite the fact that the increased and that of the surrounding area correspondingly
piles were preaugered to wíthln about 30ft of their final decreased. Observations also were made of lateral move-
elevation, high excess pore-pressures (maxinmm of. about ments, and it was found that driven piles tended to be
40 ft of water) were measured near the comer of the displaced away from subsequent dríving, with rr;ovements
existing building, even before a substantial building load continuing for a considerable length of time after com-
was carried by the piles. pletion of driving. Where large differences in elevation
Figure: 2.5 shows sorne measurements of heave and existed withln the foundatíon area, pile driving often
settlement of buildings caused by pile driving, as reportéd displaced the soil preferentially toward the areas in which
by D'Appolonia and Lambe (1971). The settlement data the lower elevations occurred.
plotted are for net settlement one to three years after the
end of construction. Larger movements than those mea- Estimation of Dísplacements
sured by Lambe and Horn were found, although the piles Lambe and Horn (1965) proposed a method of estimating
were again preaugered to within 20 to 30 ft of the final the heave and subsequent settlemeRt of the surface of soil
tip elevation. near a piJe resulting from driving of the pile. Although
From measurements of displacements resulting from the method was found to predict movements considerably
pile driving in clays, Hagerty and Peck (1971) concluded larger than those measured, it appears to be a logical pro-
that the soil displacements are less for piles driven in cedure and worthy of further application. The method
;>,
n
'-
0 "
:;l· " Not<Z ·
X
)(
z"' •
S<Zttlrzm<Znt 15 mrzasur<Zd trom
0·051--~--:.~-,A--~-=---th<Z oríg¡nol pr<Z-construct,on _ __
<ZI<Zvat1on, not trom th<2 top
Ce: Of th<Z h<ZOV<Z. 1
"' E"'
E " •
~ !l •
~~<11 • • !symbol
E 0·10
::J
E
X
o ... ·¡¡ o
)( 0·0150
Rdríg<Zratíon
Mat<Zrials
¿ •• 1
• 0·0155 Spac<Z C<Znt<Zr
1
0·15 L__ _ _ __L__ _ _ _....L_ _ _ __J._ _ _ ____¡
FIGURE 2.5 Movements of nearby buildings caused by piledriving operations (D'Appolonia and Lambe, 1971).
12 EFFECTS OF INST ALLA TION OF PILH
/
7J 3
\
\
\2
\
testing, is to divide the soíl into a number of layers, test
samples from each of these layers, and add up the resulting
displacements computed for each !ayer to obtain the overall
2 \ displacemen t.
Of course, the movement of the surface of the soil,
discussed above, is not necessarily the same as the move-
Measur<2d or ment of the top of an existing pile caused by the installa-
1 calculat<2d tion of an adjacent pile. A very simple approximate pro-
du<2 to
cedure for estimating the heave of an existíng pfle was
proposed by Hagerty and Peck ( 1971 ). This procedure~
Stc;ep 1 D<2p0s1t format•on was based on the concept that inextensible vertical piles
St<2ps ·1o ,'b Undra1nczd loading and consolidation embedded in the clay would be lifted by the relative rise
due to build1ng ( 1f appiiCOblcz)
Step 2 Pi1<2 drtv1ng SlffiUIOtiOn of the soil along the uppei part of the piJe, but that along
St12p 3 DlSSipOtlon subsequcznt to dnv•ng the Iower part of the pile, a downward force would act,
FIGURE 2.6 Test pro~edure far displacement calculatian (Lambe tending to reduce the total uplift of the pile (Fig. 2.7).
and Horn. 196S). (©Ca nada, 1965 by University af Taranta Press.)
A surface a-a can be found af which the relative movement
between soil and piJe is zero. The pile heave is considered
is b8sed on the stress-path approach advocated by Lambe to be approximately equal to the ~eave of the soil, on the
( 1964). 1t consists of estimating the effective stress-path assumption that no heave takes place be!ów a-a; that is,
for an element in the compressible soil !ayer,
(L · d¡¡) ( ·¡ h )
running a laboratory test on a sample of thís soíl such that Pile heave =--L- Sm eave (2.4)
the loading follows the stress path estimated for the field
element, and using the Jaboratory-measured value of The depth, dh, is estimated by balancing the poten tia!
vertical strain to estímate the building heave and settle- upward and downward adhesive forces on the upper and
ment. lower parts of the pile.
There are essentially three steps in the procedure, For simple distributions of plie-soil adhesion, d¡¡
which ís shown diagrammatically in Fig. 2.6: may be calculated directly (for example, for uniform
adhesion, dh = O.SL), so that the pile heave would be
l. Consolidation of the sample (usually under Ka con- estimated as one half of the soil he ave; while for linear! y
dítions) to the in-situ stresses in the !ayer prior to pile increasing adhesion, from ze:o at the surface, dh 0.67 L,
driving. Where the element under consideration is beneath and the pile heave is estimated as one third of the soil
an exísting building, as it was. in the case described by heave.
Lambe and Horn, this first stage involves simulation of the Hagerty and Peck found good agreement between
undrained loading caused by construction of the building observed piJe heave and estimates based on the above sim-
followed by consolidation, after initial K 0 consolidation ple approach.
of the element to the field stress state.
2. lncreasing the total lateral stress at constan! total
vertical stress until the pore pressure in the element equals
that measured by the field piezometers or that calculated rlczoving
soil
/
on the basis of Section 2.2.2. The vertical strain measured
during this stage al!ows the heave to be computed.
3. Dissipation of the pore pressures developed by the
pile driving, that is, consolidation of the sample. Lambe a ---
and Horn suggest that thís process might take place under Stat1onary
soH
conditions of no lateral strain, but this suggestion appears
questionable. The movement measured in this stage will
enable the total consolidation, subsequent to the heave, FIGURE 2.7 Balance of forces along driven piJe (Hagerty and Pecl
to be computed. 1971).
EFFECTS OFINSTALLATION OF PILES 13
A more refined analysis of pile movement caused by sand (Dr = 35%), the extent of movement was somewhat
adjacent pile-driving may be made using the approach larger, extending 4.5 to 5.5 diameters from the side and
described in Chapter 12 for piles in swelling or shrinking 3 .O to 4.5 diameters below . the tip. These figure.s aie in
soils. broad agreement with tft.ose found by Meyerhof (1':159).
The tests of Robinsky and Morrison also showed that
the process of sand displacement and compaction below
a pile tip is followed by sand movements adjacent to the
2.3 EFFECTS OF PILE DRIVlNG IN SANDS pile sides. These movements tend to decrease the sand
density in the immediate vicinity of the sides and thus
2.3.1 Single Piles nullify sorne of the benefits gained by the primary com-
paction. The pattern of displacements around a typical
When a piJe ís driven into sands and cohesionless soils, pile, as found from the radiographs, is shown i~ Fig. 2.8,
the soil is usually compacted by displacement and víbra- while the strain pattern deduced from these displacementS
tion, resulting in permanent rearrangement and sorne is shown in Fig. 2.9 (Vesic, 1967). The decrease in density
crushing of the particles. Thus, in loose soils, the load occurring above the tip is clearly reflected in the tensile
capacity of a piJe is increased as a result of the increase In strains, which amount to about half of the maximum
relative density caused by the driving, and installation by compressive strains below the tip. The above remarks
driving rather than boring has distinct advantages. Detailed apply to a straight-sided pile, but Robinsky and Morríson
investigations of the extent of compaction of sand and the found that the same process occurs with a tapered pile.
increase in relative density around the pile have been Their tests did not indicate that the loosening effect was
carried out by Meyerhof (1959) and Robinsky and Morrison markedly compensated by the wedging action of the pile-
(1964). taper compacting the surrounding sand. The higher load
Robinsky and Morrison conducted a careful series of capacity of a tape red píle can probably ·be attributed
model-pile tests in sand in which the displacement and com- mainly to the greater normal stresses developed between
paction around the piles was studied by means of radio- the pile and the soil when loaded by the foundation.
graphy techniques. lt was found that in an initially very On the basis of an empírica! correlation among density,
loase sand (relative density Dr = 17%), soil movement penetration resistance, and friction angle, Meyerhof (1959)
extended 3 to 4 pile diameters from the side of the pile and devised a method of estimating the extent of the zones of
2.5 to 3.5 díameters below the pile tip. In a medium dense increased density, and hence increased friction angle,
9. . . . . .
8~
H-~--
1
1
'
·.
1
1/ 7
~/
/ 1
··~/
// i
~ :
13 .....,._
:
: 1\ 1
-~
'1
1
1
l
40 20 o 20 40
Vertical Stroih o/o
FIGURE 2.8 Displacements around driven pile in sand (after FIGURE 2.9 Strains around driven pile in sand (after ,Robinsky
Robinsky and Morrison, 1964). (Reproduced by permission of the and Morrison, 1964). (Reproduced by permíssion of the National
National Research Council of Canada from the Canadian Geo- Research Council of Canada from the Canadian Geotechnical Jour-
technical Journal, Vol. l. 1964, p. 81.) nal, VoL 1, 1964, p. 81.)
14 EFFECTS OF INSTALLATION OF PILES
=
When <f¡'1 40° in Eq. (2.6), there is no change in relative
density due to pile driving.
When groups of piles are driven into a loose sand, the soil
around and between the piles becomes highly compacted,
and if the pile spacing is sufficiently close (less than about
six diameters), the ultirnate load capacity of the group
rnay be greater than the sum of the capacities of the indi-
vidual piles-that is, the efficiency of the group is greater
than l. On the other hand, if the sand is so dense that
Observed Results: F•na!._ P<Zn<Ztrot1on R<Zslstonc<Z --()
Ortgmal
pile driving causes loosening rather than cornpaction, the
Est1motad Rasults; Angi<Z ol friction ~· O group efficiency may be less than l.
Ma)Or principal stress rat1o --() An estímate of the effects of driving a group of piles
Test i'Jo 11, Worce ster, Mass into loose sand may be made by application of the
FIGURE 2.1 O Observeil vs. computed compaction of sand near approach suggested by Kishida ( 1967) for single piles,
piJe (after Meyerhof, 1959). assurning that superposition of the effects of single piles
is applicable. In applying Eq. (2.6), the value of if¡~ is the
changed value caused by previous piles. By application of
around a piJe driven in sand; a typical result compared wíth this approach, a rough estímate rnay be made of the effect
observátions is shown · in Fig. 2.1 O. These results are in on ultimate load capacity of the order of installation of
broad agreement with those of Robinsky and Morrison the piles. It has been found in practice that piles driven
( 1964), but according to Meyerhof, the amount of com- la ter have a greater load capacity than those driven earlier.
paction near the tip is greater, and that near the top of the Sorne field measurements of the amount of cornpac-
shaft is less. tion caused by the driving of a group in a granular soil, in
A sirnpler method of estimating the effectsof driving which standard penetr~tion tests have been carried out
a pile in loose sand in the vicinity of the tip is that pro- before and after driving of groups have been reported by
posed by Kishída (196 7). On the basis of field and model Phílcox (1962). The test results are shown in Fig. 2.12. In
test-results he assurnes that the diameter of the compacted case (a), the standard penetration number, N, n~ar the
zone arou;td a pile is 7d. Within this zone, he furiher center of a four-pile group, was more than doubled by
assumes that the angle of friction if¡' changes linearly with driving. In case (b ), the in crease in N for a point a little
1
distance from the original value of t/1 1 at a radius r = 3.5d away from the center of a nine-pi!e group shows a relatívely
to a maximurn value of if¡~ at !he pile tip, as shown in Fig. smaller increase (average about 75%). Cases (e) and (d)
2.11. show that the increase in N becomes less as the point
The re la tionship between if¡ 1 ' and <P/ is taken to be
considered becomes more distant from the center of the .
group. Another point apparent from Figs. 2.12b and 2.12c
(2.6) is that the effect on driving on N, and hence on soíl density,
is greater below the tip than along the shaft.
In order to relate the increase in N lo the increase in
if¡, Kishida (196 7) suggests that if¡'1 and N m ay be related
by the following expression:
(2.7)
The differences between the degree of densification
at various poínts within a group, as shown in Fig. 2.12,
suggest that the load capacity of piles near the center of
the group rnay be greater than those }1e3i the e~ge of the
group, and that the load distribution, even at working
loads, rnay be nonuniform, with larger loads being carried
FIGURE 2.11 Effect of driving on q, (Kishida, 1967). by the center piles~as predicted by Kishida's approrch.
EFFECTS OF INSTALLATION OF PILES 1,5
~ ;;
""
- ¿
¡;!
.S
e
e Q
o (j
g60 ~
g 60
"e
"
ll.
&
80 80
2'1Ó Split Spoon (140"'- 3') 2'</> Splít Spoon (140#- 30')
Kimb.,ri<>Y Rd. / Austin Avo. Kowloon Kímb<>rl<>y Rd j
Aust1n Avo J(owloon
(C) (d)
FIGURE 2.12 A comparison of N Values-bcforc and after driving piles (Philcox, 1962). (Reproduced by pennission of the lnstitution of
Structural Eng;:neers, London.)
This behavior, which is in contrast to that usually observed the soil. The adhesion has been found to be less than the
for groups in clay, has been observed in tests carried out by undrained cohesion befare installation, mainly :.•ecause of
Hanna (1963) and Beredugo (1966). As suggested by softening of the clay immedlately adjacent te the soil
Kishida ( 1967), the effects of differing compaction may surface. This softening may arise from three causes:
also explain the dependence of pile-load distribution on
the arder of dríving piles in sand. (a) Absorption of moisture from the wet concrete.
(b) Migration of the water from the body of the clay
toward the less highly-stressed zone around the borehole.
2.4 EFFECTS OF INST ALLING BORED PILES (e) Water poured into the boring to facilitate operation
2.4.1 Clay Soils of the cutting tool.
The effects of installing bored piles in clay have been stud- Factor (e) may be eliminated by good dr~lling technique,
ied largely in relation to the adhesion between the pile and · and (b) can be minimized by carrying out the drilling and
16 EFFECTS OF INSTALLATION OF PILES
p
20000
A 1 1
T
"'
K 1 1
L
w
10000
EE~
8000 G,¡- 1
1
~é>i"
6000
8 ~ ..... n ¡
5000
'V '•, .C>
4000
~
"
~
.o
6
~' ~ 6
1
.e 1 ¡ 1
1
r~~ ~
f
',;
e 2000
1 l i 1
~
:;;
: ~~""" l_
"'
.e
A-P Tunnels
!
!o~
(J) o Brandon '* """.....
1000 o Paddington
800 + Kensal G reen
¡--~ - l* Camberwe\1 f.--- - '
'
600
500
400
++ Finsbury
0 South Ba~k
o Victoria
1
1
---
Á Míllbank 1
Y Galway St. 1 i 1
: & St. Pauls 1
200 1 1 1 1 1 '
20 22 24 26 28 30 32 34 36 38 40
Water content, %
FIGURE 2.13 Relation between shear strength and water content for London clay (Liquid Limit 1 70-85) (Sk~mpton, 1959).
concreting operations as rapidly as possíble. Sorne effects A further effect of installing a bored piJe is that the
from factors (a) and (b) are consídered to be inevitable by clay just beneath the pile base may be disturbed and
Skempton (1959), but their seriousness will depend largely softened by the action of the boring tools. The effects of
on the technique employed, whether or not casing or this disturbance may result in increased settlements, espe-
drilling fluid is u sed to support the' walls during construc- cially for belled piers, in which the base carries ,a major
tion, and the time taken for construction of the pile. proportion of the load; hence, it is important to clean out
Palmer and Holland (1966) contend that softening in over· the base thoroughly. However, as stated by Skempton
consoliiiated London clay is minimized if drilling and con- (1959), base disturbance and softening should have a
creting is carried out within one or two hours. negligible effect on the ultimate bearing capacity of the
Meyerhof and Murdock (1953) measured the water base because of the comparatively large mass of clay
contents of the clay immediately adjacent to the shaft of involved when the base penetr¡¡tes the el ay. In contrast,
a bored pilc in London day and found an increase of the shearing process developed in the clay along the piJe
nearly 4% at the contact surface, although at a distance of shaft is probably restricted to the narrow softened zone.
3 in. from the shaft, the water contents had not altered. Construction problems may also arise with bored
This increase should be a maximum value, as the hole was piles, and a number of these have been described by Pandey
drilled by hand and took two to three days to complete. (1967) in relation to the foundations for a heavy industrial
An estímate of the effect of the increase in water content building, including the following:
can be made if information is available on the relation
between shear strength and water content. Such a relation- (a) Caving of the borehole, resulting in necking or misalign-
ship for London clay has been presented by Skempton ment of the piJe.
(1959) and is reproduced in Fig. 2.13. These results show
(1)) Aggregate separation within the piJe.
that an increase in water content of only 1% results in a (e) Buckling of the pile reinforcement.
20% change in the ratio Ca fcu, of undrained adhesion Such structural defects may be difficult to detect, since a
ca to original undrained strength eu, while for a 4% in crease load test may not reveal any abnormal behavior, especially
in water content, cafcu is reduced to about 0.3. Values if the load is only taken to the design load.
of Ca/cu. for bored piles are discussed in detail in Section Barker and Re ese ( 1970) investigated the influence of
3.2. drilling fluids on the performance of bored piers. They
EFFECTS OF INSTALLATION OF PI LES 1 'l
concluded that when proper construction techniques are soil for severa! feet below and around the pile. Thus, in
ernployed, drilling rnud has no detrirnental effects on the calculatíng the load capacity of a bored piJe in sand,
load-carrying characteris.tics of a borl;(d shaft. The con- Tornlinson (197 5) suggests that the ultirnate val ue of
crete properties and the concrete-placernent procedure angle of shearing resistance lji should be used, 'unless the
are the two rnost critícal factors involved in the construc- pile is forrned in a dense grave! when the "surging" effect
tion process. Elirnination of the effects of the drilling rnay not take place. If heavy cornpaction can be given to
rnud is accornplished when it is cornpletely displaced by the concrete at the base of the piles, then the disturbed
the concrete, resulting in a vigorous scouring of the bore- and loosened soíl rnay be. recornpacted and the value of
hole wall by the rising concrete. Should drilhng rnud be lji for the dense state used., However, if the shaft is
trapped between the concrete and borehole wall, it would obstructed by the reinforcing cage, such cornpaction rnay
virtually eliminate the developrnent of any shear-load not be possible.
transfer in the vicinity of the trapped drillirtg rnud. The Tests on bored piles in sand have been reported by
use of casing in placing the concrete involves a greater risk Tourna and Reese (1974) and Clernence and Brurnund
of trapping drilling rnud than does the procedure of placing (1975). Tourna and Reese found evidence of the arching
the concrete under the drilling rnud by the use _of a trernie that occurs around the pile with driven piles (see section
or concrete purnp. Barker and Reese suggested that reduc- 3 .2) and that results in the developrnent of lirniting values
tion factors of 0.6 for clay and 0.8 for sand and silts should of skin frictíon and base resistance at depth. It was also
be applied to the shear strength in the design of drilled found that the skin resistance, for piles penetrating less
shafts. However, no reliance should be placed on load than 25 ft, could be correlated with the integral around the
transfer developing within three shaft diameters of the pile periphery of a~0 tan lji' (where a~0 ""' effectíve over-
surface or one diarneter of the base. burden pressure ), using a reduction factor of about 0.7.
There were indícations that srnaller reduction factors are
2.4.2 Sands appropríate for greater penetrations. Frorn a large-scale
test on a bored pier in sand bearing on a sirnulated rock
There is relatively little quantitative inforrnation on the base, Clernence and Brurnund (1975) found that 20 to 30%
effects of installing bored piles in sands or cohesionless of the design axial load in "end-bearing" drilled piers was
soils. Such piles usually require casing or drilling fluid carríed by the pile skin. A roughly linear increase in skin
to support the walls of the hole and sinking of the hole, friction with depth was rneasured, except near the lower
and subsequent withdrawal of the casing while conc~eting part of the pier, where a sharp increase in skin friction was
the shaft is likely to disturb and loosen the soil to sorne noted, presurnably because of the confining effect of the
extent. Also, sorne loosening is hable to occur at the rock base. It was found possible to use the results of direct
bottorn of the pile as a r~sult of baling or "shelling-out" shear tests for the soil-pier interface rnateríals to predict
the hole, and when this is done under water, the upward the lirniting skin friction, except near the tip, where the
surge on withdrawal of the baler or shell can loosen the calculated skin friction was lower than that rneasured.
ULTIMATE LOAD CAPACITY OF PI LES
3.1 INTRODUCTlON shaft and base resistances, less the weight of the pile; that
is,
There are two usual approaches to the Galculation of the
(3.1)*
ultimate load capacity of piles: the "static" approach,
which uses the normal soíl-mechanics method to calculate
where
the load capacity from measured soíl properties; and the
"dynamic" approach, which estimates the load capacity
Psu ultimate shaft resistance
of driven píles from analysis of pile-driving data. The
Pbu ultimate base resistance
first approach will be described in detall in this chapter,
W weight of pile
and the second in Chapter 4.
In this chapter, a general expression for the ultimate
Psu can be evaluated by integration of the pile-soil
load capacity of a single pile is given and its application to
shear strength Ta over the surface arca of the shaft. T a
piles in clay and sand is described. Approaches for groups
is given by the Coulomb expression
of píles in clay and sánd wíll then be outlíned. Other topics
include the design of piles to rock, the use o.f in-situ tests
Ta = Ca+ On tan rPa (3.2)
such as the standard penetration test and the static cone
to estímate píle-load capacity, the calculation of uplift
where
resistance of piles and grou'ps, and the load capacity of
bent píles.
Ta pile-soil shear strength
Ca adhesion
3.2 ULTIMA TE LOAD CAPAClTY OF SINGLE PiLES On normal stress between pile and soiJ.
rPa angle of friction between piJe and soil
;3.2.1 General Expression • lt is an implicit assumption of Eq. 3.1 that shaft and base
resistance are no! interdependent. This assumption cannot be
The net ultima te load capacity, Pu, of a. single pile is strictly corree!, bu! there is !ittk doubt that it is corree! enough
generally accepted to be equal to the sum of the ultimate for practica! purposes for all norrnal-proportion piles and piers.
18
ULTIMA TE LOAD CAPACITY OF PILES 19
an ís in turn frequently related to the vertical stress av, base resistance of piles, reliance has to be placed on approx-
as imate theoretical or semiempirical methods. With regard
to: sands, these methods ha ve be en reviewed by Ve sic
(3.3)
(1967), who fmmd that the solution of Bcrezantzev et al.
(1961) gene rally fitted experimental results bes t.
where From Eqs. (3.1), (3.5), and (3.6),
Rigorous solutions for the bearing capacity of surface For píles in clay, the undrained load capacity is generally
footings using the methods of classical plasticity are now táken to be the critica! value unh~ss the clay is híghly
well-established (Prandtl, r 1923; Sokolovskii, 1965; Cox, overconsolidated. (Burland, 1973, contends, however,
1962; Lundgren and Mortensen, 1953; Davis and Booker, that an effective stress-drained analysis ís more appropríate,
1971 ), and t~e only doubts regarding the practica! validity as the rate of pore-pressure dissipation ís so rapid that
of these solutions lie in the possible effects of the differ- for normal rates of load application, drained conditions
ences between the behavior of real soil and fhat of the ideal generally prevaíl in the soil near the pile shaft.) If the clay
material assumed in the theory. At the presen t time, there is saturated, the undraiqed angle of friction 1/>u is zero,
are few if any classical plasticity solutions that are relevant and 4>a may also be taken as zero. In additíon, Nq = l
to a buried footing, and therefore, for the calculation of and N1' =O for tJ> =O, so that Eq. (3.7) reduces to
20 ULTIMA TE LOAD CAPACITY OF PILES
Penetration
Further simplification is possible in many cases, since
for piles without an enlarged base, Abavb = W, in which Case Soil Conditions Ratiob Ca/e u
case Sands or sandy soils <20 1.25
overly íng s tíff
(3.1 O) cohesíve soils >20 See Fig. 3.2
25
0.8
0.6
e,
Cu
0.4
Tomli~ '--..,__
0.2 .
.....__
---
Undrained Cohesion e" kips/sq ft
FIGURE 3.1 Adhesion factors for driven piles in cl,ay (after McC!elland, 1974).
ULTIMA TE LOAD CAPACITY OF PJLES 21
ÜL---------------~--------------~----~--------~~
1000 2000 3000 4000
Undrained shear strength !e") ltJ/ft 1
FIGURE 3.2 Adhesion factors for case l (sands and gravels overlying stiff to very stiff cohesivc soils) (Tomlinson, !970).
FIGURE 3.3 Adhesion factors for case III (stiff to very stiff clays without overlying strata) (Tomlinson, 197 0).
values of ca/cu is evident. This scatter ís often attributed be little data on appropriate values of Ca for driven piles
to the effects of "whlp" during driving. A more complete founded in very sensitive clays, and the extent to whlch
investigation of adhesion for driven piles in stiff clay has "set-up" compensates for remolding can at present only
been made by Tomlinson (1970), who found that ca/cu be determined by a load test.
may be markedly influenced by the soil strata overlying For bored piles, the available data on ca/cu is not as
the clay, as well as by the value of Cu. Tomlinson has sug- extensive as for driven piles, and much of the data that is
gested the adhesíon factors shown in Table 3.1 and Figs. available is related to London clay-" Table 3.2 gives a sum-
3.2 and 3.3 for cu > 1000 lb/sq ft (48 kPa). The most mary of adhesÍon factors, one of wlúch is expressed in
notable feature of Tomlinson's results are the high values terms of remolded strength, e,, as well as the undisturbed
of ca/cu for case I, where sand or sandy grave! overlies undrained strength, Cu. Results obtained by Skempton
the clay, because of the "carrying down" of á skin of the (1959) and Meyerhof and Murdock (1953) suggest that
overlying soil into the clay by the pile. There appears to an upper limit of Ca is 2000 lb/sq ft (96 kPa).
22 ULTIMA TE LOAD CAPACITY OF PlLES
TABLE 3.2 ADHESION FACTORS FOR BORED PILES IN CLAY A somewhat different approach to the calculation of
· the ultirnate shaft capacity Psu has been adopted by Vijay-
Adhcsíon vergiya and Fochi (1972) for steel-pipe piles. From an
Soil Type Factor Value R-eferencc examination of a number of loading tests on such piles,
they concluded that Psu can be expressed as follows:
London clay 0.25-0 7 Golder and Leonard
Average, ( 1954)
0.45 fomlínson ( 195 7)
(3.10a)
Skempton (195?)
----~·· where
Scnsitive cby Go1der (1957)
o~ mean effective verticai-·stress between ground
Highly exp~nsive Ca/e u 0.5 Mohan and Chandra
el ay (1961) surface and pile tip
Cm = average undrained shear strength along pile.
'A
o 0.1 0.2 0.3 0.4 0.5
o
•
•
tJ.
t!l
p'>l!
!-·
o.
•
1
Locctiun Synthol Source
Detrol! o r Housel
Morganz.a Ma:isur
Cleveland o Peck
Dray:on X Pe ek
:Jorth Sea t. Fox
Lemowe 0 Woodward
••
Stan-nore Torr linson
0Je\v Orit~ans Blessey
Ven ice
Alliance
•
'V
McCielland
McCielland
Do'laidsonville
MSC Hocston
• Darragh
-- McCietland
San Francisco Seed
17:J
Bt~:ish Columbia McCamrnon
Burnside
• Peck
200
l'IG!JRE .3.4 Fricti•mal capacity coefficíent le vs. pile penetration (Vijayvergiya and Focht, 1972).
ULTIMA TE LOAD CAPACITY OF PILES 23
¿¡¡
<U
o::l
4
3
3JJil1w·;77. 3.2.2.2 DRA!NED LOAD CAPACITY
A was found to be a function of piJe penetration and is Burland (1973) discusses appropriate values of the
plotted in Fig. 3.4. combined parameter {3 = Ks tan q. ~ and demonstrates that
Equation (3.10a) has been used extensively to predict a lower lirnit for this factor for normally consolidated clay
the shaft capacity of heavily Ioaded pipe-piles for offshore can be given as
structures.
{3 = (l - sin cp') tan cp' (3.13)
Bearing Capacity Factor N e
The value of Nc usually used in design is that proposed by where
Skernpton (1951) for a circular area, which increases frorn
6.14 for a surface foundation to a lirniting value of 9 for 'rp' = effective stress friction angle for the clay
length ? 4 diarneters (Fig. 3.5). The latter value of Nc
9 has been confirrned in tests in London clay (Skempton, For val u es of rp' in the range of 20 to 30 degrees, Eq.
19 59) and has be en widely accepted in practice. However, (3.13) shows that {3 varíes only between 0.24 and 0.29.
differing values have been found by other investígators; This range of values is consistent with values of {3 Ks
for example, Sowers (1961) has found 5 < Nc < 8 for tan rp ~ deduced frorn measurements of negative friction
rnodel tests, and Mohan (1961) has found 5.7 < Nc < on piles in soft el ay (see Figs. 11.26 and 11.2 7). Meyerhof
8.2 for expansivt: clays. The variations in the va1ue of (1976) also presents data that suggests similar values of
N e m ay well be associated with the influence of the stress- {3; however, there is sorne data to suggest that ¡3 decrea~es
24 ULTIM~TE LOAD CAPACITY OF PILES
.~
1
!'--......_
-........ ..........___ Pile díaleter 4.0 L. l
20
~
1
1
¡ \\ i
\ !--...
~ 60
1
e
1
\ ·, ~
! 1
"'"'1'\
1 1
1 :\ Field tests
80 ...
1
• _¡ \ lloose, moist, sánd)
! 1 \ G-41
100 : i i l. \
\'
1
0 1 1
Llose Medium dense sand \
sand IG~2) \
IG-1)
1
120
j
1
··---· 1 le
1
e
1
o\
140 1 1 1 1 1 1 1·
10 20 30 40 50 60
Point resistance (ton/ft 2 1
FIGURE 3.7 Varíation of point resistanee with piJe length (Vesic, 1967).
ULTIMA TE LOAD CAPACITY OF PIL.ES 25
assume that the vertical stresses a,. and Ovb in Eq. (3.7)
are the effective vertical stresses caused by overburden.
However, extensive research by Vesic (1967) and Kerisel
( 1961) has revealed that the unit shaft and base resistances
of a pile do not necessarily increase linearly with depth, o' ve
but instead reach ahnost constant values beyond a certain
depth (Figs. 3.6 and 3.7). These characteristics have been
confirmed by subsequent research (e.g., BCP Comm., L
1971; Hanna and Tan, 1973). Vesic also found that the
ratio of the limíting unit point and shaft resistances, fbffs.
cf a pile at depth in a homogeneous soil-mass appears to
be independent of pile size, and is a function of relative
density of the sand and method of installation of the
piles. Relationships between fblfs and angle of ínternal FIGURE 3.9 Simplified distribution of vertical stress adjacent to
friction (q/), obtained by Vesic, are shown in Fig. 3 .8. pile in sand.
The above research indicates that the vertical effective
stress adjacent to the pile is not necessarily equal to the
Sorne design approaches have effectively incorpo"ated
effective overburden pressure, but reaches a limiting
Vesic's findings by specifying an upper limit to the shaft
value at depth. This phenomenon was attributed by Vesic
and base resistances. Eor example, McClelland et al. (1969)
to arching and is sin1ilar to that considered by Terzaghi
have suggested, for medium-dense clean sand the following
(1943) in relation to tynnels. There are however other
design parameters: .p ~ = 30°; Ks = 0.7 (compression loads)
hypotheses, such as arching in a horizontal plane, which
or 0.5 (tension loads), with a maximum value of shaft
might explain the phenomena shown in Figs. 3.6 and 3.7.
resistance fs of 1 ton/ft 2 {9t kN/m 2 ); and Nq = 41, with
500.-------.--------.-------r-------r-~ a maximum base resistance fb of 100 ton/fe (9 .6 MN/m 2 ).
400 However, such approaches take little account of the nature
of the sand and may not accurately reflect the variation
of pile capacity with pile penetration, as the limiting
resistances generally will only become operative at relative-
ly large penetrations ( of the order of 30 to 40 m).
In order to develop a method of ultimate load pre-
-·-•'
u
e
~
100
diction that better represents the physical reality than
the conventional approaches, and yet is not excessively
complicated, an idealized distribution of effective vertical
,;
~
stress a~ with depth adjacent to a pile is shown in Fig.
'"e
~
o
G. / simplified, leads to the two main characteristics of behavior
a 30 observed by Ve sic: namely, that the average ultima te skin
e
·;o
ro
resistance and the ultimate base resistance become con-
a:
20 stant beyond a certain depth of penetration.
Georgia Hwy. Oept.
tests in
If the pile-soil adhesion Ca and the term eNe are taken
silts as zero in Eq. (3.7), and the term O.S¡d N 7 is neglected as
being small in relation to the term involving Nq, the ulti-
mate load capacity of a single pile in sand may be expressed
as follows:
30 40 (3.15)
Ang!e of shearing resistan ce 1degrees)
a:_. = effective vertical stress along shaft results, it may be possible to derive dífferent relationships
effective overburden stress for z ,;; Zc or limit- for dífferent pile materials.
ing value a ;,e for z > z e F or bored or jacked piles, the values of Ks tan if> ~ in
a ~·b effective vertical stress at level of pile base Fig. 3.10b are considered to befar too large, and it is sug-
Fw correction factor for tapered pile (= 1 for gested that values derived from the data of Meyerhof
uníform diameter pile) (1976) are more appropriate for design. These values are
shown in Fig. 3.10c, and have been obtained by assuming
On the basis of the test results of Vesic (1967), values if>~ 0.75q:.'. The values for bored piles are reasonably
of Ks tan 1/> ~ and the dimensionless critica! depth zc/d consistent with, although more conservative than, those
have been evaluated. Vesic's results are presented in terms recommended by Reese, Touma, and O'Neill (1976).
of the relative density D, of the sand, but the results may Also shown are val u es of Ks tan if; ~ tor dríven piles, derived
also be expressed in teirns of the angle of interna! friction from Meyerhofs data; these latter values are considerably
tp', by using a relationship such as that suggested by Meyer- smaller (typically about one half) of the values given in
hof(1956): Fig. 3.1 Ob. Sorne of thís difference m ay lie in the method
of interpretation of the data ofVesic and others by Meyer-
tp' == 28 + lSD, (3.16) hof, which leads to smaller values of Ks tan if> ~ associated
with larger values of 'ic/d.
Relationships between Ks tan ,P ~ and ,p·, and Zc /d and The bearing capacity factor Nq is plotted against 9
r?', are shown in Fig. 3.10. In a layered-soil proflle, the in Fig. 3.11, these values being based on those derived by
critica} depth Zc refers to theposiÚon of the pile embedded Berezantzev et al. (1961). Vesic (1 967) has pointed out
in the sand. lt should be emphasized that these relation- that there is a great variation in theoretical values of Nq
ships may be subject to amendment in the lig..1.t of further derived by different investigators, but the values of
test results. For example, at present, the dependence of Berezantzev et al. appear to fit the available test data best.
I<s tan tp ~ on the pile material is not defmed. Vesic's tests The solutions given by Berezantzev et al. indicate only
were carried out on steel tube piles, but the values of Ks a small effect of relative embedment depth Ljd, and the
tan</>~ derived from these tests appear to be applicable to curve in Fig. 3.11 represents an average of this small range.
other pile materials. However, in the light of future test The curves given by Meyerhof (1976) show a larger effect
3
For driv«n pll¡¡¡s 1/) = ~ í/)"1 •10 (Fig.3.10a,Fig.3.10b)
For bor¡¡¡d piiQs,í/) =01 -3 (Ftg.3.10a), í/)=í/); (F1g 3.10c)
wh<ZrQ 01 angl¡¡¡ of 1ntqrnal friction prior to
1nstallctt•on of pil¡¡¡
(o) zc/d vs íl) (b) K 5 tan 0'0 vs 0 (e) YoluQS of K5 tan 0~ Basqd
( Driv¡¡¡n Pil¡¡¡s) on Mcy,zrhof ( 1976)
20 3·0
15 2·5
10 2·0
D
''-.
V
" 5 1· 5
Nq 100
- - F-7-
/ rp ~ = angle of interna! friction prior to installation of
the pile
/
/ (b) For. the determination of Ks tan rp~ and zc/d, the
V
value of rp along the pile shaft should be taken as the mean
~
··-
1--- of the values prior to, and subsequent to, drivíng; that is,
10
25 30 35 40 45
(3.18)
0.
FIGURE 3.11 Relatiorrsilip between N q arrd <f; (after Berezantzev
et aL, 1961).
(b) Bored Pites
of L/d; however, the curve of Fig. 3.11 also Iies near the (a) For the determination of Nq and zc/d, it is suggested
middle ofMeyerhofs range. that the value of if¡ be taken as rp '1 3, to allow for the
Values of the taper correctíon factor F w are plotted possible loosening effect of installatipn (see Section 2.4).
against rp in Fig. 3.12 and have been derived from the re- (b) For Ks tan 1;~, Fig. 3.10c should be used, taking the
1
sults of the analysis developed by Nordlund (1963). val ue of rp 1 directly.
In applybg the results in Fig. 3.10 to Fig. 3.12, it is
suggested that the following values of rp be used to allow The above suggestions may also require modification
for the effects of pile installation. in the light of future investigations. Furthemwre, if jetting
is used in conjunction with driving, the shaft resistance may
decrease dramatically and be even less than the value for
a corresponding bored piJe.
McClelland (1974) has reported tests in whích the use
of jetting wíth externa! return flow followed by driving
reduced the ultimate shaft capacity by about 50%, while
installation by jetting alone reduced the ultimate shaft
capacity to only about 10% of the value for apile installed
by driving or.ly.
Another case in which caution should be exercised is
when piles are to be installed in calcareous sands. Such
sands may show friction angles of 35° or more, but have
beeQ found to províde vastly inferior su¡:iport for driven
piles than normal silica sands. In such cases, McClelland
(1974) suggests limiting the skín resistance to 0.2 tons/ft2
(19 kN/m 2) and base resistance to 50 tons/ft 2 (4800
kN/m 2 ). In such circumstances, drilled and grouted piles
may provide a· more satisfactory s_olution than wholly
1.0 2.0 driven piles.
Pile !aper angle wo
In many practica! cases, only standard penetration-
FIGURE 3.12 PHe taper factor Ft.J(after Nordlund, 1963). test data may be available. The value of rp ~ may be esti-
28 ULTIMA TE LOAD CAPACITY OF PILES
3 400;.
For drMtn pii<Zs, Ql = ~ 0;•10 For dnvrzn pilrzs • íZ\
= - -- ·
2
whrzr<Z, 0; anglrz of íntrzrnal frlction
For borrzd p1lrzs , 0 QIÍ -3
pr<or to ¡nstallaf10n of pll<l
whClrrz 0; =anglrz o! ínt<lrnal trístion
prior to installatíon of pli<l
10
00
FIGURE 3.14 Dimensionlcss ultirnate base·load capacity for· pile
in uniform sand.
10 against Q> in Fig. 3.14. The value of L/d does not generally
32 36 40 44
have a marked effect on the ultimate base load unless
<P is relatively large, that is, for dense sands.
FIGURE 3.13 Dimensíonless ultimate shaft-load capacity for pile
in uniform sand. The use of a high value of cP for very dense sands ( say,
cP > 40°) simultaneously for the shaft and the base, should
also be treated with cautlon, since the full base resistance
mated from a correlation such as that given by Peck, may well only be mobilized after a movement sufficient
Hansen, and Thorburn (1974), or by the following em- for the operative value of <P along the shaft to be signifi-
pírica! relationship suggested by Kishida (1967): cantly less than the peak.
lf the pile is founded in a relatively thin, firm stratum
(3.19) underlain by a weaker layer, the ultimate base load may
be governed by the resistance of the pile to punching into
where the weaker sotl. Meyerhof (1976) shows that if the weaker
1ayer is situated less than about 10 base diameters below
N = standard penetration number the base, a reduction in base capacity can be expected;
he suggests that in such cases, the ultimate point resistance
A more detailed discussion of the relationship between can be taken to decrease linearly from the value at lOdb
<P '1 and N, and also <P '1 and relative density D,, is given above the weaker !ayer to the value at the surfac~ of the
by de Mello (1971). weaker !ayer.
For the case of a driven pile in a uniform !ayer of The suggested approach of ultimate load calculation
sand, dimensionless values of the ultimate shaft load and has been applied to 43 reported load-tests on driven piles.
ultimate base load may be derived using Eq. (3.15) and The details of the parameters chosen for the calculations
Figs. 3.10, 3.11, and 3.12. In Fig. 3.13, the dimensionless are given in Table 3.3, and the comparison between cal-
ultimate shaft load PsuFfd 3 is plotted against cp for várious culated and measured ultimate loads is shown in Fig. 3.15.
values of L/d; ;:¡ is the effective unit weight of the soil The mean ratio of calculated to observed ultimate loads is
above the critical depth Zc. The marked increase in ultimate 0.98, with a standard deviation of 0.21. It should be
shaft load with increasing L/d and <P is clearly shown. The pointed out that the ultimate load of al! piles considered
dimensionless ultimate base load PbuF(dAb is plotted in the comparison is less than 300 tons. The use of this
úLTIMA TE LOAD CAPACITY OF PILES 29
l/lustrative Example Considering first the ultima te skin resistan ce, the
The piles considered are Piles B and A from the Power values of c/J given by Eq. (3.18) are as follows:
Plant Site, Area I, Helena, Ark. Pile B was a closed-end
steel-pipe pile, 24.4 m long and 0.32 m in diameter, driven c/J = 28.75° (0-2.4 m)
into fine sand grading to coarse and having an average c/J = 34° (2.4-18.3 m)
30 ULTIMATE LOAD CAPACITY OF PILES
From Fig. 3.l0b, the values of Kstanr/>~.are 1.00 (0-2.4 + ~:..:..::......:..;..:..:..;:.;;x 0.15 + 44.12 x oo- 2.ss)J
m), 1.30 (2.4-18.3 m), 1.18 (18.3-20.8 m), 1.35 (20.8 m).
lf it is assumed that the crítica! depth is less than 2.4 m 0 202
below the surface. then for dJ = 28.75°, = 5.0, from
'7
X rrX0.33/\4.1X 1.30 + 44.12 X -X
.. ·
4
Fig. 3.10a; that is, Zc = 5.0 X 0.31 1.56 m. Thus, the
assum ption is justified X 88
At the critical depth, the effective overburden stress = 2243 kN (252.2t)
is
"" 1.56 X 17.3 = 26.99 kN/m 2 The measured ultimate load for this pile was 2400 kN
(270 t).
Beca use the pi le has uniform diameter, F w = l.
For thc ultimate base resistance, the value of rp given
by Eq. (3.17) is 36.5°. From 3.11, the value of Nq
3.3 PILE GROllPS
ís 98.
Subsíituting into Eq. (3.15),
In examining the behavior of pile groups, it is necessary
?" = ,, X 0.32{[(O+
2 ~· 99 ) X 1.55 + 26.99 X
to distinguish between two types of group:
(a) Converse-Labarre formula, was accompanieá by the formation of vertical slip planes
joining the perimeter piles, the block of clay enclosed by
r¡, _ t Í(n-l)m+(m-l)n] the slip planes sinking with the pile relative to the general
1
~ C mn 190 (3 .21)
surface of the clay. For wider spacings, the piles penetrated
individually into the clay .The critica! spacing was found to
where increase as the number of piles in the group increased.
Although Whitaker's tests confirmed the existence of
m =· number of rows the above two types of failure, th~ transition between the
n = number of piles in a row ultimate group capacity as given by individual pile failure
~ = arctan d/s, in degrees and that given by block failure was not as abrupt as the
d "' pile diameter Terzaghi and Peck approach suggests. In order to obtain
s = center-to-center spacing of piles a more realistic estímate of the ultimate load capacity of
a group, the following empírica! relationship is suggested:
(b) Feld's rule, which reduces the calculated load capacity
of each pile in a group by l/t6 for each adjacent pile, that
is, no account is taken of the pile spacing.
(e) A rule of uncertain origin, in which the calculated load
capacity of each pile is reduced by a proportion I for each
adjacent pile where
1
I = -djs (3.22)
8
1.0 . . . . - - - - - - - - . . , - - - - - - - - , . - - - - - - - - - ,
3 2 x 30 di S Ti
2 x 12 d(SF)
3 2 x 12 d(ST! 1
G"
e
o 6 1----------+.......::--J'+--~r=
"'
u
e: 72 X 24 díWI
Ll.l
1
0.2~------~-------~-------~
2 3 4
Spaci ng/diameter
FIGURE 3.17 Relationships for freestanding groups Óf 2' to 9'!,piles oflengths 12d to 48d,from model tests (after de Mello, 1969).
ULTIMA TE LOAD CAPACITY OF PILES 33
o
0.6 0.6 0.6
!] !] !] o
o
0.2 0.2
2 3 4 2 3 4 2 3 4
d' '-d
d'
Calculated
o Experimental
(Whitaker, 1957)
lo 48
d
FIGURE 3.18 Experimental and calculated group efficiency, effect of group síze.
calculatíon appears to predict . with reasonable accuracy has measured the load carried by the piles in model free-
the effects of group size, pile spacing, and pile length. standing groups in clay by introducing a smallload gauge
It has often been assumed that all piles in a group are at the head of each pile. The results for a 3 2 group of piles
equally loaded. However, if the group supports a rigid at three different spacings are shown in Fig. 3 .20, in which
cap, the load distribution within the group is generally the average percentage of load taken by each pile is plotted
not uniform, the outer piles tending to be more heavily- against the group load as a percentage of the group load at
loaded than the piles near the center. Whitaker ( 1957) failure. At spacings of 2d and 4d, the comer piles take the
greatest load (about 13 to 25% more than the average
load) while the center pile takes the least (about 18 to 35%
less than the average). At a spaci.'1g of 8d, virtually no
group actíon was observed and the load dístribution was
r¡ uniform. The load distribution for a 5 2 group, ata spacing
of 2d, is shown in Fig. 3 .21. The comer piles reached their
maximum load at about 80% of the ultimate group load,
and carried a constant load thereafter. At failure, the comer
0.5 0.5
o 2 4 (i 8 o 2 4 6 8 piles carried about 28% more than the average load, while
~ S
(¡¡) d (b) d the center and Hghtest-loaded pile carried about 44%
l ~ 24 less. Therefore, there appears to be a tendency for the load
d
1.0 1.0 distribution to become increasingly nonuniform as the
number of piles in the group increases. A theoretícal
method for calculating the load distribution prior to
r¡ '1
ultimate failure is described in Chapter 6, and this method
also confirms the trends displayed by Whitaker's tests.
(a) (e)
s/d = 2 s/d 8
where
"C r,J
.,.e
o~
_j-
o average adhesion along pile
undraíned cohesion at leve! Cl pile tip
undrained cohesíon beneath pile cap
bearing capacity factor for piJe (see Fig. 3.5)
20 40 60 80 100 "' bearing capacity factor for rectangular cap Be X
Load on a group as a percentage
of the load at failure
Le (Le> Be)""' 5.14 (1 + 0.2 BcfLc) (Skempton
1951)
'(he first vaJue will apply for clase pile-spacings while the
A o o co sn o
B
'-../
A
• Average of plles A second will apply at wider spacings when individual action
so Eo oo so o Average of plles B can occur.
co Eo o E() co . Average of pi les C Whitaker (1960) carried out tests on model piled foun-
dations in clay and found that at close spacíngs, block fail-
• Average of pi les O
ure occurred, and that when the cap did not extend beyond
so E
o o B() D ó Average of píles E
the perimeter of the group, it added nothing to the efficien-
o Pi le F
Ao so e'O o o B A cy of the group. At greater spacings, the efficiency-versus-
spacing relationship was found to be an exter.sion of the
FIGURE 3.21 Load dis'tribu tion m 5' pite group at 2d spacing relationship for b.lock failu re, with the efficiency exceeding
(Whítaker, 1970). unity because of the effect of t.he cap. Good agreement was
l:LTIMATE: LOAD CAPAC!TY OF P!LES 35
UJ
0.6
in this figure are reasonably consistent with the data in
Table 3.4. Results of tests on somewhat larger model
piles, in groups of four and nine, carried out by Vesic
0.5 (1969), are shown in Fig. 3.24. Vesic measured t!re point
loaél separately from the shaft resistance, and in tl1~ light
of his measureménts, he concluded that when •he efficíency
0.4 - of closely spaced piles was greater than unity, this increase
was in the shaft rather than the point resístance.
The broad conclusion to be drawn from the above
data is that unle~s the sand is very dense or the píles are
Spacíng factor, s!d widely spaced, the overall efficiwcy is líkely to be
rests on freestandí:~g groups
than l.· The maxímum efficiency is reached at a spacing of
Tests on
2 to 3 diameters and generally rm!ges between 1.3 and 2.
- - Calculated for foundations,
assumíng block U.2.2 JNFLUE!VCE OF PILE C"r1PS
FIGURE 3.22 Efficiency of piled groups (Whitaker, 1970).
As can be seen in Fíg. 3.24, the piJe cap can contribute
significantly to the load capacíty of the group, partícularly
obtaíned between the model test results and the predicted in the case of tl1e srnaller four-pile groups. Hov.ever, it
efficiency from the block failure equation (Fig. 3.22). seems likcly that mobilization of the bearing capacíty of
The load-settlement behavior of piled foundatíons tbe full area of the cap requires considerably greater move-
containing a relatively smal! number of piles to reduce ment than that required to mobilize the capacity of the
settlement is considered in detall in ~hapter 10. piles themselves. This is the implication of tests by Vesic.
and for practica! purpose>, the contribution of the cap
3.3.1.3 ECCENTRIC LOADING can be taken to be thc bearing capacity of a strip footing
of half-width equal to the distance from the edge of the
Model tests on groups with small eccentrícitíes of load have capto t!Je outside of the.pile.
been carried out by Saffery and Tate (1961), who found
that for eccentricities up to two thirds of the spacing,
3.3.2.3 ECCENTRIC LOADING
the group efficiency is not noticeably affected. Meyerhof
(1963) also reported that model tests on piled foundations The influence of eccentric loading on the load capacity
showed that the load eccentricity had no effect on load of pile groups in sand has been studied by Kishida and
ca;:>acity for eccentricities up to half the group width. Meyerhof (1965) in a series of model tests. These tests
This behavior is explained by the fact that the reduced showed that small eccentricities of load have no signif-
base resistan<::e is offset by mobilization of lateral resis- icant inf!uence on the bearing capacity of freestanding
tance. The group capacíty can therefore be calculated as groups and piled groups because the applied moment is
for symmetrical vertical loading, except that for groups resisted mainly by the earth pressure moment on the
whose width is on the same order as the pile length, Meyer- sides of the group. At larger eccentricities, the load capa-
hof ( 1963) suggests that the shaft resistance can be ignored city decreases rapídly because of smaller point resistance
and the base resistance calculated in a fashion símilar to of the group by a reduction of the effective base area.
eccentrically-loaded spread foundations, that is, using a In estímating load capacity, Kishida and d!eyerhof
reduced effective base area. suggest that the moment caused by a load V at eccentricity
"'"'
Press ( 1933) Medium-grained· 6-10ft 5 & 6 in. 12-20 2-8 Various >1 Driven pilcs. Max.
moist, dense of 1.5 ats/d "'.2
11
sand 23ft 16 in. 17 2 Various <l. Bored pites
Cambefort (195 3) Humus/stiff 100 in. 2 in. 50 2-7 2 1.39 Driven piles
clay/sand/ 3 1.64 Average values of r¡.
grave! 5 1.17
9 1.07
~
240 Oa$h l1nes Loose sand
/
Rp . Rough píle
Sp· . Smooth plle
l. large group, piles 1
J7~
S . Sm(jll grn11rt, shnn pdf;S
2.51 i
§ ..!P
..o¡
o; ,¡Q'
~e t¡/1
.,a.~ oj¡l¡i
>
"e
"'
"' 2.0
>
u
e
1 /¿t ' --
] 160
<;.
a.
]'" ¡¡c:r 1 rora¡ er¡:.
'' ll.-?
-,w'"' 1
-"f -t-l(A.;,;; 0
~
041 1 ... y
).» Point eft,crenc _ =-1---
- 3:' 32. 'Í' ~ 4!'>0 1.0
-o... ..........
. 1?() •. ~5
o<'" e "¡sts)
--!'s·
~....-------
x- - - ""*- - - -32
· S p,cl¡=315'
- -~- :._ --- --- 1
.
-1
1
i .
0.5
9 2 3 4 5 6 7
Spacíng In píle diameters P!le spacing in diameters
FIGURE 3.23 Measured values of group efficiency in sands-model .ests (Lo, FIGURE 3.24 Píle group efficiencies (Vesic, 1969).
1967). (Reproduced by permission of tJ>· National Rcsearch Council of Canada
from the Canadian Geotechnical Jour.. al, Vol. 4, 1967, pp. 353-354.)
"'....
38 ULTIMATE LOAD CAPACITY OF PILES
loase Dense
Ir,', _:e 3~, 1 ¡._., .~ 4 3")
1.25 1.25 E~rt
Sand Sand
S1eel Steel
paper ~ paper
~ :,;
2
.'!
o "'
1.00 1.00
¡; " ~ '"'
u ~
u u ~
"' "'
"'e
-o e "
:J
:J Cl
Cl :J
:J
o e
DO
o o;e
00
0.50
.f" ·~
g Cl
Cl ro
ro u
u 00
00 e
e ·e: 0.25 0.25
ro
ro
"' "'
m
OJ
o
2 3 4 o 2 3 4
FIGURE 3.25 Bearing capacity of model piJe groups under eccentric load in sand: (a) freestanding plle gr~ups; {b) piled foundations
(Kishitla ami Mcyerhof, 1965 ). (Copyright Canada, 1965 by University of Taranta Press.)
e is balancerl by the moment caused by lateral forces on ing between 20% and 50% more than the average. Thes,~
the sidcs cf the group until it reaches the maximum value results are in contrast to the load distribution in groups in
corresponding to the coefficient of passive earth pressure. el ay, where the center pile carries the least load and the
Within this limit, the eccentricity of load is assumed to comer piles the most.
have no effect on the point resistance. When the moment The influence of the order of driving piles in a group
Ve is greater than can be resisted by side pressure on the 011 the load distribution has been studied by Beredugo
outer piles., the extra is considered to be taken by an (1966) and Kishida (1967). They found that when the
eccentric base resistance for the case of block failure; load on the group was relatively small, piles that had been
or, for individual pile failure, by the development ofuplift installed earlier carried less load than those that have been
resistance of some piles. The total bearing capacity then installed later; but whe11 the failure load of the group was
decreases with further i11crease in eccentricity. approached, the influence of driving order dirninished,
Comparisons between the theoretical and measured and the position of the pile in the group became the domi-
effect of load eccentricity 011 load capacity are shown in nant factor. At this stage, the piles near the center took
Fig. 3.25 for the tests carried ou't by Kishida and Meyerhof the most load and the comer piles. the least. as in Vesic's
(1965), and there is fair agreement for tests in both loose experiments.
and dense sands. Beredugo also investigated the effects of repeated
loading and found that there was a progressive reduction
3.3.2.4 LOAD DISTRIBUTION IN GROUP
of the influence of driving arder, and that for the third
The most detailed data available on load distribution and subsequent loadings, the pile position was the domi-
within groups in sane! is that reported by Vesic (1969), 11ant factor at allloads up to the ultimate of the group.
wha- made axial load measurements in individual piles
during group placement, as wcll as during loading tests.
For the four-pile groups tested, the measured load distri-
butio11 was Ji.most uniform, as expected; the maxirnum 3.4 PILES TO ROCK
deviation from the average was 3 to 7%. For the ni11e-
pile groups, significant nonuniformit.y of load was mea-
sured. The center pile carried about 36% more load than 3.4.1 Point-Bearing Capacity
the average, while the comer piles carried about 12%
less and the edge piles 3% more. Other tests on similar There are a number of possibl approaches to the cstima-
groups shmved a similar trend, with the center piles carry- tion of point-bearing capacity of piles to rock, includi11g:
ULTIMA TE LOAD C.$-PACITY OF PiLES 39·
(a) The use of bearing-capacity theoríes to calculate the cL:aws attention to the fact that the load-penetration curve
ultímate point-bearing capacity p¡1u. for rocks of medium strength or less (.;:;;; lOO MPa) has a
(b) The use of empirical data to determine the allowable large "plastic" component, despite the brittle nature of
point pressure Pba· the rock. The curve divides into two portions, with what
(e) The use of in-situ tests to estímate either ultima te appears to be a change of slope associated with the forma-
point capacity Pb 1, or al!owable point pressure Pba. tion of a crushed zone beneath the footing. The displace-
ments required to mobilize the full bearing capacity of
Bearing-Capacíty Theories such rocks are very large, and it seems that factor of a
Pells ( 1977) has classified theoretical approaches into three safety of 3 to 4 is required to lirnit the displacements to less
categories: than 2% of the footing diameter. Very brittle rocks (qum
> 150 MPa), do not exhibit this "plastic" load-penetration
l. Methods that essentially assume rock failure to be curve.
"'plastic" and either use soil mechanics-type bearing- The presence of jointing in ¡he rock will tenrl. to reduce
capacity analyses o; modifications thereof to account for the ultima te bearing capacity. The presence of dosel y·
the curved nature of the peak faiiure envelope of rack. spaced continuous tight joints may not reduce the bearing
2. Methods that idealize the z.ones of failure beneath a capacity much below that for the intact rock material. If
footing in a form that allows either the brittleness-strength the~ are open vertical joints with a spacing less than tne
ratio or the brittleness-modular ratio to be taken into width or diameter of the pile point, the point is essentially
account. supported by unconfined rock columns and the bearing
3. Methods based on límiting the maxímum stress beneath capacity may be expected to be slightly less than the aver-
the loaded arca to a value less than required to initiate age uniaxial strength of the rock. If the joint spacing is
fracture. These methods assume essentially tltat once the much wider than the footing width, Meyerhof (1953) sug-
ma:dmum strength is exceeded at any point in a brittle gests that the crushed zone beneath the footing splits the
materiaL total col!apse occurs. block of rock formed by the joints. Sowers and Sowers
{1970) present a theory for this case that generally indi-
For a typical sandstone having an effective frü:tion cates a bearing capacity slightly greater than the uniaxial
angle rf/ in excess oC 45°, effective cohesion e' of about . strength. Thus, in summary, theoretical considerations sug-
one-tenth of the uniaxial strength, qum, and a ratio of gest that the ultimate bearing capacity is unllkely to be re·
Young's modulus-to-uniaxial strength of about 200, Pells duced much below the uniaxial strengtn of the intact rock,
shows that the various theories predict an ultín1ate point- even if open vertical joints are present.
bearing capacity ranging between 4.9qum (incipient failure
theory based on the modified Griffith theory) to 56qum Use of Empírícal Data
(classical plasticity theory). Various model tests on intact Allowable bearing pressures on rock have often been
rock carried out by Pells and others indicate ultímate specified by various building codes and authorities, either
capacities ranging between 4 and ll times qum. Pells , based .on a, descrlption of the rock, or in terms of the
-t r ,¡ ~
';J~. ;? t~'. / 1•.\
..,: )::J l ¿_ ~ \. . •
1 \j
;¡.~
(
• Ur1failed
't' Failed
50r-------.-------.---~.-.-------.-------,-------.
!':
"~ 30~------~------~--
Q.
.§
"'"'
_o
]' 20~.----~~----~--
"0
"'>
u
"'
:;:
<(
10~-----+----~~----~----~~-------r-----
o1~
~1· 1
-E---4 i
o 10 20 30 40 50 60
Ur1confir1ed strength (MPa)
FIGURE 3.26 Achieved end-bearing pressures in field tests on piles to rock (Thorne, 1977).
uniaxial compressive strength Qum. Some typical values been used but may be expensive if the rock is strong and
of Qum and Óther rock properties are summarized in large loads are required. Freeman et al. (1972) have
Table 3.5 Typically, allowable pressures, Pba, ranging described the use of the Ménard Pressuremeter to estímate
between 0.2 and 0.5 times Qum have been stipulated. An the allowable point-bearing capacity, Pba, of piles in rock,
example of sti~ulated bearing pressures related to rock and suggest that Pba may be taken as the value where the
types is provide by Ordinance No. 70 in New South Wales, pressure-versus-volume relationship starts to become
Australia, in which val ues of Pba range between 430 kN/m 2 nonlinear. Satisfactory designs of caissons in sound shale
for soft shale to 321 O kN/m 2 for hard sandstone free from bedrock using the above approach have been reported by
defects toa depth of 900 mm. Freeman et al., and design pressures considerably larger
Thorne ( 1977) has collected data on recorded values than those specified by empirical relationships or building
of bearing capacity, as shown ·in Fig. 3 .26. These values codes have been used.
vary from 0.3qum to about 4qu~, an('[ most cases do not
involve failure. The few recorded failures are in swelling
shales and in fractured rocks, it is clear from these results
that the fracture spacing has an effect on the bearing capa- 3.4.2 Pile-Rock Adhesion
city, . although the data is insufficient to quantify this
effect. When piles are socketed or dríven into rock, sorne load
On the basis of the available data, an allowable point- transfer to the embedded portion of the shaft will usually
bearing pressure on the order of 0.3Qum would appear to occur. Theoretical solútions for load transfer are díscussed
be quite conservative for all but swelling shales. Reference in Sectíon 5.3, and also by Ladanyi ( 1977). The distribu·
to the tbeoretícal solutions shows that such values generally tion of applied load between side-adhesion and end-bearing
imply a factor of safety of at.least 3 in fractured or closely- at workíng loads, as gíven by theory, has be en supported
jointed rocks and 12 or more for intact rocks. by in-situ measurements at a number of sítes (Pells, 1977).
There is not a great amount of data on ultimate values of
The Use of !n-Si tu Tests pile-rock adhesion, but Thorne (1977) has summarized
A number of methods of in-situ testing of rock have been sorne of the available data, and this summary is reproduced
de'v'elopeci in recent years. Plate-load tests have frequently in 3.27. These results show that a number of failures
ULTIMA TE LOAD CAPACITY OF PILES 41
• Unfailed
• Faíled
5r------r------r-·----.-----~------r------,
Kings Park
....,c;.......o.___ _ -....i----=S; 1 shale
Normal maxímum
andesite--::.,.¡¿=M-el-bo_u_rn_e for 25 MPa concrete-:::É:~C-a-líf-or-n-ía-l
/ mudstone 1 shale and
(value approxímate) sandstone
Ca nada 1
shale • UK síltstone/mudstone UK shale
o 10 20 30 40 50 60
Unconfíned strength (MPa)
FIGURE 3.27 Adhesion attaíned in field tests on piles in rock (Thorne, 1977).
have occurred, even in relatively unjointed rocks, at values 3.5 USE OF IN-SITU TESTS
on the order of O.IQum. It should. be noted that in many
instances, concrete strength will be the limiting ·factor,
and in the few instances in which information is available 3.5 .l Sta tic Cone Penetro meter
on concrete strengths, failure has occurred at an average
shear-stress of between 0.05 to 0.2 times the ultimate com- The basis of the test is the measurement of the resistance
pressure strength of concrete, f~- However, the tests of to penetration of a 60° cone with a base area of 10 sq
Jaspar and Shtenko (1969) indicated that considerable cm. Two types of cone are commonly used; the standard
· cautíon must be exercised with piles in expansíve shales point, with which only point resistance can be measured;
that are lik:ely to be affected by water; an adhesion of and the friction-jacket point, which allows both point
only about 11 psi (75 k.Pa) was measured in these tests. resistance and local skin resistance to be measured (Bege-
Freeman et al. (1 972) suggest a design value of allowable mann, 1953 and 1965).
pile-rock adhesion of 100 to 150 psi (700 to 1000 k.Pa), In purely cohesive soils, it is generally accepted that the
depending on the quality of the rock. With such a value, cone-point resistance, Ckd, is related to the undrained cohe-
they recommended that the full calculated end-bearing sion, eu, as
capacity be added to obtain the total-design-load capacity.
On the basis of the limited information available, (3.27)
it would appear reasonable to use as a design value an allow-
able adhesion of 0.05!~ or 0.05Qum, whichever is the As· discussed in the previous section, the factor Nc may
lesser value. These values should not be applied to highly vary widely both theoietícally and in practice, and values
fractured rocks, for which values of adhesion between 75 of Nc ranging from 10 to 30 have been suggested. The
and 150 k.Pa may be more appropriate. lt must be empha- major causes of this variation are the sensitivity of the soil,
sized that care should be exercised to remove all remolded the relative compressibility of the soil, and the occurrence
soil from the socket zone. Furthermore, for upÍift loads, óf adhesion on the side of the cone. The variation in the
a reduction of the above loads(e.g., by about 30%) appears rate of strain between the cone test and other testing
to be desirable. methods also has an effect on the deduced value of Nc,
42 ULTIMA TE LOAD CAPACITY OF PILES
but the use of a constant-penetration rate minirnizes cone resistance Ckd within a distance 3.75 db above and
variations from thís cause. For design purposes, a value of db below the pile tip, where db is the diameter of the
Ne 15 to 18 appears reasonable (Begemann, 1965; pile tip.
Thomas, 1965; Blight, 1967; Thorne and Burman, 1968). Full-scale tests carried out by Vesic (1967) showed
Van der Veen ( 195 7) suggested that the ultima te resis- that the point resistance of the piles tested is comparable
tan ce of a pile point, of diameter db, could be derived from with that of the penetrometer, but the shaft resístance of
the corresponding cone-penetration curve by taking the the piles was approximately double that measured by the
average cone resistance over a distance bdb below the penetrometer. Thus, the ultimate load capacity is given by
pile point and adb above the point. Average values· of
a= 3.75 and b 1 were suggested by Van der Veen.
(3.28)
The adhesion measured by the friction jacket may
safely be taken as the skin friction for driven piles in clays
where
(Begemann, 1965). Alternatively, but less desirably, the
cohesion rnay be estimated from the point resistance and
Ckd = measured cone-point resistance at base
an appropriate reduction made to obtain the pile-soil
j~ average shaft friction along pile, as measured
adhesion (see Section 3.2.1).
on the friction jacket
For piles in sand, various attempts have been made to
relate the cone-poiút resistance to the angle of friction
and relative density of the sand (Meyerhof, 1956; Shultze For driven steel H-píles, Meyerhof (1956) suggested that
...
and Mezler, 1965; Plantema, 1957), but it has been found the above shaft resistance should be halved.
that con e resistance is very sensítive to changes in density. A comparison between the piJe and penetrometer
For p~actical use, the previously mentioned suggestion of resistances for the tests reported by Ves;c (1967) is shown
Van der Veen (1957) may be adopted; name1y, that the ~ Fig. 3.28. The upper and lower linlits of the penetro-
ultimate point resistance of the pile be taken as the average meter values are shown. Correlation with static cone tests
10
-.,
u
t
i::l
'O
e
;::¡
e
""3:
o
"'
.D
.!::
ñ.
a"'
was found by Vesíc to be better than with the results of 3.5.3 Pressuremeter Test
standard penetration tests (see oelow).
~~ or cases in wlüch separa te measurements of friction- The use of the pressuremeter in foundation design has been
jacket resístances are not made, Meyerhof ( 1956) suggested developed extensively in France in recent years. Its appli-
that for dríven concrete or timber piles, the ultimate sk:in cation to the estimation of pile load capacity has been
friction fs could be estimated from the cone point resis- summarized by Baguelin et al (1978) who present curves
tance ckd as fol!ows: relating ultimate base capacity to the pressuremeter limit
pressure, for both dríven and cast-in-sítu piles. Relation-
fs 0.005Ckd (3.29) ships are also presented between ultimate skin resistance
and limit pressure for steel or concrete piles in granular and
For driven steel H-piles, Meyerhof suggested that the cohesive soils, and for cast-in-situ piles in weath;¡red rock.
above value be halved. Sorne comparisons (Mohan et al., The following upper limits on the ultima te skin ·resistance
1963; Thorne and Burn1an, 1968) indicate that Eq. (3.29) are suggested by Ba:guelin et al for pressuremeter limit
underestimates the skin friction by a factor of about pressures in excess of 15 00 k.Pa;
2 if ckd is less than about 35 kgf/cm 2 •
In sands, it is necessary to make a distinction between concrete displacement piles in
the skín frlction for downward and upward loading. Modi- granular soil 122 k.Pa
fications for uplift resistan ce are discussed in Section 3.7.
concrete displacement piles in
cohesive soíl, or steel displace-
ment piles in granular soil 82 kPa
3.5.2 Standard Penetration Test
steel displacement piles in
Meyerhof (1956) has correlated the shaft and base resis- cohesive soil 62 kPa
tances of a pile with the results of a standard penetration
non-displacement piles in
test. For dísplacement piles in saturated sand, the ultímate
any soil 40 k.Pa
load, in U.S. tons, is given by
(3.30)
3.6 SPECIAL TYPES OF PILE
where
3.6.1 Large Bored Piers
f!p standard penetration nnmber, N, at pile base
N average value of N along pile shaft Large-diameter bored píles have come into íncreasing use
in recent years as an alternative to piJe groups. They have
For small displacement piles (e.g., steel H-piles), been constructed up to lO ft in diamcter and in lengths
exceeding lOO ft, often with an underreamed or belled
base. Such piles have found extensíve use in London clay,
(3.31)
and much of the research on large bored piers is based on
their behavior in London clay. Empírica! methods of
where desígn have been developed on the basis of extensive expe·
rienc~ and research. One of the earliest investigations was
net sectíonal are a of toe ( sq ft) in model tests on piles wíth enlarged bases, reported by
gross surface area of shaft (sq ft) (area of all sur- Cooke and Whitaker (1961). These tests ~evealed that,
faces of flanges and web for H-piles) whereas settlements on the arder of 10 to 15% of the base
diameter were required to develop the ultimate base capa-
In Eq. (3.30), the recommended upper limit of the unít city, the full shaft resistance was developed at very small
shaft resístance (ii/50) is 1 ton/ft 2 and in Eq. (3.31 ), settlements, on the order of 0.5 to 1.0% of the shaft
0.5 ton/ft 2 . díameter. (The theory given in Chapter 5 supporJ:s these
The above equations have also been used with sorne findíngs.) A considerable amount of field-test evidence has
success in stiff clays (Bromham and Styles, ·1971 ). subsequently been obtained (Whitaker and Cooke, 1966;
44 ULTIMA TE LOAD CAPACITY OF PILES
1.25r-----,------,------,-----,-----~------.------.-----.
• Source of data i
Toml1nson ( 1957) } Average values
Skempton ( 1959)
1.00 f-----".ct------+~-----t------+---r-- Mohan and Chandra ( 1961
Turner {19621 1
)f:--j
lar pile load tests
FIGURE 3.31 ! -tationship between calcu and undrained shear strength for pulling tests (Sowa, 1970). (Reproduced by permission of the
National Research Council of Canada from the Canadian Geotechnical Jo urna!, Vol. 7, 1970, pp. 482-493.)
46 ULTIMA TE LOAD CAPACITY OF PILES
Relatively f~w pulling tests on piles have been reported (a) The shear resistance of a vertical cylinder above the
in th~ literature. A surnrnary of sorne of the avaílable results base, rnultiplied by a factor k, plus the weíght of wil and
is given by Sowa (1970), who has found that the values of pile, W, above the base.
calcu agree reasc.nably well with the values for piles sub- (b) Tlie uplift capacity of the base plus W, that is,
je~ted to downward loadíng 3.3l).
For piles of uniforrn diameter in sand, the ultimate
uplift capacity rnay be calculated as the surn of the shaft
(3.34)
resistance plus the weíght of the pile. There is, however,
little data available on the skín friction for upward loadii1g, where
and the available data is to sorne exteni conflicting. For
exarnple, tests r~ported by Ireland ( 1957) on piles driven Nu uplift coefficient
inío fine sand suggest that the average skín friction for ""' Nc for downward load
uplíft loading is equal to that for downward loading, but
data surnmarized by Sowa ( 1970) and Downs and Chieurzzi Examination of the results of model and field tests Jed
(1966) indic:ates considerable variations in average skin Meyerhof and Adams to suggest the following values of
friction between different tests, although there is a ten- k:
dency for the values to be lower than for downward load-
ing, especially for cast-in-sítu piles. In the absence of other Soft clays k := t~l.25
information, a reductíon to two thirds of the calculated Medium clays k 0.7
shaft resistance for downward loading is recornmended. Stiff clays k 0.5
However, a reliable estimate is best detenr.ined by carryíng Stiff fissured clays k 0.25
out a pulling test ín-situ.
If stalíc-cor,e-penetration tests are used as a basis for The !ow values of k in the stiffer clays are partly attributed
estín1atíng ultima te uplift skin resistance, Begemann (1965) by Meyerhof and Adams to the influence of tension cracks
suggests that the calculated skin resistan(;e for downward arísing frorn premature tension-failure in the clay.
loading be adjusted by a reduction factor dependent on lt has been found that negative pore pressures may
the soil and pile type. He also suggests reduced values of occur in clays duríng uplift, particularly with shallow
skin resistance be used if the uplift load is oscillatíng. embedment depi.hs. The uplift capacity under sustained
Begemann's suggestions, however, should be viewed with loading may therefore be less than the short-term or un-
cor,siderable caulíon, as they are based onlimited data. draíned capacity, because the clay tends to soften with
Additional upUft resistance may be obtained by under- time as the negative pore pressures dissípate. The long-tenn
reaming or enlarging the base of the pile, and in such uplift capacity can be estimated from the theory for a
cases, the pile shaft may have little or no influence on the material with both friction and cohesion, using the drained
uplift capaci1y. Traditional methods of design assume the paran1eters rPd and ca of the el ay.
resistance of the enlarged base ·to be the weight of a cone For a soil with both cohesion and friction, the follow-
of earth having sides that rise either vertcally or at 30° ing expressions were obtained by Meyerhof and Adams
from the vertical. Neither of these · methods has proved for the ultima te load capacity, Pu u, of a circular base:
reliable in however. The 30° -cone method is
usually conservative at shallow depths but can give a con- (a) Shallow depths ( L < db):
siderable overestimate of uplift capacity at large depths
(Turner, 1962). Parr and Varner (1962) showed that the
vertica1-failure-surface approach did not apply to piles in (3.35)
cby, althougJ¡ it could apply to backfllled footings. Alter·
nativ0 theoríes for uplíft resistance of enlarged bases have (b) Great depths (L > H):
bcen proposed by Baila (1961), MacDonald (1963), and
Spence (1 %5)-- these theories differing in the assump-
(3.36)
tions regarcling the shape of the failure surface.
Meyerhof and Adams ( 1968) have developed an approx-
imate approach based on observations made in laboratory where
model tests. They suggest that the short-term uplift capa-
city of a piJe in clay (under undrained conditions) is given "'f soil unit weight*
by the lesser of s shape factor
ULTIMA TE LOAD CAPACITY OF PILES 47
1 + mL/db, with a maximum value of values may be appropriate to upward loading. However,
lt-mff/db the theory for failure of anchor piles with enlarged bases,
K!l = earth-pressure coefficient ( approximately 0.9. or of anchor plates more generally, has yet to be fully
0.95 for .p values between 25° a.."ld 40°) developed.
m coefficient depending on .p For use in Eqs. (3.35) and (3.36), values of H/db,
H limiting height of failure surface above base s, and m, obtained from tests results by Meyerhof and
W weight of soíl and pile in cylinder above base* Adams, are shown in Table 3.6. The ultimate uplift capa-.
cíty should be taken as the lesser value of that given by
The upper límit of the uplift capacity is the sum of the net Eq. (3.37) and the appropriate equations 3.35 or 3.36.
bearing-capacíty of the base, the side adhesion of the shaft, The results of model tests in clays, reported by Meyer-
and the weight of the pile, that is, hof and Adams ( 1968), are shown in Fig. 3.32. !3oth the
undrained and long-term pullout loads are shown; and the
(3.37)
where
20 25 30 35 40 45 48
bearing-capacity factors
2.5 3 4 5 7 9 11
ultima te shaft-shear resistan ce_
effective vertical stress at level of pile base
m 0.05 0.1 0.15 0.25 0.35 0.5 0.6
400
200
considerable reduction in load capacity with time can (b) The uplift load of an equivalent pier foundation
clearly be seen. The extent of the load-capacity decrease consisting of the· footings and enclosed soil mass.
becomes greater as the soil becomes stiffer. The predicted
Meyerhof and Adarns (1968) have presented sorne
long-term capacities of the piles show reasonable agree-
data on the uplift efficiency of groups of two and four
ment with the measured values.
model circular footings in clay. The results indica te that
The above theory can also be used to estímate the
the uplift efficiency iricreases with the spacing of the foot-
uplift capacity of piles in sand. Meyerhof and Adams have
ings or bases and as the depth of embedment decreases,
compared predicted and rneasured uplift capacities for
but decreases as the number of footings or bases in the
buried footings in sand and have found fair agreeme~t,
group increases. The uplift efficiencies are found to be
although there is a relatively wide scatter of points.
in good agreement with those found by Whitaker (1957)
for freestanding groups with downward loads.
For uplift loading on pile groups in sand, there appears
3.7.2 Pile Groups to be little data from full-scale field tests. However, Meyer-
hof and Adams ( 1968) have carried out tests on srnall
Meyerhof and Adams (1968) suggest that the ultimate groups of circular footings and rough circular shafts, and
uplift load of a group be calculated as the lesser of have analyzed the group efficiencies. For a given sand
density, the uplift efficiencies of the groups increase
(a) The sum of the uplift of the individual footings. roughly linearly with the spacing of the footings or shafts,
Out-of-
Pi le Aiignment
Reference PileType Length Soil Type atTip Type ofBend
Parsons and Composite: lower 140ft 20 ft fill, layers of 4.4 ft Gentle sweep over
Wilson 85 ft, 10%-ín. pipe, organic sil t, rnedium lower length
(1954) top 55 ft. sand, fine sand, silt
corrugated pipe with c!ay layers,
grave!, bedrock
Johnson Composite: lower 40ft 20ft silt overlying 8 ft Gentle sweep over
(1962) 40 ft, 1O% in. mediurn sand lower length
upper 50 ft,
corrugated taper pipe
Mohr (1963) 1 0 3.4-in. pipe 85 ft 80ft soft silt, 10.25 ft Gentle sweep
stiff s:ind da y,
medium den.se sand
National Precast hexagonal, 60m 50m soft clay, 1Om 11m Gentle
Swedish Hercu[es jointed clay, silt, sand,
Council rock a! 70rn
(1964)
Hanna (196 7) Steel H-section 140ft 34 ft stiff clay, 5O ft 3.0 ft Triple curvature.
14 BP73 soft clay, 64 ft stiff relatively sharp
clay, shale direction changes
Steel H-section 138ft 6.0 ft Double curvature,
14 BP 89 rela tively sharp
dírection changes
ULTIMA TE LOADCAPACITY OF PILES 49
and increase as the depth of embedment becomes smaller. largely be caused by the neglect of the structural strength
The uplift efficiency decreases as the numbt:.r of footings or of the pile shell in the design. Long, precast, hexagonal
shafts in the group increases and as the sand density in- test piles ha ve also been found to perform satisfactorily,
creases. but Hanna {1967) has found that for steel H-piles, large
stresses are induced because of bending during driving. Pile
bending is attributed by Hanna to the d~velopment of asym-
3.8 LOAD CAP ACJ.TY OF BENT PILES metrical stresses in the pile as a result of the eccentric pile-
tip reactiÓn and eccentric driving inherent in all pile-driving
A number of cases have been reported in which long, work. These eccentric stresses are considered · to be suffi-
slender piles have become bent during driving. A summary cient to initiate bending, which causes the piJe to drive off
of these measurernents is shown in Table 3.7. For con- vertical. Reverse curvature of the pile may subseq_uently oc-
crete-filled steel shell piles, load tests indicated that the cur, and this is believed to result primarily fro~ the verti-
piles could tolerate significant out-of-verticality and still cal-weight component of the inclined pile forcing the pile
carry their design load with safety. This, however, may to bend.
Deflection in inches
80 40 o 40 40 o 40
North
~
East
80
t
North
---
South
í
14BP69
North
t
E..
c. 14BP73
fiGURE 3.33 Measured deflection components of driven pile (Hanna, l968). (Reproduced by permission of the National Research Council of
Canada from the Canadian Geotechnical Journal, Vol. 5, 1968, pp. 150-1 72.)
SO ULTIMA TE LOAD CAPACITY OF PI LES
Boreho~ o 2 3 4 5
1 1 1 1
í
Sea le in ft
FIGURE 3.34 Driven positíon of pile tips (Hanna. 1968). (Reproduced by permission of the National Research Coundl of Canada fro 1 ~ the
Canadian Geotechnical Joumal, Vol. 5, 1968, pp. 150-172.)
Typical deflection profiles, ·reported by Hanna (1967), Pmax maximum allowable soil pressure
are shmvn in Fig. 3.33. These proflles ha ve been obtained Pcr buckling load of pile
from measurements on an inclinometer installed within the k modulus of subgrade reaction
H-piles. The as-driven positions of the piJe tips for every Pmax maximum lateral detlection (deviation of the
20 ft of depth are shown in Fig. 3.34. For the two piles center line of the piJe from a straight line con-
conside,red, mínimum computed radii of curvature were on necting the piJe tip and the point at which
the order of 170 ft and 190 ft at depths of lOO ft and 70 curvature of the piJe begins)
ft: these values are about six times less than the suggested
safe mínimum value for steel H-piles of 1200 ft (Bjerrum,
For the second criterion to- be satisfied, the allowable load
1957).
Pis
Methods of estimating the stresses in a piJe dueto non-
verticality ha ve been proposed by Johnson ( 1962), Broms
(1963), Parson~ and Wilson (1954), and Madhav and IÚo p"' 0.5 (b -~4c) (3.39).
(1975). Typícal of these methods ís that of Broms, who by
expressing the deflected shape of the piJe as a Fourier sine where
series and assuming the soil to be a Winkler medium, was
able to derive a simple approximate equation for the buck-
ling load on the piJe (the subject of buckling is discussed
more fully in Chapter 14). Provided that sorne information
of the departure from straightñess of the actual piles is e (3.39)
available, the maximum soil pressure along the pile and the
maximum bending moment can then be calculated. As de- Pcr pile-buckling load
sign criteria, Broms suggested that A area of piJe
Gmax allowable maximum stress in piJe
(a) The calculated maximum soil pressure along the pile Ep Young's modulus of piJe
should not exceed one third of the ultima te value. fp moment of inertia of pile
(b) The maximum stress (axial plus bending) in the piJe z pile section -modulus
.should be less than the allowable value. Rmin mínimum radius of curvature along piJe
Rmin (3.40)
where
ULTIMATE LOAD CAPACITY OF PILES 51
500
<0. 400
:;e
e:
"'
.2
"
2i
<1l
:;:
.2
<i 200
o 10--6 10- 5
10
5
104
0 oL____________2~(_)____________4LO----~------------_J60
Mínimum radíus ot curvature (ft) lv1axfmum dc1ícci;on \in,}
(.:r) On basis of steel stre:;s Un On basis of soíl pressure
FIGURE 3.35 Allowable loads for bent pites (from Broms' analysis).
For a typical steel H-pile sectíon in clay, the allowable pendent of pÜe length. For the limiting steel-stress criter-
loads from Eqs. (3.38) and (3.39) are plotted in Fig. 3.35. ion, an allowable steel stress of 18 kips/sq ft has been
For the limiting-soíl-pressure criterion, the allowable load adopted. The a!lowable load is insensitive to change in soil
in creases as the stiffness of the soil increases (K =kd = 3 3 subgrade-reaction modulus or pile ler:gth.
times the cohesion, has been assumed) buLis almost inde-
LOAD CAPACITY BY DYNAMIC
METHODS
52
LOAD CAPACITY BY DYNAMIC METHOD 53
.• W·H
Load HT ~rw·H
---·f·--·f-.~.
