FOUNDATION
ANALYSIS AND DESIGN
JOSEPH E. BOWLES Third Editi
International Student EditionStandard U.S. reinforcing bars and metric bars available selectively
War Ne
uo
bee
Lae
Area,
00s
out
00
oat
O44
0.60
0.79
1.00
156
2235
4.00
Perimeter, Size,
mm
0.788 6
L178 8
10
1571 12
4
1.963 16
18
2.356 20
2.749
3.142
3Sd4
3.990
4.430 S
40
532 4s
7.09 60
‘ched as shown.
Area,
m? x
0.028
0.050
0.079
013
0.184
0.201
0
Perimeter
mm
1X8
2
U4
qT
440
50.2
508
os
oo
OKs
Kx
942
1008
108
1287
lana
IxKSFOUNDATION
ANALYSIS
AND DESIGN
Third Edition
Joseph E. Bowles
Consulting Enginoer|Sofiware Conculiant
Engineering Computer Software
Peoria, Illinois
INTERNATIONAL STUDENT EDITION
McGRAW-HILL INTERNATIONAL BOOK COMPANY
Auckland Bogota Guatemala Hamburg Johannesburg Lishon
London Madtid Mexico New Delhi Panama Paris
San Juan Sto Paulo Singapore Sydney TolkyoCOMPUTER PROGRAM DISCLAIMER
Neither the publisher or the author warrants the included programs to execute other than the
displayed output if the programs and data ate correctly entered into a computer. Any use of the
programs to solve problems other than those displayed is the sole responsibility of the user as to
whether the output is correct or is correctly interpreted.
‘This book was set in Times Roman by Santype-Byrd,
The editors were Julienne V. Brown and Susan Hazlett:
the production supervisor was Leroy A. Young.
New drawings were done by J & R Services, Inc.
FOUNDATION ANALYSIS AND DESIGN
INTERNATIONAL STUDENT EDITION
Copyright © 1982,
Exclusive rights by McGraw-Hill Kogakusha, Ltd. for
manufacture and export. This book cannot be re-exported
from the country to which it is consigned by McGraw-Hill.
Ist printing 1982
Copyright © 1982,1977,1968 by McGraw-Hill, Ine.
All rights reserved. No part of this publication may be
reproduced or distributed in any form or by any means, or
stored in a data base or retrieval system, without the
prior written permission of the publisher.
brary of Congress Cataloging in Publication Data
Bowles, Joseph E.
Foundation analysis and design.
Bibliography: p.
Includes indexes.
1, Foundations. 2. Soil mechanics. I. Title
TAT75.B63 1982 624.15. 81-13649
ISBN 0-07-006770-8 AACR2
When ordering this title use ISBN 0-07-085192.
Kosaipo PRINTI
LID. TOKYO, JAPAS.vi CONTENTS
Chapter 3
39
3-10
3-1
3-12
313
314
315
3-16
317
3-18
Chapter 4
at
42
43
44
45
46
47
48
49
410
411
412
4-13
414
Chapter 5
Sl
5-2
53
54
55
56
5-7
Exploration, Sampling, and In Situ
Soil Measurements
Data Required
Methods of Exploration
Planning the Exploration Program
Soil Boring
Soil Sampling
Marine Sampling
‘The Standard Penetration Test (SPT)
Other Penetration Methods
Core Sampling
Water-Table Location
Depth and Number of Borings
Presentation of Data
Field Load Tests
Field Vane Testing of Soils
Measurements of In Situ Stresses and K, Conditions
Static Penetration Testing- -Dutch-Cone Penetration Test (CPT)
The Borehole Shear Test
Seismic Exploration
Bearing Capacity of Foundations
Introduction
Bearing-Capacity Equations
General Comments on Bearing-Capacity Computations
Bearing Capacity—Examples
Footings with Eccentric or Inclined Loadings
Effect of Water Table on Bearing Capacity
Bearing Capacity for Footings on Layered Soils
Bearing Capacity of Footings on Slopes
Bearing Capacity from SPT
Bearing Capacity Using Cone Penetration Test (CPT) Data
Bearing Capacity of Foundations with Uplift or Tension Forces
Bearing Capacity Based on Building Codes
(Presumptive Pressure)
Safety Factors in Foundation Design
Bearing Capacity of Rocks
Foundation Settlements
The Settlement Problem
Stresses in a Soil Mass Due to Footing Pressure
The Boussinesq Method for Evaluating Soil Pressure
Westergaard’s Method for Evaluating Soil Pressures,
Immediate (Elastic) Settlement Computation—Theory
Immediate Settlements—Application
Alternative Methods of Computing Elastic Settlements,
0
79
80
81
84
89
96
7
102
102
105
108
108
110
a
117
121
122
123
130
130
131
135
140
143
147
149
153
155
159
160
163
163
167
im
1
172
173
178
183,
187
19258
59)
5-10
Sil
Chapter 6
61
6-2
63
64
65
6-6
67
6-8
Chapter 7
Tl
7-2
73
7-4
5S
7-6
TT
78
19
7-10
7
712
73
Chapter 8
Bd
8-2
83
84
8-5
8-6
87
88
8-9
8-10
811
8-12
CONTENTS Vii
Stresses and Displacements in Layered and Anisotropic Soils
Consolidation Settlements
Reliability of Settlement Computations
Proportioning Footings for a Given Settlement or
Equal Settlements
Structures on Fills
Structural Tolerance to Settlement and Differential Settlements
Improving Site Soils for Foundation Use
Introduction
‘Compaction
Precompression to Improve Site Soils
Drainage Using Sand Blankets and Drains
Vibratory Methods to Increase Soil Density
Foundation Grouting and Chemical Stabilization
Altering Groundwater Conditions
Use of Geotextiles to Improve Soil
Factors to Consider in Foundation Design
Footing Depth and Spacing
Displaced Soil Effects
Net versus Gross Soil Pressure—Design Soil Pressures
Erosion Problems for Structures Adjacent to Flowing Water
Corrosion Protection
Water-Table Fluctuation
Foundations in Sand Deposits
Foundations on Loess
Foundations on Expansive Soils
Foundations on Clays and Silts
Foundations on Sanitary Landfill Sites
Frost Depth and Foundations on Permafrost
Environmental Considerations
Spread Footing Design
Footings—Classification and Purpose
Allowable Soil Pressures in Spread Footing Design
Assumptions Used in Footing Design
Reinforced-Conerete Design—USD
Structural Design of Spread Footings
Bearing Plates and Anchor Bolts
Pedestals
Rectangular Footings
Eccentrically Loaded Spread Footings
Unsymmetrical Footings
Wall Footings and Footings for Residential Construction
Spread Footings with Overturning Moment
196
197
200
201
203
204
208
208
209
2ut
213
215
217
218
218
221
221
224
225
226
227
227
227
228
230
233
235
236
238
240
240
241
242
243
249
258
265
269
273
282
286
290vili CONTENTS
Chapter 9
9-8
9.9
9-10
9-1
Chapter 10
10-1
10-2
10-3
10-4
10.5
10-6
10-7
10-8
10-9
Chapter 11
11
M2
113
4
Us
11-6
11-7
11-8
11-9
11-10
ttt
11-12
11-13
1-1
Chapter 12
12-1
12-2
12-3
12-4
Special Footings and Beams on
Elastic Foundations
Introduction
Rectangular Combined Footings
Design of Trapezoid-Shaped Footings
Design of Strap (or Cantilever) Footings
Footings for Industrial Equipment
Modulus of Subgrade Reaction
Classical Solution of Beam on Elastic Foundation
Finite-Element Solution of Beam on Elastic Foundation
Bridge Piers
Ring Foundations
General Comments on the Finite-Element Procedure
Mat Foundations
Introduction
Types of Mat Foundations
Bearing Capacity of Mat Foundations
Mat Settlements
Design of Mat Foundations
Finite-Difference Method for Mats
Finite-Element Method for Mat Foundations
Mat-Superstructure Interaction
Circular Mats or Plates
Lateral Earth Pressure
‘The Lateral Earth Pressure Problem
Active Earth Pressure
Passive Earth Pressure
Coulomb Earth-Pressure Theory
Rankine Earth Pressures
Active and Passive Earth Pressure Using Theory of Plasticity
Earth Pressure on Walls, Soil-Tension Effects, Rupture Zone
Reliability of Lateral Earth Pressures
Soil Properties and Lateral Earth Pressure
Earth-Pressure Theories in Retaining-Wall Problems,
Graphical and Computer Solutions for Lateral Earth Pressure
Lateral Pressures by Theory of Elasticity for Surcharges
Other Causes of Lateral Pressure
Pressures in Silos, Grain Elevators, and Coal Bunkers
Retaining Walls
Introduction
Common Proportions of Retaining Walls
Soil Properties for Retaining Walls
Stability of Walls
295
295
295
304
309
312
320
326
330
339
341
344
349
349
350
351
352
354
361
363
374
374
378
378
379
381
381
388,
392
396
399
399)
401
404
ata
419
420
431
431
433
436
43812-5
12-6
12-7
128
12-9
12-10
12-11
12-12
12-13
12-14
12-15
12-16
Chapter 13
13-1
13-2
13-3
13-4
13-5
13-6
13-7
13-9
Chapter 14
14-1
14.2
143
144
14-5
146
14-7
148
14.9
Chapter 15
15-1
15-2
15-3
15-4
15-5
15-6
15-7
Retaining-Wall Forces
Allowable Bearing Capacity
Settlements
Tilting
Design of Gravity and Semigravity Walls
Wall Joints
Drainage
Abutment Wing and Retaining Walls of Varying Height
Design of a Cantilever Retaining Wall
Design of a Counterfort Retaining Wall
Basement or Foundation Walls; Walls for
Residential Construction
Reinforced-Earth Retaining Structures
Sheet-Pile Walls—Cantilevered
and Anchored
Introduction
Soil Properties for Sheet-Pile Walls
Types of Sheetpiling
Safety Factors for Sheet-Pile Walls
Cantilever Sheetpiling
Anchored Sheetpiling: Free-Earth Support
Rowe's Moment Reduction Applied to the
Free-Earth-Support Method
Finite-Element Analysis of Sheet-Pile Walls
Wales and Anchorages for Anchored Sheetpiling
Braced, Tieback, and
Slurry Walls for Excavations
Construction Excavations
Soil Pressures on Braced Sheeting or Cofferdams
Conventional Design of Single-Wall (Braced) Cofferdams
Estimation of Ground Loss around Excavations
Finite-Element Analysis for Braced Excavations
Instability Due to Heave of Bottom of Excavation
Other Causes of Cofferdam Instability
Construction Dewatering
Slurry-Wall (or -Trench) Construction
Cellular Cofferdams
Cellular Cofferdams: Types and Uses
Celi Fill
Stability and Design of Cellular Cofferdams
Practical Considerations in Cellular Cofferdam Design
Design of Diaphragm Cofferdam Cell
Circular-Cofferdam Design
Cloverleaf-Cofferdam Design
CONTENTS ix
440
448
449
450
453
457
458
459
460
466
468
469
474
474
415
477
479
481
489
495
499
506
516
516
519
522
327
530
536
539
540
544
548
548
582
553
563
565
568
$73X CONTENTS
Chapter 16
16-1
16-2
16-3
16-4
16-5
16-6
16-7
16-8
16-9
16-10
16-11
16-12
16-13
16-14
16-15
Chapter 17
17-4
17:2
17-3
17-4
17-5
17-6
17-7
17:8
17.9
Chapter 18
18-1
18-2
18-3
18-4
38-5
18-6
18-7
18-8
18-9
Chapter 19
19-1
19:2
19-3
19-4
19-5
Single Piles—Static Capacity and
Lateral Loads; Pile/Pole Buckling
Introduction
Timber Piles
Concrete Piles
Steel Piles
Corrosion of Steel Piles
Soil Properties for Static Pile Capacity
Static Pile Capacity
Ultimate Static Pile Point Capacity
Skin Resistance Capacity
Static Pile Capacity—Examples
Piles in Permafrost
Static Pile Capacity Using Load-Transfer Load-Test Data
Tension Piles—Piles for Resisting Uplift
Laterally Loaded Piles
Buckling of Fully and Partially Embedded Piles and Poles
Single Piles—Dynamic Analysis
Dynamic Analysis
Pile Driving
The Rational Pile Formula
Other Dynamic Formulas and General Considerations
Reliability of Dynamic Pile-Driving Formulas
The Wave Equation
Pile-Load Tests
Pile-Driving Stresses
General Comments on Pile Driving
Pile Foundations—Groups
Single Piles versus Pile Groups
Pile-Group Considerations
Efficiency of Pile Groups
Stresses on Underlying Strata
Settlements of Pile Groups
Pile Caps
Batter Piles
Negative Skin Friction
Matrix Analysis for Pile Groups
Caissons Including Drilled Piers
‘Types of Caissons
Open-End Caissons
Closed-End, or Box, Caissons
Pneumatic Caissons
Drilled Caissons
575
375
582
584
589
592
592
593
598
602
610
616
620
622
623
632
638
638
638
643
647
654
656
663
665
668
on
671
om
673
676
683
687
690
690
696
705
705
706
mM
ns
1719-6
19-7
19-8
19-9
‘hapter 20
A-l
A-2
A3
A-4
A-S
B-1
B-3
CONTENTS Xi
Bearing Capacity and Settlements of Drilled Caissons
Design of Drilled Caissons
Laterally Loaded Caissons
Inspection of Drilled Caissons
Design of Foundations for Vibration Control
Introduction
Elementary Vibrations
Forced Vibrations for a Lumped Mass
Approximate Solution of Vibrating Foundation——
Theory of Elastic Half-Space
Lumped-Mass Solution of the Vibrating Foundation
Soil Properties—Elastic Constants
Coupled Vibrations
Effect of Piles to Reduce Foundation Vibrations
Other Considerations for Machinery Foundations
Appendixes
General Pile-Hammer and Pile Data Tables
H-Piles
Pile Hammers
Sheet Piles
Pipe Piles
Prestressed-Concrete Piles
Selected Computer Programs
Beam, Lateral- and Sheet-Pile Finite-Element Program
Mat Program
Three-Dimensional Pile-Group Program
References
Indexes
Name Index
Subject Index
720
725
729
730
732
732
733
738
743
749
757
759
760
761
765
766
769
mM
773
714
775
781
787
788PREFACE
This edition is the latest in the continuing process of producing an up-to-date
compendium of methods and procedures for the analysis and design of founda-
tions. As in the earlier editions the primary focus is on interfacing structural
elements with the underlying soil. This is where the major focus of foundation
engineering lies in both the author's opinion and that of many others. Engineer-
ing of dams, fills, and embankments, and flow of water in soil masses may more
properly fall under the general category of geotechnical engineering. In some
cases the latter considerations may be major factors in designing or constructing
a foundation; therefore, some background material on several of these topics has
been included.
Most engineers now recognize that it is not possible or very practical to
identify foundation engineering as soil mechanics/properties with structural en-
gineers designing the foundation elements. A foundation engineer must be versed
in both the geotechnical aspects of soils as well as the structural behavior pro-
duced by the often complex foundation-soil interaction. The latter statement
reflects the general design philosophy contained in this text.
I have undertaken a fairly extensive revision; however, none of the chapter
titles have been changed and in most cases the section headings have been
retained. Chapters 2, 8, 9, and 16 have been almost totally rewritten, with very
substantial rewriting undertaken in Chaps. 3, 4, 10, and 11. This was done to
produce a more logical sequencing of topics and to include new methodology
which was in transition when the second edition was being rewritten. The ma-
terial is about 80 percent SI to reflect both the general trend in textbooks and the
anticipated status of SI for most of the useful life of the book.
In many cases there is no “unique” design equation/methodology and one
or more of several alternatives tends to be preferred by certain engineers or in
geographical areas. I have attempted to present those alternatives where they
seem to be widely enough used to warrant the text space. Where it was practical,
xiiiXiV PREFACE
examples are analyzed using one or more of the alternatives so the reader obtains
both familarity and an opinion about the procedure. In several examples a “fee!”
of a probable answer is produced by use of the alternatives a common engineer-
ing office practice where the input data arc uncertain—cither as a direct com-
putation or as an average from the several alternatives. | have attempted to
include realistic example (and home) problems for the reader, with the examples
being somewhat less edited. About 50 percent of the home problems are new, and
more answers are included than in the earlier editions.
Several methods such as the footing on slope, grid analysis of mats, and the
sheet and lateral pile solutions using the finite-element method have been com-
pared via examples with the solutions of others or with alternative methods. This
has been done to illustrate that these new methods are adequate. In passing it
should be noted that the finite-element solution’contained herein for beams, sheet
piles, and lateral piles is extremely widely used. It is the author's opinion that the
simplest solution which produces a satisfactory and cconomical design is pre-
ferred. Solutions which require esoteric mathematics to produce miniscule com-
putational refinement in a mathematical model based on soil data which are
uncertain at best are not very practical, Many solutions of this type do not get
further than the pages of a technical journal and others soon disappear from
engineering practice after a short time.
T have included ahout S60 references so that almost every topic can be
researched in depth. | have tried to avoid using obscure references which could be
obtained only with great difficulty by the average user. Most worthwhile material
eventually gets published in some form by ASCE, ASTM, TRB, CGJ, at a spe
cialty conference, or in the proceedings of the ICSMFE, which are not very
difficult to acecss. I have retained the usc of the list of publication abbreviations
(above included) at the beginning of the bibliography to reduce its size. It is
hoped that no significant work has been omitted; however, in the interest of
space, not all the work on a given Copic is cited. Generally the most recent or
those works with the best bibliography coverage were included. I hope also that I
have not offended the junior authors of coauthored references by the use of “et
al.” when there are more than two authors.
T wish to express appreciation to the many users both in the United States
and abroad who have written or called with comments or constructive criticism
or simply to make inquiry about a procedure. I should also like to thank those
who took part in the McGraw Hill user survey to provide input for this revision
and include: Jack Bakos of Youngstown State University; William Baron of
Clemson University; William Gotolski of Pennsylvania State University; and
Roy V. Snedden of the University of Nebraska. I would also like to thank the
final manuscript reviewer William Baron of Clemson University.
Finally I should like to acknowledge the considerable contribution of my
wife Faye, who helped as usual with the typing and the myriad other operations
to produce the manuscript.
Joseph E. BowlesFOUNDATION ANALYSIS AND DI!CHAPTER
ONE
INTRODUCTION
1-1 FOUNDATIONS—DEFINITION AND PURPOSE
All structures designed to be supported by the earth, including buildings, bridges,
earth fills, and earth, earth and rock, and concrete dams, consist of two parts.
These are the superstructure, or upper part, and the substructure element which
interfaces the superstructure and supporting ground. In the case of earth fills and
dams, there is often not a clear line of demarcation between the superstructure
and substructure. The foundation can be defined as the substructure and that
adjacent zone of soil and/or rock which will be affected by both the substructure
element and its loads.
The foundation engineer is that person who by reason of experience and
training can produce solutions for design problems involving this part of the
engineered system. In this context, foundation engineering can be defined as the
science and art of applying the principles of soil and structural mechanics toge-
ther with engineering judgment (the “art") to solve the interfacing problem. The
foundation engineer is concerned directly with the structural members which
affect the transfer of load from the superstructure to the soil such that the
resulting soil stability and estimated deformations are tolerable. Since the design
geometry and location of the substructure element often have an effect on how
the soil responds, the foundation engineer must be reasonably versed in structural
design.
A number of practical considerations are a part of the engineering of a
foundation:
1. Visual integration of geologic evidence at a site with any field or laboratory
test data.2 FOUNDATION ANALYSIS AND DESIGN
2. Establishing of an adequate field exploration and laboratory testing program.
3. Design of the substructure elements so that they can be built—and as econ-
omically as possible.
4. Appreciation of practical construction methods and of likely-to-be-obtained
construction tolerances. Stipulation of very close tolerances can have an enor-
mous effect on the foundation costs.
These several items are not directly quantifiable and thus require a considerable
application of common sense.
A thorough understanding of the principles of soil mechanics in terms of
stability, deformations, and water flow is a necessary ingredient to the successful
practice of foundation engineering. Of nearly equal importance is an under-
standing of the geological processes involved in the formation of soil masses. It is
now recognized that both soil stability and deformation are dependent on the
stress history of the mass. It has been common until recently to associate founda-
tion engineering solely with soil mechanics concerns, leaving the interfacing el-
ements to the structural (or other) designer. Current trends are to recognize that
foundation engineering is a systems problem and cannot be nicely com-
partmentalized as some persons would prefer. Readers may determine the validity
of this statement as they progress through the text.
The science of soil m&chanics and its relationship to geological processes has
progressed considerably over the past fifty years. However, hecause of the natural
variability of soil and the resulting problems associated with testing, which will
be elaborated upon in Chap. 3, the design of a foundation still depends to a large
degree upon “art,” or the application of engineering judgment. A subset of this
application is the assessment of the tolerable risk associated with the foundation.
The primary focus of this text will be on analysis and design of the interfac-
ing elements for buildings and retaining structures and those soil mechanics
principles particularly applicable to these elements. These interfacing elements
include both near surface members such as footings and mats and deep elements
such as piles and caissons. Retaining structures of concrete (commonly termed
retaining walls) and metal (as sheetpiling) are considered in later chapters. Soil
mechanics principles include both stability, including soil water effects, and defor-
mation analyses. Soil stability can often be enhanced by various improvement
techniques, the most common being compaction, and several of the more popular
of these methods will be briefly considered in Chap. 6.
1-2 FOUNDATION CLASSIFICATIONS
Foundations for structures such as buildings, from the smallest residential to the
tallest high-rise, and bridges are for the purpose of transmitting the super-
structure loads. These loads come from column-type members with stress inten-
sities ranging from perhaps 140 mPa for steel to 10 mPa for concrete to the
supporting capacity of the soil, which is seldom over 500 kPa but more often onINTRODUCTION 3
the order of 200 to 250 kPa. The reader can readily note that this interface
connects materials whose differences in useful engineering strength can vary by a
tactor of several hundred. The transmission of these large superstructure loads to
the soil may be by use of:
| Shallow foundations --termed footings, spread footings, or mats. Foundation
depth is generally D < B (see Chap. 4).
2. Deep foundations—piles or caissons with D > 4 to 5B (see Chaps. 16 to 19).
Any structure used to retain a soil or similar mass such as grain, coal, or ore
1 a geometric shape other than that occurring naturally under the influence of
vravity is a retaining structure. Any foundation not classed as shallow, deep, or a
retaining structure may be termed a special foundation.
Typical foundation types are:
1. Foundations for buildings (either shallow or deep)
2. Foundations for smokestacks, radio and television towers, bridge piers, indus-
trial equipment, etc. (either shallow or deep)
3, Foundations for port or marine structures (may be shallow or deep and with
retaining structures extensively used)
4. Foundations for rotating, reciprocating, and impact machinery, and for tur-
bines, generators. etc. (either shallow or deep and may require vibration con-
trol)
5. Foundation elements to support excavations or retain earth masses as for
bridge abutments and piers, or retain grain, ore, coal, etc. (retaining walls or
sheet-pile structures)
Foundations for buildings are extremely numerous; foundations for the sev-
eral other types of superstructures are constructed in somewhat lesser numbers.
1-3 FOUNDATION SITE AND SYSTEM ECONOMICS
A building foundation must be adequate if the structure is to perform satisfac-
torily and be safe for occupancy. Other foundations must be adequate to perform
their intended functions in a satisfactory and safe manner; however, buildings
usually have more stringent criteria for safety and performance than other
structures notable exceptions being nuclear-plant facilities, turbines for power
generation, and certain types of radio-antenna equipment. Foundations for nu-
clear plants require extremely rigid design/performance criteria for safety reasons.