T
f IV
:.i ·<l ·W·H
~~- ~~-~------
s~ttkzmant
The work done during impact is approximately The left-hand si de of Eq. ( 4.1 O) represents the energy of
the hammer blow, the first term on the right-hand side is
{4.7) the energy consumed by the elastic compression of the pile,
computed as a static compression under the force Ru, and
Neglecting the elastic deformation ·of the soil, and intro- he second term is the energy absorbed by the plastic de-
ducing Hooke's Law for the pile, formation of the soil.
A summary of various practica! formulas is given in
Table 4.1. Tables 4.2, 4.3, and 4.4 give typical values of var-
(4.8)
íous quantities required for these formulas.
where
4.3.2 Reliability of Dynamic Formulas
e = ratio between actual displacement at pile top and
that given by Hooke's Law Severa! investigations have been carried out to determine
the reliability of the various pile-driving formulas by com-
From Eqs. {4.6), (4.7), and (4.8), the following equation is paring the load capacity computed from the appropriate
obtained: formula wíth the measured capacity from a pile loading
test. Sorne of the most comprehensive investígations have
been reported by Sorensen and Hansen {1957), Agerschou
(4.9)
(1962), Flaate {1964), Housel {1966), and Olsen and
Flaate (1967).
Sorensen and Hansen used data from 78 load tests on
concrete, steel, and wooden piles, most of these having
their points bearing on sand (a few were founded on hard
moraine clay). The results of their comparisons are shown
4.2.2 Practica! Driving Formulas in Fig. 4.3, in which the ratio, Jl, of the measured to the
computed load-capacity is plotted against the pcrcentage
Although the above general formula takes most practica! of load tests smaller than Jl. This plot is a probability plot,
factors into account, the validity of the law of impact is and a straight line on this plot represents a normal or
very questíonable, since the piJe is by no means the free Gaussian distribution of results. Figure 4.3 shows that all·
body that the law of in1pact assumes. As pointed out by the formulas considered, with the exception of the Eytel-
Terzaghi ( 1943), "Newton himself warned against the appli- wein formula, follow approximately a Gaussian distribu-
cation of his theory to problems involving for instance the tion. There is very little difference in the accuracy of the
impact produced by 'the stroke of a hammer ."' In addition Danísh, Hiley, and Jan bu formulas, and the theoretícal
to this basic criticism, the formula is not readily
applicable in practice, since many of the quantities are ex-
tremely diffícult to measure or estímate reliably. Conse- 98 • Jonbu formulo r
quently, rnost practicai pile-driving formulas are simplífi-
cations of the general equatíon, often incorporating em-
90
o H;ley formulo
x Eytelwe;n formula----
.c. Donish formulo - - ,
i/, ,_,,r:./
.'
/ /
,.
pírica! "comtants" and coefficients. Most of these for- a 70 • WOvt,z equotion---- t' X ~
~50 ;, /
mulas can be expressed in the following form: ,)< ' '
~ 30 __..-<-6 Q/
(4.10)_
~ 10
3 }'
~ ....... ~ ...... ~
// 1
1
1 ,/ ~-- ¡o
where
·0·8 -0·6 -0·4 -0·2 o 0·2 0·4
log ll
Sanders
WH
S
Engineering
WH e= 1.0 in. for drop hammer
S+e
News 0.1 in. for steam hammer
0.1 Wp/W in. for steam hammer
on very heavy piks
Eytclwein
(Dutch)
e¡WH • W+n'Wp See Tab!es 4.2, 4.3 and 4.4 for values
Hile y
S+Yl(e,+e,+e3 ) W+WP of e¡, e" e,, C3 , and n.
TABLE 4.2 VAL UES OF curves derived from the wave equation (see Sectíon 4.3),
HAMMER EFFICIENCY, ef but the Eytelwein fonnula is very inaccurate.
Agerschou 's investigation concentrated on the Engi-
neering News formula but also broadly confinned the con-
Hammer Type e¡ clusions of Sorensen and Han sen regarding the Hiley, ..
Janbu, and Danish formulas. Agerschou showed that des-
Drop hammer released by trígger 1.00 pite its popularity, the Engineeríng News fonnula is unre-
Drop hamm.cr actuated by rope and
friction winch . O. 75
liable. It has the highest standard deviation, and 96% of the
McKiernan-Terry single-acting hammers 0.85 allowable loads determined by this formula will have actual
Warrington-Vulcan single-acting hammers 0.75 safety factors ranging between 1.1 and 30.0. Flaate (1964)
Differential-acting hammers 0.75 inve·stigated the accuracy of the Jan bu, Hiley, and Engineer-
McKiernan-Terry, Industrial Brownhoist, ing News formulas for 116 tests on timber, concrete, and
Natíonal & Union doÚble-acting hammers 0.85
steel piles in sand. The conclusions reached by Agerschou
Diesel hammers 1.00
regarding the unrelíability of the Engineering News formula
are reinforced by Flaate's results. There ís relatively little
a From Piie Foundatíons by R. D. Chellis. © 1961 difference between the Jan bu and Hiley formulas, although
McGraw-Hill Book Com pan y, In c. Used with per.mission ·
of McGraw-Hill Book Company. the fonner is perhaps the more reliable overall and gives
56 LOAD CAPACITY BY DYNAMIC METHOD
0
TABLE 4.3 VALUES OF C., C,, C, FOR HILEY FORMULA
Material to Which Easy Driving: Medium Driving: Hard Driving:· V ery Hard Driving:
Blow ls Applied P1 500 psi P1 1000 psi P1 =
1500 P1 =2000 psi on
on Cushion on Head or Cap psi on Head Head orCap
or Pile Butt (in.) or Cap (in.) (in.)
lf No Cushion
(in.)
b The first figure represents the compression of the cap and wood dolly or packing above !he cap, whereas the second
figure represents the compression of the wood packíng between the cap and the píle head.
a From Pi/e Foundarions, by R. D. Chellis, © 1961 McGraw-Hill Book Company, lnc. Used with permission of
McGraw-Hill Book Company.
good results for timber and concrete píles. Hiley's formula formula is greater and that of the Engineeríng News for-
is also reasonable for timber piles. mula is less. This difference may well stem from the fact
The tests undertaken by the Michigan Department of that the results in Table 4.5 are predominantly for píles in
State Highways at Belleville, and reported by Housel sand, whereas the Belleville site consists largely of clay. ·
(1966), are compared in Table 4.5 with predictions from a Consequently, the reliability of pile-driving formulas at this
large number of píle-driving formula. The Engineering News si te is likely to be poor, as significan t fríctional resistan ce
formula gives predictions of u! tíma te load of between 2 and may be mobilized along the pile, whereas this resistance is
6 times the measured values, the Hiley formula gives 7 to not directly considered in the formulas. It is also interesting
30 times the measured values, and the Eytelwein formula to note that the Belleville resulis are consistent with the
gives 5 to 25 times the measured values. In comparison with analyse~ performed by Forehand and Reese (1964), which
the previous comparisons, the spread of results of the Hiley suggest that the Engineering News formula may be less uñ-
LOAD CAPACITY BY DYNAMIC METHOD 57
Ru = 9.0vfe¡H-'H (4.llb)
a After Whitaker (1970}.
X log 10 {10/S)- 27
The overall conclusions from the above comparisons are it is necessary to take into consíderation the vibrations
that, if driving formulas are to be used, those which invo]ve which are produced by the impact."
the least uncertainty are the Janbu, Danish and Hiley for-
mulas, whíle the most uncertain is the Engineering News
formula.
In conclusion, it is interesting to note the remarks made 4.3 PILE-DRIVJNG ANALYSJS BY THE WAVE EQUA-
by Terzaghi (1943) "In spite of their obvious deficiencies TIQN
and their unreliability, the pile formulas still enjoy a great
popularity among practicing engineers, because the use of The realization that pile driving could not accurately be
these formulas reduces the design of pile foundations to a analyzed by rlgid-body mechanics led to the development
very simple procedure. The price one pays for this artificial of an analysis that utilizes wave theory. This analysis takes
simplification is very high. In sorne cases the factor of safe- account of the fact that each hammer-blow produces a
ty of foundations designed on the basis of the results ob- stress wave that moves down the length of the pile at the
tained by means of pile formulas is excessive and in other speed of sound, so that the entire !ength of the pile is not
cases significan! settléments have been experienced. On ac- stressed simultaneously, as assumed in the conventional
count of their inherent defects all th(' existing pile formulas dynamic formulas.
·are uti~'rly rn·isleading as to the influence of vital condi- As previously stated, the wave-equation approach is
tions, such as the ratio between the weight of the pile and primarily used to yield a relationship between ultimate pile
the hammer, on the result of the pile dríving operations. In load and pile set, although the stresses set up in· the pile
order to obtain relíable information concerning the effect during driving are also calculated. In addition, this approach
of the impact of the hammer on the penetration of the piles enables a rational analysis to be made of the effects of var-
LOAD CAPACITY BY DYNAMIC METHOD 59
ious factors in the driving process, such as pile characteris- For a pile, the tesistance of the surrounding soil must also
tics, hammer characteristics, and cushion stiffness. lt there- be considered; in which case, Eq. (4.12) becomes ·
fore provides a convenient and logical means of determin-
ing the suitability of a given driving system for driving a
given pile, and of choosing an optimum system to obtain = (;) (~) ±R (4.13)
a desired load capacity without damaging the pile. lt has
been widely used for piles supporting offshore structures where
(McClelland et al., 1969; McClelland, 1974).
Although the safe working load of the pile is again R = soil-resístance terrr.
determined by applying a safety factor to the calculated
ultimate-load capacity, the fact that sorne account is taken Equation ( 4.13) m ay ·be sol ved, for t,he appropriate initial
of the soil characteristics removes at least part of the uncer- and boundary .tonditions, to determine the relationship
tainty in ascribing a suitable safety factor to a particular among displacement, time, and position in the pile, from
si te. which the stress variation in the pile may be determined.
The method of solution of the wave equation to obtain Because of the complications involved in practical piling
the ultima te load versus set relationship is described below, problems, analytical solutíons to Eq. (4.13) generally are
and typical values of the requisite soil-data input are dis- not' feasíble, and resort must be made to numerical means
cussed. Some solutions from the wave-equation approach of solution. A conveníent numerical method has been des-
showing the effect of various factors are given in Section cribed by Smith (1960). A simple computer program is
4.4, and a dÍscussion of the reliability of the wave equation listed in Bowles (1977).
is given in Section 4.5.
The use of the wave equation was considered by Isaacs
( 1931) and Glanville et al. (1938), but it was not until the 4.3.2 Smith's Idealization
work of Smith (1960) that the method was fully developed.
The method developed by Smilh is a finite-difference meth-
Considerable refinements of Smith's analysis have been
od in which the wave equatíon is used to determine the pile·
made, notably by Samson, Hirsch and Lowery (1963), and
set for a given ultimate pile load. The pile system is ideal-
Forehand and Reese (1964). Scanlan and Tomko (1969)
ized as shown in Fig. 4.4. and consists of
ha ve also applied wave theory, in a somewhat different ap-
proach, to estímate pi! e capacity. In the developmen t of the
l. A ram, to which an initial velocity is imparted by the
method described below, the nomenclature and notation of
pile driver.
Samson et al. (1963) and Lowery et al. (1969) will gener-
ally be employed.
where
Load
E D
Friction l1nk D
limits spring,
load '----.._ (C} Equlvalent
Rh<Zolog~eal
Spr¡ng
constant K' Mod¡zl of So1!
A load-deformation diagram such as Fig. 4.6 may be esta- C(m, t) = compression of interna! spring m at time t
blished separately for each spring, so that D(m, t) displacement of element m at time t
D' (m, t) plastic displacement of externa! spring m
' . ) _ Ru(m) at time t
K ~m - Q (m) ( 4.15)
F(m, t) force in interna! spring m at time t
g acceleration caused by gravity
wher~ K' (m) is the spring constan! during elastic deforma- J(m) soil-damping constan! at element m
tion for externa! spring m. K(m) spring constant for interna! spring m
K' (m) spring constant for externa! spring m
To allow for the effccts of dynamic loading during R(m, t) force exterted by externa! spring m on
driving in increasing the instantaneous resistance of the soil, element m at time t
the dynamic load-settlement behavior of the soil is taken V(m, t) velocity of element m at time t'
to be that shown in Fig. 4.6b, which as pointed out by W(m) = weight of element m
Lowery et al. ( 1969),. corresponds to a Kelvin rheological
model. This dynamic behavior i.s characterized by a further Equatíon (4.18) applies for the elastíc pile elements for
parameter J, the darnping constan t. The dashpot in the mo- which interna! damping is ignored. For elements such as the
del produces an additional resistíng force proportional to capblock and the cushion block, in which interna! damping
the velocity of loading (V). should be considered, the foltowing equation should be
u sed instead of Eq. (4.18):
l. The ínitíal velocíty of the ram at initial impact; Vr, 4. Values of the quakes Q(m) and Q(p) and the damping
which can .be calculated as fact~m J(m) and J(p). These quantities are discussed in Sec-
tion 4.3.4.
A cross-sectíonal arca of element m These comparisons are made in ea eh time-ínterval, tlt, and
E = Young's modulus of element m D' (m, t) ís adjusted accordlngly.
ilL = length of element m 4. The plastíc deformatíor( of the píle tip, D' (p, t), re-
mains constant, startíng at zero, unless changed by the con-
3. Values of the externa! spríng constants, K' (m), of the ditíon (see Fig. 4.7b)_
soil. This necessitates the assumption of a total pile resis-
tance Ru, the percen tage of piJe resistan ce to be used as sí de
resistance, and the distribution of síde resistance along the
píle. lf known, the soil strength and adhesion propertíes
l1lay provide a guide to the selection of the above quantí- , Displac<Zm<Znt
tíes (see Chapter 3).
If it ís assumed that the proportion of load carried by
the pile point is ~. then the interna! spring constant, K(p), {a) Pile Shoft
for the píle tip is
D' (p, t) <tD(p, t) ·· Q(p) (4.28) Because inelastic springs and material of different densities
and elastic moduli are usually involved in practica! prob-
This comparison is made at eacL tü:ne·interval and D' (p, t) lems, Samson et al. ( 1963) recommended a value of C::.t of
ís adjusted accordíngly. about half the value given by Eq. (4.29). The accuracy of
5. The soil resistances R(m, t) fC>r m = 3 topare calculated the solutíon is more sensitive to the choice of t::.t if the pile
from Eq. (4.19). lf desired, when D(m, t) D' (m, t) fírst is divided into only a few elements. The solutions of Sam-
becomes zero, Eq. (4.22) may be used. son et al. suggest that t::.L = L/ 10 is generally a reasonable
6. The spring compressíons C(m, t) for m =1 to p 1 are division· of the pile. Smíth's original suggestions on C::.t and
calculated from Eq. ·(4.17). t::.L were 1/4000 sec. and 8 to 1O ft, respectively, for most
7. The forces in the pile elements, F(m, t), are calculated practica} piles, which are consistent with the values recom-
from Eq. (4.18) for the pile elements in which no interna! mended above.
dampíng occurs, or from Eq. (4.21) for the capblock and
cushion block (m t and 2). For the capblock, which is
Modification for E[fect o[Gravity
not attached to the pile, F(l, t) can never be less than zero.
Smith's original procedure does not account for the static
For the pile cap, two cases are possíble:
weight of· the pile sin ce all springs, interna! and externa!,
are assumed to exert zero force at t = 0: that is, F(m, O) =
(a) F(2, t);;;. O if cap is not properly attached.
R(m, O) O. If the effect of gravíty is to be included, these
(b) F(2. t) can be ncgative if cap is attached to pile.
forces must be given initial values to reproduce equilibrium
of the system; in fact, these inítia1 values should be those
Case (a) or (b) must be specifíed at the start of the prob·
in effect as a result of the previous blow, but thís refine-
lem.
ment appears unjustified (Samson et al., 1963).
8. The vetocity V(m, t) ls calculated for m = l to p from
Studies by Samson et al. (1963) indicate that the gra-
Eq. (4.20).
vity effect ís relatively small, and in a typicál case, the effect
9. The tyclc is repeatr~d for successive tirne-intervals until
of gravity was to in crease the permanent set by about l 0%.
the pile segments reach their maximum downward move-
For practica! purposes, it does not appear necessary to in·
ment and rebound upward. Unless Eq. (4.22) is used for
elude the effect of gravity in the wave-equation analysis.
R(m, t), the computation can be terminated when
The permanent set of the pile típ as a result of the ram blow The soil parameters required for the wave-equation analysís
ís the maxirnum value of D' (p, t). are the ultima te soil resistance, Ru; quake, Q; and damping
10. lf the relationship between permanent set ( or its recí- factor,].
proca!, the blow count), and the ultirnate resistance Ru of
the pile is required, various values of Ru are chosen and the
Ultimate soil resistance, Ru
procedure repeated. A plot of Ru versus permanent set ( or
Various values of Ru are input into the computer program
blow count) may thus be obtaíned.
arid the corresponding permanent set determined. The main
problem wíth Ru is to determine the relative proportions of
It is obvious that the above procedure requires the use
shaft and base resístance. A reasonable estímate of these
of a computer for practica! problems. A simple computer
proportions may be made by estimating the static shaft and
program has been given by aowles:( 1977).
base resistances from the known or assumed soil propertíes,
In employing the numerical procedure described above,
¡¡.s described in Chapter 3. A typícal example of the effect
the accuracy of the resulting solution will depend on the.
values of C::.t and t::.L chosen. It has been shown that for of varying the proportions of shaft and base resistance is
shown in Fig. 4.8. A somewhat higher ultima te resistance
free longitudinal vibrations in a continuous elastic bar, the
discrete element solution is an exact solution of the partía! for a g.\ven driving resistance is obtained if some shaft resis-
tance is considered, rather than only end-bearing. As a
differential equation when
rough guide where other information is not available, values
of the percentage of shaft resistance suggested by Forehand
t::.t (4.29)
and Reese (1964) are sliown in Table 4.7.
64 LOAD CAPACITY BY DYNAMIC METHOD
600r----r---------,-----r----r----r----,---~--------~
o
o 5 10 15 20 25 30 35 40 45 50
alows pllr ¡,,eh
FIGURE 4.8 Resistance vs. set curves for 36-in.-diameter prestressed concrete Pite. Effect of percentage point load (Forehand and Reese,
1964).
TABLE 4.7 EMPIRICAL VALUES soil-resistance curves employed by Seed and Reese (1957)
OF Q, J, AND .PERCENT SIDE ADHESIONa and co'yle and Reese (1966).
An example of the effect of varying Q on the Ru ver-
Síde sus driving-resistance curve is shown in Fig. 4.9. The curves
Q J(p) Adhesion have been obtained by Hirsch et al. (1969) for a steel-pipe
Soil (in.) (sec/ft) (% of Ru) pile in a layered-soil profile consísting maínly of clays. Ru
tends to decrease as Q increases.
Coarse sand 0.10 0.15 35
Sand grave1 mixed 0.10 0.15 75-100
Damping factor, J
Fine sand 0.15 0.15 100
Sand and clay or Empirical correlations between J and soil type obtained by
1oam, at least 50% 0.20 0.20 25 Foreh~nd and Reese (1964) are shown in Table 4.7. The
of pile in sand values in this table are for the pile point [i.e., J(p)]. The
Silt and fine sand average value for the si des of the pile J(m) have been found
underlain by hard 0.20 0.20 40
strata
Sand and grave!
underlain by hard 0.15 0.15 25 350r-------------------------------------,
strata
~ 300
Cl.
0 ,;;:
After Forehand. and Reese (1964). ~· 250
"
Cl:.
.. 200
Quake, Q ¡;!
Values of Q have been obtaíned empirically to date, and B
~
the single empirical values of Q for all elements of the pile .
0:
150
value at the pile tip being greater than the values along the
shaft. Alternatively, Q could also be estimated from the FIGURE 4.9 Effect of varying Qside (Hirsch et al., 1969).
LOAD CAPACITY BY DYNAMIC METHOD 65
1·2
1000
"'
Cl.
1·0
::¿ 800
0:::
~ 600
!:: 0·8 o
FIGURE 4.10 Effect C•f J (Hirsch <:1 aL, 1969). Arkansas load test,
pile4. 0·2
to be less than J(p), and for practica! purposes, it has been 0·2 0·4
suggested that l!QUid1ty Ind<zx
0·8
On the basis of the above modified equation, Coyle and
Gibson found J 1(p) to be almost independent of velocity,
and reasonable correlatíons between 1 1(p) and soíl pro·
, ~V1ctoria Sand perties could be obtained. The relationship between J 1 (p)
0·4 -Arkonsos and q/ fór sands ís shown in Fíg. 4.11, while the relation-
Sand
shíp between J 1 (p) and the liquidity index for clays is
02 shown in Fig. 4.12.
FIGURE 4.11 Effective angle of interna! shearing resistance vs. As described in Chapter 2, driving apile into normally con·
damping constant for sand (Coyle and Gibson, 1970). solidated el ay results in the creation of excess pare pres·
66 LOAD CAPACITY BY DYNAMIC METHOD
' 1!: !
~u
~
1
1
Sto tic " Dr1V1ng :E...J-'1
\ Copoc1ty, ', R<2S1stonc<2
\
Rsr
"o. .,
', Ru
. X
o; O
e \ u" e cE
o
;o
\ o :;-
\ e ;o
o 0'\
'-
(;
\
\
¡¡:; b Stot1c
RST
(o) P•'e 1n Normoliy consol1doted Cloy (bl Pil<2 1n Sond or Ov<Zrconsolidoted Cloy
sures and subsequent consolidation and possible regam m sistance during driving (see Fig. 4.13b ), and under these cir-
soil streng!h and pile-soil adhesion. As shown in Fig. 4.13a curnstances, the driving limít of the hammer m ay be reached
(McC!elland ct al., 1969), undcr these circurnstances, apile before design penetration is reached. ·
harnmer with a driving lirnit less than design capacity may
successfully drive a pile to design penetration. A prediction
of the piJe capacity on the basis of the wave equation, how-
ever, will only give the pile capacity irnmedíately after 4.4 TYPICAL SOLUTIONS FROM WAVE-EQUATION.
driving; thus, if the pile capacity sorne time after driving is ANALYSIS
required, sorne knowledge of the arnount of "set-up" (i.e.,
in crease in mil strength and adhesion with time) is required.
If the "set-up" factor (the ratio of soil strength a consider- 4.4.1 Resistance versus Set Curves
able time after driving to that immediately after driving)
can be estimated, the final load capacity, RuF, can be cal- Effects of Pile Characteristics
culated as follows: Typical solutions showing the effect of various pile charac-
teristics on the resístance versus set curves have been pre-
l sented by McClelland ét al. ( 1969). The problem considered
"_¿,¡ D.R¡sk¡
i=l
(4.32) is shown in Fig. 4.14, together with the effects of piJe
length above the ground, ernbedded length, distribution of
soil resistance, and pile wall thickness .. The effects of the
where first two factors are relatively small, while as prevíously in~
dicated in Fig. 4.8, the ultimate resístance increases as the
M¡ immediate soíl resistance in soil type í, as calcu- proportion of load taken by the pile point decreases. The
Jated from the wave equation effect of pile wall thickness is quite marked, wíth the ulti-
sk¡ sct-up factor for soil type i mate resistance increasing as the wall thickness increases.
l number of soíl Jayers through which the pile is Bender, Lyons, and Lowery (1969) have found that a pile
driven having a varying wall thickness along üs length may be ade-
quately replaced by a pile of equivalent uniform thickness.
Lowery et al. (1969) tentatively suggest that a set-up factor
of 3 míght be appropriate for soft clays, 2 for firrn and stiff Effect of Hammer Characteristícs
clays, and 1 for other soils. For the same problem shown in Fig. 4.14, the effect of the
McClelland et al. (1969), on the other hand, consíder hamrner energy is shown in Fig. 4.15. The resistance in-
that for piles driven into hard clay or sand, a decrease of creases as the harnmer energy increases, but doubling the
soil strength and adhesion with time could occur. The final energy leads only to an increase of about 28% in this case.
static resistance of the piJe would then be less than the re- For the 60,00G-ft-lb hammer, whích is widely used in the
LOAD CAPACITY BY DYNAMIC METHOD 67
1600r---,----,---,---.
Assumlld Cond1t1ons
3
HammQr affiCiqncy = 80°/c
Aluminium- mi carta cush10n 1200 Pilo Wt
block, o= o-'so ~
e 88 tons
Viscous damping sac/rt o ~
1200 1200
400 0-20 hammor L2 = 290'
"'eo
>- 800 Rp/Ru = 5"/o
L 1 = 320'
R 0jRu = 5 "lo
o '------'------'--..L...---'
o 20 40 50 80
400 Blows par Inch
20 40 60 80
Blows' pczr [nch Blows pczr Inch
(e) Effact of P1l<2 Pczncztrat1on (d) E1fact of Pczrcantagcz Po1nt Load
United States, the maximum pile-load capacity for the pile (b) Grouting a pile into an oversized hole.
considered is on the arder of 1200 tons for a practica! driv- (e) Driving a pile concentrically with an undersized pilot
ing limit of 40 blows per in. For offshore pile installations, hole.
capacities in excess of 2000 tons may be r.equired, and ( d) Driving a pile with the aid of uncontrolled drilling or
under these circumstances, using a large blow count or in- jetting.
creasing the energy rating of the hammer are nót efficient
solutions. As shown in Fig. 4.14, increasing the pile wall These four procedures are illustrated in Fig. 4,16.
thickness m ay be effective. Alternatively, McClelland et al.
(1969) suggest four possible solutions: The effect of ram shape and hammer efficiency has
becn examined by Bender, Lyons, and Lowery (1969). For
(a) Driving an insert pile through an initially installed, larg- a given ram weight and fall, a longer ram was found to be
er pile. slightly more effective than a shorter one, although the
2400r----,-----·r----~--~
<11
e
;?.2000
en
e
~ 1600
o
L
:::J
o
O)
u
e
3<11
L· 1 = 32o'
~ 400 Pile A L2 = 290'
n:: Insert pile Controlfed Uncontrofled
f'l.p/Ru= 5°/o (drlven) dnff¡ng drill1ng
( re-drlven) (re-driven)
QL---~------L----~--~
o 20 40 60 80 (a) (b) (e) (d)
Blows per lnch
FIGURE 4.16 Installation procedures currently in use for piles that
FIGURE 4.15 Effect of hammer energy (McC!elland et al., 1969). cannot be installed by driving alone (McClelland et al., 1969).
68 LOAD CAPACITY BY DYNAMIC METHOD
presence of a cushíon tended to reduce the effect of ham- the effect of relative shaft and point resistances has been
mer shape. A reduction in hammer' efficiency was found to shown in Figs. 4.8 and 4.14.
result in considerably more blows being required to drive
to a given resistance when the resistance is- high. At lower
resístances, this effect is not as pronounced; however, 4.4.2 Pile Stresses
the net increase in total driving time per pile can be quite
signifícant, and the possibility of not being able to drive the Effect of PiJe Characteristícs
piJe to design penetration must be considered. Thus, neg]ect For typical cases ínvolving prestressed concrete piles, Sam-
of hammer maintenance can seriously reduce hammer capa- son et aL (1963) have ínvestigated the influence of the
bilities. Young's moduius of the pile and the stiffness of the cush-
The effects of both hammer and pile characteristics on íon block on the maximum tensile á"nd compressive stresses
the resistance versus set curves have also been examined by in the piJe. Higher compressive and tensile stresses are de-
Mosley and Raamot (1970), who give a series of solutíons veloped for higher values of píle modulus or increasing stiff.
for various sizes of steel and concrete piles driven by two ness of the cushion block. The high tensile stresses are of
different hammers. considerable interest, especially for a prestressed concrete
pile, and could significantly influence the design of the piJe
Effect of (ushion Stiffness and the specification of driving conditions. Samson et al.
An example of the effect of cushion stiffness is shown in (1963) have noted that the time for a tensile wave to be
Fig. 4.17 (Bender, Lyons, and Lowery, 1969). As the cush- reflected back along a typical pile is less than 0.02 sec,
ion stiffness decreases, the ability to drive against soil resis- whereas the time interval between successive blows, for a
tance decreases. Figure 4.17 suggests that there is an opti- typical rate of 105 blows per m in, is 0.57 sec- that is, about
mum cushion stiffness that can provide adequate protection 28 times greater than the time for tensile-wave reflection.
for the hammer and pile while not seriously affecting the Thus, successive blows cannot be relied upon to rfduce ten-
driving capabi!ity of the system. For example, increasing sile stresses.
· the cushion stiffness abo ve about 1000 kip/in. when driving
against 800-kíps resistance does not lower the number of Effect of Hammer Characteristics
blows per foot and wílllead to higher driving stresses with Samson et al. (I 963) have also investígated the effects on
no gain in driveability. It is clear that in practice the cushion piJe stresses of ram ve!ocity and weíght, and of explosive
should be inspected at regular intervals, so that a deterio- pressure (as might be encountered in certain types of diese!
rated cushíon, whích might adversely affect the driving pro- hammers). The ram velocíty is of primary ímportance, the
cess, may be detected and replaced. stresses increasing with increasing velocíty. However, the
effects of ram weight and explosive pressure are relatively
Effect of Soil Characteristics minor.
The effects of varying the quake Q and damping factor J
have been shown previously, in Figs. 4.9 and 4.10, whíle Effect of Soil Otaracteristics ·
From Figs. 4.9 and 4.1 O, it will be seen that an in crease in
both quake Q and damping J Iead to a decrease in Ru and
hence to lower driving stresses in a given pile.
e
4000
"'
<fJ 4.5 RELIABILITY OF WAVE EQUATION
"'e
Investigations of the reliabilíty of the wave equation in pre-
e
o dicting ultima te resistance have been inade by Sorensen and
E Hansen (1957) and Lowery et aL (1969). A statistical ana-
u"'
:;J
---D, 1600 K - - - lysis of the above comparisons is summarized in Table 4.8.
: E=2000K
Despite dífferences in application of the wave equation in
L~::d:;;;;;;;;t::=±=-.L_j_ -
1
'---'---'-----'
300 400 the two cases, the results are reasonably consistent afld in· .
Rote of P<m«trat<on ( Foot) dicate that the wave equation is at least as good as the best
FIGURE 4.17 Effect of cushion stiffness (Bender etal., 1969).
of the píle-driving formulas (see Section 4.2). Lowery et ál.
LOAD CAPACITY BY DYNAMIC METHOD 69
TABLE 4.8 STATISTICAL ANALYSES Despite the above comparisons, care should. be taken
OF WAVE-EQUATIOH RELIABILITY when attempting to use any dynamic approach, whether
it be a pile-driving formula or the_ wave equation, for esti-
Reference Standard Upper Limit for Nominal Number mating the static bearing capacity of a pile, since such ap-
Deviation 96% s~Jety if Safety of Load proaches strictly only predict the piJe capacity just after
on 11 Lower Limit Is Factor Tests
driving. In soft clays, "set-up" can greatly affect the load
LO
capacity of a piJe subsequent to driving, as described in Sec-
tion 4.3.5, and unless an appropriate allowance is made,
Hansen &
Sorensen 0.23 4.0 2.7 78 serious errors in predicted capacity could result. Also, if
(1957) piles are driven thr~ugh a compressible soil that may conso-
lidate ynder its own weight or may be subjecte~ to fill or
Lowery et al. embankment loading, a downdrag force is transmitted to
0.26 3.4 2.0 31
(1969) the pile by "negative friction" acting on the pile surface.
In considering the safe load that can be applied to such
11 ~ ratio of measured to computed load capacity piles, account should be taken of this downdrag force,
which may in sorne cases be a significant part of the ulti-
( 1969) consider that load capacity is predicted by the wave mate load capacity of the pile. The calculation of down-
equation toan accuracy as follows: drag forces caused by negative friction is treated in detail
in Chapter 11.
Piles in sand: ± 25%
Piles in da y: ± 40%
Pi! es in sand & e! ay: ± 15%
4.6 PILE IMPEDANCE
Sorne comparisons have also been made between mea-
sured and predicted stresses in the pile during driving, a de-
Parola (1970) used the concept of impedance to motion
tail that cannot be predicted by conventional pile-driving
to examine energy transmission from the ram to the pile,
formulas. A typical comparison made by Samson et al.
as a function of system impedance. His simplified analy-
(1963) is shown in Fig. 4.18 for a point 9.5 ft below the
sis simulated ram-drive head impact and energy transmission
pile head. Fair agreement is found when interna! damping
to an infinite elastic rod. From a series of analyses employ-
is considered.
ing a rar:ge of hammer-cushion-pile properties commonly
used in practice, Parola found that a range of pile/hammer
impedance values would assure at least 90% efficient ener-
i,2rJo gy-transfer to the piJe, this range being expressed as:
"'"'
G
L
.:;; -- T1rTlQ 'M1ll1sQCO~ds)
"'
Cl
130(]
5 10 '15 20 peA = (0.6 to 1.10)~mn
-¡-(K) ( 4.33)
"'
~
J,
e
i1¡
>-"' CJ
where
coo 1
p mass density of pile
'o
\;
¿ e wave velocity in pile =.VEJP
"'"' -- Expcrli'rlQn~cl
E Young's modulus of pi! e
.:;;''-" 1200 --v-- TheCJret,co'
A cross-sectional area of pile
('J :;, Wram ram weight
:> Q
"'L"'
~
The implications of thls concept are as follows. Too length) was more effective in increasing penetration per
stiff a píle, and he.nce too large a pile impedance, will cause blow than was the magnitude of the impact stress. Con-
the ram to rebound, ref1ecting input energy. Piles having versely, for hard driving conditíons, pile penetration was in-
too low an impedance absorb only a portion of the ram creased more effectívely by íncreased stress amplítude than
energy, as the ram will follow the pile and retain energy. by increased ímpact duration. Thus, in the latter case, use
Either condition causes inefficient driving and may cause of a lighter ram at hlgherimpact velocity, a stiffer cushlon,
pile damage. Pile impedance also has a significan! inf1uence and a hlgher ímpedance pile all tend to produce a hlgher
on peak driving stresses. Higher impedance piling (heavier stress wave of shorter duration, and this stress-wave shape
and/or stiffer sections) induces hlgher peak stresses and will drive piling more efficiently under hard driving condi-
shorter impact durations under otherwise similar conditions. tions. Under easy driving conditions, the selection of the
Along with matching the impedance of the pile and opposite trend in any of the above variables will produce
driving system, consideration must also be given to the more penetration per blow. The judícious selectión of a
shape o(the transmitted stress wave in order to ensure the compatible hammer-pile-soil system may therefore optimíze
most efficient driving system. Parola found that pile dríve- dríveability and minimíze installation cost. It is in pursuing
abílity is directly inl1uenced by stress-wave shape. For easy thls aim that the wave-equatíon analysis probably enjoys its
dríving conditions {low resistance to penetratíon), it was greatest success.
found that longer impact duration (longer stress-wave.
SETTLEMENT ANALYSIS
OF SINGLE PI LES
db diameter of pile base · This chapter describes these methods and discusses their.
F factor of safety (>3) on ultimate load advantages and lirnitatíons. Attention is then concentrated
on solutíons obtained "frorrí the elastíc-based analysis and
Focht (1967) has examined data frorn a number ofload theír use in predicting_the load-settlement behavior of piles.
tests and has related the observed settlernent, p, at the The estirnation of the required soil parameters is then dís-
working lo:rd to the computed column deforination Pcol at cussed, and fmally, sorne comparisons between observed
the working load. Focht hás deffned a "rnovement ratio" as and theoretical pile behayior are presented.
71
72 SETTLEMENT ANALYSIS OF SINGLE PILES
5.2 THEORETICAL METHODS OF ANALYSIS l. The pile is divided into a number of segments (for sim-
plicity, three segments are shown in Fig. 5 .lb).
5.2.1 Load-Transfer Method 2. A small tip movement, Pr. is assumed (zero may be
selected, but generally the tip undergoes·some movement,
Thís method, proposed by Coy le and Re ese (1966 ), utilizes except for end-bearing piles on rock).
soil data measured from field tests on instrumented píles 3. The point resistance, P, caused by this movement is
and laboratory tests on model píles. The relevant soíl calculated. This may be done approximately ·by assuming
data requíred in this m"ethod are curves relating the ratio the pile tip to be a rigid circular area and employing the
of the adhesion ( or load transfer) and the soil shear strength Boussinesq theory:
to the pile movement. Such curves were first developed by
Seed and Reese (1957), and a typical relationshíp is shown
in Fig. 5.1. In actual problems, a number of such relation- (5.2)
ships may be requíred to describe the load transfer along
the whole length of the pile. where
The load-transfer method may be summarized as
follows: E, v are · the average deformation parameters of the
material beneath the tip, estimated from triaxial tests
or other data
where
L3 length of segmen t 3
P3 average perimeter of segment 3
(5.4)
whcre
= Q3 + Pr
Qm 2
A3 "' area of segment 3
Ep pile modulus
FIGURE S. lb Load-transfer analysis (after Coyle and Reese, 1966) 9. The new midheight movement is then given by
SETTLEMENT ANAL YSIS OF SINGLE PI LES 73
~
_j
soil with an embedded depth not exceeding 100 ft, and the
soil shear strengths in these curves have been obtained from
o
o 0.05 0.10 0.15 0.20 0.25 0.30 unconfmed compression test~.
Pile movement (in.) From a series of tests on instrumented pile in sand,
Coyle and Sulaiman ( 1967) ha ve presented data on the
FIGURE 5.2 Design load·transfer curves for pipe píles in cla)l
(Coyle and Recse, 1966). load-lransfer-versus-movement characteristics for steel piles
in saturated sand, a summary of which is shown in Fig. 5 .3.
This data suggests that for depths of O to 20 ft, curve A,
(S .S) with an upper limit of skin friction of twice the shear
strength, can be used (considerably higher values were
obtained at shallow depths). For depths greater than 20ft,
10. p~ is compared with the estimated value of p 3 from step
the measured relationships approach curve B, with an upper
(4).
limit of skin friction of 0.5 times the shear strength.
ll. If the computed movement p~ does not agree with p 3
Reese et al. ( 1969) carried out load tests to study the
within a specified tolerance, steps (2) to (10) are repeated
load transfer hlong bored piles in el ay. On the basis of a
and a new midpoint movement calculated.
curve-fitting analysis of these test results, the followii:Ig
12. When convergence is achieved, the next segment up is
relationship between load transfer (adhesion) and pile
-considered, and so on, until a value of load (Q 0 ) and dis-
movement was developed:
placement (p 0 ) for the top of the piJe are obtained.
The procedure is then repeated using different assumed 1az = 1amax [2.0)t- (~)] (5.6)
tip movements until a series of values of Q0 · and Po are
-0-----<~------~------------~----oStt
FIGURE 5.3 Design load·transfer curves for píles in sand (after Coyle and Sulaiman, 1967).
74 SETTLEMENT ANAL YSIS OF SINGLE PILES
where
The máin difference between the various methods líes in
Taz adhesion at depth z (tons/ftl) the assumptions made regarding the distributíon of shear
Tamax = maximum adhesion that can occur at any stress along the pile. D'Appolonia and Romualdi, Thurman
depth (tons/ft 2 ) and D'Appolonia, and Salas and Belzunce assume the
p = downward movement of pile at depth z (in.) · shear stress at each element to be represented by a single-
So = 2de (in.)
point load acting on the axis at the center of each element.
d = pile diameter Nair assumes a uniformly-loaded circular area at the center
E = average failure strain, in percent, obtainea
of each element. Poulos and Davís, Mattes and Poulos,
from stress-straín curves for unconfined com- and Poulos and Mattes s;onsider a shear stress distríbuted
pression tests run on soil sarnples riear the pile uniforrnly around the circumference of the píle. The latter
tip appears to be the most satisfactory of those men!ioned,
especially for shorter piles. However, for relatively slender
Although the load-transfer method has gained quite piles, there is very little difference b~:tween solutions
wide acceptance, the following theoretical and practica! based on the three above representations of shear stress.
limitations should be recognized: In the derivations described below, the method of Poulos
and Davis (1968), among others, is followed. The basic
(a) In using the load-transfer curves, it is inherently as- problems of a floating or friction pile in a semi-infmite mass
sumed that the movement of the pile at any point is related and an end-bearing pile are considered in detail and modifi-
only to the shear stress at that point and is independent of cation,s to these analyses are described.
the stresses elsewhere on the pile. This inherent assumption
ís equivalent to that rríade when the theory of subgrade 5.2.2.1 BASJC ANALYSIS FOR SINGLE FLOATJNG PJLE
reaction ís used to analyze laterally loaded piles. Thus, no
The pile is considered to be a cylinder, of length L, shaft
proper account is taken of the continuíty of the soil mass.
diarneter d, and base diarneter db, and loaded with an
(b) The load-transfer method, beca use of its inherent dis·
axial force P at the ground surface. For the purposes of the
regard for continuity of the soil, is not suitable for analyz.
analysis, the pile is acted upon by a system of uniform
ing load-settlement characteristics of pile groups.
vertical shear stresses p around the periphery, and the base
(e) In order to obtain load-transfer curves at a site, con-
is acted upon by a uniforrn vertical stress Pb, as shown in
siderably more instrumentation is required on a pile than
Fig. 5.4. The si des of the pile are assumed to be rough. The
for a normal pile-load test. Extrapolation of test data from
soil is initially considered to be an ideal homogeneous iso-
one site to another may not always be entirely successful.
tropic elastic half-space, having elastic pararneters E5 and v 5
that are not influenced by the presence of the pile. Modifi-
5.2.2. Analysis Based on Elastic Theory cations for more realistic representation of soil behavior
will be discussed later. Unless otherwise stated, db will bé
Elastic-based analyses have been employed by severa! taken to be equal to d.
investigatórs: for example, D' Appolonia and Romualdi As in alrnost all methods of pile-settlement analysis,
(1963), Thurrnan and D'Appolonia (1965), Salas and Bel- it is assumed, that the pile and soil are initially stress-free,
zunce (1965), Nair (1967), Poulos and Davis (1968), Mattes and that no residual stresses exist in the pile resulting from
and Poulos O969), Poulos and Mattes {l969a), Butter- its installation. Holloway et al. (1975) emphasize the irn-
field and Banerjee (197la, 1971b), Banerjee and Davies portance .of residual pile-soil stresses on pile behavior and
(1977), Randolph and Wroth (1978).In most of these ap- on the interpretation of pile-load tests, and suggest a meth-
proaches, the pile fs divided in.to a number of uniforrnly- od for evaluating su eh stresses. However, in order to reduce
loaded elements, and a solution i:; obtained by irnposing the complexity of the 'analysis here, the assumption of an
compatibility between the displacements of the pile and the initia!Jy stress-free pile is adopted; as subsequently will be
adjacent soil for each element of the pile. The displace- demonstrated for predicting pile settlements, the influence
.rnents of the pile are obtained by considering the compres- of the residual stresses may be adequately taken into
sibility of the pile under axial loading. The soil displace- account by choosing appropriate values of the soil modulus.
ments are obtained in most cases by using Mi~dlin's equa- If conditions at the pile-soil interface remain elastic
tions for · the displacements within a soil mass caused by and no slip occurs, the movements of the pile and the
loadíng within the mass. adjacent soil must be e qua!. The correct values of the stress
SETTLEMENT ANAL YSIS OF SINGLE PILES 75
d
Soii-
Young's
h modulllS, E,
Shear Poisson's
stress ratio, v!
p Young's modulus, E0
0,.____ _ _ _ __
Displacernent, p / Rigid stratum
system p* and base stress Pb will be those that satisfy this piers of normal proportions very similar solutions to those
condition of displacement compatibility. ldeally, consider- from the simple analysis described below.
ation should be given to compatibility of both vertical and In order to obtain a solution for the values of p, Pb, and
radial displacements, and a normal stress system u should the displacement of the pile, it is necessary to obtain ex-
also then be imposed on the pile elements. However, as will pressions for the vertical displacement of the pile and the
be discussed, this more-complete analysis gives solutions soil at each elernent. in terms of the unknown stresses on
that are generally almost identical with those from a sim- the pile, impose the compatibility condition, and solve the
pler analysi~ that considers only vertical displacement com- resulting equations.
patibility: therefore, only the simpler analysis is described
in detail. Details of the more complete analysis are given by Soil Displacement Equations
Butterfield and Banerjee (1971) and by Mattes (1972). It Considering a typical element 1 m Fig. 5 .4, the vertical
has also been shown by Mattes (1972) that an even more displacement of the soil adjacent to the pile at i resulting
refmed analysis, one recognizing the difference between !he from the stress p¡ on an element j can be expressed as
fictitious and the real stress-systems, gives for piles and
d .
• The shear stresses (p) are fictitious in that they represen! trac- sPij = -J~p¡ (5.7)
tions applíed to the boundaries of the imaginar y surface in the half-
space representing the piJe surface, and are not n"ecessarily the actual
stresses acting on the real-pile surfaces. Once the. values of p are
where
determined, the actual stresses and displacem~ta they produce
anywhere in the half-space, including the real-piJe boundaries, may Iq = vertical-displacement factor for element i due to
be calculated. shear stress at elernent ¡
76 SETTLEMENT ANAL YSlS OF SINGLE PlLES
As a result of al! n elements and the base, the soil dis- ring to Fig. 5 .4d, consideration of vertical equilibrium of a
placement at i is small cylindrical pile element yields
(5.8) (5.11)
where where
Iib = vertical displacement factor for element i due to a == axial stress in piJe (average over the cross section)
uníform stress on the base p shear stress on piJe surface
A similar expression may be written for the base; and The axial strain of this elemen t is
for all elements on the pile, the soil displacements may be
written as a*
(5.12)
Ep
[/s] fp} (5.9)
where
(5.14)
where
or "point··bearing," but the results of severa! analyses and Soil Displacement Equations
field and laboratory measurements have shown that a signi- To properly determine the displacement in the soil sur-
f~ant proportion of the load may be transferred from the rounding the piJe, it woúld be necessary to use equations
piJe shaft to the surrounding soil. foi loading within a two-layer elastic system. Since suitable
To analyze the behavior of such piles, the analysis de- analytical solutions to this problem are not available, Mind-
scribed in the preceding section, for a floating pile, must be lín's equations for displacements caused by loading within a
modifíed to allow for the effect of the stiffer bearing-stra- half.space may be utilized in an approximate manner. To
tum. The same assumptions are again made for the pile and allow for the reduction in soil dísplacements because of
soil behavior, but in addition the bearing stratum is as- the presence of the bearíng stratum, a method ís used that
sumed to be an ideal elastic half-space with constant para- is an extension of the "mirror-image" approximation sug-
meters Eb and Vb. The problem is defined in 5.5a. To gested by D'Appo!onia and Romualdi (1963) for piles
obtaín the solution for the unknown stresses on the pile bearing on rock. Referring to Fig. "5 .Sb, pile elemént j ís
shaft and tip, and the corresponding piJe movements, com- mirrored in the soil-bearing stratum interface by an imagi-
patíbility of the vertical displacements of the pi! e and adja- nary pile element /, which is acted on by stress kpj in the
cent soíl aré again considered. opposite direction to the stress P¡ on the real element j. The
p
¡
Pii<Z
Young' s cnodulus E,
-- r-- d
Baar~ng Strotum •
Young 's modulus Eb
PQISSOn. S ratiO Vb
Int<2rf:Jc<Z batweer
h. so11 ond bear1ng
stratum
pb
(e) Stress<2s Actng on F 1112
L
!mog,nary El12m<2nts J
¡'
lirniting values of k are k·= O for a floating pile where Eb dívided into three components: displacement causeá by
in which case the stratum has no effect on soil dis- shear stresses along the pile, displacement caused by applied
placements; and k 1 for a pile resting on a rigid bearing- axial load P, and displacement caused by the finite com-
stratum (Eb oo), in which case the condition of zero pressibílity of the bearing stratum. Again assuming only
tip-disp]a¡;ement is satisfied. In general, k must be deter- a;o¡.ial compression of the píle, it may be shown (Poulos and
mined as part of the analysis, and it is one of the assump- Mattes, 1969a) that the dísplacement vector {pP} for the
tions of the analysis that it has thc .same value for all n elemen ts along the shaft is
imaginary elements.
Taking downward displacements of the soil as positive, {pP} -{Ep~A [Dp] (5.23)
the displacement sPií of the soil at i because of the stress on
the real element j and the irnaginary element j', is
+ (1T(l ¡.;b 2)) (~)(~)[X]}. {p} ·tpR:1Té {h}
d ( ,
--pi l¡¡ - kl íJ)
sPíj (5.20)
ES
where
where
Iq vertical dísplacement factor for i due to shear
stress on elementj, as before [Dp] = n X n matrix of pile displacement factors, with
vertical displacement factor for i due to shear Dp¡¡ = 4ohid for í ~j
stress on imaginary element ¡' (cai,culated for a
dístance L + h¡ from the irnaginary soil surface) or
If we make the simplifying assumption that the influence Dp¡¡ = 4oh¡d for ¡;;;:. ¡ ·
of the stress on the pile tip has a negligible effect on the o = L/n
soil displacement at i, sPi, then h¡, hi = distances from bearing stratum to points i and
j (see Fig. 5.5c)
{h} = n column vector of values of h¡
. (5.21) [X] n X n matrix, every term of which is unity
{p} · = n column vector of Pi val ues
and foral! n elements along the píle shaft (not including the
{ W} n column vector of values of unity
RA = arearatioofpile(Eq.5.l0)
pile tip),
Displacement Compatibility
(5.22) Assuming again no slip at the pile-soil interface, {pP} =
{sP}, so that from Eqs. (5 .22) and (5 .23),
where
(5.24)
{sP} and {p} vectors of soil-displacement and shear-
stress, respectively ( of arder n)
[Is - ,u;] 11 X n matrix of values off¡;;- kJ¿.
piJe base rnay then be evaluated from the equilibrium equa- 5.2.2.3 MODIFICA TION TO BASIC ANALYSES FOR SINGLE
tion PILE
n
Pile-Soil Slip
P ==
4 Pb + nf-n ""p·
.t::...J 1
(5.25)
The analyses described above require that no slip occurs at
j=l
the pile-soil interface. However, sin ce real soils ha ve a finite
Having obtained the solutions for the chosen value of shear strength and the pile-soil interface has a finite adhe-
k, a closer estimate of k for the particular piJe, soil, and sive strength, slip or local yield will occur when the shear
base parameters being considered may be obtained by ex- stress reaches th~ adhesive ( or yield) strength. By use of a
amining cornpatibility between the displacements of the method similar to that descríbed by D'Appolonía and
soil and the bearing stratum at the pile base. The soil dis- Rornualdi (1963), Salas (1965), and Poulos and Davis
placement at the piJe base is, from Eq. (5.24), (1968), the elastic analyses can be modified to take ac-
. count of possible slip, provided that the following assump·
tions are made:
sPb (5.26)
l. Local yield or slip occurs at the pile-soil interface
where when the average shear stress on any piJe elernent, calcu-
lated from the elastic analysis, reaches the limiting value
vertical-displacernent factors for the center fa.
{Y} (5.30a)
Thus, by equating Eqs. (5.27) and (5.28), the next
approxirnation for the value of k is as follows:
or
elastic. From the resulting solution for {p}, the shear-stress is more accurate than considering only a thin, enlarged
vector{p} is calculated from Eq.(5.14) or Eq.(5.9). These base.
shear stresses are then c~mpared with the specified limiting For the case of apile with a rigid cap resting on the soil
stresses r0 . At an element where the computed stress ex- surface, uniformly loaded anriular elements are included in
ceeds r 0 , the displacement-coznpatibility equation for that the analysis, to represent the pile cap. Compatibility of pile
element (i.e., the appropriate row of the. matrix [Z] in Eq. and soil displacements is considered at these cap elements,
5 .30b) is replaced by the pile-displacement equation for as well along the pile. Details -of such an analysis are given
that element (i.e., the appropriate row of the matrix in by Poulos (1968b) and Butterfield and Banerjee (1971b).
equation 5.14), putting the shear stress at that element The effects are discussed later in this chapter and again in
equal to r 0 • For example, if an element i has slipped, the Chapter 10.
elements Z¡¡ in row i of the rñatrix Z in Eq. (S .30b) are re-
piaced by the elements pl¡¡ of matrix pi, while the element 5.2.2.4 ACCURACY OF ELASTIC SINGLE-PILE SOLUTJONS
d Y¡ on the right-hand side is replace~ by E ;L ~ 2 (raí Investigations into the sensitivity of the solutíons on the
~ . P A n
number of elements used in the analysis have shown that
- Y¡), where r0 ¡ is the value of r0 at element i. the _u~men ts to _diV.iA~_!he pg~_.S.h~ft.. le.A<:b__19
The modified system of equations is now resolved and answers of acceptable accuracy unless the pile is relatively
the procedure is repeated until the computed values of long (Lid> 50) or very compressible (K< 100), in which
shear stress do not exceed the limiting values r0 • case 15 or 20 elements may be desirable. For short, stiff.
By successively increasing the applied load P until al! piles, even the use of S. elements gives accurate solutions.
elements have failed, a load-settlement curve to failure may The use of a single base element and the application of a ri-
be obtained. gidity-correction factor (see Appendix A) also appears to
Analyses taking account of pile-soil slip along the shaft be quite satisfactory, as the solutions are almost identical
have revealed that for nom1al piles having Jength-to-dia- with those obtained by the use pf S · annular elements to
meter ratios greater than about 20 and for constant r0 , the divide the base.
load-settlement curve ís substantially linear until a load of Complete solutions for the settlement of a pile, in
at least 50% of the failure load is reached. For the predic- which both vertical and radial displacement compatibilíty
. tion of settler:1ent at working loads for such piles, a linear- are considered, have been presented by Butterfield and
elastic analysis is theref;:;re adequate. For larger-diameter Banerjee (1971a), and Mattes (1969; 1972). Comparisons
piles or piers, full shaft slip may occur at relatively low between the complete solutions and solutíons in whích
Joads. For such cases, a simplified procedure for obtaining only vertical-displacement compatibility is considered,
the load-settlement curve, described m· Section 5.4., has are shown in Fig. S .6 for the shear stress along the pile
been developed. (Mattes, 1969) and in Table S .1 for the top displacement
of the piJe (Mattes, 1972). Only for relatívely short pi! es
Other Modifications (L/d<25) does the inclusion of radial-displacement com-
The basic analyses have been formulated in terms of a uni- patibility have any effect on the solutions, and even in such
form piJe with provision for a thin enlarged base. However, cases, the effect is unimportant' from a practica! point of
extensions may readily be made to allow for cases in which view. It therefore appears quite adequate to employ ana-
the shaft is not of uniform diameter or in which the pile is lyses in which only compatibility of vertical displacements
attached to a pi! e cap resting on the soil surface. is considered.
For piles having nonuniform shaft diameter, the relative Although the analysis described in Section S .2 .2 .2 is
diameters of the various shaft eleinents are considered when primari.ly developed for end-bearing piles, it may be used
calculating the pile and soil displacements. In thosé cases to obtain solutions for a floating piJe in a uniform mass
where the shaft diameter of an elementis less than that of by puÚing Eb!Es = l. The possible errors involved in the
the element above it, the stress on the annular area at the analysis are a maximum for Eb!Es 1, so that by com-
j·uhction of the two elements must be included as an addi- paring this solution with the corresponding solution from
tional unknown. Examples of the analysis of underreamed the floating-pile analysis described in Section 5.2.2.1 ,.an
and step-taper piles using the above approach have been estímate may be made of the inaximum error of the end-
given by Poulos (1969). For piles with an enlarged base of bearing piJe ana!ysis. Comparisons of the settlements of
relatively large volume, the shaft elements neár the base can the top and tip of a pile obtained from the two solutions
be considered to have an increa.sed diameter; this approach have been made by Poulos and Mattes (1969a), and reJa-
l
10
:'l0.8~ el~~
Without radial compatibility
t~Ur'
l
o 0.4 0.8
-- 1.2 16
p1rd LIP
(a) Distríbutíon of si de shear ( v, 0.5)
L
.b o d
10
10
d - - 1';
7
L
ft
1
0.8 08~
o
prrdl/P
ondl/P
(b) Distribution of side shear 1 v, O)
(e) Distríbut ton of radial stress
p
Displacernent p =-lp
LEs
SETTLEMENT ANAL YSIS OF SINGLE PILES 83
tively close agreement has been found (see Figs. 5.24 and analysis to investigate the effects on settlement of non-
5.25). The end-bearing analysis, when applied to the homogeneity of the soil that might aríse during installatio.n
floating pile, results in a slight underestimate of settlement of the píle. In a further application, Balaam et al. (I:976)
(maximum effect about ·10%) and a slight overestimate of have used thls approach to analyze the behavior of piles
the amount of load in the pi! e. From a practica! point of composed of grave!. ·
view, the errors involved in the end-bearing pile analysis are
unlikely to be significant, especially for E biEs >l.
5.2.4. Comparisons between Solutions from
5.2.3 Finite-Element Analysis Mindlin Approach and Finite-Eiement Analysis
Detailed descriptions óf the finite-element method have Balaam et al. (1975) obtained elastic solutions for the case
be en given by Zienkiewicz ( 1971) and Desai and Abe! L/d = 10, K= 1000, h/L 2, and Vs = 0.45. Twénty trian-
(1972), and its use in geotechnical problems is discussed gular elements were used for the pile and 160 triangular
comprehensively in Desai and Christian (1977). The appli- elements for the soil. A free outer boundary was assumed
cation of finite-element analysis to pile foundations has at 35 pile-diameters from the píle axis, the base underlying
been described by severa! investigators. Ellison et aL (1971-) the soil was assumed at 35 pile-diameters fro111 the piJe
have considered. a multilinear soíl stress-strain curve and axis, and the base underlying the soil was assumed to be
have introduced special joint elements at the pile interface rough and rigid. The settlement at the top of the pile was
to allow for slip. Dcsai (1974) has considered apile in sand found to be only 2.0% less than that given by the previous
with a hypcrbolic stress-strain response an~ has also used analysis utilizing Mindlin's equations. Furthermore, the
special elements for the pile-soil interface. Hyperbolic finite-element solution was identical with the conventional
stress-strain behavior has also been used by Esu and Otta- elastic-finite-element solution in which the pile and soíl are
viani (1975) for analyzing apile in clay. A very interesting analyzed together as a single mass. Decreasing the number
result of theif analysis is that the load-settlement behavior of pile elements to 1O and the soil elements to 120 in·
of a pile is substantially linear to a load we!l beyond half creases the discrepancy between the finite-elemcnt solution
the failure load, despite the fact that the soil stress-strain and the elastic solution to 3.5%. In a parametric study of
response is nonlinear. This fact suggests that elastic theory, the settlement of apile presentcd by Lee (1973), the solu-
modified for slip as previously suggested, should prqvide an tions are obtained from a finite-element analysis. Table 5.2
adequate basis for load-settlement prediction, provided shows a comparison between Lee's solutions for a f!oating
appropriate values of soil modulus are used. pile in a uniform mass and the corresponding solutions
Lee (1973) and Valliappan et al. (1974) have done elas- from the elastic analysis presented herein .. In this case, the
dc parametric studies of the influence of soil layering on finite-element solutions are slightly greater, but gene rally
settlement behavior. The superior accuracy of isoparametric there is close agreement between the two series of solutions
elements over conventional elements is also demonstrated. and such difference as does exist may well arise from
Balaam et aL (1975, 1976) have used a different type of numerical inaccuracies in one or both of the solutions.
analysis, in which the finite-element method is used to ana· ·A further comparison with Lee's solutions is shown
lyze the píle and soil mass separately and then compatibi· in Table 5.2, this time for a pi! e bearing on a stiffer stra-
lity conditions are imposed to determine the nodal forces tum. The agreement is again reasonable, and these compari-
and deflections. This approach is thus a generalization of sons suggest that the analysis based on the Mindlin equation
the elastic approaches descríbed earlier in this chapter. The s.~ould give results of adequate accuracy for practica! pur-
possibility of slip at the pile-soil interface is allowed for by poses, provided that severe variations in subsoil .conditions
specifying a limiting pile-soil shear strength, from which do not occur along the piJe (see Section 5.3.3 for further
limiting values of nodal force can be calculated. Possible discussion of the effects of soillayering).
failure within the soil mass itself is allowed for by consi- A comparison between computed load-settlement
dering the soil as a bilínear elastic or elastic-plastic materiaL c'urv:es to failure for apile in a purely cohesive soil is shown
Thls type of approach appears to hold some advantage over in Fíg. 5.7 (Pu is the ultimate load capacity). The agree-
the use of joint elements in that the rate of convergence of ment is generally reasonable, but at loads approaching the
the solutíon is much more rapíd when pile-soíl slíp or soil-. ultimate, the settl~ments given by the finite-element analy-
yield has occurred. It a]so overcomes problems that may sis are greater than those from the "elastic" approach, pro-
arise when there are extreme differences between the mo· bably because the latter uses elastic theory to calculate soil
duli of the pile and the soil. Balaam et aL {1975) used this deflections after pile soil-slip has commenced.
84 SETTLEMENT ANAL YSIS OF SINGLE PILES
TABLE 5.2 COMPARISONS BETWEEN ELASTIC- AND 5.3 THEORETICAL SOLUTIONS FOR
FINITE-ELEMENT SOLUTIONS FOR PILE SETILEMENT SETTLEMENT AND LOAD DISTRIBUTION
K = 1000
To enable rapíd practica! estimates of pile-settlement be-
(a) Floa tíng Pi! e in Semi-infinite Mass havior, it is extreme! y use fui to have available dimension-
less parametric solutions from which the effects of varia-
tions iri pile and soil properties can readily be determined.
In this section, a series of solutions is presented for the
Finite Elementa Elastic Mindlin stress and load distribution in a pile, and for the settlement
Approach of a single piJe. The soil is assumed to be homogeneous,
having constant Young's modulus and Poisson's ratio. How-
3.5 0.267 0.258 ever, the íntluence of nonhomogeneity and soillayéring is
5.0 0.211 0.205 also discussed. The solutions described have been obtained
10.5 0.115 0.112 from the analyses based on Mindlin's equations, and in
15.0 0.103 0.100
19.5
most cases, 10 elements have been used to divide the pile
0.094 0.092
shaft.
a Lee (197 3).
40 ,..........,------
/
/ " i-= 25
/ 0·4-
/ p
30 / z
{~}
/
/. L
l.
t
Ep/ = 1000 0·6
/Es
>.- d
'X. = 2 1
....j
1
1 '11
e% ·1 t
1 1
V 0-48 0·8- 1 1 v, ~o
1 1
v,,=05-
\K" 50 \K='.iOOO
•1 ''
''
4 6 8 1·0 ' '
o 1·0 2·0 3-0
'Ttdl
FIGURE 5.7 Comparíson between load-settlement curves to failure. FIGURE 5.8 Distributíon of shear stress along compressíble píle.
SETTLEMENT ANAL YSlS OF SINGLE PILES 85
where
FIGURE 5.1 ~ Poisson's ratio correction factor for base load, Cv.
a) FloatingPile
(5.33)
where
I = IoRKRhRv (5.33a)
p = settlement of pile head
P = applied axial load
100 1000
lo . = settlement-intluence factor for incompressible
K piJe in semi-infinite mass, for Vs 0.5.
FIGURE 5.12 Compressibility correctíon factor for base load, CK· RK = correction factor for pile compressibility
100
(O).!::_'
<1
1L-----~------~----~ 1 ~----~----_.----~
1 100 1,000 1 10 100 1,000
eb
FIGURE 5.14 Base modulus correction factor for base load, Cb,
87
"88 SETTLEMENT ANALYS!S OF SINGLE PILES
5 Values of K -·-+---:-c·-:-:----l
500 1,000 ~ 20,000
~ 0·81----~
'tJ.:!!
Eñ:
¡¡ §
7~
~ 0·6
e
~:::>
e ~
:::> .e
1> ¡;;O· 4 f.----+
o
'O ..J
8 "'
_Jf=
!!! 0·2~-------+
2 3
Bulb diomeler
FIGL'RE 5.14 (continued).
Shott d1ometer '
correction factor for finite depth of !ayer on a FIGURE 5.16 Influence of underreaming on tip load {incompress-
rigid base ible floating pile with single central bulb, 0.2L long).
correction for soil Poisson's ratio Vs
total depth of soillayer
a finite !ayer is to decrease settlement (Fig. 5 .20). [If the
Values of 10 , RK, Rh, and Rv are plotted in Figs. 5.18, hard base is leve! with the pile tip, case (b) should be used.]
5.19, 5.20, and 5.21. Figure 5.18 shows the decrease in A decrease in Poisson's ratio, Vs, while maintaining Es con-
settlement of a pile of constant diameter as the length in- stant leads to a decrease in settlement, as shown in Fig .
. creases. The presence of an enlarged base also decreases set- 5.21, althOugh the effect is relatively small.
tlement, although the effect ís only significant for relatively
short piles. PiJe compressibility increases settlement, espe- b) End-Bearing Pile on Stiff'er Stratum
cially for slender piles (Fig. 5.19), while the effect of having
PI.
p (5.34)
Esd
H)
0·8
.,
g.d:
u }E~
+ § 0·6
.:!! .ee 0::0::
a: :J ~ E
.e .e ~~ 0·6
o e
'O 'O 1-:::>
o 8 0·4 ~ ~
FIGURE 5.15 lnfluence of pile cap on tip load. Incompressible FIGURE 5.17 Influence of tapered or step-tapered pite on típ load
floating pile. (incompressíble floatíng pílc, top diameter =á).
SETTLEMENT ANALYSIS OF SINGLE PILES 89
1' o !
i
i 1
0·8
i
i 1 1
--
0·6
1 !
l
i :
0-4 1
\
0·2 ~ 1
\ ~ o~ Volues d01¡d
0·1 ~~··
1'\M!
~~- ~3 "'-...~'.."'-.
1
1
!
i
For ! "100
e- . !----·-·· .- lo oo254 L-
For 3;.d% ;;.1-
1
:
0·02 ! : 1 1 r
o 10 :?0 30 40 50
VS
where
I = I~xRbRv (5.34a)
Uo, Rx, Rv are defined as for Eq. 5.33 and take
the same values to sufficíent accuracy)
Rb correctíon factor for stiffness of bearing stratum
10 1000
1·0~-----------r------------~----------~
0·4
111R = ~_S_e~t_tl~em
__e~n~t~o~f~p=ile~-
are assumed to be mutually independent; for example, the
·;;:; (5.35)
effect of finite !ayer depth is assumed to be independent of Elastic shortening of pile
píle stiffness factor K. Whíle this m ay not be strictly cor-
rect, the use of the correction factor allows a convenient Theoretícal values of MR are shown in Fig. 5.23 for a pile
parametric presentation of results and should be of ade- on a rigid bearing-stratum and in Fig. 5.24 for a 25-dia-
quate accuracy for practica! purposes. meter pile resting on a nonrigid stratum. The pile-head set-
tlement is calculated as
5.3.3.1 MOVEMENT RATJOS
o o 001 L--.L--'----'---'----'----::-:":-:::'--::-::-'
100 2JO 500 1000 200C 100 200 500 1000 2000 5000 10000
K K
FIGURE 5. 23 Movcment ratio for end-bearing pile on rigid base. FIGURE 5.25 End-bearing piJe on stiffer stratum. Pile-tip move-
ment ratio.
MR lies within the range 0.5 to 2 for most practica! pile di- 5.3.3.2 EFFECT OF PILE-SOIL SLIP
mensions and this observation can be said to generally agree
For a floatíng pile in a purely cohesive homogeneous soil,
with the theoretical results in Figs. 5.23 and 5 .24.
with constant adhesion ca along the shaft, the ínfluence of
A plot of the theoretical movement ratio for the pile
slip on settlement is shown in Figs. 5.26 and 5.2 7, in terms
tip,MRt• is shown in Fig. 5.25.
of plots of a slip factor M 5 and the factor of.safety against
undrained failure, where
K="""
LO Ms V,=0·5
0·4~--~------~-~--+-----l~-----;
Cunuz
1
0·2 ····--2
3
0.2!---·--·f-------+~-~+-
oJ._--~.
100
___.__
200
_J.._ -':-:-:::7:::-:::--'
_J.._ _ _ o
2·25
1
FIGURE 5.24 Movement ratio for end-bearing pile on ~·.Lffer FIGL"RE 5.26 Settlement modification factor ,IJ,f8 for slip. Effect of
stratum. L/d and adhesion factor.
92 SETTLEMENT ANALYSIS OF SINGLE PILES
in a two-layer sol, the upper layer being of depth h 1 (h 1 In cases where the pile passes through distinct layers of
< L) and the lower !ayer being of infiníte extent. These soil, having large differences in soil modulus, the uniform
approxirnate computer solutions for pile settlernent are soil solutions may be utilized in an alternative approximate
shown in Fíg. 5.23 for L/d 25. Also shown are the solu- fashion. For example, for the simple case of apile penetrat-
tions for a point-hearing pile (h 1 "'L) for E¡/E2 2 and 5, íng one !ayer and founded in a second !ayer, the settlement
obtained frorn the analysís described in Section S.2.2, and may be estimated by treating the portian of the pile in the
finite elernent solutions frorn Valliappan et al. (1974). first !ayer as an end-bearing piJe and determining the settle-
Finally, approximate solutions are shown that use the dis- ment of this portian and the arnount of load in the pile at
placernent-fnfluen.;e factor for a hornogeneous soil and an the interface of the two layers (say, P2 ). The settlement is
average rnodulus, F;av, as foÍlows: added to the previously calculated settlernent of the upper
portian to obtain the overall settlement of the pile head.
Comparisons with the finite 'element solutíons of Lee
(5.38) (1973) indicate that this approximate approach gives a
settlement within 20% of the finite-element solution, the
Figure 5.28 shows that at least for E 1 >E 1 , the approx- accuracy increasing as the modulus of the bearing stratum
imate computer solution for h 1 "' L overestimates the set- increases relative to that of the overlying soil.
tlernent somewhat. as compared with the point-bearing pi! e Despite the apparent success with which the settlement
analysis, but the e ·ror is not great. Furthermore, the simple of a pile in a nonhomogeneous soil may be estimated
solutions employirg thc average modulus Eav are in reason- approximate methods, it must be borne in mind that such
able agreement with thc approximate computer solutions methods will probably not give an accurate picture of the
and in sorne cases are in {act in closer agreement with both distribution of load and settlement along the length of piie,
the solutions for a point-bearing pile, and the finite-element a more refined analysis is warranted and necessary íf such;;
solutions. picture is required.
On the basis of the above evídence, it is suggested that
where the soil mc•dulus varíes along the length of the pile 5.3.3.4 OTHER EFFECTS
(e.g .• where a nurrber of layers occur), and where the mo-
The effects on settlement of enlarged bulbs, tapering of the
dulus variation between successíve layers is not large, the
pile and of a pile cap resting on the soil surface have been
settlemenr, may b~ calculated from the expressions for a
investigated for an incompressible pile in a semi-infinite
pi! e in uniform soJ (Eq. 5.33 or Eq. 5 .34) using an average
mass (Poul0s, 1968a; 1969). These effects are shown in
soil modulus Eav a:; follüws:
Figs. 5.29, 5.30, and 5.31 in terms of the settlement of a
uniform-diameter freestanding pile. The presence of en-
,¡
( 1) "';'
-
L
..:;.; E¡h¡
jocj
(5 .39)
where
E¡ modulus of !ayer i
h¡ thicknes1 of !ayer i
11 = number of different soillayers along pile length
- - •., 0,5
OA
a: 0,2
o
2 4 6 8 10
0,1 0.08 0.06 0.04 0,02 o
d
FIGLRE 5.31 Effect of pile cap on settlement. Incompressible piJe in half-space, rigid cap, diameter de-
SETTLEMENT ANAL YSIS OF SINGLE FILES 95
10
01
005
01 02 03 04 05
2 15 05 o
r .b.
L r 08 o6 04 02 o
r L
FIGURE 5.32 Displacement influence factors. L r
different soíJs. lf the first !ayer is of depth h 1 (h 1 > L), the identical with those given by the Mindlin equation for a
settlement p of the piJe is given approximately a:;: poinf-load acting at a depth of 2L/3 below the wrface, that
is,
P -_ Po 1m + m""
+ -P [ ~ Ij--Ij+-
..:::.,¡ -
1( 1)] · (5.41)
L ESJn 2 E. s¡ (S .42)
where
The first term of the above equation is the sett!Ement . Equation (5.41) is similar to that suggested by Nair (1967)
of the piJe in the founding !ayer ( depth h 1 , modulus Es 1) for piles and that used by Egorov, Kuzmin, and Popov
and the second term represents the summatíon of the dis- (1957) for surface foundations, and makes use of the Stein·
placements of the underlying layers caused by the piJe. It brenner approximation that the stresses in a layered system
should. be noted that the value of Ur Ij+J) calculated for are the same as those in a uniform mass. The use of this
layer i should be that for a value of K corresponding to a equation will generally lead to an overestimate o[ the settle-
soil modulus although for deeper layers, the value of ment caused by the underlying layers if th<~ modulus of the
K used has almost no influence on the calculations. layers decreases with depth, but Poulos and Mattes (197lb)
For application of (5.41), it is convenient to have estímate that this error is not serious unless the modulus of
values of the influenc•: factor lp on the axis plotted against successive layers varíes by more than a factor of 10. If
depth, and such a plot is shown in Fig. 5.35 for three values softer layers overlie stiffer layers, little error is involved in
of L/d and for Vs"' 0.5. The effect of Lfd becomes insigni- using, . (5 .41).
ficant for HIL >
1.75. The influence factors then become For cases in which layering occurs along the pile as well
as beneath the pile, an average value of Esl should be e'sti-
1.0 mated from Eq. (5.39).
f---~5···· 1Values ol -~
1---- 1 -
i---- -
.:..-~~
p •....
···-· ··-· -
f.---~
w~
-·
¡::.=h..\ - 5.3.5 Immediate and Final Settlements'
1---f ....... ~ ~ !H -
1---c-·· ~ r·J=:!= ----
For piles in sand or unsaturated soils, the final settlement
(exduding possible creep movements) may be considered to
[', K- 1000 occur immediately on application of the load, so that the
~
v, 0.5
1
values of Es and Vs u sed in calculating the settlement of the
---· ... -~
f··
"'\ drained conditions followed by a tírne-dependent consoli·
dation settlement. After dissipation of the excess pore
pressures resulting from loading of the piJe is complete, the
f -- -- t--· \
·"")· ....
tot?l settlement <Jf the pile is PTF (pTF =Pi+ PcF, where
0.01
o
!
2 3 4 5
1
1
' ~ PCF final consolidation settlement). The imrnediate set-
t!ement Pi is calculated from the theoretical solutions by
using the· undrained Young's modulus of the soil, Eu, and
f::l 0.2 0.1 o the undrained Poisson's ratio, Vu, which is 0.5 .for a satu-
L L
H rated soil. The final settlement PTF is calculated by using
FIGURE 5.35 Innuence factors for seitlement beneath center of a the drained Young modulus of the soil skeleton, E;,and the
píer. drained Poisson's ratio v~.
SETTLEMENT ANALYSIS OF SINGLE PILES 97
0·1 o
FIGtlRE 5.36 n,e relative importance of immediate settlement for an incompressible pile in a semi-infinhe mass.
5.3.5.1 RELATIVE IMPORTANCE OF !JI1MEDIATE tends to decrease with increasing pile compressibility (i.e .,
SETTLEMENT decreasing K), but it still remains the most significant part
of the final settlement. For end-bearing piles, virtually the
It is possible to examine the relative magnitudes of the
entire settlement of the pile head is immedíate settlement;
ímmediate and final settlements of a pile if the soil is as-
sumed to be an ideal two-phase elastic homogeneous isotro·
only for compressible slende.r piles (L/d > 50, K < 500)
does the consolidation movernent exceed 1O% of the total
pie material. For such a material, the undrained and drained
final movement.
moduli may be shown to be related as follows:
Pi
(5.44)
PTF
where
·00001
e
~ 0·21----···~·-_;_=-~----~----·-·-
-m
U'J
9 041-------- ----'----
0
u
§ 0·6
o
u
o
~ 0·81--------- ----"·--------------------+-----------
:?
o
FIGURE 5.38 Approxímate solutíons for the rate of settlement of a single pile (Poulos and Davis, 1968).
of time-dependen! setdement. At higher load levels, how- Theoretícal analyses of a piJe in a viscoelastJc soil have
ever, significan! time-dependen! settlements occur, primar- been described by Booker and Poulos {1976a; 1976b). In
ily as a result of the effects of shear creep. the case of soil whose creep-response varíes linearly with
Figures 5.36 and 5.37 imply that in contras! to suéface log time, it has been shown that the logarithrnic creep rate
foundations, the consideration of the rate of settlement of e, (the slope of the settlement-versus-log-tirne relationship)
a pile is of relatively minor importance. Figures 5.36 and ís given by
5.37 also show the fallacy of calculating settlements of sin-
PlpB
gle piles by usíng one-dimensional consolidation theory. (5.45)
d
However, it should be em:)hasized that consolidation settle-
ment becomes more imp·Jrtant for piJe groups. A further where
implication in regard to pile-loading tests, is that suffídent
information on the settlernent behavior of a pile at normal P applied load
workng loads may be obtained from a constant-rat~:-of fp displacement-influence factor from elastic theory
penetration test, as ~uggested by Whitaker and Cooke (Eqs. 5 .33a or 5 .34a)
( 1966) (sec Chapter 16). B parameter in the creep function J(t) of the soil,
Although the rate of consolidation for piles is generally in which
not of great importance, some approximate solution~; for
J(t) A+Blog(I+a:t) (5.46)
the rate of consolidation settlement of an incompressible
piJe have neverthe!ess been obtained by Poulos and Davis
{1968) and are shown in Figs. 5.38 and 5.39. For a g,iven
diamcter, the rate of settlement decreases as L in creases. ·01 ·1 10
h ave finished (e .g., Muray ama and Shíba ta, 1960; Sharrnan, "'S\ 0·6
1961. Yamagata, 1963; Cambefort and Chadeisson, 1961; i3
u
Bromham and Styles, 1971). In severa] cases, it has be en o
observed that the settlement appears to increase linearly ~ oe
O'
with the logarithm of time, and consequently, sornE: in- o"'
vestigators have proposed empírica! equations relating
settlement and time (e.¡~., Cambefort and Chadeisson, FIGURE 5.39 Comparíson between rate of settlement of a piJe and
¡ 961). a surface footing (Poulos and DHis, 1968).
SETTLEMENT ANALYSIS OF SINGLE PILES 99
(5.48)
The base-load-versus-settlement relationship is there- The ultimate shaft and base resistances r1ay be cal-
fore culated from the methods described in Chapter :-1.
Consolidation Settlement
($.52)
To calcula te consolidation settlements for pEes in el ay,
it is convenient to assume that the consolidation process
It :;hould be noted that the effect of having a modulus is entirely elastic, even if sorne yielding or slip has occurred
at the pile tip, , greater than the average value along the under undrained conditions. A similar assumptiOn is made
shaft, , is taken into ::.ccount in determining the s·~ttle in Chapter 1O in relation to pile-raft systerns. Referring to
ment-influen:.:e factor I, and not by using the value c•f Eb Fíg. 5.41, the final consolidation settlement, pr:¡:.·, is
for in the denominator of (5.52). by
where
(.5.54)
working load on pile
2. The second extends from the latter point to tbe ul- displacement-ínfluence factor for the drained
timate bearing capacity of the pile, Pu Pru + Pbu, at Poísson's ratio, ~~~
which the set tlernent, Pu, is given as
Pu
Elastic 1mmedíat12
s<Ztti<Zmrznt.
V r
_Elastic total
tínal sctti<Zm<znt
Pu
1 1
1 1
f>cF <Zfas
~ Shaft compress1on r7
1 aft<Zr full slip
Load 1 1
--Load vs. ímmczdlatrz s«:~tlem¡znt
1
o Setti<Zm<Znt
(a) Frorn laboratory triaxial tests. test ( e.g., as described by Da vis and Poulos, 1963, for pad
(b) F ro m píle-loading tests. footings). At low load levels (about one third of the ulti-
(e) From empírica! correlations based on previous exper- mate), where little time-dependent settlements may occur,
ience. it m::.y be possible to use a constant-rate-of-penetration
(C .R.P.) test rather than a maintained loading test, and ob-
tain a single value of soil modulus. The interpretation of a
5.5.1 Laboratory Triaxial Tests pile-loading test to obtain the deformatíon parameters is
considered in detail in Chapter 16.
The use of such tests for settlement predictions of pad If the load test is carried to failure, the field value of
foundaLons has been described by Davis and Poulos pile-soil adhesion rnay also be determined, provided that
(1963), Lambe (1964), and Kerisel and Quatre (1966). In the strength parameters can be estimated indepeqdently.
all cases, the stress paths of typical elements in the field are Thus, a single pile-loading test may provide sufficient data
reproduced in the labo1·atory test and the resulting strains to enable both the deformation and the strength parameters
measured. However, for a pile foundation, such tests are to be estimated.
complicated by the difficulty of determining the appro-
priate stress path, both during installation of the pile and
resulting from the applied load on the pile. 5.5.3 Empirical Correlations
Conventional types of triaxial tests such as those used
for settlement prediction of shallow foundations (Da vis md In order to provide sorne information on values of Es for
Poulos, 1963; 1968) ha ve be en found to give val ues of Es situations in which pile-loading test data are not available,
that ary much too low for piles. For example, Bromham a number of published piJe-test results have been analyzed
and Styles (1971) obtained a value only about one third of and values of Es determined.
that backfigured from a field-loading test, whiie Mattes
(1972) obtained a value of only about one eighth of the 5.5.3. 1 PILES IN CLA Y
backfigured value from a field-loading test. Attempts Where possible, the calculated values of Es llave been
(Mattes, 1972) to simula te more closely the stress path of a correlated with reported values of the undrained cohesion,
soil element near a driven pile by failing the specimen in ex- eu, of the el ay, and these are plotted in Fig. 5 .42 . .Mean re-
tension, allowing consolidation and then testing for tht de- lationships between and Cu are plotted for 'Jored and
formation parameters, led to higher values of Es, but still driven píles for which two trends can be observed •
lower than the backfigured values from model-pile ksts.
The conclusion, then, is that it is not possible to measure l. For soft to medium days [cu < 17lb/sq in. (120 kPa)],
accurately the value of Es appropriate toa pile in a conven- Es for driven piles is greater than for bored piles, but for
tional triaxial test, even with relatively refined testing pro- stronger clays, Es for bored piles becomes greater. The first
cedures. At the present time, pcssibilities of other, mo're effect may be attributed to the higher excess pore pressures
appropriate forms of laboratóry testing remain to be and greater subsequent reconsolidation for the driven piles
explored. in soft clay, while the second may be attributed to the ef-
fects of "whip" in the driving of piles in stiff clay.
2. For stiff clays, Es appears to reach a limiting value--
5.5.2 Pile-Loading Tests about 6000 Ib/sq in. (40 MPa) for driven piles, and 12,000
lb/sq in. (80 MPa) for bored píles-although sorne tests on
Because many uncertainties may be associated with small- bored píles in London clay gave considerably higher values.
scale JalJoratory tests, it is desirable where possible to de-
duce the deformation and strength parameters from a full- In many of the tests, it is not possible to determine
scale pile-loading test. Such factors as the method of in- whether the value of is the undrained or the drained val-
stallation of the pile ancllayering of the soil profile are lhen ue. It is possibly reasonable to consider the values of Es in
largely taken into account. In order to determine all three Fig. 5.42 as drained values, E~, and in the absence of other
required deformation parameters from full-scale pile tests, informaEc'.", the undrained modulus Eu may be estimated
it is theoretically necessary to carry out loading testo on from the following relatíonship for an ideal isotropic elas-
two piles of different proportions and to use the appro- tic two-phase soil:
priate theoretical solutions to backfigure these parameters.
However, it is probably sufficient either to estímate the 3E'
E = S (5.59)
value of v~ or to determine it from a laboratory triaxial u 2(1 + v~)
SETTLEMENT ANALYSIS OFSINGLE PILES 103
o
TABLE 5.3 SUGGESTEDAVERAGE
50000 Average lar VALUES OF Es FOR DRIVEN PILES IN SAND
driven piles
.r ----¡;----
Range Qf
~:.. L Sand Density
Range of Relative
Density,D, Es
E, (kNím') t/ o (lbs/sq in.) (MN/rn 1 )
1
1
1 Loo se < 0.4 4000-8000 27.5-55
1 Medium 0.4-0.6 8000-10,000 55 -70
10000 t. Dense > 0.6 10,000-16,000 70 -llO
- t:..l
1
TABLE 5.4 AVERAGE VALUES surface cond,tions are given in Fig. 5.43. The site involved
OF K FOR SOLID PILES about 120 ft of naturallevee and back-swamp deposits con-
sisting of layers and laminations of clays, silts, and fine
Soil Pite Material sands, which overlay a 70-ft deep !ayer of fine silt grading
Type
to sandy grave! with depth. Two pairs of step-taper piles
Stt·el° Concrete Tim ber and one pair of steel-tube piles were driven: step-taper piles
lB and 3 were friction piles driven to within 20 ft of the
Soft clay 60,000 6000 3000 sandy stratum, step-taper piles 2A and 4 were end-bearing
Medium clay 20,000 2000 1000
Stiff day 3000
in the ~andy stratum, tube pile 9 was a floating piJe
300 150
Loose sand 15,000 1500 750 founded at a similar depth to lB and 3, and tube pile 10
Dense sand 5000 500 250 was end-bearing in the sand. Piles lB and 3 gave very.similar
load-test results, and were analyzed as floating piles in a fi-
aFor hollow or H-piles, multiply these nite !ayer to derive a backfigured soil modulus for the back-
v:dues by area n1tio RA. swamp deposits. The relevant details of these piles are as
follows:
3 18 2A 4 9 10
natural
lllVIlll (Cuo 5·5 p.s ¡)
1"--- ~- --
r--:,__ -~
----------~'
Stczp 1-- Tapczr- ~Tubcz
50 backswamp
dczpostts (Cu• 6-13 ps 1)
90
120
denscz sand
LczncJth 96 ', 1üt¡'" 126' 134' 100' 120'. N o 32- 154
Top dtc1m. 171,;4 17 '.-4 17~· 17Y~ 12 3,.-;¡" 12 3,.-;¡
Base ' 9y4· 10 3,.-;¡. 10 3/4" 10~4· 12 3/4. 12 3/¿' 150
Pile Number 2A 4 9 10
o ·4 r - - - - - r - - - - - , 05
o
o
o
0
0 t:r"'---..:-'-~--o-=--2 0·1 0·2 o~~----0~·725=---~o~·5~
0·1 0·2
5ettlement (¡nJ Settlement ( 1n ) Sattxzment (in l Sattlemcznt (in.l
Shoft Load vs. Settlemant Basa Lood vs Settlemcznt Shatt Load vs. Settlament Base Lood vs. Setti<lll'l<2nt
(a) P1la D
105
106 SETTLEMENT ANALYSIS OF SINGLE PILES
200,-----.,----· pile tests lB and 3, and it can be seen that the settlement
performance of both the floating and end-bearing piles has
,,
e: been closely predicted.
{?.
PiJe number 2 4 5 6
z PiJe type 14-in. 21-in. 17 -in. 17-in. 19-ín.
I' H-beam pipe pipe pipe pipe
Pile length (ft) 81 65 66 45 65
L/d 70 37 47 32 41
K 470 250 350 350 350
End-bearing (E.B.) E.B. E.B. E.B. F E.B.
or floating (F)
E biEs 3 3 3 1 3
Monsur & Koufmon Monsur & Kaufman Load (tons) 125 125 125 75 125
(1956) (1956)
Observed (in.) 0.13 0.13 0.16 0.10
(a) (b) settlement 0.13
Predicted (in.) 0.144 0.130 0.142 (0.100) 0.150
settlement
Predícted/ 1.10 1.00 0.89 (LOO) 1.15
t o observed settlement
T
o
NOTE: Soil type: alternating strata of silts, silty sands, sandy
silts, with interspersed clay slrata.
Bearing Stratum: Dense fine sand.
PiJe S (floating) used as control pile for predictions.
Mansur & Kaufmon Monsur & Kaufmon and measured load-distributions along the piles are com-
(1956) (1956) pared.
(e) o Mozaswr<Zd (d)
- Predrct<Zd
Figure S .48 and· Table S .6 show that quite low values of
K are possible when steel tubes or H-sections are used as
piles. In such cases, ít is likely that very little load does in
fact reach the pile base, even in nominally end-be~ring piles.
l In the case described he re, the results of a floating-pile test,
L
when combined with the results of a routine borehole test,
0·5 0·5 have allowed the accurate prediction of the load distribu-
tion along and settlement of end-bearing piles on the same
si te.
6
~
e
1·5
(Control Pilczs)
1·0
The~rcztical Curve
l
E" 1·0 0-B
_L_~---
_>!
~
o
o
"
VJ
u ·o
.;!u :;¡~ 0·5 o Jmrnczdiatcz Sczttlczmcznt, Su
0·5 Brass PIIQS
e:. F1nal Sczttlczmcznt, SrF
-6 "'
"'
L o0.
o.__2
40 o4
o L-------~--------~
10 40
TABLE 5.7 TESTS BY
D'APPOLONIA AND ROMUALDI (1963)
FIGURE 5.50 Comparison between observed and predicted ratio
P;/PTF (Mattes and Poulos, 1971).
Pilc number 1 2
Pile t)rpe 14 BP 89 14 BP 119
(H -pile) (H-pile)
Length (ft) 44 45 Model Tests
Assumed L/d 33 34 A series of carefully controlled model tests on piles in nor-
Assumed area 0.143 0.186
mally consolidated clay has been carried out by Mattes and
ratio (RA)
K 430 560 Poulos (1971) in order to examine the effects on settlement
E biEs of length-to-diameter ratio and pile compressibility, and the
Load (tons) 75 100 relative proportions of immediate and final settlement. Fig-
Obscrved (in.) 0.07 0.11 ure 5.49 shows the ratio of predicted to observed settle-
settlement ments for piles of various L/d, using brass piles having L/d
Predicted (in.) 0.09
0.06 = 25 as control piles and backfiguring the und~ained and
setl:lement
Predicted/observed settlemwt 0.86 0.82 drained Young's moduliESI1 andE~,from themeasured set-
tlements of these piles. The agreement between predicted
and measured settlements for L/d = 1O ancl 40 is reason-
NOTE: Soil: layers of fill, sandy silt, sand and grave!,
fine to medium sand, sand and grave!, ably good, indicating that the theory predicts with ade-
and sandy silt. quate accuracy the effects of pile length on settlement.
Bearing stratum: shale. Also shown are comparisons for plastic piles, of about ten
times the compressibility of the brass piles,, having L/d =
40. Again, the good agreement indicates that the theory
The bearing stratum was assumed to be rigid. In Table 5.7, gives a good prediction of the effects of pile <_:ompressi-
the piJe properties and settlement details are listed, and bility.
comparisons based on the assumed soil properties are made. Figure 5.50 shows a comparison between measured and
In Figs. 5.48g and 5.48h, the load distributions within the predicted ratios of immediate to final scttlernent :)f the m o-
piles are compared with the calculated distributiom.. In del piles, and reveals fair agreement. The test results con-
each case, quite reasonable agreement between predic tion firm the conclusion reached from the theory that the major
and observation is obtained, although better agreernent part of the settlement of a pile occurs as immediate settle-
could possibly have been obtained if the soil had been con- ment and that consolidation settlement is relatively unim-
sidered to have a modulus that increased with depth. portant at normal working loads.
SETTLEMENT OF PI LE GROUPS
where the piJe and the adjacent piJe carry the same load.
Solutíons for a: as a function of severa! variables are de-
S
scribed below. The use of interactíon factors to analyze
the settlement of general piJe groups is described in Sec-
p
tion 6.2.3.
1·0r-----,-----~----~------------~------~------~------~-------,
2 3 4 5
0·2 0·15 0·05 o
L
FIGURE 6.2 Interaction factor.1 for floating piles, d = 1O.
0·15 005 o
L
FIGURE 6.3 Interaction factors for floating piles,-= 25.
d
112 SETTLEMENT OF PILE GROUPS
L
50
d
0.05 o
4 5
0·2 0·15 0·05 o
L
FIGURE 6.5 lnteraction factors for floating piles,- 100.
d
SETTLEMENT OF PILE GROUPS 113
1.6 r---r--.---,--.,.--,------,~----....,
1 50
d
1.5
ex Ndb • o.,
(6.7)
where <> íntern factor for requíred
where
"' = íntern. factor for relative spacing a~ for ¿.d ~ 1
ao.s = interaction factor for Vs = 0.5
(6.8)
ofSlip
As pointed out in Chapter 5, slip at the interface between where
the píle shaf1 and the soil plays a negligíble role in the
set tle::ncnt behavíor of single pi! es of normal proportions FE is a factor depending on K, L/d, and Eh/E~.
at the working load, but for piers having low values of
L/d ami for pilcs with enlarged bases, slip has sorne signif- Values of for four values of L/d are shown ir. Fig. 6.14,
10
08
0·6.
aE
o 4
' '
0·2
o
o 2 4 5
0·2 0·15 0·05 o
1·0 1 !
i ...................
!
0·8 1 ~/, - 25
i Vs = 0·5
\ :
~\:.VaiU<25
............................
of :
"-, 1
"
r-:-,:_- ....
·~-
0·4
e··"--""'~~
_5(~
~
··--
--,~~~
--
----¡--
-~ ~~
0·2 ¡ - - .::u~
r- ..-.::'-
-----==~
-..___
---
----- ,_ -·~---
r·· ..
.. ---
··-
1- ::1~~ 1
t=
1
2 3 4 5
0·2 0·15 0·05 o
L
FIGURE..;:~ 1 lnteraction factors for end-bearing piles,- = 25.
d
which ~hows the transition from end-bearing to floating is reasonable io calculate the increase in settlement of pile
interact..on with /Es and K. The sma\ler the val u e of K i caused by pi! e j, L::,.pij, approximately as
or the larger the value of L/d, the srnaller the value of
Eb/Es fJr which tends to 1 (i.e., the interaction factor (6.9)
tends to tbe rigid base value ). Although the curves shown
are for s/d 5, they apply approximately for other values where
of s/d.
P¡ settlement of pile j under its own load
ex¡¡ interaction factor corresponding to the spacing
6.2.2.3 11\fTt'RACTJON BETWEEN PI LES OF DJFFERENT SIZE
between piles i and j, and for the geometrical
For two piles-í and j in 6.15- of different size, it pararneters (í.e., length and diameter) of pilej
S 0.1 o
,:¡ Q
L .
FIGURE 6.12 lnteraction factors for end-bearing piles,-; =50.
a
116 SETTLEMENT OF PILE GROUPS
1·0
·----+----~---t---
1
0·8 ,---
... . ___ L/d = 100 -------r-----"'1
"'· VaiUC2S of K 1
1
,
1
1
V5 = 0·5
1 ' 1
1
1 1 . 1 1
0·6 --10----+-----+----------r--- -----+
...__ 1 100 1
1
(lE 200 1
0·4
1
-- L- ------+----- T
1--------t
1
o
o 2 3 4 5
0·2 0·15 0·1 0·05 o
L
FIGURE 6.13 lnteraclion factors for end-bearing piles,- = 1 OO.
d
Simila rly, the increas.e :.n settlement of piJe j caus·~d by between piles i and j, and for the geometrical
pile i, 6.p¡¡, is parametrrs of piJe i.
.8 ¡-----¡-------j---j---f----i----- -f---- 1
Hl w
LL
3
ILW 2 .6r-r--~~-T-
0-1:3 e
L "25
-~-~=
D
L ·;:; d
u
.8u :J
'O v, -- 0.5
Cl ~
lL
e
e 0-G
D
_
·¡:; .4
u
º
+'
u
~
ee
l~+-
:J
D
"'
0:: 0'4
e
_1.._ =10
º
+' d '
--r---i-vs=
1
u i 1
Cl
L
$l o-s----
.S
o ~--1_---~----~--~
1 10 100 1000 =-- 10 100 1000
Eb¡
/Es
FIGURE 6.l4a lnleraction reduction factor FE. FIGURE 6.l4b Interaction reduction factor FE.
SETTLEMENT OF PILE GROUPS 117
Pilcz l Pilcz j
Ef ~
f---'~
FIGURE 6.15 Interaction between two piles of different size.
(6.11)
FIGURJ: 6.14c Interaction teduction factor FE.
where
p ti displacement of single pile j under unit load 1. In terms of the settlement ratio Rs, where
¡yki interaction factor for spacing between piles
k and j, and for the geometrical parameters of Average group settlement
of pile j ( 6.15)
Settlement of single pile at same average
load as a piJe in the group
Equ.ations (6.12) or (6.13) may be written foral! píles in
the group, giving n displacement equarions. Also, for 2. In terms of the group reduction factor Re, where
vertical load equilibrium,
Average group settlement
1/ Re
Settlement of single pile at same total ( 6.16)
= L P¡ (6.14) load as the group
j= 1
Pe = RsPavPl (6.18)
l 2
1
¡
3
-$1·
or 1
j5·
.Lh4 5 6 l
't'
Pe ( 6.19)
where 300 T
J
Pav average load on apile in the group
Pe total group load
1t ::hould be emphasized that the above analysis does 1
not diiectly consider the influence of softer compressible
layers which may lie beneath the piles unless the interaction 125'
~ ~too ~-
factors and single pile settlement are computed to allow for
Madi u m
such layers. A simple extension of the analysis to cover this Cloy 1
problern is described in Section 6.4. ~~J
To illustrate the use of the superposition principie
for calculating the settlement of a pile ¡;roup, an example
is described below. Solutions for square groups of piles are FIGURE 6.16 PiJe group configuration in example.
described in Section 6.3.
The settlement of pile l (and all type A pílesl is given
by Eq. (6.12) as
Rlustrative Examp/e
A freestanding group of si.x 12-in-diameter concrete piles PA = p¡{PA (0.27+0.42+0.25)+ P8 (0.42+0.35)+rA}
is driven into a deep !ayer of medium el ay, and ís to be
subjected to a load of 300 tons (see Fig. 6.16). A test on or,
a single pile gives a final settlement of 0.60 in. under a
load of 50 tons. Determine the final settlement of the
six:-pile group. !!..tJ.. = 1.94PA + 0.77 P8 (6.20)
P1
From Table 5.4, the value of K is about 2000. Piles
1, 3, 4, and 6 behave identically and will be called type where p 1 is the settlement of a single pite under unit load.
A, whik piles 2 and 5 w!ll be called type B. The loads on Similarly, for pile 2 (and all type B piles),
typcs A and B are PA and Ps, respectively. From Fig.
6.4 for L/d = 25, the interaction factors may be inter- PB = P1 {PA (0.42+0.42+0.35+0.35)+
polated for K 2000. The factors are tabulated in Table P8 (0.42)+Ps}
6.1.
or,
TABLE Ll
P.4 57.4 tons be interpolated from Tables 6.2 and 6.3. For groups con-
PB 35.2 tons taining more than 16 piles, it has been found that Rs varíes
PB approximately linearly with the square root of the number
134.4
P1 P1 of piles in the group. Thus, for a givef1 value of pile spacing,
K and L/d, Rs may be extrapolated from the values for a
From the piJe-load test, 16-pile group anda 25-pile group as follows:
0.60 (6.23)
O¡
50
O.O 1 in ./ton
where
Therefore
value of Rs fot 25-pile group
value of Rs for 16-pile group
PA = PB 1.66 ín.
number of piles in group
Settle:11ent rati<,
Figure 6.17 shows the group reduction factor, Ro,
plotted against s/d for vanous groups. Ro, and hence group
settlement, decreases as the number of piles in creases. How-
ever, at relatively close spacings, the use of more piles to
Group-reduction factor decrease settlement becomes increasingly ineffective if the
same spacing between the piles is retained.
In general, it is found that the settlement of a group of
piles in a relatively uniform stratum depends primaríly on
lf the piJe cap ís sufficiently flexible and the load is
the breadth or width of the grollp; hence, within a group of
uniformly distributed so that all pi les are equally loaded,
· then given breadth or width, increasing the number of piles
beyond a certain number will only marginally improve the
PA PB - 50 tons settlement performance of the group, unless the original
PA 50 X 2. 7 1 X P1 1.62 in. spacing within the group is greater than abom six día-
PB 50 X 296 X P1 1.77 in.
No ofPíles
in Group 4 9 16 25
2 1.83 2.25 2.54 2.62 2.78 3.80 4.42 .4.48 3.76 5.49 6.40 6.53 4.75 7.20 8.48 8.68
10 5 1.40 1.73 1.88 1.90 1.83 2.49 2.82 2.85 2.26 3.25 3.74 3.82 2.68 3.98 4.70 4.75
10 1.21 1.39 1.48 1.50 1.42 1.76 1.97 1.99 1.63 2.14 2.46 2.46 1.85 2.53 2.95 2.95
2 1.99 2.14 2.65 2.87 3.01 3.64 4.84 5.29 4.22 5.38 7.44 8.10 5.40 7.25 9.28 11.25
25 5 1.47 1.74 2.09 2.19 1.98 2.61 3.48 3. 74 2.46 3.54 4.96 5.34 2.95 4.48 6.50 7.03
10 1.25 1.46 1.74 1.78 1.49 1.95 2.57 2. 73 1.74 2.46 3.42 3.63 1.98 2.98 4.28 4.50
2 2.43 2.31 2.56 3.01 3.91 3.79 4.52 5.66 5.58 5.65 7.05 8.94 7.26 7.65 9.91 12.66
50 5 1.73 1.81 2.10 2.44 2.46 2.75 3.51 4.29 3.16 3.72 5.11 6.37 3.88 4.74 6.64 8.67
10 1.38 1.50 1.78 2.04 1.74 2.04 2.72 3.29 2.08 2.59 3. 73 4.65 2.49 3.16 4.76 6.04
2 2.56 2.31 2.26 3.16 4.43 4.05 4.11 6.15 6.42 6.14 6.50 9.92 8.48 8.40 10.25 14.35
100 5 1.88 1.88 2.01 2.64 2.80 2.94 3.38 4.87 3.74 4.05 4.98 7.54 4.68 5.18 6.75 10.55
JO 1.47 1.56 1.76 2.28 1.95 2.17 2.73 3.93 2.45 2.80 3.81 5.82 2.95 3.48 5.00 7.88
..,
.....
.....
N
"'
TABLE 6.3 THEORETICAL VALUES OF SETTLEMENT RATIO Rs: END-flEARlNG PlLE GROUPS, WITH RlGlD CAP, BEARlNG ON A RlGlD STRATUM
No. of Piles
in Group 4 9 16 25
L/d s/d K .10 100 1000 "" 10 100 1000 "" 10 100 1000 "" 10 100 1000
~
~ !.52 1.14 1.00 LOO ? 02 L31 1.00 1.00 2.38 1.49 LOO 1.00 2. 70 1.63 1.00 1.00
10 5 1.15 1.08 1.00 1.00 1.23 1J2 1.02 LOO 1.30 1.14 1.02 1.00 1.33 1.15 1.0) l. UU
10 1.02 1.01 1.00 1.00 1.04 1.02 LOO 1.00 1.04 1.02 1.00 1.00 1.03 1.02 LOO 1.00
2 1.88 1.62 1.05 1.00 2.84 2.57 1.16 1.00 3.70 3.28 1.33 1.00 4.48 4.13 1.50 LOO
25 s· 1.36 1.36 1.08 1.00 1.67 1.70 1.16 1.00 1.94 2.00 1.23 1.00 2.15 2.23 1.28 1.00
o 1.14 1.15 1.04 1.00 1.23 1.26 1.06 LOO 1.30 1.33 1.07 1.00 1.33 1.38 1.08 1.00
2 2.49 2.24 Í.59 1.00 4.06 3.59 1.96 1.00 5.83 5.27 2.63 LOO 7.62 7.06 3.41 1.00
50 5 1. 78 1. 73 1.32 1.00 2.56 2.56 1.72 1.00 3.28 3.38 2.16 1.00 4.04 4.23 2.63 1.00
10 1.39 1.43 1.21 1.00 1.78 1.87 1.46 1.00 2.20 2.29 1. 71 1.00 2.62 2.71 1.97 1.00
2 2.54 2.26 1.81 1.00 4.40 3.95 3.04 1.00 6.24 5.89 4.61 1.00 8.18 7.93 6.40 1.00
100 5 1.85 1.84 1.67 1.00 2.71 2.77 2.52 1.00 3.54 3.74 3.47 1.00 4.33 4.68 4.45 1.00
10 1.44 1.44 1.46 1.00 1.84 1.99 1.98 1.00 2.21 2.48 2.53 1.00 2.53 2.98 3.10 LOO
SETTLEMENT OF PILE GROUPS 123
:~: rt.---r-t----r-t---,-~--=-~
~-----;--J.
coincident with the curve for the group, can be used
,--1
-¡----+-----+-----
1 1 over a practica! range of group breadths. The dependence of
settlement on group width rather than number of piles has
been confirmed in full-scale tests by Berezantzev et aL
Design curve for !l 1 {1961), and from data collected by Skempton (1953) for
driven pi le groups L
in sand 1 v, 0.5 driven piles in sand. The relationship suggested by Skemp·
{1.7 (Skempton, 1953)-- K oo --t----+-----1
o ton (Eq. 6.1) between settlement ratio and group breadth,
a:
as reexpressed in terms of Re and plotted in Fig. 6.18,
B (1.6
agrees quite well with the theoretical curves. These results
'-'
!!
e
o
suggest, therefore, that if settlement is the sole criterion, it
"ti 0.5 is more economical to use a smaller number of piles at a
::J
'O
~ relatively large spacing, rather than a large number of píles
c. 0.4
:J
-~··-···
at el o ser spacings.
e
(.:;¡
0.3
0.2 -··--···~
6.3.1.1 EFFECTOF FIN/TE LA YER DEPTH
(6.24)
meters. This point is illustrated in Figs. 6.18 and 6.19, in
which the group reduction factor, Re, is plotted against to-
tal group breadth. These figures show that for larger groups, Thus, ~~~ is a factor by which the values of Rs for an infi-
Re does not vary greatly w ith the number of piles in the nitely deep layer in Table 6.2 are multiplied tJ obtain Rs
group. For groups containing more than 25 piles, it appears for a finite !ayer. Figure 6.20 shows that as would be ex-
that a common limiting curve of Re versus group bwadth, pected, ~~~ de creases as h/L decreases, and that the effect of
1 o
--- ---- - -
: 1
~ = 25
1
1~------
08 ~------
1/, 05 ,__ -
K =100
---· : ------~
~---~-
06 \ . -- ¡-
-
----~- --------~
\ ---+~-
~ 1"---.
:
i
04
"'~;
~~- ' 22
i
~~------ : 32 1
¡
02
1 4"52
1
o '·
o 5 10 15 20 25 30 35 40
8rvadth of Group J d
32
~h v, 0.5 Rs for group resting on bearing stratum
(6'.25)
S
42 d ce 2.5 Rs for floating group on infinitely deep
0.4
52 !ayer
0.2
l. ~b, and hence settlement ratio Rs, de creases as the rela-
i
1 tive stíffness of the bearing stratum, E biEs, increases, this
effect being most pronounced for shorter, stiffer piles.
2. For slender piles ( e.g., L/d = 100), unless the piles are
o
1 2 quite stiff, the bearing straturn has little effect on settle-
h
0.5 (1 ments, because ve1y little load reaches the pile tip under
L
1c normal working-load conditions.
h
FIGURE 6.20 Rcduction coefficient th for effect of finite !ayer
100
_ --1oo-
_.....,~--+-
---- 25
0.6
3 2 group
4
o~------~-----------L------~
10 100 1000
_!_:o<.!
d
{al/ groops)
0·6
IJ..
-o:
1-
a.
0·7
0·5~----~----~-----~----~------~----~----~-----J
o 5 10 15 20 25 30 35 40
-t
FIGURE 6.24 Relative importance of immediate settlement-pile groups with rigid cap.
126 SETTLEMENT OF PILE GROUPS
dation settlement theoretically comprises less than 10% less uniform as the pile spacing decreases, the number of
of the total final movement. piles increases, L/d íncreases, or K increases. The load dis-
tribution is also influenced by the existence of a !ayer of
finíte depth, a typical example for a 3 2 group being shown
6.3.2 Load Distribution in Groups with Rigid Cap in Fig. 6.26.
For piJe groups bearing on a rígid stratum, load
F or a wide range of values of Lfd, sfd, K, and group size, distributions are gíven in Tables 6.5a and b, while typícal
load distributions within a floating-pile group are given in dístributíons for a 3 2 group are shown in Fíg. 6.27. As
Tables 6.4a and b, the pile load being expressed as a frac- wíth the floating groups, the load dístribution generally
tion of the average load in the group. The key for identi- becomes \ less uniform as L/d in creases and spacíng
fication of the piles in each group is shown in Fig. 6.25. decreases, but in contrast to the -~orresponding floating
The greatest loads occur at the comer piles, and the least groups, íncreasing the piJe stíffness factor K causes the
at the center píles. The load distribution tends to become load dístribution to become more uniform.
s/d
3' group JO 2 1.28. 1.47 1.56 0.84 0.75 0.72 0.52 0.16 -0.15
5 1.20 1.25 1.26 0.91 0.88 0.88 0.57 0.47 0.45
10 1.10 1.13 L14 0.95 0.94 0.94 0.78 0.73 0.70
20 1.04 1.05 1.06 0.98 0.97 0.97 0.91 0.88 0.88
100 2 1.24 1.11 1.70 0.86 0.93 0.66 0.58 0.84 -0.45
5 1.22 1.17 1.37 0.90 0.92 0.81 0.53 0.61 0.24
10 1.14 1.15 1.28 0.94 0.93 0.86 0.70 0.68 0.42
20 1.07 1.10 1.21 0.97 0.95 0.90 0.86 O. 79 0.55
4' group 10 2 1.68 2.00 2.14 0.97 0.95 0.95 0.38 0.09 -0.04·
5 1.42 1.51 1.52 1.01 LOO 1.00 0.56 0.48 0.47
10 1.21 1.25 1.28 . 1.01 1.00 1.00 0.77 0.73 0.70
20 1.10 1.13 1.12 1.00 LOO LOO 0.89 0.86 0.86
100 2 1.56 1.35 2.30 0.96 0.97 1.01 0.52 0.70 . -0.15
5 1.50 1.45 1.84 1.02 LO 1 0.98 0.47 0.52 . 0.18
10 1.29 1.35 1.65 1.00 1.00 1.00 0.70 0.63 0.34
20 LIS 1.24 1.42 1.00 1.01 1.00 0.83 0.75 0.56
TABLE 6.4b LOAD DISTRIBllTIONS WITHIN 5' FLOA TING-PILE GROUPS VAL UES OF P/Pm,
K 100 1000 ~ lOO 1000 "" 100 1000 .. 100 1000 "" 100 1000 "" lOO 1000 "'
L/d -
s/d
5' !O 2 ?.1? ?.48 2 ()_<; 1.18 1.19 l.l2 1.05 1.07 1.20 0.42 0.21 -0.15 0.26 0.10 0.16 0.12 0.01 0.45
Group 5 1.64 1.75 1.79 1.13 1.14 1.15 1.07 1.09 1.10 0.60 0.53 0.49 0.55 0.4'! 0.46 0.49 0.45 0.42
10 1.31 1.39 1.41 1.07 1.09 1.09 1.03 1.04 1.05 0.81 0.75 0.74 0.76 0.71 0.69 0.73 0.68 0.64
20 1.18 1.22 1.19 1.05 1.06 1.05 1.01 1.02 1.02 0.90 0.87 0.89 0.86 0.83 0.85 0.82 0.79 0.82
25 2 1.90 2.46 2.90 1.17 1.19 1.13 1.01 1.09 1.20 0.58 0.20 -0.20 0.38 0.12 0.09 0.24 0.04 0.25
5 1.62 1.98 2.11 1.14 1.18 1.19 1.05 1.09 1.07 0.63 0.40 0.35 0.54 0.34 0.27 0.46 0.29 0.22
JO 1.39 1.63 1.73 1.10 1.14 1.16 1.04 1.05 1.07 0.77 0.64 0.56 0.70 0.55 0.47 0.64 0.45 0.37
20 1.22 1.37 1.40 1.06 1.09 1.09 1.02 1.04 1.05 0.87 0.77 0.74 0.82 0.72 0.71 0.78 0.68 0.67
-------·-·
100 2 2.06 1.75 3.00 1.15 1.14 l.! O 1.08 LOO 1.20 0.41 0.65 -0.30 0.33 0.48 0.05 0.25 0.33 0.40
5 1.77 1.78 2.34 1.18 1.18 1.22 1.07 1.06 1.09 0.54 0.55 0.21 0.48 0.42 0.14 0.33 0.30 0.07
10 1.45 1.58 2.05 1.10 1.13 1.21 1.05 1.04 1.08 0.72 0.66 0.38 0.68 0.58 0.26 0.63 0.49 O.l 7
20 1.25 1.41 1.78 1.17 1.10 l.l3 1.02 1.02 1.02 0.85 0.77 0.55 0.80 0.70 0.50 0.75 0.62 0.52
....
"'_,
128 SETTLEMENT OF PJLE GROUPS
12 r-----,-----r-----r---.....,
¡-s-.¡ To
1 1o 5
o 01
Píl<2 3
1 2
• • •
2 3 2
1
o o 01 •1 •2 •1
S
_Lo o· 03 o2 1 1 • • •
Lo 03 02
o o 03 o2
o o
o t--4~;;;.:::;=:::::::::~~;;;.;=-..:.::~
01
~f-d
1
o o o o1 p
Pav
3 2 Group
~~d 09
42 Group
il =25
rs
o o o 01 08
v,=o 5
K~100
K=1000----
o o o 04 02
01 oL-------~s--------,~o------~15~------2~o
o o :::>6 o5 o3 S
d
o o o o4 o2 FIGURE 6.27 Load distributions in 3' end-bearing group.
3r----------r--------~r---------.---------~
2·0.
u
o
o
_,j
~
ñ:
"'"'o
t..
1·5
m!"
·12·-+>-~
.2 .3 .2
tr--
·t,-·~
4d
"'
>
<l:
Incornpressíbl<2 Píl<2s
"-.....
u
e Ljd = 25
o
_,j
0·5 V5 =0·5
S!
il: K = 1000
o1
10 100 1000
40
Ety¡
Es
FIGURE 6.26 lnfluence lay•:r depth on load distribution-3' group FIGURE 6.28 lnlluence of bearing stratum on load distribution in
with rigid cap. gro u p.
SETTLEMENT OF PILE GROUPS 129
TABLE 6.5a LOAD DISTRIBUTIONS WITHIN 3' AND 4' END-BEARING PILE GROUPS
VALUES OF PfPav
corresponding group with a rigid cap, PR, is given in Table For typical groups of end-bearing piles on e rigid base,
6.6. The ratio PmaxiPR increases with increasing number values of PdiPmax are shown in Fig. 6.30 fo: K = 100,
of piles in the group, but ís almost independent of spacing For such compressible piles, relatively large dífferential
for a practica! range of spacings. The value of K has little settlements may occur, expecially for large groups of
effec on PmaxiPR. slender piles. However, the relative differential settlement
For groups of end-bearing piles on a rigid base, values decreases rapidly with increasing K and is ze·o for piles
of PmaxiPR are shown in Table 6.7, for L/d = 25 and K== that can be considered as incompressible.
lOO. For such groups, PmaxiPR decreases rapidly as K
increases and is unity for K > 2000, since no íntera.;tion
then occurs. 6.3.4 Approximation of Group as a Single Pier
The ratio PdfPmax. of the maximum differential s~ttle·
ment to the maximum settlement, is shown in Fig. 6.29 For ca!Culations relating to large structures supported by
for incompressible floating piles in a semi-infinite rnass. a number of pile groups, it may often be useful to replace
This ratio increases with increasing spacing but decreases each pile group by an equivalent single pier that settles
if the !ayer depth is decreased or L/d increased. The value an equal amount. Such an approximation is useful, for
of K has relatively little influence, example, if an analysis of intergroup interactio:J. is desired,
130 SETTLEMENT OF P!LE GROUPS
K lOO 1000 100 1000 100 1000 lOO 1000 100 1000 100 1000
L/d
s/d
~; 2 group 10 2 1.11 0.86 1.02 0.94 0.95 0.95 0.99 1.13 0.90 1.14 0.81 1.18
5 1.06 1.01 1.02 1.00 1.01 1.00 0.96 1.00 0.95 1.00 0.94 1.00
JO 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
20 1.00 1.00 1.00 1.00 1.00 LOO 1.00 1.00 1.00 1.00 1.00 1.00
25 2 1.55 0.99 l.J2 0.99 0.99 0.96 0.77 1.07 0.59 1.03 0.44 0.99
5 1.37 1.09 1.09 1.03 1.04 1.00 0.77 0.95 0.72 0.92 0.67 0.87
JO J.J5 1.03 1.04 1.01 1.02 1.01 0.90 0.98 0.88 0.97 0.86 0.97
20 1.02 1.00 1.00 1.00 1.00 1.00 0.98 1.00 0.98 1.00 0.98 1.00
100 2 2.08 1.48 1.20 l.11 l.OJ 0.96 0.48 0.83 0.25 0.63 0.36 0.44
5 1.75 1.55 1.16 1.13 1.08 1.04 0.53 0.68 0.46 0.58 0.39 0.47
10 1.42 1.39 1.08 1.09 1.04 1.04 0.73 0.77 0.70 0.71 0.67 0.65
20 1.21 1.22 1.05 1.05 J.OJ 1.01 0.88 0.87 0.83 0.82 0.79 0.78
~~
S
1.5 1.2 1.5 1.2 1.5 1.2
TABLE 6.7 VALUES OF PmaxiPR FOR or if settlements caused by underlying compressible strata
END-BEARING PlLE GROUPS. L/d = 25, are to be estimated (see Section 6.4). Two types of approxi-
K 100, vs 0.5. mation may be useful:
~ 3' 4
2
S'
l. An equivalent single pier of · the same circumscribed
plan area as the group and of sorne equivalent length,
----------"----------------
2 1.04 1.08 1.17
Le.
2. An equivalent single pier of the same length, L, as
5 1.19 1.14 1.2J the piles, but having an equivalent díameter, de.
10 1.09 1.1 o 1.13
20 1.03 1.02 1.02
4o 1.oo J.oo 1.oo For incompressible lloatíng groups, values of Le/L
for the first approxirnation, obtained by Poulos (l968b),
SETTLEMENT OF PILE GROUPS 131
0·3
h
d
0.6
...
e; S.
"
E
L
"B" 0·1
""'
h
0~--~--~---LL-~~~~~~--J--- L=oo
0 ID 00 ~ ~
5/d
FIGURE 6.29 Diffcrential ;e1tlerncnt in lloating pile groups with Ü L___ _ _ _ _ .L------"'--·---
equally loaded pites. 0 10 20
S
d
are shown in Fig. 6.31. L,:/L depends both on spacing and FIGURE 6.31 Equivalen! length of single pier for sarn~ settlernent
L/d, but is virtually independent of the numbcr of píles as píle group.
in the group. For most practica! cases, Le/L líes between
0.9 and 0.6.
e:
"E
-
S!
\ií
U1
0·2
l:í
¿
:.--
e:
"
E
"
~
U1
ó
X
CJ
¿
FIGURE 6.30 Differentíal 'ettlernent in end-bearing groups with equally loaded piles.
132 SETTLEMENT OF PILE GROUPS
3 2 group, driven
concrete piles
Layer 1
Medium el ay
,Cu "60 kPa
d 0.4 m
d.,
B
0·4
Layer 2
6m
Mediurnsoftclay, mv 0.15m 2 /MN
Layer 3
o Wml Medium clay, m., ~ 0.05 m2 /MN
o 5 10 15 20
~&7~~·
S
d
FIGURE 6.32 Diameter of equivalen! pier to represen! pile group. FIGURE 6.33 !Uustrative example.
SETTLEMENT OF PILE GROUPS 133
as indicated in Fig. 6.33. An estímate is requíred of the TABLE 6.8 CALCULATION OF ADDIT!ONAL
total final settlement of the pile group. SETTLEMENT CA USED BY UNDERL YING LA YERS
The first step is to estímate the drained Young's mod-
ulus and Poisson's ratio· of ea eh !ayer. For the first !ayer Layer lk+l b Ei!f< /k- ll<+l
in which th<c piles are situated, interpolation from Fíg. (k)
(MN/m') Eiíf<(m'fMN)
5.42 for c11 '" 60 kPa gives E~ = 17 .O MPa. v~ is assumed
to be 0.35. For the second and third layers, v~ is assumed
2 1.67 0.428 2.0 0.308 4.2 0.02 79
to be 0.35 and 0.3, respectively. Using the specified values
3 2.0 0.308 2.55 0.230 14.9 0.0052
of mv, and the following theoretical relationship between
H~andmv, l: = 0.03 .ll
a Le 18m.
b From Fig. 5.35.
r.' 1
(1- 2v~Xl + v~) :. Settlement caused by layers 2 and 3 4.5 X
18.0 0.)331
L,s ( 6. 27)
(1- v~) m¡. ~ .0083 m
8.3 mm
the values of E~ for laycrs 2 and 3 are 4.2 MPa and 14.9
MPa, respectively.
Assuming Young's modulus of the concrete to be the equivalent pier, Le/de = 18/5 3.6. The calculations
17,000 MPa, to evaluate the settlement of layers ! and 3, using Eq.
(5.41), are tabulated in Table 6.8. It is assumed that the
Pile-stiffness factor rock beneath !ayer 3 is rigid. From Tablc 6.8, the settlc-
ment caused by the underlying layers is 8.3 mm, so that
X 1.0 the estimated final settlement of the group is
K
17
p 8.9 + 8.3
1000
17.2 mm
The settlement of a single pile in the first !ayer may now be
calcuiated, the relevant dimensionless parameters beíng lf the equivalent-diameter approach is used, the settJe.·
L/d 20/0.4 = 50, K 1000, v~ 0.35, h/L 30/20 = ment caused by the underlying layers is calculated to be
· 1.5. Using Eq. 5 .33a, the single-pile influence factor 1 9.7 mm, compared with the above value of 8.3 mm.
is found to be 0.046, and for the average píle load of
4.5/9 0.5 MN, the single·pile settlement is
Piav Pav • P1 0.4 X 17.0 X 0.046 6.5 PREPARATION AND USE OF DESIGN CHARTS
.0034 m
The theoretical solutions presented in this chapter can be
3.4mm
used to prepare design charts to assíst in the s~lectíon of
The settlement ratio must now be determined. From the necessary number and spacing of piles to support a
Table 6.2, for a 3 2 group in a deep !ayer, with s/d 2/0.4 given load with a specified maxímum settlement and
5, K 1000, and Vs = 0.5, R~ = 3.51. Making allowance factor of safety against failure. The procedure is best
for the effect of the fíníte !ayer from Fig. 6.20 and for the described with reference to the following simple example.
effect of Vs being 0.35 rather than 0.5 (Fig. 6.22), the A load of 5 MN ís to be supported on a deep !ayer of
required value of Rs is cstimated to be 2.63. The settlement clay having the followíng average properties: Cu = 50
of th¡~ group in the founcling !ayer, Pon. is then given by kPa, pile-soil adhesion ca 45 kPa, E' = 15 MPa, ¡/
Eq. (IS.l8) as 0.3. It is proposed to use dríven 20-m-long concrete piles,
0.4 m in diameter, for which Ep = 15,000 MPa. It is re-
PGD = 2.63 X 3.4 quired to examine the combínations of number a:1d spacing
8.9mm
of piles that satisfy the criterion of a factor of safety of
2.5 agaínst faílure and a maximum final settlement of 15
Thc settlcment contril::ution from the underlying !ayers
must now be calculated. From Fig. 6.31, using the equiva- mm.
lent-length approach to represent the group; Le is about Consideríng first the settlement criterion, the results
0.9L, or, Le 0.9 X 20 '= 18 m. The plan area of the for Rs in Table 6.2, together with the solutions for single-
group is 4.4 .:< 4.4 = 19.36 m 2 , and hence the diam•:1er pile settlement (Chapter 5) may be used to prepare plots
of the equivalent pier is de 4.96 m-say, 5 m. Thus, for . of settlement, S, versus number of piles, n, f·lr various
134 SETTLEMENT OF PILE GROUPS
4.0 --'""T,-,-------,---..,------,-------,
'1
1 Curve 2-minimum
~) -- \ requirements for
FS •- 0.25
\
3.0 \
__,_ FS --2.5
U>
Values of specmg,
1.1.
s (rn) \
f:Jl 4.0
"'eu
\
\
o
2.0
/ a 2.0 - '\ Curve ·1-minimum
8
u " '\. '-... requirements for
"'
u..
0.. ....,_s=15mm
0.5 - '-.
.......
....... ........
0.2-
1.0
......._
----
o 11 OL_____L__ __ L_ _ _L __ __ L_ _~
2 5 10 20 50 100
o 10 20 30 40 50
No. of pi les, n
l~o. of pi les, n
FIGURE 6.35 Design chart for factor of safety of pile groups in
example. FIGURE 636 Composite design chart for piJe group example.
SETTLEMENT OF PILE GR::1UPS 135
ties. H is the simplicity and rapidity of construction of There is close correspondence between all three settlement
such charts-as compared with, say, a "one-off' finite-ele- profi!es, except in the ímmediate vicinity of the groups, in
ment solutíon-that make the elastic parametric solutions which case the e nsideration of a uniform pile-load distri-
so useful in designing pile groups. bution leads to nonuniform settlement of the ¡;roup. The
satísfactory nature of the equivalent-pier approximation is
evident from Fig. 6.37.
Further examples showing the effects of pile-stiffness
6.6 SURFACE SETTLEMENTS AROUND A GROUP factor K, length-to-diameter ratio L/d, and relative layer-
depth h/L, on the surface-displacement pro files surroundíng
The precedíng sections have dealt wíth the settlements di- a 2 2 group, are shown in Figs. 6.38, 6.39, and 6.4J.
rectly beneath the pi! e group. In sorne cases, ít m ay al so be
of interest to estímate the settlement of the ground surface
at some distance away from the group--for example, in
deterrnining the addítíonal settlement of an existing build- 6.7 OBSERVED ANO THEORETICAL GROUP
ing caused by a new structure. Such an estímate may be BEHAVIOR
obtained by usíng the solutions in Fígs. 5.32, 5.33, and
5.34 for the settlement distributíon around a pile. As with
the calculatíon of settlements caused by compressi'lle 6.7 .l Settlements
underlying strata, it ís convenient to consider the group
either as having a uniform load dístribution among the A number of comparisons between measured and theoreti-
piles, or as an equivalcnt single pier. An example of the sur- cal values of settlement ratio for t1oating-pile groups were
face-displacement profile caused by a 3 2 group with a tigid made by Poulos and Mattes (l97lb ). A summary of the
cap in a uniform semi-infinite mass, as given by Poulos ami cases consídered ís givcn in Table 6.9, and the ccmparisons
Mattes (197la), is shown in Fig. 6.37. Profiles were ob- are shown ín Fig. 6.41. In all cases, the load leve! corre-
tained by consídering: sponds to a factor of safety of at least 2 agaimt ultímate
failure of the group. With the exception of the model tests
l. The correct distributíon of pile loads P¡. by Hanna (1963) in loase sand, the agrecmcnt i:. generally
2. A uníform distríbution of loads satisfactory for both large and small val ues of K. The poor
3. A single equívalent picr. agreement for the tests in loose sand may be attributed to
32 Group
L
15 = 25
d
S
; 5
'\ d
Y, 0·5
Group wíth R•grd Cap
K 1000
Group wíth Flexible Cap
( Uní!orm load drstribution)
0·5
o 10 40
X
d
FIGURE 6.37 Comparison b<:tween correct and approximate surface·displacement profilcs.
136 SETTLEMENT OF P!LE GROUPS
35,----..,..----.,.-----.---.,-----, 25
3
22 Group 1
l
25 1
d
2·5
~.K =100
S
d = 5
lp
1 5
'\.\\ p • _!¡¡_
LE,
lp
v, =o 5
h
[
(X) \\
K= 1000 ' \
p= -lo 1 o \.\~·\
20,000
o5
\ '\ ".!!.:1'15
"
'\_L
,"
'-
' ...
' .......
..
_-- ------
o 10 20 30 40 50
2
3 Group -} = 5
K =1,000
vs = 0·5
h
-r="""'"
p
p = d~. lp
0·01
o 10 20 30 40
X
d
L
·FIGURE 6.39 lnfluenc:e of -on surface·displacement pro files.
d
7
2
V Sartary & Tatq(l961 3 Mod12l Group in 3 2 Modal Group in 3 2 Modal Grocp •n
o Whital<<ir (1957) Clay Oqnsq Sand
SowG'rs at n.t Hanna (1963)
(1961)
5
Rs Rs
\
~K~100
3
o
3 3
o
.........__
.........
o
2
3 Mod~l Group In
Cloy o
1 11
1 1
1 3 5 3 5 1 3 5 1 3 5
5/d s/d s/d s/d
(a) (b)
(a) (f)
7 - 4 15 15
2 2
3 Full S cal" Group 2 Full Se al" Grot.p 5 2 Mod~l Group in
2
5 Modal Group o
!O Dcmsq Sand .n OQMS<l Sand Dansq Sand
B<ZrazantzQ'v at al Bo2ro2zantzqv <2t al Hanna (1963)
( 1961) (1961)
5 3 10 10
~
o \
Rs R, l~s
3 5 5
1.2 ---'-- 4 12
o o
6 4 6 1 3 5 1 3 5
5/d 5/d s/d
~e) (d)
(g) (h)
Assumed Parameters
.. Test
Pi le
Material
Soil Type
for Comparisons
Remarks
L/d K Layer Depth/L
137
138 SETTLEMENT OF PILE GROUPS
_.._ ______
. .---_ 4 ~
<;
VJ
Block
•
foilurcz
h/L=1·5 L/d=25
5 x 5 Groups
3x3 Groups
2 o Stifl cloy, h/L = 1 8 L/d = 20
D Stiff cloy h/L =1·8 L/d = 20 • Soft cloy h/L =2·4 L/d = 20
• Soft cloy h/L = 2·4 L/d = 20
(al (b)
o -L----~----~----~----1----_J QL---~----~-----L----~----~----J
1·0 1·5 2·0 2·5 3·0 3·5 4·0 1·0 15 2·0 2·5 3·0 40
Pil« Spacing s/d Pilcz Spacing s/d
FIGURE 6.42 Theorctical and measured settlernent-ratios. (Barden and Monckton, 1970).
Further cvidence of the applicability of the theor.e tical rock, comparisons between measured and theoretical settle-
approach can be found in the cornparisons between theory ments ha ve been made by Poulos (1972b). A typical soil
and the data collected by Skernpton, 1953 (Fig. 6.18) and profile is shown in Fig. 6.43. The calculated values were
Barden and Monckton, 1970 (Fig. 6.42). With the eJCcep- based on a val ue of modulus backfigured from the result s of
tion of the 3 2 groups in soft clay, the results ofBarden and pile-loadir.g tests, and a typicalload-test result ís shown in
Mon.;kton are in fair agreernent with theory. Fig. 6.44. For each building, the foundation consisted of a
For two buildings at M.I.T., founded on end-bearing number of piJe groups, so that in obtaining the theoretical
piles passing through a deep deposit of clay and bearing on settlement ratios, both the interaction between the piles of
1 20
o 2 4 10 12 14
Miscel!aneous fili
l ~~
o (WL 30%1
-~
o
>
..!!!
IJJ
--60 f-
~~ o
~\
~\ o
stress,
~so.~ - :{ \ Sv éb o Ovo
x \ inferred lrom
\ OCR
-- '100 ~ - o e
¡Glacial till (ver y dense sandy grave!)
x F ield van e strength
Weathered shale o Maximum past pressure from oedometer tests
--120 (q" = 25 30 Tsf; e= Stsf, 1J = 29' 1
FIGURE 6.43 Typícal soíl pro file, MIT buíldings. (After D' Appolonía and La m be, 1971).
SETTLEMENT OF PILE GROUPS 139
Píkz
í
X
ing the settlements that occurred prior to commencement
of the settlement measurements--especially in the case of
the eastern (right-hand) side of the Space Center Building.
1 Poulos (1972b) presenteri a further comparison be-
tween predicted and observed settlements by corrparing the
theoretical relationship between settlement and number of
piles in the group with measured values reaorted by
PiPQ S<ZCtiOO'
D' Appolonia and Lambe ( 1971) for four buildbgs on the
12 · 7 5 in. OD x
M.I.T. campus. An average piJe length of lOO ft was as-
12·50•n. ID 1 sumed in deriving the theoretical relationship, which is
Dríwn w1th Boot Plot<Z
& F•ll,zd w•th 4000ps• concr¡zt.;; virtually linear and is an fair agreement with thc observed
relationship. Thls comparison is reproduced in Fig. 6.47.
o 0·5 1·0 1·5 2·5 3·0
S"ttl.;;m<Znt (in ) 6.7 .2 Load Distribution
FIGURE 6.44. Load-settlement curves from pile load test (a:'ter
D' Appolonía and La m be, 197 '! ). Comparisons between measured and theoretical oad-distri-
butions withln a model pUe group with a rigic cap have
be en made by Poulos ( l968b ). These are shov, n in Figs.
1 2 3 4 5 6 7 8 9 10 11
A r----r----,-----,-----·,----r----.-----r----r----~---~
i·40 ·32 !·35 ! !
B ~.......+---~--!---·· --!------ t·_.,.4,.-,8),__+·"'-'45"-'-J--t-f"'48"-'-)-t-1--...;...¡----+------1
·37 ·32
·40) (·39)
·40
{·41)
'__j
.
e-- +_3"'15"-----+'--4.:!.0~~,"'-4"'15'---+1 : - - - - ,- - - -
D f--- -+-
·45) ~-415) 1<·47)
... ......,...¡~_;..~-~~....:::.:___+~:.:-:____+-:.c:...:..__+--1-~-~ ·¡-- .
EL---~---L----L----L--~----~---L--~----~--~
M.zasur.;;d column S<rttl<2m<mts :n inchqs , Nov 1969
Pr<2d1ct12d línal colurnn S(lttl.zm.znts in brock<Zts.
FIGURE 6.45 Comparison bctween predícted and observed column settlements, MateriaÍs Scíence Building.
2 3 4 5 6
A -
B -
t '
e - 0·2_3_ _ _ _ --~-
0·46) r0·36l
'
8 1
)(
Poulos ( 1972b) presented a further comparison be-
_J
Dasign Load tween predicted and observed settlements by comparing the
theoretical relationship between settlement and number of
piles in the group with measured values reportcd by
p,P<2 Sectíon·
D'Appolonia and Lambe (1971) for four buildíngs on the
Driv~n ~~¡~~:~~!~ Í M.I.T. campus. An average pile length of 100 ft was as-
Plate
sumed in deriving the . theoretical rdationship. which is
Fílled with 4000psí concrete vírtually linear and is an fair agreement with the observed
relationship. This comparison is reproduced in Fíg. 6.47.
1
o 0·5 1·0 1·5 2·5 3·0
FIGURE 6.45 Comparíson betwcen prcdicted and observed col·~mn settlements, MateriaÍs Science Building.
2 3 4 5 6
A .
0·12
(0·3'1
8 -~~~-- ----fo:g~:::,.)---,.;:-
e -
but nevertheless exhíbits similar trends. For the 5 2 group, Group load,Pc 90 tons
both the magnitude of the loads, and their variation with Group-reduction factor, R G 0.40
spacing, are predicted mo.re closely by the theory. Single pile settlement at 0.20 mm/ton
unit load, p 1
Predicted settlement of 7.2 mm
6.7.3 Group Behavior Predicted from Single-PiJe Test group,PGPtRG
Result; Measured settlement of 7.1 mm
group
Tests by Koizumi and lío (1967)
A series of fuH-scale tests by Koizumi and Ito (1967) offers Although the cap of this group was in contact witL the soil,
.un excellent opportunity to study the prediction of pile·
the effect of the cap on group settlement calculated on the
group performance from the results of a single-pile test. A
basis of the analysis in Chapter 10 is in this case negligible.
single floating·pile and a nine-pile rigid-capped group of
It will be seen that there is excellent agreement between
similar piles were founded in a thíck uniform !ayer of sil::y
predicted and measured group settlement. The measured
clay overlain by a thin !ayer of sandy silt. The piles we;e
and theorétically-predicted load distributions within the
closed-end st~el tubes instrumented to allow piJe loads,
group also agree well, as·is shown in Table 6.1 O.
earth pressures, and pore pressures to be measured. Prmi-
Soil displacements near the group were recorded mainly
sion was also made for measuring displacements ar.d
pressures in the soil remote from the piles. Details of the
foundations and site conditwns are given in-Fig. 6.50.
By using the single-pile load-test· results, a soil modulus
TABU: 6.10 THEORETICAL AND MEAS-
of 2500 psi was backfigund, corresponding to a pile-stíff.
URED LOAD DISTRIBUTION TESTS Of
ness factor of 500, at a load factor against failure of ap- KOIZUMI AND JTO {196 7)
proximately 2.5. From the theoretical solutions presented
in this chapter, the settlemtmt of the rigíd-<:apped group (at Pile Load/ Average PiJe Loa<!
the same load factor) was calculated and compared with the PiJe Loca tion
measured settlement, as folbws: Theoretical Measureda
r L
u
"'n dy
~
¡m¡
!:.!L,E OETA~ '5ilt
len,;U; S.Sm
Dia(neter
Wal1
30 cm
0.)2 cm
V
Mater 1a t s tee \ zV
V lilt
/ <ia
/ Wlt
sh< u.
·/
I/
V
V
o'~o o 6/
/
T o o o
in connection with ultimate-bearing-capacity investigations groups with piles of L/d = 25, at a pile spacing of two dia•
and at working loads, are considerably smaller than those meters. In Table 6.11, the mea su red settlements are com-
predicted theoretically. pared with those predicted from the results of single-pile
tests carried out under the same conditions, and it can be
Model-Group Tests by Mattes and Poulos ( 19 71) seen that there is good agreement between observed and
Tests were carried out on 3 X 3 and 6 X 1 floating pile predícted settlements.
0
TABLE 6.11 MODEL PILE GROUP TESTS
of po:>sible approaches are described. Th~ effects of soc- = pL = Pu -the abo ve equations yíeld the following solu-
keting the tip and of pile-batter are then discussed, and tions for the depth of rotation, z,, and the ultimate lateral
the "thin-pile" analysis ís extended to piers with sígníficant ioadHu:
base resistance.
(7.3)
7 .2.1 Conventional Statical Approach
(7.4)
The simplest method of estimating the ultimate lateral resis-
tance of a floatíng pile is to consider the statics of a piJe, as
shown in Fig. 7.1 for apile with an unrestrained (or ''free")
head. The pile is subjected to a horizontal force H and a
moíuent M, and the ultimate soil pressure at any depth z P~~L = j(1 +~~y+ l- (1 + ~) (7.5)
below the soil surface is Pu. The limiting combination of H
and M, Hu and Mu, to cause failure-that is, to :mobilize
HudL is plotted against e/L in Fig. 7 .2.
the ultima te soil resistance along the pile, assuming the pile Pu
to be rigid-may be obtained by considering equilibrium of For the case of a linear variation of soil resistance with
horizontal forces and moments, and solving the resulting depth, from Po at the ground surface to PL at the pile t4p,
simultaneous equations for the unknown depth of rotation, the following equations may be derived:
z,, and the ultima te horizontal load Hu (taking the mom·~nt
Mu as He, whete e eceentricity of loading). Treating the
piJe as ~ thin strip of diameter or width d, these equatíons,
in general form, are
Zr L
+! l2po )
'(pL · Po
(!_)
L
(Zr)
L
(3 ~)ff!.o + PL)
\L VJL- Po
Hu
fo Puddz
f
Zr
Puddz (7 .1)
_ (2pL+Po\ O
PL PoJ
Zr
Mu Hu e J Pudzdz (7.2)
o
L
+ J Pudzdz 0·5r·-----¡------1~---~~--~
Zr
Uniform py
0·2
Pii<Z diomet,zr=d
Lin¡zarly
L 0 ' 1 Varying Pu
Distrlbution z ro at
~urfac!Z. pL at tip
O Pu PL /2
o 0·5 1·0 0·5
<l/L L/1Z
FIGURE 7.1 Unrestrained laterally-loaded píle. FIGURE 7.2 Ultímate lateral resistan ce of unrestrained rigíd píles.
ULTIMAn: LATERAL RESISTANCE OF PILES 145
~·::~-~--m=
1-p:1t2
'' ±;......M'H -- !
o~ ~=RJJ.~s~o~¡,=
1 1
\1-~
• ~ ---f-
ir'
N ~T"-1--l::::~ ,
~ ~ -~-
~N . '~ . .!¡__
-~ --~
1
•
i ' 1' " ' - 1
1 i '
1 h =! -H,l.---
1·0
'
''
¡
1
¡ ¡~
! "'
i -0~5 1'--' ......,8 o
~ '::-.f'.....
~~-
1...... i"{A
• ~¡ "t 1
0·5
1
1f.Jue'L
1
1
1
1
1 : ~ ::::.... Í'. ".....
1 -- 1
" "'
1
1
~
1 ........ 1
~".....
'
""" ~~- ~
l
! l
-0·5
1 ~
1 ! 1
-1 1'-- ..........
¡....._ ~
~
!
l l
i
i i
FIGURE 7.3 Ultímate lateral resístance of uruestrained rigid piles.
146 UL TI MATE LATERAL RESiSTANCE OF PILES
10
Srnoot;¡ (
Valua ganarally Co•OJ
8 assumad
0~--~--~--~----~--~--~--~----L---~--~
o 0·2 0·4 0·6 0·8 1·0
1-o o-e o-e o-4 o-2 o
b/d d/b
FIGURE 7.5 Effect of aspect ratio and adhesion ratio on lateral resistance for purely cohesive soil.
ULTIMA TE LATERAL RESISTANCE OF PILES 147
2~---L--~~--~---J
o 5 10 15 20
z z
cr a
(o) (b)
FIGURE 7.6 La tera1 resistancc factors Kq and K e (Brinch Hansen, 1961 ).
!'
1
L
D<2fl<2ction
100
Plastic
hing<2
(o)
w
9cud
D~flactton Soil R~oction Bc;~ndins Momant
",
N
60 (My~ldl
u _,.--.. Hu
"',
I 50
"e:
V
o
;;¡ 40
;¡¡
oc" JO
::'
~
o 20
-'
~
o" 10
I
~
1
::> (
o
(a) Emt>adm.,nt L•mgth L/d (b)
Mmax
&nding Momant
""', 100
u
..._
..¡' 60
"e
V 40
3
"'
;¡¡ 20
a:"
10
oL
!! 6
o
·-' 4
~
o
s 2
::>
1
34 6 10 20 40 60 100 300 600 (e)
D<:~fi<Zction Soil R4oction Banding Momant
i:bl
FIGURE 7.9 Ultimate lateral resistance in cohesive soils: (a) ~hort FIGURE 7.1 O Restrained piles, in cohesíve soil: (a) s.ho.~t; (b) ínter-
piles; (b) Ion piles (Broms, !964a). medíate; (e) long (after Broms, 1964a).
ULTIMA TE LATERAL RESISTANCE OF PILES 149
moment in the pile just below the cap is available*. Tn of predicted to measured ultimate loads is about two
Fig. 7.1 Oa, the following relationships hold for "short" piles: thirds. The distribution of soil resistance is
Solutions in dimensionless terms are shown in Fig. 7 .9a. effective vertical ·werburden pressure
For "intermediate" piles (i.e., first yield of pile ( 1 + sin <t;')/(1 - sin q/)
occurs at the head) in Fig. 7.10b, Eq. (7.9) holds, and angle of in terna! friction ( effective stress)
takíng moments about the surface,
The analysis resulting from the assumption of the above
My 'kudf(l .5d + 0.5[) (7 .13) factor of 3 is rnuch simpler than that which would follow
using Brinch Han sen 's variable factor Kq (Fig. 7.6 ).
This equation, together with the relationshipL 1.5d + f +g, Broms'~ approach is equivalent to assuming that Brinch
may be solved for fl~t. lt is necessary to check that the Hansen's Kq 3Kp for all depths. From Fig. 7.7, it can be
maxímum positíve moment, at depth f + 1.5d, is less than seen that for values of </> likely to obtain in sands, 3Kp
My; otherwise, the failure mechanism for "long" piles líes between Brinch Hansen's surface and deep values of
illustrated in Fig. 7.1 Oc holds. For the latter mechanism, Kq.
the following relationship applies:
Unrestrained or Free-Head Piles
(7 .14) Possible failure-modes, soil-resistance distributions, and
Hu = (l.Sd + 0.5[) bending-moment distributions for "long" and "short"
piles are shown in Fig. 7.11 (for constant soil unit weight 'Y
Dimensionless solutions are shown in Fig. 7.9b. along the pile). As before, the pile will act as a "short"
pile if the maxímum moment is less than the yield moment
of the section. In Fig. 7.lla, the rotation is assumed to be
about a point close to the tip, and the high pressures acting
7.2.2.2 PILES IN COHESIONLESS SOILS near this point are replaced by a single toncentrated force
at the tip. Taking moments ábout the toe,
The following assumptions are made in the analysis by
Broms (1 964b ): 3
H = 0.5 ¡dL Kp (7 .16)
u e +L
l. The active earth-pressure acting on the back of the pile
is neglected.
This relationship is plotted in Fig. 7 .12a using the dimension-
2. The distribution of passive pressure along the front of the
less parameters L/d and HulKp"fd 3 • The maxímum moment
piJe is equal to three times the Rankíne passive pressure.
occurs ata distance /below the surface, where
3. The shape of the pile section has no influence on the
distribution of ultimate soil pressure or the ultimate lateral
resistance. (7 .17)
4. The full lateral resistance is mobilized at the movernent
considered. that is,
fu)
The simplified assumption of an ultirnate soil resist-
ance, Pu, equal to three times the Rankíne passive pressure
is based on limited empírica! evidence from cornparisons
bet)Veen predicted and observed ultimate loads rnade by
f 0.82
Jl -
Kp'Y
/'7
1 1
3YdLKp
( b)
F!Gt;RE 7.11 Frec-head piles in a cohesionless soil: (a) shOit, (b) long (after Broms, 1964b).
R<Zstrom<Zd/ i
---:-7 Fr<2<2 t1<2adad
Hu ------r--r- <2
~
V
e: :~~~
tu / ~. T=O
~ .120 ¡-:----,f-+--1'-+---1
¡¡; --if-d / •
""
0::
aor----4----~~----~-~~~~~~
o
L
...o
(j
..J
o 4 20
Embadmant L.;¡ngtn, Lid
(o)
.:- 1000 . - - - - - ; - - - - . . - - - - , . - - - - , . . - - - - - . .
'h
"'' "
I
(j
g 100
o
tí
·¡¡;
0:: ""
oL
(j
+'
o
..J
ti
+'
o
E
:;;
1·0 10 100 1000 10000
S
(b)
FIGURE 7.12 Ultimate lateral resístance of píles in cohesionless soils: (a) short; (b) long (after Broms, 1964b).
(Mm 0 x).
1/>=0
Y=•J
i __
(a) D<lfl<tctiOn
l e
FIGURE 7.14 Plastícity analysís for Jaterally loaded plate.
M
en
Failur~
-2
FoíiUr<l
-1
- - ..._ ___ _ \
,~
l
FIGURE 7.15 Failure of a rigid vertical platc under moment and horizontal load (Davis, 1961}.
relevan t lo a group o f pi! es dosel y spaced in a single long vertical load, the pite weight, and the vertical component of
row. For the latter, the plane-strain plasticity analysis the lateral forces on the front and back of the píle.
plays an analogous role for the loadings considered in this Consideration of these forces, together v.ith those actíng
chapter to the role played by the block-bearing-capacity on the front and back of the piJe, leads to a quartic
analysís in consideration c,f groups of piles subject to expression for the position of the center of rotatíon.
vertical load (Chapter 3). Solution of this equation .enables the ultimate lateral load
to be calculated. Roscoe also describes a similar analysis
for tíed piers, restrained to rotate about the center of the
píle at the ground surface. In both analyses, however, it·was
7.2.4 Pites with Significant Base Resistance assumed that the ultimate lateral pressure on the pile was
the dífference between the Ranking passive and active
Satisfactory theoretícal solt::tíons to thís problem have not pressures; thís assumption may be conservative unless the
yet been obtaíned. For relatively long píles, it may be pile or pier is shallow.
adequate, if c'onservative, to add the shearing resistance of
the base of the .pile to the ultimate lateral resistance of the
píle calculated from the pre.ceding sections. F or relatively
short piers, the base may provide sígnificant moment- 7 .2.5 Socketed Piles
resistance, and this can be estimated from bearíng-capacíty
theory for eccentric and inclíned loading (e.g., Meyerhof, For piles that· are socketed into rock or whose tip is
J953). As the length-to-diameter ratio de creases, the embedded in a firmer stratum, a moáificatíon of the
center of rotation moves downward toward the base of the preceding analyses for floating piles is necessary. A typical
piJe and may even be located outside the pile. In such cases, case is illustrated in Fig. 7.16, for a free-head. pile. Here,
it may be desirable to consider alternatíve failure-mechan- assumed failure-modes and moment-distributions are shown
isms and adopt the one giving the mínimum ultímate- together with an arbitrary distribution of ultimate soil
lateral-resistance of the pier. resistance, Pu. The actual distributíon of Pu may be
A reasonable engineering approach has be en suggested by estimated from the theories described in Section 7 .2.1. It is
Roscoe (1957), who considers the presence of a horizontal again assumed that the effect of the high pressures near the
shear-resistance at th(~ interface between the base and the tip may be replaced by a single force, sínce the center of
soil and í.he effect o~· an eccentric vertical reaction acting rotation ís obviously close to the tip. Considering "short"
on the base. This reactíon is balanced by the applied piles first: Takmg moments about the típ gives
154 ULTIMA TE LATERAL RESISTANCE OF PILES
Soll
C<:mtNZ O!
Rotatíon
_F
L, where
(I0
PurZ dz (7 .24)
z vertical distance, meaSJred downward from the sur-
face
L + Lr )
+ J Puszdz The bracketed term ipur>, in Eq. (7.25) applies only if f
L, extends below the top of the stiffer stratum. The maximum
moment is then
where
f
ultímate lateral resistance of stíffer stratum d J {pus + <Pur>)zdz (7 .26)
ultimate lateral resistan ce of soil o
z vertical distance, measured upward from the tip
Hu m ay then be determined freo m Eqs. (7 .25) and (7 .26).
For tile value of Hu thus calculated,1the maximum moment, For piles with a restrained o: fixed head, similar analyses
may be carríed out; these cases may be treated as extensions
Mmax must be checked. If Mmax < My, the piJe will fail
of the restrained-pile analyses of Broms, 1964a and b
as a short piJe. I f M max >My, then the piJe will fail as a long
(Section 7 .2.2).
piJe, and Mmax must be equal toMv. The position oUHmax
(dislance f below the surface) may be determined by the
condítion of zero shear there; that is, when 7.2.6 Piles Subjected to Inclint~d Loading
axial direction, fail ure will occur essen tially by axial slip
(and also bearing failure of the tip for downward loading).
Lateral faílur« will occur when the inclinatíon of the applied
load ís large, that is, as the load becomes perpendicular to 300
2
the píle axis. The above two modes of failure will occur as Cu= 15001b/tt
follows: 20' Ca= 0·45 Cu
250 = 6751b~t:1
l. Axial failure will occur when the ultimate lateral J~
i:apacíty exceeds the horizontal componen! of the ultima te
~~:!'
inclined load: that is, when 200 Axial (uplift) --t---=La=t=o;ro~l---i
/
Load
Capocrt.y
or Ou
kips
Hu > Pu tan o (7.27)
where
.... ·:::
·: :: :·
...... '',
l ;.:.:_:.;·.
.. .
11. • . .•
. . ·....
1 1 ..
1
1
\
L \
1 \
1 \
1 1 \
1 \
1 \
d \
\
1 \
1 1 \
1 1 \
----~~~---------~
'-
jp-9lKpY~
Daflaction Earth Prassur~ Distribution
FIGURE 7.18a Earth-pressure distríbution for oblique pul! (Broms, 1965).
Puo axial capacity when the applied load acts along Lateral Failure
the pile axis If it is assumed that the vertical component of load does not
D..Pu íncrease in pullout resístance caused by the two affect the lateral resistance of the píle, then
latera:! forces, T and R, ín Fíg. 7 .18b
Qu = Hu cosec o (7 .33)
The ultimate inclined-load capacity is then
where
(7.32) ultima te lateral resista.:-tce for horizontalloading
Consíderation of the statics of the pressure distribution ín The actual load capacity is then the lesser of the values
Fig. 7 .18b cnablcs Mu, and hence Qu, to be determine d. calculated for axial and lateral faílure.
For piles with a restrained head, or "long" piles {which
may fail by failure of the pile it>elf), the approach outlined
v~a
16
above may be extended in a similar manner to that described
in Section 7 .2.2.
/Í
Broms (1965) has compared predicted ultimate load
cz
L_' H capacities with those measured in the tests by Yoshimi
{1964) and found reasonably good agreement. An example is
2/3g shown in Fig. 7.19 for an 18-in.-long model pi! e in sand
g
subjected to a load inclined a.t an angle of 30° to the
vertical, the piJe being battered at an angle {3 to the vertical
{the pile is treated as a vertical pile loaded at an angle
{3 + 30° to the vertical-see next section). Also shown is the
L
predicted ultimate load if no allowance is made for the
1 1
increase in uplift capacity caused by lateral load. The latter
1
1 prediction is obviously very conservative.
1
1
1
1 1
l...j R
7 .2.7 Battered Pi!es
500
400
\ Latarol
Failura
0 Maasurad voluas (Yoshirní.1964)
-F"n<díctad valuas (Brcrr,s ,1965)
- --Pradictad valu<Zs (no allowanc:a
..ó 300 (
0
\ tor lataral a1fact on uplift
capacity)
"O
o
o
_J Axial
\
-------{
200 Fallura\~'
..,<:;
o
E
..-'
:::>
100
1~
o
-30 -15 o 15 30
Inclmatíon ~
FIGURE 7.19 Predictcd and observed inclincd-load capacity-modcl pilc in sand.
(7 .34)
L
7.3 PILE GROUPS
.¡. e=15cm
a..~ a:' L= 50cm
'O 'O 140 D:2·0cm
"' o ~·=37· 2°
SS y :H15Pjc
~o 130
r: .~
ue
·- +'
f..
e¡¡
H>
........
" e¡¡
....ow ...."o
110 .~ .~
+' ...
FIGURE 7.21 Load capacity of batlered píles and píles subjetled lo inclined load (Awad and Petrasovits, 1968).
l. n tí mes the lateral load capacity of a single pile. Ultima te lateral-load r:apacity of group
(7.35)
2. Thc lateral load capacíty of an equivalen! single block n X ultimate lateral load capacity of
containing the piles in the group and the soil between single piJe
them.
A relatively small amount of data ís available for values of
The fírst value, representing individual piJe faílure, can be TfL. A series of tests on model pi! e groups in clay was carried
obtaincd by the methods described in Sectíon 7.2. The out by Prakash and Saran {196í') while Oteo (1972) carried
sccond value, representing block failure and occurring at out similar tests in sand; the values of TJL derived from these
re!ati·<eiy clm:e spacings, can be obtaíned as described in test results are shown in Fig. 7.22. r¡L decreases wíth
Section 7.2.4 for an equivalen! single piJe of diameter or increasing numbers of piles in a group or with decreasing
width equal to the breadth of the group perpendicular to
the dírectíon of loading. However, in using Broms's
theory for a píle group in clay, ít is clearly absurd to allow
a "dead" zonc of zero soíl-reaction of 1.5 times the group
breadth, while ígnoring such a zone may be unduly
optimistic. A reasonable compromise is to use a "dead"
zone of thc lesser of l .Sd (d = individual pi! e diameter) or
O.IL (l. embedded length of piles). Results of a limitcd
series of model tests suggest that the above procedure gives
(l
a reasonable estímate of the group capacity at close spac- :l
oL
ings. lt the group ís relatively narrow, and loadcd perpen- (?
dicular to the longer direction, the ultimate lateral load
for block faílure may be estimated from the plastícity
solutions in Fíg. 7.15. For a gro u p of fixed-head piles,
with the head embeddcd in a massive cap, the ultimate
2 3 4 5 6
load for block failure can be calculated as the sum of the
j
Spacrng Prle DiorPetcr
resistan:e of a short restrained pi! e ( e.g., see Fig. 7.9 and
7.1 2) and the shear resistan ce o f the base. So me allowance - - Model pi les incloy ~- 32
may also be made for side shear resistance of the block. (Prokosh Jnd Sornn (1967)
The concept of a group efficiency for lateralloadmg, - Model P'l''' in sond ~ -275
(Ofec (1972))
rn, can be employed as wíth group efficiency forvertical
loading, where for a group of n piles, FIGURE 7.22 Lateral group efficiency from modeltests.
ULTIMA TE LATERAL RESISTANCE OF PILES 159
spacing. Comparison with the efficiency values ,(r¡) for 7 .3.2 Groups Containing Batte1:ed Piles
axially loaded groups shows that 1)L < r¡, or in other words,
that groups are less efficient under lateral loading than As with groups of vertical piles, the ultimate lateral·
under ax1ial loading. Model tests on three- and seven-pile load capacity of groups conta:ning battered piles may be
dolphins embedded in sand o ver .ying silty clay ha ve be en taken as the lesser of
described by Tschebotarioff (1153). For the three-pile
group (a central vertical pile an d two outer battered at l. The sum of the lateral lcad capacities the individual
10°), ancl efficiency 77L of O. 7~· was calcula ted, the piJe pi les in the gro u p.
spacing at the ground line being about nine diameters. 2. The load capacity of the group acting as a single
For the seven-pile dolphin, containing a central vertical block.
pile surrounded by piles batterecl at 5°, with a spacing at
the ground line of about three diameters, 77L = 0.52. The first value can be estima te el from Section 7 .2.1 or
Because of the effects of pile batter and the layered soil Section 7.2 .2 for vertical pi les and Section 7.2 .4 for
prcifile, direct comparison between Tschebotarioffs results battered piles. The second value can be estimated in a
and those shown in Fig. 7.22 is not possible; nevertheless, similar fashion to that described for vertical groups, but
the values of 77L in the two sedes of tests appear to be now allowing for the battered piles. If an analysis of the ·
reasonab 1y consisten t. type describe el by Roscoe (19 57) is employed, su eh allow-
Finite-element analyses provide a means of theoretically ance may be made by considt:ring the resultan t forces on
estimating the lateral efficiency of a pile group. By the front and back of the group to act on inclined surfaces.
carrying out a plane-strain analysis of the pile group in if the front and back piles ate battered. The base shear-
plan, and employing a f\Onlinear stress-strain relationship resistance may also be assumed to act over the plan arej
Íor the soil, a load-deflection curve may be obtained for of the group at the leve! of the pile tips. Alternatively,
the pile group (on the assumption th1t the piles are infinitely and more simply, an equivalen! block with vertical sietes
long). By comp:1ring the maximu1:1 load capacity from this may be considered. Both of these approaches imply that
analysis with tHe corresponding ·value for a single pile, an if the group fails a5 a single block, the ultimate laterai-
estímate of the group lateral efficiency may be made. load capacity of the group dE pends on the batter of thc
Analyses of this tvpe have been performed by Yegian and outer pites only, and not on the batter of interior piles.
Wright (1973) and Moser (1973). The solutions obtained by Sorne confirmation that this inplication is reasonable may
Yegian and Wright show that the efficiency oftwo or more be seen from the results of model-pile tests in clay carried
piles in a row is considerably ess when the horizontal out by Simek (1966), who f,)tmd that little benefit was
loading is in a direction parallel to the line joining the' piles
than when it is perpendicular. For example, for two piles at
center-to-center spacing of two diameters, the efficiency is
about 0.72 for loading parallel to the piles, but 0.90 for
loading perpendicular to the piles.
In designing the individual piles in the group, it is
desirable to determine the loac distribution within the
group. Mcthods for such determinations are described in
Chapter 9 for groups subjected to a general system of loading,
and although these methods are strictly valid only for
working-load conditions, they probably give a reasonable
estímate of the failure load di:>tribution. In practice, a
widely use el and relatively simple c[,~sign-method of determin-
ing the load distribution in a group with a rigid cap is to
assume that the piles in the group carry equal proportions
of the applied horizontal and vertical load, together with
additional vertical loads that are proportional to the dis-
tances of the piles from the gro11p's center of gravity and
thereby balance the applied moment. This approach ignores
the effect of the soil and considers tJ.' oile group simply as
a structural system. FIGL'RE 7.23 Model g10ups tested by Simek (1966).
160 ULTIMA TE LATERAL RESISTANCE OF P!LES
dcrived by having battered piles additional to those at the creases significantly with increasing embedded 1ength.
encl of the group. The benefits of having the end piles Also, as indicated above, the effect of batter in increasing
battered were found to be particularly marked when the lateral-load capacity decreases as the embedded length
piles were embedded in the soil for only a relatively small increases, and for 75% ernbedment, has virtually no
dis:ance. Four group-configurations were tested by Simek, effect.
as shown in Fig. and the results of these tests are
summarized in Table 7 .1. The ultirnate lateral load capacity,
Hu, is expressed as a percentage of the total weight, W, of
the ~roup, and it is se en from Table 7.1 that Hui W in- 7.4 USE OF PI LES TO INCREASE SLOPE ST ABILITY
-
by p:le, Hu (Hu P)
Additíonal resísting
moment H, R Hu e
a piJe is installed in the slope, the portion of the píle distribution has been determired. Approximate allowance·
(length L 1 ) above the assumed failure surface wíll be can be made for th~ inclínat.on, as outlined in Section
subjected to an inclíned dísturbíng force P at sorne 7 .2.6. The eccentricity e can, as a first approximation, be
eccentricity e abo ve this surface. lgnoring any axial estimated by assum.ng full nobilization of :he pile-soil
resistance for símplicity, this disturbing force can be con- pressure above the assumed faiL.tre surface.
sidered to be resisted by tite lower portian of the pile Once the value of Hu has been thus determined, the
(length L 2 ) below the critica! failure surface. The maxi- additional resisting moment e r force caused by the pi! e
mum val u e of this resísting force. H 11 , is given by the least can be detetmined, and hence t:1e effect on the safety factor
of the following four values: can be evaluated (see Fig. 7.~4). The procedurc must be
repeated for a series of tria! failure surfaces to find the one
1. The ultimate lateral resístance of a "short" pile of with the lowest safety factor. ·::onsideration should also be
length L 2 loatled atan eccentricity e. given toa surface that passes be low the piJe tips.
2. The ultima te lateral' resistan ce of a ''long" pi le loaded With groups of piles, adju:;tments can be made to the
at an eccentricity e (this value will depend on the yield ultimate pile-soil pressures to allow for group effects, anJ
moment of the pile ). the influence of each píle can be added up to tle termine thc
3. The ultimate load tha t can be developed along the uppe r effect of the gro u p on slope stability.
part (kngth L 1 ) of the pi le if the soil t1ows past the piJe and
the ultima te pile-soil pressure is developed along this portion
of the piJe. 7.5 METHODS FOR fNCREASING TffE LATERAL RE-
4. The shear resistance of the piJe section itself. SISTANCE OF PILES
The values in 1, 2, and 3 may be obtained from the analysis Broms ( 1972} has di ;cussed so:ne rncthods of increasing the
presented in Section 7 .2, once the ultima te pile-soil pressure lateral resistan ce of píles. As shown in Fig. 7.25, most of
Sand or
gravel fíll
(a}
77
Prefabricated
concrei:e pi!e Beams
these methods rely on increasing the dimensions and/or cyclic loads. The fill gradually works itse1f down into the
stiffness. of the piles near the ground surface. The use of a clay and increases the effective diameter of the piles. The
sand m gravel fill placed around a piJe (Fig. 7 .25a) is height of the fill around the piles is limited, however, by
very effective for soft clays when the piles nr~ subjected to the bearing capacity of the unürlying soil.
LOAD·DEFLECTION PREDICTICl~N
FOR LATERALLY LOADED PILE:S
163
164 LOJ'..D-DEFLECTION PREDICTION FOR LATERALL Y LOADED PlLES
since the displacements ata point are influenced by stresses (for example only elastic), or too time consuming, or too
and forces at other points wiíhln the soil. A further dls- uncertain with regard to accuracy to be entirely suitable
advantage is that t~Jiog_!IlQ@l!!~ of the ¡;nodel (the for parametric studies.
modulus of subgrade reaction) is_depe~Lc!.w.LQllJhe si~~f In thls chapter, the application of both the subgrade-
the_fP!:l!!_datiQn. In spite of these drawbacks, the subgrade- reaction and elastic approaches w the analysis of a single
reactíon approach has been widely employed in foundation pile is described in detail, and in ea eh case, the available data
practice because it provides a relatively simple means of on the relevant soil parameters i; reviewed. The extension
analysis and erables factors such as nonlinearity, variation of the elastic approach to the e:>timation of group move-
of soil stiffness wíth depth, and layering of the soil profile ments is then described.
to be taken into account readily, if only approximately. In
addition, despíte the many difficulties in determining the
modulus of subgrade reaction of real soil, a considerable
amount of experience has been gained in applyíng the 8.2 SUBGRADE-REACTION ANAL Y SIS
theory to practica! problems, and a number of empírica!
correlations are available for determining the modulus.
Frorn a theoretical point of view, the representation of 8.2.1 Basic Theory
the soil as an elastic continuum is more satisfactory, as
account :.s then taken of the continuous nature of soil. The In the Winkler soil model, the pressure p and deflection p
use of fús model for the analysis of the settlement of at a point are assumed to be related through a modulns of
piles and pile groups, as described in Chapters 5 and 6, has subgrade reaction, which for horizontal loading, is denoted
been found to provide a convenient and relatively reliable as kh. Thus,
means of describing pile behavior under axial loading.
While the elastic model is an idealized represen tation of p =e knP ( 8.1)
real soil, it can be modified to make allowance for sciÍl
yield and can also be used to give approximate solutions for where kh has the units of force/length 3 . Equation (8.1) has
varying modulus with depth and for layered systems. Jn been restated frequently (e.g., Reese and Matlock, 1956;
addition, it has the important advantage over the subgrade- Davisson and Gill, 1963) as
reaction approach of cnabling analysis to be made of
group ac:íon of piles under lateralloads; also, it provides a w = Kp (8.la)
means of analyzing the behavior of battered piles subjected
to a g,eneral system of loading (Chapter 9). A further where
advantagt! of the elastic model is that it enables consistent
analysis of both immediate movements and total final w soíl reactíon per unit-length of piJe
movements. The majar drawback to the applicatíon of the K subgrade-reaction modulus, in units of force/
e las tic method to practica! problems is the di fficulty of length 2 (K = khd)
deterrnining !he appropriate soil rnoduli; however, this d diameter or width of píle
dífficulty also exists to a certain extent with the subgrad,>
reaction method. The pile is usually assumed to act as a thin strip whose
The ~xact solution of the problem of a laterally load,~d behavior is governed by the beam equation
f1exible pile in an elasto-plastic soil rnass is a complicat<~d
and dífficult une in tluee-dimensional continuum mecha- (8.2)
nics and does not appear to have been satisfactorily solved
at. present. Some attempts include IWo-dimensional finite
elernent treatments in the horizontal plane (Yegian and where
Wright, 1973; Baguelin and Frank, 1979; Rowe and
Poulos, 1979); a special finite element technique which is Ep modu!us of elasticity of piJe
capable of dealing with general three.dimensional loading fp moment of inertía of pi..e section
for ax1-symmetric geometries, but only for elastic cond.i- z depth in soil
tions (R.mdolph, 1977; Banerjee and Davies, 1978); and d width or diameter of píle
general bree-dimensional elastic finite element analysis but
with allowance for axial slip via the use of joint elements As in the simple theory of bending of beams, the
(Desai and Appel, 1976). Such analyses are too restrictive effect of axial load in the pile i:; ignored. (The effect of
LOAD-DEFLECT!ON PREDlCT!ON FOR LATERALLY LOADED PILES 165
that is,
(8.4)
1JL3
-p_z + 2p_¡- + PJ (8 5)
+ (k¡dp¡) o Ff;lpn:;
• -2
• 1 2. A fíxcd-head piJe, for wluch Eq. (8.5) still applies, and
also,
rotation El dp O
'P ¡:.dz
that is,
i-2
P2 - P-! =0 (8.7)
2
• i t
-shear o
L' that is,
_ L _ ___ ··--~-T,, 1 ~Pile tip
1
- Pn-1 + 2Pn - 2Pn+2 + Pn+! o (8.8)
o n.,. 2
1 n +J
and moment o
FIGURE 8.1 Finite-difference analysis of laterally loaded piles.