The other foundations support extremely expensive machinery which is often
very sensitive to small soil deformations
More recently, and after loss of life from several avoidable failures, dam
designs where soil is the principal construction material are being more carefully
made, One might note that more principles of soil mechanics and geology apply4 FOUNDATION ANALYSIS AND DESIGN
to earth dams than to the majority of foundation engineering problems. In ad
tion to the stringent criteria of the superstructure, instability and water flow
through the base soil are serious considerations. A further area of concern is the
inevitable deformation of the base soil and subsidence in the superstructure (dam
fill material). Careful attention to the occurrence of the latter deformations can
allow the designer to avoid a base crack in the dam and the resulting piping
failure, or a crest crack and the associated overtopping failure.
‘Almost any reasonable structure can be built and safely supported if there is
unlimited financing. Unfortunately, in the real situation this is seldom, if ever, the
case, and the foundation engineer has the dilemma of making a decision under
much less than the ideal condition. Also, even though the mistake may be buried,
the results from the error are not and can show up relatively soon—and probably
before any statute of limitations expires. There are reported cases where the
foundation defects (such as cracked walls or broken mechanical fixtures) have
shown up years later—also cases where the defects have shown up either during
construction of the superstructure or immediately thereafter.
Since the substructure is buried, or is beneath the superstructure, in such a
configuration that access will be difficult should foundation inadequacies develop
after the superstructure is in place, it is common practice to be conservative. A
‘one or two percent overdesign in these areas produces a larger potential invest-
ment return than in the superstructure.
The designer is always faced with the question of what constitutes a safe,
economical design while simultaneously contending with the inevitable natural
soil heterogeneity at a site. Nowadays that problem may be compounded by land
scarcity requiring reclamation of areas which have been used as sanitary landfills,
garbage dumps, or even hazardous waste disposal areas. Still another com-
plicating factor is that the act of construction can alter the soil properties con-
siderably from those used in the initial analyses/design of the foundation. These
factors result in foundation design becoming so subjective and difficult to quan-
tify that two design firms might come up with completely different designs that
would perform equally satisfactory. Cost would likely be the distinguishing fea-
ture for the preferred design.
This problem and the widely differing solutions would depend, for example,
on the following
1, What constitutes satisfactory and tolerable settlement; how much extra could,
or should, be spent to reduce estimated settlements from say 30 to 15 mm?
2. Has the client been willing to authorize an adequate soil exploration pro-
gram ? What kind of soil variability did the soil borings indicate ? Would
additional borings actually improve the foundation recommendations ?
3. Can the building be supported by the soil using,
a. Spread footings—least cost.
b. Mats—intermediate in cost.
c. Piles or caissons-—several times the cost of spread footings.
4. What are the consequences of a foundation failure in terms of public safety ?INTRODUCTION 5
What js the likelihood of a lawsuit if the foundation does not perform ad-
equately ?
+ Is sufficient money available for the foundation ? It is not unheard of that the
foundation alone would cost so much the project is not economically feasible.
It may be necessary to abandon the site in favor of one where foundation costs
are affordable.
6 What is the ability of the local construction force ? It is hardly sensible to
design an elaborate foundation if no one can build it, or if it is so different in
design that the contractor includes a large “uncertainty” factor in the bid
What is the enginccring ability of the foundation engineer ? While this factor is
listed last, this is not of least importance in economical design. Obviously
engineers have different levels of capability just as in other professions (law-
yers, doctors, professors, etc.) and in the trades such as carpenters, electricians,
and painters.
If the foundation fails because of any cost shaving (in reality implicitly ac-
cepting a higher risk), the client tends to quickly lose appreciation for the tempor-
ary financial benefit which accrued. At this point, facing heavy damages and/or a
lawsuit, the client is probably in the poorest mental state of all the involved
parties. Thus, one should always bear in mind that absolute dollar economics
may not produce good foundation engineering
The foundation engineer must look at the entire system: the building pur-
pose, probable service-life loading, type of framing, soil profile, construction me-
thods, and construction costs to arrive at a design that is consistent with the
client/owner’s needs and does not excessively degrade the environment. This must
be done with a safety factor which produces a tolerable risk level to both the
public and the owner.
Considering these several areas of uncertainty, it follows that risk and lia-
bility insurance for persons engaged in foundation engineering is very costly. In
attempts to reduce these costs as well as produce a design which could be
obtained from several engineering firms (ie., a “consensus” design) there is active
discussion (and the practice has already been undertaken in several areas) of
having the foundation engineer submit the proposed design to a board of quali-
fied engineers for a “peer review.”
1-4 GENERAL REQUIREMENTS OF FOUNDATIONS
‘A foundation must be capable of satisfying several stability and deformation
requirements such as:
1. Depth must be adequate to avoid lateral expulsion of material from beneath
the foundation particularly for footings and mats.
Depth must be below the zone of seasonal volume changes caused by freezing,
thawing, and plant growth.
N6 FOUNDATION ANALYSIS AND DESIGN.
3. System must be safe against overturning, rotation, sliding, or soil rupture
(shear-strength failure).
4, System must be safe against corrosion or deterioration due to harmful materi-
als present in the soil. This is a particular concern in reclaiming sanitary
landfills and sometimes for marine foundations.
5. System should be adequate to sustain some changes in later site or construc-
tion geometry, and be easily modified should later changes be major in scope.
6. The foundation should be economical in terms of the method of installation.
7. Total earth movements (generally settlements) and differential movements
should be tolerable for both the foundation and superstructure elements.
8. The foundation, and its construction, must meet environmental protection
standards.
1-5 FOUNDATION SELECTION
The different types of foundations shown in Table 1-1 will be taken up in some
detail in later chapters. It will be useful, however, at this point to enumerate the
several types and their potential application. Where*groundwater is present, it is
understood that if the depth is below the depth of the footing (or excavation) it
will not be a problem. If groundwater is within the construction zone, it must be
removed by pumping down the water table, using grout or concrete curtain walls,
steel shells, or other means as appropriate.
When groundwater is removed, or when construction is below the water
table such that the groundwater may become polluted, approval of appropriate
governmental agencies is generally involved to minimize the environmental ef-
fects.
1-6 SI] AND FPS UNITS
This textbook will use both Foot-pound-second (Fps) and SI units. The SI units
will be both those generally accepted and “preferred usage.” Problems will be
either SI or Fps—there will be no intermixing. In the text either set of units may
be used but the alternate system units will not be included in brackets as is
commonly done in a number of publications. This method of usage will help the
reader in the transition from Fps to SI by requiring concurrent thinking in both
sets of units,
Preferred usage will entail a number of mass and pressure units. The reason
is that most soil laboratory equipment lasts for years (scales, pressure gages,
calipers, etc.) and few laboratories (even in SI countries) have this equipment in
true SI units.
In this text we will define density as a mass unit [kg/m?, kg/cm, or g/cm?
and lb/ft? (pef)]. Unit weight is a force unit and in units of pounds or kips/ft? or
kilonewton/m? (kN/m®). The SI unit of kilonewton will be used for most if not allTable 1-1 Foundation types and typical usage
INTRODUCTION 7
Foundation type
Use
Applicable soil conditions
Spread footing, wall footings
Mat foundation
Pile foundations
Floating
Bearing
Caisson (shafts 75 em or more
in diameter) generally bear-
ing or combination of bear-
ing and skin resistance
Retaining walls, bridge abut-
ments
Sheet-pile structures
Individual columns. walls.
bridge piers
Same as spread and wall foot-
ings. Very heavy column
loads, Usually reduces differ-
ential settlements and total
settlements
In groups (at least 2) to carry
heavy column, wall loads: re-
quires pile cap
In groups (at least 2) to carry
heavy column, wall loads: re~
quires pile cap
Larger column loads than for
piles but eliminates pile cap
by using eaissons as column
extension
Permanent retaining structure
Temporary retaining structures
as excavations, waterfront
structures, cofferdams
Any conditions where bearing
capacity is adequate for
applied load, May use on
single stratum: firm layer
over soft layer or soft layer
over firm layer. Check
immediate, differential, and
consolidation settlements
Generally soil bearing value is
less than for spread footings:
over one-half area of build-
ing covered by individual
footings. Check settlements
Poor surface and near surfa
soils. Soils of high bearing
capacity 20-50 m below
basement or ground surface,
but by distributing load
along pile shaft soil strength
is adequate. Corrosive soils
may require use of timber or
concrete pile material
Poor surface and near-surface
soil of high bearing
capacity (point bearing on)
8 50m below ground sur-
soils
face
Poor surface and near-surface
soils: soil of high bearing
capacity (point bearing on) is
850m below ground sur
face
Any type of soil, but a specified
zone (Chaps. 11. 12) in back
of wall usually of controlled
backfill
Any soil: waterfront structure
may require special alloy or
corrosion protection. Coffer-
dams require control of fill
material8 FOUNDATION ANALYSIS AND DESIGN
soil quantities. This is because the newton is too small and the meganewton is
too large (except for steel and concrete stresses).
Using KN will give soil pressures such as intergranular and bearing capacities
in kN/m? (kilopascal, kPa). In the Fps system we have correspondingly the
pressure unit of kips/ft? (ksf).
A conversion factor set which is very useful to remember is that to convert
mass density in g/cm? to kN/m? or pef
9.807 KN/m? (actually 9.80655 but rounded)
1 g/em? = 62.4 pef
1 ksf = 47.88 kPa (say, 50 kPa for general use).
The unit weight of water is 62.4 pef or 9.807 KN/m?, and has a density of 1 g/cm?
or 1000 kg/m. Soil will normally vary from
1 g/cm?
Fps: 90 to 130 pef
SI: 14 to 21 kN/m*
Note: If the unit weight y is reported to 0.1 pef, then one should report to
0.01 kN/m* for comparable accuracy.
The Fps units of load in tons and pressure in tons/ft? has been rather widely
uscd in the past both in the field and for certain laboratory tests. This was
because the widely used European unit of | kg/cm? is very nearly 1 ton/ft? (see
Table 1-2) using the 2000-Ib ton. This text will use only kips, pounds, and
kilonewtons for these units, primarily for consistency. The reader should carefully
note the source and context when the “ton” unit is used. Sometimes, but not
Table 1-2 Useful SI and metric factors* for convert-
ing Fps units
To convert from To Multiply by
inch centimeter (em) 2.54
square inch square centimeter (em?) 6.45160
cubic inch cubic centimeter (cm*) 16.38706
in* em* 4162314
kilogram force, ky newton (N) 9.80665
pound force, Ib, key 0.45359
kip kilonewton (kN) 4.44822
kg/em? KN/m? (kPa) 98.0665
kip/in® (ksi) KN/m? (kPa) 6894.757
Ibyin? (psit kg/em? 0.07031
ton/ft? kg/em? 0.97653
foot + kip KN - meter 1.35582
kef kN/m3 157.08762
* Using latest (1980) ASTM factors in conversions.INTRODUCTION 9
slways, it is a metric ton of 1000 kg. In some literature the metric ton is spelled
and represents the long ton of 2204 Ib.
1-7 COMPUTATIONAL ACCURACY VERSUS DESIGN
PRECISION
The pocket and desktop calculators, and digital computers, compute with 10 to
14 digits of accuracy. This gives a fictitiously high precision to computed quan-
tities Whose input values may have a design precision only within 10 to 30
percent of the numerical value used. The reader should be aware of this actual
\ersus computed precision when checking the example data and output. The
author has attempted (o maintain a checkable precision by writing the intermedi-
sue value (when a pocket calculator was used) to the precision the user should
imput to check the succeeding steps. If this intermediate value is not used, the
computed answer can easily differ in the 0.1 position. The reader should also be
aware that typesetting, transcribing, and typing errors inevitably occur, par-
ticularly in misplaced parentheses and misreading 3 for 8, etc.
The text user should be able to reproduce most of the digits in the example
problems to 0.1 or less unless a typesetting (or other) error has inadvertently
occurred. There may be larger discrepancies if the reader uses interpolated data
and the author used “exact” data from a computer printout without inter-
polating. Generally this will be noted so that the reader is alerted of potential
computational discrepancies. The example problems have been included pri-
marily for procedure rather than numerical accuracy, and the user should be
aware of this underlying philosophy when studying them.CHAPTER
TWO
SOIL MECHANICS IN FOUNDATION
ENGINEERING
2-1 INTRODUCTION
No construction material has both engineering and physical properties which are
more variable than the ground, These properties vary both laterally and verti-
cally and often by large orders of magnitude. Those properties of particular
interest to the foundation engineer include:
1. Strength parameters (stress-strain modulus, shear modulus, Poisson's ratio,
cohesion, and angle of internal friction)
2. Compressibility indexes (for deformation/settlement)
3. Permeability
4, Gravimetric-volumetric data (unit weight, specific gravity, void ratio, water
content, etc.)
Some knowledge about these properties allows the engineer to make esti-
mates for:
1. Bearing capacity
2. Settlements including both the amount and rate
3. Earth pressures (both vertical and lateral)
4. Pore pressures and dewatering quantities
Not all engineering properties of the soil are of equal importance in the design
and construction of a given foundation. This review chapter will emphasize those
10SOIL MECHANICS IN FOUNDATION ENGINEERING 11
soil mechanics principles most applicable for the analysis and design of founda-
tions.
The practice of foundation engineering was largely empirical until 1925,
when K. Terzaghi, often called the “father of soil mechanics” published his book
“Erdbaumechanik auf bodenphysikalischer Grundlage.” From this point
onward, and with the years from 1950 onward being especially fruitful, founda-
tion engineering has developed into a more rational approach. This is not to say
that there is not still a large element of “art” and empiricism in the practice, but
the currency tendency is to rely much more on laboratory testing and recognized
principles of the behavior of elastoplastic solids than was done earlier.
Much laboratory work in the area of soil behavior has been done and
reported in geotechnical literature in recent years. Most of this work has been
done on samples prepared under ideal conditions in the laboratory. This tends to
produce samples which are rather homogeneous, uniform, and generally lacking
in geological aging so that those properties of anisotropy and cementation are
often not produced. A few laboratories attempt to reproduce anisotropy, but the
practice does not seem to be widespread at present. Reproducing aging and
environmental processes in the laboratory to obtain natural cementation effects is
generally too time-consuming to be practical. Tests on laboratory samples, how-
ever, constitute a large part of the data base on which empirical correlations and
field predictions are made. When one takes into account the actual soil makeup,
its geologically obtained properties, and the difficulties of obtaining samples
which have sufficiently small amounts of disturbance that the resulting test data
are reliable, test data on laboratory prepared samples may bear little resemblance
to field performance.
Several problems are involved with laboratory testing or field samples, in-
cluding:
1. Recovery of undisturbed samples
2. Small quantity of samples relative to the volume of soil involved
3. Limitations on laboratory test equipment (and sometimes of qualified per-
sonnel)
It is not difficult to see that “engineering judgment” will play a significant role in
the practice of foundation engineering. The term engineering judgment is not
being used in this context to give dignity to a “guess,” regardless of how it may
appear to the casual observer.
The proper application of engineering judgment requires that the foundation
engineer have available a site profile, soil property data, and sufficient geological
information to arrive at a safe, economical, and practical decision. In cases such
as one-story load-bearing wall construction used for buildings of the department
store, office, and service station type where the soil is relatively homogeneous, the
necessary information may comprise only the boring logs from four or five re-
latively shallow exploratory borings. For a 10-story building the necessary infor-
mation would normally have to be more. Where a 100-story building is involved,12 FOUNDATION ANALYSIS AND DESIGN
the amount of information would be very considerable and might cost on the
order of 0.5 to 1 percent of the total construction cost. It should be obvious that
in any of the examples cited it would be helpful if the foundation engineer had
provided recommendations and/or designs for previous projects near the current
site.
2-2 FOUNDATION MATERIALS
The foundation engineer is concerned with the construction of some type of
engineered structure on earth, generally above any water surface—but it may be
below as for marine structures located under bodies of water, or rock. The earth
is composed of a mixture of rock and soil, the latter being weathered or degraded
rock. Water in varying amounts is found in the pores, or voids, of the soil and in
more limited quantities in the cracks and pores of the rock. Air is also present in
many of the void spaces but usually is not a significiant factor in foundation
design.
Rock will be defined as that naturally occurring material composed of min-
eral particles so firmly bonded together that relatively great effort is required to
separate them (ie., blasting, heavy crushing, or ripping forces), Soil is defined as
naturally occurring mineral aggregations which can be readily separated into
elemental particles and in mass form contains numerous voids. These voids
contain air, water, or organic materials in varying quantities. The elemental soil
particles are formed from the decomposition of rock by mechanical (air, ice, wind,
and water) and chemical processes.
Soil deposits can be described as residual or transported formations. A resid-
ual soil formation is one which has been formed at the current location by
decomposition of the parent rock. A transported soil is one which has formed at
‘one location and has since been transported by wind, water, ice, or gravity to the
current site. The terms residual and transported must be taken in the proper
context, since many of the current residual soils are formed (or are being formed)
from transported soil deposits of earlier geological periods. In many cases these
earlier deposits have indurated to become rocks, and later uplifts have exposed
this material (both rock and compressed soil) to a new onset of weathering. Areas
in which these processes are ongoing include the Appalachian mountains,
Piedmont regions, and much of the Plains area of the United States. Similar
areas exist on other continents.
The initial reason for classifying soil deposits as transported or residual was
that the latter soils tended (but not always) to have better engineering properties
of strength and deformation. In this context it may be preferable to apply the
classification of “transported soil” to a deposit of relative recent (geologically
speaking) age and produced from the activity of recent glacial, windborne (loess),
and (including ongoing) water action, A number of geological survey maps are
available to locate areas where there is a high likelihood of encountering a
transported soil deposit.SOIL MECHANICS IN FOUNDATION ENGINEERING 13
2-3 SOIL VOLUME AND DENSITY RELATIONSHIPS
The more common soil definitions and mathematical relationships are presented
in the following sections. Figure 2-1 illustrates and defines a number of the terms
used in the following relationships.
Void ratio e. The ratio of the volume of voids V, to the volume of soilds ¥, in a
given volume of material, usually expressed as a decimal.
en} (2-1)
Porosity n. The ratio of the volume of voids to the total volume V, of a soil mass;
may be expressed as a percentage or a decimal.
na (2-2)
Water content w. The ratio of the weight of water W,, in a given soil mass to the
weight of soil solids W, in the same mass and expressed as a percentage.
x 100 (2-3)
Unit density p. The ratio of mass per unit of volume. In the Fps system the
values are the same as unit weight following. The SI system gives units of kg/m?,
but preferred usage units are g/cm? or kg/cm’.
Unit weight } (as distinct from density). The ratio of the weight of soil to the
corresponding volume with units of force per unit volume. The general expression
is
(2-4)
Commonly used units are pef, kef, or kN/m?. Dry unit weight is based on using the
‘tS am mr ey PS Pango] 77
of % js pee
| Woter ¥| . | (|? We Ff
oY Hes Bost oe
rhecte| Teleestt] | teed gl
tot ot : eee: md
lo) te) te}
Figure 2-1 (a) Weight/volume relationships for a soil mass; (b) volume/void relationships; (c) volumes
expressed in terms of weights and specific gravity14 FOUNDATION ANALYSIS AND DESIGN
weight of soil solids W, in Eq. (2-4), The saturated unit weight is obtained when
W, is based on a state when all the soil voids are filled with water.
Degree of saturation S. The ratio of the volume of water to the total volume of
soil voids and expressed as a percentage.
s Ge 100 (2-5)
A “saturated” soil as obtained from beneath the water table may have a com-
puted S of 95 to 100 percent.
Specific gravity G. The ratio of the unit weight of a material in air to the unit
weight of distilled water at 4°C. No serious error in soils work is introduced at
the usual laboratory temperature ranges of 15 to 25°C. The average specific
gravity of a mass of soil grains G, is computed as
(2-6)
In this equation ),, = unit weight of water, usually taken as 1.00 g/cm?*.
These six basic definitions/equations are sufficient to develop any needed
volumetric-gravimetric relationships for soil mechanics problems. For example, a
useful relationship between the void ratio e and porosity a can be obtained as
follows (see Fig. 2-16).
Since the volume values are symbolically given, let the volume of solids
V, = 1.00 and from Eq. (2-1) directly obtain e = V,. Placing these values on the
left side of Fig. 2-1b gives V, = | + e. Making the indicated substitutions into Eq
(2-2), obtain
(2-7)
VY ite
(2-8)
From Fig. 2-1a one obtains the total weight of soil by inspection as
W= Wo + W, (2-9)
From the definition of the water content the W, is obtained as wW,, and by
substitution we obtain
W, = W, + WW,
Solving for W,,
(2-10)SOIL MECHANICS IN FOUNDATION ENGINEERING 15
Now dividing through by the total volume V; and using the unit weight definition
of Eq. (2-4) obtain the very useful relationship for dry unit weight as
" Ywet
Yacy =
I. Generally K, < 1, but for
OCR greater than 4 or 5 it may be 1.0 or somewhat larger.
When site conditions change, the soil goes from the K, stress state to a new
stress state. During the transition from the existing to a new steady-state condi-
tion the soil stresses are very complex and difficult to analyze.
In the in situ K, condition the soil may be homogeneous (same material) but
is seldom isotropic (has same properties both horizontally and vertically). In
most, if not all, cases the soil is anisotropic, that is, elastic properties and per-
meability are different in the vertical and lateral directions. At the very least the
soil tends to be stratified and increases in density with depth.
4. Intact or fissured clays. The term “clay” is used to describe any cohesive soil
deposit with sufficient clay present that drying produces shrinkage with the
formation of cracks or fissures such that block slippage can occur. These soils
can be troublesome for both field sampling and laboratory testing. In labora-
tory strength tests the fissures can define failure planes and produce fictitiously
low strength predictions. In situ, however, when the potential slip block is
confined or bridged by the loaded area, the fissures may be of much less
importance. A greater potential for strength reduction exists in fissured (or
jointed) clays during construction. Opening an excavation reduces the over-
burden pressure p, so that the fissures may enlarge. Rainwater or humidity can
enter these openings with a result of softening the clay and reducing its
strength. An “intact” clay, as the name implies, is one which is free of fissures
and joints. Intact clays tend to be normally consolidated (OCR = 1), since
shrinkage cracks are an indication that drying stresses have been developed,
with some apparent preconsolidation being produced.
5. Soil water. Soil water is a major factor in foundation design from the
standpoint of construction problems and of its adverse effect on the shear
strength. It is a factor in both cohesionless and cohesive soil deposits. Water
effects in cohesionless soils include providing surface tension which may allow
low vertical cuts, making it difficult to compact unless the water content is
high and converting the mass to a viscous fluid (termed liquefaction) under
shock-load conditions when S—> 100 percent. Water effects in cohesive soils
include changes in plasticity (softening or stiffening) from changes in water
content. Shear strength is markedly affected by the presence of water via
softening (or stiffening) and also by changes in pore pressure from external
causes (load increase or decrease, lowering of water table, pumping, etc.).SOIL MECHANICS IN FOUNDATION ENGINEERING 19
2-5 ROUTINE LABORATORY TESTS
The following tests are those most commonly used for foundation design work
unless stated otherwise.
Atterberg Limits
The Atterberg limits are routine laboratory tests for arbitrary moisture contents
to determine when the soil is on the verge of being a viscous fluid (liquid limit
w,) or nonplastic (plastic limit w,). The plasticity index Ip,
is the arbitrary range of water contents (Fig. 2-2a) for which the soil is plastic.
The shrinkage limit ws is that water content beyond which no further reduction
of mass volume takes place with further drying.
‘The liquid and plastic limits are widely used for soil classification. The plasti-
city index is commonly used as a correlation factor to estimate the angle of
internal friction and settlement parameters.
The liquidity index defined as
(2-16)
Wy — We
can be used to estimate the in situ state. If /, > 1, the soil may liquefy under a
shock (the natural water content is above the liquid limit). A sudden shock may
be the result of pile driving or even heavy equipment operating in the vicinity.
2 | No further
© | reduction
| g
$
Be Ww Hod”
o
Water content w, percent Water content w, percent
(2) (8)
Figure 2-2 (a) Graphical significance of liquid and plastic limits; (6) no further reduction in void ratio
occurs at water contents below shrinkage limit ws20 FOUNDATION ANALYSIS AND DESIGN
Specific Gravity G,
Specific gravity G, may be determined in a laboratory test with moderate diffi-
culty. This test is not often performed, since the specific gravity does not have a
wide range of values for most soils, as indicated below.
Soil G
Gravel 2.65 2.68
Sand 2.65-2.68
Silt, inorganic 2.62-2.68
Clay, organic -2.58-2.65
Clay, inorganic 2.68-2.75
A value of 2.67 is commonly used for cohesionless soils and a value of 2.68 to
2.72 for inorganic cohesive soils. The specific gravity is needed to compute the
void ratio in consolidation studies and sometimes to compute unit weights.
Unit Weight 7
Unit weight » is fairly easy to evaluate for a cohesive soil by trimming a block
and placing it in a container of known volume and measuring the water necess-
ary to fill the container. If the work is done rapidly so that the sample does not
absorb water, a reliable value can be obtained—if the average of several trials is
used.
It is very difficult to determine the in situ unit weight of cohesionless depo-
sits, since undisturbed samples are difficult to recover. Estimated values of unit
weight are commonly used for these materials based on field exploration tests (see
Chap. 3). Tables 3-2 and 3-3 give a range of values for unit weight which may be
used in the absence of laboratory data. It should be evident that a “loose” soil
has a smaller unit weight than a “dense” soil.
Unit weight is necessary to compute the present overburden pressure for
consolidation studies and to estimate the net increase in bearing pressure which
can be allowed at some depth in the ground. It is also needed to compute lateral
Pressures against retaining structures and to make estimates of skin resistance for
piles.
Relative Density
Relative density is sometimes used in cohesionless soils to describe the state
condition. Relative density may be defined
Cn = Cmin
e,
D, (2-17)
Cmax — EminSOIL MECHANICS IN FOUNDATION ENGINEERING 21
(2-18)
Yomax ~ Fonin Yn
where ay
oid ratio and unit weight of soil in loosest state
void ratio and unit weight of soil in densest state
existing void ratio and unit weight
Crass eae
Carin
Holtz (1973) reports that relative density determinations can be made on all
gravelly soils with fines (—No. 200 sieve) less than 8 percent, The fines limitation
for sandy soils is between 12 and 16 percent.
The relative density D, is used by some agencies to identify potential liquefac-
tion under earthquake or other shock-type loads [Durham and Townsend (1973),
Finn et al. (1971)].
It is the author's opinion that the D, test (or criterion) is almost worthless
while using a unit weight value is a good quality-control criterion and practical
The reason is that it is very difficult to determine the three unit weights required
for Eq, (2-18). For example, a coarse, well-graded sand (SW) was tested in the
author's laboratory (with the sand replaced and remixed each time) a large
number of times, yielding the following
max = 120 pef — 3 pef
97 pef +3 pef
The maximum unit weight was 120 pef; the — 3 pcf means the average value
from all tests was 117 pef, Values have been rounded to | pef for simplicity. If we
take the existing value y, = 110 + 2 pef, what is the range of D,?
_ 108 100 (117) _ 951
117 — 100 (108)
A second possibility is
112 ~ 100 (117)
"117 ~ 100 (112)
= 0.74
These computations give a range of D, of 74 — 51 = 23 percent.
Chapter 3 gives a method of indirectly determining D, in situ. Bowles (1978)
in experiment 18 illustrates a simple laboratory procedure to either compute D,
or obtain a unit weight for quality control.
Water Content w
This is a routine laboratory determination as part of the Atterberg limits, com-
paction tests, compression tests, shear tests, etc. If it is made on soils obtained
from exploration borings, it is the natural moisture content wy.
The natural moisture content wy for a saturated soil is useful in predicting
the water table location and the location of a perched water table, This value can22 FOUNDATION ANALYSIS AND DESIGN
be used together with the Atterberg limits (wp and w,,) to estimate if the soil is
preconsolidated. A soil which has been preconsolidated is at a higher density
(lower void volume) than a normally consolidated soil. In normally consolidated
inorganic clays the liquid limit is generally close to the natural water content. In
overconsolidated material which has a higher density, the wy value must of
necessity be less and so, if closer to the wp than to w,, is an indication of
preconsolidation. It necessarily follows that the closer wy is to wp the larger the
OCR is in qualitative terms. This procedure should be used with discretion,
however, since many marine deposits and deposits underlying marsh areas
exhibit large OCRs in the upper zoncs with wy near or even exceeding w;, [sce
Ladd and Foott (1974) and Koutsoftas (1980)].
Grain Size
The grain-size-distribution test (see Fig. 2-3) is used for soil classification, al-
though for much foundation analysis the soil classification is based on a visual
inspection, supplemented by methods indicated later. The engineer is usually
concerned only with whether the soil is gravel, sand, silt, or clay, or sandy clay,
silty sand, etc, where the second term is the predominant material and the first
indicates the filler; for example, sandy clay is a clay with some sand present. This,
type of visual classification can, with some exercise, be determined quite reliably
by one of the following procedures or tests:
1. To differentiate between gravel and sand, samples of each type of material can
be prepared to include jars of fine, medium, and coarse sizes if the material is
to be further subdivided and can be kept in the laboratory. The engineer
simply makes a visual comparison.
To differentiate between fine sand and silt, both materials may appear as dust
when dry. By placing a spoonful of the soil in a test tube of water and shaking,
sand or silt can be detected, since the sand settles out in 1} min or less,
whereas the silt takes 5 or more minutes to settle (ie. water clears). One can
observe relative thickness of the sediments for subclassification, as silty sand,
ete.
3. To differentiate between siit and clay:
a. One moistens a spot on a soil lump and rubs a finger on it. If the rubbed
spot appears smooth, the material is clay, but if it appears scratched, it is
silt or silty.
b. The dispersion test is made by mixing the material in water in a test tube
and observing the time for the water to clear. Silt usually takes 10 min or
less, whereas clay may take several hours or more.
c. Mix a small quantity of soil with water to form a plastic ball and place it in
the palm of the hand and shake horizontally with jerking movements. If the
material is silt or predominantly silt, the surface will become wet and shiny
since water travels through silt particles relatively easily and the inertiaSOIL MECHANICS IN FOUNDATION ENGINEERING 23
us. British (BS.) German DIN French
Sieve no. mm Sieve no. mm_ Sieve no. mm__ Sieve no. mm
4 4% = = =
10* 2.00 Bt 2.057 - 344 2.000
20 084116 1.003 — 31, 1,000
30-0595 300.500 500 0.500 280.500
36+ 0.422 400+ = 0.400.274 ~—0.400
4+ 0.4200 — ~ _
500.297, 520.298 =
60 0.250 600.251 250.280 250,250
80 0.177 850.178 160 0.160 230.160
100 0.149 100.182 125 0.12522 0128,
200 0.074 «200 0.076 = 800.080 200.080
270 0.053.300 0.053, 50, 0.080 180.050.
* Breakpoint between sand and gravel.
+ Use for Atterberg Limits.
(a)
Ss
8
8
x Sieve analysis
| © Hydrometer
Poorly graded | a
(uniform) ed AN
Dast | Dasl (12. | |) eee |
Ole Ted tree J
10 1.0 Dis Op 0.) 0.01
Grain size, mm
(o)
Well-graded soil
$
Percent finer by weight
o
g
8
Figure 2-3 (a) Various standard sieve numbers and screen openings; (b) grain-size-distribution curves.24 FOUNDATION ANALYSIS AND DESIGN
forces cause the water to move to the surface. Clay, on the other hand,
shows no change.
d. Crushing of dry clay lumps is relatively difficult, whereas silt lumps break
quite easily.
e. Clay can be rolled out into small threads, whereas silt is much more diffi-
cult to roll into small threads and generally requires much more water,
Grain size is of considerable importance in seepage and soil-drainage prob-
lems.
2-6 SOIL CLASSIFICATION IN FOUNDATION DESIGN
It is necessary for the foundation engineer to classify the site soils for use as a
foundation for several reasons
1. To be able to use the data base of others in predicting forundation per-
formance
2. To build the geotechnical engineer's data base in the
design
3. To maintain a permanent record which can be understood by others should
problems later develop and outside parties be required to investigate the orig-
inal design.
art” application of
Normally the Unified Soil Classification system (Table 2-1) with slight modi-
fications is used in foundation design work. For example, in much foundation
work it is academic whether a sand is well or poorly graded, but its density and
the presence of gravel are of considerable interest. Whether a fine-grained cohe-
sive soil is actually a clayey silt rather than a silty clay is not as important as
identifying its strength and settlement characteristics.
In foundation work the following terms are commonly used for cohesionless
soil deposits:
Loose
Medium
Dense
Table 3-2 gives additional subdivisions with quantifications based on using field
exploration data. Noting that these terms are both subjective and comparative,
one should avoid using an excessive number of subdivisions. Similar subjective
terms used for cohesive soils are:
Very soft
Soft
Medium
Stiff
Hard‘SOIL MECHANICS IN FOUNDATION ENGINEERING 25,
Table 3-3 gives additional subdivisions but, again, with the subjective nature of
these classifications a large number of subdivisons should be avoided.
The foundation engineer relies primarily on a written description of the soil
for assistance. Where a soil may simply classify as a SM or CH in thé Unified soil
system of Table 2-1, the geotechnical engineer would say “Reddish-brown, dense,
silty sand, low plasticity, SM” or “Gray-blue stiff clay with trace of sand, high
plasticity, CH.” The coloring and other distinguishing features such as “dense,”
“stiff,” and “trace of sand” are self-explanatory, but note the ease of cross refer-
encing these da’ to the next job the engineer might have in this area where the
same type of material is encountered. The terms SM and CH would not convey
much information alone, and it is altogether possible that another SM or CH
could exist in the same boring at a different depth with very different foundation
design properties (ic., loose, gravelly, very dense, less silty, soft or hard, etc).
Other terms given in the next section may be used in the identification and
classification of the soil at a site for foundation suitability and in preparing a
report to the client on general findings and recommendations.
2-7 SOIL CLASSIFICATION TERMS
Identifying names, some of which are local in nature, are assigned to certain sizes
or types of rock or soil formations, of which some of the principal ones are as
follows.
1, Bedrock. Rock in its native location, usually extending greatly both hor-
izontally and vertically. This material is generally overlain by soil of varying
depths. If exposed, the outer portions may become weathered. Bedrock varies
from igneous rocks, generally the hardest and formed from molten magma, to
metamorphic rocks formed from metamorphizing sedimentary rocks under
great heat and pressure, to sedimentary rocks formed as a combination of
chemical action and pressure from overlying soil deposits. Rocks may be solid,
but the interface with the overlying soil may be much fractured and eroded
and may contain voids from several weathering processes. Depending on the
geologic history of the area, the rocks may be much fractured, folded, and
faulted. Various textbooks on geology should be consulted for further infor-
mation, especially concerning particular areas (¢.g., Legget (1962), Thornbury
(1965), both with extensive references].
Boulders. Smaller pieces of material which have broken away from the bed-
tock, usually 250 to 300 mm or more in dimension (Table 2-2). Pieces smaller
than boulders may be called cobbles (50 to 75 mm minimum size) or pebbles (3
to 5mm minimum size).
Gravel. Common term used to describe pieces of rock from about 150 mm
maximum to less than 5 mm dimensions. May be crushed stone when manufac-
tured, bank-run gravel when excavated from a naturally occurring deposit and
containing finer material, or pea gravel if it has been screened to sizes 5 to 3
mm (pea size). Gravel is a cohesionless material; that is, it does not possess
particle adhesion or attraction.| ¢ weqa savess8 yew ouey | sunita os | Esk 2
stoquacs “V._ 2x92 suit} Biaqiony : Aeja-pues “spurs axe.) eee Eo
enp jo asn Businbas | | ¢ _ fis! 42
s2seo auiji9pog 23% | pue £ ‘ f gee
wowing “1 uu 2002 pum a2] 38 og ¥ zg | oun s(228) 22 F
paspegur sums sma | nog sua wy |g E as-puesspurs A$ eit =
Ba q — goz g
—— Bs z ces) sé
a2 3 | a : 1 at 22° 8
ue = 20 as é fa
soy siawunbas uonspesd ye Sunzow ion | 3S z 4 ec 4
oo ee = | ‘spurs pape K00d | Bg 2s g
Bs 3 es = 25 eel wg
— 4 83 28 55 BE 23
Be. uas [ sauy ou so | | es 3 Se
axa 5 lan 2egze2 | spues <|aae8 MS vz z= 25
at = "pose soreat,/ =") | 22322 | = 28
ee era) a a5g2%e “spurs papest-1i2q 2 \3
El Eeze somata Sepp-pues Cage oe Fe
{wey saieai8 “7 yum aur, segeoo8 ; pu ives Pus fea 5
|" va snoge say] Boquny | 2 E = anes “snes Aake| eee ga
syoquiss yenp jo asn asa | peo, ses
Buamnbas saves autpzop40q gael? - E sez, 22. 3
aur ¢ pur p UsaMi9q g228 £ sounxtut | wo | <2 ge a
#) yum aut... x0q¥ 283 gee ses jones sins | GO * 28s el
gEge cs Sess
zSa528 | Oo se)
15 10} siuawasinbes uonepesd ge eaks soanistur pues-]oves8 a | ge co =
ce Eggdie | ewe pape A004 ge 25
so ee? ee ac 2S
“ax Md», a | “soumnxtur pues.yavet wo | Ue =
yur | wade — 4 = 75 ty ueys sareaid yg =") =
eee ed a | | 2ne8 papess-9Mm, | =
—_ a : SOWEU [POI | yu ks
e1ian49 uoreayissp}> €s01es04e7, ouu ppords | — | aa
Keren) apuBsdese| wonBoYy!sse[D [9S PaMU) 1-7 *N4BLsopuig sep yim ammintu puESax"I8 papEs8-198 \IO-MO.
a{dwexs 104 “sjoquis dno jo suonwuqwos iq parvutlsap are sdnoi# om Jo sonstdioRIP\D BuIssassod sflos Jo) pan “suoHEayISseE|D aUNZapIOg 4
8 UrY) sayeasd st 7H way pasn Am xyINs +553] 30 9 st #7 ay) PUR 889] 10 8 SI
Uuoy pasn p xyns synuH Sunquany Lo paseg st UOISIAIPGns “ 400 KPe
30
bedrock
Figure 2-4 Typical soil profiles at locations indicated. Values for soil properties indicate order of
magnitude —not to be used for design.
found along the Mississippi River, where damp air rising affects the density
of the air transporting the material, causing it to deposit out. Such deposits
are not, however, confined to the Mississippi Valley. Large areas of Ne-
braska, Iowa, Ilinois, and Indiana are covered by loess deposits. Large
areas of China and Russia (Siberia) and some areas of Europe are covered
with loess deposits. Loess is considered to be a transported soil.
h. Muck. A thin watery mixture of soil and organic material
i. Peat. Partly decayed organic matter; may be contaminated with soil
8. Other terms used in soil classification
a. Alluial deposits. Soil deposits formed by sedimentation of soil particles
from flowing water; may be lake deposits if found in lake beds; deltas at the
mouths of rivers; marine deposits if deposited through saltwater along and
on the continental shelf. Alluvial deposits are found worldwide. For exam-
ple, New Orleans, La., is located on a delta deposit. The low countries of
Holland and Belgium are founded on alluvial deposits from the Rhine
River exiting into the North Sea. Lake deposits are found around and
beneath the Great Lakes area of the United States. Large areas of the
Atlantic coastal plain, including the eastern parts of Maryland, Virginia, the
Carolinas, and the eastern part and most of South Georgia, Florida, South
Alabama, Mississippi, Louisiana, and Texas consist of alluvial deposits.
These deposits formed when much of this land was covered with the seas.SOIL. MECHANICS IN FOUNDATION ENGINEERING 31
Later upheavals such as that forming the Appalachian mountains have
exposed this material. Alluvial deposits are fine-grained materials, generally
silt-clay mixtures, silts, or clays and fine to medium sands. If the sand and
clay layers alternate, the deposit is a varved clay. Alluvial deposits are
usually soft and highly compressible.
Black cotton soils. Semitropical soils found in areas where the annual rain-
fall is 00 to 750 mm. They range from black to dark gray. They tend to
become hard with very large cracks (large-volume-change soils) when dry
and very soft and spongy when wet. These soils are found in large areas of
Australia, India, and Southeast Asia.
. Laterites. Another name for residual soils found in tropical areas with
heavy rainfalls. These soils are typically bright red to reddish brown in
color. These soils are formed initially by weathering of igneous rocks with
the subsequent leaching and chemical erosion due to the high temperature
and rainfall, The collodial silica is leaching downward, leaving behind
aluminum and iron which become highly oxidized and are relatively insol-
uble in the high-pH environment (greater than 7). Well-developed laterite
soils are generally porous and relatively incompressible. Lateritic soils are
found in Alabama, Georgia, South Carolina, many of the Caribbean is-
lands, large areas of Central and South America, and parts of India, South-
east Asia, and Africa.
|. Residual soil. Soil formed in place by mechanical and chemical weathering
of rocks. This soil is found over much of the Eastern part of the United
States cast of the Appalachian mountains. It is also found over much of the
Southeastern United States and large areas located in the Ozark and
Rocky mountains. It is also found on most islands of the world, in large
areas of South America, Australia, and parts of Europe.
Saprolite. Still another name for residual soils formed from weathered rock.
Soil is often characterized by soil particles to large angular stones in the
soil deposit. Check the context of use to see if the term is being used to
describe laterite soils or residual soils.
Shale. A fine-grained, sedimentary rock composed essentially of com-
pressed and/or cemented clay particles. It is usually laminated from the
general parallel orientation of the clay particles as distinct from claystone
or siltstone, which are indurated deposits of random particle orientation.
According to Underwood (1967), shale is the predominant sedimentary
rock in the earth’s crust. It is often misclassified; layered sedimentary rocks
of quartz or argillaceous materials such as argillite are not shale. Shale may
be grouped as (1) compaction shale, and (2) cemented (rock) shale. The
compaction shale is a transition material from soil to rock and can be
excavated with modern earth-excavation equipment. Cemented shale can
sometimes be excavated with excavation equipment but more generally
requires blasting. Compaction shales have been formed by consolidation
pressure and very little cementing action. Cemented shales are formed by a
combination of cementing and consolidation pressure. They tend to ring32. FOUNDATION ANALYSIS AND DESIGN
when struck by a hammer, do not slake in water, and have the general
characteristics of good rock. Compaction shales, being of an intermediate
quality, will generally soften and expand upon exposure to weathering
when excavations are opened. Shales may be clayey, silty, or sandy if the
composition is predominantly clay, silt, or sand, respectively. Dry unit
weight of shale may range from about 12.5 kN/m? for poor-quality com-
paction shale to 25.1 kN/m? for high-quality cemented shale.
Figure 2-3a displays the most commonly used numbered sieves and their
openings in millimeters. Figure 2-3b displays typical grain-size distribution curves
for a well-graded and a poorly graded soil. Note that it was necessary to use a
hydrometer analysis to obtain the grain sizes smaller than 0.074 mm. A sieve
analysis together with the Atterberg limits would be used with Table 2-1 to
classify a soil
2-8 IN SITU STRESSES AND K, CONDITIONS
The K, soil state was defined in Sec. 2-4 as being an equilibrium steady-state
condition of zero lateral and vertical strains
A soil state with no horizontal and vertical strains has no shearing stresses
on these planes and the normal stresses are principal stresses. Any new stress
conditions imposed on the soil element use the K, stress ratio as a starting point
(see Fig. 2-5).
It is extremely difficult to measure K, in situ; however, Chap. 3 presents
several methods which give results of varying degrees of accuracy. There are two
very good reasons why one should be able to ascertain K, with some degree of
accuracy.
Dp Pacis Bs
Figure 2-5 Qualitative representation for soil stresses for OCR = 1 (normally consolidated, ne and
OCR > 1, Note that the limiting equilibrium for a normally consolidated soil is o, < 04:SOIL MECHANICS IN FOUNDATION ENGINEERING 33
1. This is the in situ reference state for changes in load conditions.
2. The value is needed for laboratory testing to reproduce the in situ stress state
for the triaxial test.
Several investigators have proposed approximations for K,. One of the ear-
st, and probably the most widely used, is that of Jaky (1948) for normally
consolidated cohesionless soils (and for agricultural grains stored in bins and
silos):
== sin Fy 4 2/3 sin 6) (2-19)
1 +sin ¢
which has been simplified to
I= sing! 2-20)
where 6 = effective angle of internal friction of the material.
For normally consolidated clay Brooker and Ireland (1965) proposed
K, = 0.95 - sin ¢ (2-21)
Alpan (1967) relates K, for normally consolidated clays to the plasticity index I,
(in percent) as
K, = 0.19 + 0.233 log I, (2-22)
Experimental (and intuitive) evidence indicates that K,, increases with the OCR,
resulting in the following approximation:
K,,ocr = Kono * OCR" (2-23)
The value of the exponent n may be obtained from Fig. 2-6 for sand. The
0.50 - po
0.48
0.40
36-8 O~~«C«D~~SC~—b Figure 2-6 Exponent n for sands,
¥ [Afier Alpan (1967)34 FOUNDATION ANALYSIS AND DESIGN
value of n for cohesive soils can be obtained from the author's rearrangement of
Alpan’s (1967) plot to obtain
n = 0.54 x 107/282 (2-24)
More recently Wroth (1975) reanalyzed a number of soils reported in the
literature by others and proposed as a best fit
K, = OCR x Kaye — 7 (OCR = 1) (2-25)
1-p
where p’ = Poisson's ratio in terms of effective stresses. Wrath (1975) also pro-
posed a curve for y’ versus [, (in percent) for lightly (OCR < 5) overconsolidated
soils which has been reduced to equation form by the author to give
u' = 0.23 + 0.0031, (2-26)
There is a unique relationship between Poisson's ratio and K, for normally
consolidated soils, From Hooke’s generalized stress-strain relationship of Eq.
(2-52) and taking the strain , = 0.0, we obtain
62 — fo, — to, =O
Let 6, = 63 = K,o, and solving for K, obtain
Thus if Poisson’s ratio «is known, we should be able to estimate K,, or estimate
iif we have an approximation for K,.
From a survey of the literature, several general observations about K, can be
made:
K, tends to be higher for finer than for coarser soils.
K, will be larger for loose cohesionless soils and smaller ¢".
K, tends to decrease with an increase in overburden pressure p,
K, is larger when the soil is overconsolidated
Bene
Example 2-2 Compare K, by the several approximate methods given in this section both for a
normally consolidated clay and for the clay with OCR = 5.
Other data o=28
1, = 20 percent
Souvmion By Brooker and Ireland's Eq. (2-21),
= 095 — sin 25° = 0.53
By Eq. (2-22),
19 + 0.233 log 20 = 049
When the OCR = 5, the values of K, may be:
By Wroth’s Eq. (2-24), estimate
u! = 0.23 + 0.00320) = 0.29SOIL MECHANICS IN FOUNDATION ENGINEERING 35
and obtain
9
55 (5 ~ 1) = 245 ~ 1.63 = 082
0
(049) —
KOsP) 1-029
By Alpan’s equation, obtain m by Eg, (2-24) as
n= 0.54 x 107707! = 0.46
Inserting this value of n into Eq. (
22), obtain
K, = 0445)4° = 1.03
‘The user must now apply sume engineering judgment to decide what value to use for K,.