166 !-OAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES
(8J3a) (8.l9a)
-2M 0 (J
G~) X
Q = -------:---- [sin {3L sin {3(L -, z) (8.20)
M (8.14)
sinh {3(L - z) - sin {lL sinh {3z sin p(L - z) J. sin {3z + sin {3L sinh (Jz sin {3(L ·-· z)]
sinh p(L- z) sin ¡3z cosh {3(L- z)] sin {3L [cosh ¡3z
(8.21)
sin {3(L - z) - smh {3z cos {3(L z)])
{3 {k '14f':T, 1p )Y.
\ /¡U¡ (8.16)
--r~-t'L
X
.'!!
LL
(8.17a)
For rigíd pi le
1 4 6 12 '----'
o (8.18) 10 100 1,000 10,000 100,000
Deflact on and slope factors
[sinh (JL cosh (J( L ;; ) cos {3z + sin {3L cosh {3z cos (J(L - z) J FIGURE 8.2 Top deflection and rotation for lateral loads on ver-
tical piles for constan! kh (after Barber, 1953). (Reprínted by per-
sính 2 (JL sin 2 pL · míssion of the American Society for Testing and Materials, © 1953.)
168 LOAD-DEFLECTION PREDICTION FORLATERALLY LOADED PILES
Values fór the dimensionless coefficients Kp¡¡, KeH, 1. Rigíd píle (holds if ~L < 1.5):
and so on, are given in Table 8.2. For deflections and
rotation~; at the soil surface, convenient plots presented by 4H(l + 1.5e/L)
Barber ( 1953) are shown in Fig. 8.2. For a free-head or un- p (8.25)
khdL
restrained pi! e,
{) 6H(l + 2e/L)
~. khdL 2 (8.26)
deflection, p
-----
(JL z/~ Kpf/ K oH KMH KQH Kplv! KeM KMM KQM
2.0 0.8125 -0.1981 0.54 74 0.0468 -0 2245 0.4171 O. 16~ O 0.0698 0.1787
2.0 0.8750 -0.2659 0.5389 0.0222 -0.1665 0.4572 0.15~'8 0.0318 0.1241
2.0 0.9375 -0.3330 0.5356 0.0059 -0.0916 0.4963 0.15:.4 0.0082 0.0645
2.0' 1.0000 --0.3999 0.5351 O. -0 0000 0.5351 0.1551 0.0000 o.
3.0 O. 1.0066 1.0004 o. 1.0000 -1.0004 1.0C·8 1.0000 o.
3.0 0.0625 0.8210 0.9695 0.1543 o 6575 -0.6589 o.8U:3 0.9690 0.1545
3.0 0.1250 0.6459 0.8919 0.2508 03829 -0.3854 0.64::3 0.8913 0.2514
3.0 0.187 5 0.48:J2 0.7870 0.3018 O.l 709 -0.1743 0.48S7 o. 7862 0.3029
3.0 0.2500 0.3515 0.6698 0.3184 00141 -0.0184 0.3493 0.6684 0.3202
3.0 0.3125 0.2371 0.5514 0.3101 -0 0956 0.0905 0.23:;2 0.5491 0.31. 27
3.0 0.3750 0.1444 0.4394 0.2850 -0 1664 0.1607 0.14::9 0.4360 0.2887
3.0 0.4375 0.0716 0.3389 0.2496 o 2063 0.2002 0.071 o 0.3339 0.2544
3.0 0.5000 0.0164 0.2528 0.2091 -0.2223 0.2162 0.0168 0.2458 0.2150
3.0 0.5625 -0.0242 0.1823 0.167J -0.2205 o 2147 -0.02:!2. 0.1728 0.1744
3.0 0.6250 -0.0529 0.1271 0.1272 -0.2057 0.2011 -0.041:9 0.1148 0.1353
3.0 0.6875 -0.0727 0.0864 0.0908 -0.1819 o. 1793 -0.0661 0.0709 o "~995
3.0 0.7 500 -0.0861 0.0584 0.0594 -0.1519 0.1524 -0.0763 0.0396 0.0684
3.0 0.8125 -0.0953 0.0411 0.0340 -0 1178 0.1227 -0.0816 0.0189 0.0426
3.0 0.8750 --0.1021 0.0321 0.0154 -0.0807 0.0916 -0.0839 0.0069 0.0225
3.0 0.937 5 -o.1q11 0.0287 0.0039 -0.0414 0.0599 -,'),0846 0.0014 O.OC83
3.0 1.0000 -0.1130 0.0282 o. -0.0000 0.0282 -0.084 7 0.0000 o.
4.0 O. 1.0008 1.0015 o. 1.0000 -1.0015 1.0021 1.0000 O.
4.0 0.0625 o. 7550 0.9488 0.1926 0.5616 -0.5624 0.7567 0.94 72 0.1929
4.0 0.1250 0.5323 0.8247 0.2907 0.2411 -0.2409 0.5344 0.8229 0.2910
4.0 O.!B75 0.3452 0.6693 0.3218 0.0234 -0.(,220 0.3478 0.6673 0.3219
4.0 0.2500 0.1979 0.5101 0.3093 -0.1 !08 0.1136 0.2010 0.5082 0.3090
4.0 0.3125 0.0890 0.3641 0.271 í -0.1810 0.1855 0.0926 0.3626 0.2705
4.0 0.3 750 0.0140 0.2403 0.2226 -0.2055 0.2118 0.017 8 0.2397 0.220;
4.0 0.4.375 -0.0332 0.1419 0.1715 -0.1996 0.2079 0.0295 0.1430 0.1671
4.0 0.5000 -0.0590 0.0682 0.124:' -0.1758 0.1858 -0.0558 0.0720 0.1176
4.0 0.5625 --0.0692 0.0163 0.084:. -0.1432 0.1545 ··0.0674 0.0242 0.0749
4.0 0.6250 -0.0687 -0.0176 0.0529 -0.1084 0.1200 -0.0696 -0.0043 00406
4.0 0.6875 -0.0615 -0.0379 0.0299 -0.0756 0.0858 -0.0665 -0.0178 0.0149
4.0 o. 75p0 -0.0505 -0.0488 0.014 7 -0.0475 0.0538 -·0.0616 .. 0.0206 -0.0025
4.0' 0.8125 -0.0376 -0.0536 0.0057 -0.0255 0.0242 -0.0568 -0 0166 -0.0122
4.0 0.8750 -0.0239 -0.0552 0.0014 -0.0101 -0.0033 -0.0535 -0.0096 -0.0148
4.0 0.9375 -0.0101 -0.0555 0.0001 -0.0016 -0.0296 -0.0520 -0.0029 -0.0106
4.0 1.0000 0.0038 -0.0555 -0. 0.0000 -0.0555 -0.0517 --0.0000 -0.
.'l.O o. 1.0003 1.0003 o. 1.0000 -1.0003 l.OOC2 1.0000 O.
5.0 0.0625 0.6964 0.9214 0.2249 0.4 711 -0.4715 0.6964 0.9211 0.2250
5.0 0.1250 0.4342 O. 74 76 0.3131 0.1206 -0.1210 0.4343 O. 74 72 0.3133
5.0 0.1 S75 0.2317 0.5479 0.3155 -0.0842 0.0840 0.2320 0.5472 0.3158
5.0 0.2500 0.0901 0.3628 0.2716 -0.1817 0.1818 0.09(7 0.3620 0.2720
5.0 0.3125 0.0013 0.2121 0.2093 -0.2079 0.2084 0.0022 0.2111 0.2095
5.0 0.3750 -0.0466 0.1013 0.1461 -0.1919 0.1930 -0.0455 0.1002 0.1461
5.0 0.4 37 5 -0.0659 0.0277 0.0915 -0.1556 0.1575 -0.0644 0.0267 0.0910
5.0 0.5000 -0.0671 -0.0157 0.0494 -0.1133 0.1163 -0.0654 -0.0161 0.0482
5.0 0.5625 -0.0584 -0.0368 0.0203 -0.0738 0.0778 -0.0567 --0.0361 0.0180
5.0 0.6250 -0.0456 -0.0435 0.0026 0.0412 0.0461 -0.0444 -0.0409 -0.0012
5.0 0.6875 -0.0321 -0.0419 -0.0063 -0.0169 0.0223 -0.0321 -0.0365 -0.0117
5.0 0.7500 -0.0197 -0.0369 -0.0088 -0.0008 0.0055 -0.0221 --0.0276 -0.0159
5.0 0.8125 -0.0090 -0.0317 -0.0075 0.0081 -0.0059 -0.0150 -0.0175 -0.0157
5.0 0.8750 0.0002 -0.0279 -0.0044 0.0108 -0.0139 -0.0110 -0.0086 -0.0125
5.0 0.9375 0.0086 -0.0261 -0.0014 0.0079 -0.0201 -0.0094 -0.0023 -0.0072
5.0 1.0000 0.0167 -0.0259 -0. 0.0000 -0.0259 -0.0091 -0.0000 -0.
170 LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES
L 3 nh a'
1,000
:"
2. lnfmitely long pile (Zmax > 4.0): ~
:¡¡
;¡ 100
.!!!
2AOH (8.33) LL
p
(nh)315(E
. . p ¡p )2/5
10
Solutions for pile-head deflection and slope, obtaired Dependíng on the angular restraínt provided by the
by Barher (1953), are plotted in Fig. 8.3. The actual cap, values of M/HT will range from zero for a free-head
deflection and slope are given by Eqs. (8.22), (8.23), and case to -0.93 for the case where the cap prevents any
(8.24), except that k~zd is now replaced by n~¡L in the rotaticn of the pile head--in other words, the fixed-head
denominator of these equations. case. Davisson (1970) suggests that the degree of fixity that
A comprehensive series of solutíons for deflection, can usually be developed is I•,f/HT -0.4 to -0.5; for thís
rotation, moment, shear, and pressure along a pile have case, the positive o.nd negative moments in the píle are
been presented by Reese and Matlock (J 956). For the case approximately equaL
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES 171
r'-J
e:
"'
:~
"'uo 1.0
-'=
l}
o
1.2
Defection coefflcient, e,
FIGURE 8.4 Curves of deflection coefficient Cy for long pites (after Matlock and Recse, 1961)
8.2.2.3 c;f.'Nl:RAL D!STR!BUTfO:V OF kh WIT!J DEPT!J factors lp!J, lp,"f, and lpF for uniform k ís shown in
8.6, 8.7, and 8.8 as a function of the ratio of thickness d
Cases involving a general distribution of k~¡ wíth depth, of
upper !ayer, h 1 , to length L, md the ratio of modulus of
the form k~¡ nhz11 jd or khd = k 0 + k 1i + , have been
upper !ayer, k 1 , to that of the lower !ayer, k. These results
considered by Matlock and Reese (1960). deflections
ha ve been derived from those of Davisson and Gill (1963)
and moments are associated with larger values of n.
and apply to a pi! e of intermediate flexibility (kdL 4 / éplp
Matlock and Reese also give solutions for the case k~zd =
256): they may be used as factors to correct the uniforrn
k0 + Both deflections and moments decrease as k 0
!ayer influence factors in Fig. é .2.
in creases.
The results of Davisson and Gill's analysis may be
summarized as follows:
8.2.2.4 LA YERED SOILS
S'olutions for pllles in a two-layer system ha ve been presented l. With respect to reducing ~;urface deflection and maxi-
by Davisson and Gill (1963) and Reddy and Valsangkar mum moment, there is little :Jenefit to be gained from a
(1968). The intluence of the upper Iayer on the detlection stiff surface layer exceeding a éepth of 0.2R (a depth of the
172 LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES
0.2
0.4
0.6
0.8
1.0
1.2
"'e'
~
1.4
"'
.ü
:E
o"'
u 1.6
L case
o_
o"'
1.8
2.0
2.2
2.4
2.8 1----..i·····--+-~+--··-+·····--l----
Moment coefficient, Cm
FIGURE 8,5 Cmves of moment coefficicnt Cm for long píles (Matlock and Reese, 1961).
order of a few pile-diameters) or from a modulus ratio consider a distribution of kn in each ]ayer of the form
e.xceeding 5, where khd k 0 + k 1z + k 2 z2 . The conclusions from this analysis
are identical with those of Davísson and Gil!.
R lié'
~p
1p¡lkd] l/4
2. The soil from the ground surface to depths of 0.2R to 8.2.3. Modulus of Subgrade R·~action
0.4R exerts a controlling intluence on piJe behavior, so
that investigations to determine kh should be most thorough Determinatíon of the modulus of subgrade reaction is gener·
in thís area. In addítion, seasonal variations in moísture ally carrü:!d out by one 'of the following methods:
content may affect the upper part of the soíl profile and
hence influence the pile behavior. l. Full-scale latcral-loadíng te;t on apile.
3. Use of analytical results for a constant kh with depth 2. Plate-loacling tests.
may Jead to underestimates of moment and deflection by 3. Empírica! correlations with other soil properties.
a factor of 2.
The most direct means of using pile-loading tests is to
A further analysis of piJe behavior in a layered system instrument the pile so that the soil pressures and pile
has been made by Reddy and Valsangkar (1968), who deflections along the píle can be measured directly: This
LOAD-DEFLECTION PREDICTION FOR LAT:::RALLY LOADED PILES 173
2·0 2.0 , - - - - r - - - : - - - - r - - - - - - , r - - - - - .
Values of 1¡. 1 /k
~ó
el !,/)
o
_)
o ' - o
L
E
1·~¡
3: e:
¡.... :::>
L L
~ ~ Hl
L L
o
o ...,
.....
u u
J'Li'
el
e"'
u u
e o·~s
"'-" "' _:::J
~¡
6
to a piJe. Terzaghi (1955) consídered that for clays, the
<JVJ 5 modulus of subgrade reaction ís essentially the same both
o LE
_¡ horizontally and vertically and is independent of depth, and
' o
o- he suggested the following conservative relationshíp for
3: e
1-.::J
k¡¡:
L L
o o
..........
u u
J'J' (8.42)
whe're
FIGURE 8. 7 Effect of laye red soil on def1ection-inf1uence factor Typical values ofksl for overconsolidated clays, suggest-
(after Davisson and Gil!, 196 3).
1 pM ed by Terzaghi, are shown in Table 8.3.
174 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
TAIÍLE 8.3 VALUES OF ksl TONS/FT' FOR SQUARE TABLE 8.4 TYPJCAL VALUES OF nh FOR COHESIVE SOILS
PLATES, 1 x 1FT, ON OVERCONSOllDATED CLAYa
nh
Consístency of Clay Stiff Very Stiff Hard Soil Type (lb/in.') Reference
UnJrained shear Soft N/C clay 0.6-12.7 Reese and Matlock, 1956
0.5·1 1-2 2 1.0-2.0
strenph eu ton/ft' Davisson and Prakash, 1963
Range for ks1 S0-100 10()..200 200 N/C organic clay 0.4-1.0 Peck and Davisson, 1962
0.4-3.0 Davisson, 1970
Propcsed values of
75 100 3'üo Peat 0.2 Davisson, 1970
a 0.1·0.4 Wilson and Hilts, 1967
After Terzaghi (1 955).
Loess 2940 Bowles, 1968
4
Using a value of E 50 equal to 50 to 200 times the undrained After Terzaghi, 1955.
shear strength Cu (Skempton, 19 51),
Davisson (1970) suggests a more conservative value of Because of the nonlinearity of observed horizontal load-
deflection curves for laterally-loaded piles, resulting frorn
local yielding of the soillong befare ultima te failure occurs,
67 cu/d (8.46)
the overall modulus of sub grade reaction depends largely on
the deflection of the pile, or the applied load leveL An
For softer cohesive soils, ít is usually assumed that kh example of the variation of nh with deflection fo·r piles in
increa~es linearly with depth, that is, kh nh • zjd. Typical sand is shown in Fig. 8.9, where the value of nh, expressed
values ofnh for such soíls are shown in Table 8.4. as a fraction of n¡¡ for a ctimensionless pile displacement
LOAD-DEFLECTION PREDICT!ON FOR LA TERALL Y LOADED PlLES 175
1.5 pile and a plle group. On the basis of these expressions, the
r following values of kerrfkh are suggested for normal spacings:
1 o Tests by Alizadeh and
Davisson 119701
~ e:. Tests by Awad and
Two pites: kerrlkh 0.5.
a Petrasovits ( 1968)
Three or four piles: kerdkh
Five or more piles: kerr/kh
0.33.
0.25.
~
1.0
\ Repeated loading causes sorne deterioratíon of the soil
~ resistance, effectively reduc:ng kh. Davisson (1970) states
that the net effect is that the deflectíon observed under
\ first applícation of a load is essentially dqubled if the load
~
.~~-
noO-Wi'
0\ C;.
is cycled 50 times or more. Moments are also: increased
and occur over an increased depth of embedment. Re-
peated loading has the effect of redncing kh to approxi-
\ o
u mately 30% of the value app:icable to initialloading.
0.5 ~
o'- C;.
The combination of group effects and repeated loading
can reduce ketf to a value a:. low as l 0% of that applícable
~ to initialloading of an isolated pile (Prakash, 1962).
o o'ó'--- o
--
C;.
-
deflectfort p, and depth z:
p (8.48)
where
following equation:
where
p o (8.49)
FIGURE 8.10 Concept of p
t p curves.
(e) The deflection, p 50 , at one half the ultimate soil equal to the increase in radim. of the pressurerneter and is
resistance is calculated as obtained from the volume c:hange measurernents. Using
· this approach, Frydrnan et al. were able to reproduce with
Ps<J (8.51) quite good accuracy the measured deflection profiles of
two prestressed-concrete test piles. Bague!in et al (I 9n)
(d) Firrally, points describing the p-p curve are computed also discuss the use of pres~:uremeter test results to obtain
from t he following relationship: p-p curves and describe a nurnber of reasonably successful
applications of this technique to instrumented piles.
P/Pu (8.52)
L
A unified approach for constructing p-p curves for piles
'in clay has been developed by Sullivan et al (1979).
P,
kt alternative and perhaps more generally applicable
approach to deterrnining p-p curves was described by
Frydman et al. ( 197 5), who rnade use of pressureme ter n
test results. The p-p curves were obtained by assumlng that if
Pn
the sarne pressure is applied to the soil by the pressure- n +1
1
is eqt.al to the ratio of their diameters, or widths. The FIGURE 8.11 Floating pile. S tre>ses acting on (a) pilc; (b) so1l
lateral movement of the soil next to the pressuremet¿r is adjacent to pile.
178 LCAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
{ 1 d [ '' } (8.55)
o
sP r Es fsHP
o
where {sP f, {P} are the n + 1 column vectors of horizontal
soil •.ti>placement and horizontal loading between soil and
pile. (The stress. between pile and soil caused by externa!
loads on the piJe is p/2 compression on one side arid p/2 ten-
sion on the othtr si de.) Us] is the n + 1 by n + 1 matrix of
mil-displacement-influence factors.
Elements Iij of [IsJ are evaluated by integration over a ()
TABLE 8,,; TYPICAL VALUES OF PILE FLEXIBILITY FACTORKR FOR VARIOUS SOILS
with
o o o 00 .... 1-4 6-41
1 for 1 <i < n + 1
0.5 forj = l,n + 1 o o o o o .... o 1 -4 52
(8.59) The horizontal-equilibrium equation is identical with that
for the free-head pile (Eq. 8.58), but the moment equation,
must be _altered to take account of the 11xing moment at
where the píle head. This equatíon then becomes
180 LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES
wíth where
Solutions obtained from the above analyses are given Free-head: moment at head = applied momentM
in 1he followíng sections (8.3.2 and 8.3.3). For these Fixed-head: head rotation = O
solmíons, 21 elements were used to divide the pile. This Pinned-tip: tip displacement = O
number was generally found to gíve results of adcquate tip moment,Mt = O
accuracy, cxcept for very slender or flexible pi! es, in which Fíxed-tip: tip displacement O
case. dcflections and rotations may be underestímated. tip rotation O
Evangelista and Viggiani (1976) and Poulos and Adlér
(1978) have examined the accuracy of such solutions and The boundary condition> may be expressed in finit!:·
conclude that greater accuracy and economy may be difference form and the fint and last rows of [D 2 ] and
achieved by formulating the beam equatíon for piJe bending {C} may be determined. The inner rows are identical wíth
In t ~rms of unequal spacings between adjacent no des. those of [D] and [D 1 ] for the floatíng pile.
This leads to different matrices [D] and [D¡J in Eqs. The soil displacements at all elements along the pile may
(8.56) and (8.60). Smaller e1ements can be used near the be written as
top of the pile, where displacement, pressure, and moment
gradients are steep, and larger elements used along the d
{sP} · = [/s J{p} (8.63)
lower part of the pile.
An alternative formulation or" the analysis, using a finite
element discretization of the pile, is described by Poulos where
and Adler (1978). Non-uniform pile sections are readily-
handled with this analysis. {.~p} = the n + lcolumn vector of soíl displacements
The extensíon of the analyses to cover the case of (15 ] = the n .+1 by n + 1 m a trix of soíl-displacement·
yield of the soil is described la ter in this section. Modi· int1uence factors
fications for battered piles are discussed in Chapter 9.
In evaluating the e!ements of Us] , allowance should be
made for the effect of the rigid base in reducing soil
8.3.1 2 SOCKETED PILES
movements, and a convenient approximate means of
The assumptions made regarding pile and soil behavior are makíng this allowance is to introduce a fictitious "mirror
similar to those made in relation to floating piles, but image" of the pile, loaded by equal and opposite horizontal
the :mil is now assumed to be underlain by a rough, rigid stresses (see Fig. 8.12). This procedure is analogous to that
bearing stratum, and a force Ht and a tip momefit Mt act used in the analysis of axially-loadéd end-bearing piles
at the tip, which is restrained from moving horizontally. (Chapter 5). The displacements at all points along the pile,
The solution for the case of purely-elastic soil behavior is resulting from both the real and írnaginary elements, are
again obtaíned by equating pile and soil displacements at the again evaluated from the e):pressions derived by Douglas
node points. The piJe displacements are again obtained by and Davis (I964).
LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PI LES 181
( For fixed-head
2
pi les, M = Mto) 2
P,
Roe k
(b)
FIGURE 8.12 Sockcted píle.
The piJe and soil dísplacements from Eqs. (8.62) and loads and moments, the entire lnad-deflection curve for the
(8.63) can be equated, and the resulting equatíons, to- pile may be obtained. Thís analysis assumes that at
gether .vith the appropriate equilibrium equations, solved elements at which the soil ha:; not yielded, the soíl dis-
for the unknown pressures and displacements as well as the placement caused by elements that have yielded is stíll
típ force H¡ and ti p moment Mf. given by elastic theory. This assumption should not in vol ve
Sohtions from the above analysis are given in Section_ serious error when only a few elements have yielded, but is
8.3.4, \V>úle the extensíon of the elastic analysis to account likely to lead to inaccuracy in the load -deflection curve as
for locd yield of the soil is described below. the ultima te loador moment is approached.
The above analysís is best carned out by reexpressíng
the equations in terms of the unknown deflections rather
8.3.1.3 .tNALYSJS FOR LOCAL YIELD OFSOIL than the unknown pressures, so that difflculties regarding
displacements at yielded elements are elimínated.
Because the elastic analysis shows that high pressures are
developed near the top of the piJe, real soils are likely to
8.3.1.4 ANALYSISOFPILE-SOJLSEPARATION NEAR SUR.E4CE
yield a1. relatively low loads, and consequently, increased
displace ments will occur. Modifications of the elastic The elastic analysis assumes that the soil behind the pile
analysis to take account of soil yield have been descríbed adheres to the piJe at all times. However, because soif has
by Spillers and Stoll (1964) and Poulos (197la). For a limited ability to take tension, ít is likely that separation
specífied load and moment, a solution is first obtained will occur near the top of the pile, where large stresses,
assumir g the soíl to be elastic, and the soil pressures thus compressive in front of thé pile and tensile behind the pile,
obtained are compared. with the specified yield (ultima te are ·developed. This separation 2.nd local yield are the main
lateral) pressures, Py, at each point (see Chapter 7 for a dis· causes of the marked nonlinea.ríty in load-detlection be-
cussion of values of Py, denoted as Pu therein). At elements havior that is observed in lateral loading tests, even at low
where the elastic pressure exceeds Py, the displacement- load levels. Douglas and Davís (1964) state that this effect
compatibility eqúation is replaced by the condition that could lead to an increase in displacements and rotations of
the pre ssure equals Py. The solution is then recycled and 100% in the extreme case, but in practica! cases involvíng
the procedure repeated until the yield pressures are stiff piles, an increase of 30 to 40% appears to be a more
nowhe1e exceeded along the pile. Then by increasing the reasonable allowance.
182 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
Approximate allowance for the effects ofseparation may the linear portion of these curves depends on the pressures
be made in the elastic analysis in the following manncr: on all the elements rather than on the pressure at one
element only. It should be noted that it would be possible
1. F:om the initial elastic analysis, the strcsses along the to use other types of p-p relationshíps in conjunction with
pi le a ·e obtaíned. Assumíng in-situ horizontal stress to be elastic theory, although it is doubtful whether the added
Ksa 1,, where av is the vertical overburden stress and Ks is a complications in such an analysis would be justilied in
coefficíent of horizontal pressure (see Chapters 2 and 3), view of the approximations ínvolved in using the Mindlin
the elements a t whích the resultíng stress a t the back of the equations.
pile, K,a 1 · 0.5 p¡, is negative (Le., tensile) are determíned
(the factor of 0.5 arises because half the total force on a
pile element is compressive at the front ofthe element and 8.3.2 Solutíons for Floating PiJe in Uniform Soil
half is tensile at the back of the pile ).
2. Assuming the soíl to have zero tensile strength (if 8.3.2.1 D!SPLACEME'lvT AND RCTA TION
justifiable, a certain small tensile strength could be consider-
ed), separation is assumed to occur at these elements and the Solutions have been obtained for the case of a free-head
displaccmcnts caused by these elements are recalculated, floating pile, loaded by a horizontal force H acting at an
assummg these displacements to be twice the values given eccentricity e above the ground line. A soil having a uni-
by thc Mindlin equations. The factor of 2 is correct for form modulus Es and limiting pressure Py is considered;
loading in an infinite mass, and at the soil surface (pro- such an idealization is generally considered to be appti-
vided 1 hat l's 0.5). cable to piles in overconsolidated clay. In the solutions,
3. A ncw solution is obtained, and the procedure repeated
the influence of local yield of the soíl adjacent to the pile
until r,o resulting net-tensile stresses exist at the back of is taken into account, but no allowancc is made for the
the pik effects of pile-soil separation. A value of Vs of 0.5 has been
chosen; however, Vs has relat:.vely little influcnce on the
. Solminns showing thc effect of separation are discussed in solutions .
Section 8.3.2. The ground-!ine displacement, p, and rotation, e, for a
Th:s procedure may also be combined with the analysis free-head pile may be expressed as follows (Poulos, 1973):
for local yield of the soil to obtain a more accurate load-
deflectíon relatíonship to failure for a laterally-loaded (8.64)
pi le.
element i) are replaced by l/ k¡. The presence of the off-dia- by horizontal load and mornent, respective-
gonal dements of [lsl in the elastic theory is caused by the ly, for constart (1 0 ¡¡ = 1pM from the
abílity of an elastic material to transmit stress, in contrast reciproca! theorem)
toa Winkler material. Fp = yield-displacem~nt factor ratio of piJe
Th•= ebstic analysis modified for local yield is equival- displacement in elastic soil to píle displace-
cnt toa nontinear subgrade-reaction analysis in which the p-p ment in yielding soil, for constant Es and
curves are linear up to the yield pressure Py, but in which Py
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES 183
50
1d ~
~ ¿
~~OIUIZS of
r----. u
vs =
~~o
Valu<ZS of L/d
20
100 = 1
::o) 3 1~~'-'"' 1
10 ---
\l.'\.
10
~ ---~·--
'
IpH IaM
,_ ____ ---~-'···
5
1d -~- L .•. e
- - - ~-- ~-
! '~
¡--- +---
t-- - -- j-
¡~ t-- - -
2
J.
~~
10 -----
¡---
t - - j----- e· --+ -------
1
r--r-
-
f- .J. t - -
: - -- -- L--
1 .6 -5 -4 3 -2 1
10 10 10 10 10 10 10
KR
FIGURE 8.15 Valués of IeM-free-head 11oating pilc, constant soil
['
J 'o yield-rotation factor"' ratio of piJe rotation rnodulus.
in elastic soil to piJe rotation in yielding
soil, for constant Es and Py The e]astic influence factors lpH• lpM !eH), and
Py limiting soil pressure (also termed 'yield leM have been gíven by Poulos (l97la) and are reproduced
pressure' or 'ultimate lateral pressure' else- in Figs. 8.13, 8.14, and 8.15. Because of the Iimited num-
where in this book) ber of elements used, the solutíons may be somewhat
inaccurate for piles that are very slender or very t1exible,
- and may lead to underestimates -Jf deflcction and rotation.
The yield factors Fp and Fe are functions primarily of the
relative eccentricity of the loaó, efL. the pile-t1exibility
factor, KR, and the applied load levd, which may be
conveniently expressed dimensiJnlessly as H/ Hu, where
100
Hu is the ultirnate lateral-load capacity of the pile if
failure occurs by failure of th'! soil (!.e., if the piJe is
rigid). Values of Fp and Fo are shown in Figs. 8.16 and
8.17 as functions of e/L, KR, and H/llu, for L/d = 50.
Both Fp and Fe decréase (i.e. the effect of soíl yield
10 increases) as H/Hu increases ora> KR decreases. Howcver.
for relatively rigid piles (K R 10- 2 ), the effect of soil
yield is not great at ordinary wo:·king loads. An indication
of the effect of L/d for one val u e of KR and for e/ L O is
given in Fig. 8.18. Fp decreases as L/d decreases.
1 L-~~----~--~--~----~--~--__j The valÚes of Fp and Fe in Figs. 8.16 and 8.17 are for
1(1 6
10' 5 1Cí 3 10'2 10' 1 10
e/ L ~ 0-that is, mornent and load acting in the same
KR
direction. Values could also be deríved for the case where
FIGURE H.l4 Values of fpM and IeH-free-head 11oating pile, con- rnornent and load act in opposite directions~thal is,
stan! soil nodulus. e/L <O.
184 LOAD-DEFLECTION PRED!CT!Ot:' FOR LATERALL Y LOADED PILES
o (8.67)
(8.ti8)
o 2 4
' ~~ ~tt 1
v,
~-r--~~
= 0.5
z
0.2
0.4
Free-he3d pile
L_
d
o 25
Values of Lid L -', O" 0.5
1
111 0.6
IOF 0.8
1.0
Effect of Kr1
lpH
-·2 o 2 4 5· e 10
o ,--,----.--,-=--r--~~
L
~
--~
0.2 d /'1oo
/
/
0.4 /
1 .____ __]___
z
10 6
10 '• 10 L KR = 0.001
0.6
1 V; = 0.5
l0 •ee-head pile
FIGU¡;:E 8.19 Intluence factor fpp-fi.xed-hcad t1oating pile, con- 1
0.8
stant s' ,j[ modulus. 1
1.0 \ (b) Effect of Lid
..- Values of lpF and FpF are plotted in Figs. 8.19 and 8.20.
For this case, the ultima te load, H 11 , is delinee! as
(8.69)
0·3 ,____:_-..,.---.....,...----,----.----. creases. Figure 8.22 also gives sorne indícatíon of the
1
1ft d1a. concrata pila subjactad o possible order of error in predicted movernents caused by
horrzontal load ~:>nly errors in estimating Es.
1
For relatively short and rígid free-head plates, Douglas
1
and Davis (1964) have pres·~nted corresponding elastic·
u.
y displacernent and rotation-ínfluence factors. These factors,
"- 0·2 whlch are shown in Fig. 8.23, :;hould be more accurate than
"'"'
.e the corresponding values for short, rigid piles in Figures
"
~
8.13, 8.14, and 8.15, because the transverse distribution of
+'
e pressure is not assurned to h~ constant across the width,
"'
E as is the case with the presem anal)'sis. For L/d > 15, the
"'u difference between the two sets of solutions is negligible.
-ºo. 0·1 By the reciprocaltheorern,/0H =fpM. The srnall discrepancy
"'
Cl
in Fig. 8.23 results frorn rninor inaccuracy in the nurnerical
.8e analysis. Beca use the piJe is idealized as a thin strip, these
o
N solutions will tend to overestimate deflections and rotations
¡::
9 for a piJe of finite thickness. The errors involved in the
'thin-strip' idealization are di:;cussed by Randolph ( 1977),
20 40 60 80 100 and are generally less than 15%.
The elastic analysis enables determination of the
Píla Langth (ft)
relative arnounts of undrained rnovernent p¡ and final
FIGURE 8.22 lnl1uence of pile length and soil modulus on piJe dís-
rnovernent PTF of the piJe. p¡ is cakulated by using Es '=
placernen t.
25
-H~/1
IM 1 1 1
/
V -- 1
2.0 '-1
~
L
~/
f2/
1
Pr
/
1
V 0.5 ~-/
/ /
-Li
1 1
1 V!' _¿;',,H
~d7
L
v( Vi,H
v
- 1.5 --·
<~
(D --
lt¡M
/
V 1
b
.;V
lpM
-·-
o
u
""<.>e:
/
V ~
~~
~ 1.0 --
~V
L
:¡::
~
e
a----:
.. M_ ' '"" + E;l__ ·,':_
/ ~-
~
V p
M
¡:;-dl.
1
I,M +
_1 1
~
• !pM
I,H
0.5
·""
/
o
0.1 0.2 0.5 1.0 2.0 5.0
d/L
FIGl:RE 8.23 !nfluence coefficients for rotation and translation of rigíd plate (Douglas & Davis, 1964).
LOAD-DEFLECTION PRED!CTION FOR LA TERALL Y LOADED PI LES 187
1
Ground -líne moment only
Ground~líne horizontal load only
v; O 1
0.8
O.G
p
0.7 --~ ~~
0.4 L = 25
~'s 0.5
K, 0.5
OL-------~-----------------_J-------~
0.5 ·-----4----~----~----~--~----~----~ o 0.05 O. 10 0.15 0.20
1( 10-s w-• 10- 3 10~ 10
H
KR
FIGURE 8.24 Theoretica! ratio of immediate to total-final FIGURE 8.25 Example of effect uf pile-soil separa tion on hcad
displacem;ont. movemcnt.
188 L::>AD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
M,/HL M,/M
0.05 0.10 0.15 0.20 o 0.2 0.4 0.6 0.8 1.0
o
0.2 02
0.4 0.4
l L
L L
0.6 0.6
25 ,,, 0.5
0.8 0.8
Horízocnalload only, Moment only,
F'eehead Freehead
1o 1.0
(a) (b)
M,/HL
~o.s
0.2
10'"
0.4
.z
L
0.6
~ '''· 25
,., 0.5
0.8
1.0
(e! Fixed-head
virtwtl!y double the detlection of the piJe head under ing for stiffer piles. For moment-loading only, the maxi·
elasti;: condítions. Tltis effect may also he considered as mum moment always occurs at the surface and equals the
apprc.xima te! y equivalent to reducing the soil modulus by applied moment.
a factor of about 2. For a fixed-head piJe, unless the pile is very flexible
(KR < 10-5 ), the maxímum moment oecurs at the pile head
where the restraint is provided. The variation of thls
8.3.2.} MOMENTS IN PILE
restraining moment with KR wd L/d is shown in Fig. 8.28
Typical moment distributions along a pile in a purely elastic for the case of an elastic soil.
soil are shown in Fig. 8.26. The variation with KR and L/d An example of the ínfh.:ence of local yield on the
of the maximum moment in a free-head pile in an elastic moments in a pile ís shown in Fig. 8.29. At failure, the
soíl Stlbjected to horizontal load is shown in 8.27. Tiús maxímum vaiue of dímensiorless moment Mz/HL in thís
maximum moment typically occurs at a depth of between case ís about twice the elastíc value (i.e., that for H/Hu <
0.1L and 0.4L below the surface, the larger depths occurr 0.38). It should be noted that the elastic distribution of
LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADEDPILES 189
M,HL
0.16 o 0.05 0.1 0.15 0.2 o 25
o,.......;:::::--,---,---~---.---.,
0.14
0.12
··+i 1 L
¡values of d
0.2
1 1 50
0.1 --1--~·
0.4
z
1
L
t~
(Mz 1m;;¡~
C.08
HL 0.6
0.06
0.8
•),04
10 ' 10 ' 10 3
10 ? 10 ' 10 M,'M 0
KH
o 0.2 0.4 0.6 0.8 1.0
or---r---.----¡----,----~
0.2
moment is largely dependent on the pite flexibility, KR,
whereas the dístribution at failure ís independent of KR.
lt therefore fo!lows that the largest íncreases in moment 0.4
during local yielding may occur in relatively t1exible piles
L
in which the morilents under elastíc condítíons are small.
0.6
8.3.2.4 CO!v!PARISONS WlTH SUBGRADE-REACTION THEOR Y
1 25 v, = 0.5
d
- - Elastic solution
- - -Subgrade-reaction
solution
2
2
1
10- 5 10 4 10-3 10 2
10"' 10 10
KR KR
(a) (b)
104 20
~
3 ~
10
~
~
~ ~
7 ~
IoM 10
~ 1" F
~
~
~
::::----
10
2 -
10
KR
(d)
FIGURE 8.30 Comparison of elastic and subgrade-reaction solutions for displacements aPd rotations, constan! Es·
men ts and rotations if kh and Es are related as described stiffer píles and the overall agíeement is better than for
above. The discrepancy between the two theories will displacements (Jnd rotations.
be decreased if the relationshíp between kh and Es ís
varied as the relative flexibility of the pile varíes, as . 8.3.2.5 COMPARISONS WITH SOLUTIONS FROM A "SECANT
suggested by Vésic (1961) for strip foundations. l'rfODULUS" APPROACH
Comparisons between e!astic and subgrade-reaction so- In practica! predictions of lateral deflection of piles, it has
lutions for moments are shown in Fig. 8.31. The largest been customary practice to use the solutions obtained from
difn:rence again occurs for relatively flexible piles, for an analysis in ~hich no soil yield is assumed to occur,
wlúLh subgrade-reaction theory overestimates the moments. together with an appropriately-chosen secant modulus of
Ho..-. ever, the two solutions are in reasonable agreement for the soil. It is obvious that in such an approach, the secant
0.5
0.1 o.)
- - Subgrade-reaction
solution
o. 1
10 _, 10
KR KR
(a) Maxírnum rnoment in free-head pile lb) Fixing rnomPnt in fixed-head pile
FIGURE 8.31 Comparison of elastic and subgrade-reaction solutions for moments, constantEs.
pE,/pv d BEJPv
o 2 0.05 0.10 0.15
o
p
~
7
'/
z 0.5
z
I L
Deflection Rotation
- - - Solutions frorn
analysis with yíeld
- - - Solutíons from
secan! elastíc aoalysis
50
v, 0.5 2 3
KR ~ 10 J
Floating pi le,
constant E, & Pv
t 0.5
Bending rnoment
FIGURE 8.32a Comparisons between solutions with soil yield and secant-elastic solutions, H!Hu 0.4.
191
192 LOAD"DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
.......::
t C.5
_l_
L
1
1 o1!l.___.L___.L___..L__..L____J
. - - Solutions
50 f rom analysís
7
wiF1 yteld
05
KR 10 J
secant elastlc
Floating pile, an.:dysts
constan\ E, & P,.
_z 0.5
L
Sending rnoment
FIGURE 8.32b Comparisons bctween solutions with soil yield and secant-elastic solutions, H/Hu 0.6.
modulus (whether it be a modulus of subgrade reaction or not great, and it would appear that the secant-elastic
an elastic modulus) will decrease as the load on the pile approach will not be in serious error unless a significan!
i!1sreases. lt is of some interest now to examine the relation· amount of local yield occurs along the pile.
ship betwecn snluti0ns for pile behavior obtained from this
"secan1 clastic approach" with !hose obtained from the
analysi:; incorporating the effects of pile-soil yield, as
8.3.3 Solutions for Floating Pile in Soil with Linearly
descríbed above.
Increasing Modulus
Comparisons are shown in Figs. 8.32a, 8.32b and 8.32c
for a typical floating pile in a uniform soil, subjected to a
lateral force H at the ground lín~. Solutions are expressed For sands and soft normally consolidated clays solutions
in dimwsionless form, Es being the original "true" elastic for linearly increasing soil modulus with depth are required
modulus of the soil. The secant-elastic modulus for each since the assumption of one constant modulus may lead to
load leve! has been obtained by fitting the elastic and yield solutions of unacceptable inaccuracy. As previously men-
solutio.1 for ground-line deflection of the pile, and deter- tioned the Mindlin equation can still be used when the mo-
rnining the equivalen! soil modulus, as described in Section dulus is not constant although the resulting solutíons will
8.5. Figure 8.32 shows that the solutions, although general! y only be approxímate, and will tend to over-estímate
similar. do differ significan ti y as the applied force approach- groundline deflections and rotations somewhat (Banerjee
es the ultimate. The main differences are that the secant- and Davies, 1978). In the solutions described below, the
elastic approach underestimates the ground-line rotation soil modulus is assumed to increase linearly with depth,
and the maximum bending moment. However, at normal from zero at the ground surface, so that at any depth z,
working-load levels (e.g., H/Hu ""0.4), the differences are the modulus is
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PIL-ES 193
0.2 0.4
l.
L ~ 0.5
Deflection Rotation
L
50
d
v, 0.5
KA 10-·3
Floating pile,
- - Solution¡ from
constant E, & Pv analysis .vith yield
- - - Solutions from
secan! elastic analysis
Bending moment
FIGURE 8.32c Comparísons betwecn solutíons with soil yíeld and seéant-elastic solutions, H/Hu ~ 0.8.
(8.72)
where
The pile-flexibility factor is now defined as
/pH, /pM elastíc-influence factors for displacement
caused by horizontal load and moment,
KN (8.73) respectively, for linear! y varying Es, and
similar! y for lm and 18M
F'P == yield-displacement-factor == ratio of pile
The soil-yield strength, Py, is also assumed to vary linearly displacement in elastic soil to pile displace-
with depth, from zero at the surface to a val ue of pL at ment in yielding soil, for linearly increasing
the leve! of the pile ti p. and Py, ami similarly for F'e.
..
-o
1000r-----,------¡------,-------,------,------,------,
1
1
¡
: 1
¡--+.---+.-~
1
1-
1!
-+-
; 1 1
i .
--r-
1
1 : i
Volu~s of ¿
f
1 ¡
~~ 1-=+=-----+-¡------¡------;'
~-=r---t----..~t&t ___¡ T
t j 1 \ 1\ '\
i
~·- !
_,
!
i
j j
I~M, l~H
~---+--·
i
1,
1
lí'
1
1
1
-t-
1 '
. ! . 1 1
L
d
I
l
10 -6 -3
10 ...
10 -~-2 10 16 5 104 10 10 2 10~ 1 10
IV iO
Epip Epip
¡o;N = Nhl5 KN= NhL 5 ~
FIGURE 8.33 Va!ues of fpH-free-head floating pile, linearly-vary- FIGURE 8.34 Values of fpM and f~:IH-free-head floating pile, li-
ing soil modulus. nearly-varying soil modulus.
1
100000 ~
·····-----
r~ ···-·
r'\.'\ ' --
10000 ' ~\
1
,,
: '\'\'\
~
~
~
1
!
1
1000 \'\\
~
'ír\
100
1-------
' "' ~""-
Valuqs of
'\.~
~~5
~
·--~-L-. -
'
-~-
!
10 -5 ·5 -4 -3 -2 -1
10 10 10 10 10 10
K = Eplp
N Nh L5
KN =10"3
KN =10-4
KN=10" 5 -+----~~~--~~~----,
O· 01
o 0·2 0·4 0·6 0·8 1·0
H
H;:;' H/H
u
FIGURE 8.36 Yield-displacement factor F~-free-head floating FIGURE 8.37 Yield-rotation factor F'e-free-head 11oating pile, li-
pile, linearly-varying Es and Py- nearly varying Es and p y-
195
196 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
The elastic ínfluence factors lpH, etc., are given in Figs. LO .-r...----r-...,---.,.--.._:c--·-r---...,.-=:----.,
8.33, 8.34, and 8.35. The yield factors Fp and Fe are,
again, functions primarily of e/L, KN, and HfHu; they are
shown in Figs. 8.36 and 8.37. Hu, which in this case, may
again be obtained from statical considerations, is shown in
dimensionless form in Fig. 7.2 as a function of ef L.
For a fixed-head píle, the deflectior1_Ls
given JY
H 1 , 1
p = .N¡¡ L__2_ I pF 1F pF (8.76)
Values of
Illustrative E:xample of Construction of load-Deflection 0.001 ....__ __..._ _ ___.__
Curve for Single Pile o 0.2 0.4 0.6 0.8 10
_fi
H"
The application of the theory is illustrated in the following
example. The case considered is a free-head steel-tube piJe, FIGURE 8.39 Yield-displacement ractor F~p-fixed-head floating
piJe, linearly-varying Es and Py·
40-cm O.D. and 2-cm wall thickness, situated in a rnedium-
dense sand. The píle is 14-m long, embedded 10m in the
sand, and loaded at the top -that is, at an eccentricity of
4 rn abovc the ground surface. The sand is assumed to have
strength parameters and q/ = 34°, a saturated unít
wcight of j3LX. L0 3 kg,lcm 3 , anda Young's modulus that
varíes lincarly with depth. From Terzaghi's correlation
(Table 8 5), the coefficient of subgrade reaction, n 11 , is Taking
0.5 kg/cm 3 and the rate of increase of Young's modulus
with d~pth, N h, will be taken as equal to nh.
Th~ relationship hetween applíed horizontal load and
ground-line deflectíon will be computed. For the pile, then,
1 :
:........ ---- e--- ---··-~
·--- ··-····
~
p H '
- 2 ' 1 pF
Nh~
100 ....... ·--+·
"'
~-----
f-
""""" ~~ 1
!
i
----
r-
·-·
'-~~~ L~ 1_
""~
i
1 1
10
! ~ ~
""'~
f-
! -
---···---
~
e-
1 ' ----·-
1
1 -~
!
FIGURE 8.38 Values of l'pp-fixed-head floating pile, linearly· varying soil modulus.
LOAD-DEFLECTION PREDICTION FOR LA TERALL Y LOAD EO PILES 197
4.78 X 10 10 kg/cm 2
1
= 0.5!8H/F p cm (Hin tonnes) (8.77)
= Eplp 4.78 X 1010 The ultima te load, Hu, for failure of the soil will now be
KN 9.56 X 1o-s
NhLs o.5 ·x 1000 5 computed. From Fig. 7 .2,
Now,
0.169
L/d 1000/40 25
ejL 4/10 0.4 In this case, Py is the yield pressure halfway along the
embedded part of the piJe. It will be assumed, as suggested
From Figs. 8.33 and 8.34, by Broms, that Py = 3pp = 3Kpa~
32.6 tnnnes
04
__¡
I
....
o
E
::> O. 2 1------+---
015}--------+
10
The subsequent calculatíons for the load-deflection curve are Ljd and of the pile t1exibility factor, KN. These moments
tabulated in Table 8 .. are greater than for a uníform Es with depth (Fig. 8.27),
and occur typically at depths e f OJL to 0.45L below the
surface, the larger depths being for stiff piles.
TABLE 1!.7
The fixing moment at the head of a fixed-head piJe is
plotted as a function of KN in Fig. 8.41. Comparison with
H H Fp p (cm)
(Fig. 8.36}
Fig. 8.28 shows that larger moments are again developed
(tonnes} Hu (Eq. 8. 77)
for a linearly-increasing soíl modulus.
5 0.152 0.76 3.4
10 0.304 0.44 11.8
8.3.3.3 COMPARJSONS WITH SUBt;RADE-REACTJON THEOR Y
15 0.456 0.305 25.5
20 0.608 0.23 45.0 In relating elastic and subgrade-reactiOn theories, Nh and
25 0.760 0.18 72.0 nh ha ve been equated. For the displacement factors 1pf!,
30 0.912 0.155 100.3 comparisons shown in Fig. 8.42 reveal somewhat closer
agreement than with the ca:;e of constan t modulus.
Similar comparisons are found for the other influence
8.3.3.2 MOMENTS JN FILE
factors. The subgracle-reaction solution does not directly
For free-head piles, the maximum moment caused by take account of the effect of Lid, but in fact corresponds
horízon 'al load only ís shown ín Fig. 8.40 as a function of relatively dosel y to the elastic solutions for L/d 25.
_j
T
'-
;;;
10
- - Elastic solut':on
8.3.4.1 DISPLACEMElVTS AND ROTATIONS
- - - Subgrade-reaction
solution For a socketed pile in a uniform soil, subjected to a hori-
zontal force H at an eccentridty e above the surface, the
displacement and rotation at the ground line may be
calculated from the same exp:éessions as for a floating pile
(Eqs. 8.64 and 8.65, or Eqs. 8.66 and 8.67); but now,
different elastic-ínfluence fa,;tors, lpH. 1pM· Irm, and
leM, are used. These factors are plotted in Figs. 8.43,
100 8.44, and 8.45 for both a pinned tip and a fixed tip. The
tip boundary condition has· virtually no effect on these
factors unless the pi! e is relatively stiff (K R > 10- 2 ).
For smaller. values of KR, tht: displacemeht- and rotatíon-
influence factors are almost identícal with those for a
10L------~------J-------~------~
floating pile. Sirrúlarly, for a given load H, the yield
10-6 10 100 factors Fp and F8 for a floating pile may be applied to a
socketed pile if K R is less than 10-2 . F or larger values of
KR, the effects of local yield at normal working loads are
FIGURE 8.42 Comparison between clastic and subgrade-reaction very small, and Fp and F 0 may be taken as unity.
solutions -free-head pi le.
100
-·
f- -·
f- ! -·
Values of .b -·
~ d ':
10
-
- ~00
-:::::::::~5
1
f=------------
f-
--- :-.....-
~: ~"-..
5-- ~ ~
----
1 -·
f- ¡ -.... ~
1 1 .... .;.;,~
1 f--
'"r\, -
1=----
f- ~-J -
f----·
-· '\ -·
~
1-- -
1
0.1
!-
- - Pinned tip
- - - Fixed tip
1
1
-
\ ',_
-
f-
!-----------·
f-
\
f- 1 -
1 1
!i
0.01 1 i 1
10-6 10
KA =
E, L4
FIGURE 8.43 Influence factors IpH for free-head socketed piles in uniform soil.
..,
o
o
___ , __ f---- -~ ~
10 ~ ' i-----::
,,
r-- ~
lpM
e,
r--
-----
" R::-...~ J,M
00
lnH
1-
~~ -
t- ~~
--··· . -
··-·-
t-
t- 1--- '""\ 10
..... - -~-----
f--- Pinned
\
t1p
-~
-------
1·· - r--- fíxed
-~--
\
O. ---- ~·---¡-- ·- -- ·- ·~
=----·-----
~··
e-
--·--·-- i·-
! - · - ---
':.
1- -
'---··
1-
- ~
1
0.0 1.
0.1~-----L------~----~------~----~------~----~
10-6 1 o· 5 10-4 10-3 10 w· 10
KR E, L4 KR E, L4
FIGURE 8.44 Influence factors and 1&H for free-head socketed piles in FIGURE 8.45 lnl1uence factors IeM for free-head socketed piles
uniform soíl. ín uniform soil.
1000
~ Values of ~ 1 ···-
100 e--
1~ --
10
_..::::::::-,
~
~-50
100 ! ¡--
...
" ""'"-
"" ··---···---- i
------- --------::::
---~~ 10
""'~
5 f-·
-
- i
r--....
·-·
,~r-=- ~~ .......
~
-
-- Values of ~ -
~~'
~~~
- r-... 100 .
~
~
······-
25
~~ .,, 5
pH 10 - -- -
~
-----= - ~-
."
'~ -
I¡¡F 1-
,________ ¡---
--- ~\
~
~'
-
-------
-
!------------
·-·-
~ ,_ --
~
----
- i
----
s~s--
-
--- -- ·---- \
j_ \
\! \ r-·
1-----·
--- t----
·----
__\
\ -
0.1 f-.-------- e-- -
----- -- \- 1------- 1----- ------
-- 1 l\ \ r
----~-
-
------ 1--- \
Pinned
f------- ------- 1--\--.J; r-
- \
- --- Fixed
\ '-
J
\
0.1 \
0.01 1 10" 6 10- S 1 o 4 10' 3 10' 2 10-1 10
lO IU IU lU lU IU
KA EE_I_¡;_ KN ~ Nh L S
E,L'
FIGURE 8.46 Int1uence factor for fixed-head socketed piles in FIGURE 8.4 7 Values of /~H-free-head socketed piJe, línearly vary-
uniform soil. ingEs-
-
w
o
202 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
10,000
t=-----r-
l=--
1 1
1
-
-
~ 1
-
~
-
-~
~~
1000 ¡...-
i
"" -=-
"~
1--
-~-··~~-
"~'\
1-- -
'
~
- -
···-
100
~·
-
"r\.'-
--
-
1'\.~ -
-
---~ ... -----
-
-
~~ 1 Values of
100
~ -
10
1~ r- 25
5
-
e-
----··
\- " -
-----
:.
'1\\ -
-
-·
- -
\\
- \ -
¡----- \ -
f-- \ -
f-----
¡
--
f--
i-·
-- 1 Fixed tip
1
·--
'\ -
o1 1 \
6
10 10
KN -S,_!:,_
- -
Nn L,
FIGURE 8.48 Values of fpM and Iim-free-head socketed píles, linearly varying Es·
For a fixed-head socketed pile, the groundline disp!ace- An example of the effect of pile length, end conditions,
ment i> given by Eq. (8.68), and the e!astic influence and soil modulus on pile-head movement is shown in Fig.
factor lpF is plotted in Fig. 8.46. 8.51 for a 1-ft-diamet~r concrete pi!e. The soil is elastic and
For the case of pi!es in a soil with a linearly increasing has a uniform modulus Es with depth. As the soil modulus
modults with depth, the displacement and rotation is decreases, the píle movement increases until the limiting
given by Eqs. (8.74), (8.75), and (8.76), as for floating displacement for an unsupported pile is reached; this
piles. The elastic-influence factors (lpF. etc.) are plotted in limiting value is finite for all cases except the free-head,
Figs. 8.47 to 8.50. As with the case of a soil with uniform pinned-tip case. For the piJe considered, the tip boundary
modulus, the. yield factors Fp, F(J, and FpF for floating condition only affects movement if tlie soil has a modulus
piles may be applied to socketed ¡,1les at the same load H, less than about 40 lb/in. 2 -that is, an extremely soft soil.
provided that the flexibility factor KN is less than about The increase in the effective length of the piJe with decreas-
10-'2_ ing modulus is shown in Fig. 8.5lb.
100,000
~\ '\ 1
1
10,000 ,---M
1- \.\.\
-=\"
r-
1-
~~
1
'
-- - - - -
1000
-
-
1-
---··
' \~
~
~
'\
1-
-·
·-
'¡
-
---···
100 r-
--
'- '' t.. '
--
e-
· · - 1--- ~ -·-----···
~ ~'.....
: L
- Values of d
'"'
\ ~255
=--
1
10
- ! \ --
- ! \
- \
\
r--
\'
\
1------
'-
- \
- 1 \
Pínned típ
- - - Fíxed típ
\
1 \
1 1
\
\
0.1 1 \
10-6 10
1000 f--
! -
f-- -
r-
f.- -
'~ i
1
""'~"
- -
-
~ -25
-
-
~~
10 - -
~: 5'~~ 0.. -
'\~
-
f--
~~ -¡
1- '\~~
~
1-
\
~----'-
~ '\ -
1- \ -
1-
-- \ \ -
1- \ -
'--
\ 1\
\ \ \
0.1
-e- \
..• \
-
~---
_1
tip
\
\
-~
- - - - Fixed típ -
...
\
1
\\
0.01
10" 6 10" 5 10-4 10" 3 10-2 10-1 10
_E~
KN
Nn L5
A summary of the moments developed in a socketed pile For L/d 1O and L/d = !00, and an elas tic soil having
in a uniform elastic soil is given in. Figs. 8.52, 8.53, and uniform Es with depth, the horizontal force developed at
8.54 for L/d = 10 and L/d = 100. These figures show, as a the pile tip is shown in Fig. 8.55 for free-head piles and in
function of KR, the moment at the head of a fixed-head Fig. 8.56 for fixed-head piles. 1his force is generally very
pile, at the tip of a fixed-tip piJe, and the maximum moment small for KR < 10-3 , but in creases .rapidly in the range
in a free-head piJe subjected to horizontal load only. As KR 10-3 to 10-1 , á.nd generally is a maximum value for a
with displacements, the tip boundary condition only stiff piJe (KR ~ 1)--the exception is moment-loading on
influences the moment if KR is greater than about 10""'2. free-head, fixed-típ piles.
LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES 205
0.1
~
¡¡
E
<U
:> Legend
o
E 0.01
Cu:ve 1 Free~head, pinned-típ
2 F ree~head, fíxed-t ip
3 Fixed-head, pinned-tip - - - Fixed-head, pi>ned-tip
4 Fixed-head, fíxed-tip
A Upper limit for curves 2 and 3 1 1
0.001
8 Upper limit for curve 4
10
4 w-3 10- 2 10
KR
FIGURE 8.52 Fixing moment at pile head-socketed pite in uní-
Soil modulus, E, lbíin. 2 form soil.
(,r) Effect of soil mod"lus and end condítions
1.0
.;;,~""
'/
0.2¡-·
E, 100 lbiín. 2 0.75
~<U " ~ 1
r: 1 ft día. concrete pi le
u
e: ! subjected ta horizontal load M,
0.5
~ only HL
..,e F1xed-head, fixed-tip
~:> 0.1
o
E 0.25
;¡;
e: E 500 lbíín 2
;¡ '
.§
I
E, = 1500 lb/ín 2 0~----L----M~~~;_L-----L-----L---~
10 10-2 1o~' 10
o KR
o 20 40 60 80 100 (,,¡ Resulting from applíed horizontal load
Pile length (fri
!b) Effect of pile '.ength .Or---~--~----r---~---.--~
FIGURE 8.51 lnfluence of soil modulus, pile length, and end con- 1 l " -:,.-:-"""
O 8 1---- Free-head, ,,.. /
ditions on horizontal movement of socketed pile in uniform soil.
06¡..--~+=
o4 ~
1
1
~ :---~--
1/ l
d
1
/,
Jj
= 100-!'rf-lO--t-----1
8.3.4.4 TIP ROTAT!ON OF PINNED TIP PILES
H,
Ho
QL---~~--~----~----~----~----~
10~ 5 10 4 10~~] 10~ 2 10- 1 10
102 10 1
10
KR
KR
(a) Free~head, pinned-tip
(a) Homon::al Load Only
HL H,L
M
o~~~I__~--~--L_~
10~ s 10-4 10-o 10_, w· 1 10
KA
(b) Free-head, fixed-tip
FIGURE 8.54 Maximum momcnt in free-head soclceted piles (hori- FIGURE 8.55 Tip force-frec-head soclceted piles in uniform soil.
zontal load only)~uniform soil.
(8.79)
3 ----+
10
I<R
FIGURE 8.59 Influence factor 1 8F for laterally-loaded fLxed-head,
pinned-tip socketed pile in uníform soiL
H¡
8¡ = - -2 • f IJ H + L' 1 3 • ! 8·"
,u (8.80)
Eb L C Lb 'C
where
for
KR --EL 4
b e
A reasonable criterion for judgíng whether or not the tip "' 0.46 ft = 5.5 in.
rotation Br is satisfactorily small, is to compare it with the
tip rotation 81 of the corresponding pinned-tip piJe From Fig. 8.55, the horiwntal Ji¡, developed
(Eq. 8.78 or Eq. 8.79). For satisfactory performance as a at the piJe típ may be deterrnined.
fixed tip, should be only a small fractíon of 81 (e.g.,
Frorn Fig. 8.55a: H¡/H 0.50
although thís figure will generally depend on engíneer-
From Fig. 8.55b: H¡L/M 1.46
ing judgement in relation to the case being examined). If
8¡is found to be unacceptably large, the embedment depth, :. H¡ 0.50 X 50 + !.46 X
, should be increased until e¡ is acceptably small. 20
A~ mentioned previously, the above procedure will only 32.3 kips
be necessary for stíff piles (KR ~ 10-2 ) since only then
(M¡ = O)
does socketing the tip influence pile behavior. An example
illustrating the abo ve procedure ís described below.
Now, assuming that tlús force has to be resísted by
embedment of the piJe ínto the rock (i.e., neglecting
lllustmtit·e Example adhesion between the piJe base and the rock), it is found
The case considered ís that of a 20-ft-long free-head con- that for a factor of safety of 3 and assuming the ultirnate
crete pier, 3 ft in diameter, loaded by a horizontal force pressure of the roe k near the st..rface lo be 2cb, the requíred
of SO kips applied at an eccentricity of 2 ft above the soil embeddment depth is only 0.5 in.
surface. The pier is situated in a 20-ft !ayer of soft alluvial Finally, considering the típ rotatíon of the pinned-
silt, having an average c11 of 2.5 lb/in 2 , undrained Young's típ piJe (frorn Figs. 8.57 and 8.58),
modu.us, of 100 lb/in 2 (14.4 kip/ft 2 ), and underlain
by sandstone. The average properties of the sandstone near T 0 ¡.¡ 2.4
the rock surfacE' are Eb = 1.5 X 10 6 lb/in. 2 , Cb = 5000
TeM 2.4
lb/in.'
In this example, an examinatíon will be made of the
Therefore, from Eq.
relative merits of having a pinned tip ~o the pier and of
havin~. a fixed tip, socketed into the underlying sandstone. 50 X 2.4 2.4 X 100
Tiprotation, e, +
Considering the pile-flexibility factor, KR, it is found 14.4 X 20 2
14.4 X 20 3
that fp 82,700 in 4 , and assuming Ep = 3 X 10 6 lb/in 2 ,
0.0229 radians
8.4.1 lntroduction
H1- = -O.;t X 50 + 0.30 X iOO = -36.0 kips
20
M¡ 0.80 X 50 X 20 + 0.75 X 100 = 875 kip ft. Many published methods of analyzing the behavior of piJe
groups subjected to horizontal load and moment make use
of the theory of subgrade reaction, and have the advantage
Assumíng again a factor of safety of 2 applied to the that groups containing battered piles and subjected to
load and moment and an ultimate rock pressure of 2c, ít ís various types of loading can be readily considered. How-
found by statical consíderations that the required embedded ever, be cause the subgrade-reaction m o del is not con-
depth of the pile into the sandstone ís 1.25 ft = 15 in. tinuous, the effects of interaction between pil~s cannot
An estímate must now be made for the actual rotation properly be considered. The consideration of the soil asan
of this "fíxed" tip to determine whether it is sufficiently
elastic material provides a convenient means of examining
small for the tip to be effectively fíxed. For the embedded
group effects for laterally loaded piles.
portian of the tip,
The followíng analysis has been descríbed by Poulos
1 .2 5/3 0.42 (1971 b ). It parallels the analysis of axially-loaded piles,
6
in that interaction between two identical piles is examined
3 X 10 X 82700 first, and the analysis ís then extended to general pile
3.28
1.5 X l 0 6
X l 54 groups. A simple approach is subsequently described for
utilizing the solutíons for single piles to obtain approxi-
For thís value of KR, the típ is rigid, and hertce use may be mate load-deflection curves t::¡ failure for pile grcmps.
made of the solutions of Douglas and Da vis ( 1964) for the
rotation of rigid plates. Using these solutions, ít is found
that
8.4.2 Elastic Analysis of lntE:raction between Two Piles
Itw 0.92
Two identical, equally-loaded pites are considered, each
0.53 pile being divided into a number of elements, as for the
-(36.0 X 0.92) single pile. The center-to-center pilc-spacing is s, and the
:. rotation of "fíxed" tip
216,000 X l angle between the line joining the piJe centers and the
direction of loading is {3, termed the departure angle (see
875 X 1.53
+ Fig. 8.60). While elastíc conditions prevai! within the soil,
216,000 X 1.25 3
· the horizontal displacements of the soil and pile at each
3.08 X 10-3 radians element may be equated, and together with the relevant
equilibrium equations, solved for the unknown pressures.
This is about 14% of the típ rotation of the pinned-tip In the analysis that follows, the only interaction effect that
pile, so that the embedment of the tip 15 in. into the sand- is considered is the horizontal movcment of one pile thM
stone is unlikely to be sufficient for the fixed-tip condítion results from loading on another pile, the loading and
to be achlev~~d. lncreasing the embedment depth to 24 in. movement being in the same horizontal direction. A more
reduces the rotation to 7.5 X 10-4 radíans, or about 3.3% general analysis along the lines of the present analysis would
of the rotatíon of the pínned-tip piJe, which should be
be possible but would not be of frequent practical import-
satisfactory.
ance.
The soil deflections along pile 1 may be expressed as
The above example indicates the advantage that may be
gained by socketing the tip "to obtain. fixed-tip conditions, d
if the ·pile is relatively stiff. In socketing the píle 24 in. Es (¡[ + 2/] {p f (8.81)
a
into rock rather than 0.5 in., reduction in displacement
at the ground line from 5.5 in. to 1.05 in. is obtained. where
However, it must be emphasized again that -for piles that
ªre more flexible and have a value of KR of less than about
1o-2, no benefit is achieved at working loads by attempting (¡! + 2/] the (n + 1) by (n + 1) matrix of influend~
to "fix". the pile ti p. factors 19, where 1Iq and 2Iq are the
210 LOA,D-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
~M
H
T p,
n
j --
::::: P,
n + 1 P¡ + 1
Píle 1 Pi le 2
n ~ 1 n +1
influence factors for horizontal displace- obtained by assuming the uniform pressures on each element
ment at i caused by stress on element j of piJe 2 to be replaced by an equivaknt point-load acting
of píle 1 and pile 2, respectively at the center of the element. This procedure is justified in
view of the resulting simplificati:m in evaluation of 2 /;j and
The influence factors 1/ij are obtained as for the single is unlikely to be seriously inaccurate except for extremely
pile (Section 8.3.1). Values of 2 / 11 are most conveniently close spacings.
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PlLES 211
--
0.3
6.1
Departure angle,
FIGURE 8.61 Typical variation of interaction factor with depar-
ture angle.
1.0 1
v, 0.5
0.9
¡~·1 I.L
Values of d 1
-13
- - - ¡3 = 90°
0.8
0.7
.\\\ ~
2~
~
\
\\ 1 KR "'10 1
0.6 1\10
"oH \ '~
'\ " ~~ ~
apH 0.5
\ ,~0""" ~ !'...
0.4 --25 ~
r-;:: ..........
...... _
' ~o
1 '.....""' -
...... t'-..._ ~ \
0.3 ,,
0.2
' ...... ' ~.;: ~ 1\
' ...... ~
0.1 ~ ~\
-- c---. _::::::__,_ r--....,
2 3 4
o
o 4 5
~
2 3
S 0.15 0.1 0.05 o ~ 0.2 0.15 0.1 0.05 o
d
º- d
sJ.
FIGURE 8.63 "pH for KR "' 10-3 • FIGURE 8.65 apH for KR = 10.
se en that the variation of CXpF with ~ is sufficien tly el ose to factors is similarly close to linear. For convenience, a linear
linear to be considered as such for practica! problems, (a variation of all values of ex with {i will be subsequently
more accurate assumption is that CXpF varíes linearly with assumed, so that only values of ex for ~ = Oo and 90° then
sin 2 P; however, this refinement is probably unwarranted need be computed.
for practica! problems ). The variation of other interaction Interaction factors for displacement and rotation of
1.0
0.9
v, 05
0.8
0.7 0.7
apH
0.6
,
I
u
0.6
o.:~ e 0.5
"'
"
d
~
0.4
0.4
0.3
0.2
0.1
o
3 4 5 o 3 4 5
S 0.2 0.15 0.1 0.05 o 0.2 O. 15 O. 1 0.05 o
d
,<! _e!
FIGt.:RE 8.64 rY.pH for KR 10-1 FIGURE 8.66 "pM and aeH for KR"' 10"'.
LOAD-DEFLECTION PRED!CT!ON FOR LATERALLY LOADED P!LES 213
1.0
1-¡-·
1.0
0.9 0.9
0.8
0.7
¡-_ 0.8
0.7
0.6 J 0.6
_¡1
I
l:l'
~
""'e:; 0.5 o?J 0.5
:;
0.3 0.3
0.2 0.2
0.1 0.1
I. 3 4 5 00 2 3 4 5
S 0.2 0.15 0.1 0.05 o S 0.2 0.15 0.1 0.05 o
d
Q. d el
FIGURE 8.67 aplH and a.eH for KR" 10-'. FIGURE 8.69 a.pM and a.eH for KR lO.
the pile at the ground surface are .shown in Figs. 8.62 to The followíng characteristics of behavior may be seen:
8.77, for values of L/d of 10, 25, and 100. Four values of
KR ranging betwecn 10-5 and lO are considered. In all
cases, v., = 0.5, but since Vs has relatively little influence on l. Al1 a values decrease with increasing spacing and are
the interaction factors, the values for Vs 0.5 can be used greater for (3 = 0° than for (3 90° .
wíth little error foi all values of Vs. 2. All a values increase with increasingL/d.
1.0 1-,,--,---,
1.0 r----r,--.---,----,----,-----,--
v, 0.5 ¡
0.9 o.9 r---t--+---+--+-+--- o ' : - ·-·
fl 90' 1
0.8 o.8 t---t--+---+--
1
1
0.7 0.71---f--....---+-·+··········+---+-~lf--·-·+--··
i '
0.6 0 ·6 r--r--,---¡-,--t·¡ -t--r--t===±=:;:j-· !
-,;
e! ~\ ~ 1
! 1KR 10
5
J
<>o M 0.51---t--+---+--+---t-- -r-'====t-i=~-1----l
o?J
:;
0.5
~-r, j
ijq
2 3 4 5
2 3 4 5
S 0.2 0.15 0.1 0.05 o S 0.2 0.15 0.1 0.05 o
d d
.1 Q.
FIGURE 8.68 «pM and a.eH for KR = 10·1 . FIGURE 8.70 a.eM for KR = w-s.
214 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED P!LES
1.0 ,---,---,---.,--,--·--,---r--.,---,----.
0.91--lt-----+---+--t---+----+·----f·--···-+-~·······~
2 4 5
2 3 4 5
S 0.2 0.15 0.1 0.05 o S
d
0.2 0.15 0.1 0.5 o
d <!
<!
FIGURE 8.73 e>:aM for KR 10.
3. All a valucs generally increase with increasing pile- 5, For a free-head pile, the deflection-interaction factors
stiffnEss factor KR. are greater than the corresponcling rotation factors.
4. For a free-head pile, the ínteraction factors for moment 6. For horizontal loading only, values of O'pp are greater
are less than those for horizontalloading. than the corresponding values for a free-head pile, apH·
2 3 4 5
~ 0.2 0.15 0.1 0.05 o 2 3 4 5
d o
d
S 0.2 0.15 0.1 0.5
d
S
<!
FIGURE 8.72 eteM for KR = ¡o·•. FIGURE8.74 o:pFforKR= 10·'.
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES 215
1.0 1.0
0.9 0.9
0.8 0.8
0.7 0.7
0.6 0.6
0.4 0.4
0.3 0.3
0.2 0.2
0.1
0.1
o
o o 2 3 4 5
o 7 3 4 5
0.2 0.15 0.1 005 o S 0.2 0.15 0.1 0.05 o
~ d
d
r! <!
FIGURE 8.75 o:pF for KR = 10-'. FIGURE 8.77 a¡)F for KR 10.
.
The influence of a linearly increasing modulus with constant modulus, although the effect is small for P 90°.
depth on the interaction between two piles in a typical For practica! problems, the values of o: for cúnstant modulus
case is :shown in Fig. 8.78. The interaction factor tends to in Figs. 8.62 to ·8.77 may thus be used, assumíng that KR =
be less for the linearly increasing modulus than for the KN; their use \Vill generally overestimate interaction and
wíll thus be conservative.
1.0
1.0
11 le 25
0.9 -···-- !---- d
v, 0.5
v, J.5
0.8 1 - · -
' KR 1o '3 KN
- ;:~
oo
0.7 --- \
Values of d L
-- 90"
0.8 Fixed-head pi les
!
~ L'--.
1
90"
0.6 jKR 'l 10
\~' ~ ~~
0.6 ·t·-
apF 0.5
0.4
', '~~~ ~t---. 100........ ........
O'pf
0.3
.......~
~' '-\
....... 10
...... .........
0.2
......
0.1
~'-..:'r,\\
o 1'~
o 2 3 4 5 0 o~~-L--~--~~----5~-------L------~
S 0.2 0.15 0.1 0.05 o S 0.2 0.1 o
d d
<! <!
FIGURE 8.76 o.pF for KR Hí'. FIGURE 8.78 lnfluence of modulus distribution on interaction.
216 LOAD-DEFLECTION PREDICTJON FOR LATERALL Y LOADED PI LES
8.4A Elastic Analysis of General Pile Groups pinned to the pí!e cap. It may also be applied to fixed-
head piles by using the appropriate unit-reference displace-
An extension of the analysis for two píles to the case of a ment and interaction facto1s. Where moment loading is
four-pile group has revealed that the principie of super- applíed to the group, axial forces wíll be developed in the
position, as described in Section 6.2.3 for axially-loaded píles and thus consideratíon of both axial and lateral
groups, also applies to the laterally-loaded group. lt interaction is necessary. Thi!; more general analysis is dis-
then-fore appears reasonable to extend the use of the super- cussed in Chapter 9.
position principie to the analysis of the displacement and Whíle dírect consíderation ís given only to calculating
rota1ion of any general pile group subjected to lateral load the displacements at the ground surface, the movement at
and ~11oment. the top of a group loaded above the ground surface may
Consideration may then be given to the calculatíon of readily be evaluated. To the calculaiéd surface displacement
lateral dísplacements and rotatíons at the ground surface is added the addítional displacement caused by the rotation
for the following types of groups: e at the ground surface and the elastic deflection of the
pile at the poínt of load applkation.
1. A free-head group in whích ea eh pi le displaces equally. The group displacement rnay be conveniently expressed
2. A. free-head group in which an equal (or known) hori- in terms o(a dísplacement ratio Rp, whích is the ratio of
zontal load and/or moment acts on each piJe of the group. the group displacement to the dísplacement of a single
3. A fíxed-head group in whích ea eh pi le displaces equally. píle carryíng the same average load or moment as a piJe
in the group, and is analogous to the settlement ratio Rs
For cxample, for a group of n free-head piles subjected for axially loaded groups. Alternatively, the displacement
to h::Jrizontal load only, the displacement of a pile k in the may be expressed as a group reduction factor RR, defined
group is, hy superpositíon, as the ratio of the group dísplacement to the displacement
of a single píle carryíng the !;ame total load or moment as
the group, and is analogous to the group reduction factor
Pk (8.84)
Re for axially loaded groups.
RR is calculated as follows:
where
PG
(8.86)
PH t11c umt reference displacement, that ís, the
Hcfi
dísp!acement of a single free-head píle under where pis the appropriate u;1it-refercnce displacement and
unít horizontal load PG is the group displacement. Whíle elastíc conditions pre-
H1 the load on pile j vail in the soil, RR and Rp ar'~ related simply as follows:
C(pHki thc va!ue of o:pH for two piles, correspondíng
to the spacing between piles k and j and the (8.87)
angle ~ between the direction of loading and
the line joining the centers of piles k and }. where n is the number of piles in the group.
In practica! problems, R 0 is the more use fu! quantity;
lf the total load on the group ís H G, then but in examining the behavior of various groups theoreti-
n_ cally, the use of RR has sorne advantage, since as with
He~ ~H¡ (8.85) RG, RR always lies within ;:he range l to 1/n.
}~i Various values of RR. rray be determined, dependirg
on the type of loading, pilehead condition, and whether
In thc case of equal oisplacements, the n equatíons for deflection or rotation is considered. These values will be
pile dísplacements from Eq. (8.84) and the equilibríum denoted as follows:
equation (8.85) may be solved for the unknown loads and
the group displacement. RRpH group·reductíon factor for deflection caused
In the case of e qua! loads in all piles, Hj = HG/n, and by horizontal load.
the :lisplacement of each piJe may therefore be calculated RRpM group-reduction factor for deflection caused
directly from (8.84). · by applied moment.
The analysis described above applies to cases involving RReH = group-reduction factor for rotation caused by
horizontal loadings on a group of piles having their heads horizontal load{= RRpM)
LOAD-DEFLECTION PREDlCTION FOR LATERALLY LOADED PlLES 217
.~ = 25
d
6 8 10 2 4 6 8 10
S S
d d
FIGURE 8.79 lntluence ofpile stiffness on RRpF· L
FIGURE 8.81 lntluenee of- on R
d
RRe.41 = group-reduction factor for rotation caused
is shown in Fig. 8.79 for twl ·alues of KR and for a fixed-
by applied moment.
head group. At any spacing, . is considerably greater
RRr>F = group-reduction factor for fixed-head piJe.
for the stiffer píles. For a group of free-head piles,RRpll is
found to be smaller than RRpF for fLxed-head piles (Fig.
8.80). Also,RRpF increases a' L/d increases (Fig. 8.81).
8.4.5 Elastic Solutions for Square Groups
For a given total load, the displacement of a group
decreases as the number of p.Jes in the group increases. lf,
For , 32 , 4 2 , and groups in a uniform soil, the variat-
however, the displacement of the groups is plotted against
ion of the group-reduction factor RRpF with piJe spacing
the total group breadth, it is found that the value of the
group reduction factor RRpF is almost independent of the
number of piles in the group. Typical plots of RRpF versus
group breadth are shown in Figs. 8.82 and 8.83. With the
exception of the four-pile ar.d nine-pile groups, for which
RRpF tends to limiting values of 0.25 and 0.11, respective-
ly, at relatively small breadths, the points lie closely on a
single curve. Only for large b.:eadths do the points for indi,
vi dual groups tend to diverge from the common curve, as
RRpF tends to the limiting value of 1/n, where n ís the
number of piles in the group. The dependence of RRpF on
breadth rather than number of piles in the group parallels
the similar dependence on breadth-only of axially-loaded
groups.
Figure 8.84 shows the ratio p¡/ PTF of immediate to
total,final movement for piles with L/d "' 25 in an ideal
elastic two-phase soiL This ratio depends primarily on the
drained Poisson's ratio, v~ and is almos! independent of
factors such as the number of piles in the group, KR, and
S
the piJe spacíng. Even for the extreme case of v 's = O,
d PdPTF-= 0.72, and, for more practica! values of v's, is of
FIGURE 8.80 Intluence of head fixity on R R. the order of 0.8 to 0.9. in other words, the major part of
218 LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PI LES
B
d
FIGURE 8.82 vs. group breadth- K R 10·'
the movement (excluding creep) of a laterally-loaded pi! e 4. The nonuniformíty of load distribution generally be-
group cccurs immediately on loading. comes more pronounced as KR and L¡'d increase.
For groups situated in a uniform soil and in which al!
pi! es d:.sp1ace equally, typícal distributions of horizontal
load w1thín 3 2 and 4 2 groups are shown in Fígs. 8.85 to Illustrative Example
8.88. These figures show that:
The problem show11 in Fig. 8.89 involves the calculation
of the distribution of horizon cal load and the horizontal
l. The outer piles carry the greatest load and the center displacement at the ground JinE: of a .six-píle group of 1-ft-
piles the least. diameter c<;mcrete pilcs situatcd in a uniform medíum
2. The load distribution becomes more uniform as spacing clay. lt is assumed that the top of each piJe is rigídly
increases. attached to a massive pile cap, so that the top of each píle
3. The relative maxímum load in the group increases as is fixed and cach displacés equally. The value of Hs
the number of piles in the group increases. shown in Fig. 8.90 is the value (assumed constant with
Ji
d
FIGllRE 8.83 RRpF vs. group breadth- K R O. l.
LOAD-DEFLECTION PREDICTION FOR LA TERALL Y LOADED PI LES 219
1.0
0.9
/
V 2.5
/ •1
0.8
/ 2.0
1
•
25
V
~
d
.Pi_ v, 0.5
PTF
// KR = 1o··s
0.7
! 1.5
' .........
......... 3
1 ' ....., _
;
~;;~ -
/ "
FIGURE 8.84 Ratio of írnrnediate to total-final displacernent for / ''
fixed-head groups. 1
/
10
~
d
,. 3 2 Group FIGURE 8.86 Ty·pical horizontal load distributions in fíxed-head
2' . )
2 1 pilc group.
•••
2.0 ..
3 4 3
~ ~ ~
~
__ __,_ HG
6 is H 1 , and that in piles 2 and S is H 2 • For pi!es 1, 3, 4,
and 6, the displacement at the ground line ís given by
25
Pi!es ·"'d 0.5
1. ~~ e
10" 5
H
H;)v
where
1 .o
_;;·::-:_-:~-----~ interaction factor for deflection at píle 1
2 / /--
caused by load on pile 3, and similarly for
0.5
/' / other o: values.
/. 4//
/ displacement of a single fixed-head piJe under
/ 1
/
/ unit load
o ,/
-0.5~----~-------L------~'------~'~----~•
A similar expression may be written for the tlisplacerrient
p 2 at píles 2 and 5.
For the condítion of e qua! displacement of all pi! es,
P1 p 2 =p. Also, frow equilibrium,
o 2 4 6 8 10
~
d He = 4H 1 + 2H 2
FIGURE 8.85 Typical horizontal load distributíons in fixed-head
pile group. where He is the total applied load.
220 LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED P!LES
1 1 1 1
¡----,·---- ·----,--
4 2 Group 4 2 3roup
1 2 2 1 1 2 2 1
•~ •4 •4 •3 •3 4• •4 •3
:;·.s t-
3 • 4 •
4
• 3~HG
- •3 4• •4 •3 -;--------;- H G
•1 •2 •2 •1 •1 •2 •2 •1
1
•• •• d
25 •- •-•-•l
100
2.0 t- v, = 0.5 - d
----- .L
Pi les KA 0.1 10
el
1.5
-
1 .O --
/ ----------
3- - ......
-------
2 ------- ..-.:=
o.s .. !-------------- -_
------
Ot- -
4.7
PF 31.4 X 1o· 3 in./kip
300 O. 500 TABLE 8.8
r-~~~
1· D<lrivqd trom dqflqction
1 1 1 mqasUrqm<z-nts.
Dqrívqd trom slopq
m<lasurqmqnts.
Dtzrivtld from or
KN Vs Nh rqlationshlps
Nh
kips fcu ft
/./¡ossiwz Cap
r
-~----'---1-----HG =100k1pS
1 Soil
Surfoccz
FIGURE 8.90b :v!ethod of backfíguring modulus from load test.
Illustratíve Éxample
(8.89)
The problem considered will be one of the model tests
where reported by Oteo (1972), which is discussed in Section 8.6.