Probably K, = 0.5 is satisfactory for the normally consolidated condition. In the OCR state it is
reasonable to assume that K, is in the vicinity of 1.0, and taking into account that the OCR
‘cannot be precisely determined and 1, cannot be determined to a routine accuracy much better
than 20 4, a possible best estimate of K, is
1.03 + 082
093
My
2-9 SOIL WATER—SOIL HYDRAULICS
Water in soil adversely affects cohesive soils by reducing the cohesion by soften-
ing, and may cause bulking of cohesionless soils depending on the amount of
water present. If sufficient water is present to develop pore pressure [changes in
water height in a piezometer tube (see Fig. 2-7a)] there may be a marked re-
duction in the o' tan ¢’ component of the shear strength.
Permeability is the facility for water flow through a soil mass. It is a major
factor in soil drainage, well water supply, and construction dewatering.
All natural soil deposits contain free water in their voids. After prolonged dry
periods the amount of water may be quite small near the ground surface; how-
ever, immediately after a rain the voids may be nearly filled. In their upper zone
the natural water content wy and soil strength are transient phenomena.
At nearly all points below the soil mantle there is a zone of flowing water
called the water table, The soil below this point is saturated; however, individual
samples may have trapped air bubbles and produce S slightly less than 100
percent. The water in this zone is flowing under a hydraulic gradient from a
higher to lower energy level. The water level in a series of piezometers inserted
along the direction of flow will define the hydraulic grade line. Above the water
table is a capillary zone (S—> 100 percent) where the voids are also nearly filled
with water. This water is held in place by surface tension between water mole-
cules and soil grains and is not free to move. The water in this zone produces an
increase in the effective weight of the soil. The depth of water below the water
table produces a buoyant effect on the submerged soil.
Very considerable adverse effects on the soil can be produced by the pore
pressure in the free water zone. Note, however, the water either in the capillary
vone or from other sources may become a flow zone if the existing void ratio is36 FOUNDATION ANALYSIS AND DESIGN
W,= surchorge, pst
7
allt 2 t (S=] hak
3 Ge FS>l A>
+ Copitor le
Ele artorecee | cone he. PS<1 &<&
£ | 1 = = actual tydroulic
“Hoy | 5-100 percent } | gradient
A tera cial
lo)
te)
Figure 2-7 (a) Soil/water relationships for effective and pore-pressure concepts: (b) critical hydraulic
[gradient und concept of tafety factor against a “quick” condition: (c) computation of height of
capillary rise in a capillary tube.
sufficiently reduced that there is an excess of water for the remaining voids
Water cannot flow instantaneously; so any excess pore water will exist for some
time under a higher energy potential (or pressure).
Effective Pressure
The pore water (also called neutral) pressure reduces the total pressure on a plane
to an effective value. Effective (or intergranular) pressure is defined as the pres-
sure between the individual grains in a mass. This pressure and friction coefficient
produces a portion of the shear strength of cohesive soils and all the shear
strength of cohesionless soils. The nominal effective pressure is based on total
load area since the actual grain contact area is indeterminate.
Referring to Fig. 2-7a and neglecting any shear resistance along the sides of
the unit area, one can obtain the nominal effective pressure o’ at the surface of
the water table as
a
yihy +y2hy=q psf or kN/m (a)
Note that the capillary zone water directly affects 2 and “wet” unit weights
are used for both y, and 73.
The effective pressure at point 4 (below the water table) can be computedSOIL MECHANICS IN FOUNDATION ENGINEERING 37
using Archimedes’ principle as
o =a t yshs — Ywhs ()
It is usual practice to define the submerged or buoyant unit weight »’ = Ysa
~ Fs SO that Eq. (6) with }.,, = 73 is rewritten
qt yhy (o)
We can obtain a general expression relating the effective pressure to the pore
pressure using Eq. (b) and taking yh = «
q+ 7shy=o'+u (d)
o=o'+u
where the total pressure ¢ = q + 7's hy
Rearranging and solving for the effective stress,
o=o-
(2-27)
Equation (2-27) can be rewritten to include the effects of changes in was
—(u+ Au) (2-274)
Equation (2-27a) indicates that in increasing the neutral pressure (+ Au) the
cllective pressure is decreased by the same amount. If the neutral pressure is
sufficiently increased, the effective pressure reduces to zero; ie. the soil will, if
granular, possess no shear strength. This condition is referred to as a “quick”
condition, and conditions for its occurrence may be approximately evaluated as
follows.
From Fig. 2-7b, equating the upward and downward pressures at point 4,
(hy — Lyyy + Lay + Wy (downward) = (h + hz yyw (upward) (e
For F = 1 and with the surcharge W, =0 in Eq, (e) the hydraulic gradient
hvL is a critical value i, . For this case
(2-28)
G, for sand is approximately 2.65, and for the practical ranges in e from 0.3
to 1.0, i, ranges from 0.8 to 1.25. Considering the several uncertainties involved, it
is common to approximate i, = 1.0.
Permeability
Flow of soil water, for nonturbulent conditions, has been expressed by Darcy as
ki (2-29)
v38 FOUNDATION ANALYSIS AND DESIGN
where i = hydraulic gradient h;
k
as previously defined
coefficient of permeability as proposed by Darcy, length/time
Table 2-3 lists typical order-of-magnitude values for various soils. The quan-
tity of flow q is
q=kiA — volume/time
Two tests commonly used in the laboratory to determine k are the constant-
head and falling-head methods. Figure 2-8 gives the schematic diagrams and the
equations used for computing k. The falling-head test is usually used for k < 10°
m/s (cohesive soils) and the constant head for cohesionless soils.
Capillary Water
Capillary rise in a soil can be estimated from the equation for h, shown on Fig.
2-Te.
This equation generally overestimates the height of capillary rise consider-
ably because the soil pore system is not regular. Few laboratory observations of
capillary rise have been found to exceed 1 or 2m.
Flow Nets
The flow of water through soil under an energy potential can be mathematically
expressed by a Laplace equation as
eh, @h
stk =0
pet haz
&
oy
where k,, ky = coefficients of permeability parallel to the x, y axes, respectively
A = energy potential
The above equation is for two-dimensional flow, which with appropriate axis
rotation will apply to most seepage problems. A graphical solution of this equa-
Table 2-3 Order-of-magnitude values for permeability k, based on description of
soil and by Unified Classification, m/s
10° 10-7 10-* 10° 10%
Clean gravel Clean gravel and Sand-silt Clays
GW, GP sand mixtures mixtures |
GW, GP SM, SL, SC
SW, SP
GMSOIL MECHANICS IN FOUNDATION ENGINEERING 39
ch
oe
FTL area of
he tube=a
bain —
=m :
* Abr “ <
PE soi
| somele Seil
| | Aeora Area bet 5,
i / collected
oA Feo @ intime + ae ¥ to find @
lo (6)
Figure 2-8 Schematic for permeability determination. (a) Constant-heat permeameter; (6) falling-heat
permeameter.
tion results in families of intersecting orthogonal curves which are called a flow
net. One set of the curves represents equipotential lines (lines of constant piezo-
metric head) and the other set intersecting at right angles represents flow paths.
The flow net consists of squares of varying dimension if k, =k, and rectangles
otherwise. In general, for reasonably homogeneous soil a graphical solution of
the Laplace equation provides seepage quantities which are at least as correct as
one is likely to obtain for the coefficients of permeability.
Seepage quantity from a flow net can be computed as
Q=kH z= Wt — (ft or m3 in time ¢) (2-30)
a
where k = transformed coefficient of permeability when k, #k, and so the re-
sulting flow net consists of squares, k = \/k,k, in units of H and t
H = differential heat of fluid across system
ny, ng = numbers of flow paths and equipotential drops, respectively, in system
W = width of seepage flow
1 = time base (1 hour, I day, 1 week, ete.)
Figure 2-9a illustrates a flow net for one side of a cofferdam-type structure
which will be of most interest in this text. We may use the flow net to estimate
how much drawdown may be allowed on the construction side of the wall or how
much excavation can be performed before the construction side becomes “quick.”
For other seepage problems the user is referred to any text on soil mechanics
leg., Wu (1976), Bowles (1979)].40 FOUNDATION ANALYSIS AND DESIGN
7 impervious
(a
= Bx 1 WENA 21 (86400)= 24 5 mY day mm
rr
Figure 2-9 Typical flow nets as used for sheet-pile or cofferdam structures, (a) Single sheet-pile wall
‘oF other wall too far to influence net; (b) double-wall cofferdam as used for bridge piers, ete.
Example 2-3 From Fig. 2-9a assume the following data:
H=60m
4x 10-$ mys
Your = 19.80 KN/m? (sand)
Distances: AB = 2m, BC
m,CD=15m,DE=1m
Requinep (a) Flow quantity/day per meter of wall
(8) Effective pressure at point C
SoLUTION (a) Flow quantity (estimate ny = 4.1), Also with tailwater at the dredge line
H=6+2=8m
Q=kH e wt = 4 x 10" eT ea 400) = 14.2 m°jdaySOIL MECHANICS IN FOUNDATION ENGINEERING 41
(b) Effective pressure at C.
Total pressure at C: ¢ = (19.8) = 39.6 kPa
Static pore pressure at C= u = 29,807
Enoess pore pressure at C ~ Au ~ #18X9.807) = 9.81
(u + Au) = 39.6 ~ 29.4 = 102 kPa
e
Since o' > 0, the soil is not “quick.”
(il
2-10 CONSOLIDATION PRINCIPLES
The primary and secondary consolidation settlements of saturated cohesive soils
are generally estimated using consolidation theory. A consolidation test is per-
formed to obtain a compression parameter for the amount of settlement and a
consolidation parameter for the settlement rate estimate. The overconsolidation
ratio OCR can also be determined from this test. The test is performed on an
“undisturbed” sample which is placed in a consolidation ring available in diam-
eters ranging from 45 to 115 mm (1.8 to 4.5 in). The sample height is between 20
and 30 mm (0.75 and 1.5 in); 20 mm is the most commonly used thickness to
reduce test time. The larger-diameter samples give better parameters, since about
the same amount of disturbance is developed for any size sample, with the
relative effects less for the larger samples. The most common test diameter is 64
mm (2.5 in), since this best balances the costs of sample recovery and disturbance
effects. Tube diameters larger than 76 mm may result in a premium charge for the
sample—particularly if a larger borehole must be made.
The consolidation test proceeds by applying a series of load increments to
the sample and recording sample deformation at selected time intervals. Sufficient
laboratory data should be obtained to allow computation of the water content
wy and the specific gravity so that the void ratio at any time interval can be
computed.
The load-deformation data are plotted on either a semilogarithmic plot or \/t
as illustrated in Fig. 2-10. The purpose of these plots is to obtain the time at some
percent consolidation, The fs value (time at 50 percent) is most commonly used,
and the procedure for obtaining it is to obtain the dial reading at 100 percent
consolidation Doo and the dial reading at the start of the test Do.
‘The procedure for finding Do is to either use the actual initial dial reading at
the start of the test or, if the initial curve is parabolic, find the apparent value.
This may be necessary, since one cannot plot log time at ¢ = 0. The corrected
value of Do is found as follows:
Select a time 1, in the parabolic portion.
Select a second time t, = 4t, in the parabolic part.
Obtain the offset between ¢; and ¢.
Plot this offset distance above t, to obtain Do.
Bia ee
The Doo value is obtained as the intersection of tangents drawn to thetuorsssudwos £swpuosas Jo 2124
st
780) saan g wos,
i 9 =087 yim
seeauo>
puget Se OF
B2=yf
osc'tma 10 1020 8%
‘uw 9¢ = 001;
eile =Sc-| secon ona
9 0002 Bat
Bolg’ pu 9)
jozc’z
love'z
joe
jond'e
zg
§
008’
ww 01 x Suipees |o1g
nso UE Jf *D pure ‘1 yI0g
96) Sunutergo soye payeurensar 189) pure Sunioyd yrIm ‘7 uNeIqo oF pasn yo}d aum /~
10 02 101d Bc
uy
9G (0) "EP OUI} Ias-auIN BUNUESeId Jo sPOYTeW OIE amndiy
yw ouy
oot vo
SSA Te + 0802
==} 19692 ;
ie ane
ye aaing 2
- loec'z ©
og: % \ a
ae jones
peeenaeg *
See oermcendl 5
fee : 23
Um ft 9 01X60 = a =? par
po0} 4o weues sy eae
pila a wwei="a ‘nes 2er'2=°0
| - . 082
42SOIL MECHANICS IN FOUNDATION ENGINEERING 43
midcurve area and the end portion as shown on Fig. 2-10a. Considerable in-
erpretation is necessary if the curve does not exhibit a pair of identifiable tangent
ones. Sometimes an increase in vertical scale will improve the tangent locations.
The Dso (or other percent consolidation value) is obtained from Do and
Doo. This value projected to the settlement curve allows one to obtain fs. from
the time axis
The v/t method is also used to obtain t59. This method involves plotting the
dial reading versus \/time with a straight line estimated for the first several
points, This line is extended to the abcissa, and a point 15 percent larger is
located (see Fig, 2-106). Through this point and the intercept of the ordinate axis
4 second straight line is drawn, When the dial reading versus \/time curve
intersects this second line, the Deo value is obtained, The apparent initial dial
reading Da is obtained where the initial straight-line part of the settlement curve
intersects the ordinate axis. With Doo and Do values it is easy to obtain Dso and
co Which is most commonly used. Values of ts should compare by the two
methods, but in real soils large differences are sometimes obtained. The ,/time
method is more rapid, since the test for that load increment can be terminated
when Dgo is found; however, if secondary compression is to be estimated, the
semilog plot should be used.
The Coefficient of Consolidation c,
The fo data are used to compute the coefficient of consolidation ¢, as
TH
&
(2-31)
where T; = time factor (see Table 2-4)
H = length of longest drainage path for a particle of water; in the labora-
tory it is the half sample thickness when drainage is from both faces
1, = time for i percent consolidation to take place—tso is usually used
Case | of Table 2-4 is usually assumed in the conventional laboratory test. For 50
percent consolidation Eq. (2-31) becomes
_0.197H?
tso
(2-31a)
The several load increments give separate values of c, which can be plotted
on the void ratio or strain versus log p curve shown in Fig. 2-11a. The plot is
usually very erratic because of changes in void ratio, temperature, and S. The
urve can be smoothed somewhat by using a small vertical scale—beyond this it
san exercise in engineering judgment to determine the value of c, to use for
stimating field settlements. The time for a given settlement to take place in the44 FOUNDATION ANALYSIS AND DESIGN
Table 2-4 Time factors for indicated pressure distribution
ue Case I Case IT
0 0.000 0.000
0 0.008 0.048
20 0031 0.090
30 oon os
40 0.126 0.207
50 0.197 0281
60 0.287 0.71
70 0.403 0.488.
80 0.567 0.682
90 O48 0933
100 x %
Cose I Coxe Io Cose
fuer pet pet
7% eee S|
' |
4 H - D
Pore-pressure distribution for case I Pore-pressure distribution
usually assumed for ease 1a for case II
[Taylor (1948)]
field is obtained as a rearrangement of Eq. (2-31) to obtain
ey a
ev ti
Primary Consolidation Settlements
Primary consolidation settlements occur during the time of an excess pore-
pressure gradient caused by a stress change in the stratum of interest. At the end
of primary consolidation the excess pore pressure is very nearly zero and the
stress change has gone from a total to an effective state. Additional settlements
termed secondary compression (or consolidation) continue for some additional
time. These will be considered in a later section.
‘The sharp break in the e versus log p or ¢ versus log p curve is used to
estimate whether the soil is preconsolidated. Casagrande (1936) proposed the
method shown in Fig. 2-I1a to determine a value of p, . Steps in the method are:
1. Determine by eye the sharpest curvature and draw a tangent.
2. Draw a horizontal line through the tangent point and bisect the angle a thus
producedSOIL MECHANICS IN FOUNDATION ENGINEERING 45
3. Extend the slope of the curve to intersect the a bisector.
4. This intersection point is at the preconsolidation pressure of the in situ sample.
The value of p, obtained from this plot is compared with the existing effective
overburden pressure p, and if:
1. p, © pe (+ about 10 percent), soil is normally consolidated
2. pp > p.. sample may have excessive disturbance.
3. Py < Pe. Soil is preconsolidated; OCR = p, /p,. Use p, to define soil state
The reader should note that p,, p. are effective stresses.
‘The amount of primary consolidation (settlement) is estimated from a com-
pression index C, or compression ratio C. which are obtained from the slopes of
the straight-line part of the plots shown in Fig. 2-11. If there is no “straight-line”
segment, engineering judgment must be applied to obtain an equivalent “slope”
(Fig. 2-12),
de Ac
log p2/P. “log p2/Ps
where the terms are identified in Fig. 2-11
The plot of void ratio (void ratio at the end of a local increment) versus log
pressure may require correction for sample disturbance. This correction can be
made approximately as follows (Fig. 2-I1a, Fig. 2-124)
1. Extend the straight-line portion until it intersects the void ratio abscissa at
about 0.4.
In some manner obtain the initial void ratio of the in situ soil. The void ratio
of the sample is a rebound value (too high) and, from the curve as shown on
Fig. 2-11, is too low. Use these boundary values with wy and G, to estimate
the true value.
In some manner determine the in situ overburden pressure p,.
4. At the intersection of p, and ¢, extend a straight line to intersect point located
in step 1
5. The slope of step 4 is the corrected value of C,
We may define the coefficient of compressibility as
(a)
(6)
(o)46 FOUNDATION ANALYSIS AND DESIGN
Toad
Reroided
C0710,
Co ago
Void roto,
038
Pressure, kPa
fal
Figure 2-11 Methods of presenting settlement data to obtain settlement parameters C, and C,
Method of correcting a normally consolidated clay [after Schmertmann (1955)] is shown on (a). Data
used to plot both curves are shown on (6), Slight discrepancy between C; values due to plotting.
Substituting Ae from Eq, (a), we obtain
4 Ap
nT
m, Ap
where m, = coefficient of volume compressibility:
a «1
Ive, Ap &,
me
From Fig. 2-13 the settlement AH is by proportion
A) ie
H l+e,
or AH =m, Ap H = eH
where H = total thickness of stratum and ¢ = average unit strain in H.
fe)
vy
(9)SOIL MECHANICS IN FOUNDATION ENGINEERING 47
= 280 kPa (by eye)
a ke BH,
pe,
the amount Ap’ = p, + Ap — p, is used together with p, and cither C, or C to
compute the virgin compression
The total AH is the sum of the computed settlements along the two branches
and should be significantly less than for a normally consolidated soil
Consolidation settlements predicted on the basis of the preceding several
‘equations have been found to be reliable if the laboratory samples were of good
quality. The predicted settlement rate is generally poor. The rate prediction may
be improved by monitoring settlements in the early stages and correcting the field
lime settlement curve using these data.
Primary Settlement Index Correlations
\ number of correlations exist for C, and C’, so that settlement estimates can be
made without the time and expense of performing laboratory consolidation tests.
The settlement rate and OCR cannot be obtained from correlations; however, in
many cases these are not needed.
An early and widely used correlation for inorganic clays of medium to low
sensitivity with about +30 percent error was that of Terzaghi and Peck (1967)
CC. = 0,009, ~ 10) (2-36)
where w, is in percent.
Azzouz et al. (1976) list a number of correlations based on a statistical
alysis of a very large number of soils. From a study of this list by the author
the following were selected:
C, £0.37(¢, + 0.0030, + 0.0004 wy ~ 0.34) (2:37)
which has a reported 86 percent reliability, and
Ce = 0.135(e + 0.01w, ~ 0.002w — 0.06) (2-38)
\shich has a 76 percent reliability.
Factors favoring these equations are the use of the in situ void ratio, liquid
mit, and specific gravity. The in situ void ratio can be computed as wy G,; so
‘here is a direct relationship between wy and e,. The use of both e, and w,, should
‘ncrease the reliability and offset the additional computing effort involved.
Secondary Consolidation
condary consolidation (also secondary compression or creep) is that settlement
utinuing beyond primary consolidation. Secondary compression can be the50 FOUNDATION ANALYSIS AND DESIGN
major settlement component in organic soils but is usually quite small for inor-
ganic soils. A new K, condition is obtained at the end of secondary compression.
The slope of the secondary branch (see Fig. 2-10a) of the settlement versus
log time curve has been observed to be approximately constant for a given soil.
Based on this observation it is only necessary to take one of the load increments
sufficiently long to establish the slope of this branch of the curve. The slope of
this. branch C, can be computed as
AH, /Hi
2-39
* log ta/ty aay
or, since AH/H, = Ae,
As
24
log t/t ea
and the secondary settlement estimate is
A
AH, = HC, log #7“ (2-41)
4
thickness of stratum
laboratory sample thickness
change in laboratory sample height from t, to tz
some time At after t
ty = some time after primary consolidation
Secondary settlement estimates tend to be rather poor but may be preferable
to simply guessing
2-11 SHEAR STRENGTH
Soil failure is a combination of particle rolling and sliding. This mobilizes the
shear strength of the soil as opposed to the compressive or tensile strength. The
shear strength involves the soil strength parameters of cohesion c and angle of
internal friction ¢. The shear strength in terms of total stresses is
s=ctotang (2-42)
In terms of effective stresses the shear strength is
s=c+o'tang’ (2-43)
where s = shear strength, ksf or kPa
¢ = soil cohesion, ksf, kPa
= angle of internal friction
@ = total normal stress on shear plane, ksf, kPaSOIL MECHANICS IN FOUNDATION ENGINEERING 51
The effective stress values are obtained when a’ = o — wis used as the normal
stress on the shear surface. Equation (2-42) defines the Mohr-Coulomb failure
envelope of Fig. 2-14. This envelope constitutes the limiting states of stress which
can be imposed on a soil. In general, since two parameters are involved, two or
more tests must be performed to evaluate the envelope equation, From Fig. 2-14a
one can obtain an expression for the normal stress on the plane of interest as,
ates o1-43
Oy = A + HS cos 20 (2-44)
From the figure geometry,
20=9 +6
¢
o=45+5
72
ent on oa a=
Normal etree @ =
1 CU airect sear tests (a) CU tests on OC soit
Luncontined compression test (el U tests on NC soil
vure 2-14 Mohr's circles and rupture envelopes for several shear-strength tests shown,52. FOUNDATION ANALYSIS AND DESIGN
The deviator stress is defined as a, — 04 since this stress difference is the
instant sample load in a triaxial test where the initial cell pressure is. A similar
analogy exists in situ when the stresses on a soil element change from foundation
loads. If Eqs. (2-42) and (2-44) are solved simultaneously and
aaz(as— ©) a
tan (45 2) Dean ¢
substituted, we obtain the following two widely used equations for the principal
stresses
0, =o; tan? (4s + ] + 2c tan (ss + $) (2-45)
63 = 6, tan? (4 = $) — 2c tan (4s — ) (2-46)
Shear strength of a soil is heavily dependent on the type of test, which may
be:
1. An unconsolidated-undrained (commonly “undrained”) or U test
2. A consolidated-undrained (CU) test
3. A consolidated-drained (CD) test
These tests can be made using direct shear or direct simple shear (sometimes
called plane strain) and triaxial equipment (see Fig. 2-15) on intact samples which
should be in the same K, consolidation state as the field. Some organizations
identify consolidated-undrained compression tests at the K, state as CK,UC
tests. There is some debate over whether the direct shear test is a plane-strain test
and/or how to obtain a test with one principal strain of zero. There is also some
debate over the merits of a given test [see state-of-art and summary by Saada et
al. (1980) with their many references].