Aluminium piles 8 mm in diameter wíth an embedded
== total load on group length of 220 mm were tested, with an eccentricity of
group-reductiorr factor for deflection caused loading of 55 mm above the ground surface. The group
by horizontal load consídered was a nine-pile square group with a center-to-
group-reduction factor for deflection caúsed center spacing of four diameters, in sand having an inítial
by moment densíty of 1.80 t/m 3 . The deflection of the group at the ·
yield-displacement factor for a single píle, for poínt of load application will be.• calculated for various
loads.
H¡¡G ,H11, being given by Eq. (8.88). From a single-piJe test, the value of Nh was found to be
Hu 11 ur 4.0 kg/cm 3 (39.23 MN/m 3 ), md hence tite dimensíonless
singh>pile elastic-ínfluence factors (see Eq. pile-flexibility factor, KN = Epl pJ,VhL 5 , ís found to be
6.8X 10-4.
The deflection of the group at the point of load
Similar expressions may be derived for the rotatíon and application is given by Eq. (8.90), modified for linearly
del1ection of a pile with varying Es and for fixed-head piles. varying E5 :
If the de!1ection of the group at the point of load
application, PGa, is required, the additional deflection He
PGa = - 4 - [(L
2 1
RRpH pH + eLRRpMfpM )/F~p
c:msed by rotation is added to the ground-line deflectíon. L N¡¡
Neglecting the deflection of the piles caused by bendíng of
the freestanding portian, Pea is given by + (eLRReHf~H + e 2 RReMfÓM)(F~]
0.322
when:
RRpM RReH 0.203
RReM = 0.149
RRoM group-reduction factor for rotation caused by
moment,
Iau.loM síngle-píle elastic in!1uence factors
Sucstitution of the above values into the above equation
and the symbols are as defined for Eq. (8.89). gives
A similar expression is obtained for soil wíth linearly
varying Es, and in Eq. (8.90), Es is replaced by NhL; IPH•
1pM, and lo M are replaced by I'pH, /~M, and I'eM; and FP,
Pea He [(oAJs) + e:.:¡)]
Fe ar•; replaced by F'p, Fe.
T'1e use of Eq. (8.89) or Eq. (8.90) er 1bles the overall where PGa is in mm and He is in kgf.
load-deflection behavior of the group to b~ calculated, and From the single-pile test, the ultimate lateral load is
an example illustrating the applicatíon of this method is 2.7 kgf (at a deflection of 0.5 pile-diameters). Referring to
gíven below. Fig. 7 .22, the group lateral-efficiency factor 7JL is 0.60.
F :Kht and Koch ( 1973) developed a similar type of The estimated reduced ultimate lateral-load of a pile in
the group Hur is therefore 0.60 X 2.7 = 1.62 kgf, and the
apprcach to the calculation of load-defle~tion behavior of
group ultimate lateral-load capacity ís 9 X L62 = 14.58 kgf.
groups, by combining a nonlinear subgrade-reaction ana!ysís Values of F' P and F' e m ay n()W be determ;n.ed frorn Figs.
for a single pile with the elastic analys1s for píle interaction. 8.36 and 8.37 for efL = 55/220 = 0.25 :::;.d KN = 6.8 X
This approach also enables nonlinear load-deflection relat- 10-4. The calculatíons are tahulated in Table 8.9, and the
ionships to be obtained and has been applied to the pre- load-deflection curve thus derived is plotted in Fig. 8.100,
dictÍ(•n of deflections of offshore piJe groups. toget:ter with the measured curve.
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES 223
TABLE 3.9 CALCCLATION OF LOAD-DEFLECTION ClJRVE mising agreement between the calculated and measured
FOR GROUP behavior.
Full-scale loading tests are probably the most saiis-
He He Pea factory means of determining the soil modulus, since such
(kgt) nHur Fp Fe (mm) .factors as the effects of installation and pile-soil separation
0.1 l. O 1.0 0.26
are taken into account automatically and reflected in the
1.46
2.92 0.2 0.90 J. O 0.55 backfigured modulí. There appears to be two possible
5.84 0.4 0.68 0.88 1.39 means of interpreting pile-load results:
8.76 0.6 0.55 0.72 2.57
11.66 0.8 0.40 0.56 4.63
l. To use the ground-!ine deflection at the working load
13.14 0.9 0.36 0.52 5. 74
to backfigure a secant value of soil modulus, which may be
used with elastic theory to predict deflections at the
working load (ignoring the effects of local yield and soil-
pile separation).
8.5 D.ETERMINATION OF SOIL MODULUS 2. To use the linear portian of the load-deflection curve to
backfigure a tangent value of soil modulus, which may then
A number of methods may be employed l.o estímate the be used with the theory (including the effects of local
Young's mod ulus of the soil for use in the theoretical yield) to predict the load-deflection curve to failure.
solutions given in the preceding sections. Among these are:
The latter procedure would appear to be preferable, as a
L Laboratory tests in which the stress path of typical more relevant value of the pile-flexibility factor may be
elements of soil along the piJe are simulated. obtained. However, in sorne cases, the use of the first
2. Plate-bearing tests, preferably on vertical plates, at procedure may be more expedient if piles similar to thc
various depths. test píle are to be used in the foundation, and as shown in
3. Pressuremeter tests. Section 8.3 .2, the use of a secan t modulus wi th purcly
4. The use of full-scale loading tests to backfigure the ebstic theory should give results of adequate accuracy al
modulus. normal working loads. In either case, the principie of
S. Empírica! correlations \Víth other properties. interpretation of the load test is thc same. Considering
first the case of Es constar.t witli. depth and a free-head
Little cvidence is available at present to indicate whether piJe, the ground-line deflect:;on for an elastíc soil is, from
the first approach yields satisfactory values of modulus, Eq. (8.66)
although the simulation of the correct stress path caused H l'vf
by loading of the piJe is easier in this case than for an
p -- . f H + - -2 . fp'f (8.91)
EsL P EsL "
axially-loaded pile. Nevertheless, the problems associated
with the simulation of the effects of installation of the pile By substituting the mea;;ured values of p, H, and M in
remain. this equation, Es may be expressed as a functíon off pH ::mJ
Similarly, little information is available on whether the fpM· For various values O'c KR, lpfi and fpM rilay be
use of values of modulus determined from plate-loading obtained from the theoretical curves in Figs. 8.13 and 8.14,
tests at various depths giv0;, satisfactory load-deflection and hence a relationship between Es and Kn is obtáined
predictions for piles, although the use of such data gave from the definition KR = Epfp/EsL 4 . Símultaneous solut-
reasonable predictions in one series of full-scale tests, as ion ( e.g., by graphical means) of these two reiationships
described in Section 8.6. gives the values of Es and KR for the plle. Because the
The use of pressuremeter tests by Frydman et aL (1975) theoretícal-influence factors are insensitive to the valuc of
and Baguelin et al (1978) has already been mentioned in Poisson's ratio of soil, vs, tho~ value chosen for this quantity
Section 8.2.4, in relation to the determination of p-p is of secondary importanc,~. The above procedure may
curves, and such tests can a\so be interpreted, in terms of similarly be applied to fixed-head piles or to tht case of a
elastic theory, to give values of Young's modulus at various linearly increasing. modulus with depth, for which relation-
depths. The pile-soíl yield pressure, Py, may also be esti- ships between Nh and KN would be derived.
mated from the limit pressure measured by the pressure- A more complete del1nition of the soil moduius
meter. This procedure has been applied to the pile test can be obtained íf tests on piles of dífferent proportions
reportcd by Frydman et al. and has produced very pro- are made or if the ground-line rotation as well as the
224 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
def1ectíon of a pile is measured. The case that best fits the intersection of curves 1 and 3, giving the backfigurecl
data (constan! or linearly-increasing Es) m ay then be rnodulus from the deflection measurement, gives Nh
deterrnincd. Thls procedure is analogous to the procedure for 28 k.ips/ft 3 , whlle the value backfigured from the slope
dcterrnining the subgrade-reaction modulus described by measurement (íntersection of curves 2 and 3), gives Nh
Reese and Cox (1969). 23 k.ips/ft 3 . The closer agreem~nt between these two values
Asan example of the application ofthe above procedure, of Nh indicates that the assu:nption of a linearly varying
the result s of the test on Pi! e l-B described by Alizadeh modulus is a better approximation than a constan! modulus.
( 1969) are analyzed. Thls piJe was a Class B timber pi! e, It is interesting to note that the value of the modulus of
of embedded length about 37 ft and situated in a ·soíl subgrade-reaction nh backfígured by Alizadeh from sub-
profile consisting of 4 ft of sand and grave! underlain by grade-reaction theory is about 30 kips/ft 3 , which is in
clay <md silt strata. The consistency of the clays ranged reasonable correspondence wi th HÍ~ backfigured val ues of
from soft to medium, with an average shear-strength of Nh.
about 600 lb/ft 2 • The piJe was loaded at the ground line
by jackíng against an adjacent píle (Píle 1-A). At a load Empirical Correlations
of 1O kips, for the first load cycle, a deflection of 0.80 From a number of published load-deflection measurements
in. and a slope of 0.012 radians were measured at the on full-sca1e piles, Poulos (l971a) backfigured secant values
grounj line. of at work.ing-load levels on the assumption that is
Considering, first, the case of a soíl modulus assumed constan! with depth. For cohesive soils, the values of Es so
to be constant with depth. The def1ection and rotation !J deduced varied widely, lying within the range
may be cxpresscd as in Eqs. (8.66) and (8.67). Substítut·
ing th·~ appropriate values in these expressions, one obtains (secant) Es = 15cu to 95c¡¡ (8.92)
Es = 4.05 fpH k.ipsift 2 , from the deflection measure- with an average va1ue of 40 Cu, whcre Cu undrained
ments shear-strength of el ay. The lower val ues tended to be
associated wíth very soft clays and the higher values with
and stiff clays. Banerjee and Davies (1978) backfigured value5
of Es of between 1OOcu and l80cu. All these val u es are
E.1 "' 0.6081 oH kipsi f¡2, from the slope measurements lower than values normally as>aciated with surface founda··
tions or axially loaded piles, and this can be attríbuted to
Frorn Figs. 8.13 and 8.14, values of lpiJ and loiJ may be the effects of local yielding of the soil and pile-soil separa-
Dbtained for various values of pile-stíffness factor KR. tion near the typ of the piJe (soil anisotropy may also have
Hence, two plots of versus K R may be obtained ( shown contributed slightly.) The scatter of the backfigured results
as curves 1 and 2 in Fig. 8.90a). A further relationslúp may also reflects the effects of !o.;al yielding, since more local
bcobtained frum the definitíonofKR =oEpfp/EsL 4 ,which yield may have occurred in the softer soils at working loads
upon :mbstitutíon of the appropriate quantities, gives than in stiffer clays.
In order to obtain more-satísfactory correlations, values
of the tangent modulus shotld be determined from the
initiallinear portions of the mNLSUred load-def1ection curves.
This relationship ís plotted as curve 3 in Fig. 8.90a. The Although only a limited number of tangent values of
backllgured Es from the deflection measurement is given modulus ha ve been determined and rclated to Cu, the
by t he intersection nf curves 1 and 3, while from the slope correlation appears to be more consisten! and to suggest
measurement, it ís given by the intersection -of curves 2 values of Es in the range
and 3. These values are, respectively, 44 kips/ft 2 and 26
kips/f 2 • lf the assumptíon of constan! sml modulus was (tangent) Es (8.93)
valid, these two values should have been the same.
l\'ow considering the case of a linearly varying modulus These values are about one ha.lf of those normally associat-
with depth, the def1ection and slope rnay be expressed from ed with surface foundations (D' Appolonia et al., 1971)
Eqs. (8.74) and (8.75). As in the case of constan! modulus, and may ref1ect the effects of soil anisotropy and pile-soil
two relationshlps between Nh and KN may be plot ted; separation, as descríbed in Section 8.3.2.
these u e shown as curves 1 and 2 in Fig. 8.90b. Also, K N = The use of the correlation given by Eq. (8.93) rather
Eplpf,VhL 5 , so that another Nh·versus·KN relationship rnay than Eq. (8.92) enables a more logical predictíon of the
be plotted; this is shown ·as curve 3 in Fig. 8.90b. The · load-deflection relationship, sí~1ce the effects of local yield
LOAD-DEFLECT!ON PREDICTfON FOR LATERALLY LOADED PILES 225
may be taken into account by determining the appropriate For sand, no reduction te the static value of Es appears
value of the yield-deflection factor, Fp (Eq. 8.64). to be necessary, but the static val u e of Py is multiplied by a
As indicated in the discussion of subgrade-reaction factor which increases from about 0_25 at the surface to 1.0
theory and from the example of pile-loading test interpretat- or so at a depth of 3 diameters or grea ter.
ion deSLTibed earlier, the assumption of a linear! y varying The reduction in modulus and lateral resistance is
modulus with depth may often by more satisfactory than a likely to arise from one or both of two main causes:
constant modulus, especially for soft clays and cohesionless
soils. In these cases, it is reasonable to assume that the rate (1) Build-up of excess pare pressure (mainly with clays).
of modulus increase, N¡¡, is approximately equal to the (2) Actual degradation of particles at their contacts
modulus of sub grade reaction, nh. Approximate ranges of (mainly with granular materials, particularly calcareous
values of n¡¡ are discussed in Section 8.2.3. Banerjee and sands).
Davies (1978) found that a value of N¡¡ of about 40 lbíin 3
fitted a number of experimental results. lt should be recognized that, in predicting the behavior
At Lhe present time, it has not been possible to dís- of a prototype from small-s.;ale tests, the significance of
tinguish between values of undrained moduli and drained the first cause will be affected by the dependence of pare
moduli; however, most tests were probably carried out pressure dissipation on scale a1d period of cyclic loading.
relatívely rapidly so that the correlations quoted are best
considered as being applicable to the undrained modulus.
Also there is insuff!cient data for any distinction to be made
between values of soil modulus for various installation 8.6 COMPARISONS BETWEIEN OBSERVED ANO THEOR-
methods_ ETICAL LOAD-DEFLECTION BEHA VIOR
00 2
SM R.O.:: 75"/o
Void Ratio= 48"/o
Bulk Density
2:05 gm.fcc.
2
"'
L
....()1()1
E 3
.e
....o.
()1 Cl 4·
o l l : 48 °/o
P.l 27 OJo
P 1 = 21 °/o
5 5 .
6 '---'-----''---'----J 6L-~--~~~--~
(él) Bore Log ( b) Standard Penetration (e) Sta tic cone ( d) . Dynamic Conrz
Test Tqst T¡zst
FIGURE 8.91 Soil data (tests of Mohan and Shrivastava, 1971).
12 r-·----......,..-
Plle N2
8 f-----,---+-_,c:.-+----4 8 1------,.·------+-~.,,4/F+-,.~.
Load Load
(k N) (k N}
4 1----+,..L.......~+-,...""'--+---l 4~----~~~----4----
(b)
4 8 12 16 4 8 12 16
Dczflczction (mm ) Defl¡zctíon (mm )
16r-------~~--~--~
P illl N3
12 1--------,.
Load
(k N)
4 4
(d)
4 8 12 16
o 4 8 12 16
Dllfl¡zction (mm) D12f ltZct ion (mm)
12r--------.----r---~ 12
Pihz N6
sr---.---~----+---4 8
Load Load
(k N) (k N)
4
4 8 12 16 4 8 12 16
Deflectíon (mm.) Deflection (mm)
16
Pile 1N1
12
Load __.._ Observed curve
(k N)
- - Computed curve
8
(e)
o
o 4 8 12 16
Del lect ion (mm )
FIGURE 8.92b Comparison between observed and computed load- dcflection curves (tests of M•Jhan ond Sh:ivastava, 1971}.
TABLE 8.10 SL'MMARY OF PlLESa TESTEO BY MOHAN AND For the ínstrumented pi le IN l, comparisons between
SHRIVASTAVA (1971). the observed and theoretical dístríbutions of deflection,
slope, and mornent with depth are given ín Fig. 8.93.
Pile
Again, the agreernent is reasonable, ami is at leastas good
Pile Día. Embedded Stiffness, as that reported by Mohan md Shrivastava, using Kubu's
Test (cm) Length (cm) Eplp(kg cm 2 ) (1965) modífied subgrade-reaction approach.
N1 10 200 6.22 X 10' ;:'imílar agreement has been found by Poulos (197la)
N2 10 300 6.22 X 10' betwee,' theoretical behavíor and that reported by Kerisel ·
N3 10 400 6.22 X 10' and Adam (1967) from full-scale piJe tests in clay.
N4 10 500 622X 10'
N5 3.8 525 0.316X 10'
N6 76 525 2.46 X 10'
JN1b 10 525 3.20 X lO'
Tests of Gleser ( 1953)
~ All piles are steel pipe.
Gleser (1953) report~d measurements of deflection and
lnstrumented pile. moment along a fixed-heacl piJe in sand. In order to
TABLE 8.11 CALCULATED VALUES OF Pv AND Es FROM compare the measured distributions with the theory, the
PLATE-BEARING TESTsa BY MOHAN AND SHRlVASTAVA theoretical and measured pile-head deflectíons were equated
(1971) and the soil modulus backfigured. The test was ínterpreted
in two ways: first, as the cas·~ of a pile in~ soíl havíng con-
Depth Below Dirn. of Es Py stant soíl-modulus Es wíth depth; and second, as a píle in a
Surface (c;n) Loading (kg/cm') (kg/cm') soil whose modulus jncreases with depth. The calculated
50 Vertical 121 4.2 deflection and moment profiles are compared with the
50 Horizontal 82 3.1 measured values in Fíg. 8.94. As might be expected for a
100 Vertical 220 8.0 pile in sand, the calculated profiles for linear! y varying Es
150 Vertical 35 1.8
are in much closer agreement with the measurements than
aAl! tests on 30-cm square plate. are the profiles for constan! Es.
228 LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES
. ·~"·-r~]k
5 ._______,¡_____.¡_1_ _ _
_,
1
4
\ Appliad load H • 4·90 k N is shown in Fig. 8.97. It is ínteresting to note that a con-
siderable amount of the defle,;tion occurs immediately on
application of the load, as the theory predicts for an
ideal two-phase elastíc soil (P:mlos, 1971a ). F or the case
shown, the measured ratio of immedíate to final deflectíon
(o)Banding Momants is 0.56, whereas the theory predícts a somewhat hígher
ratio of0.73.
Daflact10n (mml Rotation (,. 10' 3lrod Load-deflection predictions were made assuming a
!:\ 10 o 5 10 15 uniform Es and a constant value of Py of 9 Cu along the
pile. Two values of Es were considered, 2500 lb/in. 2
(175 kg/cm 2 ) and 1000 lb/in. 2 (70 kg/cm 2 ), correspondíng
/
to values of Es/cu of approxirr:ately 450 and 180.
1 Comparisons between the t wo predicted load-deflection
~
E
2
1
-~- --¡--- curves and the observed curve:; are shown in Fig. 8.98 for
four of the tests. Final deflectíons (measured values after 90
0 ......._ Maosurad valu12s min) are considere d. The agreement is generally satisfactory,
3 - - Comput12d valuqs and with the exception of test 1, the predicted curve for
Es 2500 lb/in. 2 agrees more closely with the observed
1
4 L----'----' curve. Thís value of Es is of the same order as the value of
( b) Dqt 1act1on Ce) Rototion about 500 to 1000 Cu generally applícable to surface
foundations (D' Appolonía et .11., 1971 )-allowing for re-
FIGURE 8.93 Cornparisons betwecn rneasured and computed beha-
vior, pile !NI (Mohan and Shrivastava, 1971).
ductíon resulting fro·m the effects of pile-soil separation, as
prevíously mentioned~and is much higher than the average
secant value of 40cu backfigúred from reportee! field tests
Model Tests of Prakash and Saran ( 196 7) (see previous sectíon). It is interesting to observe that the
A series of tests on groups of model piles in clay was difference between the predictions for Es = 1000 lb/in? and
reported by Prakash and Saran ( 1967). Tests, ínvolvíng Es 2500 Ib/in 2 is much less than the factor of 2.5
horizontal load near ground leve!, were carried out on between the Es values themselves. In fact, it is found that
free-head single píles and groups of four and nine piles, the predictions are as sensítíve to the value of Py as to the
whích m ay be considered as effectively free-headed, since. value of Es.
significan! rotatíons occurred at the piJe caps. Compari- Comparisons between mea:;ured and predicted values of
sons between the measured and theoretical ratio ·of group ultima te load Hu are shown in Fig. 8.99. Predicted values
dísplacement to single-pile displacement, for a load equal · are given for both the simple statícal theory, assumíng a
to the average load in the groups, are shown in Fíg. 8.95. constant value of Py of 9cu along the whole length of the
The average piJe-load considered is well below the ul- piJe, and the modified approach of Broms (1964a) who
limate value for the single pile. Figure 8.95 shows reason- assumes the same Py distributicn, except that Py is taken as
able agreement between theory and observation, although zero from the surface to a depth of 1.5d (see Section 7 .2).
the theoretical displacement ratio is smaller at close spacings The predicted values from bmh methods Iie withín 25%
and larger at greater spacíngs. of the m.easured values, but the simple statícal theory tends
Comparisons between the observed and predicted load- to overestímate Hu, whereas Broms's approach tends to
deflection behavíor of two of the tests are shown in Fig. underestimate Hu and therefore is more conservative.
LOAD-DEFLECTION PREDICTION FOR LATERALL Y LOADED PILES 229
---<>-- Measured
- · - Theory (variable EJ
- fheory (constant E, 1
\
z \
L L
\
\
\
\
\
\
\
\
FIGURE 8.94 Comparisons between measured and theoretical deflections and rotations. Tests ofGleser (1953).
Measured values
"' Theoretical
5 relationshíp
3.0
"'
4 - 2.5 "'
Rp ~
3 Re 20
2 - 1.5
FIGURE 8.95 Comparison between theory and Prak:ash and S~.an's results.
15 15
o, o,
~ -"
CJ ::J
I 2 I
3 group
·6 10 "'¡;- 10
_Q "' ~= 3
d
o
a. a.
:0 :0
~ ~
c:J c:J
5 5
(a)
o 2 4 6 8
o 2 4 6 8
Deflection, p (mm) Deflection, p (mm)
FIGURE 8.96 Comparisons between measured und'predicted group ¡load-def1ection curves (tests of Prakash ancl Saran, 196 7).
0·100
E
É 0·075
e
Q
u
:!. 0·050
Q/ Test 5 Applied load
o _ _ _ _ J........___
0·025
1
a 12 24 36 48 60 72
FIGURE 8.97 Typical time-def1ection curve for piJe (Druery and Ferguson, 1969).
25
L=575"
1 Test 1 1 25 r----,---...,---~ L=5.5"
d=0.25''
1 T<2St 2 d=0.25"
• Measured e=0.75" g M12osur12d e= 1 .o"
20 -- - - Es=2500psi } Cu =5 21b/sq in __ E = ~.l} Cu= 5 91b/sq 1n
-- E5 =1000psi
Theoret1cal 20 s 2500 P Th<2oret1cal ~--
.D - - Es = 1000 p:,1 ~~--
o
o
....J
15
10
--~~~
1---·
-- ~
~
-::--- ---- ~ I
D
o
15
+-) ~4-~¡--
10 1---.......,w'---~--+---c-----r-----
/o/
-·-·-~~
D _<3
~
o.
Q
./
<1:
/ 1
¡·
1
--
"
...!
FIGURE 8.98a Comparisons between measured and theoretical FIGURE 8.98b Comparisons between measured and theoretical
load-def1ection curves (tests of Druery and Ferguson, 1969). load-deflection curves (tests of Druery.and Ferguson, 1969).
230
LOAD-DEFLECTION PREDICTION FOR LATERALLY LOADED PILES 231
25r-----,-----,-----~----,--L~=~5~2~"-----,
Test 3 d=<ÚS'' A Colculot~d from Broms' opprooch
e Colculat~d from s1mpl~ statical
• M
easure
d
.}
e•l.33"
cu= 5.81b/sq. in
200
approoch
20 --- Es- 2500ps' Theoretical 1
Es= iOOOps:
1·25
1
z 150 1 - - - - - - - - j - - -
. :--
I
::J
I
10 ---··-+---+-~~±--~~---/¡-- -g
......
o
:¡
u
100
e;
u
FIGURE 8.98c Comparisons bctween measured and theoretical FIGURE 8.99 Comparisons between measured and calculated ultí-
load-defkction curves (Tests of Druery ;md Ferguson, 1969). mate loads in model pite tests (te~ts of Druery and Ferguson, 1969).
30 1 L= 5.6
Test 4 d=0.375"
Measured e= 1.0"
• E
5 2500 psi } cu= 5 91b/sq in
25 1 - - Es = 1000 psi Th<2ore~1ca1 -
! ~
20 /
·-·~
/
:o
u
o
o
..J
15
1/
1 l
/
//
--
u
~
o.
Q 10 f /
v'
l
<(
VI
1
5
FIGURE R.98d Compárisons between me¡>sured and theoretical load-defloctíon curves (tests of Druery a:1d Ferguson, 1969).
Model Pi/e Tests of Oteo (1972) 1.8 t/m 3 , comparisons between the measured and predicted
Oteo ( 1972) carried out tests on aluminium-tube piles in load-deflection curves, shown in 8.1 00, reveal el ose
san d. The piles were 8 mm in diameter, with an embedded agreement. The basis of the prediction ís discussed in the
length of 220 mm, and various initíal densities of the sand example given in Section 8.4.6. The results of a single-pile
were use d. Both single piles and pile groups were tested. For test were used to backfigure the rate of increase of Young's
nine-pile groups ·tested in sand with an initial density of modulus with depth, Nh.
10
'¡;,
=-
I "'
..
-o'
E
a.
o" (a) 3 group, ~ = 2
1
ts 2
2 3 4 5 6
PG (mm)
20
Cl
=-v
:r:
-o
E"'
a.
::l
o
ts
10
V
"/
_...,...
--
(h) 3 2• group,
.-
·~ = 4
o
o 2 3 4 5 6 8 9
PG (mm)
20
rn
=- l?
I
-o" 10
~"'
a.
"~
(.9
oo
2 4 5 6 8 9
PG (mm) - - - - Measured
- - - Predícted
FIGURE 8,100 Comparisons between measured and predicted group load-defiectíon curves (tests of Oteo, 1972).
232
GENERAL ANALYSIS OF PI LE GROUPS
9.1 INTRODUCTION The first two methods ca.n only consider interaction
between the piles through the pile cap and not interaction
In previous chapters, the behavior of vertical pile groups through the soil as well. Therefore, they assume that once
under axial loading or lateralloading has been consídered. the loads on any piJe are known, the deOections of that pile
In general, a pile group may contain battered pilAs and may may be calculated from th€:se loads alone. The third
be subjected to simultaneous axial load, lateral load, mo- method removes this limitation and allows consideratíon of
ment, and possibly, torsional load. Methods of analyzing pile interaction through the soil; the deflections of a pile
this general problem may be broken down into three cate- are therefore not only a function of the load in that pile
gories: but also of the loads in all the piles in the group.
In this chapter, the three approaches mentíoned above
l. Simple statical methods that ignore the presence of the will be described, with empha~;is being placed on the third
soil and consider the pite group as a purely structural method, that employing elastk theory. An example will be
system. presented to compare the solutlons from each approach.
2. Methods that reduce the pile group to a structural
system but that take sorne account of the effect of the soil
by determining equivalent free-standing lengths of the piles.
The theory of subgrade reaction is generally used to deter- 9.2 SIMPLE STATICAL ANALYSIS
mine these equivalent lengths. Typical of these methods are
those described by Hrennikoff (1950), Priddle (1963), Traditional design methods have relied on the consideration
Francis (1964), Kocsis (1968) and Nair et al. (1969). This of the pile group as a simple statically-determinate system,
type of approach will be termed the ':equivalent bent ignoring the effect of the soil. One such method, whích
method,'' following Kocsis ( 1968). may be employed eíther graphically or analytically, is illus-
3. A method in which the soil is assumed to. be an elastic trated in Fig. 9 .l. Considering, for simplicity, loads and
cominuum and interaction between piles can be fully batter in the x, z plane only and piles having a pinned head,
considered. the steps in this method are as follows:
233
234 GENERAL ANALYSIS OF P!LE GROUPS
The principie of this method i:; illusl rated in Fig. 9.1 for <~
planar group. The actual group, shown in Fig. 9 .la, is acted
upon by vertical and horizontal forces and a moment. The
equivalent bent, shown in Fig. 9 .la, consists of the píle cap
Force Polygon supported by fixed-end freestanding columns or cantilevers
V
of equívalent lengths Lei, Le2 and Le3, and equivalem
cross-sectional are as A el, A e2 and Ae3. There are a number
of ways of convertíng the actuallengths and cross-sectional
areas of the piles to equivalent values for the columns and
P, V,
these are discussed in 9.3 .2 below. Once the equivalen!
lengths and areas have been det~rmined, the equivalent bent
H may be analyzed by standard structural analysio techniques
FIGURE 9.1 Simple statical method for determination of group to determine the deflections, rotations, and pile loads in the
load distribution. system.
In order to simplify the structural analysis, the piJe eap
is frequently assumed to be rigid and the piles assumed to
l. Assuming each piJe to take an equal share of the verti- behave elastically. Saul ( 1968) and Re ese et al. (1970) have
cal load on the cap and assuming the vertical load in a piJe, presented matrix analyses in 1N hich the abo ve assumptions
caused by moment in the cap, to be proportional to the are made, and in the former paper, torsional loading and
distan ce x, the vertical piJe loads.are calculated as dynamic forces may also be incorporated. However, if hand
computation is contemplated, the method described by
V Mx¡ Nair et al. (1969) is more convenient. Their analytical pro-
V·=
z n + n (9.1)
cedure is as follows:
2:(x/)
j= 1 l. Through the rígid pile ca::>, arbitrary horizontal and
vertical displacements, 77 ando, andan arbitrary rotation, e,
1. If ;he solution is done graphically, the forces V and H are ímposed. Thus, axial and lateral forces and moments
are plotted on a force polygon. The vertical pile forces, V¡, will be introduced in the pile heads-these being a function
from Eq. (9.1) are then set off. e
of r¡, o, and and of the arrangement and characteristics of
3. Th1~ force polygon is then completed by drawing lines the equivalent cantilevers, which reflect pile and soil prop-
parallel to the piJe directions. The axial force, P¡, in each erties. Expressions for these forces and moments can be
pile may thus be obtained. There is then a residual horizon- determined frorn standard structural analysis, and are given
tal force, He, whích is assumed to be equally distributed be- in the original paper.
tween each pile in the group. 2. The mornents and forces in the pilé heads are added to-
4. If desired, the design of the group m ay be amended and gether in the various coordinate dírections and equated to
the pil<~ batters adjusted to give He 0-that is no horizon- the externa! applíed forces and moments. This will give
tal load in the piles, so that each píle is axial! y loaded. three equations in three unknovms, o, r¡, and (}.
GENERAL ANALYSIS OF PILE GROUPS 235
V H H
l1!
~1 r -- '{[ {a)
~M
--
'·jl (b)
V
--
'"If-H (e)
c,jf-H
<2
{d)
,:ff~H
( b) Equivoi<Znt B<Znt
FIGURE 9.2 Principie of equivalent-bent approach.
--
3. These eq uations are sol ved for 8, r¡, and 8.
4. The mcments and forces at each piJe head are then
determined using the expressions derived in step l. The Actual P1l<2 Equival<znt Cant1l<2v<2r
necessary information for the design of the group is thus
FIGURE 9.3 Equivalen! cantilevers for laterally loaded piles.
obtained.
lateralloading of piles. U~ing this approach, sorne allowance
can be made for side shear and for group effects as de-
9.3.2 Determination of Equivalent Bent scribed below.
In puhlished methods using the above approach, the equiva- a) EquivaJent Length of Piles
lent lengths of the piles are almost invariably determined The equivalent length will depend on the boundary condi-
from a subgrade-reaction analysis. The normal deflection tion at the piJe head and on the type of loading assumed to
(or rotation) of a pile subjected to normal load (or mo- act. A number of cases have been considered, as illustrated
ment) is calc:ulated and equated to the normal deflection in Fig. 9.3, and the solutions derived for the equivalent
(or rotation) of a cantilever under the same load (or mo- cantilever lengths are summarized in Table 9.1. The group
ment), the cantilever having the same moment of inertia as effect has again been taken into account approximately by
the piJe. The required equivalent length can then be deter- applying the group displacement and rotation ratios to the
mined (e.g., Francis, 1964;Kocsis, 1968;Nairetal., 1969). single-pile movement (see Section 8.4).
The equivaJ,~nt area of each cantilever is corrÍmonly The various ratios referred to in Table 9.1 are as fol-
assumed to be that which gives the same axial deformation lows:
as the actual pile when considered as a column, the effect
of side shear from the soil thus being neglected. RpH = group-displacement ratio for free-head piles sub-
The equivalent lengths and areas of the· cantilevers in jected to horizontal load, obtained by super-
the equivalent-bent method may_ also be determined by use position of values of Ct.pH·
of the solutions given in previous chapters for vertical and RpM = group-displacement ratio for free-head piles sub-
236 GENERAL ANAL YSlS OF PlLE GROUPS
jected to moment, obtained by superposition of sonable to take either the av·~rage of LeH and LeM as the
values of apM. equivalentlength, or, conservatively, adopt LeH.
Reu group-rotation ratio for free-head piles sub-
jected to horizontal load, obtained by super- TABLE 9.2 EQUIVALENT LENGTHS Lef! AND LeM (L/d =50,
position of a8 u values (Re u RpM) Vs = 0.5, single pite, constantEs)
RoM group-rotation ratio for free-head píles sub-
jected to moment, obtained by superposition of
Cl.oM values .
RpF group-displacement ratio for fixed-head piles Leu/L 0.0406 0.0818 0.157 0.293 0.551 1.123
subjected to horizontal load, obtained by super-
LeMIL 0.0251 0.0614 0.127 0.224 0.446 1.154
position of Ct.pF values.
EP "' Young's modulus of pile and can tilever modulus of pile, lp polar moment of inertia of pile sec-
A e '" equivalent arca of cantilever tion, Gs = shear modulus of soil, and d piJe diameter. For
the case of a linearly-increasing soil modulus with depth
From Eqs. (9.2) and (9.3), (Fig. 9.5), Kr = Gplp/Ncd 5 , where Nc = rate of increase
of soil shear modulus with depth. Rotation increases a'
torsional stiffness decreases and as length-to-diameter raÚo
{9.4)
(L/d) decreases. The effect of L/d becomes less as the piJe
becomes less stiff, and for very flexible piles (small values
For a pile having an unsupported length e above the of K T ), the rotation is independent of Ljd.
ground surface, the axial det1ection of this length must 'be Although no solutíons have yet been obtained for tor-
added to that for the embedded portion. The corresponding sional interaction of piles, it is likely that such interaction
expression for A e is then is small. Figures 9.4 and 9.5 provide a basis for évaluating
the response of a piJe in a group to torsion. They can ais6
be used to evaluate the results of torsional píle-load tests
(9.5)
(see Chapter 16). When incorporating torsional movements
into an equivalent-bent analysis, it appears most convenient
to determine an equivalent torsional stiffness, Gplp, of the
If an cquiv<lent length L~ is required, rather thanAe, cantilever such that the actual pile and the cantilever of'
equivalent length and area, deduced from Section 9 .3.2,
will have the same rotation. The structural analysis of thc'
(9.6) equivalent bent may then proceed.
The abo ve expressions for A e and L ~ should apply for 9.4 ELASTIC CONTINUUM ANALYSIS OF PILE
battered piles as well as vertical piles since, as shown in BEHAVIOR
Table the axial movement of apile due to axial load is
not significantly intluenced by íts inclination. The elastic analyses described in previous chapters for piJe
movemcnts under axial and lateral loads may be extended
to cover piles and pile groups subjected to combined loads.
9 .3.3 Torsional Response of Piles
r--------,--------,--------,-------~-------,--------r-----~
1 !
~ 1
~\
! 1
1 i 1
1
\
! 1000
! T1
1
'
!
~1
10
r- \ 1
r- \ ! 100
r- 1\ ! ¡
¡
1
\"' cjJ=-_l_·l
~
1 4
1--- 1 ! 1
1
" NGd
10
""
1
1 1
1
i '\
f--
-
"'r--
"'\...
1
1
!
1
¡ 'i- 1.
V•lo" "' ~ '
-
1~ r-
"~ ~
~ ~21!
--1
Values of
i
dL
...
....
1~
10
'1 o 2
1
1 ~
~
~~
~- 1 o
2
-+----L.-
0.1
r- ~ ............. 5 1
1---
1---
~~....._
~ ~10
... 10' J
~ 25
r-
T
1
i
1
1
1
"-- ¡..._
--
.0301
1 o. 4
-
!
"' ~~l "
- 1
1 1 lOo-·
~-¡¡l
1 [ 1 1
10-5 i 1 1 1 i
1 1 1 1 1 1 1 1 1 1T
1 1 1
M~ ~~-t--+~-j_~~ 1 - 1
.00312
FIGURE 9.4 Influence factor for torsional rotation of pile head- constant Cs. FIGURE 9.5 lnfluence factor for torsional rotation of piJe head-
linearly increasing G S·
(~EN ERAL ANALYSIS OF PILE GROUPS 239
QC /~~~)(\
~ "
\>''. ' \
.¡;
axial force ís considered to mobilize only shear stresses on stresses and displacements. In evaluating rhe soil displace-
the periphery and a uniform normal stress on the base of ments, the unknown force on each element is resolved into
the plle. Thc: pi k (Fig. 9 .6a) is divided into n elements of vertical and horizontal components, and vertical and hori-
egua! length and the axial displacements of the soil at the zontal displacements caused by each of these components
center of each element are evaluated and equated, the re- are calculated using Mindlin's equations. These displace-
sulting equations being solved to obtain the unknown ments are then combined to giVE' the axial displacements.
240 GDlERAL ANALYSJS OF PILE GROUPS
9.4. j .2 BA TTERED PILE SUBJECTED TO NORMAL LOAD AND ment on batter angle appears to be unaffected by the pile-
MOi'r!ENT flexibility factor, KR, and the boundary condition at the
The analysis in this case follows closely that described in piJe head.
Section 8.3. lt is assumed that only stresses normal to the
9.4.1.3 BATTERED PILE SUBJECTED TO VERTICAL AND
pile are mobilized in the soil by the applied load and
HORIZONTAL LOADS
moment, and that the plane of the batter and of the
loading are identical. The pite ís divided into elements, as The fact that the axial and no·mal displacements ofa pile
'shown in Fig. 9.6b, and the soil and pile dísplacements are are almost independent .of the pile batter means that solu-
evaluated at each element and equated, the resulting equa- tions previously obtained for vertical and horizontal dis-
tions beíng solved for the unknown normal stresses and dis- placements of vertical piles may be applied to calcula te the
placements. The soil displacements are evaluated in a axial and normal displacements of battered piles. This, in
'Similar manner to the axially-loaded piJe. turn, leads to a relatively simple method of calculating the
Typical results for battered pile-dísplacements are given horizontal and vertical displace nents of a battered piJe sub-
in Table 9 .3a for axial load and in Table 9 .3b for normal jected to vertical and horizontal loads and moments. The
load and moment. In each case, the appropriate displace- vertical and horizontalloads V and H are first resolved into
ment-coefficient is given for piles with a batter angle, l/J, of axial and normal cornponentsP and Q, as follows:
0° (a vertical piJe) and 30°. lt should be noted that the
solutions for positive and negatíve batter-angles are identi- p V cos l/J + H sin l/J (9.7)
cal.
It is signífícant that both the axial and normal displace- Q = H cos l/J - V sin l/J (9.8)
ments are almost unaffected by the batter of the pile; the
maxirnum effect for a batter angle of 30° is approximately The axial and normal displacements, Pa and p N, m ay then
only 4%. The virtual independence of the normal displace- be calculated and resolved into vertical and horizontal com-
fpa
10 1.415 1.382
25 1.860 1.859
100 2.542 2.538
·¡o 7.29 7.35 3.22 3.37 39.89 39.78 3.90 4.05 5.81 5.92 1.04 1.09
25 9.75 9.84 3.98 4.13 54.68 54.65 4.99 5.15 7.27 7.40 1.23 1.28
100 12.21 12.33 4.79 4.95 68.28 68.32 6.16 6.33 8.67 8.82 1.44 1.49
Free-head piJe: PN - 1(
Esf.
IN·Q+IM·-
P P L
l•f)
Fíxed-head piJe: PF = _L (lpF. Q)
Ej.
GENERAL ANALYSIS OF PILE GROUPS 241
ponents. To simplífy the analysis, it is assumed that lateral fhFV (Ipa- lpp)sin¡J¡cos¡J¡
loads do not influence axial displacements and that axial fhFH lpa sin 2 ¡J¡ + fpF cos 2 ¡J¡
loads do not affect lateral displacements. Tests reported by fpF normal displaceme r1t-influence factor for fixed-
Evans (1954) show that ·this assumption is conservative in head pile
that the lateral deflection of a pile subjected to axial and
lateralloads is less than that of the pile subjected to lateral Vertical displacement of fixed-head pi! e:
load only. The following expressions are then obtained for
the battered pi! e, for the case of constant E 5 with depth:
PvF (V • lvFV + H • lvFH) {9.12b)
Vertical displacement:
where
Pv (9.9)
r· c,J
B¡.
adjacent loaded píle, and which are functions of the piJe
spacing, relative stiffness, and geometry; and for horizontal 4¡, B¡, C¡¡ (9.! 3)
loads, of the dire..::tíon of loading. By summation of the
{: J {:; J
Ae Be Ce
interaction factors for each pite in a group resulting from al!
the other piles in the group, the displacement of each pile
may be written in terms of the loads on each piJe in the
group. 1' i ' + +-
,,~3
A similar approach can be adopted for groups con·
taini ng battered piles. The first case considered will be that
Se
of :. group in which all the piles are battered in the same
L
plane and on which the horizontal load acts in the same
plane. In the interaction analysis for vertical pites, it is
implicit that normal forces produce only normal deflections
and that axial forces produce only axial deflections. Thus,
l
Two battH<Id p¡l,zs Vert1ca1 p11es ot
considering two battered piles i and j in a group, it is EqU!vclent Spoc1ng
assumed that an axial load on piJe j will cause a deflectíon
of pile i that is in the axial direction of piJe i, and equal to FIGURE 9.8 EquivalE' ni spacing of battered piles.
GENERAL ANALYSIS OF PlLE GROUPS 243
where the co<:fficients of the sub-matrices are as follows: the analysis could be extended to take these into account.
For example, the influence of the cap on vertical move-
Av¡¡ PalCt.if cos 1/1¡ cos 1/1¡ + PNQlo.pHif sin'#¡ sin'#¡ ments may be allowed for as described in Chapter 10.
Bv¡¡ if;¡ sin if;¡ PNQio.pHij sin'#¡ cos 1/1¡
'p 0 ¡o.¡¡ cos Groups in which píles are battered in different direc-
Cv¡¡ PNMio.pMif sin 1/1¡ tions can be treated approxirnately by resolving the hori-
Ah¡¡ Pa¡o.¡¡sinlj;¡cos>J;¡ PNQ!O.pH¡¡coslj;¡sin>J;¡ zontal load into two components and calculating the in-line
Bhij Palo.¡¡ sin V;¡ sin >J; ¡ + PNQ 1O.pHif eos '#¡ cos '#¡ horizontal displacements caused by each component, using
(~ij PNMJO.pMif cos '#¡ as the length of a pile its projected length in tl}e plane of
A o¡¡ -fJN¡cxoH¡¡ sin if;¡ loading. The resultant horizontal displacement can then be
B 0 ~· ONlaoH¡¡ cos '#¡ calculated from these displacernent components.
C 0 ¡¡ OMJCXOMtj A more complete analysis ¡hich avoids mapy of the
Pal axial deflection of single pile caused by unit assumptions made above has been described by Banerjee
axial load and Driscoll (1976). However, because it does not employ
PNQI normal deflection of single piJe' caused by unit interaction factors, a complet¡, re-analysis is necessary for
normal load each group configuration, whereás the present analysis
PNM! normal deflection of single pile caused by unit allows any group configuration to be rapidly analyzed once
moment values of the interaction factors and single-píle responses
ONI rotation of single pile caused by unit normal have been evaluated
load
O,m rotation of single pi! e caused by unit moment
9.4.3 Parametric Studies of Pile Groups
The above-menti<med unit deflectíons and rotations
may be calculated from the theoretical relationships in 9.4.3.1 EFFECTOF P!LE STIFFNESS ANO BATTER ANGLE
Chapters 5 and 8 if values of the soil moduli can be esti-
mated, or if pile-load test data are available, from the pile The effects of pile stiffness and batter angle on the defle<:-
deflections at the working loads. The interaction factors o. tion and load distribution withín apile groupare illustrated
may be found in Figs. 6.2 to 6.5 while values of the inter- in Figs. 9.! O to 9.14 for a typical case 9f a group of síx
action factors aPH• o.pM. and o.oM are gíven in Figs. 8.62 to piles, as shown in Fig. 9.9. The pile cap is assumed to be
8.77. rigid, and rigidly attached to piles in an elastic soil whose
The submatrices Av, By, and so on, are of order n X n, modulus is constant with depth. Three values of pile-stiff-
while the vectors V, Pv, and so on are of order n. Equation ness factor K are considered: K 100 ( corresponding to
(9.13) together with the three equations expressing vertical concrete pi! es in a stiff soil), K = 1000 ( corresponding to
and horizontal load-equilibrium and moment-equilibrium,
may be solved to obtain the 3n + 3 unknown vertical and
horizontal \oads, moments, displacements, and rotations, V
M
for the desired boundary conditions at the pile heads. _.,_H
A number of cases may be considered, including
L
and the vertical displacement of a piJe is related to its
position in the group and the rotation.
2. Piles pinned toa rigid pile-cap, which is similar tocase l
except that the pile head moments are zero.
1
3. Piles attached toa massive cap, in which case horizontal
and vertical displacements are equal but all pile-head rota-
tions are zero.
4. Piles attached to a relatively flexible piJe cap, so that
each pile is subjected to known loads and moments.
No account is taken in the above analysis of the hori- FIGURE 9.9 Pile group consídered in parametríc study of spacing
zontal shear and rotational resistance of the cap, although and batter angle.
244 G• NERAL ANALYSIS OF PILE GROUPS
- - K 100 where
- - - K 1000
K= 10000
0.1 , - - - - - - - - - , 02.---------, V = vertical lo1d on group
H = horizontal load on group
M = moment on group
I,v, ivH, etc. dimensionless deflection and rotation
----..__ influence coefficients evaluated from
--~ the analysis
o~..-- _ __.__ _ __J
o'-------'-----'
o 7.5 15 o 7.5 15 For the symmetrical group considered he re, Ih v = ! 8 v =
\t Batter angle, 1/1° ivH fvM O, or in other words, the horizontal de0ection
and rotation ca u sed by unit vertical load are zero. Sin ce Pv,
.005.---------, .005.--------, Ph and 8 define the rigid body displacement of the cap, to
--
which the píles are assumed to rigidly attached, the dis·
- placements and rotation of any individual pile are readily
calcula te d.
------ For a center-to-center spacing of 3d at the pile cap, Fig.
9.10 shows the effect on the deflection and rotation co-
o.______.__[_ __, efficients of the relative piJe stiffness and the angle of
1
0 Lo-----=-7'=.5---,-'5
o 7.5 15 batter of the outer piles. The coefficients are not greatly
Batter angle, ,¡;" affected by pile batter, but the pile-stiffness factor, K, has
a significant effect.
FJGlJRE 9.1 O Effecl .of batter angle and relative stiffness on deflec·
Corresponding solutions for the loads and moment on
tíon and ro latían coefficients: síx·píle group, L/d ~ 25; "s = 0.5;
s/d 3. the front piles (type 3) of the group are given in Figs. 9.11
and 9.12. The actual vertical load V 3 , horizontal load H 3 ,
and momen t M 3 in each pi! e ar~ given by:
concrete piles in a medium-stiff soil), and K= 10,000 (cor-
responding to concrete piles in a soft soil). For each value
VCvv + HCvu + MCv111/d (9.18)
of K, the value of pile-flexibility factor KR is related as
follows:
VCHv + HCHH + MCmf!d (9.19)
KI
(9.14) (9.20)
RA
where
where
V, H, M = the
applied loads and moment on
l moment of inertia of pile section
the group, as before
RA area ratio, defined in Eq. (5.17)
Cvv. CvH, etc. = load and moment coefficients
L pile length
Figure 9.11 shows that most load and moment coeffi-
The vertical deflection of the cen ter of the píle cap, p 1,,
cients are markedly infll¡enced by pile batter and pile-stiff-
the horizontal deflection, Ph, and the rotatíon, 8, for a
ness factor K. However, the vertical piJe load caused by
general loading system are given as follows:
vertical load on the ·group and the horizontal pile load
caused by horizontal load on the group ( coefficients Cv v
V H M
P¡; · Ivv + E d. fvu +E d2. lvM (9.15) and CHH) are almost independent of both factors.
S S
- - K=100
---K 1000.
- - - K 10000
0.2 ::::::--,
. 0.1 . - - - - - - - - - ,
- --=-== ----=
- - --·----
0.2 ::::--, ............
'--..
Cvv
0.1-
eVH 0.1-
--- -- -- 0.05
Cv~
r=--------
r------_j
QL------'-----' o o '-------'----'
o 7.5 15 o 7.5 15 o 7.5 15
Batter angle, e
o 1
7.5 15 o 7,5 15 7.5 15
lL:,----
/,. ---- CM,..
-------- --- CMM
o~~:c:=:J 0'-------'-----'
o 7.5 15 o 7.5 15 7.5 15
Batter angle,"
FIGURE 9.11 Effect of batter angle and relative pile stiffness on load and moment coefficient for pile No. 3: síx-pile group; L/d 25;vs
0.5;s/d=3.
The effect of pile spacing on the deflection and rotation spacing between the outer piles. The pile loads and mo-
coefficients for the center of the cap of a six-pile group ments are correspondingly greater in the four-pile group.
(Fig. 9 .9) is shown in Fig. 9.13 for a batter angle of 15°.
Almost all coefficients decrease with increasing spacing, a~ 9.4.3.4 EFFECT OF PILE CONFIGURATION
would be anticipated.
In order to examine the effect of pile configuration on
group rotation and deflections, the six groups shown in Fig.
9. 4.3.3 EFFECT OF NUMBER OF PILES IN GROUP
9.15 have been anaiyzed. Group A ís the one shown in Fig.
The effect of the number of piles in the group on the de- 9.9, while group B ís the same group except that the center
flection and rotation coefficients for the center of the pile two piles are removed. The other four groups have dífferent
cap is shown in Fig. 9 .14. A four-pile group and a six-pile piles battered. In all cases, the batter angle of any battered
group are compared for K = 1000 and a pile spacing of piles is 15°.
three diameters. As would be expected, deflection and rota- The deflection coefficients for the leading piles of each
tion coefficients are greater for the four-pile group, and the group are shown in Table 9 .4. The following observations
rotatíon coefficients are most affecteó because of the closer may be made:
--K 100
K 1000
---K 10000
.2 , - - - - - - - - - , .2 r - · - - - - - - - , .1 , . - - · - - - - - · - . ,
.1-
Cvv
o 1
g
o~-----~-----J
3 6 6 9 3 6 9
2_ 2. S
d d d
~------l
.2
.010
.2 ~--====--~ ---=-
'\
,"\,.
.005 :--.
~""":J
.1-
c11v
CHH
.o .O i
3 6 9 3 6 9 6 9
S ~
d d
.2 .5 ,----------=-~-, .1 o ,---·--------,
/-
-"
"'-,
- -.25 =---------
·.1
CMv
o
-
3
""'
....... .._
, ..........
6
--- ---
9
CMH
o ¡____
3
__jl~--1
6 9
S S S
d d d
FIGURE 9.12 Effect of píle spadng on load and rnoment coefficients for pilc No. 3: six-pile group; L/d 25; vs O.S;s/d = 3; batter angle
Ji; ~ 15°.
TABLE 9.4 EFFECT OF PI LE CON FIGURA TION ON DEFLECTION AND ROTA TION COEFFICIENTSa
~
fvv
~
Coel~ A
0.0391
B
0.0432
e
0.0609
D
0.0548
E
0.0451
F
0.04 76
fvu 0.0136 0.0121 0.0262 -0.00346 0.00495 0.00569
I,M 0.00615 0.00671 0.00666 0.00463 0.00486 0.00571
Il1v o o 0.0148 -0.0148 -0.00733 -0.00677
l¡¡}{ 0.1006 0.1010 0.1093 . 0.1093 0.1026 0.1025
lh~1 0.00453 0.00403 0.00380 0.00380 0.00409 0.00416
Iov o o 0.00102 -0.00102 -0.000818 O.OOOOll
hm 0.00453 0.00403 0.00380 0.00380 0.00409 0.00416
le M 0.00205 0.00224 0.00188 0.00188 0.00189 0.00190
246
GENERAL ANAL YSIS OF PILE GROUPS 247
--K~ 100
---K=IOOO
---K 10000
\
0.1 , - - - - - - - - - - ,
.05 - .1
l..,v '
""-·---~-
lt;H A B e
------·- 6 pri<ZS 4 pii<Z5 6 pd<Zs
o 1 ()
3 6 9 3 6 9
2_ S
d d
lhM
0.005
ioH
\.
... ~
\"
' ~-....:::-.._
0.005
IoM
D E
'\ F
o
3 6
2_
-- 9 6
S
9
6 pii<ZS
spoong c1 cap
6 pJI<Zs
3d ( <Zx c<2pt
Wh(iU'Q
tor B
S= 6d)
1
d d L/d 25
K 1000
FIGURE 9.13 Effect of pile spacing on dd1ection and rotation co-
Bott<Zr angi<Z 15°
efficients: six-pile group; L/d = 25; vs = 0.5; batter ang!e ·.¡; !5°.
Group
~uan~
hty A B e D E F
0
Pv (mm) 16.8 18.5 27.4 .22.1 18.9 19.4
Ph (mm) 16.8 16.6 24.0 11.3 13.7 14.0
e 0.00436 0.00444 0.00497 0.00278 0.00312 0.00404
TAB.LE9.6 EFFECTOF BOUNDARY def1ection of the group with pinned piles is larger than in
CONDITIONS AT TOP OF PI LE: GROUP A !N FIG. 9.15 the first, but the rotation is less. The group with a massive
cap sustains the smallest movements.
Piles Rigidly
Piles Rigídly Attached to
Atlachcd but Cap Piles Pínned Massive Cap~No
Coefficient Can Rotate to Cap Top Rotation
9.5 COMPARISON OF METHODS OF PILE-GROUP
I.,v 0.0391 0.0396 0.0391 ANALYSIS
l¡¡H 0.1006 0.1338 0.0906
Ihw ··1/jH 0.00453 -0.00220 o
[IJM 0.00205 0.00345 o To compare the three method:; of analysis described in thís
chapter, two simple planar pile-groups have been analyzed.
As shown in Fig. 9.16, each ;;;roup has three piles. In the
first, al! piles are vertical, while in the second, the outer
TABLE 9.7 COMPARISON OF GROUP
piles are battered. In applying the equivalent-bent method,
W!TH VARJQljS BOUNDAR Y CONDITIONS
V 1200 kN,H ~ 400 kN,M ~ 600 k~/m the equivalen! length of each member has been taken as the
L "' 10m, d ~ 0.4 m, Es = 7000 kN/m' mean of LeH and LeM (Table 9.1). A computer program
(Harrison, 1973) was used to evaluate the solution. For the
Piles Rigídly elastic continuum analysis, the single-píle vertical and hori-
Pi1es Rígidly Attached to zontal responses have been obtained from the theoretical
Attachcd but Cap Piles Pinncd Massive Cap~No
Quantíty Can Rotate to Cap Top Rotation
solutions in Chapters 5 and 8, while the corresponding
group effects have been determined from Chapters 6 and 8.
0
Pv (mm) 16 8 17.0 16.8 The piles are assumed to be rigidly attached to a rigid
Ph (mm) 16.8 18.0 13.0 cap in both cases. The loads, moments, and deflections
(J 0.00437 0.00384 o from each method of analysis are summarízed in Table 9 .8.
a For center of cap. The m a in points of interest from this table are:
ficients for the center of the pile cap for various boundary
condítions at the junctíon of the pile top and pile cap are 1·2m 1·2m
10m 2 3
shown in Table 9.6. A larger horizontal deflection occurs if
the piles are pinned, but the presence of a massive cap ap· Q-4m
pears to have relatively little influence. The vertical move-
E,= 7000kPa
ment caused by vertical load is unaffected by the boundary 6
Ep= 21 x 10 kPa
conditions. Considering the same numerical values as used
in Table 9.5, the resulting deflections and rotation are Group A Group B
shown in Table 9.7. Under this loading system, the lateral FIGURE 9.16 Pile groups considered in comparison of rnethods.
GENERAL ANALYSIS OF PILE GROUPS 249
TABLE 9.8 COMPARISON OF elastic continuum analysis. T:1e simple statical analysis
METHODS OF GROUP ANAL YSIS assumes zero rrwment in all pile:;.
3. The equivalent-berit approach predicts a considerably
Simple Equivalen!· Elastk
larger rotation than the elastic continuum analysis and a
Staticai Bt~nt Continuum
Quantity Analysis Anaiysis Anaiysis slightly larger vertical deflection of the leadíng pile, but a
smaller horizontal deflection.
Group A V, (k/V) 75 67.2 50.5
V, {kN} 200 200.0 163.4 1t should be noted that the computed rotatíon and hori-
V, (kN) 325 332.8 386.1
H, (kN)
zontal deflection in the equivalent-bent method are sensi-
66.7 66.6 .75.9
H, (kN) 66.7 66.7 48.2 tive to the equivalent length of the piles. For example, for
H, (kN~ 66.7 66.6 75.9 Group A, if the equivalent length was taken as LeJ,f ( = 1.96
M 1 (kN m) o -6.2 -39.6 m) instead .of the mean of L,M and LeH (~ 2.24 m), the
M, (kN m) o -6.2 -23.5 vertical and horizontal deflections and rotation would be
M, (kN m) o -6.2 -39.6
16.8 mm, 6.7 mm, and 0.00521, compared with 17.5 mm,
Pv, (mm) 17.5 14.8
Ph (mm) 8.9 11.8 8.9 mm; and 0.00581 in Tabk 9.8. On the other hand, if
() 0.00581 0.00248 LeH (=2.52 m) is used, the corresponding values are 18.2
mm, 11.4 mm, and 0.00639. The latter val ue of horizontal
Group B V, (kN) 75 59.3 65.4 deflection corresponds more closely to that from the ela,:¡tic
V, (kN) 200 200.3 174.8
V, (kN) 325 329.6
continuum analysis in this case.
359 8
H, (kN) 52.0 76.7 20.3 A more detailed comparison of the computed deflec-
H, (kN) S2.0 75.5 26.3 tion and rotation under the individual components of load
H, {kN) 52.0 47.8 153.3 reveals that the vertical movements caused by vertical load
M, (kN m) o -43.3 -6.4 as given by the equivalent-bent method and the elastic
M, (kN m) o -26.9 -5.1
continuum method agree closely, but that the computed ro-
M, (kN m} o 66.9 -41.8
Pv, (mm) 16.4 12.9 tation ca u sed by both horizontal load and moment is consi-
Ph (mm) 8.2 10.4 derably greater in the equivalent--bent method. The equiva-
() 0.00490 0.00233 lent-bent method also gives a larger horizontal deflection
caused by moment, but a sr:1aller horizontal deflection
caused by horizontal load.
l. The vertical pile loads from the three methods are of The above comparisons, therefore, highliEiht the díf-
the same order, although the elastic continuum analysis ficulty of attempting to characterize a complex pile-soil
tends to predict a higher maximum load. system by a structural frame. Be cause it is of a more ration-
2. There is a considerable discrepancy between the al nature, the elastic continuum analysis should give more
computed pile moments from the equivalent bent and reliable deflection predictions.
PILE·RAFT SYSTEMS
10.1 INTRODUCTION witl1out the need to resort to computer analysis. The Davis
and Poulos method of analysis and design will be described
In the design of the foundation for a large building on a in this chapter. Sorne alterna:ive approachcs to pile-raft
deep deposit of clay, it may b~ found that a raft founda- analysis will also be rnentioned.
tion would have an adequate factor of safety against ulti-
mate bearing-capacity failure, but that the settlements
10.2 ANAL YSIS
woulcl be excessíve. Tradítíonal practice (assumíng the addi-
tíon of basements to produce a t1oating foundation is
The basis of the analysís is similar to that employed for
unacceptable) would then be to pile the foundation, and to
freestanding groups (Chapter 6\ except now the basic unit
choose the number of píles to givc an adequate factor of
to be considered is a single pLe with an attached circular
safety against individual pile failure, assuming the piles take
cap resting on the soil surface, instead of th.e previous unit
all the load. However, it is clearly illogical to design the
of a single freestanding pile. The interaction of pile-cap
piles on an ultima te-load basis when they have only been in-
units can be considered in a :'ashion similar to that des-
troduced in arder to reduce the settlement of an otherwise
cribed in Section 6.2 for freestanding piles. The settlernent
satisfactory raft.
interaction between two identical, equally-loaded units can
Pile-raft foundations have been successfully used in
again be expressed in terrns of an interaction factor, a,,
Mexico City (Zeevaert, 1957) and more recently in London
where
(Hooper, J 973). In the latter case, the finíte-element meth-
od has be en employed to analyze the behavior of the foun- Additional settlernent caused by adjacent unit
ex,
dation and compare it with the measured behavior. In a Settlement of single unit
method of designing a pile-raft system proposed by Davis (10.1)
and Poulos (1972b ), the number of piles required to re-
duce the settlement to the desired amount is determined Curves relating ex, to dimemionless pile spacing s/d are
250
PILE-RAFT SYSTEMS 251
1.0l--r==:-- ::::::¡ -- -
........
\
\
' '\
\
- .........
.........
'· ..... -- ¡-
................ d
L
10
~8
\
\
,, '
' .....
\
\
'' .....
.....
0.6
"" .
........
..... ..... .........
.....
' ....... .......
''
el,
Values of ---
d
~·,
.......
0.2
.,
2 4 6 8 10
S o: 0.08 0.06 0.04 0.02 o
d
º-
FIGURE 10.1 lnteraction factors for pile-raft uníts,!.:. 10.
d
shown for various values of dc/d (de cap diameter) in occurs for Vs O tl1an for v8 0.5, but that for dc/d > 10,
Figs. LO.l, 10.2, and 10.3 for three values of L/d. In all Vshas little influence on interaction.
cases, Vs = 0.5, the piles are incompressible (K= oo), the pile The curves in Figs. 10.1 te 10.4 may be used to analyze
cap ís rigid, and the units are situated in a semi-infinite piled foundations or pile-raft systems by considering them
mass. lnteraction increases as dc/d increases, bu t the effect to consist of severa! pile-cap units, each having an equiva-
of dc/d becomes smaller for larger L/d. lent value of dc/d such that the area occupied by the unit is
Corresponding curves for Vs = O are shown in Fig. 10.4 the same as that occupied by <1 typical portian of the cap in
for L/d = 25. A comparison between these curves and those the group. For example, for ;:, square a:rrangement of piles
for Vs := 0.5 shows that for dc/d ,¡;;; 10, greater interaction in the group,
l. 25
d
V 5 =0 5
08 !Loo
L
Valwzs of .2!;
d
06
o 4
o 2
0 o~---~2------4~-~-~6~---~a------10L-----~-----L------L-.
1 0~----~-----r----~------~----,-----~----~~-----~-----.-----,
1.
d
=100
08
v ~o
5 5
06
Clr
04
o 2
oo 2 4 6 8 10
S o1 008 006 o 04 o 02 o
d
L
FIGLRE 10.3 lnleraction factors f0r pile-raft units,- ~ 100.
d
1Or-------------------------------~-------------------------------~
08
'........._ ---- ------
............................................
...........
........... ......_
o6
--- ........... ......_ Valu<.?s of de
d
---
04
o2
o o~-----------------------------L-----------------------------~
2 4 6 8 10
S 01 008 006 004 002 o
d d
5
FIGURE 10.4 lnteraction factors for pile-raft units, vs O.
PILE-RAFT SYSTEMS 25'3
the value of a,, for the equivalent val u e of de/d These are re la ted as
of unit j, corresponding to the spacing between
units i andj Rs = nRe
P¡ load on unitj
p1 the settlement of a single pile-cap unit under For practica! use, since pile tests are normally carried out
unit load on a single freestanding pile, it may be more convenient to
~xpress the settlements in tenns of the settlement of a sin-
From Section 5.3.2, p 1 can be expressed as gle freestanding pile-that is, .n terms of Rs and Re (Sec-
tion 6.2). Thus,
(10.4)
Rs =Re ·Rs (10.9)
where
and
Pt settlement of freestanding pile under unit load
Re ratio of settlement of pile..:ap unit to settlement Re =Re ·Re (10.10)
of freestanding pile (Fig. 5.31)
The settlement, p, of the system is thus given by
As before, n equations may be obtained from Eq.
(10.3) for the n pites in the group; together with the equi- Pe
librium equation P = Rs·-·p¡ (10.11)
n
n
or
LP¡ (10.5)
j=l
P =Re· Pe· P1 (10.12)
they may be solved for two limiting cases:
For immediate settlements, p 1 is the immediate settle-
l. Equal displacement of each unit (corresponding to a ment, per unit load, of a single pile at the average load of
rigid raft). a pile in the group; and for total final settlements, p 1 is the
2. Equalload (or known loads) on each unit. corresponding total final settlement per unit load.
o5
-v, O 5
Pilli'S J..= 10
o 4 d
o 3
o2
01
o o 10 20 30 40 50 60 70 80 90 100
L
FIGURE 10.5 R<; vs_ b¡cadth. rigid rJft,~ ~ 10.
d
O G
Pí les -a-= 25
V 5 =0 5
o :s
o 4
o 3
o 2
10 20
254
PILE-RAFT SYSTEMS 255
o8
0-7
0·6
Ys O 5
o 5
Pi 1<2S h- =100
04
o 3
02
10 20 30 40 50 B 60 70 80 100
o
FIGURE 10.7 Re vs. breadth, rigid L = 100. O Br---,----,--- ------,---~---.
2
3 Pll" Group
o7
"• =0 5
.tl.::. c-;J
has virtually no influence on the settlement of the piles. L
As the spacing increases, the effect of the raft becomes o6
more pronounced. The spacing, s/d, at which the raft F'raa- standing group
begins to influence the settlement, íncreases as L/d in- Group w1th rott
creases and also as the number of píles in the group in- o5
creases.
Typical curves of Re versus B/d for the case of uniform
loading are shown in Fig. 10.9 for the center of the system
and in Fig. 10.10 for the comer. For the center, there
appears to be a single mean curve for all groups between
03
2 2 and 6 2 . For the comer, Re decreases as the number of
piles increases, up to a 6 2 group. Thus, at least up to a
62 group, the differential settlement wíU tend to increase
02
as the number of piles increases.
As previously mentioned, the curves in Figs. 10.9 and
10.10 provide an approximate means of considering a per- o 1
fectly flexible raft only for systems in which the pile
spacing is not excessive or the number of piles too sma!L
The general case of a raft of any flexibilíty could also be
5 10 15 20 25 30
considered approximately by means of the interaction
S
analysis descríbed herein, but a more satisfactory analysis d
requires a new approach involving consideration of the FIGURE 10.8 Comparisons bet\~een settlement freestanding
overall action of the raft (see Section 10.5). group and píle-raft system.
o 8
07
o 6
Pllcz; ~ :25
Ys=O 5
o4
o3
o 2
o 7
O G
O· '3
.k 25
d
v,=O 5
o 3
o 2
o 1
o 10 20 30 40 50 B
--- __ __ - __
60 70
,
80
...
90 10C
d
2 o o 6
• 22
{a) .1...:25
d
o 5 0
~ =25
A v5 = O· 5
2
0. 3 ~=886
101 o 4
o(')
t:.
a: 0
o-
(')
RG A e42
a: 1 o 3 + 1!:.
+ •52
o 2
0 Rows ol 4 pi l<ls
ARows ol 5 pil<ls
+ Rows of 6 pil<ls
o 1
20 40 B 60 80 100
d
o L-----~----~----~---~----~----~~
(b) 1_ = 100 o 5 10 15 20 25 30
d No. or Pil<ls
----
For the given group breadth and value of L/d, the value
of ReofRGo.s may be interpolated from Fig. 10.11, and Eq.
1 pil<l
(10.13) then used to estímate the required value of Rev-
The use of the above approximation will be described
subsequently in relation to the calculation of consolidation
settlements of the system.
!3_ev = 1 + ( Reo _ 1\ ( 1 _
RGo.s Reo.s ') 0.5
2) (10.13)
For large systems containing a considerable number of
pÚes, .the value of B/d may be very large and that of Re
258 PILE-RAFT SYSTEMS
very small, so that the use of the curves in Figs. 10.5, 10.6,
and 1O.7 may lead to inaccura te answers. lt is then con-
venient to consider the piles as a number of small groups of
piles within the system and to replace each small group by
an equivalen! single pile or pier. This equivalent pile should
ha ve the same area as the gross plan area ofthe small group
and an equivalen! length Le/L (see Fig. 6JI). The system
is thus replaced by a smaller number of shorter, larger-
diameter pi les. As an example, a system 40-ft square com-
prising an 8 X 8 group of 1-ft-diameter, 100-ft-long piles, 0·8
A
Q
• Cl.
0·6
piles, ca eh of equivalen t diameter de =5 5.65 ft, and -5
.!:;<11
now ata spacing of 10ft. From Fig. 6.31, the equivalen!
(})
o
length of cach pile is (for L/d 100 and s/d = 5), Le
oc
92 ft. Thus, the refcrcnce pile for the equivalent
~
e E "' 0·4
0.92L "'
E,g VaiUGlS of PiiGl
"'~
system ís onc having Le = 92 ft and de= 5.65 ft, orLe/de= -
:::!/) "' Stiffn12ss Factor K
J 6.3. In terms of this reference piJe, the group breadth is "' 0·2
now 40/5 .ú5 7.1 diameters, rather than the original
group breadth of 40 diameters.
lt has been found that the settlement of the system cal- o
culat~d for thís equivalen t system agrees with that of the 10
meters, solutions for the ratio of the settlement of pile and L"'
u
The use of even a very compressíble piJe beneath the very A further example of the effect of strip flexibility is
flexible strip (Fr = 10) causes a significant settlement- shown in Fig. 10.15 for a strip with five piles, ami equal
redu(;tion as compared with the case of a strip only, and the concentrated loads above each pi le. Distributions of dimen-
use of an almost rigid pile reduces the sett!ement by almost sionless contact pressure, settkment, and bending moment
75%. in the strip are given for three values of strip stiffness.
However, for a stiff strip (Fr 10 6 ), compressible pi! es The strip and pile lengths are 50 pile diameters, and the
have virtually no influence on settlement, and even a rigid strip width is 5 diameters. Fi~;ure 10.15 shows that as the
pile only reduces settlement by less than 20%. strip flexibility decreases, the contact pressures near the
The percentage ~ 5 of the applied load taken by the pi! es decrease while the settleraents near the pi! es increase,
piíe beneath the strip of 5-diarneter width ís plotted in Fíg. The magnitude of the bending moments increases as tb.e
10.14. This percentage decreases as the píles becorne more strip beco mes stiffer.
comprcssible or the strip becomes more rigid. For other Although the results presented in Figs. !0.13, 10.14,
values of strip width b, the percentage of load ~b is given and to 10.15 are only indic2tive, they may be useful in
app roximately by suggesting orders of correction to be applied to the S0lu-
tions for a rigid raft and rigid piles.
~b =L - 2.3 log 10 (b/5d) (10.14)
PE 5 d_ 5
100P
Pp, --------- -·/ B
'
Load P
.
M/Pd 2
o 0·2 0€ 00
FIGURE 10.15 Typical solutions for strip with five pUes (Wiesner FIGURE 10.16 Simplified appr•Jach for calculation of undrained
and Brown, 1975). load-settlement curves.
260 PILE-RAFT SYSTEMS
l. The !in e OA, from zero load to the ultima te load PA of where
the piles alone, the settlement being calculated from Eq.
(10.11) or Eq. (10.12) .. Reo.s the elastic value of Re for the pile-raft systern
2. The líne AB, frorn the load PA to the ultimate ·load PE for lls llu :::: 0.5
of the whole system (piles plus raft), the settlernent being Rev' the elastic value of Re for lls v~
calculated from the equation for the settlernent of the
p !TF :::: total final settlcment of a single piJe under
raft acting alone without the piles. For exarnple, for a
unit load
square rigid raft B X B with m piles, the overall undrained
p ¡¡ :::: immediate settlernent of a single piJe under
settlernent ata working load Pw is given by
unit load
:. PCF Pw(Rev'PITF Rco:SP!i) (10.19)
Pw
Both P!TF and Pli rnay be obtained either from apile-
where the fírst term represents the settlement of the pile- loading test (the settlernent pei unít-load, at the workíng
load) or from the theoretical relationships described in
raft systern, calculated on an elastic basis for lls = 0.5, and
the second term represents the settlement of the raft acting Chapter 5.
alone. This second term will only be operative if Pw > PA , The total final settlement of the system is then the sum
that is, íf the failure load of the píles ís exceeded. of the irnrnediate settlement (taking account of possible
pi! e slip) and PCF, or
lt should be cmphasized that the above basis of cal-
culating thc ultimatc load PB of the systern as the sum of (10.20)
the capacities of the piles and the raft is only valid where
rclatively few pilcs are added to the cap or raft (Le., where
the pilc-cap units are sufficiently widely -spaced to act
indivülually). If the piJe spacing is sufficiently close for The value of Rev' may be estimated from Eq. (10.13).
block failure to occur rather than individual-unit failure, If only a few piles are add ~d to the raft or cap, the
lhe ultimate load of the group should be calculated on this failure load of these piles may well be far exceeded at the
basis. lt should also be mentioned that the sirnplified
approach discussed above does not consider the effects of
local slip along the piles or of local yield of the cap as the
load íncrcases loward failure.
In calculating consolidation settlements, it is again
assumed that the consolidation process is not affected by
any local yielding occurring under undrained conditions, so
thal tlte consolidation settlement, PCJc, is
(10.17)
Minírnum MaximuTJ Settlement
and settlement settleme1t
working load of the whole system. Despite the facCthat Considering now the settlement of the raft alone. For
these piles may have failed, they are nevertheless effective a rigid, square raft on a semi-in!'inite mass,
in reducing the settlement of the system, as is shown día·
grammatically in Fig. 10.17. The validity of the above
approach is dependent on, among other factors, the raft
maintaining contact with the underlying soil. It is con·
Considering the total final settlement under the working
ceivable that with a relatívely large number of piles, and
load of 3500 tons,
with relatively soft clay directly beneath the raft, the raft
could be "held up" by the piles because of the effects of 3500 X ( l - 0.35
2
)
nega ti ve friction and lose contact with the soil, thus effec· .947 X 63
tively reducing the system to a freestanding group. This 0.925 ft
situation wíll probably only occur when the piles are suf.
ficiently closely-spaced to act as a block, in which case the Thís settlement is excessive, and hence pites rnust be
raft would be ineffective, even íf contact were maintained. added to the raft to reduce the settlement.
However, where only a small number of pi! es are preseQt Calcula~ions will be detailed for 1-ft diameter, 100-ft-
in the system, and tl:ese pites are overloaded at the working long piles. Results for other types of piles will be sum-
load of the system, continuous contact between the raft marized.
and soil should be maintained. This latter case is the one
Immediate Settlements
beíng prit¡1arily consídered here.
Considering undrained condit.:ons first: The bearing ca-
The use of the simplífied approach described above is pacity of a single pile, assumíng (possíbly rather conserva-
demonstrated in the example of the design of a pile-raft tively) that cafcu 2/3, is found to be 145.5. tons. The·
system, given below. undrained settlement of a single piJe under unit load is
(from Eq. 5.33)
Illustrative Example I
The case of a square, rigid raft 50 ft on a side, restíng on a Pu = dEu
deep deposit of soft clay, will be considered. The total
working load on the raft is 3500 ton:;. The relevant average For this case, L/d 100 and 1 0.0254 for Vs == 0.5. There-
pararneters of the day are as follows: fore,
1 o~----~-----T------·~----~-----r----~
o 7
o 5 o7 50ft~.
2ft.~;~""· ~
PrF (ft)
~,0·4 06
+-'
100ft pilll.5, ~
~
Q~
1ft:dl0 ~
o 3 --- o 5
o 2 04
o 1 o 3
01) 2 4 6 8 10
o 2L-----~----~-----~----~----~----~
o 2 4 ; 8 10
FIGURE 10.18 lmmediate sctticment vs. number of pi!es-illustra· FIGURE 10.19 Total-final settlement vs. number of piles-illustra-
tive examplc. tive example.
1) ,¡. .OOOJ 54 ftiton Also shown in Figs. 10.18 and 10.19 are the relatíon-
ships between settlement and number of piles for two
Thcrefore, from Eq. (10.19), other, different types of píles. lt is notable that the shorter
50-ft píles, even though they ar<~ of larger díameter, are not
PcP 3500 X (Rcv0.000384- RGo. 5 0.000354) as efficient in reducing settlement as the 1-ft-diameter, 100-
1.343 Rcv 1.240RGo.sft ft-long piles. The settlement criterion cannot be satisfied by
using l·ft-diameter, 25-ft-long piles, regardless of the num-
Value> of PcF are shown in Table 10.1, together with the ber used. It is interesting to note that for such piles, the
total settlement (p¡ + PCF ). normal design procedure woul d require the use of 246
such piles to satisfy bearing-capacity requirements,
The variation of P; and PTF with the number of piles
in the system is shown in Figs. 1O.18 and 10.19. It will be
seen that in order to satisfy the specified settlement crite·
rion, only 16 of the l·ft-diameter, 100-ft·long piles are 10.5 OTHER ANALYTICAL APPROACHES
required. The undrained bearing-capacity of this system ís
12,100 tons. The traditiona1 design procedure, which deter-
In addition to the method described in detaíl in this chap-
mines the number of pi1es so1e1y on the basis of u1timate
bearing-capacity wíth no allowance made for the raft, gíves ter, a number of alterna ti ve approaches to analyzing pile-
the required number of piles as 68 (to gíve the same load raft systems can be contemphted. Those described by
capacity as the raft, i.e., 9830 tons). Thus, a very consider- Brown et al. (1975) are listed below:
able economy in design is effected. Furthermore, the use of
68 piles rather than 16 leads to a further reduction in settle- l. Strip-superposition method, in which solutions for
ment of only 0.1 ft. píle-strip footings (Brown and Wíesner, 1975) are super·
PILE-RAFT SYSTEMS 263
TABLE lO.l
posed lo obtain the settlement of the raft. Thís method simulted by a continuous annulus with an ovcrall stiffness
does not require the use of a computer but is limited to equal to the sum of the stiffnesses of the individual piles.
giving settlements only. The lreatrnent used by Desai ·~t al. was similar, excc lhat
2. "Plate on springs" analysis, in wh::ch the raft is analyzed rows of pi! es were simulated by a continuous strip .. v'hile
as a plate using the finite-element method, with the piles such approaches offer f1exibility ín being capable of taking
being replaced by springs located at appropriate nodes. The into account such factors as soíl ínhornogeneity, they
stiffness of these springs can be estimated from the eiastíc suffer from the fact that a o:onsiderable volume of data
solutions for a piJe ( allowing for interaction effects) or is required, and there will be difficulties in choosing an
from the pile-raft analysis described earlier in thís chapter. appropriate stiffness for the ring or strip simulating lhe
piles, and in dealing with pile slip.
3. "Plate on springs and continuurn" method. Here the
raft is again treated as a plate and the piles are replaced by These last four approaches require the use of a corn-
spríngs, as in the method above; but in addition, the soil is puter but have the advantage that distributions of settle-
treated as an elastic contínuum as far as support to the ment, pile load, and raft-bendhg moment can be obtained.
raft i~self is concerned. The results of analyses bmed on five methods (the first
4. "PÍa te on pi! es and contínum" method. This approach four approaches, 1 to 4, abo ve, plus the method describe el
has be en described by Haín ( 1975). It gives a el o ser repre- in detail in this chapter) were compared by Brown el al.
sentatíon of the real problem by treating the pile as in the (1975). They analyzed two relatively simple problems, one
normal pile-settlement analysis (Chapter 5), the raft as a involving concentrated loads actíng at the location of the
plate, and the soil as an elastic continuum. Interaction píles, and the other beíng a raft-pi.le system subjected lo
among the piles, raft, and soi! is then taken into account in uniform loading over the whole area; in each case, both
a logical manner. Haín and Lee ( 1978) ha ve u sed this ana- a stiff raft and 'a relatively fi,:xíble raft were considered.
lysis to successfully predict the load and settlement distri- The pla te-on-piles and con !in Jum rnethod ( 4 above) was
bution for two pile-raft systems, one in Mexico City (the assumed to give the reference solutions, as it involved the
La Azt,eca building) and the other in London (Hyde Park leas! approximation. From !he poínt of view of settlement,
Cavalry Barracks). the most satisfactory of the other four methods was found
5. S1mplified finite-element analyses; Hooper (1973) and lo be the elastic-based analysís described in this chapter.
Desai, Johnson, and Hargett ( 1974) ha ve descríbed fínite- The method of strip superpcsition overestimated settle-
element analyses of piled-foundatíon problems in which the memts, while the plate-on-spri:lgs and continuurn method
foundation, the piles, and the soil are represented by finite consistently underestimated settlement, presumably be-
elements, without performing a full three-dimensional cause ít ignores the downward movement of the contínuum
analysis. The case Hooper describecl was approximately arising from the settlement of the piles. The settlements
axially-symmetric, and each concentric ring of piles was given by the plate-on-springs nethod were generally too
264 f'ILE-RAFT SYSTEMS
large, although the error depended on the basis adopted for In summary, if only the settlement of the pile-raft
the se:ection of spring stiffnesses. system is required, the elastic-based analysis is likely to be
From the point of view of bending moments in the raft, adequate if the raft ís very stiff or very flexible. If bendíng
none of the simple methods gave accurate results when moments in the raft are required, none of the simple
compared with the values from the plate-on-piles and con- methods are satisfactory and a proper analysis of plate on
tinuurn analysis. (It should be noted that the method of pi! es and contínuum is desirable.
this chapter does not predict bending moments in the raft.)
NEGATIVE FRICTION
ON END·BEARING PILES
""'"'
TABLE 11.1 DOWNDRAG MEASUREMENTS ON PILES
.
Maximum
Negative (1) Ps
Pile Friction (2) Pp
Const. L d Load (3) t.Psp Soil Sens·
Reference Details (m) (cm) (Tons) r 12 /a~ (cm) Condition P.!. eula~ L. l. itivity Observations
J ohannesscn, Steel 53 47 -400 0.20 1) 200 Soft 25 0.15 0.5 4 ' Large negativc-frictíon loads caused
Bjerrum Pi! e 2) 10 marine· ± ± ± '!:. rock point to penetra te -1 O cm into
(1965) pi!e, 3) 190 el ay 6 0.05 0.1 2 rock, consolidation caused by a fiU
do sed 1 surcharge.
tip,
driven 57 50 300 0.18 1) 27 u , , , , Piles Il, lli driven at diffcrcnt sites
to rack 2) 5.3 from l.
u 3) 21.7
41 50 250 0.23 1) 7
, , , , , Driven ín an are a consolida ting o ver
2) 3.2 the last 70 yr. Disturbance cau~ed by
3) 3.8 pilc driving íncrcased settlement by
3-5 cm. liPsltíme befare driving 0.1
m cm/yr
Bjerrum, PileA: ~30 30 120 0.26 1) -20 7-m fill, 15 0.25. 0.7 4 Control pile, no spedal treatment, no
J ohanncssen el o sed 2) 3.3 ctayey ± ± 2 enlarged poin t.
(1969) tip 3) 16.7 silt, 5 0.05 0.2 2
steel silty
pipe el ay (Píles A-D at He roya Site)
pile
driven
to cnd
bcaring
on rock
PileB: 27 30 10 -0.03
, , , 11 , Enlargcd point (diam. = 40 cm) used
shnilar to protect asphalt-bítumen. (Bitumen
toA but l·mrn thJck, ISU/100 penetration)
with ene Bentonitc slurry around bitumcn.
larged
pbint
S~ renga ~57 50 300 0.18 1) 5 Soft 25 0.15 0.5 4 Control pile, no enlarged point, no
Sí te: 2) 2.5 marine ± ± ± ± treatment. Poínt penetrated 2.5 cm
PileC, 3) 2.5 clay 6 0.05 0.1 2 into rock.
driven
stee1-
pipe
pile to
end-
bearíng
on rock
,,
.Pi1e D, -57 50 15 -0.01 1) 5 " " " 11
Bitumen covered, bentonite stabi-
same as lized, fil1 (20m) sepa.rated from pile
Pile C by casing. Negatíve frictíon reduced
by 285 tons. Enlarged point (69 cm
u sed).
PíleE, ~57 50 210 -0.13 1) 5 Soft 25 0.15 0.5 4 Bitumen covered, enlarged point, bu t
same as marine ± ± ± no bentoníte or casing; bitumen
píleC e! ay 6 0.05 0.1 2 scraped off during dríving.
He roya 32 50 300 0.25 1) -30 7-m fill, 15 0.25 0.7 Enlarged típ?
pile 85 c1ayey ± ± +
silt, 5 0.05 0.2
silty
day
Fellenius, Precast 40 32 30 tons 0.05 Very soft, 45 0.35 1.0 18 Negatíve friction caused by consoli-
Droms reinforced (me trie) sensitive ± I (+0.5 ± dation of soil remolded by píle
(1969) píles driven marine 5 0.05 -0.1) 6 dríving.
to end· el ay
bearíng
in sand,
silt
Endo, Minou, (1) Steel- 43 61 250 to -0.35 1) 5.4 Silty 45 0.55 0.6 low Negatíve frictíon caused by consolí-
Kawasaki, pipe piles, 300 tons 2) 12;5 sand, ± ± ± datíon ofclay by pumping underlying
Shibata point (U.S.) 3) 7.1 soft 3 0.05 0.1 layers. Settlements 6.72 days after
(1969) closed, silt driving.
driven to
end·
bearing
ín
., alluvium
"'...
IV T ABLE 11.1 (Con tinued)
"'
(JJ
Maximum
Ncgative (1) p S
Pile f'riction (2) Pp
Const. L d Load (3) ilPsp Soil Sens·
Reference Details (m) (cm) (Tons) (cm) Condition p .1. L. l. i tivity Observations
Bozozuk & Composite 82 99 920 tons "-0.18 1) 37 Gray, silty 40 0.25 0.4 lO Ncgatíve frictíon causcd by embank-
Labrecque pi! e, (U.S.) 2) 1.5 clay ;; to lo ment loadíng- piJe p1accd 18 months
(1969) concrete· 3) 35.5 0.05 1.5 30 after embankment construction com-
fil!ed menced. Early pile hcave noted,
steel causcd by ecment hydration, and
tu be: electro-osmosís applied tu piJe.
pile B2
NEGATIVE FRICTION ON END-BEARING PILES 26~
11.2 HELD STUDIES ON INSTRUMENTED PILES from consolidation caused ,by pile driving is reported by
Resendiz et al. (1969).
In recent years, a number of tests on instrumented piles The variation of downdrag load with time (logarithmic
have been carried out to study the effects of negative scale) for two pi! es tested by Fellenius and Broms, is shown
friction. A summary of sorne of these tests is given in Table in Fig. 11.2. It shows the same characteristic shape as a con-
11.1 ,. and the results are díscussed below in relation to a solida tion curve.
number of aspects of behavior.
~~~~~l+l~,==~=~~=~~====~~==~~~~~rg~d-~g~-~~!~~to
_ \ Seabottom 1962: ¡,April ~~63 1 No:'1963 ~ ::: :. ~t~
.. '\~
1
1 April1963
,""
""..-·']/
t--:,...~r--i+~l----f---+----t---~
17~~~ \\
•
:!E f ~.7z'¡¡'J -Jo
~~ /1
X
1 ""
(".-;:..-· /
,.
i1 ~
1
~
-+-l/-+-/1---i -20
~-----;,~arr0-M_a_.y_- -+--············+-!1 -+-:--1-/-r--t--
o
T
,
1 96 3
A¡.rir 1964
11
: ¡ ¡ settrement
1
--_May 1963rl--+----t------t-----JI---tJ-¡·'"
11 Calculated primary 1 April 1964
\
¡ h'/ ~
E
-~
i
1
~-~~~t---r-~~-7~---~-----i--~+~~~--~----t·--.r+~~--f~-~-30 .g
flu f:J"Q~'Y
~ :! 1'
1• j
1
\~o~puted 1 ,¿rJ ~
W
r
1
-u 1 1/ j 1 Observed
,, ·4---~~--+---~f--~~-----~---;--r~~7~-t----1-40
11 ~ \~
1
1
1
~~
"-.." ~
JLV=~g:::~~:d -50
~-------------r---,
1" ''~"-
Soil Undrained shear strength,
de- tím 2
2 4 6 B 10
-~May8,
~~~~:i\963
""
L __/·s!~?~~o 6 7
'
Pile A . + Pile B Slope, April
,~ 1963 3
Symbors/ 1 • \
1
o Test pile O 10 20 2
+ Vane boring Clay
o Sarnple boring Scale ir '1leters
3
• Pi¡>e throug,_,h..:.f:.:.il:..l- - - - - - - - - - - '
A 8
$12345 6 7 3
Clay
and
sand 8
layers 3
~
"'
"'e
e
w natural water content o ""' unconfined c::>mpression test
e
o wL linuid iimit "' ~ laboratory cone test
"'
> wP = plastic 1imi!
J:
(/) Soil Propertíes
Bedrock
Symbols
¡¡. Measurement points in test pi les
o Piezorneter
- Settlernent anchor
" Cobalt capsule
FIGU.:<E 11.1 \le~surements of downdrag on steel piles (after Johannessen and Bjerrum, 1965). {il:! Canada. 1965, by University ofToronto
Press.)
210
NEGATIVE FRICTION ON END-BEARING PILES 271
Time after driving, days but the effectiveness of this treatment could not be ac-
1 10 100 1000 curately determined, although it did app!:Jar to be at best
0.------------,-------------,----------~
partially effective in the lower part of the pile.
Tests by Fellenius and Broms
{1969) In a further case in which electro-osmosis has been suc-
cessfully used to relieve negative friction (M .LT ., 1973),
treatment of a steel piJe in Boston blue clay for oníy six
hours was sufficient to eliminate virtually al! drawdrag in
v; 10 - the piJe.
e
B The theoretical analysis described in the following
;¡¡ section may be adapted to pr.;díct the rate of reduction of
E
en 15 - downdrag in apile as a result of electro-osmotic !reatment,
ro
.;:; by using the theory of eleclro-osmotíc flow iri a soil to
o
~ determine the rate of increast~ of pore pressures wíth time
o
:>o --
along the piJe, and hence the nte of decrease of adhesion.
:>5
11.3 ANALYSIS OF DOWNDRAG FORCES
P,
Deformed
surface
e;. vertical effective stress puted downdrag loads, althoug[ Zeevaert's method gave the
q•~ drained angle of friction between piJe and soil. lowest value.
A nurnber of rnethods have recently been developcd
The "alues of these pararneters are discussed below. that make use of the theory of elasticity to estímate down-
Zeevaert (1959) pointed out that the transfer ofvertical drag forces (Salas and Belzunce, 1965; Begernann, 1969;
stres! to the piJe through negatíve friction reduces the verti- Verruijt, 1969; Poulos and Mattes, 1969b; Poulos and Da vis
cal ( •vcrburden pressure on the bearing straturn. If this 1972). These rnethods have the advantage ove1 previously
stratum consists of sand, the point-bearing capacity of the described rnethods in that the dependence of downdrag on
piJe will be reduced, since it depend~ on the vertical stress. the settlernent of the soil can be studied, rather than as-
Thus, negatíve fríction will have the dual effect of inducing surníng that suffícient settlernent occurs to mobilize the ad-
dowr drag load and reducing the point-bearing capacity. In hesion along the whole length of the pile. Walkcr and
hís analysís accounting for this reduction in point capacity, Darvall (1973) have descríbed (! finite-element analysis that
Zeevaert shuwed that for soil conditions typical to Mexico can be used similarly to analyz.e the interrelationship
Cíty, the reduction in point capacity together with the in- between settlement and downdrag.
duced downdrag force severely reduced the allowable load The analysis described below follows Poulos and Mattes
capality of the pilc. However, íf the point resistance is not (1969b) and Poulos and Davis (1972), and is an extension
stron?;ly dcpendent on the effective vertical stress ( e.g., of the settlement analysis of a single pile (Chapter 5). The
ceme l!ed or cohesive cornpacted strata), the correction for analysis is divíded into two part:;:
rednced tip resistance does not apply. Furtherrnore, for a
sand bearing straturn, íf the piles are driven well into the l. Analysís of final downdrag force.
stratt:m, most of the pile load rnay be resisted by síde 2. The development of downdrag force with time during
shear. so that little load may reach the ti p. Thus, the reduc- consolidation.
tion •n pile tip capacity only occurs under certain tpecial
conditíons, and in most cases, it ís probably sufficient to
ignow the decrease in verticaJ effective stress with depth. 11.3.2 Analysis of Final Downd.rag Force
This procedure may overestirnate the base resistance, but
this will be compensated for by the tendency to over- The problem is illustrated in F:g. 11.3. The piJe is divided
cstimate the downdrag force. into elernents as for the axially-loaded end-bearing pile in
Lt)cher ( 1965) cornpared the computed downdrag load Section 5.2 .1. While elastic conditio!)S prevail withín the
in a Ji)e from Terzaghi and Peck's approach, Zeevaert's soil, soíl and pile displacements at each element are equated
appro 1ch, and an empirícal approach developed by Elmasry to obtain the unknown shear stresses along the pile. It will
(1963). There was relatively close agreement between com- be assurned that the bearing stra:um is rigid.
NEGATIVE FRICTION ON END·BEARING PILES 273
~ + I- 1]
The vertical displacement of the soil at any point arises
[ {p} Es {S} (1L6)
from two sources: the shear stresses along the pile and the d
con~olidation of the soil .\ayer itself. The soil displacements
caused by the shear stresses are given by Eq. (5.22) as
( taking downward displacements as positive and k 1)
(id) (RA )[h]
-d , where
{s¡P} =-;-;-[!- l]{p} (11.3)
bs
where
K = (~:) (RA) = the pile-stiffness factor
Pa = applied axial stress Pa/Ap
{si p} vector of soil displacements ca u sed by shear
stress Equation ( 11.6) m ay be sol ved to obtain the n unknown
{pl· vector of shear stresses shear stresses along the pile shaft. The load per unit arca,
[I - /] matrix of displacement-influence factor (in- Pb, transferred to the pile tip m ay then be obtained, from
cluding the effect of the "mirror image 1' equilíbrium considerations, as
elements, Fig. 5.5)
pro¡~ram at each time step. The effect of delaying u¡ excess pore pressure at í at time t
installatíon of the piJe for sorne time after settlement of the qt effectíve applied sur,~harge pressure at time t
clay has begun can also readily be handled by startíng the q reference value of effective surchargr: pressure
corr,putation of {S} at the time of ínstallation. For loading (e.g., the maximum value)
tha. varíes with time, or cases involvíng pore pressures
created by driving the pile or by the application of electro- (Values of Ks tan 1/J~ will be discussed in Section 11.4.7 .)
osmosis, a numerical analysis may be necessary to deter- lf the consolidating soillayer is overlain by other layers
mine the variatíon of pore pressure-and hence settlement- and has an initial effective stress p 0 acting at the top of the
with time along the pile. !ayer, it can be treated as having an equivalen! pile-soil
adhesion c~e, where
The a hove elastic analysis may be modífied to take account If thc downdrag force inducecl in a pi! e is sufficiently large,
of ·oca! yield or slip between the pile and the soil in a the crushing strength of the pile may be reached; if this
mar ner similar to that described in Section 5 .2.3. Such a occurs, the crushed portian of the piJe can sustain no addi-
moclification is very desirable in considering negative tional load and a redistribution of load in the remaining
friclion, since, as mentioned in Section 11 the field portion of the pile occurs. The additional displacement that
evidence indicates that sh~ft-soil slip is very like1y to occur accompanies the crushing of the píle now enters into the
when soils consolida te past a piJe, especially with soft clays. analysis as an additional v.niable. At each time, the
At any given time, the shear stress p on an element, as maximum axial stress in the piJe, from the analysis for an
given by the elastic analysis, is compared with the pile-soil ínfinitely strong piJe, ís checked agaínst the crushing
shea strength Ta at that element. If pis greater than Ta, it is strength, qc, of the píle: if it exceeds qc, the element can
set equal to Ta, dísplacement compatabílity now beíng con- sustain no shear stress, as the load in this crushed portion
sidered only at the elastic elements. A new solution is remains constant (assuming ideal elastoplastic behavior of
obtained and the procedure repeated until all shear stresses the pile material). Thus for the cmshed elements, the dís-
are less than or equal to Ta. The value of r 0 at any time t placement-compatability equations are replaced by equa-
can be estimated from the Coulomb expression tíons stating that the shear str<:ss is zero. A further equation
is províded by the condition that the axial stress at the top
e~ + a~ tan 1/J~ (1 1.8) of the crushed portion equals the crushing strength, qc.
Thus, for a surcharge pressure q, and crushing at the top of
where element k,
~ (Pi)-f. (qc)
effectíve stress values of pile-soil adhesion and
friction angle n ) (11.11)
effective normal stress at time t
i=l \_q -\.4L/d q
For thc case of a surcharge on the surfacc (see Fig. 11.3c), A total of (n + 1) equations is then obtained for the n shear
assumíng the water table to be at or above the surface of stresses and the unknown displacement caused by crushing.
the consolidating !ayer, Eq. (11.8) may be written in These equations are solved and the procedure repeated until
dim~nsionless form, for a point i at depth z below the the axial stress does not exceed the crushing strength of the
surface, as pile at any point. This.solution is then the required solution
for the time considered. The whole procedure can be re-
peated for a number of times after installation of the piJe.
(11.9)
q
where
11.4 THEO RE TIC AL SOLUTIONS FOR SINGLE PILE
Ks coefficient of lateral pressure, assumed to remain
constant during consolidation There are a large number of parameters that may be inves-
1' submerged unit weight of soil tigated: the piJe parameters L¡d and K; the soíl parameters
NEGATIVE FR!CTION ON END-BEAR!NG PILES 275
Vs, c;
1/q, ¡L/q, and Ks tan if¡~; the drainage conditions of the Pa = axial force in pile at top of consolidating !ayer
soil layer; and two time parameters--the time t 0 , between
the commenceinent of consolidation of the soíl and installa- The first term in Eq. (11.13) represents the maximum
tion of the pile, and the time t, between the commence- downdrag force; the acldition of the term Pa is correct if
ment of consolidation of the soil ar;d the time being con- full slip occurs, but only approximate in other cases. How-
sidered. The latter parameters may conveniently be ex- ever, the resulting overestima':e of load at the pile tip is
pressed as dimensionless time factors, generally small in most practica] cases. lt should also be
noted that Pa may ínclude the axial force caused by nega-
Cvlo tive friction of overlying soil layers, as well as applied axial
Tvo ulvf (ll.l2a)
force at the piJe head. If su eh layers are noncohesiie, it
will generally be sufficiently accurate to assume for the cal-
and culation of Pa that full slip occurs between the pite and
these layers.
TV S.~M (1l.l2b)
L2 PNFS is expressed as
L
where
PNFS = nd Jr 11 dz (11.14)
o
Cv =
coefficient of consolidation of soil
AI = 1 ( one-way drainage) or 4 ( two-way d1aínage) where
For most of the numerical results given in this chapter, ra final pile-soil adhesion (Eq. 11.2)
the effects of the parameters c~jq, ¡Ljq, and Ks tan 1>~ are
given, while the other parameters are kept constant at typi- For a uniform soíllayer,
cal values L/d = 50, K 1000, and ~-'s =O; however, sorne
indication of the effects of taking ocher values of L/d, K,
f
nd1> ta
1
+ K s tan Gla, (¡L
2 + e~\] (11.15)
and lls is also gfven.
Attention is concentrated on the maximum downdrag Correction factor N R repre~ents the ratio of the actual
force in the pi! e and the settlemen t of the top of the pi! e, maximum downdrag force to that for full slip along the
and the results are preseoted as in Poulos and Davis (1975). píle, and is plotted in Figs. 1 1.4, 1 1 and 1 1.6 as a func-
Sorne detailed distributions of force and stress along the
pile have been presented previously (Poulos and Mattes,
Contours of NA
1969b;Poulos and Davis, 1972).
An illustrative example outlíning the use of the solu-
tions is given in Section 11.4.8.
where
1d 50 v; ~O
Contours of NR
K~1000 T 0 ~0
K, tan q,; = 0.20
2 4 5 6
tíon of c~fq and -yLfq, for three values of Ks tan </>~. The
curves in Figs. 11.4, 11.5, and 1 i .6 are for L/d 50, Vs "'
K
O, and K = 1000. A decrease in K or L/d tends to decrease
FIGURE 11.7 lnfluence factors for final downdrag at pile tip- clas-
NR, but the effects are generally small and these figures tic analysís.
may be applied to most cases involving full or partial slip.
The regions in which NR is unity represen t cases in which
full slip occurs along the piJe. As c~fq and -yL/q íncrease,
NR tends to decrease, but only for high values do fully-
elastic conditions prevail (generally for c~/q > 5). Such
cases arise if the soil is stiff, or the consolidating !ayer is
situated beneath a deep overlying ]ayer, or the applied
pressure q causing consolidation of the Jayer is small.
If fully elastic conditíons are indicated, a more satis-
factory predíction may be made by using the elastic solu-
tions, in which case the values of K and Vs may become
significan t.
If the soil sett!ement is assumed to vary Iinearly from
S 0 at the surface to zero at th~ base, the elastic maxúnum
downdrag force may be expressed approximately as fol-
lows:
(11.16)
where
-~=50
d
K"'
One-way drainage
0~------J---------~--------~------~-----------~----~
For a uniform !ayer subjected to a surcharge pressure q, average value is the same in bÓth cases, it is found that
Eq. 11.16 becomes downdrag forces and piJe movement are decreased (typi-
cally by 10 to 25%) as compared with the case of a linearly-
(11.17) varying Es. Thus, the assumption of constantEs with depth
will ·be conservative when soil .:onditions are elastic. In all
Values of !¡y are plotted in Fig. 11.7. L/d and K ha ve a the ensuing solutions in this chapter, a constant Es is
majar influence on IN, but the effect of soil Poisson's ratio, assumed.
v~, is relatively small and may be approximately accom- Values of NT are shown in Figure 11.8 for one-way
modated by use of the factor R. (top) drainage and in Fig. 11.9 for two-way drainage. Nr
The effect of having a linearly-varying soil modulus represents the ratio of maxi.num downdrag force for
with depth rather than a constant value has been investi- delayed installation at time T0 to that for T0 O. For one-
gated. If the distributions of Es are chosen so that the way drainage, NT decreases as T0 increases, although only
2B NEGA TI VE FRICTION ON END-BEARING PILES
0.001 10
o
2
:::;¡
ai 0.2
ú
.2 i
ApproxiMate limiting curve
"'::: 1! full slip occurs during ·
"eS: 0.4
development of downdrag
~50
o
"
o~
e
w 0.6
E 1 = 100-0-1-----~--l-----\---\-~-l-··~~~--l
!-----~ 1
a.
o
w
> One-way drainage
"'
"o O.B
w
1::
o"'
"
10
z
:J
<U
'-' 0.2
2
"'"'
-o
e
¡; 0.4
.,o
o
e
E
Q_
o
<ü
:>
"'
1::
o.
"'o,
o'"
(11.18)
50 o
:ontours of 0;:: ---~<=o
K, tao,,; ~ 0.20 1000 o ---,i=03
p
. '¡ .
L. -----r-
100,000 ¡.-.--v~a~lu~e~o~~ l
/í 1
-----
_¡L
q
L
Contours of QR 50 Correction factor QR is p .otted in Figs. l 1.13, 1 1.14,
d
and 11.15 as a function of c~jq and ¡Lfq for three values of
Ks tan r/J~. A value of QR = 1.0 indica tes that full slip
occurs along the pile. QR tends to decrease with increasing
c~Jq and ¡Ljq, but ata slower rate than the downdrag-force
correctíon factor NR Cases in which conditions remain
entirely elastic are also indicat•;d in Fígs. 1 LI3, 1 Ll4, and
11.15, and in such cases, th~: use of elastic solutions is
preferable.
_¡L
q For elastic conditions, the axial movement of the piJe
may be expressed approximately as follows:
(11.21)
where
and 1/d, for v~ = O. One curve for v~ 0.3 and 1/d 50 caused by negative friction developed between the piJe and
also g1ives sorne indication of the effect of v~. any overlying layers.
Qr is plotted in Figs. 11.17 and 11.18 for one- and If the bearing stratum ís not rigid, the additional tip
two-way drainage. QT generally decreases as T 0 increases, settlement may be estimated conservatively by treating the
reflecting the corresponding reduction in axial force. pile tip as resting on ( or embedded in, if appropriate) the
The second terrn in Eqs. ( 11.19) and ( 11.21) represents bearing stratum and being subjected to the maximum axial
the settlernent of the pile acting as a freestanding column force in the pi! e.
under the axial load. The addition of this value to the When the consolidating soil !ayer is overlain by other
settlernent caused by downdrag will give the correct settle- layers, the settlement of the portian of the pile in these
menÍ if full slip occurs, or a slight overestimate in other layers must be added to the calculated pile-settlement at
cases. Jt should be noted that Pa includes the axial force the leve! of the top of the con:;olidating !ayer.
To =
L'
>-
0
.,
L'
11.4 .4 Rate of Development of Settlement up degree of pile settlement, for a time factor Tv =
2
Mcvte/L
As with the maximum downdrag force, the settlement of
the pile Pt at any time te after installatioh may be calcu- The first term represents the time-dependent component
lated as resulting from negative frictior and trie second the constant
value resulting from applied lo:ld. Up is plotted against time
Pt (11.22) factor Tv in Figs. 11.19 and 11.20 both for a purely elastic
soil and one in which pile-soil slíp occurs and for one· and
where two-way drainage. As with the maximum downdrag force,
the piJe movement develops :nore rapidly íf pile-soil slip
p final deflection of pile, from Eq. (11.19), or Eq. occurs or if two-way drainage corrtlitions exist. Also, L/d '·
( 11.21 ), if appropriate K, and v~ have Jittle influenc(~ on the Up·versus-Tv curve,
T :::,::¿<:
' L2
0.001 0.01 0.1 10
o
~e
K
50
1000
l
0.2 o-~··
~Q
~-
Approximate
e !imítíng curve
"'E if full slip
04 !-------~-.. occurs during - - - - ' H - - ' \ - - - - -.. --+---·...- -..
'"
-;:;
~ development of
~ downdrag
Ci
ñ 0.6
e"'
"'o"'
0.001 0.01 o. 1 10
o
l. 50
d
0.2 1-----=-.....:;::--:---------'"""'--+-------- K = 1Ql)_O_ _
v'~ =O
~ during development
of
.!'!
Ci
10
06
kr ~25
y~ o o
08 K o1000
On12·way dra1nage
qc¡q~200
lOO OC ._______,_____ _,_-=Eccla;;.:;s.;:;.ti.;:..c-=:o...:.•1_ _____.
1 0
(b}
One-woy top droinogc¡¡
~q
0-1 o 1 2 3
1
¡1
11
11
02
,,
11
~
\
'\
\
---"'"'
<:1
rr
l
0·4 1\
\\
\
\
L
•'
3 T \
<11 \ \ Volues of \
"'P.'o 2 06
1 \ Tv \
1 \ \
.<:
"
L
:l
<11
ps \o 2 o 05 01~
r _cvt 1 \ \
o1 02 03 v-1::"2 08 1
1 \ \
(a) Pottczrn of Surchargcz Loodmg 1 \ 1
1 \ 1
1 o ...__ _,___,__,_____.__ _.__ _.___ _.__,_.
(b) lsocrronczs at VariC•us Tv
11.4.6 PiJe in Soil Snbjected to Variable Loads placement of the surcharge. The soil is assumed to have a
constant permeability during 1he consolidation process, but
As an example of the further application of the analysis, to have a value of mv for elípansion or recompression of
Fig. ll.23 shows the case of a pile installed in a !ayer that one fifth of the value for virgin compression. The distribu-
is ::ubjccted to a variable surcharging loading. The deve- tion of pore pressure in the l2yer at various times has been
lopment with time of downdrag force at the tip is examined obtained by a finite-difference analysis. The resulting iso-
for various times of installation T0 after commencement of chrones are shown in Fig. 11.23.
T- Cyl, 2
- /L
00001 001
001
o 02
IN
003
004 la-= 25
K =1000
v;: O
005 Elastíc soil
The varíation of maximum downdrag load (PN = layers, in which case an equivalent v.alue, c~elq, m ay be de-
INqe) with time for various values of T 0 is shown in Fig. terrnined from 11.10. If c~/q =O, Talo~ K 5 tan . An
11.24. It will be seen that when the pile is installed during examination of values of Talo~ deduced frorn available
the peiiod of build-up of surcharge, the downdrag in creases field-test data was rnade by Dawson (1970). It was found
rapidly with time until removal of the excess surcharge. A that TaÍO~ tended to increase with increasing plasticity in-
further small increase in downdrag is followed by a decrease dex and increasing culo~, where cu = undrained shear
to a final value of load-which in this case, is very close to strength. Coating a píle with bitumen greatly reduced
the load in the pile at the time of removal of excess sur- Tala~. The correlations obtained by Dawson are shown in
charge. The later the time of installation, the less the ulti- Figs. 11.25 and 11.26 and rnay be useful for design pur-
mate downdrag load. poses when no other data is available. Walker and Darvall
If the pile is installed after removal of the excess sur- ( 1970) al so examined sorne of the sarne data as Dawson and
charge (i.e., T 0 ~ 0.15), only very small downdrag loads re- found that, as previously noted by Endo et al. (1969), Ta is
sult. The downdrag reaches a maximum value at sorne con- frequently closely-approximated by the original undrained
siderable time after installation (in this case, about Tv = 1) shear strength, cu, of the soil (Fig. 11.27). This fact a1so
and decreascs thereafter. A pile installed after Tv "'0.5 ex- emerges frorn an examination Jf Fig. 11.25, and may pro-
periences virtually no downdrag force. lt is interesting to vide a useful approximate means of estimating Ta in the
note that for T0 > 0.15, a considerable portian of the pile absence of other information.
rnay be subjected to upward stresses (positive friction) al- The estirnalion of drained Young's modulus E5 of the
though downdrag forces always ( xist in sorne position of soil presents sorne difficulties. The value apprupriate to the
the pile, general! y néar the top. case of negative friction is likely to be different from the
value appropriate to axialloacing (see Chapter 5). The cf-
fects of pile installati~n in the case of axíalloading are ve1 y
11.4. 7 Dat:1 on Pile-Soil Parameters importan!, but these effects a :e like1y to be greally rnndí-
fied by the consolidation of tf e soil. It appears likely that
The rnost important parameters in the analysis are the pile- Es wil! correspond more closely to the value for the s0il
soil shear-strength parameters and Young's rnodulus, Es and iayer subjected to surface loading, and in this case, m ay
Poisson's ratio, v5 , of the soil. The forrner parameters are of be conven.iently estirnated from the results of oedomctcr
great importance if full or partía! slip occurs, while Es :md tests vía thc following equation for an ideal elastic soil:
Vs only become irnportant when conditions are e!astic or
nearly so. In rnany cases involving soft soils, estimates of
the values and v3 wíll not be necessary. (11.23)
With nomlally-consolidated clays, it seems reasonable
to assume that c~/q = O unless the clay is overlain by other where
0.6
0.4 (
!.& 0.3
a~
0.2
0.1
FIGURE 11.25 Suggested design values of pile-soil adhesíon (after Dawson, 1970).
286 NEGA TIVE FRICTION ON END-BEAR!NG PILES
0.5 1 1
1 1 1 1 1 1
0.4 ¡-· -
~
!::. remolding o~ly
0.2 f- X Bi tu men treated
.::, Benton ite-slurry treatment
0.1 f-
+
o~~~--x~~x~--~~--~~---~~--~'-~'~~~---~
o 1o 20 30 40 50 60 70 80 90 100
Plasticity mdex
FIGURE 11.26 Píle->oil adhesion vs. plasticity inclex (after Duwson, 1970).
FIGURE 11.27 Summary of negative friction data in dimcnsionless form for sleel-pípe piles in clay (Walker and Darvall. 1970).
C~e = e~ + PoKs tant¡J~ The final maximum axial force in the piJe is tperefore
O + [(4 X 19.62) - (3 X 9.81)] X 0.3
14.7 kN/m 2 1502 + 126 = l628kN
The applied stress caused by the fill is The possibility of crushing of the pile should also be
checked. The maximum axial stress is 1628/(rr X 0.2 2 ) =
q 3 X 19.62 = 58.86 kN/m 2 12955 kN/m 2 . This is less than the normal compressive
:.c~.,/q 14.7/58.86 = 0.25 strength of concrete and hence crushing should not occur.
(19.62- 9.81) X 16
yLfq "-----=-::--::-,.....:._-- = 2. 6 7
58.86 b} Final Axial Movement of P.'le
The axial movement of the po~tion of the pile in the clay is
Applying Eq. (11.15), given by Eq. (11.19). From Eq. 11.20),
Since pile installation is delayed one year after fill place- From Figs. 11.14 and 11.15,
ment. T 0 = 18.5 X 1/16 2 = 0.072. lnterpolatíng between
Figs. 1!1.5 and 11.6forK,tan,P~=0.3,NR 0.90.•Because
conditions at the pile-soil interface a;e not e1astic, an estí-
mate of E, and v~ will no't be required, as there will be no From Fíg. ll.l6,
need to use the e1astic solutions in this case.
Assuming one-way drainage of the clay !ayer and inter-
polatíng between the available curves from 11.8, N T
=0.99. Now the axial force in the piJe at the top of the clay !ayer
1502 kN p = 7.1 mm
The downdrag caused by the overlying sand !ayer and (The component resultíng frorn the axial load of 626 k N is
fil1 must now be added. It will be assumed that e~ O, 4.>~ = 4.0 mm.)
0.7lrp' = 25°, and Ks = 0.5,so thatK 9 tanrp~ 0.23.lt will The compression of the pile above the top of the clay is
further be assumed that full slip occurs between the píle given, to sufficient accuracy, as
and these soíllayers.
At the leve! of the water tab1e, (500 + 626)/2 X 7
.0016 m
20 X 10 6 X rr X 0.2 2
4 X 19.62
L6mm
78.5 kN/m 2
The total final axial movement of the piJe head is therefore
At the leve! of the top of the clay,
7.1 + 1.6 = 8.7 mm
a~J ::: 78.5 + 3 X 9.81
e) l>-faximum Force and Defle.;tion After Six Y.:ars
107.9 kN/m 2 Six years after installation of the piJe, Tv = 18.5 X 6/!6 2
0.43. lt will be assumed that the downdrag force and de-
The downdrag force caused by these layers is then flectíon resulting from the fill and sand 1ayers have been
fully developed at this time and that only the downdrag
rr X 0.4 [(O + 78.5)/2 X 4 + (78.5 + 107.9)/2 X force caused by the clay is time-dependen t.
3] X 0.23 From Fíg. 11 .ll, interpokting between the two curves,
UN = 0.73. Applying (11 18), the maximum pile load
= 126 kN after six years ís
288 NEGATJVE FRICTION m.: END-BEARING PILES
'O
't 0.73(1 ,;28 626) + 626 J.3
1357 kN
0,25
\
pile·head movement after síx years ís )2
~
S.Omm
0.1
1 U PlLE GROUPS
In ¡:;roups of end-bearing piles in consolidating soil, down-
drag loads in individual piles are likely to be smaller than on
S
an isolatet'. píle, since the presence of additional piles tends d
to reduce the s01l se.ttlements within the groups. The inner
FIGURE 11.28 Interaction curve> for two end-bearmg piles sub-
piJe; nf a group may thus have considerably smaller down- jccted to negative friction.
dra¡; loads than the outer pilcs.
An analysis of the effects of negative friction on groups
may be madc by ex~:ending the elastic analysís for single
n
P¡ == P 1 Il ( 1 - a.d;¡) (11.25)
pi! es. combining the single-piJe analysis (Section 11.3 .3) j= 1
with the pile-group analysis (Chapter 6). A full analysis i*i
reqniies consíderation of each piJe in a group, the consider- where
ation of pile-soil displacement compatability at each ele-
mel!t of each piJe, and solution of thc subsequent equations P¡ downdrag load at pile i
for the specifl'cd boundary conclitions, for example, rigid P1 downdrag load in iwlated single pile
pile cap (in this particular case, unknown axial forces at the D'dij downdrag interaction factor for spacing between
pile heads will also need to be included in the set of equa- piles i and j
tious). Su:h an analysis will give thc stress and load distri- I1 denotes a repeated product
bution in each pile of the group. Allowance can also be
m a de for pile-mil slip or par tia! crushing of the pi! e. How- The above superposition rule is approximate on1y and
eve:, for relatively lage pile-groups, the number of equa- tends to undereslimate the pile loads, especially for stiff
tioHs to be solved m ay be large. piles at el ose spacings. It m ay be useful, however, as a guide
A simplified approach may be made by first considering to the load distribution wíthh a group. For example, for a
the interac:tion between two piles in a consolidating !ayer. square group consisting of four comer piles and a center
The rcsults of this analysis may be expressed in terms of a píle, with L/d = K = 1000, and the comer piles spaced
dowrdrag interaction factor cxd, where at five díameters, the downdrag load on the comer piles is
found to be 0.38 P 1 , and on the center pi! e, 0.19 P 1 •
The above approximate malysis takes no account of
pile-soil slip. For small group' in very soft clay, the consoli-
(11.24) datlon settlements may be sufficiently Jarge that slip will
occur along all piles in the group, so that there wi!l be no
An example of the relationship between a.d and dimension- downdrag reduction from a group effect. On the other
lesf: spacing is shown in Fig. 11.28 for L/d 25. The hand, for hrge groups of closely spaced piles, the presence
larger the valuc of piJe stiffness K, the larger the val u e of a.d of the píles will suppress the settlement of even very soft
(i.e., the greater the reductíon in downdrag load). It is also soils and it is likely that considerable reduction in down-
fot nd that ata given sjd, a.d in creases as L/d decreases. drag.Ioads will occur. However, as pointed out by Zeevaert
The two-píle analysis rnay be extended to syrnmetrical (1959), group effects may remlt in a decrease in the point-
pilE groups, and it is found that for such groups, the indivi- bearing capadty of a pile.
dué) pile loads may be estimated approximately from the Tests carried out on model píles by Koerner and
so1utions for two-pile interaction, as follows: Mukhopadhyay ( 1972) confirmed that the group effect de-
NEGATIVE FRICTION ON END-BEARING PILES 289
--:
V>
o.
e
o
·¡:; 30 •
(.)
:E • - - - - - - - - - } liHrnax e
e .g
~ (,)
E
'"> 20 •
~
"'
"' •
""e - - - - - - - - - { D.Hma< •
Vertical píle
2 3 4 5
Spacing día meter ratio, (~1
The effects of delayed installation are also present in these For the theoretical analysis, the following parameters were
cases. chosen or deduced from the published data:
E
10 J:::
a
o"'
la) Pile A
Pi le shortening (mm) Downdrag force (tons)
20
.S
_e
'5. _j
30 30 o"' '
(b) PileC
Theoretical
--- Meosured
FIGURE 11.31 Comparisons with piJe tests of Bjerrurn et al. (1969) at Heriiya.
NEGATIVE FRICTION ON END-BEARING PJLES 291
sponding measured distributions, a computer program was about 70 years and measurements showed that the conso!i-
used to obtaín the theoretical solutions. The resulting com- dation under this fill was complete. Three test piles (e, D,
parisons are s'.1own in Fig. 11.31. For both piles, the E) were installed in the newly filled area, and two (G, f{) in
agreement between measurerl and predicted downdrag-force the older area. Piles e
and G were analyzed, both being
distributions is good but the measured pile-shortening is Jess conventional uncoated steel-tube piles driven to bedrock.
than predicted. The irregularity in the measured force near For pile e, the following data were adopted after an
the tip of piJe A may be attributed to the fact that the examination of the available data given by Bjerrum et aL ·
lower part of the pile. was resting against steeply sloping
L 40 m, d = 0.5 m (8-mm wall thickness),
bedrock, thus causing load transfer to the rock by adhesion.
q 147.1 kN/m, "fsub = 9.81 kN/m 3 ,Ks tan <P~"' 0.20,
Es 9.81 MN/m 2 ,K= 1280,cv 80m 2 /yr,
Tests of Bierrum et al-Sorenga Site
t0 2.5yr,t=4.5yr.
At the second si te (Si:irenga), 1O to 15 m of fill overlay 30
to 40 m of day overlying bedrock. At one end of the site, The characteristics of the fill were assumed to be the same
fill had been placed a few years befare the piles were as at the Heroya site. Piezometer measurements indicated
installed. At the other end, the fill had been in place for two-way drainage conditions.
(a) Pi leC
lb) Pile G
- - - - Theoretical
-- - - Measured
FIGURE 11.32 Comparisons with píle tests of Bjerrum et al. (1969) at Sorenga.
292 NEGATIVE FRICTION ON END-BEARING: PILES
Comparisons between measured and theoretical distri- the Iatter layers. Three meters of fíll were placed ove.r the
butions of pile shortening and downdrag force for píle e are site and measurements were made of downdrag force, soil
shown in Fig. 11.32. which reveals a remarkable degree of settlement, and pore pressure in the soil.
agreement. The only point of disagreemenr is near the tip, In obtaining a theoreticd solution for this case, the
where the theory predicts a ~;maller downdrag force. The following parameters were selected for the el ay and piJe:
2
theory suggests that slip does not occur in this region, and L = 16.5 m (in clay), q 3.9 t/m 2 (38.3 kN/m ), 'Y!Ub =
the bearing stratum is assumed rigid in the analysis, whereas 9.81 kN/m , Ks tan rf¡~ = ü.4, Es
3
170 kgf/cm 2 (16.7
2
ín reality, the finite compressibility of the bearing stratum MN/m ). For the overlying sand, recent fill and fill mate-
may allow sufficíent movement of the pile to generate full rial, full slip was assumed to occur at the pile-soil interface,
slip near the típ. with Ks tan rf¡~ = 0.45 and = 22.6 kN/m 3 . Walker and
A t the location of píle G, fill had be en placed 70 years Darvall were of the opinion that éonsolidation under the
previously and consolidation under this fill was ahnost new fill had occurred very rapidly and that creep was in
certaínly complete. Nevertheless: a soil movement of about progress when the latest measurements were made (about
70 mm was measured in the vicinity of the piJe in a two- 250 days after commenceme 1t of filling). Accordingly, in
year period. It is reasonablc conjecture that this settle- the theoretical analysís, the measured soil-surface settle-
ment resulted from the threc-dimensional cffects of the ment of 35 mm was input in the analysis and the soil
recent filling at the othec end of the site. Calculations were settlement was assumed to va:y linearly with depth to zero
carried out for pi! e G, assuming that the settlement of the at the top of the sandy sílt lay ~r.
soil decreased linearly from 70 mm at the surface of the soil A comparison between measured and theoretical distri-
to zero at thc piJe tip. The píle and soil parameters were as bution of downdrag force in the uncoated piJe, shown in
for piJe C, except that for piJe G, L 30m. The resulting Fig. 11.33, again reveals good agreement. There ís also good
comparison between measured and theoretical downdrag agreement between the pres,ent theoretical solution and
force and piJe shortening, shown in Fig. 11.32, reveals very that of Walker and Darvall, which was based on a finite-
closc agrcement. This case provides evidence that large element analysis using a nonlinear stress-strain relationship
. downdrag forces m ay occur when soil movements are for the soiL It is significant to note that the present calcula-
caused by three-dimensional effects arising from loading at tions indicate that slip occurs only near the top of the pile
so me distance from the pi! e. in the clay. Thus, an estímate of downdrag force assuming
A further comparison was made with another pile (B), full slip along the whole length would give incorrect predic-
which was installed at the same time as the recent fill was
placed. The piJe and soil parameters were taken to be the Downdra<· force (tonnes)
same as for piJe C. except that filling and piJe installation o 50 100 150 200 250
were assumed to be simultaneous (i.e., t 0 = 0). Detailed 10~--~----~--~-----~-----~~-----~
measuremenls wcre not given for this pile, but the maxi- Sóíl
profile
mum downdrag force was reported to be about 400 tons
(3.99) MN), while the theoretical calculations gave a value
of 408 tons ( 4.07 MN), with full slip along the whole length
of the pile. The values of piJe shortening did not agree well
(100 mm measured, 29 mm calculated); however, the ~
«:::.~. . ~
average axial stress in the piJe was on the ordcr of 3200
)J
'-' " clay
kgf/cm 2 (314 MN¡m 2 ), and it is therefore probable that
yielding of the pile occurred, resulting in increased
--
shortcning of the piJe. /
-- --- --
-/
Test of Walker and Darvall Dense
A further documented case of downdrag measurement in an ·-
30
L . - - - - - L ._ ____!._ _ ___¡__ _ _ _s
__a__
nd_P'.:-.g_ra_ve_l__
end-bearing piJe has been presented by Walker and Darvall
--- Measured (Walker & Darvalll
(1973). Two stecl-shell piles of 0.76-m diameter and 11-mm
Predicted (Authors)
wall thickness ( one pi le uncoated, one coated with bitu- --- Predicted IWal cer 1'1 Oarval!¡
men) were driver\ into a soil profile consisting of 2 m of
recent fill and 7 m of sand overlying firm silty clay, sandy FrGURE 11.33 Comparison with pile tests of Walker and Darvall
silt, and dense sand and grave!, the piles being founded in (1973).
NEGATIVE FRICTION ON END·BEARING PILES 293
tions of both the magnitude and distribution of downdrag; san<l.y silt, sand, and grave! layers underlyíng the clay. [n
the use of the measured downdrag to infer Ks tan 1/J~ would the theoretical analysis, the pile was assumed to be end·
give an erroneous value. bearing on the sandy silt, and consequently the calculated.
The measurements indícate the development of positive downdrag force at this leve! is somewhat greater than the
friction and consequent reduction in downdrag force in the measured value.
PILES IN SWELLING
AND SHRINKING SOILS
12.1 INTRODUCTION ever, considerable uplift force s are induced in such piles
because of the action of the swdling soil.
Foundations in expansive clays are frequently subjected to An analogous problem arises with piles in soils under-
severe movements arising from moisture changes within the going shrinkíng or consolidation, when downdrag forces
el ay, with consequent cracking and damage caused by dis- are induced in the piles by neÉative friction. The effect of
tortion. Donaldson ( 1965) c]assifies the migration of negative friction on end-bearing piles has been discussed in
moísture bencath structures into three phases: Chapter 11. In this chapter, the effect of soil movements on
the behavior of floating piles will be considered. Beca use of
l. A primary phase caused by the erection of the struc- the similarity in approach betv·een a piJe in a swelling soil
ture, resulting in a changc in the moisture regime until an and a pile in a consolidating or shrinking soil, attention will
equilibrium state is reached under the new conditions. be concentrated on the case of a swelling soil, although
2. The fluctua tion caused by seasonal clima ti~: changes. sorne theoretícal differences tetween the two cases will
3. The results of extraneous ínfluences, such as broken be considered.
drains, leaking water pipes, local concentrations of storm- In designing piles in swelling soils, there are three
water, and gardening operations. requirements:
PUes have been used extensively for foundations in l. The pile must be able to carry the structural load safely
swelling soils in order to anchor clown the structure at a i.e. there must be adequate ultima te load capacity.
depth where changes in moisture content are negligible, 2. The piJe must have sufficient tensile strength to with-
so that movements of the struc:ture are minimized. How- stand the tensions developed in the pile due to uplift forces.
294
PILES IN SWELLING AND SHRINKING SOILS 295
-
Load
(12.1)
where
ra pile-soil adhesion
(12.2)
where
(b) Loaded pile
cffective adhesion
coefficient of horizontal pressure FIGURE 12.1 Forces in underreamed pile in expansive ~oil (¡¡fter
Collíns, 195 3).
effective vertical stress
effective angle of friction between pile and soil
tion. The probable variation of tension with depth is shown
Thus, in Fig. 12.1a.
h If a vertical load P is applied to the top of tht! pile,
T 1r Jd(c~ + Ksa~ tan <P~)dz (12.3) Eq. (12.3} is modified to
o
}¡
At the tip of the pile, there is zero tension (T 0), so P + T = 1r Jd(c~ + Ksa;, tan <P~)dz ( 12.4)
that there is a transition zone near the underreamed sec- o
296 PILES IN SWELL!NG ANO SHRINKING SOILS
Th~ probable load-distribution ís shown in Fig. 12.lb. fashion similar to that described in Chapters 5 and 6 for
Collins originally suggested that e~ and if¡~ should be piJe settlements and in Chap·er 11 for negative friction on
taken as the values of e' and ¡1/ from drained-strength tests end-bearing piles.
and that Ks could be taken as 1.0. Subsequently, Donald-
son ( 1967 b) suggested that· a reduction factor of between
0.3 and 0.7 should be applied to the measured drained 12.3.1 Basic Analysis
shear-strength to obtain the pile-soil adhesíon, r0 . A sum-
mary of suggested methods of obtaining ra is given in Sec- The problem is illustrated in Fig. 12.2. A circular piJe,
tion 12.6. length L, diameter d, and base diameter db, is situated in
Bozozuk (1972) described an analysis in which the a soil mass in which occurs, away from the piJe, a general
location of the "neutral point" (i.e., the point at which the specified distribution of movement, S ( either swelling or
maximum load occurs) was computed, assuming full slip shrinking), with depth. The píle is divídcd into n cylindrical
to occur along the pile. The maximum load on the pile dements, each with a uniform shear stress, Pj. acting on the
could thcn be obtamed. Ap extension to this analysis to periphery. In the basic analysís, the soil is assumed to be
allow for loads caused by horizontal stresses generated by homogeneous and !inearly el<<stic and it is assumed that no
an embankment was also descr.íbed. slip occurs at the pile-soil interface.
Sahzin (1968), in obtaining expressions for the move- Defining downward soil displacements and shear
ment of a pile in swelling soiL analyzed the work done by stresses as positive, as shown in Fig. 1 the soil displace-
friction forces in the upper pc-rtíon of the piJe, which tend ments along the piJe can be expressed as
to lift the pile, and the applíed load and the friction forces
in the lower half, which tend to resist uplift. Sahzín found
(sP} =- d [Is](p} + jS} (12.5)
good agrcement between piJe movements and those pre-
dictcd.from his approach.
where
The elements of [Is] are obtained by double integra- constant and the movement of the piJe subsequent to full
tion of the \.1indlin cq uations as described in Appendix A. slip will be that occurring at the point of shear reversa!.
Poissons' ratio of thc soi:l, lls, is nota very importan! para-
meter of the values of fs.
Without slip, the píle displacements pP = sP· For the 12.3.3 Compression Failure of Pile
more general case of a compressible piJe, the piJe displace-
ments must be compatible wíth the elastic properties of the Allowance can be made for cornpression fa(lure (see Chap-
piles and the analysi~¡ could proceed along the lines de- ter 11). If the load ata point i.n the pile reaches the crush-
scribed in Se,ction 5.3. He re, only the single case of an ing load, it can increase no further, and hcnce shear stresses
íncomprcssible pílc is considcrcd, and therefore pP = sP p below that point will reduce to zero and the shear stresses
=constant. It follows that above that point will redistribute, with a consequent
increase in pile displacement. The solution is rccycled until
(12.6) the load in the pile nowhere exceeds the crushingstrength.
app!íed downward load on pilc top l. For the upper fractured portion:
n
12.3.2 Pile-Soil Slip rrdL rrdl
Pi n + PbT =O (12.9)
j=m+!
The effects of pile-soil slip along the shaft can be allowed
for by specifyíng a limiting value of shear stress, fa, at
each element along the pile, and a limiting base pressure A new variable, the displacement of the fractured portíon
equal to the bearíng cápacity of the base. The solution of the pi! e, is now introduced, so that the n+ 3 equatioGs
must be recycled until the shear stresses and the base may now be solved for the n unknown shear stresses, the
pressure do not exo:eed the límiting values (see Chapter base pressure, and the displacem~nts Of the upper and lower
5). fa will usually be expressed in terms of effective stress portions of the piJe.
by the Coulomb expression. Second, the possibility of ten-
sien between the pile tip and the soil rnay be overcome by
specifying in the computer program that the base pressure, 12.3.5 Nonuniform Soil
Pb, cannot be positíve-that is if the pile ís of uniform día-
meter, the tip separa tes from the soil and hence carries no Approximate allowance for the effect of variation of. the
load. For an enlarged base pile, compressiv~ stresses will, modulus of elasticity of the soil with depth can be ma~e.by
however, act on the top of the enlarged poition of the base substituting the matrix [18 /E.d for [/8 ]/Es in Eq. (12.3).
if p b is positíve. It should be noted that if füll slip occurs By this procedure, the soil displacement at a particular
along a pile, the load distribution .in the pile wíll remain
- ""
depth will be calculated as if the modulus at that particular
. ·
298 PILES IN SWELLING AND SHRINKING SOILS
depth was also the modulus at all other depths. This ap- 1.0
prox.imation will clearly be unsatisfactory when the varia-
tion of modulus from top to bottom is very large or there
~
d
1
,, =
:,
0.3
0.25
time considered, and the solution carried out as before.
o
o 20 40 60 80 100
L
d
12.4 TYPICAL SOLUTIONS FOR PILE MOVEMENT FIGURE 12.3 Elastic solutions lor pile movement~uniform-dia
ANO LOAD meter piJe.
1
ment and the maximum load irr the pile for various length- 2.0
to-diameter ratios, L/d, of the piJe, and a number of soil-
movement profíles. In all cases, the soil movement is 0.6 /' -
assumed to decrcase linearly with depth from S 0 at the sur- 1
--
face to-zero at a depth Zs. Both a uniform diameter píle
1.0
(dbfd l) and a pi!e with an enlarged base (db/d 2) are
0.4 .....
considered. As long as elastic conditíons are preserved, the i""
solutions are applícable for swelling or shrinking soils,
except for a change in sign. In al! ~ases, 10 elements have
been u sed to divide the pi! e, and vs ís taken as O.3. 0.2 ? - 0.75
0.5
~
1
0.25
Pite Movement
The varíation of dimensionless pile movement, Pp/S 0 , with
o i
L/d and z5 /L, is shown in Figs. 12.3 and 12.4 for dbfd 1 o 20 40 6.0 80 100
and 2, respectively. These figures show that L
d
l. The movement increases as the depth of swelling FIGURE 12.4 Elastic so!utions for pile movement:-pilc with en-
in creases. larged base.
PILES IN SWELLING AND SHRINKING SOILS 299
10
_<l_t,_
~
L d -~
L 0.75
0.75
0.5 8 0.5
1.0
0.2
Pma:<~
E,ds0
QL-----~--------~------J-------L------J
o 20 40 60 80 100 20 40 60 80 100
L L
d d
FIGURE 12.5 Elastic solutions for maximum pile load-uniform· FIGURE 12.6 Elastic solutions for maximum pite load-piJe with
día meter pilc. enlarged base.
mum effect ís obtained when zs/L = 1 that is, the pile is between soil and pile or within the pile, the solutions for
founded at the bottom of the zone of movement. shrinking and swelling soils are no longer interchangeable.
Attention will be focused on piles in swelling soils, although
Maximum Pite Load sorne comparisons are made between the behavior of a pile
The variation of dímension1ess pile load, Pmax1EsdS0 , in a swelling rnd shrinking soil. Severa! of the solutions
with L/d and zs/L, is shown in Fig. 12.5 and 12.6. An given have been presented by Poulos and Davis (1973).
examination of these figures reveals that Negatíve values of pile force imply tension, while negative
values of Pp indica te upward piJe movement.
l. In most cases, the lárgest piJe load ís developed when
Zs/L is about 0.75. The Effect ofPile Length and Base Diameter
2. The maximum piJe load generally increases as L/d For a given soil profile with a linear distribution of soil
increases. swelling from S 0 at the surface to zero at a depth of lüd,
3. The presence of an enlarged base on relatively short the variation of piJe movement and maximum pile force
piles (L/d 20) results in significantly increased loads as wíth íncreasing soil movement is shown in Fíg. 12.7 for
compared wíth the case of a uniform-diameter pile. three different piJe lengths. For each length, both a
uniform-diameter· píle and a pile wíth a base diameter twice
the shaft diameter are considered. The píle-soil interface
12.4.2 Solutions Incorporating Pile-Soil Slip shear-strength, Ta, varíes linearly from zero at the surface
to O.OlEs ata depth of ~Od, and the base bearing-capacity
The above elastíc solutíons are useful in giving an indication is assumed to vary frÓm 0.36Es for L "' Sd, to 0.64Es at
of the influence ofpile slenderness, base diameter, and soil- L = 20d; these values correspond approximately to a soil
movement profile on pile behavior. However, for real soils, having q/ =30°.
these e las tic so1u tions ,are of very limited validity, as it is Figure 12.7 shows that, as would be .expected, pile
found that slip between soíl and pile commences at very movement decreases as pile length increases. When the pile
small soil movements. In order to illustrate the influence of is entirely situated in the swelling zone (L = Sd), movement
various factors on pile behavior, a number of solutions for of the pile continues after full slip has occurred along the
relatively idealized cases are examined. When failu~e occurs shaft. For piles founded below the zone of swelling, a limit-
300 PILES IN SWELLING ANO SHRINKING SOILS
So
~
--ó'-j ~ T,m
Pi le Soil Pile~soil
movement strength
-~0.06
~0.02
- - - - Enlarned
, base ' d ':!?. 2
ing piJe rnovernent ís reached after a certain soil rnove- enlarged base has the greatest influence when the pile is
ment occurs. The advantage of founding a pile below the situated at or near the bottom of the swelling zone, and
swclling zone is obvious. The rnaximum tensile load in the that the most efficient means of reducing piJe movements ís
piJe generally increases markedly as the length increases; either to use a uniform-díameter pile founded well below
relativeJy. smallloads are developed when the piJe is entirely the swelling zone ( of length abo u t twíce the depth of thís
within the swelling zone. zone) or to use an underreamed pi! e founded at or just
The preseHce of an enlarged base leads to a decrease in below the bottom of the swelling zone.
pile movement, although the effect is relatively small, For the uniform-diameter piles considered in Fíg.l2.7,
especially for L = 5d and L = 20d, or in other words, when the load distributions are shown in Fig. 12.8 for various
the pile is en tirely in the swelling zone, or ancho red well values of dimensionless soil movement S 0 /d. As the pile
below the swelling zone. In the latter case, the enlarged length increases, the load in the pile increases and the
base has virtually no effect. The corresponding maximum distribution of load also changes; the relative position of
loads are considerably greater for the enlarged base piles, the maximum load moves toward the top of the pile. For
except for the L = 20d pi! e. lt is therefore apparen t that the the L =5d and lOd piles, slip cccurs along the whole length
PILES IN SWELLING ANO SHRINKING SOILS 301
10d
L
E, 100 fam
Pi le Soil Pile-soil
movement strength
o...-------,-,----,
, 1\ Values of z
L
0.5 ~
1 "''
)>d 0.01
So
;;¡,QQ3
L
0.5
_P_
2
T;¡cn d
of the piJe at relatively small soil movements. For the L "' The Effect of Pite Shaft Diameter
20d pile, no change in load occurs after S 0 /d reaches about For a given pi le length and suil-swelling profile, the effect
0.12, but slip only occurs along the upper half of the piJe. of piJe diameter is shown in Fig. 12.9. As the diameter
L
(¡¡) Pi le movement d
L:
(b} Maximum pile load
FIGURE 12.'1 Influence of pi! e diameter.
302 PtLES IN SWELL!NG AND SHR!NKING SOtLS
~~ ~ ~
ITl\~-~~ r· j)
L= ~lu L~~
2
-;.J ~d ~~
1
01 (ií)
r
(ííí)
Dístríbutíons e! soíl
Píle-soíl 0.008
swelling wíth depth
strength
Po
E,= '100 Tam d
~
(iíi)
--0.006 -7.5
(ííi) (í)_./
-0004 /
/ .,.,- -- 0.1 0.2 0.3
Pp
///¡;;)
d So
1 1 d
-0.002 -2.5 1 (a) Pile novements
1
1
o ______j ___ _ j o
o . 0.05 ' -o. 1o o -0.05 -0.10
S,1 So
d d -15
FIGURE 12.10 lnfluence of soíl-swelling profile.
-10
\ncreases, piJe movement decreases, but the maximum load (ii)
increases. However, the rate of decrease of pi! e movement íamd2 /
--------
is almost negligible for diameters greater than about 0.03L, /
-5 /
and even a slender pile (d 0.011) moves only about 20%
1
,more than a relatively large-diameter piJe (d =: 0.2L). The 1
theory therefore suggests that small-diameter piles founded
well below the swelling zone can satisfactorily suppress up-
ward movements in swelling soils. Donaldson (1967a), has
s,
described the successful use of small-diameter piles to d
support conventional brick buildings on expansive soils in (b) maximwn pile loads
South Africa.
t
0.1
load required to prevent upward movement is only about
E,= 100 r,, half of the maximum tensile load developed in the un-
~~d ~ ioaded pi! e. Also shown in Fig. 12.12 are the pile move-
Pile-soil Soil
strength movement ments calculated on the assumption that the effects of axial
load and soil swelling can be superposed. Although super-
position is not strictly valid in this case, beca use ~lip occurs
along part of the pile shaft, it nevertheless provides an
approximate estímate of píle movement. Taking account of
Po pile-soil slip in the solution for axial load would lead to
d increased movements and closer correlation with the com-
plete solution.
, Superposition of solution
" lar swelling only and
" elastic solution for
Complete "axial load only 12.4.3 Effect of Tensile Failure of the Pile
solutíon "-
-0·15 -0·15-
-0-10 -0·10
Uooar Portian Uopar Port1on
Pp
d
-0·05 Lowar Portio.o. • 0 ' 05
----
~,....-_,...._..
o
o -0·05 -0·10 -0·15 o o -0·05 -0·10 -0·15
~ .§;¡_
d d
(a) Tt/E5 d
2 =0·2 (b) Tt /E 5 d 2=0·3
Pmax Pp
L/d db/d -¡¡ pile in a swelling-soil profile Both a uniform pile-soil shear
E¿¡L.'
strength, T a, and linear! y increasing r a with depth, ha ve
Swellin~ Consolidation Swelling Consolidation
been considered. These curves are shown in Figs. 12.14 and
S -0.0096 o 0173 -0.0556 0.0547
12.15 for línearly increasing Ta with depth, as functions of
2 -0.0196 0.0196 -0.0525 0.0509
10 1 -0.0385 0.0543 -0.0179 0.0164 SoHs
the dirnensionless rnaximurn soil·rnovement, - d and the
2 -0.0785 0.0785 -0.0139 0.0081 1 am
20 1 -O.Q785 0.0503 -0.0048 0.0024
2 -0.0785 0.0503 -0.0047 0.0023
a Soil movcmcnt decreases línearly from S 0 at surface to zero at
depth 10d. Valuats of Zs/L
2
0·20
12.5 DESIGN CURVES
Dimensionless curves have been prepared to enable rapid FIGURE 12.15 PiJe movement in swellíng soíl. Línearly íncreasing
estirnates of the rnaximurn pile load and movernent for a pile-soil shear strength with depth
PILES IN SWELLING ANO SHRlNKlNG SOlLS 305
L
p FS = Jo' Tarrddz {12.10)
1·0
Val u es of 2 5 / L
2 fllustrative Example
05
r- --- -.;¡
. 1
-· -- A 15-ft concrete pile, 1 ft in diameter, is situated in a soil ín
which the upper 1O ft is subjected to seasonal movement.'
·- -·
r-· ---..;
L_
·- ..--o-6_7¡ --- --- -- The maximum soil heave is 3 in. The maximum force and
the movement of the pile will be estimated for two condi-
0·2 r-·
'<>:9'"~ -- ·- tions:
p
'5'o 0·1 ~--.-=· l. Zero axial load.
2. An axial load of 30 kips.
005
The following data is assumed:
- - db/d=1
0·02 dt/d = 2
Es 2000 lb/in.Z 288 kíps/ft 2 ( constant with depth)
v9 = 0.4
0·01 Ta = 10 lb/in. 2 1.44 kips/ft 2 (constant with depth)
o 0·2 0·3
Ep 3 X 10 6 lb/in. 2
FIGURE 12.17 PiJe movement in swelling soíL Constant pile soil (a) Zero Axial Load
shear strength with depth. From Eq. (12.10),
306 PILES IN SWELLING AND SHRINK!NG SOILS
15
30,000 X 0.107
PFs - J 1a1Tddz
p
!2X 2000
o
0.13 in. (downward)
L44XrrX1Xl5
- 67.8 kips Thus, the net movement camed by axial load and swellíng
is 0:13 0.57 = -0.44 in. (i.e., net upward movement)
From Fig. 12.16, fordb/d =1, zs/L =10/15 0.67 and
e:)G:) =SO.
F = factor indicating the relative decrease in heave 12.6.2 Püe-Soil Interface Strength
at depth z to that at surface
PE potential expansiveness lt is commonly assumed that the shear strength between
z 1 , z2 = depth to bottom and top of expansíve !ayer soil and pile increases with depth, approximately in pro-
(h=z¡ z2) portian to the overburden pressure. In situations where the
soil is normally consolidated and is going to remain satu-
rated (zero air voids but not necessarily posítive pore water
T ABLE 12.2 VAL UES OF POTENTIAL pressure ), the limited experience from measurement of
EXPANSIVENESS (VAN DER MERWE, 1964) negative friction on piles in soft, normally consolidated
clays (see Chapter 11, Fíg. 11.25) may be taken to be rea-
Degree of Expansiveness PE in./ft of soü
sonably applicable to the case when the pile is affected by
Very high 1 shrlnking or consolidation of the soil. This experience sug-
High 0.5 gests that, in the equation
Medium 0.25
Low O
(12.13)
where
F is giveni>y 1
Ca adhesion
F ¿/k (12.12)
308 PI LES l:"J SWELLING AND SHRINKING SOILS
rp~ effective pile-soil fric1ion angle of K 0 for overconsolidated soils, a higher value of Ks may
a~ effective overburden :;tress be expected. South African experience (Collins, 1953;
Ks a coefficient of horizontal earth pressure Baikoff and Burke, 1965; Donaldson, 1967b) appears to
suggest in such situations a range in Ks of 1.0 to 1.5. The
the combined term Ks tan !/J~ usually lies within the range cohesion terrn also is not necessarily negligible. For soils
0.2 to 0.3, and that e~ can be neglected. For !/!~:::::: 20, this swellíng from an initially unsaturated condítion, the effec-
corresponds lo a range in Ks 'Jf 0.5 to 0.8, or somewhat tive-stress concept is of doubtful validity, and as a practica!
greater than the coefficient of earth pressur.; at rest, K 0 • approach, ít may be better to use the total instead of the
For swelling situations, especially when swelling occurs effective overburden stress in Eq. (12 .13) and appropríately
under relatively arid conditions, the soil rnust be in an over- deterrnined va1ues of adhesion and pile-soil friction angle.
consolidated state, and corresponding to the higher values For exarnple, Collíns (1953) recornrnends the use of a value
10
Depth
( ft)
Clay with
slíckensídes
20
30
5 10 15 o 5 10
Yellow E, (lb/in') S (in.)
bedded
el ay
40 fa) Assumed parameters in analysis
Varved
el ay
Organíc clav
50 Shale
S01l profile
Pile
length
(ft)
J Píle
length
(ft)
20
30~L----~------'
- - - Theoretícal
0 Measured (Co:líns, 1958)
13.2 ANALYSIS
100
rr: (Es) and Poisson's ratio (vs) that are unaffected by the
presence of the pi! e. The stcesses developed between the
piJe and the soil are assumed to act normal to the face of
the piJe, and no account is ttken of possible shear stresses
• horiz:::mtai movement d:t pi le ! op 1
set: lr.ment ot enba nk ment developed between the soil ar.d the sídes of the pile.
of In order to better approximate real soil · behavior, ít
0.1 0.2 0.3 0.4 is assumed that the Young's modulus, Es, may vary along
the pile and that the horizontal pressure between piJe and
soil (i.e., the dífference between the soil pressures on the
FlGt:RE 13.1 Rclativc disp!acement as a func'.ion of relativc !1exi- opposite faces of the píle) ha~; a limiting value, Py, that m ay
bilíty (after Marche and Lacroix, 1972). (Rcproduced by pcrmission
al so vary along the pi! e.
of the Nationa! Research Council of Canada, from thc Canadian
Geoteclmica/ Journal, vol. 9, 1972, pp.l·-24.) A solution to the problem is obtained by imposing dis-
placement compatibility between the pile and the adjacent
soil. The piJe displacements are obtained frorn the equatíon
examination will be made of sorne cf the factors in· of flexure of a thín strip. By writing this equation in finite-
fluencing the devclopment of piJe moments and displace- dífference form for each node point along the piJe, the piJe
mcnts, and sorne comments are made regarding values of displacements may be expres:.ed as
t:J UJ
{Pe}
vector of values
)
ments are equated:
0.4
( 13.3) z
L
06
where
In addition to the above equations, the horizontal-load FIGURE 13.3 "Standard" soil-movement profile for theoretical
and mornent-equilibriurn equations may be written in terms soiutions.
314 PILES IN SOIL UNDERGOING LATERAL MOVEMENT
13.3 TYPICAL RESULTS are assumed (Figs. 13.3 and i3.7). Actual distributions can
usually be approximated in .his way. The distributions of
In order to examine the effect of various factors on píle Es and Py along the pile are :aken to be either constant or
behavior, a number of solutions ha ve been obtained for linearly increasing with depth. Twenty-one elements have
idealized cases. Simple distributions of horizontal move- been used to divide the pile.
ments with depth in the soil nass in the vicinity of the píle
1
1
''
L 0.5
j !-
L ....
''
""',',...
'l
:\
1
- - Soil
- - - Pile Kq :e 10· 5
3
- - Pile KR 10
PileKR 10
z_
L
13.3.1 Effect of Relative PiJe Flexibility ¡o·s (a very flexible piJe), 10"3 (a pile of medium flexi-
bility), and 10- 1 (a relatively stiff piJe). For the very
For a free-head, pinned-tip socketed pile in a soil having flexible pile, the piJe displacement follows the soil dis-
constant modulus Es and yield pressure Py, and subjected placement almost exactly, and in consequence, small
to the "standard" externa! soil movements (Fig. 13.3), the pressures and moments are developed. A~ the pile becomes
distributions of pile movement, pressure, and bending- stiffer, pressures and moments in crease and the pile-
moment are shown in Fig. 13.4 for three values of KR, movement distribution changes. It is ínteresting to note
_l, = 25 v, = 0.5
d Socketed píle,
Constan! p, and E, restraíned ¡:;:nned-h,<ad,
pinned t1p
~-'-
p
= 10 E.
d p, P,
o1 02 03 04 05 1.0 o 1.0
o
---~',~
0.--.--,--r--~~--~--r--,-~---~
~:--1 "1 1 1
""-. -
. --~~ -
' '~
-
\ ~)>
l 0.5
/
b
'/
-
-
-
.¡_
L 0.5
----
¿
;:::? -
/ -
1.0 l/" 1 1
3
---Soíl - - - Píle KR = 10-
5
Pi le K R = 10- - - - Pile KA= 10·
!e
L
that in this case, where no head restraint is provided, the 13.3.3 Effect of Soil-Movement Distribution
rnovernent of the top of a stiff pile is substantíally greater
than the surface soil rnovernent, so that it is an advantage For the sarne piJe as that considered in Fig. 13.6, with al1
in this case to ha ve a more flexible pile. · unrestrained pinned head and a inned tip, movement and
Corresponding distributiom of rnovernent, pressure, and rnornent distributions for three soil-rnovement profiles are
rnornent for a píle with a fully-restrained pinned head are shown in Fig. 13.7. In al! three cases, the maxirnum move-
shown in Fig. 13.5. In this cas·~, íncreasing the píle stiffness rnent is the sarne. For relatively flexible piles, such as those
reduces the subsurface rnovernents of the pile, but leads to considered here, the head movernent is largely dependent
generally greater pressures and mornents than in an Un· on the magnitude of the soil surface movement. The
restrained pile, as well as to a relatively large restraining maxirnurn moment in the pile is greatest for the uniforrn
force at the head. soil-rnovernent profile. The momeWts tend to decrease as
the soil-movernent profile tends to a triangular distribu-
tion with zero rnovement at the base of the !ayer and maxi·
13.3.2 Effect of Boundary Conditions murn movement at the top of the !ayer; in the latter case,
the pile and soil rnovements are identical, if the piJe is
An exarnple of the effect of head and tip boundary- pinned at the tip and unrestrained at top, and no rnoments
counditions is shown in Fig. 13.6 for a pile of rnedium are developed in the piJe.
flexíbilíty (KR = 10-3 ) subje:cted to the standard soil-
rnovernent profile. As previously indicated, the provision
of head restraint reduces the pile rnovements near the 13.3.4 Effect of Magnitude of Soil Movernent
surface, but also increases the rnornent. The provision of
tip fixity, in this case, has virtually no effect on pile rnove- Fig. 13.8 shows the variation in displacement and'moment
ments except near the tip, wh<:re a relatively large rnornent distributions along a free-head socketed pile with increasing
is developed. magnítude of soil movement. In this case, conditions re-
o~-
0.2
0.2~-
0.4 -
1 -t 0.41
/
0.6 e-
1 /
// 06 ~
1 _,/'
1
08~
0.8 f-
1 ,/
o
o
0.2 0.2
0.4 0.4
z
l.
L l
0.6 0.6
0.8 0.8
V0.45d V
FIGURE 13.7 Effect of distribution of soil rnovement.
. Soil-movement
KR 0.001 _[, = 25
d
v, = 0.5 S=
p,
10 . disr.ribution
0.4
l z
l l
0.€
317
318 PI LES JN SOJL UNDERGOJNG LATERAL MOVEMENT
Cor'Jsta'it Es a;,c E
v, c. 0.5 10 ~ ~ 8000
P.. .X ll
L
0.4
l
L 1
0.6- //
. 1 .;?
:: ~-'-':.r)
_J____..J.___jc.___l.___j
cJ (/¡) Momcnt·;
.! ! ~ .06L
e
d
~
25 ·r o*fl~ /
Soil-r1ovement
d1st· 1but1o~
V
FIGURE 13.9 Eflect 0fpíle díamclcr.
main elastic until the maximum soil movement reaches 13.3.6 Effect of Es andpy Di~tributions
about 0.4d. For greater soil movements, the elastic solu-
tíon tends to overestimate both pile det1ection and mo-
ment, although the diffcrences between the purely elastic The foregoing solutions have heen for the case of constant
solutions and those including yield are r,ot great, even for soil modulus E, and yield pcessure Py in the soil ]ayer.
relatively large magnitudes of :mí! movement (e .g., Psm/d"' Typical solutions for a sockeled piJe in a soil with Es and
0.9). Thus, Fig. 13.8 indicates that in some cases, especially Pv varying linearly with depth from zero at the surface, are
those involving relatívely flexible piles and small soil move- shown in 13 .10. In this ose, the pile-11exibility factor
ments, an elastic analysis may be adequate or at worst, is defíned as KN = E plp/Nh r, 5 , where N11 is the rate of
should provide an overestimate of piJe deflection and increase of Young's modulus with depth (see Chapter 8).
moment. For a given value of KN, the pile-movement pro file appears
to be similar to that for the case of uniform but for a
pile with a larger KR value. For example, the displacement
profile for KN = 1o·.> (in a linearly varying Es) is very
13.3.5 Effect of Pite Diameter similar to that for KR = 10-1 (uniform see Fig. 13.4).
The moments are expressed in dimensionless form as
If the diameter of a pile is changed but the length remains M/PybL 3 , where Pyb is the soil-yield pressure at the leve! of
constan!, the L/d ~atio changes and the pile-flexibi!ity the pile tip. A comparison b1~tween the moments in Fig.
factor, KR, also changes. A typical example of the effects J3.9b and those for a uniform soil in Fig. 13 .4c reveals that
of changing the diameter of a socketed pile is shown in Fig. for comparable values of Py in each case, considerably
13.9. Eomparison with Fig. 13.4 shows that the principal larger moments are developed when Py is uniform. These
effect of changing diameter is to decrease KR. The effect of larger moments arise because the values of Py along the
changing the value of L/d is of secondary importance. upper portien of the pi! e are considerably larger, and in
PILES IN SOIL UNDERGOING LATERAL MOVEMENT 319
l;;.
d
25 v, = 0.5 E, and Pv linear! y increasing wíth depth
. Py
~ 10
Socketed pile •. free~head, pinned-tip
Standard soil~moveme~t profile
M 4
p X 10
3
Pyb L
d
0.2 0.4 0.6 0.8 1.0 o 2 3 4 5
o
Values of KN
0.2
0.4
;:
L l
1 L
- - - Kn = 10-
- - - KR 10- 3
--- KR 10" 5
FIGURE 13.10 Pile in soil with lincarly varying modulus and yie!d pressure.
general, the dístríbution of Py has a greater ínfluence on that the predicted values were often very much smaller than
pile behavior than does the distributíon ofEs. the observed, despite the fact that reasonable agreement
was obtained between predicted and observed settlements.
Because predicted horizontal movements are much more
sensitive than settlements to such factors as soíl anisotropy
13.4 APPLICATION OF ANALYSIS TO PRACTICAL and nonhomogeneity, it appears unlikely that horizontal
PROBLEMS movements can be predícted with confidence at the present
time. Thus, it ís desirable, if possible, to use fíeld measure-
In addition to data on pile geometry and boundary condi- ments of horizontal movements as input data. Such
tions, the analysis requires the following data as input: measurements shoul¡:l be for the surface at least, but prefer-
ably should include values at depth obtained with inclíno-
l. The distríbution of horizontal soíl movement Pe wíth meters or slope índícators.
depth.
2. The distríbution of soíl modulus Es wíth depth.
3. The distribution of soil yield pressure Py with depth. 13.5 COMPARISONS WITH FIELD MEASUREMENTS
The estimation of Es and Py has been discussed previously, There are only a few reported measurements of pile be-
in Chapters 7 and 8. However, sorne comments on horizon- havior in the presence of lateral soil movements in which
tal soil movements are necessary. sufficíent data is available to enable comparisons between
Ideally, the horizontal soil movements resulting from an observed and predicted behavior to be made. Three such
embankment or foundatíon can be estimated either from series of measurements are considered herein.
the appropriate elastíc theory if the soil profile is reason-
ably uníform (see, for example, Da vis and Taylor, 1962), or Heyman and Boersma ( 1961)
from a finite-element analysís in more complicated cases. Heyman and Boersma (1961) ha ve described tests on three
However, a number of comparisons between measured and instrumented steel-box piles, 30 cm square and 12.5 m
theoretical horizontal movements (Poulos, 1972c) revealed long, founded in sand, peat, and clay layers underlain by
320 PILES IN SOlL UNDERGOJNG LATERAL MOVEMENT
Dístance from toe of slope (m) Dístance from toe of si:Jpe (m)
O
o.--------r--------,----------,.
20 o 10 20 30
o
5 25 o
8
o
e
"'E -
o
E ro ".<:""' 5
E 2'
:J
o.
s
X
;;;
e
:;;;" ,g
15 u 7.5
cr::"'"' o
10
o
Calculatcd (const. E,)
12.5 ' - - - 4 - - - - -
Calculated (lín. var. E,)
(b) Reaction at píle head
o Measureu IHeyman and Boersrna, 1961}
FIGURE 13.11 Comparison betwec:n measured and theoretical pite moment and reaction.
sand at about 11 m. The piJe heads were propped against a Comparisons between theoretical and measured maximum-
heavy concrete beam founded on an eight-pile bent to moment, 1Wmax• and pile-hea(~ reaction, H, are shown in
restrain hcad movement. A 7-m-high road embankment of Fig. 13.11. The theory tends to overestimate M max and
hydraulíc fill was constructed in stages, with the embank- underestimate H, although the variation with distance of
ment toe originally 30m from the píles. The distributíon of the pile from the toe is reasonably well-predicted. The
lateral soil movement wíth depth was measured by an assumed distribution has relatively líttle influence on the
inclinometer, while readíngs of the moment distribution solutions. For constant E 5 , the value of Es also has rela-
and the heau reaction in the piJe were also recorded. The tívely little influence on the solutions; a larger Es leads toa
embankment was progressively extended in the direction of smaller H and M max, but e ven a four-fold in crease in Es
the piles in steps of 5 m, with readings being taken at each only gives a decrease of about 12% in Mma.x and about 8%
step. A period of two to three weeks rest was allowed at in H.
each step, but relativdy little time-dependency of moment
or head reaction was noted. Heyman (1965)
In obhiníng theoretical solutions, the measured soil The test piles and soil pro file ::lescribed by Hcyman (I 965)
movements were used as input and the following assump- are very simil.ar to those described by Heyman and Boersma
tions were made: (1961 ). Two test pi! es 12.5 m long were instrumented, one
situated at the toe of the embankment and the other at 12
l. Both the piJe tip and the pile head were pínned but re-
m from the toe. The embankment was constructed in three
strained from moving horizontally.
stages to a maximum height of 4 m. The measured horizon-
2. The values of yield pressure Py were assumed to be 9
tal soil movements were not described in detail, but it was
cu for cohesíve soils and three times the Rankine passive
stated that for the full embankment heíght, the movements
pressure for cohesíonless soils, as recommended by Broms,
were almost constant with depth, varying from 1.5 cm at
1964a,b (see Chapter 7).
12m from the toe to 3 cm at the toe.
3. Two distributíons of Es were used:
In the theoretical analysis, the above soil ¡novements
(a) C'onstant E, with depth of 500 t/m 2 .
were used as input and the same assumptions were made
(b) Es varying linear! y from zero at the surface to 1500
re&ardíng soil and píle paramcters as in the Heyman and
t/m 2 at the leve! of the pile tip.
PILES IN SOIL UNDERGOING LATERAL MOVEMENT 321
Mom-mt (tm.)
- - Theory-lin. var. E, o
- - - Theory-const. E, 1
- l > - Meosured
1
.e
Cl
•¡;¡
/ 2.0
.c.
~
e
E 2.o -
-"'
e Oepth
"'
.o (m) 4.0
E
LU
Pite 11
6.0
- - - Lin. var. E,
5.0 - - - Const. E,
- - - 6 - Measured
Reaction force at pite head (t)
E 3.o ··
.e
Cl
2.0
·¡¡;
.<:
~ 2.0 -
E
-"' Oepth _
e 40
"'
.o Pite 1 lml
E Pite 1
UJ
6.0
Boersma tests. It was further assumed that the horizontal Leussink and Wenz ( 1969j
soil movements íncreased linearly with embankment height. A test pile was built up of four channel sections to form a
Comparisons between theoretical and measured head- box 0.85 m wide and 30 m long and installed in a soíl
reaction are shown in Fig. 13.1:2, while the moment distri- profile of sand and organic clay and peat overlying sand at
butions for the maximum embankment height are shown in 20 m depth. The pile head was hinged to an almost rigjd
Fig. 13.13. The theoretical solutions for head reaction, as- súpport. A rectangular test-embankment of ore was con-
suming a linearly varying Es with depth, are in excellent structed to a maximum height of about 6 m and measure-
agreement with the measured values, while the solutions for ments taken of the horizontal soil movements at various
constant Es overestimate H. Reasonable agreement is also locations beneath the embankment and of the test pile
found between the rnoments when a linearly varying Es is situated adjacent to the embankment.. It was found that
assumed in the theory, although there is a tendency for the the soil movements were sufficiently large (maximum of
theory to underestimate the moment in the lower portian about 80 cm) to cause failure of the test pi! e. Soil move-
of each pile. ments taken just prior to failure enabled a comparison to
322 PILES IN SO!L UNDERGO!NG LATERAL MOVEMENT
14.1 INTRODUCTION lutíons from this approach and the subgrade-rcaction analy-
sis are described.
Early investigations óf the buckling of pilcs (Granholm,
1929) showcd that for piles of normal dimensions driven
through soft soil, buckling should not take place except in 14.2 FULL Y EMBEDDED PIU:S
extrernely soft soil. However, with the increasing use of
very slendcr piles (e.g., Bjerrum, l95i Brandtzaeg and
Harboe, 1957) and long piles that cxtend for a considerable l4.2.l Basic Subgrade-Reaction Theory
distan ce above the ground !in e, the possibility of buckling
has had to be reconsidered, and a considerable amount of For an elastic pile ernbedded in a "Winkler" rnedium and
research has been carried out in recent years in an attempt subjected to an axial load P at the head, the governing dif-
to obtain more accurate estimates of pile-buckling loads. ferential equation for the horizontal deflection p along the
The majority of analytical methods proposed have employ- pile is(e.g.,Hetenyí, 1946):
ed subgrade-reaction theory, although elastic-continuum
theory has also been used more recen ti y.
In this chapter, the subgrade-reaction analysis is des-
o (l4.1)
cribed and available solutions for the elastic buckling load
of fully-embedded and partially-ernbedded piles are pre- where
sented. The effects of various practical complications such
as piJe imperfections, inelastic buckling, and group effects Epip flexúral stiffness of pi! e
are also discussed. A brief discussion of the elas tic approach kh modulus of subgrade reaction
to the problem is given and sorne comparisons between so- d piJe diameter or width
323
. 324 BUCKLING OF SLENDER PILES
j_. ~.,
buckling load is the lowest value. Use ·of the numerical
approach enables any distribution of axial load and modu-
lus of subgrade reaction along the piJe to be considered.
A number of solutions have been obtained for various
combinations of head and tip boundary condiiíons and for
the cases of constan! kh with depth and linearly increasing
FIGURE 14.1 Finite-difference representation of pile. kh with depth. These solutions are described below. The
influence of various practica] complications is considered in
Section 14.4.
The solution to the above equation involves the determina-
tion of the characteristic values of P for instability of the
pile. 14.2.2 Solutions for Constan! kh
Analytical solutions can be obtained using variational
methods (Timoshenko, 1936; Toakley, 1965; Reddy and One of the earliest solutions wa~: obtained by Timoshenko
Valsangkar, 1970). Alternatively, finite-difference methods (1936) for a pin-ended pile alo:1g which no load-transfer
can be used to determine the eigenvalues. For a typical occurs. The critica! load, Pe,., is given by
point ion the pile (Fig. 14.1), Eq. (14.1) can be written as
(14.4)
Pí-2 + [Po(;) 4] Pi-1 + [6 (l4.2)
where
rr Eplp
where
L piJe length
m number of buckled half-waves
n 2Eplp
P0 axii:tlload at pile head The value of m is obtained from the condition that Pcr is
P¡ axial load at point i a mínimum. Thus, when ~ = O, Pcr is a mínimum for m= 1
khi "' modulus of subgrade reaction at i and Pcr = PE. As ~ increases- that is, k¡¡ increases-the
khr a reference value of kh number of buckled half-waves, and hence Pcr• increases.