Ihe undrained test 1s performed by loading the sample to failure with drain-
age outlets closed. A consolidated sample is obtained by applying a consolidation
stress to the sample with the drainage outlet open, When drainage halts and/or
no further volume change is measured, the drainage outlet is closed for an un-
drained test or left open for a drained test. At the end of consolidation the excess
pore pressure in the sample is zero. An excess pore pressure develops when
loading is applied during a consolidated-undrained test.
All three tests will give about the same strength parameters for dry soils;
however, plane-strain tests in dense sands tend to angles | to 4° larger than
triaxial values [Lee (1970)]. When water is present, the parameters depend
heavily on the type of test, the permeability characteristics of the soil, the amount
of water, and stress history. Generally the lowest compressive strength values are
obtained in U tests on saturated soils—in fact the shear strength is zero for both
U and CU tests on saturated sands unless the sample is undisturbed and has
measurable natural cementation, Only a cohesion parameter can be measured in
U tests for saturated clays.| MECHANICS IN FOUNDATION ENGINEERING 53
:
a
| ae
g b= rubber membrane,
Ha
k ©
son mes
tg Aen
(a)
Somple inside o
we reinforced.
fuboer membrone
Figure 2-18 Laboratory strength test equipment. (a) Triaxial test; (b) direct shear test; (c) direct
simple shear test. [Bjerrum and Landa (/966).]
Triaxial extension (CK, UE) tests are sometimes performed for important
projects. Extension tests are made by maintaining the lateral pressure and re-
ducing the axial pressure in increments, Extension test shear strengths tend to be
30 to 50 percent less than from compression tests at the same OCR. Direct simple
shear tests tend to produce shear strengths that are intermediate between CK,
UC and CK,UE tests, Figure 2-16 illustrates several field loadings and the
laboratory tests which best describe the soil strength for these conditions
The symbol s, is commonly used for the shear strength obtained from U
tests, In the particular case of triaxial tests on saturated cohesive soils this results
in an undrained shear strength defined as
Consolidated-undrained tests tend to produce small to nearly true @ angles
and a measured cohesion intercept depending on the type of soil and
degree of saturation. If pore-pressure measurements are taken, the effective par-
imeters g' and c’ can be computed. These tests are relatively expensive to per-
orm if the undisturbed sample is to be reconsolidated to the in situ K, state
without excessive damage to the soil structure. A number of small load in-
crements of up to 24 h duration each may be required. This may take 8 to 10
lays per sample for reconsolidation. A compression test at a low strain rate may
‘ake I to 2 days and an extension test twice this length of time.54 FOUNDATION ANALYSIS AND DESIGN
OQ}
Figure 2-16 Strength tests corresponding to field shear. [After Ladd (1977) and Johnson (1975). The
unconfined compression test is commonly used for all cases.
Consolidated-drained tests give the drained or effective ¢’ angle and cohesion
c. These tests require a strain rate that is so low that the strength is not
significantly affected by the small cxecss pore pressures which develop. This
required strain rate is very time-consuming and the test is very expensive to
perform. For these two reasons CD tests are seldom performed except for special
problems; the same information can be obtained from CU tests by measuring the
pore water pressure.
‘Other factors contributing to the shear strength of a soil include
The effective or intergranular pressure o' (contributes a’ tan d).
Interlocking of particles; angular particles give more interlocking and higher
strength than bank-run sand and gravel (affects # angle).
Particle packing (unit weight effect)
Particle attraction (clay minerals) and cementation (contributes cohesion).
Sample quality (disturbance, intact, fissured, etc.)
OCR (stress history).
Strain rate of test. High strain rates as in the unconfined compression test may
give strengths 20 to 30 percent higher than obtained from the in situ strain
rate.
8. Laboratory environment and technician skill.
Aaa ee
Isotropically consolidated triaxial tests are most commonly used in the lab-
oratory when a project justifies the expense. Isotropic consolidation occurs when
the sample is placed in a triaxial cell and subjected to some value of cell pressure
and with drainage permitted. This type of consolidation generally does not pro-
duce a sample at the end of consolidation which is the same as in situ because K,UNDATION ENGINEERING 55
conditions (0, <<,) are not reproduced. A special triaxial cell is required to
reproduce K,, conditions. It is probable that isotropic consolidation produces
better strength parameters than the unconfined compression test. It is also
probable that the isotropically consolidated parameters will not describe the field
performance as accurately as one would like. Anisotropically consolidated tri
axial and direct shear tests may be used for important projects. In these tests
typically the sample is K, consolidated (consolidated with zero lateral strain). The
resulting triaxial specimens are then tested in either compression or extension
tests depending on the estimated location of the sample with respect to field
failure (see Fig. 2-16).
Strength tests have principal significance on saturated cohesive soils; in par-
ticular the U test represents the “worst case” condition. Strength tests are rou-
tinely performed on nonsaturated soils as the soil borings progress from the
ground surface downward. These tests serve the purpose of giving a strength
indication of these soils at that water content. If the water table does not fluctu-
ate and rainfall is moderate, these tests produce suitable values for design. When
the soil can saturate (and the geotechnical consultant may not be able to control
this), the strength determination should be made on saturated samples. This is
particularly true for dam foundations but may be equally applicable for some
building foundations.
Normally Consolidated Clay ($~» 100 percent)
The U test on a normally consolidated clay gives
and for CU tests
ctotang
where @ is usually on the order of 3 to 10°; where $—» 100 percent.
Consolidated-drained tests (or CU tests with pore-pressure measurements) give
s=o' tang’
ie, the cohesion intercept is practically zero (see Fig. 2-17)
Overconsolidated Intact Clay (S -> 100 percent)
The U test gives a higher strength value than for the same clay in the normally
consolidated state, The difference includes the effects of both increased density
nd the negative pore pressures developed when the sample expands from loss of
overburden during sample recovery.
The CU test will give higher values if the cell pressure 0, 4*
since negative pore pressures will be present either from sample expansion or
from shear dilation on the failure plane. This strength may be unsafe for design
use since similar stressing in the field in the presence of water will result in the56 FOUNDATION ANALYSIS AND DESIGN
Figure 2-17 Qualitative rupture envelopes for three OCR ratios. All Mobr’s circles to produce
rupture line not shown, Note the initial branch of rupture line is usually curved for OCR > |
material softening, If 5
p,, the sample reacts to the application of devi-
ator stress as if the clay is normally consolidated, and in a CD test or CU test
with pore-pressure measurements we obtain
o tan ¢
Figure 2-17 illustrates the qualitative Mohr rupture envelopes for a clay at
several OCR values.
The s,/p’ Ratio
‘A number of researchers have found a nearly linear relationship between the
undrained shear strength s, and the effective overburden pressure p,. Normalized
behavior is obtained when a parameter of significance divided by another par-
ameter gives this relationship. Many clay soils exhibit normalized behavior—that
is, sy/p, is nearly a constant with some scatter attributed to normal soil hetero-
geneity and variations in water content. It appears, however, that quick or nat-
urally cemented clays with a high degree of structure do not exhibit normalized
behavior. Ladd et al. (1977) give an extensive discussion of this and other beha-
vior of clays.
‘The normalizing parameter for s, is the effective consolidation pressure p;
Usually this is taken as the existing effective overburden pressure, with the corre-
sponding laboratory value being the effective pressure used to consolidate the
sample to the test state, Figure 2-18 is a normalized plot based on an analysis of
several clays so that the s, value for OCR > 1 can be estimated. The clays used toSOIL, MECHANICS IN FOUNDATION ENGINEERING 57
Y |
Aj2 0 stu exaning overburden sve |
4 A :
Bae ihoes
aoe
Toes?
Bra oce®
a gh oe? i
a 4P aasacs™
i tromple Gen OCR StHeld)
: pe ON |
| T lyfe ORItob)
0 HP) |
Fe Ns wae |
I 8 Sone
f 12130" Baie
23a $6 6 7 6 9
OCR gle
Figure 2-18 Normalized plot 10 obtain the ratio of undrained shear strength at some pre~
consolidation pressure to a normally consolidated value for several OCR. [After Ladd et al. (1977)]
Note that above plot is based on direct simple shear (DSS) tests.
obtain this curve all had high wy and w,, but with five clays involved a trend
seems to be established so that the curve might be used for clays with lesser wy
and wy (less than 50 percent), In order to use this plot it is necessary to obtain an
undisturbed sample for consolidation testing. From the consolidation test obtain
p|. Remold a sample and K, consolidate it to the present in situ p, state. Obtain
s, from this sample. Now with s,, p, and p; one can compute the OCR, enter
Vig. 2-18, and obtain the ratio B/4. Since A is known, we can compute B and
using B compute s, at the present OCR as
Syoce = B(ps)
Several equations have been formulated by the author from curves for s,/p\,
s soil index properties for normally consolidated clays presented in the
Iuerature. A relationship between s,/p, and plasticity index 1, (Bjerrum and
Simons (1960)
“t= 0.45(1,)"? 1, > 5 percent (2-47)
Po
The scatter is on the order of +25 percent of the computed value,
This reference also gave a relationship between s,/p, and the liquidity index
1, . which in equation form is
O18)? 1, > 0.5 (2-48)
Pe
Ihe seatter is on the order of +30 percent. Karlsson and Viberg (1967) presented58 FOUNDATION ANALYSIS AND DESIGN
a curve relating s,/p, and the liquid limit, which in equation form is
wz > 20 percent (2-49)
The scatter is on the order of +30 percent. The s,/p, ratio may be computed
from the several equations and an average value used for preliminary design. If
large differences are obtained from the three equations, consideration should be
en to tests on undisturbed samples.
Unconfined Compression Strength 4,
This is a triaxial test with the confining (cell) pressure at zero gage pressure. In
practice, the triaxial cell is seldom used; rather devices are available to perform
this test directly after trimming the sample ends square and with a length such
that the L/d ratio is 2*. The Mohr's circle plot for this test is shown in Fig. 2-14c.
Three or more tests should be made with a best average q, obtained to draw a
Mohr's circle and give
Prior to about 1970 this test was always called the unconfined compression test.
Present tendency is to call this an undrained test, but in any case it is
common to use q, to denote the peak failure stress for this test.
A body of opinion holds that s, is too low from the unconfined compression
test. Values as much as 60 to 80 percent of the true value are commonly reported,
Since sample disturbance can account for a 20 to 50 percent strength reduction, it
would appear some compensating effects occur. These effects may be partly due
to the negative pore pressure which develops when a recovered sample loses
overburden and attempts to expand. Also the strain rate in the laboratory test is
higher than the ficld strain, with the result being a higher measured strength
value.
Current practice tends to a widespread use of the unconfined compression
test for all routine soil recommendations. Only when the strength is very low
and/or the project represents a considerable financial investment are more elab-
orate soil testing procedures used. We might note that U tests are used for most
foundation design work as they tend to reflect field loading more accurately.
Residual Strength
After soil failure occurs, there is a residual strength remaining in the remolded
material. The remolded material has negligible cohesion remaining but an angle
of internal friction @. This residual angle of internal friction can be defined as
s,=o tan 4, (2-50)SOIL MECHANICS IN FOUNDATION ENGINEERING 59
The residual angle of internal friction can be estimated using the residual devi-
ator stress to plot a reduced Mohr’s circle as illustrated in Fig. 2-19.
In Situ Strength Testing
When the soil is extremely sensitive to sampling disturbance so that it is nearly
impossible to obtain a sample that is suitable for strength testing, it may be
advisable to use in situ methods. Several of these are described in Chap. 3. There
's some debate at present whether the in situ methods are an improvement over
laboratory methods and whether they are being correctly interpreted.
Shear Strength/Parameter Correlations
Laboratory shear tests are expensive and so are avoided when possible. A
number of correlations exist which can sometimes be used to obtain strength
estimates. Correlations should be used cautiously, and if used, attempts should be
made to use more than one set so that as much checking as possible is obtained
irefer to Example 2-2). One of the major drawbacks to using correlations is that
the user tends to be conservative. If the geotechnical consultant is overly conserv-
itive, an ethical question may become involved.
Table 2-5 gives representative values for the ¢ angle for several soils. Obvi-
ously @ is a total stress value for U tests and an effective (¢’) value for CD tests.
Inspection of this table indicates ¢ = 30 to 34° is a good estimate for any but
loose sand (which is not likely to be used without densification) and ¢ = 0°
(undrained) is a good estimate for clays.
Figure 2-20 is an empirical correlation for ¢ and the plasticity index 1,
or britle
Figure 2-19 Residual soil strength. Stress-strain plot applicable for any soil, Mohr’s circle qual-
ely shown for a dense sand. For “loose” or “soft” soils 4,44, May be defined at a specified strain
20 percent),60 FOUNDATION ANALYSIS AND DESIGN
Table 2-5 Representative values for angle of internal friction
Type of test*
Unconsolidated- Consolidated-_ Consolidated-
undrained undrained drained
Soil u cu cD
Gravel
Medium size 40-55" 40.55
Sandy 35-50 35-50
Sand
Loose dry 28-44
Loose saturated 28-34
Dense dry 35-46 43-50"
Dense saturated 1-2 less than 43.50
dense dry
Silt or silty sand
Loose 20-22 27-30
Dense 25-30 30-35
Clay © if saturated 3.20 20-42
* See u luboratory manual on soil testing for a complete description of these tests,
ex, Bowles {1978}
2-12 SENSITIVITY AND THIXOTROPY
A considerable portion of the shear strength of a cohesive soil can be lost on
remolding or other similar disturbance. Sensitivity is defined as the ratio
__ undisturbed compression strength Ww
~ temolded compression strength — q, remolded
(2-51)
where q, is the maximum compressive strength from an unconfined compression
test. The remolded sample should be at the same water content as the original.
Most clays are insensitive, with S, ranging from 2 to 4. A sensitive soil has
values from 4 to 8 and an extrasensitive soil has S, > 8. Marine and lake soils
with high natural water contents often have no measurable remolded strength.
Remolded clays with $, < 16 usually regain a significant portion of the orig-
inal shear strength with elapsed time. This regain of strength with time is termed
thixotropy. Piles driven into a sensitive clay with thixotropic properties may
initially have a very low load capacity. A substantial load capacity may develop
in several hours to | or 2 months. In these cases load tests should be delayed to
allow the strength to stabilize. Remolded quick (S, > 16) clays tend to recover
very little of the original strength in reasonable time lapses on the order of under
4 months [Skempton and Northey (1952)]SOIL MECHANICS IN FOUNDATION ENGINEERING 61
Figure 2-20 Correlation between @" and plasticity index J, for normally consoiidated clays. Approxi-
imately 80 percent of data falls within one standard deviation . Typical extreme scatter values are
shown, [Datu from several sources» DM-?, Ladd et al. (1977), Bjerrum and Simons (1960), Kanja and
W olle 1977)
2
3 STRESS PATHS
Stress paths are sometimes used to plot changes in stresses on a soil. Sometimes a
better insight into soil behavior can be obtained using stress paths, There are a
number of ways to plot stress paths, but the simplest is that proposed by Lambe
(1964, 1967) and Lambe and Whitman (1979). This method uses p and q coordi-
nates defined as
a +63
P 2
a3
2
The stress coordinates p, q may be either total or effective stress values. Since
and subtracting
gives62 FOUNDATION ANALYSIS AND DESIGN
Thus, the use of effective stress coordinates simply shifts the p, q plot along the x
axis by the magnitude of the pore pressure. The p, q coordinates are the origin
and radius of the Mohr circle, respectively.
The stress path for a triaxial test and effective stresses is idealized in Fig.
2-21. Figure 2-21a is the conventional Mohr circle representation of the sample
under some cell pressure ¢, and deviator stress Ag, such that K, conditions are
obtained as shown by the K, line. This line is obtained as
P
The K, line should pass approximately through the origin for normal consoli-
dated soils. If the above q/p’ ratio is for failure stresses, the locus of points defines
the K, line. In general, K, falls below the Mohr-Coulomb rupture envelope and
the K, line is below the K, line. The latter is obvious since K, is an equilibrium
in situ stress state and K ; is a failure state.
Figure 2-21b is a set of Mohr circles and corresponding p’, q coordinates
from a triaxial test as follows:
1. Begins with a confining cell pressure ¢, and p’, q coordinates p' =, and
q=0.
2. Point Bis the p’, q coordinates for some Ac, stress increment.
Figure 2-21 (a) Coordinates of a point on stress path; (b) stress path for a series of Moh circles for a
single test; (c) stress path of (6) drawn without Mohr circlesSOIL MECHANICS IN FOUNDATION ENGINEERING 63
3. Point C is the p’, q coordinates for a larger Ao,
Point D is obtained by holding the major principal stress «, at some value and
decreasing the cell pressure until failure occurs.
5. That part of the stress path A, B, C is always inclined at an angle of 45° to the
horizontal axis.
Figure 2-21c is the same stress-path plot using only the p, q coordinates to
improve readability.
Figure 2-22a displays the relationship between K, and the ¢ line as follows
oe
m=
tanf tang
4
—41_ stan p=sin
a
a
and
ca cos @
Stress paths for the four basic triaxial tests are shown on Fig. 2-22b. The K, lines
shown in the figure are obtained from the following data:
Tet oy tykPx q
1 470 118294 176
2 ue 470 823353,
3 us 41 294 — 176
4 410 «1176823 353
For:
70 kPa, Decrease lateral pressure and maintain
}70 kPa (compression test with decreasing lat-
Test 1: Initial cell pressure
constant vertical pressure
eral pressure.
Vest 2: Initial cell pressure = 470 kPa. Increase vertical pressure with lateral
pressure — 470 kPa (standard compression test)
Fest 3; Initial cell pressure = 470 kPa. Decrease vertical pressure and hold lat
eral pressure constant (extension test~—decreasing vertical pressure).
est 4 : Initial cell pressure = 470 kPa. Increase horizontal pressure to 1176 kPa
and hold vertical pressure constant (extension test with constant vertical
pressure).
‘The stress path for a consolidation test can be established (refer to Fig. 2-23)
as follows:
|. Take point A as the state of effective stress at the end of primary consolidation
for any load increment.64 FOUNDATION ANALYSIS AND DESIGN
i
2 * 100. kPa
at ox
= 200
At ton: 2 -035-5ng
: p= sivl035°205"
ams 75h
|b} Stress paths for the 4 basic tn
tests
Figure 2-22 Stress paths.
2. Suddenly apply the next load increment. Until drainage can occur after some
time clapses, there is no change in volume and correspondingly no increase in
any shearing stress including the maximum value q. Thus, q = (01 — 03)/2 is
unchanged, causing the diameter of the circle to be unchanged but displaced
to the right an amount equal to the change in vertical pressure, In the case of
saturated soil Ag, = Au according to consolidation theory.
3. As drainage occurs, the pore pressure decreases and the Mohr circle of
effective stress increases as indicated at some elapsed times by circles C and D.
Circle D is the effective stress state when primary consolidation under this load
increment has occurred.
Example 2-4 A normally consolidated clay soil with K, = 0,95 — sin = 0.56 is consolidated in
an oedometer (consolidation) test. At the end of primary consllidation under a $0 kPa total
load, the principal stresses (neglecting any ring-side friction effects) are
o,=S0kPa 03 =K,o, = 28kPaSOIL MECHANICS IN FOUNDATION ENGINEERING 65
(a)
)
Figure 2-23 (a) Stress path for consolidation test from load increment 50 to 100 kPa plotted on a
Mohr circle; (6) same plot using p, q coordinates.
and the corresponding p, q coordinates are
39
These coordinates plot point 4 on the Mohr circle as shown in Fig. 2-234
The next load increment of Ap/p = 1 gives a total stress on the sample of 100 kPa. This
load increment, at t = 0, moves circle A horizontally a distance of 100 — 50 = SO kPa (this is
also the instantaneous pore pressure) to circle B with p, q coordinates of
(50 + 50) + (28 + 50)
7 =89kPa (also 50 + 39)
q= 11 as before
After primary consolidation is complete, the effective pressure o', is equal to the initial total
pressure:
0; = 100 kPa
oy = K,o, = 56 kPa
and =7% and q=q'=22kPa
The latter p, q coordinates represent point D of Fig. 2-23. Curve C is at some arbitrary
intermediate consolidation stress state. Line ABCD represents the total stress path of the soil in
consolidating under the 50-kPa load increment. Line AD represents the effective stress path. The
horizontal distance between AD and BCD represents the excess pore pressure at some instant in
time; CC’ is the excess pore pressure at a time when the total stress path is at p, q = point C
(83.5, 16.5) with a value
‘Au = 83,5 — 58,5 = 25 kPa
or 50 percent of the excess pore pressure has dissipated,
Line AD is a segment of the K, line for this soil. If this line is extended, it will intercept the
origin at p = q=0, since the clay is normally consolidated and at the beginning of the test66 FOUNDATION ANALYSIS AND DESIGN
= 0, = 0) = 0. The slope of this K, line is
1-k,
+K,
tan p=
wit
2-14 ELASTIC PROPERTIES OF SOIL
The stress-strain modulus (also modulus of elasticity), shear modulus, and Pois:
on’s ratio are the principal elastic properties of interest. Both the stress-strain
modulus E, and Poisson’s ratio p are of use in evaluating foundation settlements.
They may also be used to back-compute the modulus of subgrade reaction k,
The shear modulus G is used in soil dynamics problems to compute amplitudes
of vibrations.
The stress-strain modulus is computed as shown in Fig. 2-24a. Typical value
ranges for several soils are given: in Table 2-6.
Poisson’s ratio is defined as the ratio of lateral strain ¢, to longitudinal
strain ¢, when the applied stress is uniaxial (Fig. 2-24b) or
(a)
Typical y values for several materials are given in Table 2-7.
Inti tangent modulus
Ye ONG oy Envelope ——__
4
Deviator stress, Ao
cyclic test
Note increase in Ey
oneyclic test
Strain, « Strain
@)Stress- strain modulus
.
rr #
‘ld og +E use iii
Ja i tangent oF
vt | secant line
Hae
1 rar
a cu
ae Determation 8
(b) Poisson's ratio, « {c}Modulus of subgrade reaction
Figure 2-24 Elastic properties of soilSOIL MECHANICS IN FOUNDATION ENGINEERING 67
Table 2-6 Typical range of values for the static
stress-strain modulus E, for selected soils
Field values depend on stress history, water content,
slensity, ete
ksf
Clay
Very soft 50-250
Soft 100- 500
Medium 300-1000
Hard 1000-2000,
Sandy 500-5000
Glacial ill
Loose 200-3200
Dense 3000-15 000
Very dense 10.0030 000
Loess 300-1200
Sand
Silty 150-450
Loose 200-500
Dense 1000-1700
Sand and gravel
Loose 1000-3000
Dense 2000-4000
Shale 3000-300 000
sit 40-400
Mpa Wa fee
25 vee
5.25 rhs Be
15-50
50-100
25-250
10-153
144-720
478-1440
1457 ae
12 *
10-24
48-81 reo
48-144
96-192 geal tns
14414400 ge crite
2-20 an ener
Vable 2-7 Typical range of values for Poisson's ratio jz
Type of soil
Cay, saturated
Clay, unsaturated
Sandy elay
sin
Sand (dense)
Coarse (void ratio = 0.4-0.7)
Fine-grained (void ratio = 0.4.0.7)
Rock
Loess
lee
vonerete
#
0405
01-02
0203
03-038
02-04
os
025
01.1 04 (depends somewhat on
type of rock)
0103
036
068 FOUNDATION ANALYSIS AND DESIGN
The modulus of subgrade reaction is defined as the ratio of stress to defor-
mation as shown on Fig. 2-24c. The units of k, are the same as unit weight.
The shear modulus G is defined as the ratio of shear stress to shear strain. It
is related to E, and pas
* +H ©
The shearing strain ¢, is the change in right angle at any corner of an element as
in Fig, 2-246 such that
2, — angle BCD ~ angle B'C’D’ G)
Another concept occasionally used is the volumetric strain, defined as
AV
v @
The volumetric strain was used to plot strain versus log p of Fig. 2-11b. The bulk
modulus E, is defined as the ratio of hydrostatic stress to the volumetric strain ¢,:
(e)
Since G and E, cannot be negative, Eqs. (b) and (e) set the limits of as
—lsy4s05
It appears that the range of j« for soils is 0 to 0.5. Saturated soils have > 0.5
and dry soils have > 0.
Hooke’s generalized stress-strain law in terms of principal strains ¢ and
stresses a can be written in matrix notation as
a= De (2-52)
where the matrix D contains Poisson’s ratio as
rg | 2 [3
1 1 | se) =
D= |
@ | Sal) a
ea) elles
For the CD or CU triaxial test with a cell pressure 3 and the deviator stress Acy
acting, we have
1
6) = 5 (Aon ~ Quo) (2-53)
If we plot the stress-strain data and draw a smooth curve through the points,
we should be able to solve Eq. (2-53) for E, and jt by taking Ac, and ¢, at closely
spaced intervals so a linear variation can be assumed in the tangent modulus E,SOIL MECHANICS IN FOUNDATION ENGINEERING 69,
The result is the tangent modulus and y« for that stress level. If this is done on a
large number of closely spaced points and the initial value of ¢, is very close to
0.0, we find that Poisson's ratio exceeds 0.5 at very small strain values. This can
be interpreted that most of the stress-strain curve is in the plastic range of strains.
Vhis type of analysis also verifies that both E, and j1 are stress-dependent.
Equation (2-53) indicates that the unconfined (¢3 = 0) compression test will
produce larger axial strains ¢, at the same stress level compared with triaxial
values. This is why the unconfined compression test produces smaller values of
1, -often too small by a factor of 4 or 5
The initial tangent modulus is most often used for £,. This is for several
reasons
1. Soil is elastic only near the origin
>. There is less divergence between all plots in this region
\. The largest values are obtained—often three to five times larger than a tangent
or secant modulus from another point along the curve
In spite of these several shortcomings for E, the value along the curve is
commonly used in finite-element analyses based on the computed stress level.
This may require that the problem be iterated several times until the computed
tress level matches the stress level that was used on the previous cycle to ob-
‘in E,
A’number of investigators [Leonards (1968), Soderman et al. (1968), Makh-
louf and Stewart (1965), Larew and Leonards (1962)] have proposed that a better
vnitial tangent modulus is obtained by cycling the deviator stress to some stress
vel about five times and then failing the sample. The initial tangent modulus by
this method is somewhat higher than on the first cycle.
The method just described is to obtain a static (or resilient) stress-strain
modulus value, Cyclic tests where the cycles are in terms of low-amplitude strains
vad frequencies in the range of } to 10 Hz are used to obtain dynamic values of
1 and G. The dynamic stress-strain modulus may be two to ten times the static
value. The dynamic moduli will be considered in more detail in Chap. 20.
Both the stress-strain modulus and Poisson’s ratio are heavily dependent on:
! Method of performing the compression tests (unconfined, confined, com-
pression, extension, U, CU, or CD).
Confining cell pressure a. E, tends to increase nonlinearly with an increase in
vonfining pressure
8 Overconsolidation ratio OCR.
+ Soil density—£, increases with particle packing.
Water content of soil—lower water contents tend to higher values, Brittle
fractures at low strains occur at low water contents.
Strain rate—at low strain rates the modulus value can be lower by a factor of
2 or more compared with the value obtained at a high test rate [Richardson
and Whitman (1963)].
Sample disturbance.70 FOUNDATION ANALYSIS AND DESIGN
The stress-strain curve for all soils is nonlinear for all except a possible short
segment near the origin. Kondner (1963) proposed that the stress-strain curve
(Fig. 2-252) could be represented by a hyperbolic equation of the form
e
o,-03=
oe ae oe
which could be rewritten with Ao, = 0, — ¢3 in linear form as
é
aan athe (2-54)
The left side of Eq. (2-54) can be computed for various values of deviator stress
and the corresponding strain to make a linear plot as shown in Fig. 2-25b.
Extension of the plot across the discontinuity at ¢ > 0 gives the coefficient a, and
the slope is b. While Kondner proposed this procedure for clay soils, it should be
applicable for all soils with similar stress-strain curves [see Duncan and Chang
(1970)
The following empirical correlations may be used to estimate E, for cohesive
soils
Normally consolidated sensitive clay:
E, = (200 to 500) x s, (2-55)
Normally consolidated insensitive and lightly overconsolidated clay
E, = (750 to 1200) x s, (2-56)
Heavily overconsolidated clay
E, = (1500 to 2000) x s, (2-57)
Several equations will be presented in the next chapter using in situ testing which
may be used for both cohesive and cohesionless soils to compute E,.
2-15 ISOTROPIC AND ANISOTROPIC SOIL MASSES
An isotropic material is one in which the elastic properties (E, and 42) are the same
in all directions. The elastic properties for anisotropic materials are different in
the different directions. A material is homogeneous when the physical and com-
positional properties such as 7, void ratio, clay, and silt or clay content are the
same throughout the volume of interest.
Almost all naturally occurring soil deposits are anisotropic and nonhomoge-
neous. The anisotropy is produced from a combination of particle placement
during deposition/formation (also called geometrical or inherent anisotropy) and
from overburden pressures. In natural soils this commonly results in horizontal
bedding planes having both strength and elastic properties different for samples
stressed perpendicular and parallel to the bedding planes. This property of an-tony Figure 2-25 (a; Usual stress-strain
plot—hyperbolic-curve approxi-
mation; (b) transformed stress-
strain representation of stress
strain—gives approximate linear
curve as shown, [4fier Kondner
(a) (v) (1963)}
isotropy has been known for some time [Casagrande and Carrillo (1944)], but
only in more recent times have attempts been made to quantify the effects [see
Yong and Silvestri (1979), Law and Lo (1976), Arthur and Menzies (1972), and
Yamada and Ishihara (1979)]. Figure 2-26 illustrates anisotropy and the possible
range in strength which occurs when the stress orientation is at some angle with
respect to the bedding plane. This figure should also be compared with Fig. 2-16
1o see how anisotropy can qualitatively affect in situ shear resistance depending
on the intersection angle between the bedding plane and the potential shear
plane.
Nonhomogeneous deposits are produced from particle packing versus depth,
mass contamination during deposition, and lenses or strata of different materials
in the depth of interest. The increase in particle packing and confining pressure
with depth always produces a stress-stfain modulus increase with depth which is
usually nonlinear. It has been common, however, to assume a soil mass is semi-
infinite, homogeneous, and isotropic, even in layered deposits, as a computational
convenience. The current state-of-art is such that a soil mass can be somewhat
more realistically modeled than this, albeit at some additional time and expense.
Anisotropy is an important consideration in finite-element analyses of soils,
since elastic properties are input parameters. Where two elastic constants define
the stress-strain relationship [Eq. (2-52)] of an isotropic material, five constants
are required when a homogeneous soil is deposited in layers so that one can
issume symmetry about a vertical axis. A soil deposit which meets this criterion
Anisotrany rave R* Suv! Suh
Sua? yn 114(R-1)cos?a] Figure 2-26 Undrained shear
5. undrained shear sirength (see above for orientation) strength for anisotropic soils.72 FOUNDATION ANALYSIS AND DESIGN
is termed cross-anisotropic, Strictly, a soil is not cross-anisotropic because of a
depth variation, but this is a simplification which may not introduce serious
computational errors. This simplification has the effect of reducing 21 elastic
constants of the general case to 7.
The seven elastic constants for a cross-anisotropic material (actually only five
are independent) are defined as follows (the xz plane of isotropy is horizontal and
the y axis is vertical):
stress-strain modulus in the vertical direction
Ey, = stress-strain modulus in the horizontal plane, i.
../é, when the applied stress is
2, /e, when the applied stress is o,
,/e, when the applied stress is o,
shear modulus in the horizontal plane
Gy = shear modulus in the vertical plane
in the plane of isotropy
En
But i
: (1 + pa)
Ha _ Bs
and RE @
So the five elastic constants for a cross-anisotropic material are Gy. Ev. Ew. ts.
and jz. A more detailed discussion on cross-anisotropic behavior of soil deposits
can be found in Bhatacharya (1968).
The generalized Hooke’s law for cross-anisotropic material takes the follow-
ing form:
()
For problems of plane strain (when ¢,
O, = W190, + Mr ()
Ho,
Ey’
Substituting Eq, (c) in Eqs. (b), using Eq. (a) to obtain 1, and noting that
=0, the following form of the generalized Hooke’s law for crossSOIL MECHANICS IN FOUNDATION ENGINEERING 73
nisotropic material in plane strain is obtained:
&, = Bo, + Co,
ty =
et = Gy
where
: : @
Coa nate
Ey Ey
Hence, the D matrix for plane-strain problems of cross-anisotropic materials is
| als
e
1 |A] Bl O
D 2 clo (2)
t—}—_}-—
1
3/010
G
Thus for plane-strain problems of cross-anisotropic materials it is only
necessary to know the four parameters 4, B,C, and Gy, which can be determined
| Chowdhury (1972)] as follows:
|. Perform a set of plane-strain triaxial tests with a constant cell pressure on a
sample with the plane of isotropy horizontal.
Plot the deviator stress vs. axial strain.
\. Plot the deviator stress vs. lateral strain. The lateral strain can be computed
from the axial strain and volume-change measurements,
4. Compute: 1/B = slope of curve of step 3.
I/C = slope of curve of step 2.
S. Perform a plane-strain triaxial test with a constant cell pressure on a sample
with the plane of isotropy vertical such that the direction of plane strain is
parallel to the plane of isotropy.
+. Plot steps 2 and 3 above to obtain a second set of curves.
Compute: 1/B = slope of curve of step 3 (should check reasonably with step 4).
1/A = slope of curve of step 2.
S. Test a sample with the plane of isotropy inclined at 45° to the horizontal
(samples may be difficult to obtain except from a test pit).74 FOUNDATION ANALYSIS AND DESIGN
9. Plot the deviator stress vs. axial strain. The slope do/de of the curve is related
to Gy by the following equation
1
Gv = Vilope — (A + 2B + ©) “
Thus the four constants required to solve the plane-strain problems of cross-
anisotropic soil can be obtained from three sets of plane-strain triaxial tests; one
set of tests is on soil samples with the plane of isotropy horizontal; the second set
is on samples with the plane of isotrupy vertical; and the third set is on samples
with the plane of isotropy 45° inclined to the horizontal. Effective or total stresses
may be used as appropriate, but all values should be consistent. Since the value
of Gy is particularly critical [Raymond (1970)], all four constants A, B, C, and Gy
must be correctly determined if one wants to consider the cross-anisotropy of the
soil. If the correct evaluation of each of the four constants is not possible, the soil
should be treated as an isotropic material.
PROBLEMS
(Note: Select partial answers are purposely not identified.)
2-1 A soil has a unit weight of 20.54 kN/m?, For G, = 268 and w= 125 percent. find: y,,. void
ratio e, porosity n, and degree of saturation S.
Partial answer : 0.305, 76.1 percent
2-2 A soil has a unit weight of 122.5 pef. IFG,
porosity n, and water content w.
Partial‘answer : 0.767, 28.4 percent
2-3 A soil has an in situ void ratio e, = 1.87, wy = 600 percent, and G, = 2.75. What is 7... and 5?
What kind of “soil” might have this unit weight?
Answer : 15.04 kN/m? (95.8 pef), S = 88.3 percent
2-4 A sample of saturated clay has a mass of 1853.5 g and 1267.4 g dry. The dry unit weight is 14.71
KN /m?. What is (a) wet unit weight; (b) void ratio ¢; (c) specific gravity G, ; and (d) wet unit weight for
'S = 50 percent? Comment on the type of clay this might be.
2-5 Redo Example 2-1 if the sample weighs 2360 g, All other data are the same. Would you say this
isa “saturated” soil?
26 Classification tests were performed on a light-brown sandy soil which visually has several pieces
of gravel larger than 6 mm, The following laboratory data were obtained:
70 and the soil is saturated, find:7,y, void ratio e,
Sieve No. Percent passing
4 98.0
40 36.5
200 20.8
we = 332 226 percent
Required: Classify this soilSOIL, MECHANICS IN FOUNDATION ENGINEERING 75
2-7 Data were obtained from a relative-density test using information from six laboratory tests:
Limiting y Average 7, KN/m?
ie 1807 1752
ea 1477 15.56
Shea = 1697 (average from two tests)
Kequired: Compute the range of D,
Answer : 60 to 83 percent
2-8 A consolidation test was performed on a sample with initial dimensions of H
ing diameter
20.00 mm and
63.00 mm. At the end of the test the sample height was 14.30 mm and the oven-dry
\seight of the soil cake was 95.63 g. The G, of the soil solids is 2.66.
Required: (a) Initial and final void ratios ¢,, e, ;(b) total sample strain é
Partial Answer + 0.735, 0.285
2-9 The in situ p, = 200 kPa. A consolidation test gives p. = 500 kPa, A U test on a field sample
.¢ 5, = 80 kPa, What is the expected strength of a normally consolidated laboratory sample?
Answers, = 37 kPa
10 The in situ effective preconsolidation pressure p, is 450 kPa. The in situ effective overburden
essure p, = 200 kPa. The s, for a normally consolidated laboratory test on remolded soil is 55 kPa.
stimate the in situ s
DA Estimate K, and i for the soil af Prob 7.6 What is an estimate for Kf p/p, = 4?
2-12 A CU triaxial test on a g-c soil yielded the following data
Test No. o,.kPa Ay, KPa
1 80 120
2 160 170
3 240 200
Required: Find @ and ¢
Partial answer: ¢ = 36 kPa
2.13 An unconfined compression (U) test was performed on a cohesive sample with the following
ita obtained: 1. = 10.0 mm; diameter = 50.0 mm, AL = 8.0 mm: Praituee = 0.133 KN,
Required: What are the soil parameters?
Partial answer :¢ = 31.4kPa
2-14 ACU triaxial test was performed, with the following data obtained:
Test No. Ao, Au kPa
1 100238 36
2 200307 108
3 300 389 197
kequired: Total and effective stress parameters
Partial answer : $ = 15°; 6’ = 41°16 FOUNDATION ANALYSIS AND DESIGN
2-15 Plot the data of Prob. 2-14 using a p,q diagram for both the total and effective stresses, and find
the soil parameters,
2-16 Plot Example 2-4 in Sec. 2-13 and verify that the next load increment (200 kPa) will fall on the
K, line. Show the effective and pore pressures for several intermediate consolidation intervals.
2-17 ACU direct shear test was performed on a 50-mm-square sample with the following data
Test No. P,, KN Py. KN
1 0.05 0.043
2 000 0.061
3 030 0.096
Required: Plot the data and find ¢ and
Partial answer : @ = 13.4°
2-18 Given the data of Table P2-15 from a consolidation test.
Required for each assigned load inerement
1, Plot dial reading vs. log time and find fs
2. Plot dial reading vs... find ts, and compare with item 1
3. Assume two-way drainage and the initial H = 0800 in (at beginning of $ ton/ft” load), and
compute, and C,
Partial answer : For ton/ft?:t55 = 1.2 (log), 1.6 mins¢, = 0.0257 in?/min
2-19 Three consolidation tests were performed on separate layers of soft clay underlying the site
(refer to Fig, P2-19}, The samples were consolidated from an initial = 20.00 mm in a 630-mm-
diameter ring, The following data were obtained
‘Table P2-15 Dial readings ( x 0.0001)
Time, min fron? ton/ft® 1 ton/ft®
2188 2127
2180 2119
2172 213
2162 2105
2153 2004
2144 2083
2139 2073
2135 2062
2132 2055
231 2050
2130 2087
2129 2046
2128 2085
2127 2084SOIL MECHANICS IN FOUNDATION ENGINEERING 77
Void ratio ¢
Test load, KPa Soil No.3 Soil No. $ Soil No.7
° 1.405 Liss List
8 139s 1190 1.140
16 1.393 1187 Las
2 1.390 1.180, 1130
64 1385 L175 Los
140 1380 112s 1.080
240 1360 1080 060
560 1.180 0925 0.965
680 10
1020 092s 0.760 0.805
2040 0.725 0.625 0.680
wy 520% 69.0%
% 260-240
@ 2702.66
& 1181.12 (as weG))
aH, 0.0832 mm 0.0456 0.1023 (0-8 kPa)
AH 5.6550 mm 5.1936 4.3794 (0-2040 kPa)
Soil 1 oe 7
ckNim? 1572 1831 * 1862, R73 *
* Compute from ¢, and G, given above,
igure P2-19 Soil No. 1, organic silt and clay; soil No. 2, medium dense
sind; soils No. 3, 5, and 7, clay; soils No. 4 and 6, thin silt seams.
Required: Plot e versus log p curves as assigned and find C, . pp and comment.
Plot e versus log p curve for soil assigned and find C; and compare computed and C, from ¢ versus
lint p curve. Use sample height and definition of Ae together with given values of AH to compute
remaining values of «
Partial answer : Soil 3: C, = 1373 C, 2032
7, = 161.2
r= 1919p = 250
20 For Fig. P2-20: (a) Estimate hat which the sand would be expected to become “quick”, (b) if
‘= 0.25 m, what isthe effective pressure at A”
Answer : (a) 0.45 m;(b) 1.95 kPa
Sond
| sam) 92.65
2 = 015m
Wy =I5kg
ey Figure P2-20,78 FOUNDATION ANALYSIS AND DESIGN
2-21 What H in Example 2-3 will produce a “quick” condition at C?
Answer: H = 143 m
2-22 Plot the assigned triaxial test data of Prob. 5-14. Make a smooth curve through the points and
starting with a strain = 0.005 compute E, and y using Eq. (2-53). Note the stress and strain level
where is either >0.5 or negative. Can you draw any conclusions about these computations?
2.23 Plot the assigned triaxial test data of Prob. 5-14 and obtain @,. What is @ in this test?CHAPTER
THREE
EXPLORATION, SAMPLING, AND IN SITU SOIL
MEASUREMENTS
3-1 DATA REQUIRED
Investigation of the underground conditions at a site is prerequisite to the econ-
comical design of the substructure elements, It is also necessary to obtain sufficient
information for feasibility and economic studies for a proposed project. Public
building officials may require soil data together with the recommendations of the
geoteotechnical consultant prior to issuance of a building permit. This is true
particularly if there is a chance that the project will endanger the public health or
safety or degrade the environment.
Elimination of the site exploration, which usually ranges from about 0.5 to
1.0 percent of total construction costs, only to find after construction has started
that the foundation must be redesigned is certainly false economy. This is gener-
ally recognized, and it is doubtful if any major structures are currently designed
without site exploration being undertaken. Small structures are often designed
without site exploration; however, the practice is not recommended. The condi-
tion of the adjacent structures is an indication, but certainly no guarantee, that
the site is satisfactory. With the scarcity of building sites in urban areas and with
considerable urban renewal and the accompanying backfill, often with no quality
control, the underground conditions can have significant variation within a few
meters in any direction.
The elements of a site investigation depend heavily on the project but gener-
ally should provide:
I. Information to determine the type of foundation required (shallow or deep)
2. Information to allow the geotechnical consultant to make a recommendation
on the allowable load capacity of the foundation element (or soil)
080 FOUNDATION ANALYSIS AND DESIGN
Sufficient data/laboratory tests to make settlement predictions
4. Location of the groundwater level (or determination of whether it is in the
construction zone)
5. Information so that the identification and solution of excavation problems
(sheeting and dewatering) can be made
6. Identification of potential problems (settlements, cracks, etc.) concerning ad-
jacent property
7. Identification of environmental problems and their solution
‘An exploration program may be initiated on an existing structure where
additions are contemplated. The current safety of an existing structure may re-
quire investigation if the performance is not as anticipated (excessive settlements
or cracks) so that required remedial measures may be undertaken based on new
found information or the new evidence and a reinterpretation of the original
data.
A part of the exploration program may include on-site monitoring, both
during and after construction, to make certain recommendations are being fol-
lowed. Where the excavation reveals conditions requiring design changes, moni-
toring of progress will ensure change orders are initiated early enough to keep
costs to a minimum. Postconstruction monitoring of building performance is
particularly desirable from the geotechnical consultant's view, since this allows
for a review of the design procedures and builds a data base for future work.
Unfortunately, few owners are willing to make this investment or even allow
property entry should the foundation consultant be willing to underwrite the
cost.
While the primary focus of this chapter is on-site exploration for buildings
and other structures where the cost per unit of area is high (compact site), many
of the methods are applicable to roads, airfields, and water, sewer, pipe and
power lines, and other extended sites. Extended site exploration is useful to
establish line and grade, locate groundwater level and rock line, delineate zones
of poor-quality soil, and establish borrow pits.
3-2 METHODS OF EXPLORATION
‘The most widely used method of subsurface investigation for compact sites and
most extended sites is boring holes into the ground from which samples may be
collected for either visual inspection or laboratory testing. Several procedures are
commonly used to drill the holes and to obtain the soil samples. These will be
taken up in more detail later.
Aerial photographs, in conjunction with drill holes or test pits, may be es-
pecially useful for a soil-exploration program encompassing large areas. Color
photographs are an added cost, but they tend to be more sensitive to soil color
changes and are generally more useful. Minard and Owens (1962, with a bibli-
ography) present several articles on using aerial photographs for large-scaleEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 81
exploration programs. For large areas, statistical techniques in conjunction with
serial photographs may prove helpful in reducing the amount of sample col-
lection and testing, This technique was applied to a large-scale mapping project
in South Africa [see Kantey and Morse (1965)].
Geophysical methods are also used with primary application on extended-
site exploration. These methods fall in two general categories, namely, seismic and
resistivity methods. The use of geophysical methods is usually limited to es-
tablishing the location of bedrock underlying softer materials or locating gravel
(or sand) deposits.
Visual inspection and sample collection of the exposed subsoil in erosion
ditches, construction excavations, or test pits are sometimes satisfactory, Load
tests in test pits (or in the site excavation) can be used to establish soil bearing
capacity, but are relatively costly. Table 3-1 summarizes various soil-exploration
methods available and discussed in the following sections.
3-3 PLANNING THE EXPLORATION PROGRAM
The purpose of the exploration program is to determine, within practical limits,
the stratification and engineering properties of the soils underlying the site. The
principal properties of interest will be the strength, deformation, and hydraulic
characteristics, The program should be planned so that the maximum amount of
information can be obtained at minimum cost.
It may be more economical to provide a conservative building design than to
expend large sums on an elaborate exploration and testing program. On the
other hand, sufficient exploration should be undertaken so that the foundation
consultant is not in the position of making an expensive recommendation to
protect against uncertainties which could have been detected by a reasonable
program. It should be understood that a recommendation made simply to pro-
tcct the consultant. when an adequate exploration has been undertaken is not
cthieal
If the soil is highly erratic, there should only be sufficient borings to establish
1 general picture of the underground conditions. An extensive boring (and lab-
oratory testing) program is not justified in crratic soils, and the final design
should be conservatively based on the properties of the poorer soils, Again a
question of ethics is involved if an excessive number of borings are taken under
these circumstances.