4 4
R = Eplp/L kh,.d For an infinitely long pi! e, buckling occurs in half-wave-
n = number of elements in piJe lengths of
and
(14.5)
Legend
(14.6) f free
p "~ pinned
ft fixed, tra:1sla_1íng ·
J.R
52,000 lb/ín. 2 , 1,-,oo;
buckling is only likely if fp/A 2 <O .30. This only occurs for FIGURE 14.2 Buckling load vs. length for kh constan! (Davísson,
shapes such as round and square steel-bars and tram-rails 1963).
such as are used in underpinning operations.
Solutions for various boundary conditions have been
Equivalent length, le, is a function of the boundary condi-
presented by Davisson (1963). The axial load is again as-
tions at the pile top and tip, and of the distríbutíon of k~¡
sumed to be constan! in the pÜe~that is, no load transfer
along the pile. Solutions for left' are shown in Fig. 14.3 for
occurs and the pi! e initially is perfectly straigh t. The solu-
various boundary conditions, where
tions are shown in 14.2 in dimensionless form, as a plot
of Ucr versus lmax, where
1' = (14.11)
PcrR2
Ucr --- (14.7)
E:Jp that is,
Figure 14.2 shows that the boundary conditions exert a Buckling loads calculated from the use of Fig. 14.3 are
controlling influence on Ucr• with the lowest bucklíng loads identical with those from Fig. 14.2.
occurring for pHes with free (unrestrained) ends. Methods of determining k~¡ and typical vaiues for var-
An alternative presentation of solutiorrs has been given ious soíl types are discussed in Section 8.2.
by Francis et al. (1965) and Toakley (1965), in which the
pile is considered as 2: pin-ended strut in air of equivalent
length, le, so that the buckling load is 14.2.3 Solutions for Linearly Varying kh
11
(O)
f'-!
Pilo:
(b)
tix~d top
1o•nn~d
(e) 'd)
PtiQ: trw to
and bottom sway at top,
FIGCRE 14.3 Piles in uniform m<dium; various end conditions; plot of le/r vs. Lit (Francis et al., 1965).
where
PcrT
Ef>lp
(14.12)
\J p·p
p
ft
l free
= pinned
lixed, translating
Note: Upper end
conditíon losted
fírst
(14.13)
L
Zmax (14.14)
r T
( 1.6 , . . . - - - - , - - - - , --...,..--·---,---,
1
Lu
L
:t SR
(o) Actual Pd<z ( b) Equovol<?nt Cantoi<?V<?r FIGURE 14.6 Dimensionless deptí1 of fixíty for bucklíng. Constant
k h (after Davisson and Robinson, 1965 ). (© Canada, 1965, by Unív-
FIGURE 14.5 Partially embedded piles. ersity of Toronto Press.)
(14.16)
1.6 Lo_ _.....L2_._.J4L__..J6L---:a:------:1o:::---~12
Ls equivalen! free length of embedded portian of
piJe (see Fig. 14.5)
Lu "' unsupported piJe 1ength FIGURE 14.7 Dimensionless depth of fixity for buckling. Línearly
varying kh (after Davisson and Robínson, 1965). (© Ca nada, 1965,
and R is defined in Eq. (14.8), with L embedded length. by Universíty ofToronto Press.)
328 BUCKLING OF SLENDER PlLES
Lu
Jr = - (14.19) Fíxed-fixed
T
2
rr Epfp
(14.20)
4 (ST + hYrt
Lee (1968) carried out model tests on ~-in.-to-n-in.
diameter piles in dry sand and found good agreement be-
tween the rneasured buckling loads and those predicted·by
Davisson and Robinson's solutions.
and
(l4.2lb)
where
For both cases, the value of ¡J; could be estimated from the This aspect has been investigated by Francís et al. (1965).
elastic solutíons for base load given in Chapter 5. If the soil is assumed to have a constant modulus of sub-
grade reaction, kh, that is independent of lateral deflectir ·..
a pile free of imperfections remains straight until the criti-
. cal load, Pcr• is reached. A pilt with imperfections deflects
laterally from the onset of loading into a form governed by
the imperfections but gradually moves into the theoretical
bucklíng mode as Pcr is approached.
2.0
1.5
Free-free
0.5~--------<J..----o-----o
> 1.0
(a)
o
2.0[
1.5
Fixed-free
with sway
>" 1 .O
f'ixed-··'ree
1---~~-w~i~th~s-w~ay~-o-----o
Free-free 0.5
o
oL--~--~----L--._J
0.4167 0.8333 1.250 1.4667
(b) "'
(b)
FIGURE 14.9 Effect of skin friction for partially supported pile FIGURE 14.10 Effect of skin friction for parti~lly supported pile
for constant soil modulus: (a) n 0.2, lmax 0.4; (b) n 0.4, lrr.ax for linear soil modulus: (a) n 0.2, Zmax 4; (b) n =0.4, Zmax = 4
4 (Reddy and Vaísangkar, 1970). (Reddy and Valsangkar, 1970).
330 BUCKLING OF SLENDER PILES
In real soils, kh decreases with increasing lateral deflec- tests, so that (14.24) may be expected to usually give a
tíon, and ultimately t 1te soil becomes plastic, that is, kh be- conservative estímate of Pu¡1 , except for very soft soils.
comes zero for further increase in pite deflection. For a
fully embedded pin-ended piJe, the buckling load in a uní-
form plastic medium is equal to the Euler load of the pin- 14.4.4 Group Effects
. ended strut in air. This is one half the value for the piJe in a
uniform mass with constant kh, providing the half-wave Model tests carried out by Toakley (1 964) with groups of
length of the pi! e does not alter. two and three steel-strip piles in soft silt showed that the
An eccentricity ís a general representation of an imper- critica! load is reduced by .srouping when the piles are
fection in a pile. A series of model piJe tests on fully em· closely spaced. The test results are summarized in Table
bedded pin-ended piles, carríed out by Hoadley (1964), 14.1, together with tests on isolated piles. Full-scale tests
confirmed that loading eccentricity decreases Pcr· These reported by Hoadley, Franci.s, and Stevens (1969), how-
tests also revealed that on the average, the post-buckling ever, show little interaction between closely-spaced piles.
load capacity of the pi! e ( after elastic buckling had com- As an approximate means of estimating group effects, the
menced) was 0.5 3 times the el:tstic buckling load. Assuming value of kh may be reduced arbitrarily, as suggested in Sec-
that the soil along the pilc has bccome fully plastic subse- tion 8.2. Alternatively, an elastic analysis of group inter-
qucnt to elastic buckling, the above relationship between action under lateral loading (Section 8.4) may provide a
post-buckling and e!astic critica! loads is in excellent agree- more ratio na! reduction facto:· for kh.
ment with the theoretical prediction that the critica! load
in a plastíc medíum is one half that in an elastic mass.
TABLE 14.1 EXPERIMENTAL RESULTS ON
P1LE GROUPS (TOAKLEY, 1964)
14.4.3 lnelastic Buckling
1/8-inch Piles
For pi! es that fail ínelastícally, Granholm ( 1929) suggested
~o.of Spacing Av. Pile Load Group
the followíng interaction equation: Pi! es (in.) at F;dlure Efficiency
(lb) (%)
1 12B 100
Golder and Skipp ( 1957) carried out model pi! e tests 2 3 8J 64.9
that showcd that the above expression overestimated Pu!t 2 6 10! 79.0
for very soft soíls (cu < ll5lb/ft 2 ) but underestímated Pu1t
for c11 > 115 lb/ft 2 • Hoadley, Francis, and Stevens (1969)
found that Pult is overestimated for a soil with an average
c11 of about 500 lb/ft 2 . Francis et al. (1965), however,
found good agreement between measured and predicted ul- 14.5 ANALYSIS USING ELASTIC THEORY
timate load for a full-scale hollow steel pile in soft silty
clay, but in another case, involving a prestressed concrete The elastic analyses of pi! e b éhavior under axial and lateral
píle in silty clay, the measured load was about 20% higher loads have been extended Madhav and Davis (1974) to
than the predicted value. In both the laoter cases, the soil examine the problem of buckling of a pi] e in an ideal elastic
was stiffer than the very soft soil in Golder and Skipp's soil medium. As previously discussed, this representation of
BUCKLING OF SLENDER PI LES 331
a soil should be more realistic than the simpler Winkler Assuming the piJe to be a thín bearn, the basic equatíon
spring model. of bending is
=M (14.27)
14.5.1 Analysis
For the basic analysis, the sarne assurnptions are made as in where
the analysis of a laterally-loaded socketed piJe (Chapter 8):
that the soil is a homogeneous, isotropic, elastic !ayer Eplp flexura! rigidity o.: piJe
underlain by a rigid base, and that the piJe is of length L, pP = lateral deflection of piJe axis
diameter ( or wid th) d, with the tip resting on the rigid base, l'rf = bending rr"rnent
and is divided imo n + 1 elements (se e Fig. 14.11 ).
The lateral displacements of the soil adjacent to the pile At any point i, the moment /1{ can be expressed as
elernents are given in Chapter 8:
-dL 2
(14.25) M¡ = - -{C¡}{p} + Hoz¡ + Mo
n2
L
Epip- fl~xural
rigiditry
of pilfl.
where
1:>5 -sh<iear s~s
on píl~
where
H¡ load at típ
{E}= vector,withE(j)= 1 forj=2ton;E(l)=E(n+ El
KR = .J!....E_ = pile-stiffness factor
l) 0.5 Es L4
{ F} row vector
with where
1000r-----~-----r------r-----~-----r----~------,
1 Cosa l
Pila hingc:!d atl both ands
---->i~·~ -··-j-----·---t--¡ . . 1 --- ---;----j
i 1
!v, • O· !51 K • !500
i ;
10 ---··
10
bottom, are shown in Fígs. 14.12 and 14.13. In all cases, creasing slenderness, L/d. The effect of the axial load dis-
10 elements have been used to divide the püe. The buck- tribution is shown in Fig. 14.13, where values of PcrfPE are
ling load, Pcr• is expressed in dimensionless form as a ratio plotted for various values of pile-stiffness factor K. The
of the Euler load Pe = rr 2 Epfp/L 2 (Le., the buckling load. lower the value of K, the greatcr is the load transfer to the
of a column without soil support). PcrfPE in creases with soil along the pile shaft. PcrfPE in creases as K decreases,
íncreasing pile-flexibility (Le., decreasing KR ), and with in- thus confirming·the conclusión reached from subgrade-reac-
334 BUCKLING OF SLENDER PILES
1000r-----~----~------,-----,------r----~-----,
1 Pil~
1
fíxqd
Cas~
at
n.
both qnds
K• 100
!
i
1
pe 1 -
'··+ +-
Bucklíng load 1 of
column fíxqd 1 at
both ¡¡nds. ·
: 500 1 i
"" (Smooth or 1
·
'Y'~k- lncompr¡¡ssíblil
L~
T
KR KR
FIGURE 14.15 Equivalent column lengths of fully embedded pin- FIGl!'RE 14.16 Equivalent column lengths of ful!y embedded fixed
pin piles. pi! es.
BUCKLING OF SLENDER PILES 335
1000 . - - - - - , - - - - - , . - - - . , . . - - - . . . , - - - - - . - - - - . - - - - ,
COSIZ
hing~Zd ot . both ~Znds
V5 • O· 5
1
--- Subgrod~Z R~Zoction Th~Zory
1
- Elostic Th~Zory
it. The effect of the soil in increasing the buckling loa'd is 14.5 .3 Comparison with Subgr:1de-Reaction ~olutions
less for this case than for the piJe pinned at both ends.
Alternati'le presentations of solutions from the elastic A comparison between elastic and subgrade-reaction solu-
, analysis are shown in Figs. 14.15 and 14.16, where the tions for the buckling load of 1 piJe pinned at the top and
equivalent length of the piJe L,., as a ratio of L, is plotted tip is shown in Fig. 14.17. In )rder to make this compari-
against KR for various values of L/d. son, the modulus of subgradt reaction, kh, and Young's
lt should be noted that for a particular piJe section, KR modulus of the soil, Es, have t een related by equating the
and K are not independent quantities, and it may readily be solutions for a rigid pi! e, free 't the top and pinned at the
shown that ti p. For L/d = 25, this gives ¡;hd ""' 0.8 Es. Figure 14.17
shows ~hat the elastic theory gives larger buckling loads
than the subgrade-reaction ti- eory for relatively flexible
(14.34) piles (KR < 10- 3 ), but for stiffer piles, the solutions are
almost identical. The differen( e between the solutions be-
comes more marked as L/d decJ eases.
For a solid circular pile, Eq. (14.34) reduces to
Figure 14.17 indicates that the estimation of the buck-
ling load from the simpler st bgrade-reaction theory will
(14.35) generally give a conservative value, but unless the pile is
extremely flexible, the error wi:l not be serious.
DYNAMIC LOADS ON PI LES
336
DY:t- AMIC LOA OS ON PILES 337
0.05
~· 0.02 -
0.01
<{
,; 0.005 -
"O
3
"-
E
"'
~
~ 0.002
E
w
u
'"
~
o 0.001
0.0002
strains that will vary with location and soil type. Nair dynamic soil behavior. The us~ of such models to analyze
(1969) classifies three methods of accounting for earth- the response of a pile-foundatic'n system is discussed briefly
&~Uake forces: in Section 15.6. However, it should be emphasized here
that although such analyses m1y give useful qualitative in-
l. Equivalen! statíc load at >urface, taking either this formation, it may be unwise to place too great a reliance on
load as a certain percentage (e.g., 10%) of the vertical the quantitative results, since the random nature of earth-
static load, or as the base shear utilized in the seismic quake motíon makes the satísfactory prediction of earth-
analysis of the structure, or as a force based on an average quake performance an extremely difficult task.
ground acceleration (a seismic coefficient times gravita-
tional acceleration).
2. Equivalen! dynamic load applied at the surface, often
assumed to be sinusoidal; that is, F(t) F 0 sin wt, where 15.3 PILE RESPONSE TO AX!lAL LOADS
F 0 is the equivalent static load and w is a frequency cor-
responding to the predominant frequency of the earth-
quake, 15.3.1 End-Bearing Piles
3. One component of earthquake acceleration at bedrock.
Earthquakes introduce two components of motion in the The problem of a píle bearing or1 a rígid stratum with no tip
horizontal and one in the vertical plane, the amplitude of deforrnation and no load transfer to the soil along the shaft
the latter usually being considerably less. Since the two can be analyzed by solving 1.he wave equation on the
horizontal components are usually similar, the earthquake assumption that lateral deform< tions of the pile can be neg-
motion is usually applied in the form of a prescribed hori- lected. The solutíon to this equation may be written as
zontal acceleration. follows:
The use of equivalen! static or dynamic loads, though ALp = AL)' _ 2rrfnL (2rrfnL)
m W --v-tan -.-v- (15.3)
convenient, does not have a rational basis. For example,
Seed and Martín (1966) have suggested that in relation to
embankment design, the equivalen! seismic coefficient is where
between 0.4 and 0.25, dependíng on the position of the
'potential failure circles. However, the seismic coefficient A cross-sectional area of pile
varies with depth, so the use of a single value may be inac- L length of pile
curate, especially for pile foundations. p mass density of pile material
The use of one component of the earthquake accelera- m added mass at top of pile
tion applied at bedrock has .the most rational basis. The )' unit weight of pile material = pg
soil mass and the embedded piles overlying the bedrock are W weight of added mass 1t top of pile mg
assumed to be set in motion by the imposed acceleration, fn natural frequency of pile
which is usually taken to vary only with time, not spatially. v = longitudinal wave velo.:ity in pile
Analyses giveií in soil-mechanics líterature have generally
been based on using o!d earthquake records, although From Eq. (15.3), solutions may be recovered fór both
accelerograms can be generated artificially by stochastic the natural frequency, In,and the maximum dynami~ pi).~
processes. 4 '
force, Pm (which can be shown to occur always at the pile
Because the deformations produced in the soil by a tip ). For the limitin& cases of J weightless pile or the piJe .
horizontal base excitation are primarily shear deformations, only wíth no added mass, these soh.¡t1ons Jeduce to very
the real system may be considered as discrete system based simple forms, while the solutío:1s for the general case may
on a column of soil having a unit cross-sectional area anda be expressed ln terms of these limiting solutions. The solu-
height equal to the depth of the soillayer (Idriss and Seed, tions may be summarízed as folbws:
1968; Penzien et al., 1964). The response of the en tire soil
mass may then be analyzed by assuming the mass to be l. Weightless pile (!' O):
lumped at discrete points down the depth of the !ayer and
the linkages connecting adjacent masses to consist of a
system of springs and dashpots. This system may be non-
fn(/'=0) {15.4a)
linear and can be adapted to any desired representation of .
340 DYNAMIC LOADS ON PILES
aAEp For the general case, solutions are plotted in Fig. 15.3 in
-L- (15.4b)
terms of the values for zero top mass.
For a number of typical cases, the influence of axial
2. Pile only with no added mass (W 0): loading on the resonant frequency of end-bearing piles
to rock is shown in Fig. 1 S .4 (Richart, 1962). The top
three curves illustrate the resonant frequencies of unloaded
(15.5a)
steel, concrete, and wooden piles (Eq. 15 .3). As the axial
load is íncreased on a píle of given length, the resonant fre-
7r
=- (aAEP\ (15 .5b) quencyisreduced.
2 \ L ) The above solutions are applicable only to point-bearing
piles along which no load transfer occurs so that the pro-
3. General case:
perties of the embedding soil have no ínfluence on response
of the pile to dynamic axial load. The more general case of
fn (:') fn(W=O) = ~ ifn(')'=O)] (I5.6a)
3
E{lb/in. 2 )
p (~)p -
111 \nsinA m(W-O)
= (~)p
\sin;\,
-
m(')'-O)
(I5.6b)
Material
480
)
;;:
"'E
u
8
o
e
2
2 rr
Q '
[~
S
o
e
2
t
1
,t............L..-1,....._-'------l_-'----J.._..............___J
o 0.5 1.0 1QQL---.~--~---L--L-L-L~I-L____~--J
20 30 40 60 80 100 150 200
Top mass/Pile mass 1.0 0.5 o
Pile length (ftl
Pi le mass/Top mass
FIGURE 15.4 Resonant frequency of vertical oscillation for a
FIGURE 15.3 Solutions for natural frequency and_maxímum force point-bearing pile carrying a static load W-loaded stratum is rígid
in an end-bearing pile with no load transfer. (from Richart. 1962).
DYJ.IAMIC LOADS ON PILES 341
/
15.3.2 Floating Piles or End-Bearing Píles with Load ,1
~
Transfer · Casing
/
/
/
/
For floating piles, it would clearly be absurd to attempt an /
/
/
analysis on the assumption that no load transfer from the
shaft to the soil occurs. There would appear to be ~t least
four methods that could be employed to examine the re-
sponse of floating piles to vertical loads, and these are
listed below in descendíng order of soplústícation:
(a) Pi le and ~.oil system
l. A three-dimensional analysis using the finite ele-
ment method) in which the propagation of waves through Q Q,, sin 21f ft
the pile and soil is considered.
2. An approximate elastic analysís in which the problem
is simplified to one of plane strain and it is assumed that m
the elas tic waves only propaga te horizontally.
3. Solution of the one.-dimensional wave equation, for
example, in a similar manner to the solution of this equa-
tion to analyze the pile-driving process.
4. An analysis of the response of a lurnped-parameter
rnass-spring-dashpot system representing the pile and soil.
1 ! ¡
o
1 !
1'1
Value; of damping ratio, r
4o --- -·· ;
1
i 1 i 0.01 ' 1
3o
i
--····--
!
1
1 1
2o
!
1 ,.v-0.02
1
1
1
o
8
'(, ---- -·
r¡ ~0.05
6 -· 1·-
! /¡
5
~~ ~~o
o -
....., ~ . -.........
¡....- • ............
1
8
'
~ ~ 0-.""''\ ~
~~
O. 6 e . --· ~~ i
O. 5¡---- • f----· ~ 0-.~ &
____ _, ~....
0.40~
O. 4 != ----~
~
! ¡--
O. 3
O. 1 1 1 1 1
o 0.2 0.4 0.6 0.8 1.0 1.4 1.6 1.8 2.0
f
Frequency rat' o,{~
In using the lumped-parameter approach, Barkan As mentioned previously, Novak (1974) obtained ana-
(1962) has ignored damping (i.e., assumed e = O) and lytical solutions for the stíffness and damping constants.
assumed the equivalen! mass m to be the mass of the His solutions demonstrated that these constants are not
structure ( or machín e) and the supporting piles. The equiva- highly frequency-dependent, and that the most importan!
len! spring constant cfj (tenned the "coefficient of elastic paranieters governing the response are the slenderness ratio
resistance" by Barkan) used in Barkan's analysis is plotted and the ratio of shear-wave velocities in the pile and the
in Fig. 15.8 as a function of piJe length, but it is not speci- soil. Ignoring the frequency dependence, Novak's solutions _ ·
fied whether these values are relevant to truly floating or for the stiffness and damping constants of a pile with no tip
predominantly end-bearing piles. In terms of the pile-settle- movement (Le., end-bearing pile) are shown in Table 15.2
ment theory in Chapter 5, C8 is the ratio of load P to settle- for concrete and timber piles. His subsequent investigations
ment p and can be evaluated from Eq. (5.33) or Eq. (5.34). (Novak, 1977) índicated that the damping of floatíng piles
DYNAMIC LOADS ON PILES 343
r---·-r----~--~----,---~----~--~--~~--~-----
-~----~-----~--~---.t----~---~----~-----~----
1
v alues of damping r~tio, 1
0.01
l
~----4---~---4-----~'--#ffi---~-----~---~---+---· •
i ~Ó.02
~--~-T~------~--~~~~~---4----~--~---
10
1 i
¡ {!\¡!e- 0.05
1
8
11 \ 1
6
(~
1-------+----+-------+--------+-----l-,1-,.k- 1
3
r¡ ~
E 1 ru•
.¿'E
2
12 1.4 16 1.8
f
Frequency rat10. ~~
is larger than that of end-bearing piles, and that the vertical equivalen! mass m to be only the above-ground mass of the
dynamic response of a footing supported by floating piles oscillator, piJe cap, and static load_ They carried out a series
can be much smaller than if it is supported by end-bearing of tests on steel H-piles and concrete-filled pipe piles, in
piles. Sorne typical solutions for stiffness and damping con- silty sand and clay overlying sand, to determine the rela-
stants of floating piles are given in the latter reference. In tionship between frequency and displacement. From the
using Novak's solutions to evaluate the response of a foot- test results, values of equivalent stiffness C8 and damping
ing or structure supported by the pUes, the mass in the ratio ~ were backfigured. At' resonance, the dynamic value
equivalen! lumped-parameter system is taken to be that of of Cs was found to be greater than the static stiffness for
the footing or the structure itself. comparable piles, but it was suggested that the use of the
Maxwell et al. (1969) have also employed the lumped- static stiffness would be adequate for practica] purposes.
parameter m o del in Fig. 15.5, but have considered the Damping was found to be slight, the computed damping
344 DYNAMIC LOADS ON fiLES
TABLE 15.1 SUMMARY OF RELATIONS TABLE l5.2a STIFFNESS AND DAMPING CONSTANTS FOR A
FOR SlNGLE-DEGREE-OF-FREEDOM VIBRATION SINGLE VERTICALL Y LOADED END-BEARING PlLE 0
Critica! damping ee = 2,/c¡,m
Con ere te Pi! es Tim ber Pi! es
e Vs
Damping ntio !" -;¡-¡ {,.,,
ce Lfd
'···· '"·' f ....
Undamped natural
frequency
fn
U;)fl 10 0.01
0.02
0.03
0.050
0.051
0.052
0.001
0.005
0.010
0.050
0.054
0.059
0.003
0.013
0.029
0.04 0.055 0.018 0.066 0.050
Qo
Sta tic displacement Ps = ·e¡, 0.05 0.057 0.029 0.075 0.073
25 0.01 0.021 0.002 0.022 0.008
Amplitude-magniflcation M 0.02 0.023 0.01 ¡ O.G30 0.029
factor during vibration 0.03 0.027 o 024 0.040 0.054
0.04 0.032 0.032 0.053 0.077
0.05 0.038 0.038 0.067 0.098
For Constant-force Excitation For Rotating-Mass Excitation
Q = Q 0 sin 2rrft Q moew'sinwt (w=2rrJ) 50 0.01 0.011 0.005 0.015 0.015
0.02 0.016 0.020 0.027 0.039
Amplitudé at frequency f 0.03 0.023 0.035 0.040 0.060
0.04 0.030 0.048 0.053 0.079
mee([)' 0.05 0.038 0.061 0.067 0.099
Jñ\fn M
a Reproduced by permission of the Nation<tl Research Council of
Maximum amplitude of víbration Canada, from the Canadian Geotechníca/ Journal, Vol 11, 1974, pp.
574-598. .
b After Novak (1974).
11
V8 shear-wave velocity in soil = [Esg/(2(l+vshsH '
where
whcrc
A area of pile cross section
1 mass momer¡t of inertia about appropríate axis.
Ep Young's moduJus of piJe
"fs = soil unít weight
lp = pile unit weight
ratio r
for single piles being on the order of 0.04 to 0.05.
However, it was also found that both the stiffness and the Note: for solid piJe,
damping ratio varied with frequency, so that the use of a
single frequency-independent va!ue of each of these para- ~-
Ve.
(2. __-)y, I
K 'Ys 2(l+vs)
meters would not lead to an accurate prediction of pile
response for all frequencies. In particular, the response at where K = pile stíffness factor.
2.0 .------.----,---,----¡---,----,
40 -L 1
1
1 o
• Pile-driving
o Pulling-out
i oj u]J
1.5
30
1 -·~1 J
X
E
~
-"' 20
'
•
]}-- 1
/ ~
• 1 0.5
o 25 r - - - - - r - - ·
1.0 1.'> 2.0 2.5 3.0
0
1
o
y
--4=l
1
10
>/ 1
0.201---+-
• 1
o
o
2,0
b
350r------~--~~--~-~--~~~--~--~
úfJ
Legend , 1 si 1,5
.o
~
1
1 Excíting o
1 Force (tons)
~ _;:¡p<:P 1 '
~
':? 8 V '-'lt'~;-oo..,tl !">
f 1.0
o ro G~
~
o
0.5
o
r\
o 1 1
o 0,5 1.0 1.5 2.0 2.5 3.0
.25
!
i
.20 ·- ........... ~-
"V ¡¡
Jj¡ o
'2.
ro
en
e:
.15 -
o
o
¡le ov
o.
4 18 20 E .10 - ...
8 12 24
Frequency Hz
.05
V
l e
E
i for the single piJe, the value of ~ being about 0.1 at the
'"m" 1 l resonant frequency as compared with about 0.04 for the
~ 600
e 1 i 1
!
single pile.
1
Nova k and Grigg ( 1976) carried out tests on mcdel piles
400
\ i
and pile groups subjected to both vertical and horizontal
200
/Jl.d ~ 1
excitation. Group effects were allowed for by the modifi·
cation of the single-piJe analyses through the use of inter-
¡~ ..,
action factors (see Chapters 6 and 8). The theory was found
to predict all the qualitative features of response of the pile
o
o 4 8 12 18 20 24 groups. The use of the soil modulus derived from a static
Frequency Hz single-piJe test was found to yield reasonable predictions of
natural frequency and resonant amplitud e.
FIGURE 15.11 Results of constant-force test on uncased H-píle
For relatively large groups of piles or groups having a
D-4 after cxcavatíon of soil under cap (Maxwell et al. 1969). (Re·
printed by permissíon of !he Ameriean Socíety for Testing and breadth than the piJe length, a simple approach may
Materials, 1969.) be employed for determinin,g the amplitude of vibration, by
DY.-~AMIC LOADS ON PILES 347
ªCl.
¡::-
;¡_
where
Aumax ~ (Z.:) 0.85 (!~ O.l8fz (15.8) resonant frequency and the 1mplitude of vibration of the
cantilever may than be detellníned by standard methods.
However, no information can be obtained on the moments,
where stresses, and displacements abng the length of the piJe for
dynamíc loads.
amplitude of disturbing force 2. The pile ís considered as a beam on an elastic founda-
equivalen! spring stiffness tion subjected to tíme-dependent loadíng and analyzed by
. mo(l - Vs) finíte differences. Moments, stresses, and dísplacements
B mass ra t 10 = ----''-'-:::-or
pD2 along the length of the píle u ay be analyzed, and irnpact
mass of foundatíon loads as well as harmonic load> can be consídcred.
mass density of soíl 3. The approximate analytical techníque developed by
Poisson 's ratio of soil :--iovak (1974) and descríbed previously for vertical
response can be used. This approach derives stiffness and
For foundatíon shapes other than circular, the mass ratio of damping constan ts for piles and pi! e groups, and thus en-
a circle of equal area may be used to compute . C8 ables use of the simple "lumped-parameter" approach to
can probably be determined with sufficient accuracy from a determine the lateral response.
348 DYNAMIC LOADS ON PILES
Mass lumped
H at end
'
)¡¡
""' 3.13
. r::-
Hz (15.1 1)
yp¡
Mass, m, dístributed
along length where
F lexural rígídity.
En lp
p1 = static deflection of top under the static load, in
in.
rigidity,
Ep lp The above solution may be applied in cases where the static
lateral load is large in relation to the weight of the oile.
For the simple case of a llo<iting pile in a uniform
Winkler medium, Warburton (1964) derived the following
(a) Actual pile lb) Equ1valent can\ílever expression for natural frequency:
The problem is íllustrated in Fig. 15.14a, and a general kh horizontal modulus of subgr2de reaction (see
representation of the equivalent cantilever is shown in Section 8 .2)
Fig. 15 .14b. The mass of the pile m ay be considered to d = pile diameter or width
be uniformly distributed along its length with a mass p mass density of pile
lumped at the encj. The length Le of the equivalen! canti- A crosS-sectional area of píle
lever may be estimated from the consideration of the beam
on an elastic foundatíon subjected to a horizontal static For rnore-complicated problems, the frequency-dis-
load, as u sed by Davisson and Robinson (! 965) and Nair placernent relationship and resonant frequency of the canti-
et al. (1969)(See Chapter 14.). lever may be obtained by solving the appropriate equation
For the case of a cantilcver of uníform cross-section of motion. 1f the forced vibration has a frequency very dif-
with no concentrated static mass at the top, the natural ferent from the resonant frequency of the system or any of
frequencies, fn, are given by its harrnonics, the amplitude of motion under the forced
vibratíon is unlikely to be rnuch greater than that given by a
static analysis using the maxirnurn and mínimum values of
(15.10)
applied load.
In an alternative approach described by Hayashi et al.
where (1965), nonlinear and viscous behavior of the soil is con-
sidered. Results presented by them indicate that the non-
Epfp stíffness of cantilever in bending linear effects introduced by the soil have a significan! effect
m mass of cantilever on the pile- response.
Le equivalen! length of cantilever
Cw coefficient
22.03, and 61.70 for the first three 1S .4.2 F inite-Difference Analysis
rnodes of vibration
This analysis, presented by Tucker (! 964), is an extensíon
The above solution rnay be used for cases in which the of the static analysis described in Section 8.2. The represen-
static lateral load is negligible in relation to the weight of tation of the piJe is shown in Fig. 15.15. In the general case,
the piJe. the width or diameter of the piJe, and hence the bending
For the case of a weightless cantílever having a static stiffness Epfp~denoted here as F-rnay vary along the piJe.
load at the top, the solution for the lowest natural frequen- In the static case, lateral and axialloads H and P may act.
cy, fn, is approxirnate by the Southwell-Dunkerley expres- The soil is assumed to be a Winkler material and provides a
sion (Richart et al., 1970): lateral resistance, S¡, per unit length at any point i. To
DYNAMIC LOADS ON PILES 349
. t p d = pile diameter
L~o The inclusion of dynamic effects involves the addition
of terms to the above equation, representing inertial and
damping characteristics of the pile-soil system. Inertial
properties of the soil medium are assumed to be included in
the elastic restraints, S. Inertial properties of the pile and
soil damping are then represented as additional forces on
the píle at each poínt i. Thus the resulting lateral load at
i ís
r-1
1
ri (15.14)
t 1- 1 where
allow for rotational restraints along the pi!e (e.g., at the pile
( 15 .15)
head), a rotational elastic restraint, R¡, is assumed at the
poínt i (R¡ =O if no restraint is provided at i).
The finite-difference expression for the static behavior = J.1 (op¡\ (15.16)
of the piJe at a point i may be expressed as follows ar)
(Matlock and Ingram, 1963):
where
a¡Pi-2 + b¡p¡__ , + c¡p¡ + d¡Pi+t + e;Pi+2 =!; (15.13)
W¡ weight of piJe element at point i
where g gravitational accelera:íon
1¡ damping factor at i
a¡
. a2p¡
b¡ 2(FH + F¡) The acceleratwn, ot 2 , and the velocíty, at m ay be
e¡ F,-_ 1 + 4F¡ + F¡+l + ó 3S¡ +% {R¡_ 1 + óPH) + written in terms of higher-onler backward differences so
that Eqs. (15.15) and (15.16), for a time t, anda time in-
1(Ri+l +óP¡+¡) crement ot, become
d¡ - 2(F¡ + F¡+¡} W·
e¡ F¡+ 1 -
o
¡(Ri+l + oP¡+ 1 )
Hau = g(o ;)2 (2Pi.t 5Pi,t-l + 4p~t-2 - Pi,t-3)
(15.17)
Ji = o H13
left-hand side and all known values on the right-hand side, To solve a particular problem, an ínitial set of condi-
the resulting expression ís tions (deflection, velocity and acceleration) must be
specified and (15.19) and the associated boundary-
A¡Pi-2¡ + B¡Pi-l,t + C¡Pi,t + D¡Pi+l,t + E'¡Pi+2,t = G¡ condition and equilibrium equatíons must be solved for
(15.19) each time íncrement ót.
Tucker (1964) presented two examples íllustrating the
where use of the above approach . The first ínvolved impact
ó loading of a 70-ft-long piJe with two section changes and its
A¡ F¡_¡ - -¡(RH,t + ilPí-l,r) lower 55 ft embedded in a rnaterial having a linearly ín-
2(F¡_ 1 -t- F¡) creasing modulus of subgrade reaction wíth depth. Time
increments of 0.02 seconds were uséd in the solution. The
e-¡ F¡_¡ + 4Fi + Fi+t + ó3Si,t +! (Rí-t,t + óPi-t,r) + problem is shown ín Fíg. 15.16, together wíth the S()iutíons
for displacement versus time and moment-distribution at
~ (Rí+l,t -t- óP¡+l,t) -t- ó 3 (2W¡/(ot) 2 -t-J¡/68t) two times. The second example, one of forced vibrations on
D¡ -2(F¡ -t- F¡+ 1 ) a free-head piJe, simulates the model piJe tests of Gaul
(1958). A sinusoidally varying load was applied at the mud
E¡ = F¡+ 1 - %(R¡+t,t + fíP¡+ I,t) line, and. lO elements were used for the ·piJe. Each load
cycle was represented by 20 time-increments, and severa!
H¡,r -t- W¡(SP~r-1 4Pi.t-2 + P~t-3)
3
G¡ <'> + frequencies of load application were considered, varying
J.ó3 ' fro.11 1 to 200 cycles per second. Calculatíons at each fre-
6t3t (18Pü-l- 9Pü-2 + quency were continued un ti! the oscillation of the piJe had
stabilízed into a pattern. The problem and the solutions for
The above equation can be applied to all interior-node amplitude versus frequency and peak moment versus depth
points of the pile, and these, together with the equations for three frequencíes, are shown in Fig. 15.17. The latter
describing the top and tip boundary conditions and the curves show that píle bendíng moments increase signifi-
horizontal-load and momcnt-equilibrium equatíons, cantly -as the vibration frequency approaches the natural
describe the behavior of the píle ata time t. frequency of the system (about 80 to 85 cps here). The
(d)
(a)
o 77~7//77' .
6 2
111:F6 86 X 10 lb 1n
1 Wt 1 264 lbift
''¡ E,
1 48 os¡
'1
1
!, cps
10 lb)
le)
la)
FIGURE 15.17 Example of forced vibration of a model free-head pile (Tucker, 1964).
peak moment curve for a frequency of 1 cps agrees well 15 .4.3 Novak 's Analysis
with that obtained experimentai!y by Gaul.
Further developmcnts of the above method of analysis, Nova k ( 1974) has derived lateral stíffness and darnping con-
and also a number of test resu!ts, have been described by stants for single píles, including values for coupled rotatim1
Chan and Matlock ( 1973), Agarwal (1973), and Prakash and translatíon; Table 15.4 reproduces sorne of these solu-
and Chandrasekaran ( 1973). tíons. 1t is again found that the frequency dependence of
An alternative analysis for pile response to dynamic stiffness and darnping can generally be ígnored, and that the
lateral loads, employing a fínite-element formulation for important parameters are the ratio of shear-wave velocíties
the pile, has been clescríbed by Ross (1971 ). Nonlinear in the pile and soil, and the slenderness ratio L/d. The
soíl-resístance propertíes and geometrícal nonlinearity of effect of static load was investígated and found to be signif-
the piles were taken ínto account, anda study of a píle sub- ícant only wíth extremely poor soils. Most stiffness and
jected to wavc forces was made. It was found that maxi- darnping parameters were reduced by the presence of axial
mum displacemcnts based on the dynamic analysís were load, but the damping caused by rotation was increased.
considerably greatcr than maximum displacements by static
analysis, especially along the upper portien of the píle. It
was also found that the magnitude and distribution with 15.4.4 Pile Groups
depth of the subgrade-reaction modulus of the soil ín-
fluenced th<:: dynamic response considerably. The methods currently available for the analysis of pile
All the above analyses are based on subgrade-reactíon groups for dynamic surface loads are usually extensions
theory, with the soil being represented as a Winkler of methods of structural analy-sís in which the piles are
materiaL It is also possíble to extend the statíc-elastic reduced to equivalen! cantilevers. Such analyses, described
analysis described in Chapter 8 by íncorporating ínertia and by Shubinski et al. (1 967), Nath and Harleman (1 967), and
dampíng terms in the basic pile-bending equation. Such an Saul (1968), can take account of vertical and torsional
approach, whíle allowing for continuity of the soil rnass, is, loading as well as lateral loads. However, sorne uncertainty
however, still approximate, as it relíes on static-elastic rnay aríse in some of these analyses regarding the structural
theory to give the dynarníc response of the pile-s::>il systern. approximation of the group. An approximate estímate of
Prelirninary calculations índicate that the natural frequency the behavior of a group under dynarnic lateralloading rnay
of a pile in a uniforrn soil given by such an analysis is quite be obtained by analyzing a sirgle pile under such loading
similar to that given by a subgrade-reaction analysis. and then allowing for group effects on the basís of a static
TABLE 15.4a STIFFNESS AND DAMPING CONSTANTS
FOR LATERAL RESPO~SE OF PILES HAVING L/d > 12.5b
4Epfp
Translatíon damping constan!, Cxx = - , - (/11 , 1 )
d Vs
4Eplp
Cross-stiffness constan t, -,-(/,,,}
d
Cross-damping constan!,
where
352
DYNAMIC LOADS ON PILES 353
analysis of pile interaction Chapter 8). Novak and the soil medium at that point when acting alone. The beha-
Grigg (1976) have used this approach to modify Novak's vior of the structural system may be readily obtained from
single-pile solutions and have obtained good agreement structural analysis. Hence, soil and system displacements,
between the response so calculated and the response of velocitíes and accelerations m ay be equated at each discrete
groups of model piles. point and the resulting equatio:1s solved for each time-step
considered.
Penzien et al. (1964) consider a relatively complicated
structural system comprising a bridge deck supported on
piles, but for the simple case of a single piJe, the pile beha-
15.5 PILE RESPONSE TO EARTHQUAKE FORCES
vior may be t>btained from the beam equation. The piJe
response may then be analyzed by the method described in
An analysis of this problem has been described by Penzien
Section 15.4, using the calculated soil acceleration at each
et al. (1964). The analysis is divided into two parts: point at each time for calculating the applied lateral force,
or alternatively, usi;g the calculated soil deflections as an
l. The determination of the dynamic response of the soil input into the analysis.
medium alone when excited through its lower boundary by A study by Penzien (1970 l has shown that during the
a prescribed horizontal seismic motion. critica! time-periods of the earthquake, the piles move
2. The determínation of the ínteraction of the entire essentially with the soil, inte1 action effects being small.
structural system (structure and piles) with the moving Hence, th<i pile curvatures are essentially controlled by the
soil medium. motíon of the surrounding soil. In a layered soil in which
considerable differences in strain may develop b'etween
Dynamic Response ofSoil Alone
The soil is assumed to be of infinite horizontal extent and layers, large curvatures could be introduced in the pi! e (e .g.,
of constant depth, and is idealized as a discrete-mass system piles driven through a soft-clay !ayer into a dense-sand
based on a column of soil having a unit cross-sectional area !ayer). The curvatures and deformatíons so íntroduced
and a height equal to the soiJ ..Jayer depth. The mass is could well exceed the capacity of the pile section if it were
lumped at discrete poínts along the depth of the !ayer, the designed for vertical loads only, and it would appear desir-
number of points being dependent on the accuracy re- able to allow for moments arre lateral deformations in the
quired. Use of this idealization permits the consideration of design of piles in earthquake areas.
a layered-soil profile and nonlinear and hysteretic soil prop- It is possible that greater pile-soil interaction could oc-
erties. The idealízation used by Penzien et al. (1964) con- cur with shorter, stiffer píles m a stiffer, heavier superstruc-
sists of masses linked by a bilinear hysteretic spring and ture than that considered by Penzien. It must also be
nonlinear dashpot corirrected in parallet, which are then emphasized that this analysis was based on a particular
connected in series to a second nonlinear dashpot. Sugges- earthquake record (El Centw 1940, N-S component).
tions for determining the parameters of this model are Somewhat different conclusio11s might be reached for dif-
given. ferent acceleration-time inputs.
Alternative models of soil behavior may be used, and For a single pile, an indicaban of the bending moments,
for the simple case of a linear homogeneous isotropic reactions and deflections caused by earthquake loading may
elastic model, an exact analysis can be performed (Idríss be obtained from a static analysis of piles subjected to hori-
and Seed, 1968). zontal soil movements (Chapter 13), if the veriical distribu-
tion of maximum horizontal movements caused by an
Interaction of Structural System and Soil earthquake can be estimated. In fact, in many practica!
The consideration of interaction of the soil and structural cases, the natural frequency cf the piles may be consider-
system withín the soil is similar to that for static analyses, ably higher than the. predominant frequency of the lateral
in that each discrete point along the system is subjected to soil movements, so that a stati: analysis as described above
an acceleration-time history that is equivalent to that of wíll a quite reasonable estímate of the pile response.
PI LE LOAD TESTS
16.1 INTRODUCTION de•·eloped for carrying out pile load tests; among the most
common procedures for compressíon tests are
P!le load tests are usually carríed out for one or more of the
1. Maintained loading tests.
folluwing reasons:
2. Constant-rate-of-penetration (C.R.P.) tests.
3. Metliod of equilibrium.
1. To serve as a proof test to ensure that faílure does not
occur before a selected proof load ís reached, this proof In this chapter, these procedures and their interpreta·
load being the mínimum requíred factor times the working tion are reviewed, and the possible effccts of the loading
load. system on the measured settlement of the piJe are examined
2. To determine the ultimate bearíng capacity as a check theoretically. La ter al and torsional testing are al so de·
on the value calculated from dynamíc or statíc approaches, scríbed briefly.
or to obtaín backfigured soil data that wíll enable other It must be emphasized that in many cases, the rcsults
piles t o be desígned . of a test on a single píle cannot be extrapolated directly to
3. To determine the load-settlement behavior of a píle, predict the behavior of píle groups or other piles. As
especial! y in the regían of the anticipated working load. pointed out by Chellis (1962), the volume of soil in·
Tnís data can be used to predict group settlements and fluenced by a single pile is much less than that of a large
se! tlements of other piles. group, so the influence of deep-seated compressible layers
4. To indica te the structural soundness of the pile. may not be apparcnt in a pile load test, aithough such
layers may critically affect l.he behavior of a group. Pile
The most common type of test ís a compression test, al- l6ad tests should therefore be accompaníed by detailed
though uplift, lateral-load, and even torsíon-load tests are site ínvestigation to define accurarely the entire soil
also performed. A varíety of test procedures have been pro file.
354
PILE LOAD TESTS 35 5
16.2 MAINTAINED LOADING TEST tied down to the heads of tht anchor piles and spanning
the test pite. In testing piles installed for the actual struc-
ture rather than for special tesr piles, it is often convenient
16.2.1. Proced~·re to test an interior member of a group in this manner. A
hydraulic jack on the head of the test pile applíes the load
This is the usual method of carrying out a test, especial!y if and obtains a reaction agaínst the undersíde of the beam.
the load-settlement relationship is required. The proc~dure This method is sometimes called the '~bootstrap" method.
is to apply the load in stages, the load at each stage being Whitaker ( 1970) recommends that any anchor piJe should
maintained constant until the resulting settlement of the be at least three test-pile diameters from the test piJe,
pi!e virtually ceases before applying the next increment. center to center, and in no cas:: less than 1.5 m ( 5 ft ). For
The Civil Engineering Code of Practice No. 4 (1954) takes a íJiles with enlarged bases, the spacing should be the greater
rate of movement of 0.012 in./hr (0.305 mm/hr) as the of twice the base diameter or four times the shaft diameter
limiting rate befo re additíon of the next increment. Cooling of the test pile. However, even with these spacings, con-
and Packshaw (1950) recommend 0.0033 in./hr (0.084 siderable interaction between the anchor piles and the test
mm/hr), while A.S.T.M. Dll43-57T requires a rate of pi! e m ay occur, resulting in an inaccurate inclicatíon of the
settlement less than 0.012 in./hr (0.305 mm/hr) or until settlement of the pi! e ( the measured val u e will be less than
2 hr has elapsed, whíchever occurs fírst. lt is perhaps doubt- the correct val ue).
ful whether a time interval of 2 hr is always adequate to 4. Ground anchors that usually transfer the reaction to
ensure completion of consolidatíon settlement. However, as stiffer strata below the leve! of the pile típ. Because the
is shown theoretically in Chapter 5, the major proportion upper portian of an anchor c<1ble does not usually trans-
of the settlement of a pile occurs as immediate settlement, fer load to the soil, ground anchors can be placed closer
so relatively short intervals betw•een load increments should to the test pile than can reaction piles.
be acceptable--at least at load levels not approaching
failure. The last three methods ma'( affect the measured settle-
The usual procedure is to increase the load in stages ment of the test pile, in sorre cases sígnificantly, and íf
until the proposed workíng load is reached, and then to load-settlement is required, steps should be taken to mini-
unload and to Ieave the load off until the rise or rebound míze the effects of the test-load reactions or to correct for
substantially ceases. The pile is then reloaded to the them. A theoretical examination .of possible crrors arising
working load or to the next higher stage, and the test con- from methods 3 and 4 is descnbed in Section 16.5, and the
tinued to the maxímum load. The unloading of the pile approach could be adapted for method 2.
from the maximum load is often carried out in stages, with The settlement of the pile head may be measured by
a pause at each stage until rebound virtually ceases before dírect leveling with reference to a fixed datum, or by a wíre
unloading to the next stage. The precise loading and un- held under constan! tension between two supports and
loading procedure may often be specífied by building codes passing across a se ale attached to the test pile, or by dial
or established practíce in a particular organization. gauges attached to a beam supported on two foundations
The following methods are commonly used to apply the sufhciently far from the test pie for the reaction system to
loador downward force on the pile: be unaffected by the groucd movement. In order to
measure pile movernents and •,oads at various points along
l. A platform is constructed on the head of the pile, on the pile, displacement rods (sometimes termed "tell-tales")
which a mass of heavy material, termed "kentledge," is or strain gauges m ay be installed. This type of instrumenta-
placed. tion can be installed in almost al! types of conventional
2. A bridge, carried on temporary supports, is constructed piles, but more readily in cast ín-situ concrete piles. In
over the test pile and loaded with kentledge. The ram of a general, tell-tales are simple to install, read, and maintain,
hydraulic jack, placed on the pile head, bears on a cross- and Vijayvergiya (1 969) and Tomlinson (1 977) give details
head beneath the bridge beams, so that a total reactíon of the tell-tale system and its interpretation. As pointed out
equal to the weíght of the bridge and its load may be ob· by Fuller and Hay (1970), the installation of tell-tales or
tained. Whitaker (1970) recommends that the supports be strain gauges results in a physical change ín the cross section
more than 1.25 m ( 4 ft) away from the test pite, to of the pile, and thus its · elastic properties; thus, it m ay
minimize the effect of the supports on pile settlement. sometimes be advisable not to install too much instrumen-
3. Anchor piles capable of withstanding an upward force tation along the pile. In many cases, a tell-tale at the pile
are constructed on each side of the test píle, wíth a beam típ may give sufficient information.
· 356 PILE LOAD TESTS
16.2.2 Interpretation of Load Tests the ultirnate load capacity of the pile, may be recognized.
Among the commonly-used definitions of the ultima te load
16.2.2.1 EMPIRICAL METHODS FOR WORKING LOADS capacity are:
A considerable number of arbitrary or empírica! rules have
l. The load that causes a settlement equal to 1O% of the
been used or are contained in codes to serve as criteria for
pile diameter (Terzaghi, 1942).
determining the allowable working load from load-test
2. The load at which the rate of settlement continues un-
results. A number of these rules have been summarized by
diminished without further increment of load, unless this
Chellis (1961 ), and a few are quoted below.
rate is so slow as to indicate that settlement may be a result
of consolidation of the soil (Civil Engineering Code of
l. The test load shall be twice the design load and shall be
Practice No. 4, !954).
maintained constant for at least 24 hr ,·and untíl settlement
or rebound does not exceed 0.22 in. in 24 hr. The design
Whitaker ( 1970) considers the latter definition inadequate
load shall not exceed one half the maximum applied load,
to define failure, especially with piles in cohesive soil.
provided the load-settlement curve shows no signs of
Van Weele (1957) has $Uggested a method of cyclic
failure and the permanent settlement of the top of the píle,
loading to provide sorne indication of the distribution .of
after completion of the test, does not exceed ~in. (Boston
load between adhesion and end-bearing. A plot of the
Building Code).
elastic recovery at each unloading cycle versus load appliéd
2. Tests shall be made with 200% of the proposed load,
at that cycle is used to separate the two components. The
and considered unsatisfactory if after standing 24 hr, the
curve usually becomes a straight line soon after the early
total settlement after rebound is more than 0.01 in. per ton
load increments. The distance between this curve and a line
of total test load (building laws of the City of New York).
drawn through the origin and parallel to the straight part of
3. Observe the point at which the gross settlement begins
the curve, represents the portion of load carried by
to exceed 0.03 in. per ton of additionalload, and divide by
adhesion. This procedure is approximate only. An alterna-
. a factor of safety of 2 for statíc loads, or 3 for vibratory
ti ve method is outlíned by Woodward et al. (1972).
loads (W. H. Rabe).
The importance of residual stresses in driven piles on
4. Observe the point at which the gross settlement begins
to exceed 0.05 in. per ton of addítional load, or at which piJe-test interpretation has been stressed by Holloway et
the plastic settlement begins to exceed 0.03 in. per ton of al. (1975). Compressive residualloads are likely to exist in
additional load, and divide by a factor of safety of 2 for the lower part of the pile, andthese appear to depend on
static loads, and 3 for vibratory loads (R. L. Nordlund). the pile-soil system only, independent of the impact pile-
5. Take two thirds of the maxim!.lm test load in a case driving apparatus used. When a residual point-load remains
where settlement is not excessive and where load and settle- after driving, a portian of the point-bearing capacity has
already been mobilized; however, if load distribution
ment were proportional and the curve remained a straight
measurements are made, the gauges are generally zeroed at
line. Where the test load was carried to failure, take two
the start of the test, and the residual loads are ignored in
thirds of the greatest load at which settlement was not ex-
the test interpretation. ln compression load tests to failure,
cessive and at which loads and settlement were proportion-
the measured point-bearing value in such cases is only that
ate (United States Steel Co.).
mobilized from the start of the load test. The actual point
capacity is the measured value plus the residual point load.
· Chellis éonsiders rules 3 and 4 the most reasonable al-
Conversely; in the tensile load tests, the effect of residual
though 3 may be too conservatíve. Rules such as 5 are,un-
compressive loads is to cause an apparent tensile resistance
reliable, as various irnpressions of the steepness of the load-
at the point. While the effects of residual loads are not
settlement curves may be obtained by varyíng the scale of
readily taken into account, recognition of their effects m ay
the graph, so that a finite lirnit of the change-of-load to
at least.resolve apparent anomalies in sorne load tests..
change-of-settlement ratio is desirable.
In carrying out a maintained load test to determine the load The load-settlement relationship may be used directly to
capacity of a piJe, Whitaker (1970) suggests that it is neces- determine the single-pile settlement at the working load. In
sary first to estímate the load capacity so that a suitable estimating the settlement of a group or the settlement of a
loa~ing and reaction system may be provided, and then to píle of different proportions, the average soil modulus along
defme sorne physical event by which "failure," and hence the pile may be determined by fitting the measured lo~d-
!'ILE LOAD TESTS 3S7
settlement behavior ro the theoretical behavior. Knowing 16.1. The graphical solution ís shown in Fig. 16.1, from
the soil modulus, the stiffness of the relative to the soil which the required undrained value of Es ís found to be 73
may be determined, whereby the appropriate theoretícal kgf/cm 2 .
settlement"influence factor may be determined for a single Considering now the determination of the drained value
pile (Chapter 5), or the appropriate theoretical group" of Y oung's modulus, it will be assumed that the drained
Poísson's ratio of the soil is 03. From Fig. 5.21, R..,= 0.93.
settlement ratio (ratio of group settlernent to single pile
The other factors in (5.33) remain unchanged, so that
settlement at the average pile load) rnay be determined for substitution into this equation, usíng now the final value of
a pile group (Chapter 6). In the latter case especially, the p of 13 mm, gives
additional effects of any deep-seated compressible strata
must be carefully considered in the settlement estima te. (16.4)
An example illustrating the use of the theoretical
soíutions to backfigure the average soil rnodulus from a pile Equation (16 .3) remaíns valid, and graphical solution of
load test is given below. Eqs. (16.3) and (16.4) in Fig. 16.1 gives the drained value
of Es as about 50 kgf/cm 2 •
As an aJternative to the above procedure, the theoreti-
Example of lnterpretation of Pi/e Load Test to Backfigure cal relationship between settlement and Es can be plotled
Soil Modulus
As a simple example, the case of 0.3-m"diam., 15-m-long
floating concrete test piJe in a 30-m-thíck layer of clay wíll TABLE 16.1
.be consídered. It will be assumed that ata lq_ad of 50 metríc
ton, an ímmediate settlement cf 1O mm and a final settle- Undrained Drained
ment of 13 mm is recorded. In order to backfigure the Es ES Es
RK (kg/cm') (kg/cm') (kg/cm')
average undrained and drained val u es of of the clay, use
K (Fig. 5.19) (Eq. 16.1) (Eq. 16.3) (Eq. 16.4)
is made of Eq. (S .33).
Considering first the undrained modulus, from Fig. 10000 1.02 65.0 20 46.5
5.18, 5000 1.08 68.9 40 49.2
2000 1.19 76.0 100 54.3
1000 1.37 87.5 200 62.5
.044 for Ljd 15/0.3 50 and db/d
500 1.68 107.2 400 76.6
and
(16.1)
Draíned
(16.2) Eq. 116.4)
Sol u tion of Eqs. ( 16.1 ) and ( 16.3) is most easily carried . FIGURE 16.1 Graphical solution for soil moduli from pile load
out graphically, using a tabulation such as is given in Table test.
358 PILE LOAD TESTS
u sed for full-scale piJe tests (Whítaker and Cooke, 1961; (.1)
shown ir; 16.2a . The values of force reached at the FIGURE 16.2 C.R.P. test (aftcr W~itaker, 1970).
points mark0d A represent the uhímate load in each case.
The force-penetratíon curve for an end-bearing piJe ís
simibr to that shown in Fig. 16.2h. The upper part of the reasonable settlement data. The principie is to apply to the
curve is straight, or substantially straight, and shows a pile, at each of the t~st, a load slíghtly higher than the
steady increase in !'orce with increasing penetratíon; the required load and then to decrease the load to the desired
ultim:Jte load is taken as the point A, which represents the value. By this means, the rate of settlement diffiinishes
bcginning of this straight portion. This line is found to be a much more rapidly than with the maintained load proce-
continuatíon of the force-settlement relationship for instal- dure and equi!ibrium is reached in a matter of minutes
lation of thc píle from the surface of the bearing stratum rather than hours. The procedure suggested by Mohan, Jain,
en tire! y by C.RP. technique. ldentification of the point and Jaín is first to apply about one tenth of the estimated
A is often difficult in practice, and Whitaker (1970) ultima te load by hydraulic jack in a period of three to five
suggests that ít is usually satisfactory to take the ultimate minutes. lt is maíntained for about 5 min and then allowed
load as the force required to cause a penetration of 10% of to reduce itself vía downward rnovement of the pile. With-
the piJe diameter. in a few minutes, a state of equilibrium is generally reached.
The next increment of load is then applied and the process
is repeated. For higher loads, it is desirable to maintain the
ínitial load for a period oflO to 15 min before it is allowed
16.4 METHOD ÜF EQUILIBRIUM to relax. The total time required by this method is generally
reduced to about one third of that required in a main·
Tll!S procedure, which has been described by Mohan, Jain, tained-load test. At each stage, a cycle of loading and un·
and Jain (1967), is primarily designed to determine the loading may also be adopted and the elastic rebound of the
ultima te load capacíty, although it al so appears to provide píle top measured, in order to separate the side adhesion
P'~E LOAD TESTS 359
16~
120
~~~e~-ll
table 1 ¡
" '" 4
Fill with 1 .32~
.48l-
o-c-o Sta tic cone resistan ce
......._ Number of blows (N) o Me!hod of equ,tibrium
Silty clay '"
.e
u
·e:: .96
;:
"
E 1.12
"'
-¡::;
;;::-
(/)
Clay with
kankar
----j-
1.44 Total settlement
Silty clay _____ _
160
i 1
1.76
L!'l=±="-_,_ ___Li_j___l___l_L_
12 18 24 30 36
N
FIGURE 16.3 Load-settlcment curves from maintained load test and mcthod of equilibríum (after '.-lohJn d al., 1967)
and point-bearing capacities (Van Weele, 1957; Jain and l. The use of a reference beaw to rneasure the settlemer..t
Kumar, 1963). (Fig. 16.4a).
A number of tests were carrkd out by Mohan, Jain, and 2. The use of anchor piles to provide reactíon for the test
Jain (1967) to compare this rnethod with the rnaintained- load (Fig. I6.4b).
load pn.-";edure. A typical cornparison, shown in Fig. 16.3, 3. The use of ground anchors to provide reaction for the
reveals excellent agreement, in regard to both ullirnate load test load (Fig. 16.4c).
capacity and load-settlement behavior.
16.5 SOURCES OF ERROR IN SETTLEMENT 16.5.1 Errors Resulting from Use of Reference Beam
MEASUREMENTS lN PILE LOAD TESTS
With this system of settlemer:t measurement, the bearn
Sorne of the loading and settlement procedures commonJy supports settle because of the loaded pile. A theoretical
used in pile load tests may lead to inaccuracies in the assessment of the resulting errors involved in settlement
rneasurement of the settlement of a test pile: A theoretical rneasured may be made by using the solutions for the
examination of such inaccuracies caused by the following settlement of a point on the surface of the soil caused by
procedures (see Fig. 16.4) has been made by Poulos an·.1 a loaded pile (Section 5.3.3). From Eq. (5.40), this settle-
Mattes (1975): ment, Ps, ís given by
360 PILE LOAD TESTS
Joc ~
"..j
J ------LE
- - ¡-----
Ha
L
- r!- r!- !
S
1
(16.10)
where fp ís plotted in Figs. 5 5.33, and 5.34. The true
settlement of the piJe itself, p, resulting from the applied
load, is given from Eq. (5.33) as
The correction factor Fe '~valuated for a floating pile is
plotted in Figs. 16.5 and 16.6 for the cases of a piJe in a
p (16.6)
deep !ayer and a piJe in a finite ]ayer. Figure 16.5 indicates
that serious errors (i.e ., values of Fe) can arise in
where settlement measurements on a test píle in a deep soillayer
unless each support of the reference beam is placed about
1 = set tlement-influence factor 0.5 to 1 píle-length away from the pile. In terms of the
dimensionless distance rfL, the effect is more severe for
The measured settlement, Pm, is therefore shorter piles. Figure 16.6 shows how the effect of the
support-beam movement diminishes with decreasing soíl-
Pm = P ~ Ps layer thickness. However, even for an end-bearing p,ile
(H/L = 1), it is desirable to ha ve the supports 0.3 to 0.5
pile-lengths away fwm the test pile.
(16.7)
It is convenient now to define a correction factor, Fe, 16.5 .2 Errors Resulting from .Jacking Against Anchor Piles
to be applíed to the measured settlement, Pm, to obtain the
true settlement p~that is, With this method of load application, the upward loads on
the anchor piles cause an upward movement of the test pile
(16.8) because of interaction. As a result, if the settlement of the
test piJe is measured with reference to a remate benchmark,
or, the correction factor Fe is defined as the measured settlement will be less than the true settle-
ment. A theoretical examina~ion of the magnitude of this
under-registration may be made by using the pile-settle-
Fe (16.9)
ment interaction solutions described in Section 6.2.
PILE LOAD TESTS 361
2·5,---·--,------,------,-----.
2·0
1·5
1 r
1·0 -----------------
L
K = 1000
0·5 Vs= 0·5
FIGURE 16.5 Correction factor F" for floating pite in deep !ayer of soil.
The true settlement of the test pile resulting from the O!¡ interaction factor for two piles at a spacmg of s,
applied load is again given by Eq. (16.6). The upward move- where s is the distance between the test pile and
ment of the test pile because of the reaction on the anchor each reaction pile (Fig. 16.4b).
piles is given (using Eq. 6.12) as
The measured settlement, Pm. relative to a remote bench-
PI mark, is therefore ·
t::..p = dE . o:¡ (16.11)
S
Pm P - !::..p
where
(:f:J (1 - 0:¡) (16.12)
(16.13)
SoH
Es
K ~ 1000
3'0 .----r---~--- .......---r-----.
Ljd = 25
K = 1000
v, ~ 0·5
Fe
2
Fe
~~----~------~----~------~----~
o 5 10 15 20 25
-m7'2
d
.p d
(!2
FIGURE 16.9 Correction factor Fe. Effect of bearing stratum for
S .~ L end-bearing piJe jackcd against two .reaction piles.
/7.. .
Rig1d Stratum
not occur at normal spacings unless the piles are relatívely
s1ender and compressible.
The effect of the relative stiffness of the bearing
stratum on Fe ís shown in Fig. ! 6.9. As the bearing stratum
becomes stiffer, interaction decreases and hence Fe de-
1 ·8 K 200
--K~ 1000 creases hr a given pile spacing. However, signifícant t~rrors
in settlement measurement may still occur at normal piJe
Values of L¡d
spacings unless the bearing stratum has a stiffncss about lO
100 times ( or more) greater than the overlying soil.
Figures 16.7, 16.8 and 16.9 suggest that if settlement
measurements are made wíth reference to a rcmote bench-
1. 4
mark, the usual spacing of abo u t three díameters m ay result
in signifícant undermeásurement of the sett:~ment of the
test pile. Increasing the spacing to at least five díameters
would appear most desírable, especially for long piles ín
deep, soft deposits.
An alternative means of settlement measurement is
possible with the anchor-pile system- by measuring the
settlement of the test pile with reference to the reaction
piles, that is, by fLxing a dial gauge to the cross beam
FIGURE 16.8 Correction factor Fe for end-bearing pile on rigid JO!mng the reaction piles. The consequences of this pro-
stratum :3cked agaínst two reaction pil~s. cedure may again be examined usíng píle-interaction
PILE LOAD TESTS 363
Pa
PI _
dE (0.::> - a 1 + O.S<tz) (16 14)
S 2·0....---.-.,-
where
1·5
1 single-pile settlement :.nfluence factor
ínteraction factor for two pile, ata spacing of s
ínteractíon factor for two piles at a spacing of 2s F¿
1·0
Defining a correctíon factor}~ as FIGURE 16.10 Correction factor for floating píle in a deep
layer jacked against two reactior, piles-settleme:1t measured in
F = True settlement of loaded piJe
' (16.16)
relation to anchor piles
e Measured settlement re1ative
to reaction piles ment, in that it gíves a settlement eíther closer to or larger
than the real settlemenL However, in any such pile test,
it may be shown from Eqs. (16.15) and (16.16) that measurement of the settlement by both the alternatíve me-
thods is desirable so that a b·~tter assessment or' the true
settlement may be made.
Fe = .
(1.5- 2a 1 + 0.5o: 2)
fl6.17)
All the above solutions apply for a homogeneous soil
stratum. The expressíons in Eq. 16.13 and Eq. 16.17 also
Values of Fe 're plotted against dimensionless spacing apply for non-homogeneous smls, provided that appropriat~
s/d in Fíg. 16.10 for a floating pile in a deep soíl !ayer. values of the ínteraction factors are used. Sínce these
Comparison with Fig. 16.7 shows that Fe
ís generally less factors tend to be smaller for non-homogeneous soils than
than , in other words, less correction of the measured for homogeneous soils, the ,mors involved in the test
settlement is required if measurement ís made wíth respect procedure will be correspondingly smaller; hnwever, the
to the anchor píles. For piles of medium compressibílity general characteristícs of behaYiour and variation t)f Fe and
(K= 1000), Fíg. 16.10 shows that Fe is above unity ata Fe with spacing remain similar.
spacing of about five diameters. It must be poínted 9ut,
however, that at larger spacings or in cases where 1ittle
interaction is likely to occur between the test pile and the 16.5.3 Errors Resulting from Jacking Against Ground
reactíon piles, Fe will be less than one-that is, the mea- Anchors
sured settlement will be greater than the true sett1ement.
In such cases, the soil modulu:> backfigured {rom the un- The upward reaction on each ground anchor will tend to
corrected measured settlement would be too small in con· reduce the settlement of the test pile. Because the cables
trast to the va,lue that wou1d be obtained from the settle- for the ground anchors are generally cased and the anchors
ment measured with reference to a remote point, which themselves are srnall in relation to the test pile, it is reason-
would be too large. Thus, measurement of the test-pile able to approximate each anchor as an upward point load
settlement re1ative to the reactíon piles would appear to acting at the center of thé anchor. To simplify calculations,
have advantages over other means of settlement measure- · it is then assumed that the effect of the ground anchor on
364 PILE LOAD TESTS
the test piJe is the same as its effect on a point located half !ayer. Figure 16.1 1 shows that if the anchors are located
way along the piJe. With ~he above approximations, the 1.5 pile-lengths or more below the surface, Fe is less than
upward movement, !J.p, of the test pile caused by the 1.2, in other words, the error in the measured settlement is
ground anchors can be written as less than 20%. Beyond an anchor depth of about 2L, the
radial distance of the anchors from the piles has little effect
on the measured settlement.
( 16.18)
The case considered in Fig. 16.11 is not likely to occur
frequently in practice, since to obtain adequate load
where 1M is the vertical displacement factor for a buried capacity, the anchors are usually secured into a stiffer !ayer
point load. 1M may be evaluated most readily from at or below the leve! of the pile tip. In such a case, the up-
Mindlín's equation for a p • t load within a semi-infiníte ward movements caused by the anchors would be le;ss than
elastíc mass. given by Eq. (16.18), so that Fe will be less than indicated
The true piJe settlement, Pr, is again given by Eq. in Fig. 16.11.
( 16.6 ), so the measured settlement Pm of the test pile is Figure 16.11 will also generally give an overestirnate of
F'c for an end-bearing test-pilc bearirtg on a stiff )ayer. The
extreme case of a piJe through very soft soil and bearing on
Pm (16.19)
a stiff !ayer may be examined by considering the pile tip as
a rigid circular area carrying the total applied load, and the
Defíning the correction factor F""c as in Eq. ( 16 .9), anchors as point loads acting on the surface of a semi-
infinite mass of modulus Eb (the bearing stratum). For
this case, Fe is plotted in Fig. 16.12 together with the other
Fe (16.20)
lMt!_) limiting case of a piJe in a homogeneous deep )ayer, with
1 L anchors at the leve! of the piJe tip (the curve for Ha/L = 1.0
Lfd ~ 25 Vs <0·5
K= 1000
1 ·3
Fe
1 ·2 1·1
1·0 ¡___ _..J...__ _...J...._ _.......L_ ___¡ 1· o .____ ..J..__ __ , _ _ __¡__ _.......L_ ____..J
o 0·5 2·0 o 5 10 15 20 25
FIGURE 16.11 Correction factor Fe for floating pile in a deep FIGURE 16.12 Correction factor Fe for end-bearing piJe jacked
!ayer jacked against ground anchors. against ground anchors.
PILE LOAD TESTS 365
in Fig. 16.11 ). It m ay be seen that when the pile bears on Lateral load tests are usually used to índícate the load-
to very stiff rock through very soft soil, (Eb/Es-+ ""),Fe is def1ectíon behavior of a pile. The allowable design load is
extremely small even for very closely-spaced anchors, often taken as the load required to produce a specified
whereas the corresponding value for the homogeneous layer det1ection (e.g., 0.25 in) divided by the requíred factor of
(Eb/Es 1) is considerab1y grea:er. In practice, the value of safety (McNulty, 1956). The values of the subgrade-reac-
Fe would lie between these two limiting values. tion modulus of the soil or the elastic modulus ofthe sóil
Figure 16.12 indicates tha1: when anchors are to be may also be backfigured from 1 test by fitting the observed
fixed at the leve! of the pile tip, the spacing between the behavior to the theoretical, as described in Section 8.5. This
test pile and the anchors should be as great as possible, and procedure enables predictions to be made of movements of
preferably 10 diameters or greater. Greater spacings may be piles of other dimensions or of groups of laterally loaded
achieved most readily by installing inclined anchors. piles. In such cases, the influence of any soft Iaxers under-
Comparison with the other two test systems shows that lying the pile tips is o f much less significan ce than in the
Fe for the anchor system is generally much less, that is, case of verticalloading.
less error ís involved in settlement measurements when
anchors are used.
16.6 LATERAL LOAD TESTS Axial load tests presently represent thé only certain method
of determining the ultimate axial load capacity of indm-
The usual method of carrying out lateral load tests is to dual piles, but conventional procedures and equipment are
insta!] a pair of piles and jack their heads apart. A descrip- relatively costly and inconvement, especially if high load
tion of the method is given by Wagner ( 1953). The piles capacítiesare antícipated.
should be placed sufficíently fa.r apart so as not to obtain · The possibility of carrying out torsional loading tests
significant interaction btttween the movements of each on piles has been examined by Stoll ( 1972), who devised
pile, and hence a horizontal beam is frequently inserted a simple field torque-shear-load test that could be applied
between the piles and the jack reacts against one of the [1ile to cylindrical piles. The piles must, of course, be.capable of
heads and the beam to the other pile head. If necessary, the cárrying the required torque wíthout failure of the pile
effects of interaction with less than ideal spacing between material ítself, so that in stiffer soils, relatively short or
the piles can be estimated, as explained for vertical tests in stiff piles may be necessary. The key feature pf the test
Section 16.5, using the theory given in Chapter 8, Section apparatus is that torque is applied by small-capacity hy-
8.4. Examples of loading arrangements for both vertical and draulic jacks reacting horizontal! y against adjoining job
battered piles have been described by Alizadeh and piles, utilizing the large mechanical advantage available
Davisson (1970) and Tomlinson (1977). at the usual spacing (i.e., 3 to 5 ft or wider).
The procedure employed for the test varíes, but typi- Stoll was of the opinion that the pile-soil shear strength
cally the load is applied in a number of increments and each from torsional tests would not exceed the value for axial
increment is left on until a specified rate of movement is loading.' Model tests reported by Poulos O9'7 5p) showed
reached. Alizadeh and Davisson (1970) used for each incre- that in clay, the values deduced from axi<,1i and torsional
ment a mínimum period of one hour, or un til the pi! e head tests were in fact very similar. H was also found possible in
movement was less than 0.01 in per hr. Tests we re carried these tests to backfigure the shear modulus of the soil from
to lateral det1ections approaching 2 in. the measured to.rque-rotation relationship, using .the elastic
Lateral deflection of the pile head is usually measured theory described in Section 9.3 .3, and to use this val u e to
with a dial gauge. Strain gauges are also frequently installed predict the settlement of a pile loaded axially. Thus, there
along the embedded portión of the pile to measure flexura! is sorne evidence that torsionalload tests may be useful for
stresses whereby the bending moments may be obtained. predicting the behavior of axially loaded piles in clay. For
With steel piles, inclinometers may also be installed with sands, however, beca use of the dependence of the pile-soil
the pile to measure the variation of lateral def1ection with s..'I-J.ear strength on the stress state, torsional tests may give
depth along the piJe. Hanna (1967) has also employed misleading results; for example, for model piles in ,sand,
inclinometer readings along steel H-piles to indicate the Broms and Silberman (1964) obtained considerably tower
bending of piles during driving. values from torsional tests than from axial tests.
APPENDIXA
JNTEGRATION OF MINDLJN'S EQUATIONS FOR PILE SETTLEMENT ANALYSJS
ió rr(2
Iq j J pi dO de (A I)
(j- 1)b o
where
L
j_.
From Mindlin's equation, PI is given by '
2 2
( 1+v) {. z 1 ( 5- 1 2v + 8v )
1 = ---- -- + (-3 ··-4v)
- + .:._____,...:.
P ,Q.n(l-v) R 3 D R
. 1 '' l 2
2 2 2
;;..;('-3_-_4_v'-)z_
[ _ _2,-c_·z_+_2_c....::] + [ 6ez (z- e)]}
+ 3
(A2)
5
R2 R2
where
z h +e
Z¡ }¡ e
d2 Figure Al Single pile--basic geomctry.
R~ -- + - xdcosO + z2
4
d2
Id =
4
+ - xd cosO + z 1 2 where
and
(A3)
z from h+(j- 1)ó to h+ió
366
APPENDIX 367
2 2
(l+v) 1 z1 (rA - Ro )
];plrdr = 87T(i-v) (Ro )
711
\ + (3 - 4v)[y'x;; + A In (2y'x;; + 2r - 2A))
surta ce
\ + [(3 4v)z 2
+ 6ez 2 (z- e) {
2ez + :?e
-
1
3y'X¡ 3
+A [
2
]
r- A
(B
3(B-A 2 )fft
where
2rr
lbj "' 1T J
(j- l)
pi de (A6)
1 [¡
d fo (A4)
l;he integral in Eq. (A6) is given in Eq. (A3), with h = L,
D2 z 2 +d 2 /4andDi =z 1 2 +·d 2 /4.
where For the vertical displacement of the base resulting from
the base itself, it is desirable t o rnake an approxirnate allow-
pi is given in Eq. (A2), and for this case, ance for the effect of the rigidity of the base by rnultiplying
the displacernent of the centt:r of the uniformly-loaded dr-
e no =·L
R~ z 2 + + r2 - 2rx ces e cular base by a factor of ¡; this is the ratio of the surface
R12 z 1 2 + x 2 + r 2 - 2rx eos ()
z z 1 + 2e displacement of a rigid circle on the surface of a half-spaée
368 APPENDIX
to the center displacement of a corresponding uniformly The integral in Eq. (A 7) can readily be evaluated analyti-
loaded circle, and may be assumed to apply approximately cally, and gives
to embedded areas. Thus,
rr(l+v) ~ db
rb lbb = (3-4v)- + (5- !2v+Bv 2 )(R z)
:::: -·
1r
4
Jpi r dr (A7)
16(1-v)d 2
o
(A8)
where
r
o z = Z¡ + 2c = 2L
APPENDIXB
ELASTJC EQUA T!ONS USED FOR LATERALLY-LOADED P!LE ANALYSIS
2
Q [(3- 4v) + _1_ +.::..:_ + (3 4v)x
Px 16JTG(l-v) R1 R 2 Rf R~ Px 32
1Tf:;1__ v) [(3- 4v)F 1 +F 2+40 2v) (83)
2
2cz( 3x )+4(1-v)(l 2v)X X (1- P)F3]
+ -RJ21 - R¡ R 2 +z+c
where
(Bl)
2c 1
K¡
b
Doug1as and Da vis (1 964) integrated this equation o ver
a rectangular area, and obtained the followíng so1utíon. 2c 2
At the upper corners -;1 and B (see Fig. B2), for a K2
b
uniform horizontal pressure p,
F¡ -(K¡-K2) In
[ (K¡-K2)
2+,/4+(K¡-K2)2
j
- 2 In [ , . 2
(K 1 -K2)+.J4+(K 1 - K2)
2 J
F2 21n
[ 2(K 1 +,/1 +K~ )
•
J
+(K 1 -K 2 )X
(K 1 +K 2)+'\1'4+(K 1 +K2r
\ o
J
Surface z
l
¡
zl Q
1
_L_
1
lA %c• •
B
o~~ e
(Loading in x direc~ionl
369
370 APPENDIX
ln (K :K X
1 2
)
- (K 1 +K 2 )]
[V4+(K¡+K2) 2 - (K.+Kz)l-Kz(K2-V!+K~)
K.(vll+Kr- K¡)
F
4
=-"~In[ 2 2
(K +y'i"+K1) l+ For the displacement at other points in the same plane, the
- <K 1 +K 2 )+V4+(K 1 +K 2 ) 2j principie of superposition may be employed.
REFERENCES
Adams, J. l. & Hanna, T. H. 1970. "Ground Movements Banerjee, P. K., & Dríscoll, P. M. 1976. "Three-Di"mensíonal
Due to Pite Driving." Conf. on Behavior of Piles, Inst. Ana1ysís of Raked Píle Groups." Proc. Instn. Civ. Engrs,
Civ. Engrs., London. Part 2, vol. 61: 653-671.
Agarwal, S. L. 1973. "Discrete Element Analysis and Its Banerjee, P. K., & Davies, T. G. 1978. "The Behaviour of
Experimental Verification for Vertical Piles Under Dy- Axially and Laterally Loaded Single Piles Embedded in
namic Lateral Loads." Proc. 8lh Int. Conf. S.M. & F.E:, Non-Homogeneous Soils." Geot., vol. 28, no. 3; 309-
Moscow, vol. 2.1, paper 3/2: 9-12. 326.
Agerschou, H. A. 1962. "Analysis of the Engineering News Banerjee, P. K. 1978. "Analysis of Axially and Laterally
Pile Formula." J.S.M.F.D., ASCE, 88, SM5: 1-11. Loaded File Grol'ps," in Dcvelopments in Soil Me-
Airhart, T. P., Coyle, H. M., Hirsch, T. J. & Buchanan, S. J. chanics, Ed. C. R. Scott, Ch. 9. London, Applied
1969. "Pile-Soil System Response in a Cohesive Soil." Science Publíshers.
ASTA!, STP 444: 264-294. Barber, E. S. 1953 Discussíon to Paper by S. M. Gleser.
Aitchison, G. D. & Woodburn, J. A. 1969. "Soil Suctroo in ASTl\-1, STP 154: 96-99.
Foundation Design." Proc. 7th Int. Conf. S.M. & F.E., Barden, L. & Monckton, M. F. 1970. "Tests on Model Pile
vol. 2: 1-8. Groups in Soft and Stiff Clay "Gevt., vol. 20: 94-96.
Alizadeh, M. 1969. "Lateral Load Tests on Instrumented Barkan, D. D. 1962. Dynamícs of Bases and Fatmdations.
Timber Piles." ASTA!, STP 444; 379-394. New York: McGraw-Hill.
Alizadeh, M. & Davisson, M. T. 1970. "Lateral Load Tests Barker, W. R. & Reese, L. C. 1970. "Load-Carrying Charac-
on Piles-Arkansas River Project." J.S.M.F.D., ASCE, terístics of Drilled Shafts Constructed with the Aid of
vol. 96, SM5: 1583-1604. Drilling Fluids." Res. Rep. 8'l-9, Center for High. Res.,
Avery, S. B. & Wilson, S. D. 1950. Discussion to Paper by Univ. of Texas, Austín.
Cummings, Kerkhoff, & Peck. Proc. ASCE, vol. 75, B.C.P. Committee 1971. "Field Tests on Piles in Sand."
119Q-1199. Soils and Fndns., ¡·ol. 11, no. 2: 29.
Awad, A. & Petrasovits, G. 1968. "Considerations on the Begemann, H. K. S. 1953. "Impnved Method of Determin-
Bearing Capacity of Vertical and Batter Pites Subjected ing Resistance to Adhesion by Sounding Through a
to Forces Acting in Different Directions." Proc. 3rd Loase Sleeve P1aced Behind the Cone." Proc. 3rd Int.
Budapest Conf. S.M. & F.E .. Budaoest: 484-497. Conf. S.M. & F.E., vol. 1: 21>217.
Baguelin, F., Jézéquel, J. F., & Shields, D. H. 1978. "The Begemann, H. K. S. 1965. "The Maximum Pulling Force on
Pressuremeter and Foundation Engineering." Clausthal: a Single Tension PiJe Calcula red on the Basis of Results
Trans Tech Publications. of the Adhesion Jacket Cone." Pro c. 6th In t. Con f.
Baguelin, F., & Frank, R. 1979. "Theoretical Studies of S.M. & vol. 2: 229-233.
Piles Using the Finite Element Method." Proc. Con f. Begemann, H. S. 1969. ''Negative Skin Friction of a Sin-
Num. Methods in Offshore Piling, London. Inst. Civ. gle PiJe." Spec. Session no. 8, 7th In!. Conf. .S.M. &
Engrs., Paper no. 11. F .E., paper no. l.
Baikoff, E. M. A. & Burke, T. J. 1965. "Determínation of Bender, C. H., Lyons, C. G., & Lowery, L. L. 1969. "Appli-
Foundations for Heaving Soils." Trans. So. African cations <Of Wave-Equation Analysis to Offshore Pile
Instn. Civ. Engrs., vol. 7, no. 8 Foundations." l st Offshore Tech. Con f., Texas, paper
Balaam, N. P., Poulos,H.G.,&Booker,J.R. 1975. "Finite OTC 10S5.
Element Analysis of the Effects of Installation on PiJe Beredugo, Y. O. 1966. "An Experimental Study of the
Load-Settlement Behavior." G'eot. Eng., voL 6, no. 1: Load Dístribution in PiJe Groups in San d." an. Ceo t.
33-48. Jnl.,vol.3,no.3: 145-166.
Ba1aam, N. P., Booker, J. R., & Poulos, H. G. 1976. "Ana- Berezantzev, V.G., Khristoforov, V., & Golubko , V. 1961.
1ysis of Granular Pi1e Behavior Usíng Finite E1ements." "Load Bearing Capacity and Deformatioo o.f Piled
Univ. of Sydney Civ. Eng. Res. Rep. R290, Aust. Foun~ations." Proc. 5th lnt. Conf. S.M. & F.E., vol. 2:
Baila, A. 1961. "The Resistance to Breaking Out of Mush- 11-15.
room Foundations for Py1ons." Pro c. 5th In t. Conf. Bishop, R. F., Hill, R. & Mott, N. F. 1945. "The T!\eory of
S.M. & F. E., vol. 1: 569-576. Indentation and Hdrdness Test." Proc. Phys. Soc., 57:
Banerjee, P. K. and Davies, T. G. 1977. "Analysis of Pile 147-159. .
Groups Embedded in Gibson Soil." Proc. 9th In t. Con f. Bjerrum, L. 1957. "Norwegian Experiences with Steel Piles
S. M. & F. E., Tokyo, voL 1: 381-386. to Rack." Geot., vol. 7: 73-96.
371
372 REFERENCES
Bjerrum, L, Johnson, W., & Ostenfe1d, C. 1957. "The Set- Broms, B. B. 1966. "Methods of Calculating the Ultimate
tlement of a Bridge Abutment on Frictíon Piles." Proc. Bearing Capacity of Piles-A Summary." Sols-Soils no.