In planning the program the foundation consultant must have a good know-
ledge of current and accepted methods of both field exploration and laboratory
testing and their limitations. A competent consultant will also have sufficient
understanding of equipment function and soil behavior to make adjustments so
that nonstandard equipment or test methods can be used, if necessary, to obtain
the desired information,
In planning the program full advantage should be taken of any existing
information, including the geotechnical consultant's own data base for the area.82. FOUNDATION ANALYSIS AND DESIGN.
Table 3-1 Summary of soil-exploration methods
Method Depths Applicability
No samples taken
Geophysical Usually less than 35 m Locating firmer material underlying softer
seismic material. Certain equipment is adapted
resistivity to determination of density and soil
moisture
Vane shear Limited by torque stresses on _In situ shear strength of sensitive cohesive
rod nongravelly soils
Sounding Limited by encountering rock Locating soft material and rock by
and driving equipment probing, using solid rods as opposed to
a split spoon
Dutch cone Same as “sounding”
Pressuremeter Usually less than 35 m In situ E,, 4 Ko
Glitz! cell Same as pressure meter In situ Ky
Fracture apparatus Same as pressure meter In situ Ko
Disturbed samples taken
Auger boring. Depends on equipment and All soil where hole will maintain wall
time available, practical without casing
depths being up to about
35m
Rotary drilling Depends on equipment, most Al soils. Some difficulty may be
Wash boring equipment can drill to depths encountered in gravelly soils. Rock
Percussion drilling of 70 m or more requires special bits, and wash boring
is not applicable. Penetration testing
is used in conjunction with these
methods. and disturbed samples are
recovered in the split spoon. Pene-
tration counts are usually taken at 1- to
2-m increments of depth
Test pits and open As required, usually less than All soils
cuts 6 m; use power
equipment
Undisturbed samples taken
Rotary drilling, per- Depends on equipment, as for ‘Thin-walled tube samplers and various
cussion drilling, disturbed-sample recovery piston samplers are used to recover
wash boring samples from holes advanced by these
methods. Commonly, 5- to 10-em-
diam samples can be recovered
Test pits Same as for disturbed samples Hand-trimmed samples. Careful trim-
ming of sample should yield the least
sample disturbance of any methodEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 83
It is obviously most helpful to have done site exploration on adjacent sites, or at
lcast in the general area. It will also be most advantageous to have made the
initial borings if this is a part of the detailed site exploration follow-up from an
earlier feasibility study. Even if the consultant does not have a data base to work
from, considerable underground information may exist—particularly in urban
areas—in various public and utility offices, the owner’s files, or the files of the
engineer/architect who has retained the geotechnical consultant. In any case the
borings should be used for a correlation and extension of the existing data base if
at all possible. In an undeveloped area where no data base currently exists the
program is in fact “exploratory.”
The actual planning of a subsurface exploration program includes some or
all of the following steps:
|. Assembly of all available information on dimensions, column spacing, type and
use of the structure, basement requirements, and any special architectural
considerations of the proposed building. Foundation regulations in the local
building code should be consulted for requirements which may be a local
peculiarity. For bridges the soil engineer should have access to type and span
lengths as well as pier loadings. This information will indicate probable soil
loadings and settlement limitations.
». Reconnaissance of the area. This may be in the form of a field trip to the site
which can reveal information on the (ype and behavior of adjacent structures:
such as cracks, noticeable sags, and possibly sticking doors and windows. The
type of local existing structure may influence, to a considerable extent, the
exploration program and the best foundation type for the proposed adjacent
structure, Since the existing structures must be maintained in an “as is” condi-
tion, nearby excavations or construction vibrations will have to be carefully
controlled. Erosion in existing cuts (or ditches) may also be observed, but this
information may be of limited use in the foundation analysis of buildings. For
highways, however, runoff patterns, as well as soil stratification to the depth of
the erosion or cut, may be observed. Rock outcrops may give an indication of
the presence or the depth of bedrock.
The reconnaissance may also be in the form of a study of the various
sources of information available, some of which are:
Geological maps. Either U.S. government or state geological survey maps.
Agronomy maps. Published by the Department of Agriculture (state or U.S.).
Aerial photographs. May require special training to interpret soil data, but
terrain features are easily recognized by the nonspecialist.
Warer-well logs (and oil wells).
Hydrological data, Data collected by the U.S. Corps of Engineers on stream-
flow data, tide elevations, and flood levels.
State highway department soil manuals.
State university publications. These are usually engineering experiment station84 FOUNDATION ANALYSIS AND DESIGN
publications, Information can be obtained from the state university if itis
not known whether a state study has been undertaken and published.
3. A preliminary site investigation. In this phase a few borings are made or a test
pit is opened to establish in a general manner the stratification, types of soil to
be expected, and possibly the location of the groundwater table. One or more
borings should be taken to rock, or competent strata, if the initial borings
indicate the upper soil is loose or highly compressible. This amount of ex-
ploration is usually the extent of the site investigation for small structures. A
feasibility exploration program should include enough site data and sample
recovery to approximately establish the foundation design and identify the
construction procedures. Construction procedures (sheeting, bracing, tiebacks,
slurry walls, etc.) can represent a very significant part of the foundation cost
and should be identified as early as practical.
It is common at this stage to limit the number of quality samples recovered
and rely heavily on strength and settlement correlations using index properties
such as liquid limit, plasticity index, and penetration test data together with
unconfined compression tests on samples recovered during penetration testing.
4. A detailed site investigation. Where the preliminary site investigation has es-
tablished the feasibility of the project, a more detailed exploration program is
undertaken. The preliminary borings and data are used as a basis for locating
additional borings, which should be confirmatory in nature, and determining
the additional samples required. It should be noted that if the soil is relatively
uniform in stratification, a rather orderly spacing of borings at locations close
to critical supersturcture elements should be made. On occasion additional
borings will be required to delineate zones of poor soil, rock outcrops, fills,
and other areas which can influence the design and construction of the foun-
dation.
Sufficient additional soil samples should be recovered to refine the design
and for any construction procedure required by the contractor to install the
foundation. This should avoid an excessive (uncertainty factor) bid for the
foundation work, cost overruns, and/or damage to adjacent property owners
from unanticipated soil conditions discovered when the excavation is opened.
3-4 SOIL BORING
Exploratory holes into the soil may be made by hand tools (Fig. 3-1), but more
commonly mounted power tools (Figs. 3-2 and 3-3) are used.
Hand Tools
The earliest method of obtaining a test hole was to excavate a test pit using a
pick and shovel. Because of economics, the current procedure is to use power-
excavation equipment such as a backhoe to excavate the pit and then to useEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 85
hand tools to remove a block sample or shape the site for in situ testing. This is
the best method at present for obtaining quality undisturbed samples or samples
for testing at other than vertical orientation, For small jobs, where the sample
listurbance is not critical, hand or powered augers held by one or two persons
can be used. Hand-augered holes can be drilled to depths of about 35 m, al-
though great depths (say, greater than 8 to 10 m) are usually not practical,
Commonly, the depths run 2 to 5 m, as on roadway, airport, or small-structure
investigations.
Mounted Power Drills
For numerous borings to greater depths and to collect samples that are undis-
turbed, the only practical method is to use power-driven equipment. Wash boring
1s a term used to describe one of the most common methods of advancing a hole
into the ground. A hole is started by driving casing (Fig. 3-2) to a depth of 2 to
3.5 m. Casing is simply a pipe which supports the hole, preventing it from caving
in. The casing is cleaned out by means of a chopping bit fastened to the lower
end of the drill rod. Water is pumped through the drill rod, and exits at high
selocity through holes in the bit. The water rises between the casing and drill rod,
carrying suspended soil particles, and overflows at the top of the casing through a
| connection into a container, from which the effluent is recirculated back
through the drill rod. The hole is advanced by raising, rotating, and dropping the
hit into the soil at the bottom of the hole. Drill rods, and if necessary casing, are
udded as the depth of the boring increases. Usually 6 m or less of casing is
‘equired at a hole site, This method is quite rapid for advancing holes in all but
the very hard soil strata.
Rotary drilling is another method of advancing test holes. This method uses
sotation of the drill bit, with the simultaneous application of pressure to advance
the hole. Rotary drilling is the most rapid method of advancing holes in rock
unless it is badly fissured; however, it can also be used for any other type of soil.
i this method is applied in soils where the sides of the hole tend to cave in, a
uvilling mud may be used. The drilling mud is usually a water solution of a
'hixotropic clay (such as bentonite), with or without other admixtures, which is
‘ced into the sides of the hole by the rotating drill. This provides sufficient
length in conjunction with the hydrostatic pressure of the mud suspension
(7 ~ 1,1 to 1.2 g/em?) against the soil so that it maintains the hole. The mud
pressure also tends to seal off the water flow into the hole from the permeable
s.iter-bearing strata. Various drill heads are available, such as auger heads for
hallow highway and borrow-pit exploration, grinding heads for soil and rock,
ind coring bits for taking cores from rock, as well as from concrete and asphalt
pavements
Continuous-flight augers are probably the most popular method of soil ex-
ploration at present (Fig, 3-3), The flights act as a screw conveyor to bring the
| to the surface. The method is applicable in all soils, although in saturated
wud under several feet of hydrostatic pressure, the sand tends to flow into theEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 87
lead sections of the auger, requiring a washdown prior to sampling. Borings up to
nearly 100 m can be made with these devices, depending on the driving equip-
ment, soil, and auger size. The augers may be hollow-stem or solid, with the
hollow-stem type generally preferred as penetration testing or tube sampling may
be done through the stem. Borings do not have to be cased using continuous-
flight augers for obvious reasons, and this is a decided economic advantage over
other boring methods. Continuous-flight augers are available in 3- to 5-ft sections
and in several diameters including the following (in inches):
Solid stem
oD Ps 3 4 4h 5$ 6 7
Hollow stem |24 x 64 2x7 3x8 34x9 4x10 5x10 6x 12
ID x OD
Inspection of this list of auger diameters indicates that a wide range of tube
sample diameters may be used. Tube samples are generally limited to about
100 mm diameter, however, to obtain the best balance between sample quality
and the cost of drilling the hole.
The actual hole diameter will be on the order of 12 mm larger than the auger
size. In practice a cutting head is attached to an auger flight, and with or without
a head plug depending on the soil, and the hole is advanced. At the desired depth
the plug is removed (if used) and a penetration test performed and/or a tube
sample recovered. If a plug is not used, the soil cuttings may have to be removed
from the bottom so that the test can be made or a sample recovered which is
from undisturbed soil. Caution should be exercised in removing the plug below
the water table, since a difference in water level inside and outside the tube may
create a temporary quick condition in the soil in the bottom of the hole until the
water level stabilizes. Auger flights are added as required and the hole is ad-
vanced to a depth deemed adequate for the purpose of the exploration.
Percussion drilling is still another method of forming a hole. In this method
the drill is lifted, rotated slightly, and dropped onto the bottom of the hole.
Water is circulated to bring the soil cuttings to the ground surface; casing is
required as well as a pump to circulate the water.
3-5 SOIL SAMPLING
The most important engineering properties required for foundation design are
strength, compressibility, and permeability. Reasonably good estimates of these
properties for cohesive soils can be made by laboratory tests on undisturbed
samples which can be obtained with moderate difficulty. It is nearly impossible to
Figure 3-1 Hand tools for soil exploration. (a) Posthole, or iwan, auger; (b) small helical auger;
(} gasoline-engine-powered hand auger equipped with a continuous-flight auger. Note additional
suger flights in the foreground.(Suodwo3 a 42ypy 24,1) woods 1
stoquinu uorextouad ayer 01 pue BuIse9
Buog ysom Og,
2504 12407
\duund s210nh
\poayioyEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 89
a
d
ERE R EET Gt OE
ua
PE
—weee
Figure 3-3 Rotary drilling using a
* Tg. continuous-flight auger. (The
% Acker Drill Company’)
obtain a truly undisturbed sample of soil; so in general usage the term “undis-
turbed” means a sample where some precautions have been taken to mi imize
disturbance or remolding effects. In this context, the quality of an “undisturbed”
sample varies widely between soil laboratories. The following represent some of
the factors that make an undisturbed sample hard to come by
1. Sample is always unloaded of the in situ confining pressures, with some un-
known resulting expansion,90 FOUNDATION ANALYSIS AND DESIGN
2. Samples collected from other than test pits are disturbed by volume displace-
ment of the tube, or other collection device, which has some finite volume to
cause grain displacement when pushed or driven into the soil. The presence of
gravel greatly aggravates sample disturbance.
3. Sample friction on the sides of the collection device tends to compress the
sample during recovery. Note in passing that most sample tubes are swaged so
that the cutting edge is slightly smaller than the inside tube diameter to reduce
friction effects.
4, Some unknown changes in water content may occur if the sample is collected
from below the water table, or brought from a unsaturated zone through a
perched water table.
5. Loss of hydrostatic pressure may cause gas bubble voids to form in the
sample.
6. Handling and transporting a sample from the site to the laboratory and
transferring the sample from sampler to testing machine.
7. Quality and attitude of drilling crew, laboratory technicians, and the super-
vising engineer
8. Working environment. On very hot or cold days samples may dehydrate or
freeze if not protected on site. Worker attitudes may deteriorate in temper-
ature extremes.
It is much more difficult (close to impossible) to obtain “undisturbed” sam-
ples of cohesionless material for strength testing. Sometimes samples of reason-
able quality can be obtained using thin-walled piston samplers in medium- to
fine-grained sands. In gravelly materials, and in all dense materials, samples with
minimal disturbance are obtained only with extreme difficulty. Dilation occurs in
dense sands as a combination of volume displacement of the sampler and any
Pieces of gravel which catch on the cutting edge to give it a larger apparent
volume. Some attempts have been made to recover cohesionless materials by
freezing the soil, freezing a zone around the sample (but not the sample), or
injecting asphalt which is later dissolved from the sample, but most commonly
thin-walled piston samplers are used to obtain “undisturbed” samples. A test pit
may be used to recover a quality sample but the large amount of hand work will
make it difficult to justify the expense. The devices shown in Fig. 3-4 can be used
to recover disturbed samples from a boring for visual classification, sieve analy-
ses, and chemical tests.
The primary use of “undisturbed” cohesionless samples is to obtain the unit
weight (or relative density). The weight of soil in the known volume of the
sampler allows a reasonable determination of unit weight, and even if the sample
has been later disturbed by transporting it from the site to the laboratory. An
attempt to transfer a cohesionless sample from a tube to a testing machine for
strength determination is not likely to meet with much success. A sample rebuilt
in the laboratory to the in situ unit weight is lacking in both natural cementation
and anisotropy; these may be significant factors in both strength and per-
meability estimates, Some laboratories are of the opinion that anisotropic sam-EXPLORATION, SAMPLING, AND IN SITU SOI MEASUREMENTS 91
te)
Figure 3-4 Special sampling tools. (a) Sand-pump sampler which utilizes pumping action to recover
sample; (4) spiral-slot sampler that is filled by rotation of device; (c} slot sampler is filled by rotation
Uevice.
ples can be built to duplicate in situ and that the samples can be “aged” to
recover some natural cementation. Even assuming this can be done, few projects
could justify the expense for the small increase in confidence level obtained,
Since it is nearly impossible to recover “undisturbed” samples from cohe-
sionless deposits, density, strength, and compressibility estimates are usually ob-
tained from penetration tests or in situ measurements. Permeability may be
estimated from well pumping tests, or approximately, by bailing the boring and
observing the time for the water level to rise some amount (see Sec. 3-10).
Disturbed samples (all soils) are adequate to locate suitable borrow where
compaction characteristics and index tests for classification are usually sufficient.
In this case a larger-diameter auger (usually only shallow depths) may be used so
that bags of representative soil may be obtained for laboratory compaction tests,
sieve analyses, and the Atterberg limits.
Recognizing the difficulty, and resulting expense, of obtaining undisturbed
samples it is common practice on most foundation projects to rely on penetration
tests and recovery of disturbed samples for obtaining estimates of the soil condi-
tions. The standard penetration test of Sec. 3-7 is most commonly used, since a
disturbed sample is recovered, but other types, particularly cones, are also widely
used. The latter devices do not recover a soil sample.
Figure 3-5 illustrates the sampling spoon (also called a split spoon) commonly
used. It is made up of a driving shoe, to ensure a reasonable service life from
uriving into the soil, and a barrel. The barrel consists of a piece of tube split (split
spoon) lengthwise with a coupling on the upper end to connect the drill rod to
he surface. Inserts (see Fig. 3-Sd) are used when samples of thin mud and sand92 FOUNDATION ANALYSIS AND DESIGN
| Bie sorter shoe, the
es exible fingers
open to admit the
le) Sond then close
then the tube is
withdrown
Dring ag vents with minimum diameter of in,
p shoe a covating— ot
sing lite [Se bore | |
aac Fea
kez sss} f ats i sh
ein. tind} be] eins minimum
a ain tnininum——e |
+
i)
Sond 62-100
= Clay CEIO=IS a
op
a retainer vsed to
recover muds and
wotery somples
a
te) la)
Figure 35 Soil-sampling tools. (a) Standard split-spoon sampler; (6) dimensions of the standard
split-sampler assembly; (c) thin-wall (Shelby tube) sampler; d) split-spoon sampler inserts,
are to be recovered. In a test the sampler is driven into the soil a measured
distance, with the blows recorded. The sampler is then slightly twisted to shear
the soil at the base of the tube and withdrawn. The shoe and coupling are
unscrewed and the two halves of the barrel are opened to expose the sample.
On-site unconfined compression tests are routinely made on any cohesive sam-
ples recovered. Inspection of Fig. 3-Sb indicates any samples recovered by this
device are likely to be highly disturbed.
Representative samples from the soil on the sampler barrel are stored in
sample jars and returned to the laboratory for inspection and classification. The
field technician marks the jar lid with the job and boring number, sample depth,
and penetration blow count.
These jar samples are usually large enough to provide sufficient material for
the Atterberg limits and natural water content. In routine work these index
properties used with correlation tables and charts, together with q, , are sufficient
to select the foundation type, estimate the allowable bearing capacity, and makeEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 93
some kind of estimate of probable settlement in cohesive soil. This is particularly
true if the soil is stiff, above the water table, or is overconsolidated and fissured
where it is difficult to push a thin-walled sample tube and/or obtain an intact
sample for a compression test. The penetration number (a measure of resistance)
is usually sufficient for making both strength and settlement estimates in cohe-
sionless soils. Where the geotechnical consultant has had sufficient experience to
build a reasonable data base, strength/settlement predictions made in this
manner are quite adequate for about 85 to 90 percent of foundation work.
It is that other 10 to 15 percent of the work which causes geotechnical
insurance rates to be the highest among consultants and which taxes the ingen-
uity of the engineer. Recognizing both the difficulty of obtaining a quality sample
and of trying to return it to the K, condition for a laboratory test, in situ tests
described in later sections may be used. This is particularly true for important
structures founded on fine to medium sands and where very soft cohesive and/or
organic soils are present. Only thin seams of these latter soils may be sufficient to
cause great problems. In any case, unless competent lower strata are close
enough to decide a viable foundation alternative immediately, some testing of (or
in) these poor soils will be required,
As the field boring progresses and soft layers are encountered in the zones
which may influence the foundation selection/design, undisturbed samples are
usually taken so that consolidation and more refined laboratory strength tests
can be made. If the soil is extremely soft or experience indicates in situ tests
should be made (and the necessary equipment is available), only a few “undis-
turbed” tube samples for consolidation tests should be taken. As a general rule,
tube samples for consolidation tests should be at least 12 mm larger than the
consolidation ring; in practice, a 76-mm tube sample is often collected for use in
the 64-mm (24-in) diameter consolidometer. Sometimes a 50-mm tube sample is
used with a 48-mm (1.875-in) diameter consolidometer, but this practice is not
recommended. Tube samples larger than 76 mm can be obtained but if much
larger than 100 mm a premium may be charged for the extra drilling effort and
cost of tube.
Recovery of “undisturbed” samples in cohesive soils is accomplished by
replacing the split spoon on the drill rod with specially constructed thin-wall (16-
to 20-gage) seamless brass or steel tubing which is driven, but preferably pushed,
into the soil. The term “Shelby tube” is widely used to describe any thin-wall
tube used in this manner. Actually, Shelby tube was the trade name for hard-
drawn seamless-steel tube manufactured by the National Tube Division of the
US. Steel Corporation until several years ago. The tube is slightly rotated, or a
special cutting device is attached, to cut the sample off. Friction holds the sample
in the tube as the sample is withdrawn; however, there are also special valve or
piston (Fig. 3-6) arrangements which use a pressure differential (suction) to retain
the sample in the tube.
‘A special sampler termed a foil sampler (Fig. 3-6c) was developed in Sweden
[see Hvorslev (1949, p. 269), Kjellman (1948)} to overcome two principal defi-
‘encies of the usual sampling tubes and piston samplers. These deficiencies are
hort sample length and side friction between wall and soil as it is forced into the94 FOUNDATION ANALYSIS AND DESIGN
Wire rope or
Tight chain
saming
be
rast
overshot
Spear
ead
Locking in
ott
sca
Pushing
Hos
I)
Foil chamoers
“Hoiding ning
Stoel foils
Shoe ong
cattng edge
Sormping
- Drilt rod ‘ | :
5 — Top piston k
e-- Water under E
5 Pressure eae
5 cylinder g kK
= bs
4 -— Hollow ston f
2 rod 4
Fixed piston 4 3
(o- Thin-walled
Beginning sampling tube
operction
Top piston is fully
advanced ond semple
te) tube con be recovered
Figure 36 Typical piston. samplets. (a) Stationary-piston sampler for recovery of “undisturbed”
samples of cohesive soils, Piston remains stationary on soil and tube is pushed into the soil; piston is
then cl: nped and sample is recovered: (b) Swedish foil sampler; (c) Osterberg piston sampler.
[Heorster (1949).EXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 95
sampler. Reducing friction required reducing the sampler length. If one is in a soil
suspected of being particularly troublesome, it may be necessary to take con-
tinuous samples. This is not practical with samplers of, say, 1 m maximum length
because of continually pulling the drill rods to attach a new tube. The foil
sampler is a means of allowing sample recovery of from 10 to 20 m with a
minimal friction effect. The interested reader should consult the cited references
for exact details, but essentially the sampler operates by first placing it on the
bottom of the borehole. Next it is pushed into the soil; as the sample enters the
tube it is surrounded by 16 foil (thin metal strips about 13 mm wide hy 05 to
1.0 mm thick) strips which carry the sample up the tube. Friction between soil
and foils results in reducing the compressive stress in the sample as the length of
recovered sample increases in the tube.
Liners are available, with certain split-spoon samplers which are simply thin-
walled tubes placed inside the barrel portion. The sample is collected, the exterior
barrel opened, and the liner containing the sample is removed. Liner-recovered
samples are of doubtful quality if an undisturbed sample is the intent, the major
reason being that the sample is obtained using the driving shoe, which has a
rather large volume displacement, to trim the sample.
Although sample disturbance depends on factors such as rate of penetration
of spoon, whether the cutting force is obtained by pushing or driving, and pre-
sence of gravel, it also depends on the ratio of the volume of soil displacement to
the volume of the collected sample, expressed as an area ratio A,:
3-1)
outside diameter of tube
inside diameter of cutting edge
where D,
D,
Well-designed sample tubes should have an area ratio of less than about
10 percent [the widely used 2-in thin-wall (Shelby) tube has an A, of about
13 percent as computed in Example 3-1].
Another term used in estimating the degree of disturbance of a cohesive or
rock-core sample is the recovery ratio L,:
actual length of recovered sample
theoretical length of recovered sample
(3-2)
\ recovery ratio of 1 (recovered length of the sample = the length sampler was
forced into the stratum) indicates that, theoretically, the sample did not become
compressed from friction on the tube. A recovery ratio greater than 1.0 would
indicate a loosening of the sample from rearrangement of stones, roots, removal
f preload, or other factors.
In the final analysis, however, engineering judgment must be relied upon to
cMtrapolate the results of tests on “undisturbed samples to the prediction of field
ochavior.96 FOUNDATION ANALYSIS AND DESIGN
Example 3-1, What is the area ratio of the 2-in thin-wall (Shelby) tube?
SoLUTION Using nominal dimensions from the supplier's catalog, obtain OD = 2 in.
ID
4 in (The actual ID of the tube is slightly larger than the ID of the cutting edge to reduce
side friction on the sample as the tube is pushed into the soil,
pi-D? 2 - 1.873?
SOE x 100 = = x 100 = 1
Bi Tavs? * 100 = 136 percen
3-6 MARINE SAMPLING
The current interest in offshore construction, primarily for energy recovery but
for other purposes as well, has created a demand for a means to obtain under-
water samples. Obviously marine sampling is going to be orders of magnitude
more difficult than land sampling. The present state of art is such that about all
one can reliably obtain is a general overview of the subsurface conditions. This is
used in conjunction with conventional onshore methods (such as pile driving or
footing design) to make a foundation design. Usually piles are driven to refusal,
or bearing capacities are simply taken as small values. As the design data base
improves, more rational methods of design will surely be used.
There are a number of methods for sampling soils from the ocean floor that
are in the research stage, but one of the more popular methods uses mounted
conventional drilling equipment on a boat or a barge specially modified for the
purpose and drills through a casing extending to the ocean floor. This method
has been used considerably in the Gulf of Mexico and along the shallower waters
along the continental shelves on both the Atlantic and Pacific coasts. Penetra-
tion, vane, and pressuremeter tests can be made in the borings.
In deeper water, alternative methods are used including submarine-type ves-
sels and projectile-type devices which are lowered to the ocean floor. Servo-
mechanisms commanded from the surface may be used to force a sample tube
into the soil using the weight of the vessel or device for a reaction. A projectile
device may contain a gas or explosive charge to propel a sample tube into the
soil, again using the weight of the total device as a reaction
The most promising method for deep water seems to be to establish a con-
nection between a boat and the ocean floor using a casing with flexible joints to
allow for wave action. A sample tube with a drive weight attached is lowered on
a cable through the casing to the ocean floor. When the sample tube is on the soil
the drive weight is lifted up a guide rod (attached to the top of the sampler) a
distance of 1 to 2 m and dropped. Effectively this is a type of penetration test as
well as a means to recover a soil sample. The tube wall is relatively thin so that
sample quality from tube volume is not greatly degraded. Continuous sampling
can be done in increments of 0.6 m at a rather rapid rate as the great advantage
is in being able to draw the sample to the surface by winding the cable rather
than withdrawing a string of drill rods. The cable can produce very significantEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 97
time savings in water depths of 100 to 1000 m. The principal disadvantage of this,
method is in knowing when the cutting edge of the tube engages the soil. While
electronic methods can rather precisely locate the ocean floor, the cable elonga-
tion (PL/AE) resulting from the suspended weights combined with wave induced
casing movements creates great difficulty in defining when the tube is on soil.
Unless this is known, there is no reliable means of determining the recovery ratio
(and the sample quality).
Another rather formidable problem in great depths of water is that the loss of
hydrostatic pressure can result in about 0.1 percent sample expansion in the
deeper waters. This deteriorates the sample an unknown amount. Probably a
greater effect is that the loss of a great amount of hydrostatic pressure increases
the likelihood of gas bubbles forming sample cavities. It is not known at present
how much the change in salt concentration affects a sample recovered from a
great depth. Finally we might note that while it is possible to recover from great
depths a sample that is isolated from the surface environment, it has explosive
potential because of the high captive internal pressure and must be handled with
great care.
A number of marine sampling methods are described in ASTM (1971) and
among the references cited by Focht and Kraft (1977), which the interested reader
may wish to consult.
3-7 THE STANDARD PENETRATION TEST (SPT)
Section 3-5 discussed sampling methods and indicated that penetration testing
was a most useful method to determine the soil conditions underlying a site. The
standard penetration test is currently the most popular and economical means to
obtain this subsurface information. Some persons feel that this procedure should
not be dignified by calling it a test. Others are of the opinion that the data should
be called an index. Nonetheless the method has been standardized by ASTM
1D1586 as “Standard Method for Penetration Test and Split-Barrel Sampling of
Soils” and is commonly called the standard penetration test. This test con-
s of:
|. Driving the standard split-barrel sampler of dimensions shown in Fig. 3-5 a
distance of 460 mm (18 in) into the soil at the bottom of the boring
>. Counting the number of blows to drive the tube the last 305 mm (12 in) to
obtain the N number
4. Using a 63.5-kg (140-Ib) driving mass falling free from a height of 760 mm
(30 in)
The sampler is driven a distance of 150 mm to seat it on undisturbed soil
with the blow count recorded. The blow count for each of the next two 150-mm
increments is used as the penetration count unless the last increment cannot be98 FOUNDATION ANALYSIS AND DESIGN
completed (either from encountering rock or because the blow count exceeds
100). In this case the blow count for the last 305 mm is computed and used for N.
The boring log shows “refusal” if the blow count exceeds 100 and may show
a ratio as 70:100 or 50:100, indicating that 70 (or 50) blows resulted in a
penetration of 100 mm (4 in). Excessive equipment wear, as well as greatly re-
duced drilling meterage, results when blow counts are high. ASTM standardiza-
tion of refusal at 100 blows allows all the drilling organizations to standardize
costs so that higher blow counts result in a negotiation for higher costs/length of
boring or going to some type of coring operation.
The SPT has been considerably studied and reported on [Schmertmann
(1975), De Mello (1971), Bazaraa (1967)]. The basic conclusions are that the test
is difficult to reproduce. Some of the factors which affect the reproducibility are:
1. Effect of overburden pressure. Soils of the same density will give smaller
counts near the ground surface.
2. Variations in the 760-mm free fall of the drive weight, since this is often done
by eye on older equipment using a rope wrapped around a power takeoff
(cathead) from the drill motor. Newer equipment does this automatically.
3. Interference with the free fall of the drive weight by the guides or the hoist
rope. New equipment eliminates rope interference.
4. Use of a drive shoe that is badly damaged or worn from too many drivings to
5. Failure to properly seat the sampler on undisturbed material in the bottom of
the boring.
6. Inadequate cleaning of loosened material from the bottom of the boring.
7. Failure to maintain sufficient hydrostatic pressure in the borehole so that the
test zone becomes “quick.” This can happen when using the continuous-flight
auger with the end plugged and maintaining a water level in the hollow stem
below that in the hole.
Driving a stone ahead of the sampler.
9. Careless work on the part of the drill crew.
~
At one time it was thought that the stiffness and weight of the drill rods
connecting the sampler to the ground surface where driving takes place would
affect the blow count; however, studies by Gibbs and Holtz (1957) and more
recently by Brown (1977) indicate this is not a factor. Studies by Marcusson and
Bieganousky (1977) indicate that it is possible to reproduce N in the same soil if
the overburden pressure is kept constant. It appears from these several studies
that soil type, density, and overburden pressure are the most significant factors
affecting the reproducibility of N (assuming good workmanship and equipment).
Regardless of the impressive list of shortcomings, the SPT is not likely to be
abandoned for several reasons
1. The test is too economical in terms of cost per unit of information.
2. As commonly done at every 760 mm (24 ft) of depth a tube recovery length ofEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 99
460 mm (18 in), including seating length, produces a visual profile of around
60 percent of the boring depth.
3. The test results in recovery of very disturbed samples, but they can still be
tested for strength and index properties, and visually examined.
4. Long service life of the enormous amount of equipment in use.
5. The accumulation of a large SPT data base which is continually expanding.
6. The fact that other methods can be readily used to supplement the SPT when
the borings indicate more refinement in sample/data collection.
Corrections to the SPT number for overburden effects were recognized and
proposals made by Gibbs and Holtz (1957). These have been widely used but
now as the data base has expanded are considered to be too conservative.
Bazaraa (1967) made an analysis of a large number of borings of others to
propose the following corrections to the actual blow count N based on overbur-
den pressure:
For p, S75 kPa (1.5 ksf),
AN
— (3-3)
1+ x1P0
For p, > 75 kPa,
4Nn (3-4)
nia ON
3.25 + x23.
where 7, = 0.04 for SI units; = 2.0 for Fps units
{2 = 0.01 for SI units; = 0.5 for Fps units
p, = effective overburden pressure, kPa or ksf
‘There are a number of approximations in the literature which give approxi-
mately the N’ of the above equations. These are usually in the form of curves
where one enters N and p, to obtain the corrected SPT value N’. A set of office
design curves can be readily made using a programmable calculator and these
equations.
The following comments are made regarding correcting the actual blow
count N toa corrected value N’
N’ is increased from the actual blow count when p, < 75 kPa.
N' is decreased from the actual blow count when p, > 75 kPa.
When the blow count indicates a relative density D, < 0.5 (see Table 3-2), do
not use these equations.
4, The author suggests that N’ should not be more than 2N or much less than N.
5, Use these equations cautiously
In very fine, or silty, saturated sand Terzaghi and Peck (Ist ed., p. 426)
recommended that the penetration number N be corrected to N’ if N was greater100 FOUNDATION ANALYS
AND DESIGN
Table 3-2 Empirical values for ¢, D,, and unit weight of granular soils based on the
standard penetration number with corrections for depth and for fine saturated sands.
Very
Description Very loose Loose Medium Dense dense
Relative density D,* ous 035 06s 085 1.00
Standard penetra
tion no. N x10 ss 10-40 2070 335
Approx. angle {
‘of internat
fiction @°t 25 “30 232 30-38 35-40 38-43
Approx. range
‘of moist unit
weight 7, pet 70-1003 90-115 110-130 10-140 130-150
(Nim) 16) 14-18) (aaj | ees (20-23)
* Depends on p, ranging from 70 to $00 kPa. Low value of N corresponds to lesser p,
+ After Meyerhof (1956). = 25 + 25D, with more than 5 percent fines and ¢ = 30+ 25D, with less than 5
percent fines, Use larger values for gramular material with S percent or less fine sand and silt. See also Eq, (4-10) for
estimate of @.
Et should be noted that excavated material or material dumped from a truck will weigh 70 to 90 pe
Material must be quite dense and hard to weigh much over 130 pef. Values of 105 to 115 pef for nonsaturated soils
‘are common.
than 15 as
N'= 154+ HN ~ 15) (3-5)
Bazaraa (1967) also made a recommendation for correcting N in these soils as
N’ =0.6N (3-5a)
where N > 15. The rationale for these corrections was that the soil must be dense
if the blow count is greater than 15; therefore, the volume displacement at a high
penetration rate of the SPT would produce high pore pressures which would
further increase the blow count. It appears that not many persons use any correc
tion for the blow count in these soils at present.
The penetration test in gravel or gravelly soils requires careful interpretation,
since pushing a piece of gravel can greatly change the blow count. Generally if
one uses a statistical average of the blow count in the stratum from the several
borings, either excessively high or very low (pushing a gravel or creating a void
space) values will be averaged out so that approximately the correct blow count
can be estimated for design.
The SPT was originally developed for cohesionless soils so that samples
would not have to be taken. The test has evolved to the current practice of
routinely determining N for all soils supplemented with on-site determinations of
qu for cohesive strata, The test is usually performed every 2.5 ft or | m of depth
starting at about 2.5 ft or 1 m below the ground surface. The boring is extended
to two or three times the estimated base width of the foundation below theEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS. 101
‘able 3-3 Empirical values for g,* and consistency of cohesive soils based on the standard
penetration number.
Very Very
Consistency | soft sot | Medium | suit sit Hard
dy ks Oo Os 10 20 40 80.
(kPa) (25) (50) (100) (200) (400)
N, standard |
penetration | |
fasiance : 4 ‘ 1
sa et 100-120 110-130 120-140
(kN/m?) (16-19) {17-20}, (19-22)
L
* These values should be used as a guide only. Local cohesive samples should be tested, and the relationship
\iween WV and the unconfined compressive strength 4, established as q, = KN.
estimated foundation base. Sometimes the borings are extended to bedrock; how-
ever, at greater depths the SPT may be done at 5-ft or 2-m increments. The term
“estimated is used extensively here since these data are not precisely known at
the time of the exploration.
Empirical correlations, such as Table 3-2, between N and various soil proper-
ties have been made for cohesioniess soils. The low values for N’ correspond
roughly to p, of 75 kPa, the higher to about $00 kPa. Tables such as Table 3-2
should always be used with caution,
Table 3-3 is an empirical correlation between N and unconfined compressive
strength. These should be used even more cautiously than for cohesionless soils.
It is preferable to measure q, directly on samples than use a table such as this.
Alternatively, a table might be developed locally which has some validity by
measuring q, and N and solving
qu=KN
for the constant of proportionality K. The reason for poor correlation of g,
versus N tables is that inadvertent changes in water content, excess pore pres-
sures during sampling, and other factors may give very low blow counts, say 6 to
10, on soils where the measured q, may be on the order of 300 to 600 kPa. Note
in Table 3-3 the value of K is approximately 0.25.
There are varied opinions on how to use N when it is obtained. In a given
stratum of finite thickness the values may range by a factor of 2 or more. Early
recommendations were to use the smallest value in the stratum—which is cer
tainly conservative. Current thinking is to use a statistical average for N (as from
several adjacent borings) and only for the zone of interest. That is, rather than
use the entire stratum, use that zone just above to about two times the footing
width below the footing. Of course, one should not ignore low blow counts in a
layer which may cause settlement problems.102 FOUNDATION ANALYSIS AND DESIGN
Example 3.2 The penetration test was performed on a sand deposit for which the unit of the
soil is 17.26 kN/m? (average for first 3.5 m) and no water. At elevation —3.0 m the blow count
was 8. What is the value of N to use in the bearing-capacity equations of Chap. 4?
SOLUTION A, = 30(17.26) = 51.8 < 75 kPa; use Eq. (3-3), From Table 3-2 estimate D, = 0.35
< 05; 0 strictly we should not make a correction: however, for illustration,
48)
1+ 0.04(51.8)
3-8 OTHER PE
TRATION METHODS
Cone-type penetrometers are quite popular in Europe. The Dutch-cone
penetrometer (Fig. 3-7) is one form that is rather widely used, especially in the
Low Countries, and is becoming widely used worldwide. This device will be
discussed in some detail in Sec. 3-15. Palmer and Stuart (1957) report correlations
on a British device (Fig. 3-7g) for which the penetration number is about the
same as for the standard method. This particular device is primarily for testing in
gravels, where the problem of pushing a large stone is found.
Other penetrometers and sounding devices are also used in the United States,
The split-spoon samplers are available from 2 to 44 in in diameter by $-in
increments. A device termed a peat sampler, which is primarily for hand insertion
into very soft deposits (under 35 ft), is available for collecting 3-in-diameter
samples. Sounding rods are also used to probe for rock. In the simplest form
these are drill rods equipped with a special point.
For the larger split spoons and sounding rods, drop weights weighing 250 to
375 Ib are available. These weights may also be used to drive the casing into the
borehole. Sowers (1954) published a correlation factor for converting the pene-
tration obtained by nonstandard procedures to the standard test. This factor was
1.0 for all samplers and drive weight combinations except a 1.3-in-OD sampler
with a 140-Ib weight where the factor was N = 1.5N', In light of recent studies
concerning hammer input energy, the factors may no longer be 1, and new
evaluations may be required.
3-9 CORE SAMPLING
In rock, except for very soft or partially decomposed sandstone or limestone, very
high blow counts are necessary to get any penetration, and the term refusal is
commonly used to describe this condition. For these cases core samples may be
necessary, for several reasons, one principal reason being that the boring oper-
ation is usually charged on the basis of per meter of hole drilled, and high blow
counts reduce the daily meterage considerably. High blow counts also tend to
damage the drilling equipment, and in rock or very hard soil the impact tends toEXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 103
25mm
fF 0mm 200m
ao
30-33mm
200 mm
el (6)
10-60"
(Bo%suol)
= ks2-comm
\)
Figure 3-7 Penetration and sounding soil-exploration devices. With the exception of (g) the devices
utilize static methods. (a) Danish penetration device. Penetration is recorded for 25, $0, 75, and
11W0 kg: the device is rotated, and the penetration is noted for each 25 half-turns. [After Godskesen
11936),] (b) Swedish penetrometer. Same essential method of operation as the Danish penetrometer.
| Dahiberg (1974).] (c) Terzaghi wash-point penetrometer. The cone is jacked 10 in into the soil, and
the penetration resistance is measured. Water is then pumped through the system and circulated to
the surface, where the soil is collected. The sleeve is then pushed down to the top of the cone, and the
process is repeated. [After Hoorslev (1949). (d) Swiss penetrometer. [Crettaz and Zeindler (1974) ]
‘c] Early Dutch-cone penetrometer. (Barentsen (/936).] () Waterways Experiment Station penetrom-
ter for shallow soundings. The device is pushed into the soil by one person, and the force necessary
| penetrate 2 in is observed. The device has a load ring which is used to obtain the penetration
resistance. (g) Modified split spoon. The conical point is for use in granular soil so that large pieces of
sione are not trapped, thus causing an increase in the blow count. [After Palmer and Stuart (1957)]
reduce the sample to chips, so that the quality of the material may not be
determined.
If rock is encountered close to the ground surface, it may be necessary to
letermine if it is bedrock or a suspended boulder, unless prior geological know-104 FOUNDATION ANALYSIS AND DESIGN
ledge in the area can be relied upon. In this case cores may be taken, although
probing in the vicinity of the drill hole may indicate the size of the rock.
Rock cores are necessary if the soundness of the rock is to be established;
however, cores smaller than BX size (see Table 3-4) tend to break up in the
boring process. Cores larger than BX size also tend to break up (rotate inside the
core barrel and degrade)—especially if the rock is soft or fissured. Drilling small
holes and injecting colored grout (water-cement mixture) into the seams can
sometimes be used to recover relatively intact samples. Colored grout present in
the recovered core indicates fissures, as well as fissure size, and with cores from
several adjacent borings indicates the fissure orientation,
Unconfined and high-pressure triaxial tests can be performed on recovered
cores to determine the elastic properties of the rock. These tests are performed on
pieces of sound rock from the core sample and may give much higher com-
pressive strengths in laboratory testing than the “effective” strength available
from the rock mass. This is because the rock mass may be fractured or fissured.
Figure 3-8 illustrates several commonly used drill bits, which are attached to
a piece of hardened steel tube (core barrel) 0.6 to 3 m long. In the drilling
operation the bit and core barrel rotate while pressure is applied, thus grinding a
groove around the core. Water under pressure is forced down the barrel and into
the bit to carry the rock dust out of the hole as the water is circulated.
The recovery-ratio term used earlier also has significance for core samples. A
recovery ratio near 1.0 usually indicates good-quality rock. In badly fissured or
soft rocks the recovery ratio may be 0.5 or less.
Rock quality designation (RQD) is an index or measure of the quality of a
rock mass [Stagg and Zienkiewicz (1968)] used by many engineers. RQD is
computed from recovered core samples as
& lengths of intact pieces of core > 100 mm
nope length of core advance.
For example, a core advance of 1500 mm produced a szample length of 1310 mm
consisting of dust, gravel, and intact pieces of rock. The sum of lengths of pieces
100 mm or larger (pieces vary from gravel to 280 mm) in length is 890 mm. The
recovery ratio L, = 1310/1500 = 0.87 and RQD 890/1 500 = 0.59,
‘The rating of rock quality may be used to approximately establish field
Table 3-4 Standard de
nation and sizes for drill rods and casing
Drill Casing and Core-barrel-bit Approx. diam of Diam of core
rod OD, in core barrel OD. in borehole.* in sample, in
E 1% EX 4 q
A a AX 2 H
B q BXt 4 R
N 4 NX ny 3 2
* Diameter of borehole is very nearly the 1D of the casing.
+ In soft or fractured rock, BX or larger cores are preferred,EXPLORATION, SAMPLING, AND IN SITU SOIL MEASUREMENTS 105,
double-tube
core barrel
la)
Figure 3-8 (a) Core barrels to which (6) coring bits ate attached to obtain rock cores. (The Acker
Drill Company:)
reduction of modulus of elasticity and/or compressive strength and the following
may be used as a guide.
RQD Rock description E/E
<025 Very poor 01s
025-050 Poor 020
050.075 Fair 028
075-090 Good 0307
>090 Excellent 07-10
* Approximately for field/laboratory compres-
sion strengths als.
3-10 WATER-TABLE LOCATION
Since groundwater affects many elements of foundation design and construction,
its location should be established as accurately as possible if it is within the
probable construction zone; otherwise, it is necessary only to determine where it
is not. This can be done with considerably less accuracy. It is generally deter-106 FOUNDATION ANALYSIS AND DESIGN
mined by measuring to the water level in the borehole after a suitable time lapse.
A period of 24 h is a value which is Widely used. In soils with high permeability,
such as sands and gravels, the lapse of several hours is usually sufficient unless
the hole has been sealed with drilling mud, In soils with low permeability, such as
silts, fine sands, and clays, it may take several days to weeks to determine pre-
cisely the water level. For these cases it may be necessary to resort to indirect
means to establish the approximate location of the water table, as follows:
1. Plot degree of saturation 5 with depth if it is possible to obtain reliable data to
compute S. A direct plot of water content may be useful but for S = 100
percent can decrease as the void ratio decreases from overburden pressure.
2. Another method is to fill the hole and bail it out. After bailing a quantity,
observe if the water level in the hole is rising or falling. The true level is
between the bailed point where the water was falling and the bailed point
where it was rising.
3. Apply a computational method proposed by Hvorslev (1949, p. 77). In this
method measure the rise (or fall) for two or more equal time intervals AT’
—Ty=AT for height of rise hy
T,—T,=AT for height of rise hy
T,—T,=AT for height of rise hy
The distance to the rising water surface from the stabilized groundwater level
is Ho to Ty; Hy to hy + hz; Hz to hy +h, +hy; etc. (refer to Fig. 3-9 and
Cosing
Oh= bh =O
by thy # hy
Figure 39 Method of computation of location of the stabilized groundwater level by measuring rise
of water in the borehole for equal time intervals of At. [After Heorslev (1949).]EXPLORATION, SAMPLING, AND IN SITU SOI. MEASUREMENTS 107
Example 3-3). The distances H are computed from the measured changes in water
level
a Ae ny
i re, 7 hh
Ho
If reliable data are necessary, a piezometer should be installed in a borehole
and periodically inspected over a longer period of time until the groundwater
level stabilizes,
Artesian pressures and perched water levels can create an interpretation
problem for the unwary. If the groundwater is under pressure (artesian water),
deeper borings tend to raise the water level. Perched water may be indicated if
the water level tends to disappear when the boring extends deeper, especially
through a relatively impermeable material into an underlying material, e.g, a clay
stratum overlying a sand deposit.
Example 3.3 It is desired to establish the location of the groundwater table in a clayey material
The borehole was bailed to a depth of 35 ft below the ground surface, and the water rise was
recorded on three successive days as follows:
hy = 21h ath
hy =19 at 2h
hye 17h at 2h
(2.7
Souunox Ho = mat
21-19 Cosing
iF comercrry| fem
Hyg Bt
7?
Tyga Mat
Referring to the figure, on which the H values just com-
uted have been placed, the depth to the water table is as
follows
First days D,+ 21-358
p= 1298
Second day: Dy +181 + 21419 =35 ft | ote
p,=1298 4} _| taro
Third day Dy+ 148417419421 = 358
Figure E3-3
D, = 149 ft
Averaging results, we obtain a depth to the water table of
DL= B6K