4th Int. Conf. S.M. & F.E., vol. 2: 14-19. 18-19: 21-32.
Bjerrum, L., Brinch Hansen, J., & Sevaldson, R. 1958. Broms, B. B. 1972. "Stability of Flexible Structures (Piles
"Geotechnical Investigatíons for a Quay Structure ín and Pile Groups)." Proc. 5th Eur. Conf. Soil Mech.
Horten." N.G.L, pub. no. 28: 1-17. Fndn. Eng., Madrid, vol. 2: 239-269.
Bjerrum, L. & Johannessen, I. J. 1960. "Po re Pressures Re- Brons, K. Amesz, A. W., & Rinck, J. 1969. "The Nega-
su1tíng from Driving Piles in Soft C1ay." Conf. on Pore tive Skin Frictíon Along the Shaft of a Foundation
Press. & Suction in Soil, Butterworths, Sydney, Aust.: Píle." Spec. Sess. No. 8, 7th Int. Conf. S.M. & F.E.,
14-17. • paper no. 2.
Bjerrum, L. & Simons, N. E. 1960. "Comparison of Shear Brown, P. T., Poulos, H. G., & Wíesner, T. J. 1975. "Piled
Strength Characteristics of N ormally Consolídated Raft Foundation Design." Proc. of Symp. on Raft
Clays." Proc. ASCE Res. Con f. on Shear Strength of Foundations, CSIRO, Aust.
Cohesive Soils: 711-726. Brown, P. T. & Wiesner, T. J. 1975. "Thel3ehaviorofUni-
Bjerrum, L, Johannesson, L J., & Eíde, O. 1969. "Reduc- formly Loaded Piled Strip F ootings." Soils and Fndns,
tion of Skin Fríction on Steel Piles to Rock." Proc. 7th vol. 15, no. 4: 13-21.
Int. Conf. S.M. & F.E., vol. 2: 27-34. Burland, J. B. (1973). "Shaft Friction of Piles in Clay -A
Blight, G. E. 1965. "The Time-Rate of Heave of Structures Simple Fundamental Approach." Ground Eng., voL 6,
on Clays." ittoisture Equilibria and liJoisture no. 3, May: 30-42.
Changes in Soils Beneath Covered Areas, Butterworths, Burland, J. B., Butler, F. G., & Dunican, P. 1966. "The Be-
Sydney, Aust.: 78. haviour and Design of Large-Díameter Bored Piles in
Blight, G. E. 1967. "Observations on the Shear Testing of Stiff Clay." Proc. Symp. on Large Bored Piles, Inst. Civ.
Indurated Fissured Clays." Proc. Geot. Conf., Oslo: 97. London: 51-71.
Bogdanovic, L. J. l 961. 'The Use of Penetration Tests for Butterfield, R. and Banerjee, P. K. 1970. "A Note on the
Determining the Bearing Capacíty of Piles." Proc. 5th Problem of a Píle-Reinforced Half S pace." Geot., vol.
Int. Conf. S.M. & F.E., vol. 2: 17-22. 20, no. 1: 100-103.
Booker, J. R. & Poulos, H. G. 1976a. "Analysís of Creep Butterfield, R. and Banerjee, P. K. 197la. "The Elastic
Settlement of I'ile Foundations." Jnl. Geot. Eng. Divn., Analysis of Compressible Píles and PiJe Groups." Geot.,
ASCE,vol. 102,no.GT!: I-14. vol. 21, no. 1: 43-60.
Booker, J. R. & Poulos, H. G. 1976b. "Fíníte Element Ana- Butterfield, R. and Banerjee, P. K. 1971 b. "The Problem of
lysís of Piles in Viscoelastic Soil." Proc. 2nd In t. Con f. PiJe Group and Pile Cap Interaction." Geot., voL 21,
on Numerical Methods in Geomechanics, Blacksburg, no. 2: 135-142.
vol. 1· 425-437. Cambefort, H. 1953. "La Force Portante des Groupes de
Bowles, J. E. 1977. Foundation Analysis and Design. 2nd Pieux." Proc. 3rd Int. Conf. S.M. & F .E., voL 2: 22-29.
Ed. New York: McGraw-Hill. Cambefort, H. and Chadeisson, R. 1961. "Critere pour
Bozozuk, M. & Labrecque, A. 1969. "Downdrag Measure- I'Evaluation de la Force Portante d'un Pieu." Proc. 5th
ments on 270-Ft Composite Piles." ASTM, STP 444: Int. Conf. S.M. & F.E., vol. 2: 23-31.
15-40. Chae, Y. S. 1970. "Dynamic Behaviour of Embedded Foun-
Bozozuk, M. 1972. "Downdrag Measurements on a 160-Ft dation-Soil Systems." Hígh. Res. Rec. no. 323: 49-59.
Floating Pipe Test Pile in Marine Clay." Can. Geot. Jnl., Chan, J. H. C. & Matlock, H. 1973. "A Discrete-Eiement
voL 9, no. 2: 127-136. . Method for Transverse Vibratíons of Beam Columns
Brandtzaeg, A. & Harboe, E. 1957. "Buckling Tests of Slen- Resting on Linearly Elastic or Inelastíc Supports." Pwc.
der Steel Piles in Soft, Quick Clay." Proc. 4th Int. Conf. 5th Offshore Tech. Conf., Houston, VoL 2: 205-218.
S.M. & F.E., vol. 2: 19-23. Chandler, R. J. 1966. Discussion to paper by Whitaker and
Brinch Hansen, J. 1961. "The Ultímate Resistance of Rigid Cooke. Proc. Symp. on Large Bored Piles: 95-97.
Piles Against Transversal Forces." Geoteknísk Institut. Chandler, R. J. 1968. "The Shaft Friction of Piles in Cohe-
Bull. No. 12, Copenhagen. sive Soils in Terms of Effective Stress." Civ. Eng. and
Bromham, S. B. & Styles, J. R. 1971. "An Ana!ysis of Pile Pub. Wks. Rev., Jan.: 48- 51.
Loading Tests in a Stiff Clay." Proc. 1st Aust.-N.Z. Chellis, R. D. 1961. Pi/e Foundatíons. New York: McGraw-
Conf. Geomechs., Melbourne, vol. 1: 246-253. Hill.
Broms, B. B. 1963. "Allowable Bearing Capacíty of Initially Chellis, R. D. 1962. '~Pile Foundations," ch. 7 in Founda-
Bent Piles.'' J.S.M.F.D., ASCE, vol. 89, SM5: 73-90. tions Engineering, ed. by G. A. Leonards. New York:
Broms, B. B. & Silberman, J. O. 1964. "Skin Friction Resis- McGraw-Hill.
tance for Pi! es in Cohesionless Soils." Sois Soils, no. 10: Chellis, R. D. 1969. "Piles and Píle Structures," in Hand-
33. book of Ocean and Underwater Engineering, ed. by
Broms, B. B. 1964a. "Lateral Resistance of files in Cohe- Myers, Holm, & McAllister. McGraw-Híll, New York:
síve Soils." J.S.M.F.D., ASCE, vol. 90, SM2: 2 7-63. 8.56-8.98.
Broms, B. B. 1964b. "Lateral Resistan ce of Piles in Cohe- Claessen, A. l. M. & Horvat, E. 1974: "Reducing Negative
. sionless Soils." J.S.,ll,f.F.D., ASCE, vol. 90, SM3: 123- Friction with Bitumen Slip " Jnl. Geot. Eng.
156 .. Divn., ASCE, vol. lOO, no. GT8: 925-944.
Broms, B B. 1965. Discussion to paper by Y. Yoshimi. Clemence, S. P. & Brumund, W. F. 1975. "Large-Scale
J.S.M.F.D., ASCE, vol. 91, SM4: 199-205. Model Test of Drilled Pier in Sand." Jnl. Geot. Eng.
REFERENCES 373
Div, ASCE, voL 101, No. GT6, 537-550. ticity to Foundation Problems-Limit Analysis." Post-
Coates, D. F. & Gyenge, M. 1956. "Plate-Load Testing on Grad. Course on Fndn. Eng., Univ. of Sydney, Aust:
Roe k for Deformation and Strength Properties." Test· Davis, E. H. & Booker, J. R. 1971. "The Bearing Capacity
ing Tech. for Rack Mechs., ASTM: 19-25. of Strip Footings from tl1e Standpoint of Plasticity
Coates, D. F. 1967. Rock Mechanics Principies. Can. Mines Theory." Proc. 1st At1st.-New Zeal. Conf. on
Branch Monograph 874, Ottawa. Geomechs., Melbourne, Vol. 1: 276-282.
Collins, L. E. 1953. "A Pre1iminary Theory for the Design Davis, E. H. & Poulos, H. G. 1,?63. "Triaxial Testing and
of Underreamed Piles." Trans. So. African Instn. Civ. Three-Dimensional Settlement Analysis." Proc. 4th
Engrs., vol. 3, no. 11. Aust.-N.Z. Conf. S.M. & F.E.: 233-246.
Collins, L. E. 1958. "Sorne Observations on the Movement Davis, E. H. & Poulos, H. G. 1968. "The Use of E1astic
of Buildings on Expansive Soil in Vereeniging and Theory for Sett1ement Prediction Under Three-Dimen-
Odendaalrus." Trans. So. African lnstn. Civ. Engrs., siona1 Conditions." Geot., vol. 18: 67--91.
vol. 7, no. 9. Da vis, E. H. & Poulos, H. G. 1972. "Rate of Settlement
Cooke, R. W. & Whitaker, R. 1961. "Experiments on Model Under Two- and Three-Dim:~nsional Conditions." Geot.,
Piles with Enlarged Bases." G'eot., vol. 11: l-13. voL 22: 95-114.
Cooke, R. W. 1966. Discussíon to paper by Burland et al. Davis, E. H. & Poulos, H. G. 1972. 'The Ana1ysis of Pile-
Proc. Symp. on Large Bored Piles: 92. Raft Systems." Aust. Geomechs .. Jnl., vol G2, no. l:
Cooling, L. C. & Packshaw, S. 1950. "Notes of Pile-Loading 21-27.
Tests." Chart. Civ. Engr., May: 16-20. Davis, f.. H. & Taylor, H. 1962. "The Movement of Bridge
Cox, A. D. 1962. "Axially Symmetric Plastic Deformation Approaches and Abutments on Soft Foundation Soils.';
in Soils--IL Indentation of Ponderable Soils." lnt. J. Proc. 1st Aust. Road Res. Bd. Conf, Canberra, p. 740.
Mech. Sci., vol. 4: 371. Davisson, M. T. & Gill, H. L. 1963. "Laterally Loaded Piles
Cox, A. D., Eason, G., & Hopkins, H. G., 1961. "Axially in a Laye red Soil System." J.S.M.F.D., ASCE, vol. 89,
Symmetric Plastic Deformation in Soils." Phil. Trans. SM3: 63-94.
Roy. Soc. London, series A, 254: l-45. Davisson, M. T. 1963. "Estimating Buckling Loads for
Coyle, H. M. & Reese, L. C. 1966. "Load Transfer for Piles." Proc. 2nd Pan-Amer. Conf. on S.M. & F.E.,
Axially Loaded Piles in C1ay." J.S.M.F.D., ASCE, vol. Brazil, vol. 1; 351-371
92, SM2: 1-26. Davisson, M. T. & Prakash, S. 1963. "A Review ofSoii-Pile
Coyle, H. M. & Sulaíman, l. H. 1967. "Skin Friction for Behavior." High Res. Rec., No. 39: 25-48.
Stee1 Piles in Sand." J.S.M.F.D., ASCE, vol. 93, SM6: Davisson, M. T. & Robinson, K. E. 1965. "Bending and
261-278. Buckling of Partially Emb<;dded Piles." Proc. 6th Int
Coy1e, H. M., Bartoskewitz, R. E., & Lowery, L. L. 1970. Conf. S.M. & F.E., voL 2: 253-246.
"Prediction of Static Bearing Capacity from Wave Equa- Davisson, M. T. 1970. "Lateral Load Capacity of Piles."
tion Ana1ysis." 2nd Offshore Tech. Conf., Texas, paper Hígh. Res. Rec., No. 333: 104-112.
OTC 1202. . Dawson, A. W. 1970. "Downdrag on Pile Foundations."
Coyle, H. M. & Gibson, G. C. 1970. "Empírica! Damping M. Se. Eng. project, Dept. Civ. Eng., M.I.T.
Constants for Sands and Clays." J.S.M.F.D., ASCE, De Beer, E. E. & Wallays, M. '1972. "Forces lnduced in
vol. 96, SM3: 949-965. Piles by Unsymmetrical Surcharges on the Soil Around
Crooke, R. C. 1955. "Re-Analysis of Existing Wave Force the Pile." Proc. 5th Eur. Conf. S.M. & F.E., Madrid, voL
Data on Mode1 Piles." U.S. Army Corps Engrs., Beach 1: 325-332.
Erosion Board Tech. Memo 7 l. De Bruijn, C. M. A. 1961. "~ welling Characteristics of a
Cumming; D. A. & Gerrard, C. M. 1964. "Computatíon of Transported Soil Profile at Leeuhof, Vereeniging
Stresses in Pavements." Proc. 2nd Conf. Aust. Rd. Res. (Transvaal)." Proc. 5th lnt. Conf. Soil Mechs. Fndn.
Bd., 2, p.(. 2: 729. Eng., vol. 1: 43.
Cummíngs, A. E., Kerkhoff, G. 0., & Peck, R. B. 1950.
de Mello, V. F. B. 1969. "Foundations of Buildings on
"Effect of Driving Piles in Soft C1ay." Trans. ASCE,
Clay.'' State of the Art Report: 49-136. Proc. 7th Int.
vol. 115: 275-286.
Conf. S.M. & F.E., Mexico C::ity.
D'Appo1onia, E. & Romualdi, J. P. 1963. "Load Transfer in
End-Bearing Stee1 H-Piles." J.S.M.F.D., ASCE, vol. 89, de Mello, V. F. B. 1971. "The Standard Penetration Test."
SM2: 1-25. Proc. 4th Pan-Amer. Conf. S.M. & F.E., Puerto Rico,
D'Appolonia, E. & Hribar, J. A. 1963. "Load Transfer in a vol. 1: 1-86.
Step-Taper Pile." J.S.M.F.D., ASCE, vol. 89, SM6: 57- Desai, C. S. 1974. ''Numerical Design-Analysis for Piles in
77. · Sands." Jnl. Geot. Eng. Divn., ASCE, vol. 100, no.
D'Appolonia, D. J. & Lambe, T. W. 1971. "Performance of GT6: 613-635.
Four Foundations on End-Bearing Piles." J.S.M.F.D., Desai, C., & Abe!, J. F. 1972. Introduction to the Finite
ASCE, vol. 97, SM 1: 77-93. Element Method. New York: Van Nostrand Reinhold.
D'Appolonia, D. J., Pou1os, H. G., & Ladd, C. C. 1971. Desai, C. S., Johnson, L. D., & Hargett, C. M. 1974. "Ana-
"Initial Settlement of Structures ori Clay ." J.S.M.F.D., lysis of Pile Supported Gravity Lock." Jnl. Geot. Eng.
ASCE, voL 97, SMlO: 1359-1377. Divn., ASCE, voL 100, no. GT9: 1009-1029.
Darragh, R. D. & Bell, R. A. 1969. "Load Tests on Long Desai, C. S., & Appel, G. C. 1976. "3-[) Analysis of Later-
Bearing Piles." ASTlvf STP 444: 41-67. ally Loaded Structures." 2nd Int. Conf. Num. Methods
Davis, E. H. 1961, "The Application of the Theory of Plas- in Geomechs, Blacksburg. ASCE, vol. 1: 405-418.
374 REFERENCES
Desai, C. S. & Christian, J. T. 1977. Numerica/ Methods in Felleníus, B. 1955. "Results of Tests on Piles at Gothen-
Geotechníca/ Engíneering. New York: McGraw-Hill. burg Railway Station." Bu/l. No. 5, Geot. Dept., Sw.
Dona1dson, G. W. 1965. "A Study of Leve) Observatíons on State R.
Buildíngs as Indícatíons of Moísture Movements ín the F ellenius, B. H. & Broms, B. B. 1969. "Negative S k in Fric-
Underlyíng Soíl" in Moisture Equilíbria and Moisture tion for Long Píles Dríven in Clay'." Proc. 7th In t. Conf.
Changes in Soils Beneath Covered Areas. Butterworths, S.M. & vol. 2: 93-98.
Sydney, Aust.: 156-164. Fellenius, B. H. & Broms, B. B. 1969. "Negative Skin Fric-
Donaldson, G. W. 1967a "The Use of Smaii-Diameter Piles tion for Long Piles Driven in Clay." Spec. Sess. No. 8,
in Expansive Soil." Proc. 4th Reg. Conf. for Afríca on 7th Int. Conf. S.M. & F.E., paper no. 12 (see also vol. 2
Soil Mechs. & Fndn. Eng., vol. 1: 249. · of Conf. Proc..).
Donaldson, G. W. 1967b. "The Measurement of Stresses in Flaate, K. S. 1964. "An lnvestigation of the Validity of
Anchor Piles." Pro c. 4th Reg. Con f. for A frica on Soil Three Pile Driving Formulae in.• Cohesíonless Material."
Mechs. & Fndn. Eng., vol. l: 253. Pub. No. 56, N.G.I., Oslo, Norway.
Doroshkevich, N. M. & Boím, V. P. 1967. "ln-situ Study of Focht, J. A. 1967. Discussion to paper by Coyle and Reese.
the Bearíng Capacíty of Driven Piles in Expansive Soil." J.S.M.F.D., ASCE, Vol. 93, SMI: 133-138.
Proc. 3rd Asían Conf. S.M & F.E.: 81~83. Focht, J. A. & Koch·; K. J. 1973. "Rational Analysis of the
Douglas, D. J. & Davís, E. H. 1964. "The Movement of Lateral Performance of Offshore Píle Groups." Proc.
Buried Footíngs Due to Moment and Horizontal Load 5th Offshore Tech. Conf., Houston, Vol. 2, paper OTC
and the Movement of Anchor Plates." Ceo t., vol. 14: 1896: 701-·708.
115-132. Forehand, P. W. & Reese, J. L. 1964. "Prediction of Pile
Downs, D. I. & Chieurzzí, R. 1966. "Transmíssion Tower Capacíty by the Wave Equation.'' J.S.Jf.F.D., ASCE,
Foundatíons." J. Power Divn., ASCE, vol. 92, P02: vol. 90, SM2: 1-25.
91-114. Francis, A. J. 1964. "Analysís of Pile Groups with Flexura!
Druery, B. M. & Ferguson, R. A. 1969. "An Experimental Resístance." J.S.M.F.D., ASCE, vol. 90, SM3: 1~32.
Investígatíon of the Behavíour of Laterally Loaded Francís, A. J., Stevens, L. K., & Hoadley, P. J. 1965. Paper
Piles." B. E. thesis, Dept. Civil Eng., Unív. of Sydney, to Symp. Soft Ground Eng., Brísbane. lnst. Engrs., Aust.
Aust. Franx, C. & Boonstra, G. C. 1948. "Horizontal Pressures on
Duncan, N. & Hancock, K. E. 1966. "The Concept of Con- PiJe Foundatíons." Proc. 2nd Int. Conf. S.M. & F.E.,
tact Stress in the Assessment of the Behaviour of Rock vol. 1: 131-135.
Masses as Structural Foundatíons." Pro c. 1st Conf. In t. Freeman, C. F., Klajnerman, D., & Prasad, G. D. 1972. "De-
Soc. of Rock Mechs., Lisbon, vol. 2: 487-492. sign of Deep Socketed C:üssons into Shale Rock." Can.
Dvorak, A. 1966. "Tests of Anisotropíc Shale for Founda- Geot.Jnl.,vol.9,no.1: 105-114.
tions of Large Bridges." :•roe. 1st Cong. In t. Soc. of Frydman, S., Sha'al, B., and Mazurik, A. 1975. "Analysis of
Rock Mechs.,·Lisbon, vol. 2: 537. an Instrumented Laterally Loaded Pile." Proc. 5th
Egorov, K. E., Kuzmin, P. G., & Popov, G.P. 1957. "The Asían Reg. Conf. Soíl Mechs. Fndn. Eng., Banga1ore,
Observed Settlements of Buildings as Compared with India.
Prelimínary Calculations.'' Proc. 4th lnt. Conf. S.M. & Fukuoka, M. 1977. "The Effects of Horizontal Loads on
F.E., vol. 1: 29!-296. Piles Due to Landslides." Proc. 10th Spec. Session,
Eide, 0., Bjerrum, L., & Landva, A. 1961 "Short- and 9th Int. Conf. Soi1 Mechs. Fndn. Eng., Tokyo.
. Long-Term Test Loadíng of a Friction Pile in Clay." Fuller, F. M. & Hay, H. E. 1970. "Píle Load Tests Including
Proc. 5th Int. Conf. S.M. & F.E., vol. 2:45. Quick-Load Test Method, Conventional Methods and
Ellison, R. D., D'Appo1onia, E., & Thíers, G. R. 1971. Interpretatíons." High Res. Rec., No. 333: 74-86.
"Load-Deformation Mechanísm for Bored Piles.'' J.S.M. Gau1, R. D. 1958. "Model Study of a Dynamically Laterally
F.D., ASCE, vol. 97, SM4: 661-678. Loaded PiJe." J.S.M.F.D., ASCE, Feb. 1958, paper
Elmasry, M. A. 1963. "The Negative Skín Frictíon ofBear- 1535.
ing Piles." Thesis presented to Swiss Fed. Inst. of Tech., Gibson, R. E. & Lumb, P. 1953. "Numerical Solution of
Zurich. Sorne Problems in the Consolídation of Clay ." Proc.
Endo, M., Minou, A., Kawasaki, K., & Shibata, T. 1969. lnst. Civ. Engrs., London: 182.
"Negative Skín Friction Acting on Steel Pipe Piles in Glanville, W. H., Grime, G., Fox, E. N., & Davíes, W. W.
Clay." Proc. 7th Int. Conf. S.M. & F.E., vol. 2: 93-98. 1938. "An Investígatíon of the Stresses in Reinforced
Esu, F. & Ottaviani, M. 1975. Discussion to paper,by C. S. Concrete Piles During Dríving." Br. Bldg. Res. Bd.,
Desai. Proc. Geot. Eng. Dívn., A.S.C.E., vd. 101, no. tech. paper no. 20, D.S.I.R.
GT7: 693-695. G1eser, S. M. 1953. "Lateral Load Tests on Vertical Fixed-
Evangelista, A. and Víggiani, C. 1976. "Accuraty of Numer- He..ad and Free-Head Piles," ASTi\1, STP 154: 75-93.
ical Solutions for Laterally Loaded Piles in Elastic Half- Goble, G. G., Scan1an, R. H., & Tomko, J. J. 1967.
Space." Proc. 2nd Int. Conf. Numerical Methods in "Dynamic Studies on the Bearíng Capacity of Piles."
Geomechanics, Blacksburg, vol. 3. 1367-1370. High Res. Rec., no. 67.
Evans, L. T. 1953. "Bearíng Piles Subjected to Horizontal Golder, H. Q., & Leonard, lvl. W. 1954. "Sorne Tests on
Loads." ASTlvf, STP 154. Bored Piles in London Clay." Geot., vol. 4: 32-41.
Feagin, L. B. 1937. "Lateral Pile Loading Tests." Trans. 0 Golder, H. Q. 1957. "A Not1~ on Piles in Sensítíve Clays."
ASCE, vol. 102: 236-254. Geot., vol. 7: 192-195.
REFERENCES 375
Golder, H. Q. & Skipp, B. O. 1957. "The Buckling of Piles C1ay." Proc. 2nd Int. Conf. S.M. & F.E., vol. 5: 146-
in Soft C1ay." Proc. 4th lnt. Conf. S.M. & F.E., voL 2: 154.
35-39. House1, W. S. 1966. <'Pi1e Load Capacity ~Estima tes and
Golder, H. Q. & Osler, J. C. 1968. "Settlement of a Furnace Test Results." J.S.M.F.D., ASCE, vol. 92, SM4: 1-29.
Foundation, Sore1, Quebec." Can. Geot. Jnl., vol. 5, Hrennikoff, A. 1950. "Ana1ysis of Pile Foundations wíth
no. 1: 46-56. Batter Piies." Trans. ASCE, vol. 115: 351.
Granholm. H. 1929. "On the Elastic Stability of Piles Sur- Hutchinson, J. N. & Jensen, E. V. 1968. "Loading Tests on
rounded by a Supporting Medium." lngen. Vetenk. Piies Dríven into Estuarine Clays at Port Kharramashahr,
Akad. Handl., no_ 89, Stockho!m. Iran, and Observatíons of the Effect of · Bítumen
Griffe1, W. 1966. Handbook of Formulas for Stress and Coatings on Shaft Bearing C1pacity." N.G.I.. pub. no.
Straín. New York: Ungar. 78: I-12.
Hagerty, D. J. & Peck. R. B. 1971. "Heave and Lateral Idriss, l. M. & Seed, H. B. 1968. 'Seismic Response of Hori-
Movements Due to Pile Driving." J.S.M.F.D., ASCE, zontal Soil Layers." J.S.M.F.D., ASCE, vol. 94, SM4:
voL97,SMll: 1513-1532. 1003-1031.
Hain, S. J. 1975. "Analysis of Rafts and Raft Pile Founda- lreland, H.O. 1957. "Pulling Tests on Piles in Sand." Proc.
tions," in Soil Mechs: Recent Developrnents. Proc. 4th lnt. Conf. S.M. & F.E., vol. 2: 43-54.
Symp., Univ. N.S.W., Aust.: 213-254. Isaacs, D.V. 1931. "Reinforced Concrete PiJe Formulae."
Hain, S. J. & Lee, I. K. 1978. "The Analysis of Flexible Trans. Instn. Engrs. Aust., 12: 313-323.
Pile-Raft Systems." Geot., vol. 28, no. 1: 65-83. Ito, T. & Matsui, T. 1975. "Me~.hods to Estímate Lateral
Hanna, T. H. 1963. "Model Studies of Foundations Groups Force Acting on Stabilizing Píles." Soils and Foundns,
in Sand." Geot., vol. 13, no. 4. 334. vol. 15, no. 4:43-59.
Hanna, T. H. 1967. "The Measurement of Pore Water Pres- Jain, G. S. &.Kumar, V. 1963. "Calculations for Separating
sures Adjacent to a Driven PiJe." Can. Geot. Jn/., vol. Skin Friction and Point Bearing in Pi!es." ASTM Jn.
4,no.3:313. l!rfatería/s, Research & Standards, vol. 3, no. 4: 290-
Hanna, T. H. 1968. "The .Bending of Long H-Section Pi!es." 293.
Can. Geot. Jnl., vol. 5, no. 3: 150-172. Jampel. S. 1949. "An Analysís of Groups of Piies, 1 & IL"
Hanna, T. H., & Tan, R. H. S. 197.3. "The Behavior of Long Concrete and Constructional Eng., vol. 44,7: 201-20g
Piles Under Compressive Loads in Sand." Can. Geot. and 8: 253-257.
Jnl., vol. lO, no. 3: 311-340. Jaspar, J. L. & Shtenko, V. W. 1969. "Foundation Anchor
Harrison, H. B. 1973. Cornputer Methods in Stroctural Píles in Clay-Shales." Can. Geot. Jn/., vol. 6: 159-174.
Analysis. Englewood Cliffs, N.J.: Prentíce-Hall. Johannessen, l. J. & Bjerrum, L. 1965. "Measurement of
Hayashí, S., Miyazawa, N., & Yamashita, I. 1965. "Lateral the Compression of a Steel Pile to Rock Due to Settle-
Resistance of Steel Píles Against Statical and Dynamical ment of the Surrounding Clay." Proc. 6th lnt. Conf.
Loads." Proc. 3rd World Conf. on Earthquake Eng., vol. S.M. & F .E., vol. 2: 261-264.
2. Johnson, S. M., 1962. "Determining the Capacity of Bent
Hetenyi,- M. 1946. "Beams on Elastic Foundations." Ann Piles." J.S.M.F.D., ASCE, vol 88, SM6: 65-79.
Arbor, Mich.: Univ. of Mich. Press. Kerisel, J. 1961. "Fondations Profondes en Mili e u
Heyman, L. & Boersma, L. 1961. "Bending Moments in Sableux." Proc. 5th lnt. Conf. S.M. & F.E., vol. 2: 73-
Piles Due to Lateral Earth Pressures." Proc. 5th In t. 83.
Conf. S.M. & F.E., vol. 2: 425·-429. Kerisel, J. 1965. "Vertical and Horizontal Be2ring Capacity
Heyman, L. 1965. "Measurement of the Influence of of Deep Foundations in Clay ." Symp. on Brg. Cap. and
Lateral Earth Pressure on Píle Foundations." Proc. 6th Settl. of Fndns., Duke Univ .. 45-52.
Int. Conf. S.M. & F.E., voL 2: 257-260 Kerisel, J. & Adam, M. 1967. "Calcul des F orces Hori-
Hirsch, T. J., Lowery, L. L., Coyle, H. M., & Samson, C. H. zontales Applícables aux For:dations Profondes dans les
1970. "Pi1e-Driving Analysis by One-Dimensional Wave Argiles et Limons." Annales L.T.B.T.P., No. 239: 1653.
Theory: State of the Ah." High Res. Rec., no. 333: Kerísel, J. & Quatre, J. M. 1966. "Tassements sous les
33-54. Fondations." Annales des Ponts et Chaussees, 3: 143-
H~ ·¡dley, P. J. 1964. M. Eng. Se. thesis, Univ. of Melbourne,
164.
Aust.
Hoadley, P. J., Francís, A. J., & Stevens, L. J. 1969. "Load Kezdi, A. 1957. "The Bearíng Capacity of Piles and Pile
Testing of Slender Steel Piles In Soft Clays." Proc. 7th Groups." Proc. 4th Int. Conf. S. M. & F.E., vol. 2:
In t. Conf. S.M. & F .E., Mexico, vol. 2: 123-130. 46-51.
Holloway, P. M., Clough, G. W., & Vesic, A. S. 1975. Kíshida, H. & Meyerhof, G. G. 1965. "Bearing Capacity of
"The Mechanics of Píle-Soil Interaction in Cohesion- Pile Groups Under Eccentric Loads in Sand." Proc 5th
less Soils." Duke Univ. School of Eng., Soil Mechs. Int. Conf. S.M. & F.E., vol. 2: 270-274.
series, no. 39. . Kishida, H. 1967. "Ultimate Bearing Capacity of Piles
Hooper, J. A. 1973. "Observations on the Behavíor of a Driven into Loose San d." So/l and Fndtú., vol. 7, no. 3:
Piled Raft Foundation on London Clay ." Proc. Instn. 20-29.
Civ. Engrs., pt. 2, vol. 55: 855. K1ohn, E. J. & Hughes, G. T., 1964. "Buckling of Long Un-
Housel, W. S. & Burkey, J. R. 1948. "lnvestigation to supported Timber PUes." I S.M.F.D., ASCE, vol. 90,
Determine the Driving Characteristics of PUes in Soft SM6: 107-123.
376 REFERENCES
Kocsis, P. 1968. Lateral Loads on Piles. Chicago: Bureau of ASCE, vol. 97, SMI: 295-299.
Eng. Madhav, M. R., Rao, N. S. V. K., & Madhavan, K. 1971.
Koerner, R. M. & Mukhopadhyay, C. 1972. "Behavíor of "Laterally Loaded PiJe in Elasto-Plastíc Soil." Soils
Negative Skin Friction on Model Piles in Medium- andFndns.,vol.ll,no.2: I-15.
Piasticity Silt." High. Res. Rec., no. 405': 34-44. Madhav, M. R. & Davis, E. H. 1974. "Bucklíng of Finite
Koerner, R. M. & Partos, A. 1974. "Settlement of Building Beams in Elastic Continuum." Jnl. Eng. Mechs. Dív.,
· on PiJe Foundation in Sand." Jnl. Geot., Eng. Divn., ASCE,vol.l00,no.EM6: 1227-1236.
ASCE, vol. 100, no. GT3 · 265. Madhav, M. R. & Rao, K. K. 1975. "Analysis of Initially
Koizumi, Y. & Ito, K. 1967. "Field Tests with Regard to Bent Piles" in Anal. of Soíl Beh. and lts App. to Geot.
Pi1e Driving and Bearing Capacity of Piied Founda- Structs. Proc Symp., Unív. N.S.W., Aust.: 145-155.
tions." Soils and Fndns., no. 3: 30. · Mansur, C. l. & Kaufman, R. l. 1956. "Pile Tests, Low-Sill
Kubo, J 1965. "Experimental Study of the Behaviour of Structure, 01d River La." J.S.M.F.D. ASCE, vol. 82,
Laterally Loaded Piles." Proc. 6th lnt. Conf. S.M. & SM4, proc. paper 1079.
F.E., vol. 2: 275-279. Marche, R. & Lacroix, Y. 1972. "Stabilite des Culées de
Ladanyi, B. 1963. "Expansion of a Cavity in a Saturated Ponts Establies sur des Pieux Traversant une Couche
Clay Medium." J.S.M.F.D., ASCE, vol. 89, SM4: 127- Molle." Can. Geot. Jnl., vol. 9, no. 1: 1-24.
161. Mat1ock, H. & Rípperberger, E. A. 1958. "Measurement of
Ladanyi, B. 1977. Discussion. Can. Geot. Jnl., vol. 14, Soil Pressure on a Laterally Loaded Pile." Proc ASTM,
no. 1 153-155. vol. 58: 1245-1259.
Ladd, C. C. 1 96 5. "Stress-Strain Behavior of Anisotropi- Matlock, H. & Reese, L. C. 1960. "Generalízed Solutions
cally Consolídated Clays during Undrained Shear ." for Laterally Loaded Pi1es." J.S.M.F.D. ASCE, vol 86,
Proc. 6th 1nt. Conf. S.M. & F.E., vol. 1: 282. SM5: 63-91.
Lambe, T. W. 1964. "MPthods of Estimating Settlements." Matlock, H. & Reese, L. C. 1961. "Foundation Analysis of
J.S.M.F.D., ASCE, vol. 90, SM5: 43. Offshore PiJe Supported Structures." Proc. 5th lnt.
Lambe, T. W. & Horn, H. M. 1965. "The lnfluence on an Conf. S.M. and F .E., vol 2: 91-97.
Adjacent Building of PiJe Driving for the M.I.T. Matlock, H. & lngram, W. B. 1963. "Bendingand Buckling
Materials Center." Proc. 6th Int. Conf. S.M. & F.E., of Soil-Supported Structural Elements." Proc. 2nd Pan-
voL 2: 280. Amer. Con f. S.M. & F .E., Brazil.
Lee, l. K. 1973. "Application of Finite Element Method in Matlock, H. 1970. "Correlations for Desígn of Laterally
Geot. Engg., Part 1- Lin<:ar Analysis" Ch. 17 in Finite Loaded Piles in Soft Clay ." Proc. 2nd Offshore Tech.
Elemcnt Techniques-A Short Course of Fundamentals Con f., Houston, vol. 1: 5 77-594.
and Applícation, Univ. of N.S.W., Aust. Mattes, N. S. and Poulos, H. G. 1969. "Settlement of Single
Lee, K. L. 1968. 'Buckling of Partially Embedded Piles in Compressíble Pile." J.S.M.F.D. ASCE, vol. 95, SM 1:
Sand." J.S.M.F.D., ASCE, voL 94, SMI: 255-270. 189-207.
Lenci, C., Maurice, J., & Madignier, F. 1968. "Pieu Vertical Mattes, N.S. 1969. "The lnfluence of Radial Displacement
Sollicité Horízontalement." Annales des Ponts et Compatability un Pile Settlements." Ceo t., . vol. 19.:
Chaussees, VI, Nov-Dec.: 337-383. 157-159.
Leussink, H. & Wenz, K. P. 1969. "Storage Yard Founda- Mattes, N. S. & Pou1os. H. G. 1971. "Model Tests on Pi1es
tions on Soft Cohesive Soils." Proc. 7th lnt. Conf. S.M. in Clay." Proc. 1st Aust.-N.Z. Conf. on Geomechs.,
& F.E., vol. 2: 149-155. Melbourne: 254-259.
Lo, K. Y. & Stermac, A. G. 1965. "lnduced Pore Pressures Mattes, N. S. 1972. "The Analysis of Sett1ement of Piles
During PiJe Driving Operations." Proc. 6th lnt. Conf. And PiJe Groups ín Clay Soils." Ph.D. thesis, Univ. of
S.M. & F.E., vol. 2: 285. Sydney, Aust.
Lo, K. Y. 1968. Discussion to Paper by Orrje and Broms, Matthewson, C. D. 1969. "The Elastic Behavior of a
J.S.M.FD., ASCE, voL 94, SM2: 606-608. Lateratly Loaded Pile." Ph.D. thesis, Univ. of Canter-
Lo, M B. 1967. Discussíon to Paper by Y. O. Beredugo. bury, Christchurch, N.Z.
Can. Geot. Jnl., vol. 4: 353-354. Maxwell, A. A., Fry, Z. B., & Poplin. J. K. 1969. "Víbra-
Locher. H. G. 1965. "Combíned Cast-in-P1ace and Precast tory Loading of Pile Foundations." ASTM, STP 444:
Píles for Reduction of Negatíve Fríction Caused by 338-361.
Embankment Fill." Proc. 6th Int. Conf. S.M. & F.E., McCammon, N. R. & Golder,_H. Q. 1970. "Sorne Loading
voL 2: 290-294. Tests on Long Pipe Piles." Geot., vol20: 171-184.
Lowery, L. L., Hirsch, T. J., Edwards, T. C., Coy1e, H. M., McClelland, B. & Focht, J. A. 1958. "Soil Modulus for
& Samson, C. H., 1969. "Use of the Wave Equation to Laterally Loaded Piles." Trans. ASCE, vol. 123: 1049.
Predíct Soil Resístance on a Pile During Driving." Spec. McC1elland, B., Focht, J. A., & Emrich, W. J. 1969.
Sessíon No. 8, 7th Int. Conf. S. M. & F.E., Mexico. ·'Problems in Design and lnstallation of Offshore Piles."
Lundgren, H. & Mortensen, K. 1953. "Determination by J.S.M.F:D., ASCE, voL 95, SM6: 1419-ISI4.
the Theory of Plasticity of the Bearing Capacity of McC!elland, B. 1972. "Design and Performance of Deep
Continuous Footings on Sand." Proc. 3rd Int. Conf. Foundations." Proc. Spec. ConL on Perf. of Earth and
Soil Mech. Fndn. Eng., vol. 1: 409-412. Earth-supp. Structs., ASCE, vol. 2: 111
MacDonald, H. F. 1 963. "Uplift Resistan ce of Caisson Piles McClelland, B. 1974. 'De:;ign of Deep Penetratíon Piles
in Sand." M. Se. thesis, ~ova Scotia Tech. College, Can. for Ocean Structures." -Jnl. Geot. Eng. Div., ASCE,
Madhav, M. R. & Rao, N. S. V. K. 1971. "Model for vol. l 00, no. GT7: 705-74 7.
Machine-Píle Foundation-Soil System." J.S.M.F.D., McKemie, l. M. 1969. "Foundation Load Tests on Sydney
REFERENCES 377
Sandstone." Roch Mechs. Symp., Inst. Engrs. Aust., Stability of Deep Foundations" in Soil Mechanics--
Sydney, Aust.: 132-134. Selected Tapies, ed. by l. K. Lee. Sydney, Aust.:
McNu1ty, J. F. 1956. "Thrust Loadirig on Piles." J.S.},f.F.D., Butterworths: 528~609.
ASCE, vol. 82,no. SM4, paper 1081. Morison, J. R. O'Brien, M. P., Jchnson, J. W., & Schaff, S.
Meyerhof, G. G. 1953a. "The Bearing Capacity of Founda- A. 1950. "The Force Exer ed by Surface Waves on
tions Under Eccentric and Inclined Loads." Proc. 3rd Piles." Petroleum Trans. Amer. lnst. Min. Met. & Pet
Int. Conf. S.M. & F.E., vol. 1: 440. Engrs, vol. 189: \49-J54.
Meyerhof, G. G. 1953b. "The Bearing Capacity of Concrete Moser, M. A. 1973. Discussion. Proc. 8th Int. Conf. Soil
and Rock.'' Mag. of Concrete Research, Apri1: 107- Mechs. Fndn. Eng., Moscow, 10!. 4.3: 252-253.
116. Mosley, E. T. & Raamot, T. J970. "Pile-Driving Formulas."
Meyerhof, G. G. & Murdock. L. J. 1953. "An Investigation High. Res. Rec., no. 333: 23-32.
of the Bearing Capacity of Sorne Bored and Driven Piles Murayama, S. & Shibata, T. 1%0. "The Bearing Capacity
in London Clay.'' Geot., vol. 3: 267. of a PiJe Driven ínto Soil and Its New Measuring
Meyerhof, G. G. 1956. "Penetration Tests and Bearing Methods." Soils and Fndns., vol. J, no. 2: 2-·IJ.
Capacity of Cohesion1ess Soils." J.S.1'rf.F.D., ASCE, vol. Myers, J. J., HoJm, C. H., & McAllister, R. F. 1969. Hand-
82,SM1: J-19. book o[ Ocean and Underwater Engineering. New
Meyerhof, G .G. 1959. "Compaction of Sands and Bearing York: McGraw-HiJl.
Capacity of Píles." J.S.M.F.D., ASCE, vol. 85: SM6: Nair; K. J 96 7. "Load -Settlemen 1: and Load Transfer Char-
I-29 acteristics of a Friction PJ1e Subject to a Vertical
Meyerhof, G. G. 1963. "Sorne Recent Research on the Load." Proc. 3rd Pan-Amer. Conf. S.M. & F.E., 1:
Bearing Capacity of Foundations." Can. Geot. Jnl., vol. 565-590.
l,no.l:l6. Nair, K., Gray, H., & Donovan, ':'1. J969. ''Ana!ysis of PiJe
Meyerhof, G. G. & Adarns,J. l. 1968. "The Ultimate Uplift Group Behavior,'' ASTM, STJ' 444: 118~JS9
Capacity of Foundations." Can. Geot. Jnl., vol. S, no. Nair, K. J969. "Dynamíc and Earthquake Forces on Deep
4: 225-244. ' Foundations." ASTM, STP444: 229-261.
Meyerhof, G. G. 1976. "Bearing Capacity and Settlement Nat. Swedísh Council. 1964. "Driving and Test Loading of
of PiJe Foundations." Jnl. Geot. Eng. Div., ASCE, vol. Long Concrete Piles, Tests at Gubbero, Gothenburg."
102, no. GT3: 195-228. IV A Subcomm. on Files. Rep 99.
Milligan, V., Soderman, L., & Rutka. A. 1962. "Experience Nath, J. ll. & Harlernan, D. R. f. 1967. "The Dynamics of
with Canadian Varved Clays." J.S.lvf.F.D., ASCE, vol. Fixed Towers in Deep Water Random Waves." Proc.
88, SM4: 32-67. Civil Eng. in the Oceans, ASCE Conf., San Francisco.
Mindlin, R. D. 1936. "Force at a point in the Interior of NewJand, P. L. 1968. "The Behécvior of a Pier Foundation
a Semi-Infinite Solid." Physics 7: 195. in Swelling Ciar." Proc. 4th Aust. Road Res. Conf.,
M.l. T. 1973. "Proceedings of a ~:ymposium on Downdrag Melbourne, paper no. 499S.
of Piles," ed. by J. E. Garlanger and T. W. Lambe. Nicu, ~- D., Antes, D. R., & KessJer, R. S. J97J. "Field
M.I.T. Soils Pub. no. 331. Measurements on Instrumen ted Files Under an Over-
Mitchell, J. K. 1970. "In-Place Treatnient of Foundation pass Abutment." High Res. Rec., no. 345.
Soils." J.S.M.F.D., ASCE, vol. 96, SM 1: 73-11 O. Nishida, Y. 1963. "Pore Pressure in CJay lnduced by PiJe
Mohan, D. & Chandra, S. 1961. "Frictional Resistance of Friction." Proc. 2nd Pan-Amer. Conf. S.M. & F.E.,
Bored Piles in Expansive Clays ... Geot., vol. 11: 291. Brazil, vol. 2: 2 2 S--23 3.
Mohan, D. & J ain, G. S. 1961 . "B~:aring Capacity of Piles in Nordlund, R. L. 1963. "Bearing Capacity of P1les in Co-
Expansive Clays." Proc 5th Int. Conf. S.M. & F.E., vol. hesionless Soils." J.S.M.F.D., ASCE, vol. 89, SM3:
2: 117. ]-35.
Mohan, D., Jain, G. S., & Kumar, V. 1963. "Load Bearing Novak, M. & Beredugo, Y. O. 972. "Vertical Vibratíon
Capacity of Piles.'' Geot., vol. l3, no. 1: 76-86. of Embedded Footings." J.S.M.F.D., ASCE, vol. 98,
Mohan, D., Jain, G. S., & Jain, M. P. 1967. "A New SMI2: 1291-1310.
Approach to Load Tests." Geot., vol. 17: 274-283. Novak, M. 1974. "Dynamic S1:iffness and Damping of
Mohan, D., Jain, G. S., & Sha.cma, D. J967. "Bearing Píles." Can. Geot. Jnl., vol. Jl, no. 4: 574~598.
Capacity of Multiple Under-Reamed Bored Files." Proc. Novak, M. & Grigg, R. F. 1976. "Dynarnic Experiments
3rd Asían Con f. S.M. & F .E., Haifa, vol. J : 103-J 06. with Small PiJe Foundations.'' Can. Geot. Jnl., vol. 13.
Mohan, D., Murthy, V. N. S., & Jain, G. S. J969. "Design no. 4: 372-395.
and Construction of Multi-Under Reamed Piles." Proc. Nova k, M. and Howell, J. F. 1971. TorsionaJ Vibrations
7th In t. Con f. S.M. & F .E., vol 2: l83-J86. of Pile Foundations." Jnl. Geot. Eng. Dívn., ASCE, vol.
Mohan, D. and Shrivastava, S. P. 1971. "Nonlinear Beha- J03, no. GT4: 27!-285.
viour of Single Vertical Pile Under Lateral Loads." 3rd , · Novak, M. 1977. "Vertical Vibration of Floating Piles."
Annual Offshore Tech. Conf., Houston, vol. 2, paper Jnl. Eng. Mechs. Div., ASCE, vol. 103, no. EMJ: 153-
OTC 1485: 677-684. 168.
Mohr, H. A. 1963. Discussion on Paper by S. M. Johnson. Olsen, R. E. & Flaate, K. S. 1967 "Pile-Driving Formulas
J.S.M.F.D., ASCE, vol. 89, SM4: 2J3. for Friction Piles in Sands." J,S.M.F.D., ASCE, vol. 93,
Moore, P. J. & Spencer, G. K. 1969. "Settlement of SM6: 279-296.
Building on Deep Compressib1e Soil." J. S .M .F .D., Orrje, O. & Broms, B. B. 1967. "Effects of Pile Driving on
ASCE, vol. 95, SM3: 769. SoiJ Propertíes." J.S.M.F.D., ASCE, vol. 93, SMS: 59-
Morgan, J. R. & Poulos, H. G. 1968. "Settlement and 73.
378 REFERENCES
Oteo, C. S. 1972. "Displacements of a Vertical Pi!e Group Poulos, H. G. & Mattes, N. S. 1969a. "The Behaviour of
Subjected to Lateral Loads." Proc. 5th Eur. Conf. S.M. Axially-Loaded End-Bearing Piles." Ceo t., vol. 19:
& F.E., Madrid, vol. 1:397-405. 285-300.
Palmer, L. A. and Thompson, J. B. 1948. "The Earth Poulos, H. G. & Mattes, N. S. 1969b. "The Analysis of
Pressure and Deflection A1ong the Embedded Lengths Downdrag in Eng-Bearing Piles Due to Negative Fric-
of Piles Subjected to Lateral Thrusts:·• Proc. 2nd Int. tion." Proc. 7th Int. Conf. S.M. & F.E., vol. 2: 204-
Conf. S.M. and F.E., Rotterdam, voL 5: 156-161. 209.
Palmer, D. J. & Holland, G. R. 1966. "The Construction of Poulos, H. G. 197la. "Behaviour of Lateral! y Loaded Piles:
Large Diameter Bored Piles with Particular Reference to I-Single Piles." J.S.M.F.D., ASCE, voL 97, no. SMS:
London Clay ." Symp. on Large Bored Piles, Inst: Civ. 711-731.
Engrs: 105-120. Poulos, H. G. 197lb. "Behaviour of Laterally-Loaded Piles:
Pandey, V. J. 1967. "So me Experiences with Bored Piling." II-Pile Groups." J.S.M ..F.D., ASCE, vol. 97, no. SMS:
J.S.M.F.D., ASTE, vol. 93, SMS: 75-87. 733-751. -
Parola, J. F. 1970. "Mechanícs of Impact Pile Drivíng." Po u los, H. G. 197lc. Discussion to "Load-Deformation
Ph.D. dissertation, Univ. of Illinois, Urbana, 111. Mechanísm for Bored Piles," by R. D. Ellison, E.
Parr, R. G. & Varner, M. J. 1962. "Strength Tests on Ovet- D'Appolonia, & G. R. Thiers.Jn/. Soil Mechs. & Fndns.
head Line Tower Foundations." Elec. Res. Assoc. Rep. Div.,ASCE,vol97,no.SMI2: 1716.
O/T28. Poulos, H. G. & Mattes, N. S. 197la. "Displacements in a
Parsons, J. D. & Wilson, S. D. 1954. "Safe Loads on Dog- Soil Mass Due to Pile Groups." Aust. Geomechs. Jn/,
Leg Piles." Proc. Sep. No. 475, ASCE, vol. 80. vol.G1,no. 1:29-35.
Peck, R. B. 1958. "A Study of the Comparative Behavior Poulos, H. G. & Mattes, N. S. l97lb. "Settlement and Load
of Friction Piies." High. Res. Bd. Spec. Rep. 36. Distribution Analysis of Pile Groups." A ust. Geomechs.
Peck, R. B. & Davisson, M. T. 1962. Discussion. Trans. Jnl.,vol.G1,no.1: 18-28.
ASCE, vol. 127, pt. 4: 413. Poulos, H. G. & Madhav, M. R. 1971. "Analysis of the
Peck, R. B., Hansen, W. & Thorburn, T. H. 1974. Movement of Battered Piles." Proc. 1st Aust.-N .z.
Foundation Engineering, 2nd e d. New York: Wiley. Conf. on Geomechs., Melbourne: 268-275.
Peck, G. M. 1969 "The Rational Design of Large Diameter Poulos, H. G. & Davis, E . H. 1972. "The Development of
Bored Piles Founded on Rock." Rock Mechs. Symp., N egative Friction wíth Time in End-Bearing Piles."
lnst. Engrs. Aust., Sydney: 48--51. Aust. Geomechs. Jnl., vol. G2, no. 1: 11·-20.
Pells, P. J. N. 1977. "Theoretical and Model Studies Poulos, H. G. 1972a. "Behaviour of Laterally Loaded Piles:
Re1ated to the Bearing Capacity of Rock." Paper III-Socketed Piles." J.S.M.F.D., ASCE, vol. 98; no.
presented to Sydney Group of Aust. Geomechs. Soc., SM4: 341-360.
Inst. Engrs. Aust. Poulos, H. G. 1972b. "Seúlement Analysis of Two Build-
Penzíen, J., Schaffey, C. F., & Parmelee, R. A. 1964. ings on End-Bearing Piles." Proc. 3rd SE Asían Conf.
"Seismic Analysis of Bridges on Long Piles." J: Eng. Soil. Eng.,Hong Kong: 129-134.
Mechs. Div. ASCE, EM3: 223-254. Pou1os, H. G. 1972c. "Diffkulties in Prediction of Horizon-
Penzien, J. 1970. "Soil-Pile Foundation Interaction" in tal Deformations of Foundations." J.S.M.F.D., ASCE,
Earthquake Engineering, ed. by R. L. Wiegel, Engle- vol. 98, SM8: 843-848.
wood Cliffs, N. J.: Prentice-Hall. Poulos, H. G. 1972d. "Loacl-Settlement Prediction for Piles
Philcox, K. T. 1962. "Sorne Recent Developments in the and Piers." J.S.M.F.D., ASCE, vol. 98, SM9: 879-897.
Desígn of Higl1 Buildings in Hong Kong." Struct. Eng., Poulos, H. G. 1973a. "Load-Deflection Prediction for
vol. 40, Oct.: 303-323. Laterally Loaded Piles." A ust. Geomechs. J ni., vol. G3,
Pichumani, R. & D' Appolonia, E. 196 7. "Theoretical Distri- no. 1: 1-8.
butíon of Loads Among the Pi! es in a Group." Proc. 3rd Poulos, H. G. 1973b. "Analysís of Piles in Soil Undergoíng
Pan-Amer. Conf. S.M. & F.E., Caracas, Venezuela. Lateral Movement." J.S.M.F.D., ASCE, vol. 99, SMS:
Plantema, I. G. 1957. "Influence of Density of Sounding 391-406.
Results in Dry, Moist and Saturated Sands." Proc. 4th Poulos, H. G. & Davis, E. H. 1973. "Theory of Piles in
In t. Con f. S.M. & F .E., vol. 1: 23 7-240. Swelling and Shrinking Soils." Pro c. 8th In t. Con f.
Poorooshasb, H. B. & Bozozuk, M. 196 7. "S k in Friction on S.M. & F.E., Moscow, v-ol. 2.2: 169-176.
a Single Píle to Bedrock." Pro c. 3rd Pan-Amer. Conf. Poulos, H. G. 1974. "Analysis of PiJe Groups Subjected to
S.M. & F.E., vol. 1:613-621. Vertical and Horizontal Loads." A ust. Geomechs. Jnl.,
Poulos, H. G. & Davis, E. H. 1968. "The Settlement Beha- vol. G4, no. l: 26-32.
vious of Single Axially-Loaded Incompressible Piles and Poulos, H. G. & Davís, E. H. 1974. Elastic Solutions for
Piers.'' Geot. voL 18:351-371. Soil and Rock Mechanics. New York: Wiley.
Poulos, H. G. 1968. "The Influence of a Rigid Pile Cap on Poulos, H. G. & Mattes, N. S. 1974. ''Settlement of Pile
the Settlement Behaviour of an Axially-Loaded Pile." Groups Bearing on Stiffer Strata." Jnl. Ceo t. Eng. Div.,
Civ. Eng. Trans., lnst. Éngrs. Aust., voL CElO, no. 2: ASCE, vol. 100, no. GT2: 185--190.
206-208. Poulos, H. G. 1975a. "Lat•~ral Load-Deflection Prediction
Poulos, H. G. 1968. "Analysis of the Settlement of Pile for Pi! e Groups." Jnl. Ceo t. Eng. Dív., ASCE, vol. 1O1,
Groups." Geot. vol. 18: 449-471. no. GT 1: 19-34.
Poulos, H. G. 1969. "The Settlement of Under-Reamed and Poulos, H. G. l975b. "Torsional Response of Piles." Jnl.
Step-Taper Piles." Civ. Eng. Trans., Inst. Engrs. Aust., Geot. Eng. Div., ASCE, vol. 101. no. GT!O: 1019-
vol. CE 11: 83--87. 1035.
REFERENCES 379
Poulos, H. G. & Davis, E. H. 1975. "Prediction of Down- Tech. Con f., Houston, paper OTC 2080: 4 73-483.
Forces in End-Bearing Piles." Jnl. Geot. Eng. Reese, L. Cox, W. R., & Koop, F. D. 1975. "Field
Div., ASCE, vol. 1O1, no. GT2: 189-204. Testing and Analysis of L<terally Loaded Piles in Stiff
Poulos, H. G. & Mattes, N. S. 1975. "A Theor.etical Exami- Clay." Proc. 7tl¡ Offshore Tech. Conf., Houston, paper
nation of Errors in Measured Settlements of Test Piles." OTC 2312: 671~690.
Proc. 2nd Aust.-N.Z. Conf. en Geomechs: 174-178. Reese, L. C. & Welch, R. C. 1975. "Lateral Loadings of
Poulos, H. G. 1976. "Behaviour of Late rally Loaded Piles Deep Foundations in Stiff Clay." Jnl. Geot. Eng. Div.,
Near a Cut or Slope." Aus!. Geomechs. Jnl., vol. G6, ASCE, vol. 1Ol, no. GT7: !í33-649.
no. 1: 6-12. Reese, L. C. 1975. "Laterally Loaded Piles." Proc. of the
Poulos, H. G., & Adler, M. A. 1978. "Analysis of Lateral Seminar Series, Design, Construction and Performance
Response of Non-Uniform Section Piles." Civ. Eng. of Deep Foundations; G·:ot. Group and Continuing
Res. Re p. R330, Univ. of Syjney, Australia. Ed~cation Committee, San Francisco section, ASCE;
Poulos, H. G. 1979. "Settlem<:nt of Single Piles in Non- Umv. of Calif., Berkeley.
homogeneous Soil." Jnl. Geot. Eng. Div., ASCE, Vol. Reese, L. C., Touma, F. T. & D'Neill, M. W. 1976. "Behav-
105, no. GT5: 627-641. ior of Drilled Piers Unde1 Axial Loading." Jn/. Geot.
Prakash, S. 1962. "Behaviour of Pile Groups Subjected to Eng. Div., ASCE, vol. 102, GTS: 493-S 1O.
Lateral Loads." Ph.D. thesis, Univ. of Illinois, .Urbana. Reese, L. C. 1977. "Laterally Loaded Piles: Program Docu-
Prakash, S. & Saran, D. 1967. "Behaviour of Laterally- mentatíon." Jnl. Geot. Eng. Div., ASCE, vol. 103, no.
Loaded Piles in Cohesive Scils." Proc. 3rd Asían Conf. GT4: 287-305.
S.M.: 235-238. Rehnman, S. E. & Broms, B. B. 1970. ''Bearing Capacity of
Prakash, S. & Chandrasekaran, V. 1973. "Pile Foundation End-Bearing Piles Driven to Rock." Proc. 2nd Cong. of
under Lateral Dynamic Loads." Proc. 8th Int. Conf. In t. Soc. of Rock Mechs, Beograd, vol 2: 1S-22.
S.M. & F.E., Moscow, vol. 2.1, paper 3(31: 199-202. Resendiz, D., Auvinet, G., and Silva, C. 1969. "Conceptíon
Prandtl, L. 1923. "Anwendungsbeispiele zu Einem et Comportement des Fondations des Palais des Sports
Henchyschen Satz Ueber das Plastische Gleichgewicht." de la Ville de Mexico en Presence de Frottenment Nega-
Z. Angew. ll·fath.l'dech., vol. 3: 468. tif." Spec. Sess. No. 8, '?tli Int. Conf. S.M. & F.E.,
Press, H. 1933. "Die Tragfah.tgkeit von Pfallgnippen in paper 18.
Beziehund zu der des Einze1pfahles." Bautechnik, voL Richards, B. G. 1965. "An An 1lysis of Subgrade Conditions
11: 625--627. at the Horsham Experimer. tal Road Si te Using the Two-
Priddle, R. A. 1963. "Load Distribution in Piled Bents." Dimensíonal Dtffusion Eq·..1ation on a High-Speed Digi-
Trans., Inst. Engrs. Aust., vol. CE5, no. 2: 43-54. tal Computer," in Moisture Equilibria and Moisture
Radugin, A. E. 1969. "lncrease of Bearing Capacity of Changes in Soils Beneath Covered Arcas. Sydney:
Short Piles with Time." Soil Mechs. & Fndn. Eng., Butterworths, 243.
March-April: 103-107. Richart, F. E. 1962. "Foundation Vibrations." Trans.
Randolph, M. F. 1977. "A Theoretical Study of the ASCE, vol. 127, pt. 1: 863-898.
Performance of Piles." Ph.D. Thesis, Cambridge Univ., Richart, F. E., Hall, J. R., & Woods, R. D. 1970. Vibrations
U.K. of Soils and Foundatiom. Englewood Cliffs, N. J.:
Randolph, M. F. & Wroth, C. P. 1978. "Analysis of Defor- Pren tice-Ha1L
mation of Vertically Loaded Piles." Jnl. Geot. Eng. Div, Robinsky, E. l. & Morrison, C. E. 1964. "Sand Displace-
ASCE, vol. 104, no. GT12: 1465-1488. ment and Compaction A1 ound Model Friction Piles."
Reddy, A. S. & Valsangkar, A. J. 1968. "An Analytical Can. Geot. Jnl., vol. 1, no. 2: 81.
Solution for Lateral! y Loaded Pi! es in Laye red Soils." Roscoe, K. H. 1957. "Comparison of Tied and Free Pier
Sols-Soils, no. 21: 23-28. Foundations." Proc. 4th Int. Conf. S.M. & F.E., vol.
Reddy, A. S. & Valsangkar, A. J. 1970. "Buckling of Fully 2: 419.
and Partially Embedded Pil.es." J.S.M.F.D., ASCE, vol. Ross, H. E. 1971. ''Dynamic Response of Offshore Piling."
96, SM6: 1951-1965. 3rd Offshore Tech. Conf., Texas, papér OTC 1480.
Reese, L. C. & Cox, W. R. 1969. "Soil Behaviour from Ana- Rowe, P. W. 1956. "The Sing)e Pile Subject to Horizontal
lysis of Tests on Uninstrumented Piles Under Lateral Force." Geot., vol. 6: 70-:l5.
Loading." ASTM, STP 444: 160~176. Rowe, R. K. & Poulos, H. G. 1979. "A Method for Predict- ·
Reese, L. C .. & Matlock, H. 1956. "Non~Dimensional Solu- ing the Effect of Piles on Slope Behaviour." Proc. 3rd
tions for Laterally Loaded Piles with Soil Modulus As- Int. Conf. Num. Methods in Geomechs, Aachen.
sumed Proportional To De?th." Proc. 8th Texas Conf. Balkema. vol. 3: 1073-!0HS.
S.M. and F.E., Spec. Pub. 29, Bureau of Eng. Res., Rowe, R. K., Booker, J. R. & Balaam, N. P. 1978. "Appli-
Univ. of Texas, Austin. cation of the Initial Stn·ss Method to Soil-Stmcture
Reese, L. C., Hudson, B: S. & Vijayvergija, B. S. 1969. "An Interaction." In t. Jnl. for Num. Meth. in Eng., vol. 12:
Investigation of the Interaction Between Bored Piles 873-880.
and Soil." Proc. 7th Int. Conf. S.M. & F.E., vol. 2: Rutledge, P. C. 1950. Discussion on Paper by Cummings,
211~215. Kerkhoff, and Peck. Trans. ASCE, vol. 115.
Reese, L. C., O'Neill, M. W., and Smith, R. E. 1970. "Gene- Saffery, M. R. & Tate, A. P. K. 1961. "Model Tests on Pile
ralízed Ana1ysis of Pile Foundations." J.S.M.F.D., Groups in a Clay Soil with Particular Reference to the
ASCE, vol. 96, SM 1: 235. Behaviour of the Group When It Is Loaded Eccentri-
Reese, L. C., Cox, W. R., & Koop, F. D. 1974. "Analysis of cally." Proc. 5th Int. Conf. S.M. & F.E., vol. 2: 129-
Lateral1y Loaded Piles in San d." Proc. 6th Offshore 134.
380 REFERENCES
Sahzin, V. S. 1968. "Method of Determiníng the Magni- Equation." J.S.Af.F.D., ASCE, vol. 86, SM4: 35-61.
tude of Rise of Piles Following Flooding of Swelling Soderberg, L. 196 2a. "Consolidation Theory Applied to
Soils." Soíl Mechs. & Fndn. Eng., No. 1: 149. Foundation Pile Time Effects." Geot., vol. 12: 217.
Salas, J. A.. J. & Beizunce, J. A. 1965. "Resolution Soderberg, L. O. 1962b. Discussion to Paper by E. A. L.
Theoríque de la Distribution des Forces dans les Smith. Trans. ASCE: 1171-1174,.
Pieux." Proc. 6th Int. Conf. S.M. & F.E., vol. 2: 309- Sokolovskíi, V. V. 1965. Sta ti es of Soil Medie¡. London:
3 13. Butterworths.
Salas, J. A. J. 1965. Discussion on Div. 4, Proc. 6th Int. Sorensen, T. & Hansen, B. 1957. "Pile Dríving Formulae,
Conf. S.M. & F.E., vol. 3: 489-492. An Investigation Based on Dimensional Considerations
Samson, C. H., Hirsch, T. J.,& Lowery, L. L. 1963. "Com- anda Statistical Analysis." Proc. 4th In t. Conf. S.M. &
puter Study of Dynamic Behaviour of Piling." J. Struú F.E., vol. 2:61-65.
Divn, ASCE, vol. 89, ST4: 413-449. Sowa, V. A. 1970. "Pulling Ca.pacity of Concrete Cast in
Saul, W. E. 1968. "Sta tic and Dyr:amic Analysis of Pile Situ Bored Piles." Can. Geot. Jnl, vol. 7: 482-493.
Foundations." J. Struct. Div., ASCE, vol. 94, ST5: Sowers, G. F., MarÚn, C. B., \1/ilson, L. L., & Fauso1d, M.
1077-1100. 1961. "The Bearing Capacity of Friction Pile Groups
Sawko, F. 1968. "A Símplifíed Approach to the Analysis of in Homogeneous C1ay from Mode1 S tudies." Proc. 5th
Piling Systems." Struct. Eng., vol. 46, no. 3: 83-86. Int. Conf. S.M. & F.E., vol. 2: 155-159.
Scanlan, R. H. & Tomko, J. J. 1969. "Dynamic Prediction Sowers, G. B. & Sowers, G. F. 1970. Jntroductory Soil
of Pí!e Static Bearing Capacity'' J.S.M.F.D., ASCE, vol. Mechanics and Foundations, 3rd ed. New York:
95, SM2: 583-604. Macmillan.
Schmertmann, J. H. 1969. Discussion. Proc. 7th In t. Con f. Spence, B. E. 1965. "Uplift Resistance of Piles with En-
S.M. & F.E., vol. 3: 250-25 L larged Bases in Clay ." M.Sc. thesis, Nova Scotia Tech.
Seed, H. B. & Reese, L. C. 1957. "The Action of Soft Clay College, Can.
Along Frictíon Piles." Trans. ASCE, vol. 122: 731-754. Spillers, W. R. & Stoll, R. D. il964. "Lateral Response of
Seed, H. B. & Martín G. R. 1966. "The Seismic Coefficient Piles." J.S.M.F.D., ASCE, vol. 90, SM6: 1-9.
in Earth Dam Design." J.S.M.F.D., ASCE, vol. 92, SM3: Stein brenner, W. 1934. "Tafe1n zur Setzungberechnung."
25-58. Die Strasse, 1: 221.
Sharman, F. A. 1961. "The Anticipated and Observed P~ne Stoll, U. W. 1972. "Torque Shear Test on Cy1indrica1 Fric-
tration Resistance of Sorne Friction Piles Entirely in tion Piles." Cf¡;_ Eng., ASCE, vol. 42, no. 4: 63-64.
C1ay." Proc. 5th Int. Conf. S.M. & F.E., voL 2: 135- Sullivan, W. R., Reese, L. C., & Fenske, C. W. 1979. "Uni-
141. fied Method for Analysis of Laterally Loaded Piles in
Sherman, W. C. 1969. "Instrumented Pile Tests in a Stiff C1ay ." Conf. on Num. Mcthods in Offshore Piling,
C1ay." Proc. 7th Int. Conf. S.M. & F.E., vol. 2: 227- London. Inst. Civ. Engrs., Paper no. 17.
232. Tavenas, F. A. 1971. "Load Test Results on Friction Piles
Shubinski, R. P., Wilson, E. L., & Se1na, L. G. 1967. "Dy- in Sand." Can. Geot. Jnl., vol. 8: 7-22.
namic Response of Deepwater Structures." Proc. Civil Tay1or, D. W. 1948. Fundamentals of Soil Mechanícs. New
Engr. in the Oceans, ASCE Conf., San Francisco. York: Wiley.
Shultze, E. & Mezler, K. J. 1965. "The Determínation of Terzaghí, K. 1942. "Discussion of the Progress Report of
the Density and the Modulus of Compressibility of the Committee on the Bearing Value of Pile Founda-
Non-Cohesive Soils by Soundings." Proc. 6th Int. Conf. tíons." Proc. ASCE, vol. 68: 311-3 23.
. S.M. & F.E., vol. 1: 354-358. Terzaghí, K. 1943. Theoretical Soil Mechanics. New York:
Simek, J. 1966. "Resultats d'Observations de 1'Influence Wiley.
d'une Force Horizonta1e sur des Groupes de Pieux." Terzaghí, K. 1955. "Evaluation of Coefficients of Sub-grade
Sols·Soils, no. 18-19: 11-18. Reaction." Geot., vol. 5: 297.
Singh, A. & Verma, R. K. 1973. "Lateral Resistance of a Terzaghi, K. & Peck, R. B. 1967. Soíl Mechanics in Engi·
Field Mode1 of Pile Group in Sand and Its Comparison neering Practíce. New York: Wiley.
with a Laboratory Model." Jnd. Geot. Jnl., vol. 3, no. Thorrias, D. 1965. "Static Penetratíon Tests in London
2:113-127. Clay"Geot.,vol.lS: 174.
Skempton, A. W. & Bishop, A. W. 1950. "The Measurement Thorne, C. P. & Burman, B. 1968. "The Use of the Static
of the Shear Strength of So'ils." Geot., vol. 1, no. 2: 90. (Dutch) Cone Penetrometer for the ln-Situ Testing of
Skempton, A. W. 1950. Discussion to Paper by G. Wilson, Soils." Proc. Symp. on Field Measurements, paper no.
J. Jnst. Civ. Engs., London, vol. 34, no. S: 76-81. 498S, Aust. Road Res. Board, Me1boume.
Skempton, A. W. 1951. "The Bearing Capacity of Clays."
Thorne, C. P. 1977 "The Allowable Loadíngs of Founda-
Build. Res. Congress, London. Inst. Cív. Engrs., div. 1:
tions on Sha1e and Sandstone in the Sydney Region.
180.
Part 3. Field Test Results." Paper presented to Sydney
Skempton, A. W. 1953. Discussion: Pí1es and Píle Founda-
Group of Aust. Geomechs. Soc., Inst. Engrs. Aust.
tions, Settlement of Pile Foundations. Proc. 3rd lnt.
Conf.S.M.&F.E.,vol.3: 172. Thurman, A. G: & D'Appolonia, E. 1965. "Computed
Skempton, A. W. 1959. "Cast-In-Sítu Bored Píles in Movement of Friction and End-Bearing Piles Embedded
London Clay." Geot., vol. 9: 158. in Unifor:m and Stratified Soils." Proc. 6th lnt. Conf.
Skempton, A. W. 1966. "Summíng Up." Symp. on Large S.M. & F.E., vol. 2: 323-327.
Bored Piles, London: !55. · Timoshenko, P. S. 1936. Theory of Elastíc Stability ·' New
Smith, E. A. L. 1960. "Pile Driving Analysis by the Wave York: McGraw-Hül.
REFERENCES 381
Toak1ey, A. R. 1964. M.Eng.Sc. thesis, Univ. of Melbourne, Vesíc, A. S. 1967. "A Study of Bearing Capacity of Deep
Aust. Foundations." Final Rep., Proj. B-189, School of Civil
Toak1ey, A. R. 1965. "Buckling Lo.ads for Elastically Sup- Eng., Georgia lnst. Tech., Atlanta, Ga.
ported Struts." J. Eng. Mechs. Div. ASCE, vol. 91, Vesic, A. S. 1969. "Experiments with Instrurnented Pile
EM3 205-231. Groups in Sand." ASTM, STP 444: 177-222.
Tornlinson, M. J. 1957. "The Adhesíon of Piles Dríven in Vesic, A. S. 1969b. Discussion. Proc. 7th In t. Con f. S.M.
C1ay Soíls." Proc. 4th Int. Conf. S.M. & F.E. voL 2: & F.E., voL 3: 242-244.
66-71. Vesic, A. S. 1972. "Expansion of Cavities in Infinite Soil
Tornlinson, M. J. 1970. "Sorne Effects of Pile Drívíng on Mass." J.S.M.F.D., ASCE, vol. 98: 265-290.
Skin Friction." Conf. on Beh. of Piles, Inst. Civ. Engrs., Vesic, A. S. 1975. "Principies of Pile Foundation Design."
London: 59-66. . Duke Univ. School of Eng., Soil Mechs., series no. 38:
Tornlinson, M. J. 1975 Foundation Design and Construc- Vijayvergiya, V. N. 1969. "Load Distribution Along a
tíon. 3rd Ed. London: Pitrnan. Bored Pile Deterrnined by Tell-Tales." Paper presented
Tornlinson, M. J. 1977. Pi/e Design and Construction Prac- to Soil Mechs. and Fndn. Grp., Texas Section, ASCE,
tice. London: Viewpoint Publications. Lubbock, Texas.
Tourna, F. T. & Reese, L. C. 1974. "Behaviour of Bored Vijayvergiya, V. N. & Focht, J. A. Jr. 1972. "A New Way
Piles in Sand." Jnl. Geot. Eng. Div., ASCE, vol. 100, no. to Predict the Capacity of Piles in C1ay." 4th Annual
GT7. 749-761. Offshore Tech. Conf., Houston, vol. 2: 865-874.
Trofirnenkov, J. G. & Maríupolskii, L. G. 1965. "Screw Walther, R. E. 1962. "Prestre::sed Rock Anchors." Civil
Piles Used for Mast and Tower Foundations." Proc. 6th Eng., ASCE, vol. 13, no. l.
Int. Conf. S.M. & F.E., vol. 2: 328-332. Wagner, A. A. 1953. "Lateral :~oad Tests on Piles for De-
Tschebotarioff, G. P. 1951. Soíl Mechanics Foundations sign Inforrnation." Symp. on Lateral Load Tests on
and Earth Structures. New York: McGraw-Hill. Piles. ASTM, STP 154: 59-7 2.
Tschebotarioff, G. P, 1953. "The Resistance to Lateral Walker, L. K. & Darvall, P. Le !P. 1970: "Sorne Aspects of
Forces of Single Pil<:s and Pile Groups." ASTM, STP Dragdown on Piles." Proc. :!nd SE Asían Conf. on Soil
154: 38. Eng.,Singapore: 121-137.
Tucker, R. L. 1964. "Lateral Analysis of Piles with Dyna- Walker, L. K. & Darvall, P. le P. 1973. "Dragdown on
rnic Behaviour." Proc. Conf. on Deep Fndns., Mexico Coated and Uncoated Piles." Proc. 8th In t. Con f. S.M.
City, vol. 1: 15 7-171. & F.E., Moscow, vol. 2.1: 2~ 7-262.
Turner, E. A. 1962. "Uplift Resistance of Transrnission Warburton, G. 1964. The Dynamic Behaviour of Structures.
Tower Footings." J. Power Div., ASCE, vol. 88, papet Oxford: Pergamon.
3187. Whitaker, T. 1957. "Experintents with Model Piles in
Va11iappan, S., Lee, l. K., & Boon1ualohr, P. 1974. "Settle- Groups." Geot., voL 7: 147-167.
rnent of Piles in Layered Soils." Proc. 7th Biennial
Conf., Aust. Rd. Res. Bd., Adelaide, voL 7, pt. 7: 144- Whitaker, T. 1960. "Sorne Experiments IJn Model Piled
153. Foundations in C1ay." Proc. Syrnp. on Pile Fndns., 6th
Van der Merwe, D. H. 1964. "The Prediction of Heave frorn Conf. Int. Assoc. Bridge ar.d Struct. Eng., Stockholm:
124-139.
the P1asticity Index and the Percentage Clay Fraction."
The Civil Engr. in So. A frica, vol. 6, no. 6: 103. Whitaker, T. & Cooke, R. W. 1961. "A New Approach to
Van der Veen, C. & Boersrna, L. 1957. "TheBearingCapa- Pile Testing." Proc. 5th Inl. Conf. S.M. & F.E. vol. 2:
city of a Pile Predeterrnined by a Cone Penetration 171-176.
Test." Proc. 4th Int. Conf. S.M. & F.E., vol. 2: 72-75. Whitaker, T. 1963. "The Constant Rate ofPenetratíon Test
Van der Veen, C. 1965. "Loading Test on an Unorthodox for the Determination of the Ultirnate Bearing Capacity
Concrete Cuff Pile." Proc. 6th Int. Conf. S.M. & F.E., of aPile." Proc. Instn. Cív. Engrs, vol. 26: 119-123.
vol. 2: 333-337. Whitaker, T. & Cooke, R. W. 1966. "An Investigation of
Van Weele, A. F. 1957. "A Method ofSeparating the Bear- the Shaft and Base Resistances of Large Bored Piles in
ing Capacity of a Test Pile into Skín Friction and Point London Clay ." Proc. Symp. on Large Bored Piles: 7-
Resistance." Proc. 4th Int. Conf. S.M. & F.E., vol. 2: 49.
76. Whitaker, T. 1970. The Design Piled Foundations.
Verruijt, A. 1969. "A Simp1ifio~d E1astic Method for the Oxford: Pergamon.
Ca1cu1ation of Negative Skin Friction of Piles." Spec. White, R. 1943. "Heavy Foundations Drilled ínto Rock."
Sess. No. 8, 7th Int. Conf. S.M. & F.E., paper 5. ·Civ. Eng., ASCE, vol. 13, no. l.
Vesic, A. S. 1961. "Bending of :Bearn Resting on lsotropic Whítman, R. V. & Richart, F. E. 1967. "Design Procedure
E1astic So lid." Jnl. Eng. Mechs. Div., ASCE, vol. 87, for Dynamically Loaded Foundations." J.S.M.F.D.,
EM2: 35-53. ASCE, voL 93, SM6: 169-193.
Vesic, A. S. 1964. "lnvestigations of Bearing Capacity of Wiesner, T. J. & Brown, P. T. 1975. "Behaviour of Piled
Piles in Sand." Proc. No. Amer. Conf. on Deep Fndns., Strip Footings Subjected te- Concentrated Loads." Civ.
Mexico City, vol. J: 197-224. Eng. Res. Rep. R275., Univ of Sydney, Aust.
Vesic, A. S. 1965. "Ultirnate Loads and Settlernents of Wilson, G. 1948. "A Re1axatiort Method for the Solution of
Deep Foundations in Sand." Proc. Syrnp. on Bearíng Problems Concerning Axially Symrnetrical Distributions
Cap·acity and Sett1ement of Foundations, Duke Univ. of Load in an Elastic Medium." Jnl. Inst. Cív. Engrs.,
53-68. London, No. 6, April: 149-166. ·
382 REFERENCES
Wilson, G. 19 50. "The Bearing Capacity of Screw Pi! es and tance Displacement Relationships for Pile Foundations
Screwcrete Cylinders." J. lnst. Civ. Engrs., London, vol. in Soft Clays". 5th Annual Offshore Tech. Conf.
34: 4-93. Houston, Paper OTCI893 . Vol. 2, 663-676.
Wilson, S. D. & Hilts, D. E. 1967. "How to Determine La- Yoshimi, Y. ( 1964) "Piles in Cohesionless Soils Subjerted
teral Load Capacity of Piles." in Pile-Fou"ndation Know- to Oblique Pull". J.S.M.F.D., ASCE, Vol. 90, SM6, 11.
How (41-45). Washington: Amer. Wood Pres. Inst. Zeevaert, L. ( 1957)- "Foundation Design and Behaviour of
Wroth, C. P., Carter, J. P. & Randolph, M. F. 1979. "Stress Tower Latino Americana in Mexico City". Geotech-
Changes Around a Pile Dríven into Cohesive Soil." nique, Vol. 7, No. 1, liS.
Conf. on Recen! Devel. in the Design and Constrn. of
Piles. Inst. Civ. Engrs., London. Zeevaert, L., {1959) "Reduction of Point Bearing Capa-
Woodward, R. & Boitano, J. (1961) "Pi! e Loading Tests in city Because of Negative Fríctíon". Proc. 1st Pan-Amer.
Stiff Clays". Proc. 5th Int. Conf. S.M. & F.E., Vol. 2, Conf. S.M. & F.E., Vol. 3, 1145.
177. Zeevaert, L. ( 1973) - "Foundatioii Engineering for Diffi-
Yamagata, K. 1963 "The Yield-Beanng-Capacity of cult Su bsoil Conditions." Van Nostrand Reinhold,
Bearing Piles". Proc. Int. Conf. S.M. & F.E. Budapest, New York.
325-342. Zienkiewicz, O. C. (1971) - "The Finite Element Method in
Yegian, M. and Wright, S. G. (1973) "Lateral Soil Resis- Engineering Science." McGraw-Hill, London.
AUTHORINDEX
383
. 384 AUTHOR INDEX
Moment loading on piles, 167, 168, 184, 188, Penetrometer, see Static cone penetrometer
207 Pier:
Morison theory, 338 equivalent for group, 120
Movement ratio, 90 settlement beneath center, 96
Multimass vibrators, 338 simplified load-settlement analysis, 99
Pile cap, see Cap
Piled groups, se e Pi!e-raft foundation
Natural frequency of piles, 339-353 Pile flexibility factor, 179, 193
Negative friction: Pile-raft foundatíon:
afterdriving, 89, 265, 269 consolidation settlement, 260
battered piles, 289 elastic analysis, 250-258
case histories, 266-269 examp1e, 261-262
comparisons with theory, 289-293 finite element analysis, 263
elastic solutions, 276, 280, 281, 298 plate ana1ysis, 262
in end-bearing piles, 265-293 sirnplified analysis, 258-261
in floating piles, 294-310 ultimate load capacity, 33-35, 260
group effects, 288 Piles, classification of, 6
methods of reduction, 269-271 Piles to rock criteria, see Críteria
parametric solutions, 274-282, 298-305 Pile stiffness factor, 77
rate of development, 278-279 ty pica! va1ues, 104
settlement to mobilize, 2 78 P1ane-strain solution for lateralloading, 152-153
Neutral point, 296 P1asticity index, effect on negative friction,
Newtonian impact, 53 286
Non-homogeneous soil: P1asticity so1utions for bearing capacity, 19
effect on settlement ratio, 125 P1ate analysis for pile-raft, 262
interaction factors, 113 Plate loading tests, 172-173, 223
load test errors, 363 Point load capacity, see Base load capacity
pile buck!ing loads, 327 Poisson's ratio of soil:
pile in swelling soíi, 297 effect on interaction factors, 113
settlement ana1ysis, 77 effect on settlement, 89
solutions for lateral deflection, 170-1 73, 192-199, 215, effect on settlement ratio, 1 25
313,318 effect on tip load, 86
solutions for settlement, 92 typical values, 103
Non!inear analysis for lateral deflections, 175-177, 181, Pore pressures:
182 dueto driving, 7-9, 269, 274
Non-uniform piles, 81, 93, 180 due to surcharging, 284
effect of electro-osmosis, 271, 274
site measurements, 291
Offshore piles: Terzaghi solution, 273
driving ana1ysis, 59 Potential expansiveness, 307
load capacity, 23 Preaugured píles, 11
resistance versus set curves, 67 Pressuremeter:
Overconsolidation, effect on skin friction, 24 for lateral soil modulus, 223
for piles to rock, 40
for p-y curves, 177
Pad foundat:ion, see Shallow foundations for skin friction, 43
Parameters of soil: Probability plot, 54
driving ana1ysis, 64 Proof test, 354
empirica1 correlations for settlement, 1O1-103 Punching of piles into softer strata, 28
interpretation from load tests, 357,365 p-y (p-p) analysis, 175-177, 182,223, 225
for lateralloading, 172-175, 223-225, 227, 320-321
load capacity in clay, 20-22
load capacity in sand, 26-28 Quake:
for negative friction, 285-286 definition', 60
Partially embedded piles: effect on resistance curves,
buckling loads, 327-329 64
lateral deflections, 18 7, 2 22 typical values, 64
SUBJECT INDEX 395